diff options
Diffstat (limited to 'tools')
160 files changed, 19997 insertions, 5737 deletions
diff --git a/tools/Makefile b/tools/Makefile new file mode 100644 index 000000000000..3ae43947a171 --- /dev/null +++ b/tools/Makefile @@ -0,0 +1,77 @@ +include scripts/Makefile.include + +help: + @echo 'Possible targets:' + @echo '' + @echo ' cpupower - a tool for all things x86 CPU power' + @echo ' firewire - the userspace part of nosy, an IEEE-1394 traffic sniffer' + @echo ' lguest - a minimal 32-bit x86 hypervisor' + @echo ' perf - Linux performance measurement and analysis tool' + @echo ' selftests - various kernel selftests' + @echo ' turbostat - Intel CPU idle stats and freq reporting tool' + @echo ' usb - USB testing tools' + @echo ' virtio - vhost test module' + @echo ' vm - misc vm tools' + @echo ' x86_energy_perf_policy - Intel energy policy tool' + @echo '' + @echo 'You can do:' + @echo ' $$ make -C tools/<tool>_install' + @echo '' + @echo ' from the kernel command line to build and install one of' + @echo ' the tools above' + @echo '' + @echo ' $$ make tools/install' + @echo '' + @echo ' installs all tools.' + @echo '' + @echo 'Cleaning targets:' + @echo '' + @echo ' all of the above with the "_clean" string appended cleans' + @echo ' the respective build directory.' + @echo ' clean: a summary clean target to clean _all_ folders' + +cpupower: FORCE + $(QUIET_SUBDIR0)power/$@/ $(QUIET_SUBDIR1) + +firewire lguest perf usb virtio vm: FORCE + $(QUIET_SUBDIR0)$@/ $(QUIET_SUBDIR1) + +selftests: FORCE + $(QUIET_SUBDIR0)testing/$@/ $(QUIET_SUBDIR1) + +turbostat x86_energy_perf_policy: FORCE + $(QUIET_SUBDIR0)power/x86/$@/ $(QUIET_SUBDIR1) + +cpupower_install: + $(QUIET_SUBDIR0)power/$(@:_install=)/ $(QUIET_SUBDIR1) install + +firewire_install lguest_install perf_install usb_install virtio_install vm_install: + $(QUIET_SUBDIR0)$(@:_install=)/ $(QUIET_SUBDIR1) install + +selftests_install: + $(QUIET_SUBDIR0)testing/$(@:_clean=)/ $(QUIET_SUBDIR1) install + +turbostat_install x86_energy_perf_policy_install: + $(QUIET_SUBDIR0)power/x86/$(@:_install=)/ $(QUIET_SUBDIR1) install + +install: cpupower_install firewire_install lguest_install perf_install \ + selftests_install turbostat_install usb_install virtio_install \ + vm_install x86_energy_perf_policy_install + +cpupower_clean: + $(QUIET_SUBDIR0)power/cpupower/ $(QUIET_SUBDIR1) clean + +firewire_clean lguest_clean perf_clean usb_clean virtio_clean vm_clean: + $(QUIET_SUBDIR0)$(@:_clean=)/ $(QUIET_SUBDIR1) clean + +selftests_clean: + $(QUIET_SUBDIR0)testing/$(@:_clean=)/ $(QUIET_SUBDIR1) clean + +turbostat_clean x86_energy_perf_policy_clean: + $(QUIET_SUBDIR0)power/x86/$(@:_clean=)/ $(QUIET_SUBDIR1) clean + +clean: cpupower_clean firewire_clean lguest_clean perf_clean selftests_clean \ + turbostat_clean usb_clean virtio_clean vm_clean \ + x86_energy_perf_policy_clean + +.PHONY: FORCE diff --git a/tools/lib/traceevent/Makefile b/tools/lib/traceevent/Makefile new file mode 100644 index 000000000000..3d69aa9ff51e --- /dev/null +++ b/tools/lib/traceevent/Makefile @@ -0,0 +1,303 @@ +# trace-cmd version +EP_VERSION = 1 +EP_PATCHLEVEL = 1 +EP_EXTRAVERSION = 0 + +# file format version +FILE_VERSION = 6 + +MAKEFLAGS += --no-print-directory + + +# Makefiles suck: This macro sets a default value of $(2) for the +# variable named by $(1), unless the variable has been set by +# environment or command line. This is necessary for CC and AR +# because make sets default values, so the simpler ?= approach +# won't work as expected. +define allow-override + $(if $(or $(findstring environment,$(origin $(1))),\ + $(findstring command line,$(origin $(1)))),,\ + $(eval $(1) = $(2))) +endef + +# Allow setting CC and AR, or setting CROSS_COMPILE as a prefix. +$(call allow-override,CC,$(CROSS_COMPILE)gcc) +$(call allow-override,AR,$(CROSS_COMPILE)ar) + +EXT = -std=gnu99 +INSTALL = install + +# Use DESTDIR for installing into a different root directory. +# This is useful for building a package. The program will be +# installed in this directory as if it was the root directory. +# Then the build tool can move it later. +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +prefix ?= /usr/local +bindir_relative = bin +bindir = $(prefix)/$(bindir_relative) +man_dir = $(prefix)/share/man +man_dir_SQ = '$(subst ','\'',$(man_dir))' +html_install = $(prefix)/share/kernelshark/html +html_install_SQ = '$(subst ','\'',$(html_install))' +img_install = $(prefix)/share/kernelshark/html/images +img_install_SQ = '$(subst ','\'',$(img_install))' + +export man_dir man_dir_SQ html_install html_install_SQ INSTALL +export img_install img_install_SQ +export DESTDIR DESTDIR_SQ + +# copy a bit from Linux kbuild + +ifeq ("$(origin V)", "command line") + VERBOSE = $(V) +endif +ifndef VERBOSE + VERBOSE = 0 +endif + +ifeq ("$(origin O)", "command line") + BUILD_OUTPUT := $(O) +endif + +ifeq ($(BUILD_SRC),) +ifneq ($(BUILD_OUTPUT),) + +define build_output + $(if $(VERBOSE:1=),@)$(MAKE) -C $(BUILD_OUTPUT) \ + BUILD_SRC=$(CURDIR) -f $(CURDIR)/Makefile $1 +endef + +saved-output := $(BUILD_OUTPUT) +BUILD_OUTPUT := $(shell cd $(BUILD_OUTPUT) && /bin/pwd) +$(if $(BUILD_OUTPUT),, \ + $(error output directory "$(saved-output)" does not exist)) + +all: sub-make + +gui: force + $(call build_output, all_cmd) + +$(filter-out gui,$(MAKECMDGOALS)): sub-make + +sub-make: force + $(call build_output, $(MAKECMDGOALS)) + + +# Leave processing to above invocation of make +skip-makefile := 1 + +endif # BUILD_OUTPUT +endif # BUILD_SRC + +# We process the rest of the Makefile if this is the final invocation of make +ifeq ($(skip-makefile),) + +srctree := $(if $(BUILD_SRC),$(BUILD_SRC),$(CURDIR)) +objtree := $(CURDIR) +src := $(srctree) +obj := $(objtree) + +export prefix bindir src obj + +# Shell quotes +bindir_SQ = $(subst ','\'',$(bindir)) +bindir_relative_SQ = $(subst ','\'',$(bindir_relative)) + +LIB_FILE = libtraceevent.a libtraceevent.so + +CONFIG_INCLUDES = +CONFIG_LIBS = +CONFIG_FLAGS = + +VERSION = $(EP_VERSION) +PATCHLEVEL = $(EP_PATCHLEVEL) +EXTRAVERSION = $(EP_EXTRAVERSION) + +OBJ = $@ +N = + +export Q VERBOSE + +EVENT_PARSE_VERSION = $(EP_VERSION).$(EP_PATCHLEVEL).$(EP_EXTRAVERSION) + +INCLUDES = -I. -I/usr/local/include $(CONFIG_INCLUDES) + +# Set compile option CFLAGS if not set elsewhere +CFLAGS ?= -g -Wall + +# Append required CFLAGS +override CFLAGS += $(CONFIG_FLAGS) $(INCLUDES) $(PLUGIN_DIR_SQ) +override CFLAGS += $(udis86-flags) + +ifeq ($(VERBOSE),1) + Q = + print_compile = + print_app_build = + print_fpic_compile = + print_shared_lib_compile = + print_plugin_obj_compile = + print_plugin_build = + print_install = +else + Q = @ + print_compile = echo ' CC '$(OBJ); + print_app_build = echo ' BUILD '$(OBJ); + print_fpic_compile = echo ' CC FPIC '$(OBJ); + print_shared_lib_compile = echo ' BUILD SHARED LIB '$(OBJ); + print_plugin_obj_compile = echo ' CC PLUGIN OBJ '$(OBJ); + print_plugin_build = echo ' CC PLUGI '$(OBJ); + print_static_lib_build = echo ' BUILD STATIC LIB '$(OBJ); + print_install = echo ' INSTALL '$1' to $(DESTDIR_SQ)$2'; +endif + +do_fpic_compile = \ + ($(print_fpic_compile) \ + $(CC) -c $(CFLAGS) $(EXT) -fPIC $< -o $@) + +do_app_build = \ + ($(print_app_build) \ + $(CC) $^ -rdynamic -o $@ $(CONFIG_LIBS) $(LIBS)) + +do_compile_shared_library = \ + ($(print_shared_lib_compile) \ + $(CC) --shared $^ -o $@) + +do_compile_plugin_obj = \ + ($(print_plugin_obj_compile) \ + $(CC) -c $(CFLAGS) -fPIC -o $@ $<) + +do_plugin_build = \ + ($(print_plugin_build) \ + $(CC) $(CFLAGS) -shared -nostartfiles -o $@ $<) + +do_build_static_lib = \ + ($(print_static_lib_build) \ + $(RM) $@; $(AR) rcs $@ $^) + + +define do_compile + $(print_compile) \ + $(CC) -c $(CFLAGS) $(EXT) $< -o $(obj)/$@; +endef + +$(obj)/%.o: $(src)/%.c + $(Q)$(call do_compile) + +%.o: $(src)/%.c + $(Q)$(call do_compile) + +PEVENT_LIB_OBJS = event-parse.o trace-seq.o parse-filter.o parse-utils.o + +ALL_OBJS = $(PEVENT_LIB_OBJS) + +CMD_TARGETS = $(LIB_FILE) + +TARGETS = $(CMD_TARGETS) + + +all: all_cmd + +all_cmd: $(CMD_TARGETS) + +libtraceevent.so: $(PEVENT_LIB_OBJS) + $(Q)$(do_compile_shared_library) + +libtraceevent.a: $(PEVENT_LIB_OBJS) + $(Q)$(do_build_static_lib) + +$(PEVENT_LIB_OBJS): %.o: $(src)/%.c + $(Q)$(do_fpic_compile) + +define make_version.h + (echo '/* This file is automatically generated. Do not modify. */'; \ + echo \#define VERSION_CODE $(shell \ + expr $(VERSION) \* 256 + $(PATCHLEVEL)); \ + echo '#define EXTRAVERSION ' $(EXTRAVERSION); \ + echo '#define VERSION_STRING "'$(VERSION).$(PATCHLEVEL).$(EXTRAVERSION)'"'; \ + echo '#define FILE_VERSION '$(FILE_VERSION); \ + ) > $1 +endef + +define update_version.h + ($(call make_version.h, $@.tmp); \ + if [ -r $@ ] && cmp -s $@ $@.tmp; then \ + rm -f $@.tmp; \ + else \ + echo ' UPDATE $@'; \ + mv -f $@.tmp $@; \ + fi); +endef + +ep_version.h: force + $(Q)$(N)$(call update_version.h) + +VERSION_FILES = ep_version.h + +define update_dir + (echo $1 > $@.tmp; \ + if [ -r $@ ] && cmp -s $@ $@.tmp; then \ + rm -f $@.tmp; \ + else \ + echo ' UPDATE $@'; \ + mv -f $@.tmp $@; \ + fi); +endef + +## make deps + +all_objs := $(sort $(ALL_OBJS)) +all_deps := $(all_objs:%.o=.%.d) + +define check_deps + $(CC) -M $(CFLAGS) $< > $@; +endef + +$(gui_deps): ks_version.h +$(non_gui_deps): tc_version.h + +$(all_deps): .%.d: $(src)/%.c + $(Q)$(call check_deps) + +$(all_objs) : %.o : .%.d + +dep_includes := $(wildcard $(all_deps)) + +ifneq ($(dep_includes),) + include $(dep_includes) +endif + +tags: force + $(RM) tags + find . -name '*.[ch]' | xargs ctags --extra=+f --c-kinds=+px + +TAGS: force + $(RM) TAGS + find . -name '*.[ch]' | xargs etags + +define do_install + $(print_install) \ + if [ ! -d '$(DESTDIR_SQ)$2' ]; then \ + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$2'; \ + fi; \ + $(INSTALL) $1 '$(DESTDIR_SQ)$2' +endef + +install_lib: all_cmd install_plugins install_python + $(Q)$(call do_install,$(LIB_FILE),$(bindir_SQ)) + +install: install_lib + +clean: + $(RM) *.o *~ $(TARGETS) *.a *.so $(VERSION_FILES).*.d + $(RM) tags TAGS + +endif # skip-makefile + +PHONY += force +force: + +# Declare the contents of the .PHONY variable as phony. We keep that +# information in a variable so we can use it in if_changed and friends. +.PHONY: $(PHONY) diff --git a/tools/lib/traceevent/event-parse.c b/tools/lib/traceevent/event-parse.c new file mode 100644 index 000000000000..554828219c33 --- /dev/null +++ b/tools/lib/traceevent/event-parse.c @@ -0,0 +1,5079 @@ +/* + * Copyright (C) 2009, 2010 Red Hat Inc, Steven Rostedt <srostedt@redhat.com> + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License (not later!) + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + * + * The parts for function graph printing was taken and modified from the + * Linux Kernel that were written by + * - Copyright (C) 2009 Frederic Weisbecker, + * Frederic Weisbecker gave his permission to relicense the code to + * the Lesser General Public License. + */ +#define _GNU_SOURCE +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <stdarg.h> +#include <ctype.h> +#include <errno.h> + +#include "event-parse.h" +#include "event-utils.h" + +static const char *input_buf; +static unsigned long long input_buf_ptr; +static unsigned long long input_buf_siz; + +static int is_flag_field; +static int is_symbolic_field; + +static int show_warning = 1; + +#define do_warning(fmt, ...) \ + do { \ + if (show_warning) \ + warning(fmt, ##__VA_ARGS__); \ + } while (0) + +static void init_input_buf(const char *buf, unsigned long long size) +{ + input_buf = buf; + input_buf_siz = size; + input_buf_ptr = 0; +} + +const char *pevent_get_input_buf(void) +{ + return input_buf; +} + +unsigned long long pevent_get_input_buf_ptr(void) +{ + return input_buf_ptr; +} + +struct event_handler { + struct event_handler *next; + int id; + const char *sys_name; + const char *event_name; + pevent_event_handler_func func; + void *context; +}; + +struct pevent_func_params { + struct pevent_func_params *next; + enum pevent_func_arg_type type; +}; + +struct pevent_function_handler { + struct pevent_function_handler *next; + enum pevent_func_arg_type ret_type; + char *name; + pevent_func_handler func; + struct pevent_func_params *params; + int nr_args; +}; + +static unsigned long long +process_defined_func(struct trace_seq *s, void *data, int size, + struct event_format *event, struct print_arg *arg); + +static void free_func_handle(struct pevent_function_handler *func); + +/** + * pevent_buffer_init - init buffer for parsing + * @buf: buffer to parse + * @size: the size of the buffer + * + * For use with pevent_read_token(), this initializes the internal + * buffer that pevent_read_token() will parse. + */ +void pevent_buffer_init(const char *buf, unsigned long long size) +{ + init_input_buf(buf, size); +} + +void breakpoint(void) +{ + static int x; + x++; +} + +struct print_arg *alloc_arg(void) +{ + struct print_arg *arg; + + arg = malloc_or_die(sizeof(*arg)); + if (!arg) + return NULL; + memset(arg, 0, sizeof(*arg)); + + return arg; +} + +struct cmdline { + char *comm; + int pid; +}; + +static int cmdline_cmp(const void *a, const void *b) +{ + const struct cmdline *ca = a; + const struct cmdline *cb = b; + + if (ca->pid < cb->pid) + return -1; + if (ca->pid > cb->pid) + return 1; + + return 0; +} + +struct cmdline_list { + struct cmdline_list *next; + char *comm; + int pid; +}; + +static int cmdline_init(struct pevent *pevent) +{ + struct cmdline_list *cmdlist = pevent->cmdlist; + struct cmdline_list *item; + struct cmdline *cmdlines; + int i; + + cmdlines = malloc_or_die(sizeof(*cmdlines) * pevent->cmdline_count); + + i = 0; + while (cmdlist) { + cmdlines[i].pid = cmdlist->pid; + cmdlines[i].comm = cmdlist->comm; + i++; + item = cmdlist; + cmdlist = cmdlist->next; + free(item); + } + + qsort(cmdlines, pevent->cmdline_count, sizeof(*cmdlines), cmdline_cmp); + + pevent->cmdlines = cmdlines; + pevent->cmdlist = NULL; + + return 0; +} + +static char *find_cmdline(struct pevent *pevent, int pid) +{ + const struct cmdline *comm; + struct cmdline key; + + if (!pid) + return "<idle>"; + + if (!pevent->cmdlines) + cmdline_init(pevent); + + key.pid = pid; + + comm = bsearch(&key, pevent->cmdlines, pevent->cmdline_count, + sizeof(*pevent->cmdlines), cmdline_cmp); + + if (comm) + return comm->comm; + return "<...>"; +} + +/** + * pevent_pid_is_registered - return if a pid has a cmdline registered + * @pevent: handle for the pevent + * @pid: The pid to check if it has a cmdline registered with. + * + * Returns 1 if the pid has a cmdline mapped to it + * 0 otherwise. + */ +int pevent_pid_is_registered(struct pevent *pevent, int pid) +{ + const struct cmdline *comm; + struct cmdline key; + + if (!pid) + return 1; + + if (!pevent->cmdlines) + cmdline_init(pevent); + + key.pid = pid; + + comm = bsearch(&key, pevent->cmdlines, pevent->cmdline_count, + sizeof(*pevent->cmdlines), cmdline_cmp); + + if (comm) + return 1; + return 0; +} + +/* + * If the command lines have been converted to an array, then + * we must add this pid. This is much slower than when cmdlines + * are added before the array is initialized. + */ +static int add_new_comm(struct pevent *pevent, const char *comm, int pid) +{ + struct cmdline *cmdlines = pevent->cmdlines; + const struct cmdline *cmdline; + struct cmdline key; + + if (!pid) + return 0; + + /* avoid duplicates */ + key.pid = pid; + + cmdline = bsearch(&key, pevent->cmdlines, pevent->cmdline_count, + sizeof(*pevent->cmdlines), cmdline_cmp); + if (cmdline) { + errno = EEXIST; + return -1; + } + + cmdlines = realloc(cmdlines, sizeof(*cmdlines) * (pevent->cmdline_count + 1)); + if (!cmdlines) { + errno = ENOMEM; + return -1; + } + + cmdlines[pevent->cmdline_count].pid = pid; + cmdlines[pevent->cmdline_count].comm = strdup(comm); + if (!cmdlines[pevent->cmdline_count].comm) + die("malloc comm"); + + if (cmdlines[pevent->cmdline_count].comm) + pevent->cmdline_count++; + + qsort(cmdlines, pevent->cmdline_count, sizeof(*cmdlines), cmdline_cmp); + pevent->cmdlines = cmdlines; + + return 0; +} + +/** + * pevent_register_comm - register a pid / comm mapping + * @pevent: handle for the pevent + * @comm: the command line to register + * @pid: the pid to map the command line to + * + * This adds a mapping to search for command line names with + * a given pid. The comm is duplicated. + */ +int pevent_register_comm(struct pevent *pevent, const char *comm, int pid) +{ + struct cmdline_list *item; + + if (pevent->cmdlines) + return add_new_comm(pevent, comm, pid); + + item = malloc_or_die(sizeof(*item)); + item->comm = strdup(comm); + if (!item->comm) + die("malloc comm"); + item->pid = pid; + item->next = pevent->cmdlist; + + pevent->cmdlist = item; + pevent->cmdline_count++; + + return 0; +} + +struct func_map { + unsigned long long addr; + char *func; + char *mod; +}; + +struct func_list { + struct func_list *next; + unsigned long long addr; + char *func; + char *mod; +}; + +static int func_cmp(const void *a, const void *b) +{ + const struct func_map *fa = a; + const struct func_map *fb = b; + + if (fa->addr < fb->addr) + return -1; + if (fa->addr > fb->addr) + return 1; + + return 0; +} + +/* + * We are searching for a record in between, not an exact + * match. + */ +static int func_bcmp(const void *a, const void *b) +{ + const struct func_map *fa = a; + const struct func_map *fb = b; + + if ((fa->addr == fb->addr) || + + (fa->addr > fb->addr && + fa->addr < (fb+1)->addr)) + return 0; + + if (fa->addr < fb->addr) + return -1; + + return 1; +} + +static int func_map_init(struct pevent *pevent) +{ + struct func_list *funclist; + struct func_list *item; + struct func_map *func_map; + int i; + + func_map = malloc_or_die(sizeof(*func_map) * (pevent->func_count + 1)); + funclist = pevent->funclist; + + i = 0; + while (funclist) { + func_map[i].func = funclist->func; + func_map[i].addr = funclist->addr; + func_map[i].mod = funclist->mod; + i++; + item = funclist; + funclist = funclist->next; + free(item); + } + + qsort(func_map, pevent->func_count, sizeof(*func_map), func_cmp); + + /* + * Add a special record at the end. + */ + func_map[pevent->func_count].func = NULL; + func_map[pevent->func_count].addr = 0; + func_map[pevent->func_count].mod = NULL; + + pevent->func_map = func_map; + pevent->funclist = NULL; + + return 0; +} + +static struct func_map * +find_func(struct pevent *pevent, unsigned long long addr) +{ + struct func_map *func; + struct func_map key; + + if (!pevent->func_map) + func_map_init(pevent); + + key.addr = addr; + + func = bsearch(&key, pevent->func_map, pevent->func_count, + sizeof(*pevent->func_map), func_bcmp); + + return func; +} + +/** + * pevent_find_function - find a function by a given address + * @pevent: handle for the pevent + * @addr: the address to find the function with + * + * Returns a pointer to the function stored that has the given + * address. Note, the address does not have to be exact, it + * will select the function that would contain the address. + */ +const char *pevent_find_function(struct pevent *pevent, unsigned long long addr) +{ + struct func_map *map; + + map = find_func(pevent, addr); + if (!map) + return NULL; + + return map->func; +} + +/** + * pevent_find_function_address - find a function address by a given address + * @pevent: handle for the pevent + * @addr: the address to find the function with + * + * Returns the address the function starts at. This can be used in + * conjunction with pevent_find_function to print both the function + * name and the function offset. + */ +unsigned long long +pevent_find_function_address(struct pevent *pevent, unsigned long long addr) +{ + struct func_map *map; + + map = find_func(pevent, addr); + if (!map) + return 0; + + return map->addr; +} + +/** + * pevent_register_function - register a function with a given address + * @pevent: handle for the pevent + * @function: the function name to register + * @addr: the address the function starts at + * @mod: the kernel module the function may be in (NULL for none) + * + * This registers a function name with an address and module. + * The @func passed in is duplicated. + */ +int pevent_register_function(struct pevent *pevent, char *func, + unsigned long long addr, char *mod) +{ + struct func_list *item; + + item = malloc_or_die(sizeof(*item)); + + item->next = pevent->funclist; + item->func = strdup(func); + if (mod) + item->mod = strdup(mod); + else + item->mod = NULL; + item->addr = addr; + + pevent->funclist = item; + + pevent->func_count++; + + return 0; +} + +/** + * pevent_print_funcs - print out the stored functions + * @pevent: handle for the pevent + * + * This prints out the stored functions. + */ +void pevent_print_funcs(struct pevent *pevent) +{ + int i; + + if (!pevent->func_map) + func_map_init(pevent); + + for (i = 0; i < (int)pevent->func_count; i++) { + printf("%016llx %s", + pevent->func_map[i].addr, + pevent->func_map[i].func); + if (pevent->func_map[i].mod) + printf(" [%s]\n", pevent->func_map[i].mod); + else + printf("\n"); + } +} + +struct printk_map { + unsigned long long addr; + char *printk; +}; + +struct printk_list { + struct printk_list *next; + unsigned long long addr; + char *printk; +}; + +static int printk_cmp(const void *a, const void *b) +{ + const struct func_map *fa = a; + const struct func_map *fb = b; + + if (fa->addr < fb->addr) + return -1; + if (fa->addr > fb->addr) + return 1; + + return 0; +} + +static void printk_map_init(struct pevent *pevent) +{ + struct printk_list *printklist; + struct printk_list *item; + struct printk_map *printk_map; + int i; + + printk_map = malloc_or_die(sizeof(*printk_map) * (pevent->printk_count + 1)); + + printklist = pevent->printklist; + + i = 0; + while (printklist) { + printk_map[i].printk = printklist->printk; + printk_map[i].addr = printklist->addr; + i++; + item = printklist; + printklist = printklist->next; + free(item); + } + + qsort(printk_map, pevent->printk_count, sizeof(*printk_map), printk_cmp); + + pevent->printk_map = printk_map; + pevent->printklist = NULL; +} + +static struct printk_map * +find_printk(struct pevent *pevent, unsigned long long addr) +{ + struct printk_map *printk; + struct printk_map key; + + if (!pevent->printk_map) + printk_map_init(pevent); + + key.addr = addr; + + printk = bsearch(&key, pevent->printk_map, pevent->printk_count, + sizeof(*pevent->printk_map), printk_cmp); + + return printk; +} + +/** + * pevent_register_print_string - register a string by its address + * @pevent: handle for the pevent + * @fmt: the string format to register + * @addr: the address the string was located at + * + * This registers a string by the address it was stored in the kernel. + * The @fmt passed in is duplicated. + */ +int pevent_register_print_string(struct pevent *pevent, char *fmt, + unsigned long long addr) +{ + struct printk_list *item; + + item = malloc_or_die(sizeof(*item)); + + item->next = pevent->printklist; + pevent->printklist = item; + item->printk = strdup(fmt); + item->addr = addr; + + pevent->printk_count++; + + return 0; +} + +/** + * pevent_print_printk - print out the stored strings + * @pevent: handle for the pevent + * + * This prints the string formats that were stored. + */ +void pevent_print_printk(struct pevent *pevent) +{ + int i; + + if (!pevent->printk_map) + printk_map_init(pevent); + + for (i = 0; i < (int)pevent->printk_count; i++) { + printf("%016llx %s\n", + pevent->printk_map[i].addr, + pevent->printk_map[i].printk); + } +} + +static struct event_format *alloc_event(void) +{ + struct event_format *event; + + event = malloc_or_die(sizeof(*event)); + memset(event, 0, sizeof(*event)); + + return event; +} + +static void add_event(struct pevent *pevent, struct event_format *event) +{ + int i; + + if (!pevent->events) + pevent->events = malloc_or_die(sizeof(event)); + else + pevent->events = + realloc(pevent->events, sizeof(event) * + (pevent->nr_events + 1)); + if (!pevent->events) + die("Can not allocate events"); + + for (i = 0; i < pevent->nr_events; i++) { + if (pevent->events[i]->id > event->id) + break; + } + if (i < pevent->nr_events) + memmove(&pevent->events[i + 1], + &pevent->events[i], + sizeof(event) * (pevent->nr_events - i)); + + pevent->events[i] = event; + pevent->nr_events++; + + event->pevent = pevent; +} + +static int event_item_type(enum event_type type) +{ + switch (type) { + case EVENT_ITEM ... EVENT_SQUOTE: + return 1; + case EVENT_ERROR ... EVENT_DELIM: + default: + return 0; + } +} + +static void free_flag_sym(struct print_flag_sym *fsym) +{ + struct print_flag_sym *next; + + while (fsym) { + next = fsym->next; + free(fsym->value); + free(fsym->str); + free(fsym); + fsym = next; + } +} + +static void free_arg(struct print_arg *arg) +{ + struct print_arg *farg; + + if (!arg) + return; + + switch (arg->type) { + case PRINT_ATOM: + free(arg->atom.atom); + break; + case PRINT_FIELD: + free(arg->field.name); + break; + case PRINT_FLAGS: + free_arg(arg->flags.field); + free(arg->flags.delim); + free_flag_sym(arg->flags.flags); + break; + case PRINT_SYMBOL: + free_arg(arg->symbol.field); + free_flag_sym(arg->symbol.symbols); + break; + case PRINT_TYPE: + free(arg->typecast.type); + free_arg(arg->typecast.item); + break; + case PRINT_STRING: + case PRINT_BSTRING: + free(arg->string.string); + break; + case PRINT_DYNAMIC_ARRAY: + free(arg->dynarray.index); + break; + case PRINT_OP: + free(arg->op.op); + free_arg(arg->op.left); + free_arg(arg->op.right); + break; + case PRINT_FUNC: + while (arg->func.args) { + farg = arg->func.args; + arg->func.args = farg->next; + free_arg(farg); + } + break; + + case PRINT_NULL: + default: + break; + } + + free(arg); +} + +static enum event_type get_type(int ch) +{ + if (ch == '\n') + return EVENT_NEWLINE; + if (isspace(ch)) + return EVENT_SPACE; + if (isalnum(ch) || ch == '_') + return EVENT_ITEM; + if (ch == '\'') + return EVENT_SQUOTE; + if (ch == '"') + return EVENT_DQUOTE; + if (!isprint(ch)) + return EVENT_NONE; + if (ch == '(' || ch == ')' || ch == ',') + return EVENT_DELIM; + + return EVENT_OP; +} + +static int __read_char(void) +{ + if (input_buf_ptr >= input_buf_siz) + return -1; + + return input_buf[input_buf_ptr++]; +} + +static int __peek_char(void) +{ + if (input_buf_ptr >= input_buf_siz) + return -1; + + return input_buf[input_buf_ptr]; +} + +/** + * pevent_peek_char - peek at the next character that will be read + * + * Returns the next character read, or -1 if end of buffer. + */ +int pevent_peek_char(void) +{ + return __peek_char(); +} + +static enum event_type force_token(const char *str, char **tok); + +static enum event_type __read_token(char **tok) +{ + char buf[BUFSIZ]; + int ch, last_ch, quote_ch, next_ch; + int i = 0; + int tok_size = 0; + enum event_type type; + + *tok = NULL; + + + ch = __read_char(); + if (ch < 0) + return EVENT_NONE; + + type = get_type(ch); + if (type == EVENT_NONE) + return type; + + buf[i++] = ch; + + switch (type) { + case EVENT_NEWLINE: + case EVENT_DELIM: + *tok = malloc_or_die(2); + (*tok)[0] = ch; + (*tok)[1] = 0; + return type; + + case EVENT_OP: + switch (ch) { + case '-': + next_ch = __peek_char(); + if (next_ch == '>') { + buf[i++] = __read_char(); + break; + } + /* fall through */ + case '+': + case '|': + case '&': + case '>': + case '<': + last_ch = ch; + ch = __peek_char(); + if (ch != last_ch) + goto test_equal; + buf[i++] = __read_char(); + switch (last_ch) { + case '>': + case '<': + goto test_equal; + default: + break; + } + break; + case '!': + case '=': + goto test_equal; + default: /* what should we do instead? */ + break; + } + buf[i] = 0; + *tok = strdup(buf); + return type; + + test_equal: + ch = __peek_char(); + if (ch == '=') + buf[i++] = __read_char(); + goto out; + + case EVENT_DQUOTE: + case EVENT_SQUOTE: + /* don't keep quotes */ + i--; + quote_ch = ch; + last_ch = 0; + concat: + do { + if (i == (BUFSIZ - 1)) { + buf[i] = 0; + if (*tok) { + *tok = realloc(*tok, tok_size + BUFSIZ); + if (!*tok) + return EVENT_NONE; + strcat(*tok, buf); + } else + *tok = strdup(buf); + + if (!*tok) + return EVENT_NONE; + tok_size += BUFSIZ; + i = 0; + } + last_ch = ch; + ch = __read_char(); + buf[i++] = ch; + /* the '\' '\' will cancel itself */ + if (ch == '\\' && last_ch == '\\') + last_ch = 0; + } while (ch != quote_ch || last_ch == '\\'); + /* remove the last quote */ + i--; + + /* + * For strings (double quotes) check the next token. + * If it is another string, concatinate the two. + */ + if (type == EVENT_DQUOTE) { + unsigned long long save_input_buf_ptr = input_buf_ptr; + + do { + ch = __read_char(); + } while (isspace(ch)); + if (ch == '"') + goto concat; + input_buf_ptr = save_input_buf_ptr; + } + + goto out; + + case EVENT_ERROR ... EVENT_SPACE: + case EVENT_ITEM: + default: + break; + } + + while (get_type(__peek_char()) == type) { + if (i == (BUFSIZ - 1)) { + buf[i] = 0; + if (*tok) { + *tok = realloc(*tok, tok_size + BUFSIZ); + if (!*tok) + return EVENT_NONE; + strcat(*tok, buf); + } else + *tok = strdup(buf); + + if (!*tok) + return EVENT_NONE; + tok_size += BUFSIZ; + i = 0; + } + ch = __read_char(); + buf[i++] = ch; + } + + out: + buf[i] = 0; + if (*tok) { + *tok = realloc(*tok, tok_size + i); + if (!*tok) + return EVENT_NONE; + strcat(*tok, buf); + } else + *tok = strdup(buf); + if (!*tok) + return EVENT_NONE; + + if (type == EVENT_ITEM) { + /* + * Older versions of the kernel has a bug that + * creates invalid symbols and will break the mac80211 + * parsing. This is a work around to that bug. + * + * See Linux kernel commit: + * 811cb50baf63461ce0bdb234927046131fc7fa8b + */ + if (strcmp(*tok, "LOCAL_PR_FMT") == 0) { + free(*tok); + *tok = NULL; + return force_token("\"\%s\" ", tok); + } else if (strcmp(*tok, "STA_PR_FMT") == 0) { + free(*tok); + *tok = NULL; + return force_token("\" sta:%pM\" ", tok); + } else if (strcmp(*tok, "VIF_PR_FMT") == 0) { + free(*tok); + *tok = NULL; + return force_token("\" vif:%p(%d)\" ", tok); + } + } + + return type; +} + +static enum event_type force_token(const char *str, char **tok) +{ + const char *save_input_buf; + unsigned long long save_input_buf_ptr; + unsigned long long save_input_buf_siz; + enum event_type type; + + /* save off the current input pointers */ + save_input_buf = input_buf; + save_input_buf_ptr = input_buf_ptr; + save_input_buf_siz = input_buf_siz; + + init_input_buf(str, strlen(str)); + + type = __read_token(tok); + + /* reset back to original token */ + input_buf = save_input_buf; + input_buf_ptr = save_input_buf_ptr; + input_buf_siz = save_input_buf_siz; + + return type; +} + +static void free_token(char *tok) +{ + if (tok) + free(tok); +} + +static enum event_type read_token(char **tok) +{ + enum event_type type; + + for (;;) { + type = __read_token(tok); + if (type != EVENT_SPACE) + return type; + + free_token(*tok); + } + + /* not reached */ + *tok = NULL; + return EVENT_NONE; +} + +/** + * pevent_read_token - access to utilites to use the pevent parser + * @tok: The token to return + * + * This will parse tokens from the string given by + * pevent_init_data(). + * + * Returns the token type. + */ +enum event_type pevent_read_token(char **tok) +{ + return read_token(tok); +} + +/** + * pevent_free_token - free a token returned by pevent_read_token + * @token: the token to free + */ +void pevent_free_token(char *token) +{ + free_token(token); +} + +/* no newline */ +static enum event_type read_token_item(char **tok) +{ + enum event_type type; + + for (;;) { + type = __read_token(tok); + if (type != EVENT_SPACE && type != EVENT_NEWLINE) + return type; + free_token(*tok); + *tok = NULL; + } + + /* not reached */ + *tok = NULL; + return EVENT_NONE; +} + +static int test_type(enum event_type type, enum event_type expect) +{ + if (type != expect) { + do_warning("Error: expected type %d but read %d", + expect, type); + return -1; + } + return 0; +} + +static int test_type_token(enum event_type type, const char *token, + enum event_type expect, const char *expect_tok) +{ + if (type != expect) { + do_warning("Error: expected type %d but read %d", + expect, type); + return -1; + } + + if (strcmp(token, expect_tok) != 0) { + do_warning("Error: expected '%s' but read '%s'", + expect_tok, token); + return -1; + } + return 0; +} + +static int __read_expect_type(enum event_type expect, char **tok, int newline_ok) +{ + enum event_type type; + + if (newline_ok) + type = read_token(tok); + else + type = read_token_item(tok); + return test_type(type, expect); +} + +static int read_expect_type(enum event_type expect, char **tok) +{ + return __read_expect_type(expect, tok, 1); +} + +static int __read_expected(enum event_type expect, const char *str, + int newline_ok) +{ + enum event_type type; + char *token; + int ret; + + if (newline_ok) + type = read_token(&token); + else + type = read_token_item(&token); + + ret = test_type_token(type, token, expect, str); + + free_token(token); + + return ret; +} + +static int read_expected(enum event_type expect, const char *str) +{ + return __read_expected(expect, str, 1); +} + +static int read_expected_item(enum event_type expect, const char *str) +{ + return __read_expected(expect, str, 0); +} + +static char *event_read_name(void) +{ + char *token; + + if (read_expected(EVENT_ITEM, "name") < 0) + return NULL; + + if (read_expected(EVENT_OP, ":") < 0) + return NULL; + + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + + return token; + + fail: + free_token(token); + return NULL; +} + +static int event_read_id(void) +{ + char *token; + int id; + + if (read_expected_item(EVENT_ITEM, "ID") < 0) + return -1; + + if (read_expected(EVENT_OP, ":") < 0) + return -1; + + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + + id = strtoul(token, NULL, 0); + free_token(token); + return id; + + fail: + free_token(token); + return -1; +} + +static int field_is_string(struct format_field *field) +{ + if ((field->flags & FIELD_IS_ARRAY) && + (strstr(field->type, "char") || strstr(field->type, "u8") || + strstr(field->type, "s8"))) + return 1; + + return 0; +} + +static int field_is_dynamic(struct format_field *field) +{ + if (strncmp(field->type, "__data_loc", 10) == 0) + return 1; + + return 0; +} + +static int field_is_long(struct format_field *field) +{ + /* includes long long */ + if (strstr(field->type, "long")) + return 1; + + return 0; +} + +static int event_read_fields(struct event_format *event, struct format_field **fields) +{ + struct format_field *field = NULL; + enum event_type type; + char *token; + char *last_token; + int count = 0; + + do { + type = read_token(&token); + if (type == EVENT_NEWLINE) { + free_token(token); + return count; + } + + count++; + + if (test_type_token(type, token, EVENT_ITEM, "field")) + goto fail; + free_token(token); + + type = read_token(&token); + /* + * The ftrace fields may still use the "special" name. + * Just ignore it. + */ + if (event->flags & EVENT_FL_ISFTRACE && + type == EVENT_ITEM && strcmp(token, "special") == 0) { + free_token(token); + type = read_token(&token); + } + + if (test_type_token(type, token, EVENT_OP, ":") < 0) + goto fail; + + free_token(token); + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + + last_token = token; + + field = malloc_or_die(sizeof(*field)); + memset(field, 0, sizeof(*field)); + field->event = event; + + /* read the rest of the type */ + for (;;) { + type = read_token(&token); + if (type == EVENT_ITEM || + (type == EVENT_OP && strcmp(token, "*") == 0) || + /* + * Some of the ftrace fields are broken and have + * an illegal "." in them. + */ + (event->flags & EVENT_FL_ISFTRACE && + type == EVENT_OP && strcmp(token, ".") == 0)) { + + if (strcmp(token, "*") == 0) + field->flags |= FIELD_IS_POINTER; + + if (field->type) { + field->type = realloc(field->type, + strlen(field->type) + + strlen(last_token) + 2); + strcat(field->type, " "); + strcat(field->type, last_token); + free(last_token); + } else + field->type = last_token; + last_token = token; + continue; + } + + break; + } + + if (!field->type) { + die("no type found"); + goto fail; + } + field->name = last_token; + + if (test_type(type, EVENT_OP)) + goto fail; + + if (strcmp(token, "[") == 0) { + enum event_type last_type = type; + char *brackets = token; + int len; + + field->flags |= FIELD_IS_ARRAY; + + type = read_token(&token); + + if (type == EVENT_ITEM) + field->arraylen = strtoul(token, NULL, 0); + else + field->arraylen = 0; + + while (strcmp(token, "]") != 0) { + if (last_type == EVENT_ITEM && + type == EVENT_ITEM) + len = 2; + else + len = 1; + last_type = type; + + brackets = realloc(brackets, + strlen(brackets) + + strlen(token) + len); + if (len == 2) + strcat(brackets, " "); + strcat(brackets, token); + /* We only care about the last token */ + field->arraylen = strtoul(token, NULL, 0); + free_token(token); + type = read_token(&token); + if (type == EVENT_NONE) { + die("failed to find token"); + goto fail; + } + } + + free_token(token); + + brackets = realloc(brackets, strlen(brackets) + 2); + strcat(brackets, "]"); + + /* add brackets to type */ + + type = read_token(&token); + /* + * If the next token is not an OP, then it is of + * the format: type [] item; + */ + if (type == EVENT_ITEM) { + field->type = realloc(field->type, + strlen(field->type) + + strlen(field->name) + + strlen(brackets) + 2); + strcat(field->type, " "); + strcat(field->type, field->name); + free_token(field->name); + strcat(field->type, brackets); + field->name = token; + type = read_token(&token); + } else { + field->type = realloc(field->type, + strlen(field->type) + + strlen(brackets) + 1); + strcat(field->type, brackets); + } + free(brackets); + } + + if (field_is_string(field)) + field->flags |= FIELD_IS_STRING; + if (field_is_dynamic(field)) + field->flags |= FIELD_IS_DYNAMIC; + if (field_is_long(field)) + field->flags |= FIELD_IS_LONG; + + if (test_type_token(type, token, EVENT_OP, ";")) + goto fail; + free_token(token); + + if (read_expected(EVENT_ITEM, "offset") < 0) + goto fail_expect; + + if (read_expected(EVENT_OP, ":") < 0) + goto fail_expect; + + if (read_expect_type(EVENT_ITEM, &token)) + goto fail; + field->offset = strtoul(token, NULL, 0); + free_token(token); + + if (read_expected(EVENT_OP, ";") < 0) + goto fail_expect; + + if (read_expected(EVENT_ITEM, "size") < 0) + goto fail_expect; + + if (read_expected(EVENT_OP, ":") < 0) + goto fail_expect; + + if (read_expect_type(EVENT_ITEM, &token)) + goto fail; + field->size = strtoul(token, NULL, 0); + free_token(token); + + if (read_expected(EVENT_OP, ";") < 0) + goto fail_expect; + + type = read_token(&token); + if (type != EVENT_NEWLINE) { + /* newer versions of the kernel have a "signed" type */ + if (test_type_token(type, token, EVENT_ITEM, "signed")) + goto fail; + + free_token(token); + + if (read_expected(EVENT_OP, ":") < 0) + goto fail_expect; + + if (read_expect_type(EVENT_ITEM, &token)) + goto fail; + + /* add signed type */ + + free_token(token); + if (read_expected(EVENT_OP, ";") < 0) + goto fail_expect; + + if (read_expect_type(EVENT_NEWLINE, &token)) + goto fail; + } + + free_token(token); + + if (field->flags & FIELD_IS_ARRAY) { + if (field->arraylen) + field->elementsize = field->size / field->arraylen; + else if (field->flags & FIELD_IS_STRING) + field->elementsize = 1; + else + field->elementsize = event->pevent->long_size; + } else + field->elementsize = field->size; + + *fields = field; + fields = &field->next; + + } while (1); + + return 0; + +fail: + free_token(token); +fail_expect: + if (field) { + free(field->type); + free(field->name); + free(field); + } + return -1; +} + +static int event_read_format(struct event_format *event) +{ + char *token; + int ret; + + if (read_expected_item(EVENT_ITEM, "format") < 0) + return -1; + + if (read_expected(EVENT_OP, ":") < 0) + return -1; + + if (read_expect_type(EVENT_NEWLINE, &token)) + goto fail; + free_token(token); + + ret = event_read_fields(event, &event->format.common_fields); + if (ret < 0) + return ret; + event->format.nr_common = ret; + + ret = event_read_fields(event, &event->format.fields); + if (ret < 0) + return ret; + event->format.nr_fields = ret; + + return 0; + + fail: + free_token(token); + return -1; +} + +static enum event_type +process_arg_token(struct event_format *event, struct print_arg *arg, + char **tok, enum event_type type); + +static enum event_type +process_arg(struct event_format *event, struct print_arg *arg, char **tok) +{ + enum event_type type; + char *token; + + type = read_token(&token); + *tok = token; + + return process_arg_token(event, arg, tok, type); +} + +static enum event_type +process_op(struct event_format *event, struct print_arg *arg, char **tok); + +static enum event_type +process_cond(struct event_format *event, struct print_arg *top, char **tok) +{ + struct print_arg *arg, *left, *right; + enum event_type type; + char *token = NULL; + + arg = alloc_arg(); + left = alloc_arg(); + right = alloc_arg(); + + arg->type = PRINT_OP; + arg->op.left = left; + arg->op.right = right; + + *tok = NULL; + type = process_arg(event, left, &token); + + again: + /* Handle other operations in the arguments */ + if (type == EVENT_OP && strcmp(token, ":") != 0) { + type = process_op(event, left, &token); + goto again; + } + + if (test_type_token(type, token, EVENT_OP, ":")) + goto out_free; + + arg->op.op = token; + + type = process_arg(event, right, &token); + + top->op.right = arg; + + *tok = token; + return type; + +out_free: + /* Top may point to itself */ + top->op.right = NULL; + free_token(token); + free_arg(arg); + return EVENT_ERROR; +} + +static enum event_type +process_array(struct event_format *event, struct print_arg *top, char **tok) +{ + struct print_arg *arg; + enum event_type type; + char *token = NULL; + + arg = alloc_arg(); + + *tok = NULL; + type = process_arg(event, arg, &token); + if (test_type_token(type, token, EVENT_OP, "]")) + goto out_free; + + top->op.right = arg; + + free_token(token); + type = read_token_item(&token); + *tok = token; + + return type; + +out_free: + free_token(*tok); + *tok = NULL; + free_arg(arg); + return EVENT_ERROR; +} + +static int get_op_prio(char *op) +{ + if (!op[1]) { + switch (op[0]) { + case '~': + case '!': + return 4; + case '*': + case '/': + case '%': + return 6; + case '+': + case '-': + return 7; + /* '>>' and '<<' are 8 */ + case '<': + case '>': + return 9; + /* '==' and '!=' are 10 */ + case '&': + return 11; + case '^': + return 12; + case '|': + return 13; + case '?': + return 16; + default: + do_warning("unknown op '%c'", op[0]); + return -1; + } + } else { + if (strcmp(op, "++") == 0 || + strcmp(op, "--") == 0) { + return 3; + } else if (strcmp(op, ">>") == 0 || + strcmp(op, "<<") == 0) { + return 8; + } else if (strcmp(op, ">=") == 0 || + strcmp(op, "<=") == 0) { + return 9; + } else if (strcmp(op, "==") == 0 || + strcmp(op, "!=") == 0) { + return 10; + } else if (strcmp(op, "&&") == 0) { + return 14; + } else if (strcmp(op, "||") == 0) { + return 15; + } else { + do_warning("unknown op '%s'", op); + return -1; + } + } +} + +static int set_op_prio(struct print_arg *arg) +{ + + /* single ops are the greatest */ + if (!arg->op.left || arg->op.left->type == PRINT_NULL) + arg->op.prio = 0; + else + arg->op.prio = get_op_prio(arg->op.op); + + return arg->op.prio; +} + +/* Note, *tok does not get freed, but will most likely be saved */ +static enum event_type +process_op(struct event_format *event, struct print_arg *arg, char **tok) +{ + struct print_arg *left, *right = NULL; + enum event_type type; + char *token; + + /* the op is passed in via tok */ + token = *tok; + + if (arg->type == PRINT_OP && !arg->op.left) { + /* handle single op */ + if (token[1]) { + die("bad op token %s", token); + goto out_free; + } + switch (token[0]) { + case '~': + case '!': + case '+': + case '-': + break; + default: + do_warning("bad op token %s", token); + goto out_free; + + } + + /* make an empty left */ + left = alloc_arg(); + left->type = PRINT_NULL; + arg->op.left = left; + + right = alloc_arg(); + arg->op.right = right; + + /* do not free the token, it belongs to an op */ + *tok = NULL; + type = process_arg(event, right, tok); + + } else if (strcmp(token, "?") == 0) { + + left = alloc_arg(); + /* copy the top arg to the left */ + *left = *arg; + + arg->type = PRINT_OP; + arg->op.op = token; + arg->op.left = left; + arg->op.prio = 0; + + type = process_cond(event, arg, tok); + + } else if (strcmp(token, ">>") == 0 || + strcmp(token, "<<") == 0 || + strcmp(token, "&") == 0 || + strcmp(token, "|") == 0 || + strcmp(token, "&&") == 0 || + strcmp(token, "||") == 0 || + strcmp(token, "-") == 0 || + strcmp(token, "+") == 0 || + strcmp(token, "*") == 0 || + strcmp(token, "^") == 0 || + strcmp(token, "/") == 0 || + strcmp(token, "<") == 0 || + strcmp(token, ">") == 0 || + strcmp(token, "==") == 0 || + strcmp(token, "!=") == 0) { + + left = alloc_arg(); + + /* copy the top arg to the left */ + *left = *arg; + + arg->type = PRINT_OP; + arg->op.op = token; + arg->op.left = left; + + if (set_op_prio(arg) == -1) { + event->flags |= EVENT_FL_FAILED; + /* arg->op.op (= token) will be freed at out_free */ + arg->op.op = NULL; + goto out_free; + } + + type = read_token_item(&token); + *tok = token; + + /* could just be a type pointer */ + if ((strcmp(arg->op.op, "*") == 0) && + type == EVENT_DELIM && (strcmp(token, ")") == 0)) { + if (left->type != PRINT_ATOM) + die("bad pointer type"); + left->atom.atom = realloc(left->atom.atom, + strlen(left->atom.atom) + 3); + strcat(left->atom.atom, " *"); + free(arg->op.op); + *arg = *left; + free(left); + + return type; + } + + right = alloc_arg(); + type = process_arg_token(event, right, tok, type); + arg->op.right = right; + + } else if (strcmp(token, "[") == 0) { + + left = alloc_arg(); + *left = *arg; + + arg->type = PRINT_OP; + arg->op.op = token; + arg->op.left = left; + + arg->op.prio = 0; + + type = process_array(event, arg, tok); + + } else { + do_warning("unknown op '%s'", token); + event->flags |= EVENT_FL_FAILED; + /* the arg is now the left side */ + goto out_free; + } + + if (type == EVENT_OP && strcmp(*tok, ":") != 0) { + int prio; + + /* higher prios need to be closer to the root */ + prio = get_op_prio(*tok); + + if (prio > arg->op.prio) + return process_op(event, arg, tok); + + return process_op(event, right, tok); + } + + return type; + + out_free: + free_token(token); + *tok = NULL; + return EVENT_ERROR; +} + +static enum event_type +process_entry(struct event_format *event __unused, struct print_arg *arg, + char **tok) +{ + enum event_type type; + char *field; + char *token; + + if (read_expected(EVENT_OP, "->") < 0) + goto out_err; + + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto out_free; + field = token; + + arg->type = PRINT_FIELD; + arg->field.name = field; + + if (is_flag_field) { + arg->field.field = pevent_find_any_field(event, arg->field.name); + arg->field.field->flags |= FIELD_IS_FLAG; + is_flag_field = 0; + } else if (is_symbolic_field) { + arg->field.field = pevent_find_any_field(event, arg->field.name); + arg->field.field->flags |= FIELD_IS_SYMBOLIC; + is_symbolic_field = 0; + } + + type = read_token(&token); + *tok = token; + + return type; + + out_free: + free_token(token); + out_err: + *tok = NULL; + return EVENT_ERROR; +} + +static char *arg_eval (struct print_arg *arg); + +static unsigned long long +eval_type_str(unsigned long long val, const char *type, int pointer) +{ + int sign = 0; + char *ref; + int len; + + len = strlen(type); + + if (pointer) { + + if (type[len-1] != '*') { + do_warning("pointer expected with non pointer type"); + return val; + } + + ref = malloc_or_die(len); + memcpy(ref, type, len); + + /* chop off the " *" */ + ref[len - 2] = 0; + + val = eval_type_str(val, ref, 0); + free(ref); + return val; + } + + /* check if this is a pointer */ + if (type[len - 1] == '*') + return val; + + /* Try to figure out the arg size*/ + if (strncmp(type, "struct", 6) == 0) + /* all bets off */ + return val; + + if (strcmp(type, "u8") == 0) + return val & 0xff; + + if (strcmp(type, "u16") == 0) + return val & 0xffff; + + if (strcmp(type, "u32") == 0) + return val & 0xffffffff; + + if (strcmp(type, "u64") == 0 || + strcmp(type, "s64")) + return val; + + if (strcmp(type, "s8") == 0) + return (unsigned long long)(char)val & 0xff; + + if (strcmp(type, "s16") == 0) + return (unsigned long long)(short)val & 0xffff; + + if (strcmp(type, "s32") == 0) + return (unsigned long long)(int)val & 0xffffffff; + + if (strncmp(type, "unsigned ", 9) == 0) { + sign = 0; + type += 9; + } + + if (strcmp(type, "char") == 0) { + if (sign) + return (unsigned long long)(char)val & 0xff; + else + return val & 0xff; + } + + if (strcmp(type, "short") == 0) { + if (sign) + return (unsigned long long)(short)val & 0xffff; + else + return val & 0xffff; + } + + if (strcmp(type, "int") == 0) { + if (sign) + return (unsigned long long)(int)val & 0xffffffff; + else + return val & 0xffffffff; + } + + return val; +} + +/* + * Try to figure out the type. + */ +static unsigned long long +eval_type(unsigned long long val, struct print_arg *arg, int pointer) +{ + if (arg->type != PRINT_TYPE) + die("expected type argument"); + + return eval_type_str(val, arg->typecast.type, pointer); +} + +static int arg_num_eval(struct print_arg *arg, long long *val) +{ + long long left, right; + int ret = 1; + + switch (arg->type) { + case PRINT_ATOM: + *val = strtoll(arg->atom.atom, NULL, 0); + break; + case PRINT_TYPE: + ret = arg_num_eval(arg->typecast.item, val); + if (!ret) + break; + *val = eval_type(*val, arg, 0); + break; + case PRINT_OP: + switch (arg->op.op[0]) { + case '|': + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + if (arg->op.op[1]) + *val = left || right; + else + *val = left | right; + break; + case '&': + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + if (arg->op.op[1]) + *val = left && right; + else + *val = left & right; + break; + case '<': + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + switch (arg->op.op[1]) { + case 0: + *val = left < right; + break; + case '<': + *val = left << right; + break; + case '=': + *val = left <= right; + break; + default: + do_warning("unknown op '%s'", arg->op.op); + ret = 0; + } + break; + case '>': + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + switch (arg->op.op[1]) { + case 0: + *val = left > right; + break; + case '>': + *val = left >> right; + break; + case '=': + *val = left >= right; + break; + default: + do_warning("unknown op '%s'", arg->op.op); + ret = 0; + } + break; + case '=': + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + + if (arg->op.op[1] != '=') { + do_warning("unknown op '%s'", arg->op.op); + ret = 0; + } else + *val = left == right; + break; + case '!': + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + + switch (arg->op.op[1]) { + case '=': + *val = left != right; + break; + default: + do_warning("unknown op '%s'", arg->op.op); + ret = 0; + } + break; + case '-': + /* check for negative */ + if (arg->op.left->type == PRINT_NULL) + left = 0; + else + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + *val = left - right; + break; + case '+': + if (arg->op.left->type == PRINT_NULL) + left = 0; + else + ret = arg_num_eval(arg->op.left, &left); + if (!ret) + break; + ret = arg_num_eval(arg->op.right, &right); + if (!ret) + break; + *val = left + right; + break; + default: + do_warning("unknown op '%s'", arg->op.op); + ret = 0; + } + break; + + case PRINT_NULL: + case PRINT_FIELD ... PRINT_SYMBOL: + case PRINT_STRING: + case PRINT_BSTRING: + default: + do_warning("invalid eval type %d", arg->type); + ret = 0; + + } + return ret; +} + +static char *arg_eval (struct print_arg *arg) +{ + long long val; + static char buf[20]; + + switch (arg->type) { + case PRINT_ATOM: + return arg->atom.atom; + case PRINT_TYPE: + return arg_eval(arg->typecast.item); + case PRINT_OP: + if (!arg_num_eval(arg, &val)) + break; + sprintf(buf, "%lld", val); + return buf; + + case PRINT_NULL: + case PRINT_FIELD ... PRINT_SYMBOL: + case PRINT_STRING: + case PRINT_BSTRING: + default: + die("invalid eval type %d", arg->type); + break; + } + + return NULL; +} + +static enum event_type +process_fields(struct event_format *event, struct print_flag_sym **list, char **tok) +{ + enum event_type type; + struct print_arg *arg = NULL; + struct print_flag_sym *field; + char *token = *tok; + char *value; + + do { + free_token(token); + type = read_token_item(&token); + if (test_type_token(type, token, EVENT_OP, "{")) + break; + + arg = alloc_arg(); + + free_token(token); + type = process_arg(event, arg, &token); + + if (type == EVENT_OP) + type = process_op(event, arg, &token); + + if (type == EVENT_ERROR) + goto out_free; + + if (test_type_token(type, token, EVENT_DELIM, ",")) + goto out_free; + + field = malloc_or_die(sizeof(*field)); + memset(field, 0, sizeof(*field)); + + value = arg_eval(arg); + if (value == NULL) + goto out_free; + field->value = strdup(value); + + free_arg(arg); + arg = alloc_arg(); + + free_token(token); + type = process_arg(event, arg, &token); + if (test_type_token(type, token, EVENT_OP, "}")) + goto out_free; + + value = arg_eval(arg); + if (value == NULL) + goto out_free; + field->str = strdup(value); + free_arg(arg); + arg = NULL; + + *list = field; + list = &field->next; + + free_token(token); + type = read_token_item(&token); + } while (type == EVENT_DELIM && strcmp(token, ",") == 0); + + *tok = token; + return type; + +out_free: + free_arg(arg); + free_token(token); + *tok = NULL; + + return EVENT_ERROR; +} + +static enum event_type +process_flags(struct event_format *event, struct print_arg *arg, char **tok) +{ + struct print_arg *field; + enum event_type type; + char *token; + + memset(arg, 0, sizeof(*arg)); + arg->type = PRINT_FLAGS; + + field = alloc_arg(); + + type = process_arg(event, field, &token); + + /* Handle operations in the first argument */ + while (type == EVENT_OP) + type = process_op(event, field, &token); + + if (test_type_token(type, token, EVENT_DELIM, ",")) + goto out_free; + free_token(token); + + arg->flags.field = field; + + type = read_token_item(&token); + if (event_item_type(type)) { + arg->flags.delim = token; + type = read_token_item(&token); + } + + if (test_type_token(type, token, EVENT_DELIM, ",")) + goto out_free; + + type = process_fields(event, &arg->flags.flags, &token); + if (test_type_token(type, token, EVENT_DELIM, ")")) + goto out_free; + + free_token(token); + type = read_token_item(tok); + return type; + + out_free: + free_token(token); + *tok = NULL; + return EVENT_ERROR; +} + +static enum event_type +process_symbols(struct event_format *event, struct print_arg *arg, char **tok) +{ + struct print_arg *field; + enum event_type type; + char *token; + + memset(arg, 0, sizeof(*arg)); + arg->type = PRINT_SYMBOL; + + field = alloc_arg(); + + type = process_arg(event, field, &token); + if (test_type_token(type, token, EVENT_DELIM, ",")) + goto out_free; + + arg->symbol.field = field; + + type = process_fields(event, &arg->symbol.symbols, &token); + if (test_type_token(type, token, EVENT_DELIM, ")")) + goto out_free; + + free_token(token); + type = read_token_item(tok); + return type; + + out_free: + free_token(token); + *tok = NULL; + return EVENT_ERROR; +} + +static enum event_type +process_dynamic_array(struct event_format *event, struct print_arg *arg, char **tok) +{ + struct format_field *field; + enum event_type type; + char *token; + + memset(arg, 0, sizeof(*arg)); + arg->type = PRINT_DYNAMIC_ARRAY; + + /* + * The item within the parenthesis is another field that holds + * the index into where the array starts. + */ + type = read_token(&token); + *tok = token; + if (type != EVENT_ITEM) + goto out_free; + + /* Find the field */ + + field = pevent_find_field(event, token); + if (!field) + goto out_free; + + arg->dynarray.field = field; + arg->dynarray.index = 0; + + if (read_expected(EVENT_DELIM, ")") < 0) + goto out_free; + + free_token(token); + type = read_token_item(&token); + *tok = token; + if (type != EVENT_OP || strcmp(token, "[") != 0) + return type; + + free_token(token); + arg = alloc_arg(); + type = process_arg(event, arg, &token); + if (type == EVENT_ERROR) + goto out_free_arg; + + if (!test_type_token(type, token, EVENT_OP, "]")) + goto out_free_arg; + + free_token(token); + type = read_token_item(tok); + return type; + + out_free_arg: + free_arg(arg); + out_free: + free_token(token); + *tok = NULL; + return EVENT_ERROR; +} + +static enum event_type +process_paren(struct event_format *event, struct print_arg *arg, char **tok) +{ + struct print_arg *item_arg; + enum event_type type; + char *token; + + type = process_arg(event, arg, &token); + + if (type == EVENT_ERROR) + goto out_free; + + if (type == EVENT_OP) + type = process_op(event, arg, &token); + + if (type == EVENT_ERROR) + goto out_free; + + if (test_type_token(type, token, EVENT_DELIM, ")")) + goto out_free; + + free_token(token); + type = read_token_item(&token); + + /* + * If the next token is an item or another open paren, then + * this was a typecast. + */ + if (event_item_type(type) || + (type == EVENT_DELIM && strcmp(token, "(") == 0)) { + + /* make this a typecast and contine */ + + /* prevous must be an atom */ + if (arg->type != PRINT_ATOM) + die("previous needed to be PRINT_ATOM"); + + item_arg = alloc_arg(); + + arg->type = PRINT_TYPE; + arg->typecast.type = arg->atom.atom; + arg->typecast.item = item_arg; + type = process_arg_token(event, item_arg, &token, type); + + } + + *tok = token; + return type; + + out_free: + free_token(token); + *tok = NULL; + return EVENT_ERROR; +} + + +static enum event_type +process_str(struct event_format *event __unused, struct print_arg *arg, char **tok) +{ + enum event_type type; + char *token; + + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto out_free; + + arg->type = PRINT_STRING; + arg->string.string = token; + arg->string.offset = -1; + + if (read_expected(EVENT_DELIM, ")") < 0) + goto out_err; + + type = read_token(&token); + *tok = token; + + return type; + + out_free: + free_token(token); + out_err: + *tok = NULL; + return EVENT_ERROR; +} + +static struct pevent_function_handler * +find_func_handler(struct pevent *pevent, char *func_name) +{ + struct pevent_function_handler *func; + + for (func = pevent->func_handlers; func; func = func->next) { + if (strcmp(func->name, func_name) == 0) + break; + } + + return func; +} + +static void remove_func_handler(struct pevent *pevent, char *func_name) +{ + struct pevent_function_handler *func; + struct pevent_function_handler **next; + + next = &pevent->func_handlers; + while ((func = *next)) { + if (strcmp(func->name, func_name) == 0) { + *next = func->next; + free_func_handle(func); + break; + } + next = &func->next; + } +} + +static enum event_type +process_func_handler(struct event_format *event, struct pevent_function_handler *func, + struct print_arg *arg, char **tok) +{ + struct print_arg **next_arg; + struct print_arg *farg; + enum event_type type; + char *token; + char *test; + int i; + + arg->type = PRINT_FUNC; + arg->func.func = func; + + *tok = NULL; + + next_arg = &(arg->func.args); + for (i = 0; i < func->nr_args; i++) { + farg = alloc_arg(); + type = process_arg(event, farg, &token); + if (i < (func->nr_args - 1)) + test = ","; + else + test = ")"; + + if (test_type_token(type, token, EVENT_DELIM, test)) { + free_arg(farg); + free_token(token); + return EVENT_ERROR; + } + + *next_arg = farg; + next_arg = &(farg->next); + free_token(token); + } + + type = read_token(&token); + *tok = token; + + return type; +} + +static enum event_type +process_function(struct event_format *event, struct print_arg *arg, + char *token, char **tok) +{ + struct pevent_function_handler *func; + + if (strcmp(token, "__print_flags") == 0) { + free_token(token); + is_flag_field = 1; + return process_flags(event, arg, tok); + } + if (strcmp(token, "__print_symbolic") == 0) { + free_token(token); + is_symbolic_field = 1; + return process_symbols(event, arg, tok); + } + if (strcmp(token, "__get_str") == 0) { + free_token(token); + return process_str(event, arg, tok); + } + if (strcmp(token, "__get_dynamic_array") == 0) { + free_token(token); + return process_dynamic_array(event, arg, tok); + } + + func = find_func_handler(event->pevent, token); + if (func) { + free_token(token); + return process_func_handler(event, func, arg, tok); + } + + do_warning("function %s not defined", token); + free_token(token); + return EVENT_ERROR; +} + +static enum event_type +process_arg_token(struct event_format *event, struct print_arg *arg, + char **tok, enum event_type type) +{ + char *token; + char *atom; + + token = *tok; + + switch (type) { + case EVENT_ITEM: + if (strcmp(token, "REC") == 0) { + free_token(token); + type = process_entry(event, arg, &token); + break; + } + atom = token; + /* test the next token */ + type = read_token_item(&token); + + /* + * If the next token is a parenthesis, then this + * is a function. + */ + if (type == EVENT_DELIM && strcmp(token, "(") == 0) { + free_token(token); + token = NULL; + /* this will free atom. */ + type = process_function(event, arg, atom, &token); + break; + } + /* atoms can be more than one token long */ + while (type == EVENT_ITEM) { + atom = realloc(atom, strlen(atom) + strlen(token) + 2); + strcat(atom, " "); + strcat(atom, token); + free_token(token); + type = read_token_item(&token); + } + + arg->type = PRINT_ATOM; + arg->atom.atom = atom; + break; + + case EVENT_DQUOTE: + case EVENT_SQUOTE: + arg->type = PRINT_ATOM; + arg->atom.atom = token; + type = read_token_item(&token); + break; + case EVENT_DELIM: + if (strcmp(token, "(") == 0) { + free_token(token); + type = process_paren(event, arg, &token); + break; + } + case EVENT_OP: + /* handle single ops */ + arg->type = PRINT_OP; + arg->op.op = token; + arg->op.left = NULL; + type = process_op(event, arg, &token); + + /* On error, the op is freed */ + if (type == EVENT_ERROR) + arg->op.op = NULL; + + /* return error type if errored */ + break; + + case EVENT_ERROR ... EVENT_NEWLINE: + default: + die("unexpected type %d", type); + } + *tok = token; + + return type; +} + +static int event_read_print_args(struct event_format *event, struct print_arg **list) +{ + enum event_type type = EVENT_ERROR; + struct print_arg *arg; + char *token; + int args = 0; + + do { + if (type == EVENT_NEWLINE) { + type = read_token_item(&token); + continue; + } + + arg = alloc_arg(); + + type = process_arg(event, arg, &token); + + if (type == EVENT_ERROR) { + free_token(token); + free_arg(arg); + return -1; + } + + *list = arg; + args++; + + if (type == EVENT_OP) { + type = process_op(event, arg, &token); + free_token(token); + if (type == EVENT_ERROR) { + *list = NULL; + free_arg(arg); + return -1; + } + list = &arg->next; + continue; + } + + if (type == EVENT_DELIM && strcmp(token, ",") == 0) { + free_token(token); + *list = arg; + list = &arg->next; + continue; + } + break; + } while (type != EVENT_NONE); + + if (type != EVENT_NONE && type != EVENT_ERROR) + free_token(token); + + return args; +} + +static int event_read_print(struct event_format *event) +{ + enum event_type type; + char *token; + int ret; + + if (read_expected_item(EVENT_ITEM, "print") < 0) + return -1; + + if (read_expected(EVENT_ITEM, "fmt") < 0) + return -1; + + if (read_expected(EVENT_OP, ":") < 0) + return -1; + + if (read_expect_type(EVENT_DQUOTE, &token) < 0) + goto fail; + + concat: + event->print_fmt.format = token; + event->print_fmt.args = NULL; + + /* ok to have no arg */ + type = read_token_item(&token); + + if (type == EVENT_NONE) + return 0; + + /* Handle concatenation of print lines */ + if (type == EVENT_DQUOTE) { + char *cat; + + cat = malloc_or_die(strlen(event->print_fmt.format) + + strlen(token) + 1); + strcpy(cat, event->print_fmt.format); + strcat(cat, token); + free_token(token); + free_token(event->print_fmt.format); + event->print_fmt.format = NULL; + token = cat; + goto concat; + } + + if (test_type_token(type, token, EVENT_DELIM, ",")) + goto fail; + + free_token(token); + + ret = event_read_print_args(event, &event->print_fmt.args); + if (ret < 0) + return -1; + + return ret; + + fail: + free_token(token); + return -1; +} + +/** + * pevent_find_common_field - return a common field by event + * @event: handle for the event + * @name: the name of the common field to return + * + * Returns a common field from the event by the given @name. + * This only searchs the common fields and not all field. + */ +struct format_field * +pevent_find_common_field(struct event_format *event, const char *name) +{ + struct format_field *format; + + for (format = event->format.common_fields; + format; format = format->next) { + if (strcmp(format->name, name) == 0) + break; + } + + return format; +} + +/** + * pevent_find_field - find a non-common field + * @event: handle for the event + * @name: the name of the non-common field + * + * Returns a non-common field by the given @name. + * This does not search common fields. + */ +struct format_field * +pevent_find_field(struct event_format *event, const char *name) +{ + struct format_field *format; + + for (format = event->format.fields; + format; format = format->next) { + if (strcmp(format->name, name) == 0) + break; + } + + return format; +} + +/** + * pevent_find_any_field - find any field by name + * @event: handle for the event + * @name: the name of the field + * + * Returns a field by the given @name. + * This searchs the common field names first, then + * the non-common ones if a common one was not found. + */ +struct format_field * +pevent_find_any_field(struct event_format *event, const char *name) +{ + struct format_field *format; + + format = pevent_find_common_field(event, name); + if (format) + return format; + return pevent_find_field(event, name); +} + +/** + * pevent_read_number - read a number from data + * @pevent: handle for the pevent + * @ptr: the raw data + * @size: the size of the data that holds the number + * + * Returns the number (converted to host) from the + * raw data. + */ +unsigned long long pevent_read_number(struct pevent *pevent, + const void *ptr, int size) +{ + switch (size) { + case 1: + return *(unsigned char *)ptr; + case 2: + return data2host2(pevent, ptr); + case 4: + return data2host4(pevent, ptr); + case 8: + return data2host8(pevent, ptr); + default: + /* BUG! */ + return 0; + } +} + +/** + * pevent_read_number_field - read a number from data + * @field: a handle to the field + * @data: the raw data to read + * @value: the value to place the number in + * + * Reads raw data according to a field offset and size, + * and translates it into @value. + * + * Returns 0 on success, -1 otherwise. + */ +int pevent_read_number_field(struct format_field *field, const void *data, + unsigned long long *value) +{ + if (!field) + return -1; + switch (field->size) { + case 1: + case 2: + case 4: + case 8: + *value = pevent_read_number(field->event->pevent, + data + field->offset, field->size); + return 0; + default: + return -1; + } +} + +static int get_common_info(struct pevent *pevent, + const char *type, int *offset, int *size) +{ + struct event_format *event; + struct format_field *field; + + /* + * All events should have the same common elements. + * Pick any event to find where the type is; + */ + if (!pevent->events) + die("no event_list!"); + + event = pevent->events[0]; + field = pevent_find_common_field(event, type); + if (!field) + die("field '%s' not found", type); + + *offset = field->offset; + *size = field->size; + + return 0; +} + +static int __parse_common(struct pevent *pevent, void *data, + int *size, int *offset, const char *name) +{ + int ret; + + if (!*size) { + ret = get_common_info(pevent, name, offset, size); + if (ret < 0) + return ret; + } + return pevent_read_number(pevent, data + *offset, *size); +} + +static int trace_parse_common_type(struct pevent *pevent, void *data) +{ + return __parse_common(pevent, data, + &pevent->type_size, &pevent->type_offset, + "common_type"); +} + +static int parse_common_pid(struct pevent *pevent, void *data) +{ + return __parse_common(pevent, data, + &pevent->pid_size, &pevent->pid_offset, + "common_pid"); +} + +static int parse_common_pc(struct pevent *pevent, void *data) +{ + return __parse_common(pevent, data, + &pevent->pc_size, &pevent->pc_offset, + "common_preempt_count"); +} + +static int parse_common_flags(struct pevent *pevent, void *data) +{ + return __parse_common(pevent, data, + &pevent->flags_size, &pevent->flags_offset, + "common_flags"); +} + +static int parse_common_lock_depth(struct pevent *pevent, void *data) +{ + int ret; + + ret = __parse_common(pevent, data, + &pevent->ld_size, &pevent->ld_offset, + "common_lock_depth"); + if (ret < 0) + return -1; + + return ret; +} + +static int events_id_cmp(const void *a, const void *b); + +/** + * pevent_find_event - find an event by given id + * @pevent: a handle to the pevent + * @id: the id of the event + * + * Returns an event that has a given @id. + */ +struct event_format *pevent_find_event(struct pevent *pevent, int id) +{ + struct event_format **eventptr; + struct event_format key; + struct event_format *pkey = &key; + + /* Check cache first */ + if (pevent->last_event && pevent->last_event->id == id) + return pevent->last_event; + + key.id = id; + + eventptr = bsearch(&pkey, pevent->events, pevent->nr_events, + sizeof(*pevent->events), events_id_cmp); + + if (eventptr) { + pevent->last_event = *eventptr; + return *eventptr; + } + + return NULL; +} + +/** + * pevent_find_event_by_name - find an event by given name + * @pevent: a handle to the pevent + * @sys: the system name to search for + * @name: the name of the event to search for + * + * This returns an event with a given @name and under the system + * @sys. If @sys is NULL the first event with @name is returned. + */ +struct event_format * +pevent_find_event_by_name(struct pevent *pevent, + const char *sys, const char *name) +{ + struct event_format *event; + int i; + + if (pevent->last_event && + strcmp(pevent->last_event->name, name) == 0 && + (!sys || strcmp(pevent->last_event->system, sys) == 0)) + return pevent->last_event; + + for (i = 0; i < pevent->nr_events; i++) { + event = pevent->events[i]; + if (strcmp(event->name, name) == 0) { + if (!sys) + break; + if (strcmp(event->system, sys) == 0) + break; + } + } + if (i == pevent->nr_events) + event = NULL; + + pevent->last_event = event; + return event; +} + +static unsigned long long +eval_num_arg(void *data, int size, struct event_format *event, struct print_arg *arg) +{ + struct pevent *pevent = event->pevent; + unsigned long long val = 0; + unsigned long long left, right; + struct print_arg *typearg = NULL; + struct print_arg *larg; + unsigned long offset; + unsigned int field_size; + + switch (arg->type) { + case PRINT_NULL: + /* ?? */ + return 0; + case PRINT_ATOM: + return strtoull(arg->atom.atom, NULL, 0); + case PRINT_FIELD: + if (!arg->field.field) { + arg->field.field = pevent_find_any_field(event, arg->field.name); + if (!arg->field.field) + die("field %s not found", arg->field.name); + } + /* must be a number */ + val = pevent_read_number(pevent, data + arg->field.field->offset, + arg->field.field->size); + break; + case PRINT_FLAGS: + case PRINT_SYMBOL: + break; + case PRINT_TYPE: + val = eval_num_arg(data, size, event, arg->typecast.item); + return eval_type(val, arg, 0); + case PRINT_STRING: + case PRINT_BSTRING: + return 0; + case PRINT_FUNC: { + struct trace_seq s; + trace_seq_init(&s); + val = process_defined_func(&s, data, size, event, arg); + trace_seq_destroy(&s); + return val; + } + case PRINT_OP: + if (strcmp(arg->op.op, "[") == 0) { + /* + * Arrays are special, since we don't want + * to read the arg as is. + */ + right = eval_num_arg(data, size, event, arg->op.right); + + /* handle typecasts */ + larg = arg->op.left; + while (larg->type == PRINT_TYPE) { + if (!typearg) + typearg = larg; + larg = larg->typecast.item; + } + + /* Default to long size */ + field_size = pevent->long_size; + + switch (larg->type) { + case PRINT_DYNAMIC_ARRAY: + offset = pevent_read_number(pevent, + data + larg->dynarray.field->offset, + larg->dynarray.field->size); + if (larg->dynarray.field->elementsize) + field_size = larg->dynarray.field->elementsize; + /* + * The actual length of the dynamic array is stored + * in the top half of the field, and the offset + * is in the bottom half of the 32 bit field. + */ + offset &= 0xffff; + offset += right; + break; + case PRINT_FIELD: + if (!larg->field.field) { + larg->field.field = + pevent_find_any_field(event, larg->field.name); + if (!larg->field.field) + die("field %s not found", larg->field.name); + } + field_size = larg->field.field->elementsize; + offset = larg->field.field->offset + + right * larg->field.field->elementsize; + break; + default: + goto default_op; /* oops, all bets off */ + } + val = pevent_read_number(pevent, + data + offset, field_size); + if (typearg) + val = eval_type(val, typearg, 1); + break; + } else if (strcmp(arg->op.op, "?") == 0) { + left = eval_num_arg(data, size, event, arg->op.left); + arg = arg->op.right; + if (left) + val = eval_num_arg(data, size, event, arg->op.left); + else + val = eval_num_arg(data, size, event, arg->op.right); + break; + } + default_op: + left = eval_num_arg(data, size, event, arg->op.left); + right = eval_num_arg(data, size, event, arg->op.right); + switch (arg->op.op[0]) { + case '!': + switch (arg->op.op[1]) { + case 0: + val = !right; + break; + case '=': + val = left != right; + break; + default: + die("unknown op '%s'", arg->op.op); + } + break; + case '~': + val = ~right; + break; + case '|': + if (arg->op.op[1]) + val = left || right; + else + val = left | right; + break; + case '&': + if (arg->op.op[1]) + val = left && right; + else + val = left & right; + break; + case '<': + switch (arg->op.op[1]) { + case 0: + val = left < right; + break; + case '<': + val = left << right; + break; + case '=': + val = left <= right; + break; + default: + die("unknown op '%s'", arg->op.op); + } + break; + case '>': + switch (arg->op.op[1]) { + case 0: + val = left > right; + break; + case '>': + val = left >> right; + break; + case '=': + val = left >= right; + break; + default: + die("unknown op '%s'", arg->op.op); + } + break; + case '=': + if (arg->op.op[1] != '=') + die("unknown op '%s'", arg->op.op); + val = left == right; + break; + case '-': + val = left - right; + break; + case '+': + val = left + right; + break; + case '/': + val = left / right; + break; + case '*': + val = left * right; + break; + default: + die("unknown op '%s'", arg->op.op); + } + break; + default: /* not sure what to do there */ + return 0; + } + return val; +} + +struct flag { + const char *name; + unsigned long long value; +}; + +static const struct flag flags[] = { + { "HI_SOFTIRQ", 0 }, + { "TIMER_SOFTIRQ", 1 }, + { "NET_TX_SOFTIRQ", 2 }, + { "NET_RX_SOFTIRQ", 3 }, + { "BLOCK_SOFTIRQ", 4 }, + { "BLOCK_IOPOLL_SOFTIRQ", 5 }, + { "TASKLET_SOFTIRQ", 6 }, + { "SCHED_SOFTIRQ", 7 }, + { "HRTIMER_SOFTIRQ", 8 }, + { "RCU_SOFTIRQ", 9 }, + + { "HRTIMER_NORESTART", 0 }, + { "HRTIMER_RESTART", 1 }, +}; + +static unsigned long long eval_flag(const char *flag) +{ + int i; + + /* + * Some flags in the format files do not get converted. + * If the flag is not numeric, see if it is something that + * we already know about. + */ + if (isdigit(flag[0])) + return strtoull(flag, NULL, 0); + + for (i = 0; i < (int)(sizeof(flags)/sizeof(flags[0])); i++) + if (strcmp(flags[i].name, flag) == 0) + return flags[i].value; + + return 0; +} + +static void print_str_to_seq(struct trace_seq *s, const char *format, + int len_arg, const char *str) +{ + if (len_arg >= 0) + trace_seq_printf(s, format, len_arg, str); + else + trace_seq_printf(s, format, str); +} + +static void print_str_arg(struct trace_seq *s, void *data, int size, + struct event_format *event, const char *format, + int len_arg, struct print_arg *arg) +{ + struct pevent *pevent = event->pevent; + struct print_flag_sym *flag; + unsigned long long val, fval; + unsigned long addr; + char *str; + int print; + int len; + + switch (arg->type) { + case PRINT_NULL: + /* ?? */ + return; + case PRINT_ATOM: + print_str_to_seq(s, format, len_arg, arg->atom.atom); + return; + case PRINT_FIELD: + if (!arg->field.field) { + arg->field.field = pevent_find_any_field(event, arg->field.name); + if (!arg->field.field) + die("field %s not found", arg->field.name); + } + /* Zero sized fields, mean the rest of the data */ + len = arg->field.field->size ? : size - arg->field.field->offset; + + /* + * Some events pass in pointers. If this is not an array + * and the size is the same as long_size, assume that it + * is a pointer. + */ + if (!(arg->field.field->flags & FIELD_IS_ARRAY) && + arg->field.field->size == pevent->long_size) { + addr = *(unsigned long *)(data + arg->field.field->offset); + trace_seq_printf(s, "%lx", addr); + break; + } + str = malloc_or_die(len + 1); + memcpy(str, data + arg->field.field->offset, len); + str[len] = 0; + print_str_to_seq(s, format, len_arg, str); + free(str); + break; + case PRINT_FLAGS: + val = eval_num_arg(data, size, event, arg->flags.field); + print = 0; + for (flag = arg->flags.flags; flag; flag = flag->next) { + fval = eval_flag(flag->value); + if (!val && !fval) { + print_str_to_seq(s, format, len_arg, flag->str); + break; + } + if (fval && (val & fval) == fval) { + if (print && arg->flags.delim) + trace_seq_puts(s, arg->flags.delim); + print_str_to_seq(s, format, len_arg, flag->str); + print = 1; + val &= ~fval; + } + } + break; + case PRINT_SYMBOL: + val = eval_num_arg(data, size, event, arg->symbol.field); + for (flag = arg->symbol.symbols; flag; flag = flag->next) { + fval = eval_flag(flag->value); + if (val == fval) { + print_str_to_seq(s, format, len_arg, flag->str); + break; + } + } + break; + + case PRINT_TYPE: + break; + case PRINT_STRING: { + int str_offset; + + if (arg->string.offset == -1) { + struct format_field *f; + + f = pevent_find_any_field(event, arg->string.string); + arg->string.offset = f->offset; + } + str_offset = data2host4(pevent, data + arg->string.offset); + str_offset &= 0xffff; + print_str_to_seq(s, format, len_arg, ((char *)data) + str_offset); + break; + } + case PRINT_BSTRING: + trace_seq_printf(s, format, arg->string.string); + break; + case PRINT_OP: + /* + * The only op for string should be ? : + */ + if (arg->op.op[0] != '?') + return; + val = eval_num_arg(data, size, event, arg->op.left); + if (val) + print_str_arg(s, data, size, event, + format, len_arg, arg->op.right->op.left); + else + print_str_arg(s, data, size, event, + format, len_arg, arg->op.right->op.right); + break; + case PRINT_FUNC: + process_defined_func(s, data, size, event, arg); + break; + default: + /* well... */ + break; + } +} + +static unsigned long long +process_defined_func(struct trace_seq *s, void *data, int size, + struct event_format *event, struct print_arg *arg) +{ + struct pevent_function_handler *func_handle = arg->func.func; + struct pevent_func_params *param; + unsigned long long *args; + unsigned long long ret; + struct print_arg *farg; + struct trace_seq str; + struct save_str { + struct save_str *next; + char *str; + } *strings = NULL, *string; + int i; + + if (!func_handle->nr_args) { + ret = (*func_handle->func)(s, NULL); + goto out; + } + + farg = arg->func.args; + param = func_handle->params; + + args = malloc_or_die(sizeof(*args) * func_handle->nr_args); + for (i = 0; i < func_handle->nr_args; i++) { + switch (param->type) { + case PEVENT_FUNC_ARG_INT: + case PEVENT_FUNC_ARG_LONG: + case PEVENT_FUNC_ARG_PTR: + args[i] = eval_num_arg(data, size, event, farg); + break; + case PEVENT_FUNC_ARG_STRING: + trace_seq_init(&str); + print_str_arg(&str, data, size, event, "%s", -1, farg); + trace_seq_terminate(&str); + string = malloc_or_die(sizeof(*string)); + string->next = strings; + string->str = strdup(str.buffer); + strings = string; + trace_seq_destroy(&str); + break; + default: + /* + * Something went totally wrong, this is not + * an input error, something in this code broke. + */ + die("Unexpected end of arguments\n"); + break; + } + farg = farg->next; + param = param->next; + } + + ret = (*func_handle->func)(s, args); + free(args); + while (strings) { + string = strings; + strings = string->next; + free(string->str); + free(string); + } + + out: + /* TBD : handle return type here */ + return ret; +} + +static struct print_arg *make_bprint_args(char *fmt, void *data, int size, struct event_format *event) +{ + struct pevent *pevent = event->pevent; + struct format_field *field, *ip_field; + struct print_arg *args, *arg, **next; + unsigned long long ip, val; + char *ptr; + void *bptr; + + field = pevent->bprint_buf_field; + ip_field = pevent->bprint_ip_field; + + if (!field) { + field = pevent_find_field(event, "buf"); + if (!field) + die("can't find buffer field for binary printk"); + ip_field = pevent_find_field(event, "ip"); + if (!ip_field) + die("can't find ip field for binary printk"); + pevent->bprint_buf_field = field; + pevent->bprint_ip_field = ip_field; + } + + ip = pevent_read_number(pevent, data + ip_field->offset, ip_field->size); + + /* + * The first arg is the IP pointer. + */ + args = alloc_arg(); + arg = args; + arg->next = NULL; + next = &arg->next; + + arg->type = PRINT_ATOM; + arg->atom.atom = malloc_or_die(32); + sprintf(arg->atom.atom, "%lld", ip); + + /* skip the first "%pf : " */ + for (ptr = fmt + 6, bptr = data + field->offset; + bptr < data + size && *ptr; ptr++) { + int ls = 0; + + if (*ptr == '%') { + process_again: + ptr++; + switch (*ptr) { + case '%': + break; + case 'l': + ls++; + goto process_again; + case 'L': + ls = 2; + goto process_again; + case '0' ... '9': + goto process_again; + case 'p': + ls = 1; + /* fall through */ + case 'd': + case 'u': + case 'x': + case 'i': + /* the pointers are always 4 bytes aligned */ + bptr = (void *)(((unsigned long)bptr + 3) & + ~3); + switch (ls) { + case 0: + ls = 4; + break; + case 1: + ls = pevent->long_size; + break; + case 2: + ls = 8; + default: + break; + } + val = pevent_read_number(pevent, bptr, ls); + bptr += ls; + arg = alloc_arg(); + arg->next = NULL; + arg->type = PRINT_ATOM; + arg->atom.atom = malloc_or_die(32); + sprintf(arg->atom.atom, "%lld", val); + *next = arg; + next = &arg->next; + break; + case 's': + arg = alloc_arg(); + arg->next = NULL; + arg->type = PRINT_BSTRING; + arg->string.string = strdup(bptr); + bptr += strlen(bptr) + 1; + *next = arg; + next = &arg->next; + default: + break; + } + } + } + + return args; +} + +static void free_args(struct print_arg *args) +{ + struct print_arg *next; + + while (args) { + next = args->next; + + free_arg(args); + args = next; + } +} + +static char * +get_bprint_format(void *data, int size __unused, struct event_format *event) +{ + struct pevent *pevent = event->pevent; + unsigned long long addr; + struct format_field *field; + struct printk_map *printk; + char *format; + char *p; + + field = pevent->bprint_fmt_field; + + if (!field) { + field = pevent_find_field(event, "fmt"); + if (!field) + die("can't find format field for binary printk"); + pevent->bprint_fmt_field = field; + } + + addr = pevent_read_number(pevent, data + field->offset, field->size); + + printk = find_printk(pevent, addr); + if (!printk) { + format = malloc_or_die(45); + sprintf(format, "%%pf : (NO FORMAT FOUND at %llx)\n", + addr); + return format; + } + + p = printk->printk; + /* Remove any quotes. */ + if (*p == '"') + p++; + format = malloc_or_die(strlen(p) + 10); + sprintf(format, "%s : %s", "%pf", p); + /* remove ending quotes and new line since we will add one too */ + p = format + strlen(format) - 1; + if (*p == '"') + *p = 0; + + p -= 2; + if (strcmp(p, "\\n") == 0) + *p = 0; + + return format; +} + +static void print_mac_arg(struct trace_seq *s, int mac, void *data, int size, + struct event_format *event, struct print_arg *arg) +{ + unsigned char *buf; + char *fmt = "%.2x:%.2x:%.2x:%.2x:%.2x:%.2x"; + + if (arg->type == PRINT_FUNC) { + process_defined_func(s, data, size, event, arg); + return; + } + + if (arg->type != PRINT_FIELD) { + trace_seq_printf(s, "ARG TYPE NOT FIELD BUT %d", + arg->type); + return; + } + + if (mac == 'm') + fmt = "%.2x%.2x%.2x%.2x%.2x%.2x"; + if (!arg->field.field) { + arg->field.field = + pevent_find_any_field(event, arg->field.name); + if (!arg->field.field) + die("field %s not found", arg->field.name); + } + if (arg->field.field->size != 6) { + trace_seq_printf(s, "INVALIDMAC"); + return; + } + buf = data + arg->field.field->offset; + trace_seq_printf(s, fmt, buf[0], buf[1], buf[2], buf[3], buf[4], buf[5]); +} + +static void print_event_fields(struct trace_seq *s, void *data, int size, + struct event_format *event) +{ + struct format_field *field; + unsigned long long val; + unsigned int offset, len, i; + + field = event->format.fields; + while (field) { + trace_seq_printf(s, " %s=", field->name); + if (field->flags & FIELD_IS_ARRAY) { + offset = field->offset; + len = field->size; + if (field->flags & FIELD_IS_DYNAMIC) { + val = pevent_read_number(event->pevent, data + offset, len); + offset = val; + len = offset >> 16; + offset &= 0xffff; + } + if (field->flags & FIELD_IS_STRING) { + trace_seq_printf(s, "%s", (char *)data + offset); + } else { + trace_seq_puts(s, "ARRAY["); + for (i = 0; i < len; i++) { + if (i) + trace_seq_puts(s, ", "); + trace_seq_printf(s, "%02x", + *((unsigned char *)data + offset + i)); + } + trace_seq_putc(s, ']'); + } + } else { + val = pevent_read_number(event->pevent, data + field->offset, + field->size); + if (field->flags & FIELD_IS_POINTER) { + trace_seq_printf(s, "0x%llx", val); + } else if (field->flags & FIELD_IS_SIGNED) { + switch (field->size) { + case 4: + /* + * If field is long then print it in hex. + * A long usually stores pointers. + */ + if (field->flags & FIELD_IS_LONG) + trace_seq_printf(s, "0x%x", (int)val); + else + trace_seq_printf(s, "%d", (int)val); + break; + case 2: + trace_seq_printf(s, "%2d", (short)val); + break; + case 1: + trace_seq_printf(s, "%1d", (char)val); + break; + default: + trace_seq_printf(s, "%lld", val); + } + } else { + if (field->flags & FIELD_IS_LONG) + trace_seq_printf(s, "0x%llx", val); + else + trace_seq_printf(s, "%llu", val); + } + } + field = field->next; + } +} + +static void pretty_print(struct trace_seq *s, void *data, int size, struct event_format *event) +{ + struct pevent *pevent = event->pevent; + struct print_fmt *print_fmt = &event->print_fmt; + struct print_arg *arg = print_fmt->args; + struct print_arg *args = NULL; + const char *ptr = print_fmt->format; + unsigned long long val; + struct func_map *func; + const char *saveptr; + char *bprint_fmt = NULL; + char format[32]; + int show_func; + int len_as_arg; + int len_arg; + int len; + int ls; + + if (event->flags & EVENT_FL_FAILED) { + trace_seq_printf(s, "[FAILED TO PARSE]"); + print_event_fields(s, data, size, event); + return; + } + + if (event->flags & EVENT_FL_ISBPRINT) { + bprint_fmt = get_bprint_format(data, size, event); + args = make_bprint_args(bprint_fmt, data, size, event); + arg = args; + ptr = bprint_fmt; + } + + for (; *ptr; ptr++) { + ls = 0; + if (*ptr == '\\') { + ptr++; + switch (*ptr) { + case 'n': + trace_seq_putc(s, '\n'); + break; + case 't': + trace_seq_putc(s, '\t'); + break; + case 'r': + trace_seq_putc(s, '\r'); + break; + case '\\': + trace_seq_putc(s, '\\'); + break; + default: + trace_seq_putc(s, *ptr); + break; + } + + } else if (*ptr == '%') { + saveptr = ptr; + show_func = 0; + len_as_arg = 0; + cont_process: + ptr++; + switch (*ptr) { + case '%': + trace_seq_putc(s, '%'); + break; + case '#': + /* FIXME: need to handle properly */ + goto cont_process; + case 'h': + ls--; + goto cont_process; + case 'l': + ls++; + goto cont_process; + case 'L': + ls = 2; + goto cont_process; + case '*': + /* The argument is the length. */ + if (!arg) + die("no argument match"); + len_arg = eval_num_arg(data, size, event, arg); + len_as_arg = 1; + arg = arg->next; + goto cont_process; + case '.': + case 'z': + case 'Z': + case '0' ... '9': + goto cont_process; + case 'p': + if (pevent->long_size == 4) + ls = 1; + else + ls = 2; + + if (*(ptr+1) == 'F' || + *(ptr+1) == 'f') { + ptr++; + show_func = *ptr; + } else if (*(ptr+1) == 'M' || *(ptr+1) == 'm') { + print_mac_arg(s, *(ptr+1), data, size, event, arg); + ptr++; + break; + } + + /* fall through */ + case 'd': + case 'i': + case 'x': + case 'X': + case 'u': + if (!arg) + die("no argument match"); + + len = ((unsigned long)ptr + 1) - + (unsigned long)saveptr; + + /* should never happen */ + if (len > 31) + die("bad format!"); + + memcpy(format, saveptr, len); + format[len] = 0; + + val = eval_num_arg(data, size, event, arg); + arg = arg->next; + + if (show_func) { + func = find_func(pevent, val); + if (func) { + trace_seq_puts(s, func->func); + if (show_func == 'F') + trace_seq_printf(s, + "+0x%llx", + val - func->addr); + break; + } + } + if (pevent->long_size == 8 && ls) { + char *p; + + ls = 2; + /* make %l into %ll */ + p = strchr(format, 'l'); + if (p) + memmove(p, p+1, strlen(p)+1); + else if (strcmp(format, "%p") == 0) + strcpy(format, "0x%llx"); + } + switch (ls) { + case -2: + if (len_as_arg) + trace_seq_printf(s, format, len_arg, (char)val); + else + trace_seq_printf(s, format, (char)val); + break; + case -1: + if (len_as_arg) + trace_seq_printf(s, format, len_arg, (short)val); + else + trace_seq_printf(s, format, (short)val); + break; + case 0: + if (len_as_arg) + trace_seq_printf(s, format, len_arg, (int)val); + else + trace_seq_printf(s, format, (int)val); + break; + case 1: + if (len_as_arg) + trace_seq_printf(s, format, len_arg, (long)val); + else + trace_seq_printf(s, format, (long)val); + break; + case 2: + if (len_as_arg) + trace_seq_printf(s, format, len_arg, + (long long)val); + else + trace_seq_printf(s, format, (long long)val); + break; + default: + die("bad count (%d)", ls); + } + break; + case 's': + if (!arg) + die("no matching argument"); + + len = ((unsigned long)ptr + 1) - + (unsigned long)saveptr; + + /* should never happen */ + if (len > 31) + die("bad format!"); + + memcpy(format, saveptr, len); + format[len] = 0; + if (!len_as_arg) + len_arg = -1; + print_str_arg(s, data, size, event, + format, len_arg, arg); + arg = arg->next; + break; + default: + trace_seq_printf(s, ">%c<", *ptr); + + } + } else + trace_seq_putc(s, *ptr); + } + + if (args) { + free_args(args); + free(bprint_fmt); + } +} + +/** + * pevent_data_lat_fmt - parse the data for the latency format + * @pevent: a handle to the pevent + * @s: the trace_seq to write to + * @data: the raw data to read from + * @size: currently unused. + * + * This parses out the Latency format (interrupts disabled, + * need rescheduling, in hard/soft interrupt, preempt count + * and lock depth) and places it into the trace_seq. + */ +void pevent_data_lat_fmt(struct pevent *pevent, + struct trace_seq *s, struct pevent_record *record) +{ + static int check_lock_depth = 1; + static int lock_depth_exists; + unsigned int lat_flags; + unsigned int pc; + int lock_depth; + int hardirq; + int softirq; + void *data = record->data; + + lat_flags = parse_common_flags(pevent, data); + pc = parse_common_pc(pevent, data); + /* lock_depth may not always exist */ + if (check_lock_depth) { + struct format_field *field; + struct event_format *event; + + check_lock_depth = 0; + event = pevent->events[0]; + field = pevent_find_common_field(event, "common_lock_depth"); + if (field) + lock_depth_exists = 1; + } + if (lock_depth_exists) + lock_depth = parse_common_lock_depth(pevent, data); + + hardirq = lat_flags & TRACE_FLAG_HARDIRQ; + softirq = lat_flags & TRACE_FLAG_SOFTIRQ; + + trace_seq_printf(s, "%c%c%c", + (lat_flags & TRACE_FLAG_IRQS_OFF) ? 'd' : + (lat_flags & TRACE_FLAG_IRQS_NOSUPPORT) ? + 'X' : '.', + (lat_flags & TRACE_FLAG_NEED_RESCHED) ? + 'N' : '.', + (hardirq && softirq) ? 'H' : + hardirq ? 'h' : softirq ? 's' : '.'); + + if (pc) + trace_seq_printf(s, "%x", pc); + else + trace_seq_putc(s, '.'); + + if (lock_depth_exists) { + if (lock_depth < 0) + trace_seq_putc(s, '.'); + else + trace_seq_printf(s, "%d", lock_depth); + } + + trace_seq_terminate(s); +} + +/** + * pevent_data_type - parse out the given event type + * @pevent: a handle to the pevent + * @rec: the record to read from + * + * This returns the event id from the @rec. + */ +int pevent_data_type(struct pevent *pevent, struct pevent_record *rec) +{ + return trace_parse_common_type(pevent, rec->data); +} + +/** + * pevent_data_event_from_type - find the event by a given type + * @pevent: a handle to the pevent + * @type: the type of the event. + * + * This returns the event form a given @type; + */ +struct event_format *pevent_data_event_from_type(struct pevent *pevent, int type) +{ + return pevent_find_event(pevent, type); +} + +/** + * pevent_data_pid - parse the PID from raw data + * @pevent: a handle to the pevent + * @rec: the record to parse + * + * This returns the PID from a raw data. + */ +int pevent_data_pid(struct pevent *pevent, struct pevent_record *rec) +{ + return parse_common_pid(pevent, rec->data); +} + +/** + * pevent_data_comm_from_pid - return the command line from PID + * @pevent: a handle to the pevent + * @pid: the PID of the task to search for + * + * This returns a pointer to the command line that has the given + * @pid. + */ +const char *pevent_data_comm_from_pid(struct pevent *pevent, int pid) +{ + const char *comm; + + comm = find_cmdline(pevent, pid); + return comm; +} + +/** + * pevent_data_comm_from_pid - parse the data into the print format + * @s: the trace_seq to write to + * @event: the handle to the event + * @cpu: the cpu the event was recorded on + * @data: the raw data + * @size: the size of the raw data + * @nsecs: the timestamp of the event + * + * This parses the raw @data using the given @event information and + * writes the print format into the trace_seq. + */ +void pevent_event_info(struct trace_seq *s, struct event_format *event, + struct pevent_record *record) +{ + int print_pretty = 1; + + if (event->pevent->print_raw) + print_event_fields(s, record->data, record->size, event); + else { + + if (event->handler) + print_pretty = event->handler(s, record, event, + event->context); + + if (print_pretty) + pretty_print(s, record->data, record->size, event); + } + + trace_seq_terminate(s); +} + +void pevent_print_event(struct pevent *pevent, struct trace_seq *s, + struct pevent_record *record) +{ + static char *spaces = " "; /* 20 spaces */ + struct event_format *event; + unsigned long secs; + unsigned long usecs; + unsigned long nsecs; + const char *comm; + void *data = record->data; + int type; + int pid; + int len; + int p; + + secs = record->ts / NSECS_PER_SEC; + nsecs = record->ts - secs * NSECS_PER_SEC; + + if (record->size < 0) { + do_warning("ug! negative record size %d", record->size); + return; + } + + type = trace_parse_common_type(pevent, data); + + event = pevent_find_event(pevent, type); + if (!event) { + do_warning("ug! no event found for type %d", type); + return; + } + + pid = parse_common_pid(pevent, data); + comm = find_cmdline(pevent, pid); + + if (pevent->latency_format) { + trace_seq_printf(s, "%8.8s-%-5d %3d", + comm, pid, record->cpu); + pevent_data_lat_fmt(pevent, s, record); + } else + trace_seq_printf(s, "%16s-%-5d [%03d]", comm, pid, record->cpu); + + if (pevent->flags & PEVENT_NSEC_OUTPUT) { + usecs = nsecs; + p = 9; + } else { + usecs = (nsecs + 500) / NSECS_PER_USEC; + p = 6; + } + + trace_seq_printf(s, " %5lu.%0*lu: %s: ", secs, p, usecs, event->name); + + /* Space out the event names evenly. */ + len = strlen(event->name); + if (len < 20) + trace_seq_printf(s, "%.*s", 20 - len, spaces); + + pevent_event_info(s, event, record); +} + +static int events_id_cmp(const void *a, const void *b) +{ + struct event_format * const * ea = a; + struct event_format * const * eb = b; + + if ((*ea)->id < (*eb)->id) + return -1; + + if ((*ea)->id > (*eb)->id) + return 1; + + return 0; +} + +static int events_name_cmp(const void *a, const void *b) +{ + struct event_format * const * ea = a; + struct event_format * const * eb = b; + int res; + + res = strcmp((*ea)->name, (*eb)->name); + if (res) + return res; + + res = strcmp((*ea)->system, (*eb)->system); + if (res) + return res; + + return events_id_cmp(a, b); +} + +static int events_system_cmp(const void *a, const void *b) +{ + struct event_format * const * ea = a; + struct event_format * const * eb = b; + int res; + + res = strcmp((*ea)->system, (*eb)->system); + if (res) + return res; + + res = strcmp((*ea)->name, (*eb)->name); + if (res) + return res; + + return events_id_cmp(a, b); +} + +struct event_format **pevent_list_events(struct pevent *pevent, enum event_sort_type sort_type) +{ + struct event_format **events; + int (*sort)(const void *a, const void *b); + + events = pevent->sort_events; + + if (events && pevent->last_type == sort_type) + return events; + + if (!events) { + events = malloc(sizeof(*events) * (pevent->nr_events + 1)); + if (!events) + return NULL; + + memcpy(events, pevent->events, sizeof(*events) * pevent->nr_events); + events[pevent->nr_events] = NULL; + + pevent->sort_events = events; + + /* the internal events are sorted by id */ + if (sort_type == EVENT_SORT_ID) { + pevent->last_type = sort_type; + return events; + } + } + + switch (sort_type) { + case EVENT_SORT_ID: + sort = events_id_cmp; + break; + case EVENT_SORT_NAME: + sort = events_name_cmp; + break; + case EVENT_SORT_SYSTEM: + sort = events_system_cmp; + break; + default: + return events; + } + + qsort(events, pevent->nr_events, sizeof(*events), sort); + pevent->last_type = sort_type; + + return events; +} + +static struct format_field ** +get_event_fields(const char *type, const char *name, + int count, struct format_field *list) +{ + struct format_field **fields; + struct format_field *field; + int i = 0; + + fields = malloc_or_die(sizeof(*fields) * (count + 1)); + for (field = list; field; field = field->next) { + fields[i++] = field; + if (i == count + 1) { + do_warning("event %s has more %s fields than specified", + name, type); + i--; + break; + } + } + + if (i != count) + do_warning("event %s has less %s fields than specified", + name, type); + + fields[i] = NULL; + + return fields; +} + +/** + * pevent_event_common_fields - return a list of common fields for an event + * @event: the event to return the common fields of. + * + * Returns an allocated array of fields. The last item in the array is NULL. + * The array must be freed with free(). + */ +struct format_field **pevent_event_common_fields(struct event_format *event) +{ + return get_event_fields("common", event->name, + event->format.nr_common, + event->format.common_fields); +} + +/** + * pevent_event_fields - return a list of event specific fields for an event + * @event: the event to return the fields of. + * + * Returns an allocated array of fields. The last item in the array is NULL. + * The array must be freed with free(). + */ +struct format_field **pevent_event_fields(struct event_format *event) +{ + return get_event_fields("event", event->name, + event->format.nr_fields, + event->format.fields); +} + +static void print_fields(struct trace_seq *s, struct print_flag_sym *field) +{ + trace_seq_printf(s, "{ %s, %s }", field->value, field->str); + if (field->next) { + trace_seq_puts(s, ", "); + print_fields(s, field->next); + } +} + +/* for debugging */ +static void print_args(struct print_arg *args) +{ + int print_paren = 1; + struct trace_seq s; + + switch (args->type) { + case PRINT_NULL: + printf("null"); + break; + case PRINT_ATOM: + printf("%s", args->atom.atom); + break; + case PRINT_FIELD: + printf("REC->%s", args->field.name); + break; + case PRINT_FLAGS: + printf("__print_flags("); + print_args(args->flags.field); + printf(", %s, ", args->flags.delim); + trace_seq_init(&s); + print_fields(&s, args->flags.flags); + trace_seq_do_printf(&s); + trace_seq_destroy(&s); + printf(")"); + break; + case PRINT_SYMBOL: + printf("__print_symbolic("); + print_args(args->symbol.field); + printf(", "); + trace_seq_init(&s); + print_fields(&s, args->symbol.symbols); + trace_seq_do_printf(&s); + trace_seq_destroy(&s); + printf(")"); + break; + case PRINT_STRING: + case PRINT_BSTRING: + printf("__get_str(%s)", args->string.string); + break; + case PRINT_TYPE: + printf("(%s)", args->typecast.type); + print_args(args->typecast.item); + break; + case PRINT_OP: + if (strcmp(args->op.op, ":") == 0) + print_paren = 0; + if (print_paren) + printf("("); + print_args(args->op.left); + printf(" %s ", args->op.op); + print_args(args->op.right); + if (print_paren) + printf(")"); + break; + default: + /* we should warn... */ + return; + } + if (args->next) { + printf("\n"); + print_args(args->next); + } +} + +static void parse_header_field(const char *field, + int *offset, int *size, int mandatory) +{ + unsigned long long save_input_buf_ptr; + unsigned long long save_input_buf_siz; + char *token; + int type; + + save_input_buf_ptr = input_buf_ptr; + save_input_buf_siz = input_buf_siz; + + if (read_expected(EVENT_ITEM, "field") < 0) + return; + if (read_expected(EVENT_OP, ":") < 0) + return; + + /* type */ + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + free_token(token); + + /* + * If this is not a mandatory field, then test it first. + */ + if (mandatory) { + if (read_expected(EVENT_ITEM, field) < 0) + return; + } else { + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + if (strcmp(token, field) != 0) + goto discard; + free_token(token); + } + + if (read_expected(EVENT_OP, ";") < 0) + return; + if (read_expected(EVENT_ITEM, "offset") < 0) + return; + if (read_expected(EVENT_OP, ":") < 0) + return; + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + *offset = atoi(token); + free_token(token); + if (read_expected(EVENT_OP, ";") < 0) + return; + if (read_expected(EVENT_ITEM, "size") < 0) + return; + if (read_expected(EVENT_OP, ":") < 0) + return; + if (read_expect_type(EVENT_ITEM, &token) < 0) + goto fail; + *size = atoi(token); + free_token(token); + if (read_expected(EVENT_OP, ";") < 0) + return; + type = read_token(&token); + if (type != EVENT_NEWLINE) { + /* newer versions of the kernel have a "signed" type */ + if (type != EVENT_ITEM) + goto fail; + + if (strcmp(token, "signed") != 0) + goto fail; + + free_token(token); + + if (read_expected(EVENT_OP, ":") < 0) + return; + + if (read_expect_type(EVENT_ITEM, &token)) + goto fail; + + free_token(token); + if (read_expected(EVENT_OP, ";") < 0) + return; + + if (read_expect_type(EVENT_NEWLINE, &token)) + goto fail; + } + fail: + free_token(token); + return; + + discard: + input_buf_ptr = save_input_buf_ptr; + input_buf_siz = save_input_buf_siz; + *offset = 0; + *size = 0; + free_token(token); +} + +/** + * pevent_parse_header_page - parse the data stored in the header page + * @pevent: the handle to the pevent + * @buf: the buffer storing the header page format string + * @size: the size of @buf + * @long_size: the long size to use if there is no header + * + * This parses the header page format for information on the + * ring buffer used. The @buf should be copied from + * + * /sys/kernel/debug/tracing/events/header_page + */ +int pevent_parse_header_page(struct pevent *pevent, char *buf, unsigned long size, + int long_size) +{ + int ignore; + + if (!size) { + /* + * Old kernels did not have header page info. + * Sorry but we just use what we find here in user space. + */ + pevent->header_page_ts_size = sizeof(long long); + pevent->header_page_size_size = long_size; + pevent->header_page_data_offset = sizeof(long long) + long_size; + pevent->old_format = 1; + return -1; + } + init_input_buf(buf, size); + + parse_header_field("timestamp", &pevent->header_page_ts_offset, + &pevent->header_page_ts_size, 1); + parse_header_field("commit", &pevent->header_page_size_offset, + &pevent->header_page_size_size, 1); + parse_header_field("overwrite", &pevent->header_page_overwrite, + &ignore, 0); + parse_header_field("data", &pevent->header_page_data_offset, + &pevent->header_page_data_size, 1); + + return 0; +} + +static int event_matches(struct event_format *event, + int id, const char *sys_name, + const char *event_name) +{ + if (id >= 0 && id != event->id) + return 0; + + if (event_name && (strcmp(event_name, event->name) != 0)) + return 0; + + if (sys_name && (strcmp(sys_name, event->system) != 0)) + return 0; + + return 1; +} + +static void free_handler(struct event_handler *handle) +{ + free((void *)handle->sys_name); + free((void *)handle->event_name); + free(handle); +} + +static int find_event_handle(struct pevent *pevent, struct event_format *event) +{ + struct event_handler *handle, **next; + + for (next = &pevent->handlers; *next; + next = &(*next)->next) { + handle = *next; + if (event_matches(event, handle->id, + handle->sys_name, + handle->event_name)) + break; + } + + if (!(*next)) + return 0; + + pr_stat("overriding event (%d) %s:%s with new print handler", + event->id, event->system, event->name); + + event->handler = handle->func; + event->context = handle->context; + + *next = handle->next; + free_handler(handle); + + return 1; +} + +/** + * pevent_parse_event - parse the event format + * @pevent: the handle to the pevent + * @buf: the buffer storing the event format string + * @size: the size of @buf + * @sys: the system the event belongs to + * + * This parses the event format and creates an event structure + * to quickly parse raw data for a given event. + * + * These files currently come from: + * + * /sys/kernel/debug/tracing/events/.../.../format + */ +int pevent_parse_event(struct pevent *pevent, + const char *buf, unsigned long size, + const char *sys) +{ + struct event_format *event; + int ret; + + init_input_buf(buf, size); + + event = alloc_event(); + if (!event) + return -ENOMEM; + + event->name = event_read_name(); + if (!event->name) { + /* Bad event? */ + free(event); + return -1; + } + + if (strcmp(sys, "ftrace") == 0) { + + event->flags |= EVENT_FL_ISFTRACE; + + if (strcmp(event->name, "bprint") == 0) + event->flags |= EVENT_FL_ISBPRINT; + } + + event->id = event_read_id(); + if (event->id < 0) + die("failed to read event id"); + + event->system = strdup(sys); + + /* Add pevent to event so that it can be referenced */ + event->pevent = pevent; + + ret = event_read_format(event); + if (ret < 0) { + do_warning("failed to read event format for %s", event->name); + goto event_failed; + } + + /* + * If the event has an override, don't print warnings if the event + * print format fails to parse. + */ + if (find_event_handle(pevent, event)) + show_warning = 0; + + ret = event_read_print(event); + if (ret < 0) { + do_warning("failed to read event print fmt for %s", + event->name); + show_warning = 1; + goto event_failed; + } + show_warning = 1; + + add_event(pevent, event); + + if (!ret && (event->flags & EVENT_FL_ISFTRACE)) { + struct format_field *field; + struct print_arg *arg, **list; + + /* old ftrace had no args */ + + list = &event->print_fmt.args; + for (field = event->format.fields; field; field = field->next) { + arg = alloc_arg(); + *list = arg; + list = &arg->next; + arg->type = PRINT_FIELD; + arg->field.name = strdup(field->name); + arg->field.field = field; + } + return 0; + } + +#define PRINT_ARGS 0 + if (PRINT_ARGS && event->print_fmt.args) + print_args(event->print_fmt.args); + + return 0; + + event_failed: + event->flags |= EVENT_FL_FAILED; + /* still add it even if it failed */ + add_event(pevent, event); + return -1; +} + +int get_field_val(struct trace_seq *s, struct format_field *field, + const char *name, struct pevent_record *record, + unsigned long long *val, int err) +{ + if (!field) { + if (err) + trace_seq_printf(s, "<CANT FIND FIELD %s>", name); + return -1; + } + + if (pevent_read_number_field(field, record->data, val)) { + if (err) + trace_seq_printf(s, " %s=INVALID", name); + return -1; + } + + return 0; +} + +/** + * pevent_get_field_raw - return the raw pointer into the data field + * @s: The seq to print to on error + * @event: the event that the field is for + * @name: The name of the field + * @record: The record with the field name. + * @len: place to store the field length. + * @err: print default error if failed. + * + * Returns a pointer into record->data of the field and places + * the length of the field in @len. + * + * On failure, it returns NULL. + */ +void *pevent_get_field_raw(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + int *len, int err) +{ + struct format_field *field; + void *data = record->data; + unsigned offset; + int dummy; + + if (!event) + return NULL; + + field = pevent_find_field(event, name); + + if (!field) { + if (err) + trace_seq_printf(s, "<CANT FIND FIELD %s>", name); + return NULL; + } + + /* Allow @len to be NULL */ + if (!len) + len = &dummy; + + offset = field->offset; + if (field->flags & FIELD_IS_DYNAMIC) { + offset = pevent_read_number(event->pevent, + data + offset, field->size); + *len = offset >> 16; + offset &= 0xffff; + } else + *len = field->size; + + return data + offset; +} + +/** + * pevent_get_field_val - find a field and return its value + * @s: The seq to print to on error + * @event: the event that the field is for + * @name: The name of the field + * @record: The record with the field name. + * @val: place to store the value of the field. + * @err: print default error if failed. + * + * Returns 0 on success -1 on field not found. + */ +int pevent_get_field_val(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + unsigned long long *val, int err) +{ + struct format_field *field; + + if (!event) + return -1; + + field = pevent_find_field(event, name); + + return get_field_val(s, field, name, record, val, err); +} + +/** + * pevent_get_common_field_val - find a common field and return its value + * @s: The seq to print to on error + * @event: the event that the field is for + * @name: The name of the field + * @record: The record with the field name. + * @val: place to store the value of the field. + * @err: print default error if failed. + * + * Returns 0 on success -1 on field not found. + */ +int pevent_get_common_field_val(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + unsigned long long *val, int err) +{ + struct format_field *field; + + if (!event) + return -1; + + field = pevent_find_common_field(event, name); + + return get_field_val(s, field, name, record, val, err); +} + +/** + * pevent_get_any_field_val - find a any field and return its value + * @s: The seq to print to on error + * @event: the event that the field is for + * @name: The name of the field + * @record: The record with the field name. + * @val: place to store the value of the field. + * @err: print default error if failed. + * + * Returns 0 on success -1 on field not found. + */ +int pevent_get_any_field_val(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + unsigned long long *val, int err) +{ + struct format_field *field; + + if (!event) + return -1; + + field = pevent_find_any_field(event, name); + + return get_field_val(s, field, name, record, val, err); +} + +/** + * pevent_print_num_field - print a field and a format + * @s: The seq to print to + * @fmt: The printf format to print the field with. + * @event: the event that the field is for + * @name: The name of the field + * @record: The record with the field name. + * @err: print default error if failed. + * + * Returns: 0 on success, -1 field not fould, or 1 if buffer is full. + */ +int pevent_print_num_field(struct trace_seq *s, const char *fmt, + struct event_format *event, const char *name, + struct pevent_record *record, int err) +{ + struct format_field *field = pevent_find_field(event, name); + unsigned long long val; + + if (!field) + goto failed; + + if (pevent_read_number_field(field, record->data, &val)) + goto failed; + + return trace_seq_printf(s, fmt, val); + + failed: + if (err) + trace_seq_printf(s, "CAN'T FIND FIELD \"%s\"", name); + return -1; +} + +static void free_func_handle(struct pevent_function_handler *func) +{ + struct pevent_func_params *params; + + free(func->name); + + while (func->params) { + params = func->params; + func->params = params->next; + free(params); + } + + free(func); +} + +/** + * pevent_register_print_function - register a helper function + * @pevent: the handle to the pevent + * @func: the function to process the helper function + * @name: the name of the helper function + * @parameters: A list of enum pevent_func_arg_type + * + * Some events may have helper functions in the print format arguments. + * This allows a plugin to dynmically create a way to process one + * of these functions. + * + * The @parameters is a variable list of pevent_func_arg_type enums that + * must end with PEVENT_FUNC_ARG_VOID. + */ +int pevent_register_print_function(struct pevent *pevent, + pevent_func_handler func, + enum pevent_func_arg_type ret_type, + char *name, ...) +{ + struct pevent_function_handler *func_handle; + struct pevent_func_params **next_param; + struct pevent_func_params *param; + enum pevent_func_arg_type type; + va_list ap; + + func_handle = find_func_handler(pevent, name); + if (func_handle) { + /* + * This is most like caused by the users own + * plugins updating the function. This overrides the + * system defaults. + */ + pr_stat("override of function helper '%s'", name); + remove_func_handler(pevent, name); + } + + func_handle = malloc_or_die(sizeof(*func_handle)); + memset(func_handle, 0, sizeof(*func_handle)); + + func_handle->ret_type = ret_type; + func_handle->name = strdup(name); + func_handle->func = func; + if (!func_handle->name) + die("Failed to allocate function name"); + + next_param = &(func_handle->params); + va_start(ap, name); + for (;;) { + type = va_arg(ap, enum pevent_func_arg_type); + if (type == PEVENT_FUNC_ARG_VOID) + break; + + if (type < 0 || type >= PEVENT_FUNC_ARG_MAX_TYPES) { + warning("Invalid argument type %d", type); + goto out_free; + } + + param = malloc_or_die(sizeof(*param)); + param->type = type; + param->next = NULL; + + *next_param = param; + next_param = &(param->next); + + func_handle->nr_args++; + } + va_end(ap); + + func_handle->next = pevent->func_handlers; + pevent->func_handlers = func_handle; + + return 0; + out_free: + va_end(ap); + free_func_handle(func_handle); + return -1; +} + +/** + * pevent_register_event_handle - register a way to parse an event + * @pevent: the handle to the pevent + * @id: the id of the event to register + * @sys_name: the system name the event belongs to + * @event_name: the name of the event + * @func: the function to call to parse the event information + * + * This function allows a developer to override the parsing of + * a given event. If for some reason the default print format + * is not sufficient, this function will register a function + * for an event to be used to parse the data instead. + * + * If @id is >= 0, then it is used to find the event. + * else @sys_name and @event_name are used. + */ +int pevent_register_event_handler(struct pevent *pevent, + int id, char *sys_name, char *event_name, + pevent_event_handler_func func, + void *context) +{ + struct event_format *event; + struct event_handler *handle; + + if (id >= 0) { + /* search by id */ + event = pevent_find_event(pevent, id); + if (!event) + goto not_found; + if (event_name && (strcmp(event_name, event->name) != 0)) + goto not_found; + if (sys_name && (strcmp(sys_name, event->system) != 0)) + goto not_found; + } else { + event = pevent_find_event_by_name(pevent, sys_name, event_name); + if (!event) + goto not_found; + } + + pr_stat("overriding event (%d) %s:%s with new print handler", + event->id, event->system, event->name); + + event->handler = func; + event->context = context; + return 0; + + not_found: + /* Save for later use. */ + handle = malloc_or_die(sizeof(*handle)); + memset(handle, 0, sizeof(*handle)); + handle->id = id; + if (event_name) + handle->event_name = strdup(event_name); + if (sys_name) + handle->sys_name = strdup(sys_name); + + handle->func = func; + handle->next = pevent->handlers; + pevent->handlers = handle; + handle->context = context; + + return -1; +} + +/** + * pevent_alloc - create a pevent handle + */ +struct pevent *pevent_alloc(void) +{ + struct pevent *pevent; + + pevent = malloc(sizeof(*pevent)); + if (!pevent) + return NULL; + memset(pevent, 0, sizeof(*pevent)); + pevent->ref_count = 1; + + return pevent; +} + +void pevent_ref(struct pevent *pevent) +{ + pevent->ref_count++; +} + +static void free_format_fields(struct format_field *field) +{ + struct format_field *next; + + while (field) { + next = field->next; + free(field->type); + free(field->name); + free(field); + field = next; + } +} + +static void free_formats(struct format *format) +{ + free_format_fields(format->common_fields); + free_format_fields(format->fields); +} + +static void free_event(struct event_format *event) +{ + free(event->name); + free(event->system); + + free_formats(&event->format); + + free(event->print_fmt.format); + free_args(event->print_fmt.args); + + free(event); +} + +/** + * pevent_free - free a pevent handle + * @pevent: the pevent handle to free + */ +void pevent_free(struct pevent *pevent) +{ + struct cmdline_list *cmdlist, *cmdnext; + struct func_list *funclist, *funcnext; + struct printk_list *printklist, *printknext; + struct pevent_function_handler *func_handler; + struct event_handler *handle; + int i; + + if (!pevent) + return; + + cmdlist = pevent->cmdlist; + funclist = pevent->funclist; + printklist = pevent->printklist; + + pevent->ref_count--; + if (pevent->ref_count) + return; + + if (pevent->cmdlines) { + for (i = 0; i < pevent->cmdline_count; i++) + free(pevent->cmdlines[i].comm); + free(pevent->cmdlines); + } + + while (cmdlist) { + cmdnext = cmdlist->next; + free(cmdlist->comm); + free(cmdlist); + cmdlist = cmdnext; + } + + if (pevent->func_map) { + for (i = 0; i < pevent->func_count; i++) { + free(pevent->func_map[i].func); + free(pevent->func_map[i].mod); + } + free(pevent->func_map); + } + + while (funclist) { + funcnext = funclist->next; + free(funclist->func); + free(funclist->mod); + free(funclist); + funclist = funcnext; + } + + while (pevent->func_handlers) { + func_handler = pevent->func_handlers; + pevent->func_handlers = func_handler->next; + free_func_handle(func_handler); + } + + if (pevent->printk_map) { + for (i = 0; i < pevent->printk_count; i++) + free(pevent->printk_map[i].printk); + free(pevent->printk_map); + } + + while (printklist) { + printknext = printklist->next; + free(printklist->printk); + free(printklist); + printklist = printknext; + } + + for (i = 0; i < pevent->nr_events; i++) + free_event(pevent->events[i]); + + while (pevent->handlers) { + handle = pevent->handlers; + pevent->handlers = handle->next; + free_handler(handle); + } + + free(pevent->events); + free(pevent->sort_events); + + free(pevent); +} + +void pevent_unref(struct pevent *pevent) +{ + pevent_free(pevent); +} diff --git a/tools/lib/traceevent/event-parse.h b/tools/lib/traceevent/event-parse.h new file mode 100644 index 000000000000..ac997bc7b592 --- /dev/null +++ b/tools/lib/traceevent/event-parse.h @@ -0,0 +1,804 @@ +/* + * Copyright (C) 2009, 2010 Red Hat Inc, Steven Rostedt <srostedt@redhat.com> + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License (not later!) + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + */ +#ifndef _PARSE_EVENTS_H +#define _PARSE_EVENTS_H + +#include <stdarg.h> +#include <regex.h> + +#ifndef __unused +#define __unused __attribute__ ((unused)) +#endif + +/* ----------------------- trace_seq ----------------------- */ + + +#ifndef TRACE_SEQ_BUF_SIZE +#define TRACE_SEQ_BUF_SIZE 4096 +#endif + +#ifndef DEBUG_RECORD +#define DEBUG_RECORD 0 +#endif + +struct pevent_record { + unsigned long long ts; + unsigned long long offset; + long long missed_events; /* buffer dropped events before */ + int record_size; /* size of binary record */ + int size; /* size of data */ + void *data; + int cpu; + int ref_count; + int locked; /* Do not free, even if ref_count is zero */ + void *private; +#if DEBUG_RECORD + struct pevent_record *prev; + struct pevent_record *next; + long alloc_addr; +#endif +}; + +/* + * Trace sequences are used to allow a function to call several other functions + * to create a string of data to use (up to a max of PAGE_SIZE). + */ + +struct trace_seq { + char *buffer; + unsigned int buffer_size; + unsigned int len; + unsigned int readpos; +}; + +void trace_seq_init(struct trace_seq *s); +void trace_seq_destroy(struct trace_seq *s); + +extern int trace_seq_printf(struct trace_seq *s, const char *fmt, ...) + __attribute__ ((format (printf, 2, 3))); +extern int trace_seq_vprintf(struct trace_seq *s, const char *fmt, va_list args) + __attribute__ ((format (printf, 2, 0))); + +extern int trace_seq_puts(struct trace_seq *s, const char *str); +extern int trace_seq_putc(struct trace_seq *s, unsigned char c); + +extern void trace_seq_terminate(struct trace_seq *s); + +extern int trace_seq_do_printf(struct trace_seq *s); + + +/* ----------------------- pevent ----------------------- */ + +struct pevent; +struct event_format; + +typedef int (*pevent_event_handler_func)(struct trace_seq *s, + struct pevent_record *record, + struct event_format *event, + void *context); + +typedef int (*pevent_plugin_load_func)(struct pevent *pevent); +typedef int (*pevent_plugin_unload_func)(void); + +struct plugin_option { + struct plugin_option *next; + void *handle; + char *file; + char *name; + char *plugin_alias; + char *description; + char *value; + void *private; + int set; +}; + +/* + * Plugin hooks that can be called: + * + * PEVENT_PLUGIN_LOADER: (required) + * The function name to initialized the plugin. + * + * int PEVENT_PLUGIN_LOADER(struct pevent *pevent) + * + * PEVENT_PLUGIN_UNLOADER: (optional) + * The function called just before unloading + * + * int PEVENT_PLUGIN_UNLOADER(void) + * + * PEVENT_PLUGIN_OPTIONS: (optional) + * Plugin options that can be set before loading + * + * struct plugin_option PEVENT_PLUGIN_OPTIONS[] = { + * { + * .name = "option-name", + * .plugin_alias = "overide-file-name", (optional) + * .description = "description of option to show users", + * }, + * { + * .name = NULL, + * }, + * }; + * + * Array must end with .name = NULL; + * + * + * .plugin_alias is used to give a shorter name to access + * the vairable. Useful if a plugin handles more than one event. + * + * PEVENT_PLUGIN_ALIAS: (optional) + * The name to use for finding options (uses filename if not defined) + */ +#define PEVENT_PLUGIN_LOADER pevent_plugin_loader +#define PEVENT_PLUGIN_UNLOADER pevent_plugin_unloader +#define PEVENT_PLUGIN_OPTIONS pevent_plugin_options +#define PEVENT_PLUGIN_ALIAS pevent_plugin_alias +#define _MAKE_STR(x) #x +#define MAKE_STR(x) _MAKE_STR(x) +#define PEVENT_PLUGIN_LOADER_NAME MAKE_STR(PEVENT_PLUGIN_LOADER) +#define PEVENT_PLUGIN_UNLOADER_NAME MAKE_STR(PEVENT_PLUGIN_UNLOADER) +#define PEVENT_PLUGIN_OPTIONS_NAME MAKE_STR(PEVENT_PLUGIN_OPTIONS) +#define PEVENT_PLUGIN_ALIAS_NAME MAKE_STR(PEVENT_PLUGIN_ALIAS) + +#define NSECS_PER_SEC 1000000000ULL +#define NSECS_PER_USEC 1000ULL + +enum format_flags { + FIELD_IS_ARRAY = 1, + FIELD_IS_POINTER = 2, + FIELD_IS_SIGNED = 4, + FIELD_IS_STRING = 8, + FIELD_IS_DYNAMIC = 16, + FIELD_IS_LONG = 32, + FIELD_IS_FLAG = 64, + FIELD_IS_SYMBOLIC = 128, +}; + +struct format_field { + struct format_field *next; + struct event_format *event; + char *type; + char *name; + int offset; + int size; + unsigned int arraylen; + unsigned int elementsize; + unsigned long flags; +}; + +struct format { + int nr_common; + int nr_fields; + struct format_field *common_fields; + struct format_field *fields; +}; + +struct print_arg_atom { + char *atom; +}; + +struct print_arg_string { + char *string; + int offset; +}; + +struct print_arg_field { + char *name; + struct format_field *field; +}; + +struct print_flag_sym { + struct print_flag_sym *next; + char *value; + char *str; +}; + +struct print_arg_typecast { + char *type; + struct print_arg *item; +}; + +struct print_arg_flags { + struct print_arg *field; + char *delim; + struct print_flag_sym *flags; +}; + +struct print_arg_symbol { + struct print_arg *field; + struct print_flag_sym *symbols; +}; + +struct print_arg_dynarray { + struct format_field *field; + struct print_arg *index; +}; + +struct print_arg; + +struct print_arg_op { + char *op; + int prio; + struct print_arg *left; + struct print_arg *right; +}; + +struct pevent_function_handler; + +struct print_arg_func { + struct pevent_function_handler *func; + struct print_arg *args; +}; + +enum print_arg_type { + PRINT_NULL, + PRINT_ATOM, + PRINT_FIELD, + PRINT_FLAGS, + PRINT_SYMBOL, + PRINT_TYPE, + PRINT_STRING, + PRINT_BSTRING, + PRINT_DYNAMIC_ARRAY, + PRINT_OP, + PRINT_FUNC, +}; + +struct print_arg { + struct print_arg *next; + enum print_arg_type type; + union { + struct print_arg_atom atom; + struct print_arg_field field; + struct print_arg_typecast typecast; + struct print_arg_flags flags; + struct print_arg_symbol symbol; + struct print_arg_func func; + struct print_arg_string string; + struct print_arg_op op; + struct print_arg_dynarray dynarray; + }; +}; + +struct print_fmt { + char *format; + struct print_arg *args; +}; + +struct event_format { + struct pevent *pevent; + char *name; + int id; + int flags; + struct format format; + struct print_fmt print_fmt; + char *system; + pevent_event_handler_func handler; + void *context; +}; + +enum { + EVENT_FL_ISFTRACE = 0x01, + EVENT_FL_ISPRINT = 0x02, + EVENT_FL_ISBPRINT = 0x04, + EVENT_FL_ISFUNCENT = 0x10, + EVENT_FL_ISFUNCRET = 0x20, + + EVENT_FL_FAILED = 0x80000000 +}; + +enum event_sort_type { + EVENT_SORT_ID, + EVENT_SORT_NAME, + EVENT_SORT_SYSTEM, +}; + +enum event_type { + EVENT_ERROR, + EVENT_NONE, + EVENT_SPACE, + EVENT_NEWLINE, + EVENT_OP, + EVENT_DELIM, + EVENT_ITEM, + EVENT_DQUOTE, + EVENT_SQUOTE, +}; + +typedef unsigned long long (*pevent_func_handler)(struct trace_seq *s, + unsigned long long *args); + +enum pevent_func_arg_type { + PEVENT_FUNC_ARG_VOID, + PEVENT_FUNC_ARG_INT, + PEVENT_FUNC_ARG_LONG, + PEVENT_FUNC_ARG_STRING, + PEVENT_FUNC_ARG_PTR, + PEVENT_FUNC_ARG_MAX_TYPES +}; + +enum pevent_flag { + PEVENT_NSEC_OUTPUT = 1, /* output in NSECS */ +}; + +struct cmdline; +struct cmdline_list; +struct func_map; +struct func_list; +struct event_handler; + +struct pevent { + int ref_count; + + int header_page_ts_offset; + int header_page_ts_size; + int header_page_size_offset; + int header_page_size_size; + int header_page_data_offset; + int header_page_data_size; + int header_page_overwrite; + + int file_bigendian; + int host_bigendian; + + int latency_format; + + int old_format; + + int cpus; + int long_size; + + struct cmdline *cmdlines; + struct cmdline_list *cmdlist; + int cmdline_count; + + struct func_map *func_map; + struct func_list *funclist; + unsigned int func_count; + + struct printk_map *printk_map; + struct printk_list *printklist; + unsigned int printk_count; + + + struct event_format **events; + int nr_events; + struct event_format **sort_events; + enum event_sort_type last_type; + + int type_offset; + int type_size; + + int pid_offset; + int pid_size; + + int pc_offset; + int pc_size; + + int flags_offset; + int flags_size; + + int ld_offset; + int ld_size; + + int print_raw; + + int test_filters; + + int flags; + + struct format_field *bprint_ip_field; + struct format_field *bprint_fmt_field; + struct format_field *bprint_buf_field; + + struct event_handler *handlers; + struct pevent_function_handler *func_handlers; + + /* cache */ + struct event_format *last_event; +}; + +static inline void pevent_set_flag(struct pevent *pevent, int flag) +{ + pevent->flags |= flag; +} + +static inline unsigned short +__data2host2(struct pevent *pevent, unsigned short data) +{ + unsigned short swap; + + if (pevent->host_bigendian == pevent->file_bigendian) + return data; + + swap = ((data & 0xffULL) << 8) | + ((data & (0xffULL << 8)) >> 8); + + return swap; +} + +static inline unsigned int +__data2host4(struct pevent *pevent, unsigned int data) +{ + unsigned int swap; + + if (pevent->host_bigendian == pevent->file_bigendian) + return data; + + swap = ((data & 0xffULL) << 24) | + ((data & (0xffULL << 8)) << 8) | + ((data & (0xffULL << 16)) >> 8) | + ((data & (0xffULL << 24)) >> 24); + + return swap; +} + +static inline unsigned long long +__data2host8(struct pevent *pevent, unsigned long long data) +{ + unsigned long long swap; + + if (pevent->host_bigendian == pevent->file_bigendian) + return data; + + swap = ((data & 0xffULL) << 56) | + ((data & (0xffULL << 8)) << 40) | + ((data & (0xffULL << 16)) << 24) | + ((data & (0xffULL << 24)) << 8) | + ((data & (0xffULL << 32)) >> 8) | + ((data & (0xffULL << 40)) >> 24) | + ((data & (0xffULL << 48)) >> 40) | + ((data & (0xffULL << 56)) >> 56); + + return swap; +} + +#define data2host2(pevent, ptr) __data2host2(pevent, *(unsigned short *)(ptr)) +#define data2host4(pevent, ptr) __data2host4(pevent, *(unsigned int *)(ptr)) +#define data2host8(pevent, ptr) \ +({ \ + unsigned long long __val; \ + \ + memcpy(&__val, (ptr), sizeof(unsigned long long)); \ + __data2host8(pevent, __val); \ +}) + +/* taken from kernel/trace/trace.h */ +enum trace_flag_type { + TRACE_FLAG_IRQS_OFF = 0x01, + TRACE_FLAG_IRQS_NOSUPPORT = 0x02, + TRACE_FLAG_NEED_RESCHED = 0x04, + TRACE_FLAG_HARDIRQ = 0x08, + TRACE_FLAG_SOFTIRQ = 0x10, +}; + +int pevent_register_comm(struct pevent *pevent, const char *comm, int pid); +int pevent_register_function(struct pevent *pevent, char *name, + unsigned long long addr, char *mod); +int pevent_register_print_string(struct pevent *pevent, char *fmt, + unsigned long long addr); +int pevent_pid_is_registered(struct pevent *pevent, int pid); + +void pevent_print_event(struct pevent *pevent, struct trace_seq *s, + struct pevent_record *record); + +int pevent_parse_header_page(struct pevent *pevent, char *buf, unsigned long size, + int long_size); + +int pevent_parse_event(struct pevent *pevent, const char *buf, + unsigned long size, const char *sys); + +void *pevent_get_field_raw(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + int *len, int err); + +int pevent_get_field_val(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + unsigned long long *val, int err); +int pevent_get_common_field_val(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + unsigned long long *val, int err); +int pevent_get_any_field_val(struct trace_seq *s, struct event_format *event, + const char *name, struct pevent_record *record, + unsigned long long *val, int err); + +int pevent_print_num_field(struct trace_seq *s, const char *fmt, + struct event_format *event, const char *name, + struct pevent_record *record, int err); + +int pevent_register_event_handler(struct pevent *pevent, int id, char *sys_name, char *event_name, + pevent_event_handler_func func, void *context); +int pevent_register_print_function(struct pevent *pevent, + pevent_func_handler func, + enum pevent_func_arg_type ret_type, + char *name, ...); + +struct format_field *pevent_find_common_field(struct event_format *event, const char *name); +struct format_field *pevent_find_field(struct event_format *event, const char *name); +struct format_field *pevent_find_any_field(struct event_format *event, const char *name); + +const char *pevent_find_function(struct pevent *pevent, unsigned long long addr); +unsigned long long +pevent_find_function_address(struct pevent *pevent, unsigned long long addr); +unsigned long long pevent_read_number(struct pevent *pevent, const void *ptr, int size); +int pevent_read_number_field(struct format_field *field, const void *data, + unsigned long long *value); + +struct event_format *pevent_find_event(struct pevent *pevent, int id); + +struct event_format * +pevent_find_event_by_name(struct pevent *pevent, const char *sys, const char *name); + +void pevent_data_lat_fmt(struct pevent *pevent, + struct trace_seq *s, struct pevent_record *record); +int pevent_data_type(struct pevent *pevent, struct pevent_record *rec); +struct event_format *pevent_data_event_from_type(struct pevent *pevent, int type); +int pevent_data_pid(struct pevent *pevent, struct pevent_record *rec); +const char *pevent_data_comm_from_pid(struct pevent *pevent, int pid); +void pevent_event_info(struct trace_seq *s, struct event_format *event, + struct pevent_record *record); + +struct event_format **pevent_list_events(struct pevent *pevent, enum event_sort_type); +struct format_field **pevent_event_common_fields(struct event_format *event); +struct format_field **pevent_event_fields(struct event_format *event); + +static inline int pevent_get_cpus(struct pevent *pevent) +{ + return pevent->cpus; +} + +static inline void pevent_set_cpus(struct pevent *pevent, int cpus) +{ + pevent->cpus = cpus; +} + +static inline int pevent_get_long_size(struct pevent *pevent) +{ + return pevent->long_size; +} + +static inline void pevent_set_long_size(struct pevent *pevent, int long_size) +{ + pevent->long_size = long_size; +} + +static inline int pevent_is_file_bigendian(struct pevent *pevent) +{ + return pevent->file_bigendian; +} + +static inline void pevent_set_file_bigendian(struct pevent *pevent, int endian) +{ + pevent->file_bigendian = endian; +} + +static inline int pevent_is_host_bigendian(struct pevent *pevent) +{ + return pevent->host_bigendian; +} + +static inline void pevent_set_host_bigendian(struct pevent *pevent, int endian) +{ + pevent->host_bigendian = endian; +} + +static inline int pevent_is_latency_format(struct pevent *pevent) +{ + return pevent->latency_format; +} + +static inline void pevent_set_latency_format(struct pevent *pevent, int lat) +{ + pevent->latency_format = lat; +} + +struct pevent *pevent_alloc(void); +void pevent_free(struct pevent *pevent); +void pevent_ref(struct pevent *pevent); +void pevent_unref(struct pevent *pevent); + +/* access to the internal parser */ +void pevent_buffer_init(const char *buf, unsigned long long size); +enum event_type pevent_read_token(char **tok); +void pevent_free_token(char *token); +int pevent_peek_char(void); +const char *pevent_get_input_buf(void); +unsigned long long pevent_get_input_buf_ptr(void); + +/* for debugging */ +void pevent_print_funcs(struct pevent *pevent); +void pevent_print_printk(struct pevent *pevent); + +/* ----------------------- filtering ----------------------- */ + +enum filter_boolean_type { + FILTER_FALSE, + FILTER_TRUE, +}; + +enum filter_op_type { + FILTER_OP_AND = 1, + FILTER_OP_OR, + FILTER_OP_NOT, +}; + +enum filter_cmp_type { + FILTER_CMP_NONE, + FILTER_CMP_EQ, + FILTER_CMP_NE, + FILTER_CMP_GT, + FILTER_CMP_LT, + FILTER_CMP_GE, + FILTER_CMP_LE, + FILTER_CMP_MATCH, + FILTER_CMP_NOT_MATCH, + FILTER_CMP_REGEX, + FILTER_CMP_NOT_REGEX, +}; + +enum filter_exp_type { + FILTER_EXP_NONE, + FILTER_EXP_ADD, + FILTER_EXP_SUB, + FILTER_EXP_MUL, + FILTER_EXP_DIV, + FILTER_EXP_MOD, + FILTER_EXP_RSHIFT, + FILTER_EXP_LSHIFT, + FILTER_EXP_AND, + FILTER_EXP_OR, + FILTER_EXP_XOR, + FILTER_EXP_NOT, +}; + +enum filter_arg_type { + FILTER_ARG_NONE, + FILTER_ARG_BOOLEAN, + FILTER_ARG_VALUE, + FILTER_ARG_FIELD, + FILTER_ARG_EXP, + FILTER_ARG_OP, + FILTER_ARG_NUM, + FILTER_ARG_STR, +}; + +enum filter_value_type { + FILTER_NUMBER, + FILTER_STRING, + FILTER_CHAR +}; + +struct fliter_arg; + +struct filter_arg_boolean { + enum filter_boolean_type value; +}; + +struct filter_arg_field { + struct format_field *field; +}; + +struct filter_arg_value { + enum filter_value_type type; + union { + char *str; + unsigned long long val; + }; +}; + +struct filter_arg_op { + enum filter_op_type type; + struct filter_arg *left; + struct filter_arg *right; +}; + +struct filter_arg_exp { + enum filter_exp_type type; + struct filter_arg *left; + struct filter_arg *right; +}; + +struct filter_arg_num { + enum filter_cmp_type type; + struct filter_arg *left; + struct filter_arg *right; +}; + +struct filter_arg_str { + enum filter_cmp_type type; + struct format_field *field; + char *val; + char *buffer; + regex_t reg; +}; + +struct filter_arg { + enum filter_arg_type type; + union { + struct filter_arg_boolean boolean; + struct filter_arg_field field; + struct filter_arg_value value; + struct filter_arg_op op; + struct filter_arg_exp exp; + struct filter_arg_num num; + struct filter_arg_str str; + }; +}; + +struct filter_type { + int event_id; + struct event_format *event; + struct filter_arg *filter; +}; + +struct event_filter { + struct pevent *pevent; + int filters; + struct filter_type *event_filters; +}; + +struct event_filter *pevent_filter_alloc(struct pevent *pevent); + +#define FILTER_NONE -2 +#define FILTER_NOEXIST -1 +#define FILTER_MISS 0 +#define FILTER_MATCH 1 + +enum filter_trivial_type { + FILTER_TRIVIAL_FALSE, + FILTER_TRIVIAL_TRUE, + FILTER_TRIVIAL_BOTH, +}; + +int pevent_filter_add_filter_str(struct event_filter *filter, + const char *filter_str, + char **error_str); + + +int pevent_filter_match(struct event_filter *filter, + struct pevent_record *record); + +int pevent_event_filtered(struct event_filter *filter, + int event_id); + +void pevent_filter_reset(struct event_filter *filter); + +void pevent_filter_clear_trivial(struct event_filter *filter, + enum filter_trivial_type type); + +void pevent_filter_free(struct event_filter *filter); + +char *pevent_filter_make_string(struct event_filter *filter, int event_id); + +int pevent_filter_remove_event(struct event_filter *filter, + int event_id); + +int pevent_filter_event_has_trivial(struct event_filter *filter, + int event_id, + enum filter_trivial_type type); + +int pevent_filter_copy(struct event_filter *dest, struct event_filter *source); + +int pevent_update_trivial(struct event_filter *dest, struct event_filter *source, + enum filter_trivial_type type); + +int pevent_filter_compare(struct event_filter *filter1, struct event_filter *filter2); + +#endif /* _PARSE_EVENTS_H */ diff --git a/tools/lib/traceevent/event-utils.h b/tools/lib/traceevent/event-utils.h new file mode 100644 index 000000000000..08296383d1e6 --- /dev/null +++ b/tools/lib/traceevent/event-utils.h @@ -0,0 +1,80 @@ +/* + * Copyright (C) 2010 Red Hat Inc, Steven Rostedt <srostedt@redhat.com> + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License (not later!) + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + */ +#ifndef __UTIL_H +#define __UTIL_H + +#include <ctype.h> + +/* Can be overridden */ +void die(const char *fmt, ...); +void *malloc_or_die(unsigned int size); +void warning(const char *fmt, ...); +void pr_stat(const char *fmt, ...); +void vpr_stat(const char *fmt, va_list ap); + +/* Always available */ +void __die(const char *fmt, ...); +void __warning(const char *fmt, ...); +void __pr_stat(const char *fmt, ...); + +void __vdie(const char *fmt, ...); +void __vwarning(const char *fmt, ...); +void __vpr_stat(const char *fmt, ...); + +static inline char *strim(char *string) +{ + char *ret; + + if (!string) + return NULL; + while (*string) { + if (!isspace(*string)) + break; + string++; + } + ret = string; + + string = ret + strlen(ret) - 1; + while (string > ret) { + if (!isspace(*string)) + break; + string--; + } + string[1] = 0; + + return ret; +} + +static inline int has_text(const char *text) +{ + if (!text) + return 0; + + while (*text) { + if (!isspace(*text)) + return 1; + text++; + } + + return 0; +} + +#endif diff --git a/tools/lib/traceevent/parse-filter.c b/tools/lib/traceevent/parse-filter.c new file mode 100644 index 000000000000..dfcfe2c131de --- /dev/null +++ b/tools/lib/traceevent/parse-filter.c @@ -0,0 +1,2261 @@ +/* + * Copyright (C) 2010 Red Hat Inc, Steven Rostedt <srostedt@redhat.com> + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License (not later!) + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + */ +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <stdarg.h> +#include <errno.h> +#include <sys/types.h> + +#include "event-parse.h" +#include "event-utils.h" + +#define COMM "COMM" + +static struct format_field comm = { + .name = "COMM", +}; + +struct event_list { + struct event_list *next; + struct event_format *event; +}; + +#define MAX_ERR_STR_SIZE 256 + +static void show_error(char **error_str, const char *fmt, ...) +{ + unsigned long long index; + const char *input; + char *error; + va_list ap; + int len; + int i; + + if (!error_str) + return; + + input = pevent_get_input_buf(); + index = pevent_get_input_buf_ptr(); + len = input ? strlen(input) : 0; + + error = malloc_or_die(MAX_ERR_STR_SIZE + (len*2) + 3); + + if (len) { + strcpy(error, input); + error[len] = '\n'; + for (i = 1; i < len && i < index; i++) + error[len+i] = ' '; + error[len + i] = '^'; + error[len + i + 1] = '\n'; + len += i+2; + } + + va_start(ap, fmt); + vsnprintf(error + len, MAX_ERR_STR_SIZE, fmt, ap); + va_end(ap); + + *error_str = error; +} + +static void free_token(char *token) +{ + pevent_free_token(token); +} + +static enum event_type read_token(char **tok) +{ + enum event_type type; + char *token = NULL; + + do { + free_token(token); + type = pevent_read_token(&token); + } while (type == EVENT_NEWLINE || type == EVENT_SPACE); + + /* If token is = or ! check to see if the next char is ~ */ + if (token && + (strcmp(token, "=") == 0 || strcmp(token, "!") == 0) && + pevent_peek_char() == '~') { + /* append it */ + *tok = malloc(3); + sprintf(*tok, "%c%c", *token, '~'); + free_token(token); + /* Now remove the '~' from the buffer */ + pevent_read_token(&token); + free_token(token); + } else + *tok = token; + + return type; +} + +static int filter_cmp(const void *a, const void *b) +{ + const struct filter_type *ea = a; + const struct filter_type *eb = b; + + if (ea->event_id < eb->event_id) + return -1; + + if (ea->event_id > eb->event_id) + return 1; + + return 0; +} + +static struct filter_type * +find_filter_type(struct event_filter *filter, int id) +{ + struct filter_type *filter_type; + struct filter_type key; + + key.event_id = id; + + filter_type = bsearch(&key, filter->event_filters, + filter->filters, + sizeof(*filter->event_filters), + filter_cmp); + + return filter_type; +} + +static struct filter_type * +add_filter_type(struct event_filter *filter, int id) +{ + struct filter_type *filter_type; + int i; + + filter_type = find_filter_type(filter, id); + if (filter_type) + return filter_type; + + if (!filter->filters) + filter->event_filters = + malloc_or_die(sizeof(*filter->event_filters)); + else { + filter->event_filters = + realloc(filter->event_filters, + sizeof(*filter->event_filters) * + (filter->filters + 1)); + if (!filter->event_filters) + die("Could not allocate filter"); + } + + for (i = 0; i < filter->filters; i++) { + if (filter->event_filters[i].event_id > id) + break; + } + + if (i < filter->filters) + memmove(&filter->event_filters[i+1], + &filter->event_filters[i], + sizeof(*filter->event_filters) * + (filter->filters - i)); + + filter_type = &filter->event_filters[i]; + filter_type->event_id = id; + filter_type->event = pevent_find_event(filter->pevent, id); + filter_type->filter = NULL; + + filter->filters++; + + return filter_type; +} + +/** + * pevent_filter_alloc - create a new event filter + * @pevent: The pevent that this filter is associated with + */ +struct event_filter *pevent_filter_alloc(struct pevent *pevent) +{ + struct event_filter *filter; + + filter = malloc_or_die(sizeof(*filter)); + memset(filter, 0, sizeof(*filter)); + filter->pevent = pevent; + pevent_ref(pevent); + + return filter; +} + +static struct filter_arg *allocate_arg(void) +{ + struct filter_arg *arg; + + arg = malloc_or_die(sizeof(*arg)); + memset(arg, 0, sizeof(*arg)); + + return arg; +} + +static void free_arg(struct filter_arg *arg) +{ + if (!arg) + return; + + switch (arg->type) { + case FILTER_ARG_NONE: + case FILTER_ARG_BOOLEAN: + case FILTER_ARG_NUM: + break; + + case FILTER_ARG_STR: + free(arg->str.val); + regfree(&arg->str.reg); + free(arg->str.buffer); + break; + + case FILTER_ARG_OP: + free_arg(arg->op.left); + free_arg(arg->op.right); + default: + break; + } + + free(arg); +} + +static void add_event(struct event_list **events, + struct event_format *event) +{ + struct event_list *list; + + list = malloc_or_die(sizeof(*list)); + list->next = *events; + *events = list; + list->event = event; +} + +static int event_match(struct event_format *event, + regex_t *sreg, regex_t *ereg) +{ + if (sreg) { + return !regexec(sreg, event->system, 0, NULL, 0) && + !regexec(ereg, event->name, 0, NULL, 0); + } + + return !regexec(ereg, event->system, 0, NULL, 0) || + !regexec(ereg, event->name, 0, NULL, 0); +} + +static int +find_event(struct pevent *pevent, struct event_list **events, + char *sys_name, char *event_name) +{ + struct event_format *event; + regex_t ereg; + regex_t sreg; + int match = 0; + char *reg; + int ret; + int i; + + if (!event_name) { + /* if no name is given, then swap sys and name */ + event_name = sys_name; + sys_name = NULL; + } + + reg = malloc_or_die(strlen(event_name) + 3); + sprintf(reg, "^%s$", event_name); + + ret = regcomp(&ereg, reg, REG_ICASE|REG_NOSUB); + free(reg); + + if (ret) + return -1; + + if (sys_name) { + reg = malloc_or_die(strlen(sys_name) + 3); + sprintf(reg, "^%s$", sys_name); + ret = regcomp(&sreg, reg, REG_ICASE|REG_NOSUB); + free(reg); + if (ret) { + regfree(&ereg); + return -1; + } + } + + for (i = 0; i < pevent->nr_events; i++) { + event = pevent->events[i]; + if (event_match(event, sys_name ? &sreg : NULL, &ereg)) { + match = 1; + add_event(events, event); + } + } + + regfree(&ereg); + if (sys_name) + regfree(&sreg); + + if (!match) + return -1; + + return 0; +} + +static void free_events(struct event_list *events) +{ + struct event_list *event; + + while (events) { + event = events; + events = events->next; + free(event); + } +} + +static struct filter_arg * +create_arg_item(struct event_format *event, const char *token, + enum event_type type, char **error_str) +{ + struct format_field *field; + struct filter_arg *arg; + + arg = allocate_arg(); + + switch (type) { + + case EVENT_SQUOTE: + case EVENT_DQUOTE: + arg->type = FILTER_ARG_VALUE; + arg->value.type = + type == EVENT_DQUOTE ? FILTER_STRING : FILTER_CHAR; + arg->value.str = strdup(token); + if (!arg->value.str) + die("malloc string"); + break; + case EVENT_ITEM: + /* if it is a number, then convert it */ + if (isdigit(token[0])) { + arg->type = FILTER_ARG_VALUE; + arg->value.type = FILTER_NUMBER; + arg->value.val = strtoull(token, NULL, 0); + break; + } + /* Consider this a field */ + field = pevent_find_any_field(event, token); + if (!field) { + if (strcmp(token, COMM) != 0) { + /* not a field, Make it false */ + arg->type = FILTER_ARG_BOOLEAN; + arg->boolean.value = FILTER_FALSE; + break; + } + /* If token is 'COMM' then it is special */ + field = &comm; + } + arg->type = FILTER_ARG_FIELD; + arg->field.field = field; + break; + default: + free_arg(arg); + show_error(error_str, "expected a value but found %s", + token); + return NULL; + } + return arg; +} + +static struct filter_arg * +create_arg_op(enum filter_op_type btype) +{ + struct filter_arg *arg; + + arg = allocate_arg(); + arg->type = FILTER_ARG_OP; + arg->op.type = btype; + + return arg; +} + +static struct filter_arg * +create_arg_exp(enum filter_exp_type etype) +{ + struct filter_arg *arg; + + arg = allocate_arg(); + arg->type = FILTER_ARG_EXP; + arg->op.type = etype; + + return arg; +} + +static struct filter_arg * +create_arg_cmp(enum filter_exp_type etype) +{ + struct filter_arg *arg; + + arg = allocate_arg(); + /* Use NUM and change if necessary */ + arg->type = FILTER_ARG_NUM; + arg->op.type = etype; + + return arg; +} + +static int add_right(struct filter_arg *op, struct filter_arg *arg, + char **error_str) +{ + struct filter_arg *left; + char *str; + int op_type; + int ret; + + switch (op->type) { + case FILTER_ARG_EXP: + if (op->exp.right) + goto out_fail; + op->exp.right = arg; + break; + + case FILTER_ARG_OP: + if (op->op.right) + goto out_fail; + op->op.right = arg; + break; + + case FILTER_ARG_NUM: + if (op->op.right) + goto out_fail; + /* + * The arg must be num, str, or field + */ + switch (arg->type) { + case FILTER_ARG_VALUE: + case FILTER_ARG_FIELD: + break; + default: + show_error(error_str, + "Illegal rvalue"); + return -1; + } + + /* + * Depending on the type, we may need to + * convert this to a string or regex. + */ + switch (arg->value.type) { + case FILTER_CHAR: + /* + * A char should be converted to number if + * the string is 1 byte, and the compare + * is not a REGEX. + */ + if (strlen(arg->value.str) == 1 && + op->num.type != FILTER_CMP_REGEX && + op->num.type != FILTER_CMP_NOT_REGEX) { + arg->value.type = FILTER_NUMBER; + goto do_int; + } + /* fall through */ + case FILTER_STRING: + + /* convert op to a string arg */ + op_type = op->num.type; + left = op->num.left; + str = arg->value.str; + + /* reset the op for the new field */ + memset(op, 0, sizeof(*op)); + + /* + * If left arg was a field not found then + * NULL the entire op. + */ + if (left->type == FILTER_ARG_BOOLEAN) { + free_arg(left); + free_arg(arg); + op->type = FILTER_ARG_BOOLEAN; + op->boolean.value = FILTER_FALSE; + break; + } + + /* Left arg must be a field */ + if (left->type != FILTER_ARG_FIELD) { + show_error(error_str, + "Illegal lvalue for string comparison"); + return -1; + } + + /* Make sure this is a valid string compare */ + switch (op_type) { + case FILTER_CMP_EQ: + op_type = FILTER_CMP_MATCH; + break; + case FILTER_CMP_NE: + op_type = FILTER_CMP_NOT_MATCH; + break; + + case FILTER_CMP_REGEX: + case FILTER_CMP_NOT_REGEX: + ret = regcomp(&op->str.reg, str, REG_ICASE|REG_NOSUB); + if (ret) { + show_error(error_str, + "RegEx '%s' did not compute", + str); + return -1; + } + break; + default: + show_error(error_str, + "Illegal comparison for string"); + return -1; + } + + op->type = FILTER_ARG_STR; + op->str.type = op_type; + op->str.field = left->field.field; + op->str.val = strdup(str); + if (!op->str.val) + die("malloc string"); + /* + * Need a buffer to copy data for tests + */ + op->str.buffer = malloc_or_die(op->str.field->size + 1); + /* Null terminate this buffer */ + op->str.buffer[op->str.field->size] = 0; + + /* We no longer have left or right args */ + free_arg(arg); + free_arg(left); + + break; + + case FILTER_NUMBER: + + do_int: + switch (op->num.type) { + case FILTER_CMP_REGEX: + case FILTER_CMP_NOT_REGEX: + show_error(error_str, + "Op not allowed with integers"); + return -1; + + default: + break; + } + + /* numeric compare */ + op->num.right = arg; + break; + default: + goto out_fail; + } + break; + default: + goto out_fail; + } + + return 0; + + out_fail: + show_error(error_str, + "Syntax error"); + return -1; +} + +static struct filter_arg * +rotate_op_right(struct filter_arg *a, struct filter_arg *b) +{ + struct filter_arg *arg; + + arg = a->op.right; + a->op.right = b; + return arg; +} + +static int add_left(struct filter_arg *op, struct filter_arg *arg) +{ + switch (op->type) { + case FILTER_ARG_EXP: + if (arg->type == FILTER_ARG_OP) + arg = rotate_op_right(arg, op); + op->exp.left = arg; + break; + + case FILTER_ARG_OP: + op->op.left = arg; + break; + case FILTER_ARG_NUM: + if (arg->type == FILTER_ARG_OP) + arg = rotate_op_right(arg, op); + + /* left arg of compares must be a field */ + if (arg->type != FILTER_ARG_FIELD && + arg->type != FILTER_ARG_BOOLEAN) + return -1; + op->num.left = arg; + break; + default: + return -1; + } + return 0; +} + +enum op_type { + OP_NONE, + OP_BOOL, + OP_NOT, + OP_EXP, + OP_CMP, +}; + +static enum op_type process_op(const char *token, + enum filter_op_type *btype, + enum filter_cmp_type *ctype, + enum filter_exp_type *etype) +{ + *btype = FILTER_OP_NOT; + *etype = FILTER_EXP_NONE; + *ctype = FILTER_CMP_NONE; + + if (strcmp(token, "&&") == 0) + *btype = FILTER_OP_AND; + else if (strcmp(token, "||") == 0) + *btype = FILTER_OP_OR; + else if (strcmp(token, "!") == 0) + return OP_NOT; + + if (*btype != FILTER_OP_NOT) + return OP_BOOL; + + /* Check for value expressions */ + if (strcmp(token, "+") == 0) { + *etype = FILTER_EXP_ADD; + } else if (strcmp(token, "-") == 0) { + *etype = FILTER_EXP_SUB; + } else if (strcmp(token, "*") == 0) { + *etype = FILTER_EXP_MUL; + } else if (strcmp(token, "/") == 0) { + *etype = FILTER_EXP_DIV; + } else if (strcmp(token, "%") == 0) { + *etype = FILTER_EXP_MOD; + } else if (strcmp(token, ">>") == 0) { + *etype = FILTER_EXP_RSHIFT; + } else if (strcmp(token, "<<") == 0) { + *etype = FILTER_EXP_LSHIFT; + } else if (strcmp(token, "&") == 0) { + *etype = FILTER_EXP_AND; + } else if (strcmp(token, "|") == 0) { + *etype = FILTER_EXP_OR; + } else if (strcmp(token, "^") == 0) { + *etype = FILTER_EXP_XOR; + } else if (strcmp(token, "~") == 0) + *etype = FILTER_EXP_NOT; + + if (*etype != FILTER_EXP_NONE) + return OP_EXP; + + /* Check for compares */ + if (strcmp(token, "==") == 0) + *ctype = FILTER_CMP_EQ; + else if (strcmp(token, "!=") == 0) + *ctype = FILTER_CMP_NE; + else if (strcmp(token, "<") == 0) + *ctype = FILTER_CMP_LT; + else if (strcmp(token, ">") == 0) + *ctype = FILTER_CMP_GT; + else if (strcmp(token, "<=") == 0) + *ctype = FILTER_CMP_LE; + else if (strcmp(token, ">=") == 0) + *ctype = FILTER_CMP_GE; + else if (strcmp(token, "=~") == 0) + *ctype = FILTER_CMP_REGEX; + else if (strcmp(token, "!~") == 0) + *ctype = FILTER_CMP_NOT_REGEX; + else + return OP_NONE; + + return OP_CMP; +} + +static int check_op_done(struct filter_arg *arg) +{ + switch (arg->type) { + case FILTER_ARG_EXP: + return arg->exp.right != NULL; + + case FILTER_ARG_OP: + return arg->op.right != NULL; + + case FILTER_ARG_NUM: + return arg->num.right != NULL; + + case FILTER_ARG_STR: + /* A string conversion is always done */ + return 1; + + case FILTER_ARG_BOOLEAN: + /* field not found, is ok */ + return 1; + + default: + return 0; + } +} + +enum filter_vals { + FILTER_VAL_NORM, + FILTER_VAL_FALSE, + FILTER_VAL_TRUE, +}; + +void reparent_op_arg(struct filter_arg *parent, struct filter_arg *old_child, + struct filter_arg *arg) +{ + struct filter_arg *other_child; + struct filter_arg **ptr; + + if (parent->type != FILTER_ARG_OP && + arg->type != FILTER_ARG_OP) + die("can not reparent other than OP"); + + /* Get the sibling */ + if (old_child->op.right == arg) { + ptr = &old_child->op.right; + other_child = old_child->op.left; + } else if (old_child->op.left == arg) { + ptr = &old_child->op.left; + other_child = old_child->op.right; + } else + die("Error in reparent op, find other child"); + + /* Detach arg from old_child */ + *ptr = NULL; + + /* Check for root */ + if (parent == old_child) { + free_arg(other_child); + *parent = *arg; + /* Free arg without recussion */ + free(arg); + return; + } + + if (parent->op.right == old_child) + ptr = &parent->op.right; + else if (parent->op.left == old_child) + ptr = &parent->op.left; + else + die("Error in reparent op"); + *ptr = arg; + + free_arg(old_child); +} + +enum filter_vals test_arg(struct filter_arg *parent, struct filter_arg *arg) +{ + enum filter_vals lval, rval; + + switch (arg->type) { + + /* bad case */ + case FILTER_ARG_BOOLEAN: + return FILTER_VAL_FALSE + arg->boolean.value; + + /* good cases: */ + case FILTER_ARG_STR: + case FILTER_ARG_VALUE: + case FILTER_ARG_FIELD: + return FILTER_VAL_NORM; + + case FILTER_ARG_EXP: + lval = test_arg(arg, arg->exp.left); + if (lval != FILTER_VAL_NORM) + return lval; + rval = test_arg(arg, arg->exp.right); + if (rval != FILTER_VAL_NORM) + return rval; + return FILTER_VAL_NORM; + + case FILTER_ARG_NUM: + lval = test_arg(arg, arg->num.left); + if (lval != FILTER_VAL_NORM) + return lval; + rval = test_arg(arg, arg->num.right); + if (rval != FILTER_VAL_NORM) + return rval; + return FILTER_VAL_NORM; + + case FILTER_ARG_OP: + if (arg->op.type != FILTER_OP_NOT) { + lval = test_arg(arg, arg->op.left); + switch (lval) { + case FILTER_VAL_NORM: + break; + case FILTER_VAL_TRUE: + if (arg->op.type == FILTER_OP_OR) + return FILTER_VAL_TRUE; + rval = test_arg(arg, arg->op.right); + if (rval != FILTER_VAL_NORM) + return rval; + + reparent_op_arg(parent, arg, arg->op.right); + return FILTER_VAL_NORM; + + case FILTER_VAL_FALSE: + if (arg->op.type == FILTER_OP_AND) + return FILTER_VAL_FALSE; + rval = test_arg(arg, arg->op.right); + if (rval != FILTER_VAL_NORM) + return rval; + + reparent_op_arg(parent, arg, arg->op.right); + return FILTER_VAL_NORM; + } + } + + rval = test_arg(arg, arg->op.right); + switch (rval) { + case FILTER_VAL_NORM: + break; + case FILTER_VAL_TRUE: + if (arg->op.type == FILTER_OP_OR) + return FILTER_VAL_TRUE; + if (arg->op.type == FILTER_OP_NOT) + return FILTER_VAL_FALSE; + + reparent_op_arg(parent, arg, arg->op.left); + return FILTER_VAL_NORM; + + case FILTER_VAL_FALSE: + if (arg->op.type == FILTER_OP_AND) + return FILTER_VAL_FALSE; + if (arg->op.type == FILTER_OP_NOT) + return FILTER_VAL_TRUE; + + reparent_op_arg(parent, arg, arg->op.left); + return FILTER_VAL_NORM; + } + + return FILTER_VAL_NORM; + default: + die("bad arg in filter tree"); + } + return FILTER_VAL_NORM; +} + +/* Remove any unknown event fields */ +static struct filter_arg *collapse_tree(struct filter_arg *arg) +{ + enum filter_vals ret; + + ret = test_arg(arg, arg); + switch (ret) { + case FILTER_VAL_NORM: + return arg; + + case FILTER_VAL_TRUE: + case FILTER_VAL_FALSE: + free_arg(arg); + arg = allocate_arg(); + arg->type = FILTER_ARG_BOOLEAN; + arg->boolean.value = ret == FILTER_VAL_TRUE; + } + + return arg; +} + +static int +process_filter(struct event_format *event, struct filter_arg **parg, + char **error_str, int not) +{ + enum event_type type; + char *token = NULL; + struct filter_arg *current_op = NULL; + struct filter_arg *current_exp = NULL; + struct filter_arg *left_item = NULL; + struct filter_arg *arg = NULL; + enum op_type op_type; + enum filter_op_type btype; + enum filter_exp_type etype; + enum filter_cmp_type ctype; + int ret; + + *parg = NULL; + + do { + free(token); + type = read_token(&token); + switch (type) { + case EVENT_SQUOTE: + case EVENT_DQUOTE: + case EVENT_ITEM: + arg = create_arg_item(event, token, type, error_str); + if (!arg) + goto fail; + if (!left_item) + left_item = arg; + else if (current_exp) { + ret = add_right(current_exp, arg, error_str); + if (ret < 0) + goto fail; + left_item = NULL; + /* Not's only one one expression */ + if (not) { + arg = NULL; + if (current_op) + goto fail_print; + free(token); + *parg = current_exp; + return 0; + } + } else + goto fail_print; + arg = NULL; + break; + + case EVENT_DELIM: + if (*token == ',') { + show_error(error_str, + "Illegal token ','"); + goto fail; + } + + if (*token == '(') { + if (left_item) { + show_error(error_str, + "Open paren can not come after item"); + goto fail; + } + if (current_exp) { + show_error(error_str, + "Open paren can not come after expression"); + goto fail; + } + + ret = process_filter(event, &arg, error_str, 0); + if (ret != 1) { + if (ret == 0) + show_error(error_str, + "Unbalanced number of '('"); + goto fail; + } + ret = 0; + + /* A not wants just one expression */ + if (not) { + if (current_op) + goto fail_print; + *parg = arg; + return 0; + } + + if (current_op) + ret = add_right(current_op, arg, error_str); + else + current_exp = arg; + + if (ret < 0) + goto fail; + + } else { /* ')' */ + if (!current_op && !current_exp) + goto fail_print; + + /* Make sure everything is finished at this level */ + if (current_exp && !check_op_done(current_exp)) + goto fail_print; + if (current_op && !check_op_done(current_op)) + goto fail_print; + + if (current_op) + *parg = current_op; + else + *parg = current_exp; + return 1; + } + break; + + case EVENT_OP: + op_type = process_op(token, &btype, &ctype, &etype); + + /* All expect a left arg except for NOT */ + switch (op_type) { + case OP_BOOL: + /* Logic ops need a left expression */ + if (!current_exp && !current_op) + goto fail_print; + /* fall through */ + case OP_NOT: + /* logic only processes ops and exp */ + if (left_item) + goto fail_print; + break; + case OP_EXP: + case OP_CMP: + if (!left_item) + goto fail_print; + break; + case OP_NONE: + show_error(error_str, + "Unknown op token %s", token); + goto fail; + } + + ret = 0; + switch (op_type) { + case OP_BOOL: + arg = create_arg_op(btype); + if (current_op) + ret = add_left(arg, current_op); + else + ret = add_left(arg, current_exp); + current_op = arg; + current_exp = NULL; + break; + + case OP_NOT: + arg = create_arg_op(btype); + if (current_op) + ret = add_right(current_op, arg, error_str); + if (ret < 0) + goto fail; + current_exp = arg; + ret = process_filter(event, &arg, error_str, 1); + if (ret < 0) + goto fail; + ret = add_right(current_exp, arg, error_str); + if (ret < 0) + goto fail; + break; + + case OP_EXP: + case OP_CMP: + if (op_type == OP_EXP) + arg = create_arg_exp(etype); + else + arg = create_arg_cmp(ctype); + + if (current_op) + ret = add_right(current_op, arg, error_str); + if (ret < 0) + goto fail; + ret = add_left(arg, left_item); + if (ret < 0) { + arg = NULL; + goto fail_print; + } + current_exp = arg; + break; + default: + break; + } + arg = NULL; + if (ret < 0) + goto fail_print; + break; + case EVENT_NONE: + break; + default: + goto fail_print; + } + } while (type != EVENT_NONE); + + if (!current_op && !current_exp) + goto fail_print; + + if (!current_op) + current_op = current_exp; + + current_op = collapse_tree(current_op); + + *parg = current_op; + + return 0; + + fail_print: + show_error(error_str, "Syntax error"); + fail: + free_arg(current_op); + free_arg(current_exp); + free_arg(arg); + free(token); + return -1; +} + +static int +process_event(struct event_format *event, const char *filter_str, + struct filter_arg **parg, char **error_str) +{ + int ret; + + pevent_buffer_init(filter_str, strlen(filter_str)); + + ret = process_filter(event, parg, error_str, 0); + if (ret == 1) { + show_error(error_str, + "Unbalanced number of ')'"); + return -1; + } + if (ret < 0) + return ret; + + /* If parg is NULL, then make it into FALSE */ + if (!*parg) { + *parg = allocate_arg(); + (*parg)->type = FILTER_ARG_BOOLEAN; + (*parg)->boolean.value = FILTER_FALSE; + } + + return 0; +} + +static int filter_event(struct event_filter *filter, + struct event_format *event, + const char *filter_str, char **error_str) +{ + struct filter_type *filter_type; + struct filter_arg *arg; + int ret; + + if (filter_str) { + ret = process_event(event, filter_str, &arg, error_str); + if (ret < 0) + return ret; + + } else { + /* just add a TRUE arg */ + arg = allocate_arg(); + arg->type = FILTER_ARG_BOOLEAN; + arg->boolean.value = FILTER_TRUE; + } + + filter_type = add_filter_type(filter, event->id); + if (filter_type->filter) + free_arg(filter_type->filter); + filter_type->filter = arg; + + return 0; +} + +/** + * pevent_filter_add_filter_str - add a new filter + * @filter: the event filter to add to + * @filter_str: the filter string that contains the filter + * @error_str: string containing reason for failed filter + * + * Returns 0 if the filter was successfully added + * -1 if there was an error. + * + * On error, if @error_str points to a string pointer, + * it is set to the reason that the filter failed. + * This string must be freed with "free". + */ +int pevent_filter_add_filter_str(struct event_filter *filter, + const char *filter_str, + char **error_str) +{ + struct pevent *pevent = filter->pevent; + struct event_list *event; + struct event_list *events = NULL; + const char *filter_start; + const char *next_event; + char *this_event; + char *event_name = NULL; + char *sys_name = NULL; + char *sp; + int rtn = 0; + int len; + int ret; + + /* clear buffer to reset show error */ + pevent_buffer_init("", 0); + + if (error_str) + *error_str = NULL; + + filter_start = strchr(filter_str, ':'); + if (filter_start) + len = filter_start - filter_str; + else + len = strlen(filter_str); + + + do { + next_event = strchr(filter_str, ','); + if (next_event && + (!filter_start || next_event < filter_start)) + len = next_event - filter_str; + else if (filter_start) + len = filter_start - filter_str; + else + len = strlen(filter_str); + + this_event = malloc_or_die(len + 1); + memcpy(this_event, filter_str, len); + this_event[len] = 0; + + if (next_event) + next_event++; + + filter_str = next_event; + + sys_name = strtok_r(this_event, "/", &sp); + event_name = strtok_r(NULL, "/", &sp); + + if (!sys_name) { + show_error(error_str, "No filter found"); + /* This can only happen when events is NULL, but still */ + free_events(events); + free(this_event); + return -1; + } + + /* Find this event */ + ret = find_event(pevent, &events, strim(sys_name), strim(event_name)); + if (ret < 0) { + if (event_name) + show_error(error_str, + "No event found under '%s.%s'", + sys_name, event_name); + else + show_error(error_str, + "No event found under '%s'", + sys_name); + free_events(events); + free(this_event); + return -1; + } + free(this_event); + } while (filter_str); + + /* Skip the ':' */ + if (filter_start) + filter_start++; + + /* filter starts here */ + for (event = events; event; event = event->next) { + ret = filter_event(filter, event->event, filter_start, + error_str); + /* Failures are returned if a parse error happened */ + if (ret < 0) + rtn = ret; + + if (ret >= 0 && pevent->test_filters) { + char *test; + test = pevent_filter_make_string(filter, event->event->id); + printf(" '%s: %s'\n", event->event->name, test); + free(test); + } + } + + free_events(events); + + if (rtn >= 0 && pevent->test_filters) + exit(0); + + return rtn; +} + +static void free_filter_type(struct filter_type *filter_type) +{ + free_arg(filter_type->filter); +} + +/** + * pevent_filter_remove_event - remove a filter for an event + * @filter: the event filter to remove from + * @event_id: the event to remove a filter for + * + * Removes the filter saved for an event defined by @event_id + * from the @filter. + * + * Returns 1: if an event was removed + * 0: if the event was not found + */ +int pevent_filter_remove_event(struct event_filter *filter, + int event_id) +{ + struct filter_type *filter_type; + unsigned long len; + + if (!filter->filters) + return 0; + + filter_type = find_filter_type(filter, event_id); + + if (!filter_type) + return 0; + + free_filter_type(filter_type); + + /* The filter_type points into the event_filters array */ + len = (unsigned long)(filter->event_filters + filter->filters) - + (unsigned long)(filter_type + 1); + + memmove(filter_type, filter_type + 1, len); + filter->filters--; + + memset(&filter->event_filters[filter->filters], 0, + sizeof(*filter_type)); + + return 1; +} + +/** + * pevent_filter_reset - clear all filters in a filter + * @filter: the event filter to reset + * + * Removes all filters from a filter and resets it. + */ +void pevent_filter_reset(struct event_filter *filter) +{ + int i; + + for (i = 0; i < filter->filters; i++) + free_filter_type(&filter->event_filters[i]); + + free(filter->event_filters); + filter->filters = 0; + filter->event_filters = NULL; +} + +void pevent_filter_free(struct event_filter *filter) +{ + pevent_unref(filter->pevent); + + pevent_filter_reset(filter); + + free(filter); +} + +static char *arg_to_str(struct event_filter *filter, struct filter_arg *arg); + +static int copy_filter_type(struct event_filter *filter, + struct event_filter *source, + struct filter_type *filter_type) +{ + struct filter_arg *arg; + struct event_format *event; + const char *sys; + const char *name; + char *str; + + /* Can't assume that the pevent's are the same */ + sys = filter_type->event->system; + name = filter_type->event->name; + event = pevent_find_event_by_name(filter->pevent, sys, name); + if (!event) + return -1; + + str = arg_to_str(source, filter_type->filter); + if (!str) + return -1; + + if (strcmp(str, "TRUE") == 0 || strcmp(str, "FALSE") == 0) { + /* Add trivial event */ + arg = allocate_arg(); + arg->type = FILTER_ARG_BOOLEAN; + if (strcmp(str, "TRUE") == 0) + arg->boolean.value = 1; + else + arg->boolean.value = 0; + + filter_type = add_filter_type(filter, event->id); + filter_type->filter = arg; + + free(str); + return 0; + } + + filter_event(filter, event, str, NULL); + free(str); + + return 0; +} + +/** + * pevent_filter_copy - copy a filter using another filter + * @dest - the filter to copy to + * @source - the filter to copy from + * + * Returns 0 on success and -1 if not all filters were copied + */ +int pevent_filter_copy(struct event_filter *dest, struct event_filter *source) +{ + int ret = 0; + int i; + + pevent_filter_reset(dest); + + for (i = 0; i < source->filters; i++) { + if (copy_filter_type(dest, source, &source->event_filters[i])) + ret = -1; + } + return ret; +} + + +/** + * pevent_update_trivial - update the trivial filters with the given filter + * @dest - the filter to update + * @source - the filter as the source of the update + * @type - the type of trivial filter to update. + * + * Scan dest for trivial events matching @type to replace with the source. + * + * Returns 0 on success and -1 if there was a problem updating, but + * events may have still been updated on error. + */ +int pevent_update_trivial(struct event_filter *dest, struct event_filter *source, + enum filter_trivial_type type) +{ + struct pevent *src_pevent; + struct pevent *dest_pevent; + struct event_format *event; + struct filter_type *filter_type; + struct filter_arg *arg; + char *str; + int i; + + src_pevent = source->pevent; + dest_pevent = dest->pevent; + + /* Do nothing if either of the filters has nothing to filter */ + if (!dest->filters || !source->filters) + return 0; + + for (i = 0; i < dest->filters; i++) { + filter_type = &dest->event_filters[i]; + arg = filter_type->filter; + if (arg->type != FILTER_ARG_BOOLEAN) + continue; + if ((arg->boolean.value && type == FILTER_TRIVIAL_FALSE) || + (!arg->boolean.value && type == FILTER_TRIVIAL_TRUE)) + continue; + + event = filter_type->event; + + if (src_pevent != dest_pevent) { + /* do a look up */ + event = pevent_find_event_by_name(src_pevent, + event->system, + event->name); + if (!event) + return -1; + } + + str = pevent_filter_make_string(source, event->id); + if (!str) + continue; + + /* Don't bother if the filter is trivial too */ + if (strcmp(str, "TRUE") != 0 && strcmp(str, "FALSE") != 0) + filter_event(dest, event, str, NULL); + free(str); + } + return 0; +} + +/** + * pevent_filter_clear_trivial - clear TRUE and FALSE filters + * @filter: the filter to remove trivial filters from + * @type: remove only true, false, or both + * + * Removes filters that only contain a TRUE or FALES boolean arg. + */ +void pevent_filter_clear_trivial(struct event_filter *filter, + enum filter_trivial_type type) +{ + struct filter_type *filter_type; + int count = 0; + int *ids; + int i; + + if (!filter->filters) + return; + + /* + * Two steps, first get all ids with trivial filters. + * then remove those ids. + */ + for (i = 0; i < filter->filters; i++) { + filter_type = &filter->event_filters[i]; + if (filter_type->filter->type != FILTER_ARG_BOOLEAN) + continue; + switch (type) { + case FILTER_TRIVIAL_FALSE: + if (filter_type->filter->boolean.value) + continue; + case FILTER_TRIVIAL_TRUE: + if (!filter_type->filter->boolean.value) + continue; + default: + break; + } + if (count) + ids = realloc(ids, sizeof(*ids) * (count + 1)); + else + ids = malloc(sizeof(*ids)); + if (!ids) + die("Can't allocate ids"); + ids[count++] = filter_type->event_id; + } + + if (!count) + return; + + for (i = 0; i < count; i++) + pevent_filter_remove_event(filter, ids[i]); + + free(ids); +} + +/** + * pevent_filter_event_has_trivial - return true event contains trivial filter + * @filter: the filter with the information + * @event_id: the id of the event to test + * @type: trivial type to test for (TRUE, FALSE, EITHER) + * + * Returns 1 if the event contains a matching trivial type + * otherwise 0. + */ +int pevent_filter_event_has_trivial(struct event_filter *filter, + int event_id, + enum filter_trivial_type type) +{ + struct filter_type *filter_type; + + if (!filter->filters) + return 0; + + filter_type = find_filter_type(filter, event_id); + + if (!filter_type) + return 0; + + if (filter_type->filter->type != FILTER_ARG_BOOLEAN) + return 0; + + switch (type) { + case FILTER_TRIVIAL_FALSE: + return !filter_type->filter->boolean.value; + + case FILTER_TRIVIAL_TRUE: + return filter_type->filter->boolean.value; + default: + return 1; + } +} + +static int test_filter(struct event_format *event, + struct filter_arg *arg, struct pevent_record *record); + +static const char * +get_comm(struct event_format *event, struct pevent_record *record) +{ + const char *comm; + int pid; + + pid = pevent_data_pid(event->pevent, record); + comm = pevent_data_comm_from_pid(event->pevent, pid); + return comm; +} + +static unsigned long long +get_value(struct event_format *event, + struct format_field *field, struct pevent_record *record) +{ + unsigned long long val; + + /* Handle our dummy "comm" field */ + if (field == &comm) { + const char *name; + + name = get_comm(event, record); + return (unsigned long)name; + } + + pevent_read_number_field(field, record->data, &val); + + if (!(field->flags & FIELD_IS_SIGNED)) + return val; + + switch (field->size) { + case 1: + return (char)val; + case 2: + return (short)val; + case 4: + return (int)val; + case 8: + return (long long)val; + } + return val; +} + +static unsigned long long +get_arg_value(struct event_format *event, struct filter_arg *arg, struct pevent_record *record); + +static unsigned long long +get_exp_value(struct event_format *event, struct filter_arg *arg, struct pevent_record *record) +{ + unsigned long long lval, rval; + + lval = get_arg_value(event, arg->exp.left, record); + rval = get_arg_value(event, arg->exp.right, record); + + switch (arg->exp.type) { + case FILTER_EXP_ADD: + return lval + rval; + + case FILTER_EXP_SUB: + return lval - rval; + + case FILTER_EXP_MUL: + return lval * rval; + + case FILTER_EXP_DIV: + return lval / rval; + + case FILTER_EXP_MOD: + return lval % rval; + + case FILTER_EXP_RSHIFT: + return lval >> rval; + + case FILTER_EXP_LSHIFT: + return lval << rval; + + case FILTER_EXP_AND: + return lval & rval; + + case FILTER_EXP_OR: + return lval | rval; + + case FILTER_EXP_XOR: + return lval ^ rval; + + case FILTER_EXP_NOT: + default: + die("error in exp"); + } + return 0; +} + +static unsigned long long +get_arg_value(struct event_format *event, struct filter_arg *arg, struct pevent_record *record) +{ + switch (arg->type) { + case FILTER_ARG_FIELD: + return get_value(event, arg->field.field, record); + + case FILTER_ARG_VALUE: + if (arg->value.type != FILTER_NUMBER) + die("must have number field!"); + return arg->value.val; + + case FILTER_ARG_EXP: + return get_exp_value(event, arg, record); + + default: + die("oops in filter"); + } + return 0; +} + +static int test_num(struct event_format *event, + struct filter_arg *arg, struct pevent_record *record) +{ + unsigned long long lval, rval; + + lval = get_arg_value(event, arg->num.left, record); + rval = get_arg_value(event, arg->num.right, record); + + switch (arg->num.type) { + case FILTER_CMP_EQ: + return lval == rval; + + case FILTER_CMP_NE: + return lval != rval; + + case FILTER_CMP_GT: + return lval > rval; + + case FILTER_CMP_LT: + return lval < rval; + + case FILTER_CMP_GE: + return lval >= rval; + + case FILTER_CMP_LE: + return lval <= rval; + + default: + /* ?? */ + return 0; + } +} + +static const char *get_field_str(struct filter_arg *arg, struct pevent_record *record) +{ + const char *val = record->data + arg->str.field->offset; + + /* + * We need to copy the data since we can't be sure the field + * is null terminated. + */ + if (*(val + arg->str.field->size - 1)) { + /* copy it */ + memcpy(arg->str.buffer, val, arg->str.field->size); + /* the buffer is already NULL terminated */ + val = arg->str.buffer; + } + return val; +} + +static int test_str(struct event_format *event, + struct filter_arg *arg, struct pevent_record *record) +{ + const char *val; + + if (arg->str.field == &comm) + val = get_comm(event, record); + else + val = get_field_str(arg, record); + + switch (arg->str.type) { + case FILTER_CMP_MATCH: + return strcmp(val, arg->str.val) == 0; + + case FILTER_CMP_NOT_MATCH: + return strcmp(val, arg->str.val) != 0; + + case FILTER_CMP_REGEX: + /* Returns zero on match */ + return !regexec(&arg->str.reg, val, 0, NULL, 0); + + case FILTER_CMP_NOT_REGEX: + return regexec(&arg->str.reg, val, 0, NULL, 0); + + default: + /* ?? */ + return 0; + } +} + +static int test_op(struct event_format *event, + struct filter_arg *arg, struct pevent_record *record) +{ + switch (arg->op.type) { + case FILTER_OP_AND: + return test_filter(event, arg->op.left, record) && + test_filter(event, arg->op.right, record); + + case FILTER_OP_OR: + return test_filter(event, arg->op.left, record) || + test_filter(event, arg->op.right, record); + + case FILTER_OP_NOT: + return !test_filter(event, arg->op.right, record); + + default: + /* ?? */ + return 0; + } +} + +static int test_filter(struct event_format *event, + struct filter_arg *arg, struct pevent_record *record) +{ + switch (arg->type) { + case FILTER_ARG_BOOLEAN: + /* easy case */ + return arg->boolean.value; + + case FILTER_ARG_OP: + return test_op(event, arg, record); + + case FILTER_ARG_NUM: + return test_num(event, arg, record); + + case FILTER_ARG_STR: + return test_str(event, arg, record); + + case FILTER_ARG_EXP: + case FILTER_ARG_VALUE: + case FILTER_ARG_FIELD: + /* + * Expressions, fields and values evaluate + * to true if they return non zero + */ + return !!get_arg_value(event, arg, record); + + default: + die("oops!"); + /* ?? */ + return 0; + } +} + +/** + * pevent_event_filtered - return true if event has filter + * @filter: filter struct with filter information + * @event_id: event id to test if filter exists + * + * Returns 1 if filter found for @event_id + * otherwise 0; + */ +int pevent_event_filtered(struct event_filter *filter, + int event_id) +{ + struct filter_type *filter_type; + + if (!filter->filters) + return 0; + + filter_type = find_filter_type(filter, event_id); + + return filter_type ? 1 : 0; +} + +/** + * pevent_filter_match - test if a record matches a filter + * @filter: filter struct with filter information + * @record: the record to test against the filter + * + * Returns: + * 1 - filter found for event and @record matches + * 0 - filter found for event and @record does not match + * -1 - no filter found for @record's event + * -2 - if no filters exist + */ +int pevent_filter_match(struct event_filter *filter, + struct pevent_record *record) +{ + struct pevent *pevent = filter->pevent; + struct filter_type *filter_type; + int event_id; + + if (!filter->filters) + return FILTER_NONE; + + event_id = pevent_data_type(pevent, record); + + filter_type = find_filter_type(filter, event_id); + + if (!filter_type) + return FILTER_NOEXIST; + + return test_filter(filter_type->event, filter_type->filter, record) ? + FILTER_MATCH : FILTER_MISS; +} + +static char *op_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + char *str = NULL; + char *left = NULL; + char *right = NULL; + char *op = NULL; + int left_val = -1; + int right_val = -1; + int val; + int len; + + switch (arg->op.type) { + case FILTER_OP_AND: + op = "&&"; + /* fall through */ + case FILTER_OP_OR: + if (!op) + op = "||"; + + left = arg_to_str(filter, arg->op.left); + right = arg_to_str(filter, arg->op.right); + if (!left || !right) + break; + + /* Try to consolidate boolean values */ + if (strcmp(left, "TRUE") == 0) + left_val = 1; + else if (strcmp(left, "FALSE") == 0) + left_val = 0; + + if (strcmp(right, "TRUE") == 0) + right_val = 1; + else if (strcmp(right, "FALSE") == 0) + right_val = 0; + + if (left_val >= 0) { + if ((arg->op.type == FILTER_OP_AND && !left_val) || + (arg->op.type == FILTER_OP_OR && left_val)) { + /* Just return left value */ + str = left; + left = NULL; + break; + } + if (right_val >= 0) { + /* just evaluate this. */ + val = 0; + switch (arg->op.type) { + case FILTER_OP_AND: + val = left_val && right_val; + break; + case FILTER_OP_OR: + val = left_val || right_val; + break; + default: + break; + } + str = malloc_or_die(6); + if (val) + strcpy(str, "TRUE"); + else + strcpy(str, "FALSE"); + break; + } + } + if (right_val >= 0) { + if ((arg->op.type == FILTER_OP_AND && !right_val) || + (arg->op.type == FILTER_OP_OR && right_val)) { + /* Just return right value */ + str = right; + right = NULL; + break; + } + /* The right value is meaningless */ + str = left; + left = NULL; + break; + } + + len = strlen(left) + strlen(right) + strlen(op) + 10; + str = malloc_or_die(len); + snprintf(str, len, "(%s) %s (%s)", + left, op, right); + break; + + case FILTER_OP_NOT: + op = "!"; + right = arg_to_str(filter, arg->op.right); + if (!right) + break; + + /* See if we can consolidate */ + if (strcmp(right, "TRUE") == 0) + right_val = 1; + else if (strcmp(right, "FALSE") == 0) + right_val = 0; + if (right_val >= 0) { + /* just return the opposite */ + str = malloc_or_die(6); + if (right_val) + strcpy(str, "FALSE"); + else + strcpy(str, "TRUE"); + break; + } + len = strlen(right) + strlen(op) + 3; + str = malloc_or_die(len); + snprintf(str, len, "%s(%s)", op, right); + break; + + default: + /* ?? */ + break; + } + free(left); + free(right); + return str; +} + +static char *val_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + char *str; + + str = malloc_or_die(30); + + snprintf(str, 30, "%lld", arg->value.val); + + return str; +} + +static char *field_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + return strdup(arg->field.field->name); +} + +static char *exp_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + char *lstr; + char *rstr; + char *op; + char *str; + int len; + + lstr = arg_to_str(filter, arg->exp.left); + rstr = arg_to_str(filter, arg->exp.right); + + switch (arg->exp.type) { + case FILTER_EXP_ADD: + op = "+"; + break; + case FILTER_EXP_SUB: + op = "-"; + break; + case FILTER_EXP_MUL: + op = "*"; + break; + case FILTER_EXP_DIV: + op = "/"; + break; + case FILTER_EXP_MOD: + op = "%"; + break; + case FILTER_EXP_RSHIFT: + op = ">>"; + break; + case FILTER_EXP_LSHIFT: + op = "<<"; + break; + case FILTER_EXP_AND: + op = "&"; + break; + case FILTER_EXP_OR: + op = "|"; + break; + case FILTER_EXP_XOR: + op = "^"; + break; + default: + die("oops in exp"); + } + + len = strlen(op) + strlen(lstr) + strlen(rstr) + 4; + str = malloc_or_die(len); + snprintf(str, len, "%s %s %s", lstr, op, rstr); + free(lstr); + free(rstr); + + return str; +} + +static char *num_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + char *lstr; + char *rstr; + char *str = NULL; + char *op = NULL; + int len; + + lstr = arg_to_str(filter, arg->num.left); + rstr = arg_to_str(filter, arg->num.right); + + switch (arg->num.type) { + case FILTER_CMP_EQ: + op = "=="; + /* fall through */ + case FILTER_CMP_NE: + if (!op) + op = "!="; + /* fall through */ + case FILTER_CMP_GT: + if (!op) + op = ">"; + /* fall through */ + case FILTER_CMP_LT: + if (!op) + op = "<"; + /* fall through */ + case FILTER_CMP_GE: + if (!op) + op = ">="; + /* fall through */ + case FILTER_CMP_LE: + if (!op) + op = "<="; + + len = strlen(lstr) + strlen(op) + strlen(rstr) + 4; + str = malloc_or_die(len); + sprintf(str, "%s %s %s", lstr, op, rstr); + + break; + + default: + /* ?? */ + break; + } + + free(lstr); + free(rstr); + return str; +} + +static char *str_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + char *str = NULL; + char *op = NULL; + int len; + + switch (arg->str.type) { + case FILTER_CMP_MATCH: + op = "=="; + /* fall through */ + case FILTER_CMP_NOT_MATCH: + if (!op) + op = "!="; + /* fall through */ + case FILTER_CMP_REGEX: + if (!op) + op = "=~"; + /* fall through */ + case FILTER_CMP_NOT_REGEX: + if (!op) + op = "!~"; + + len = strlen(arg->str.field->name) + strlen(op) + + strlen(arg->str.val) + 6; + str = malloc_or_die(len); + snprintf(str, len, "%s %s \"%s\"", + arg->str.field->name, + op, arg->str.val); + break; + + default: + /* ?? */ + break; + } + return str; +} + +static char *arg_to_str(struct event_filter *filter, struct filter_arg *arg) +{ + char *str; + + switch (arg->type) { + case FILTER_ARG_BOOLEAN: + str = malloc_or_die(6); + if (arg->boolean.value) + strcpy(str, "TRUE"); + else + strcpy(str, "FALSE"); + return str; + + case FILTER_ARG_OP: + return op_to_str(filter, arg); + + case FILTER_ARG_NUM: + return num_to_str(filter, arg); + + case FILTER_ARG_STR: + return str_to_str(filter, arg); + + case FILTER_ARG_VALUE: + return val_to_str(filter, arg); + + case FILTER_ARG_FIELD: + return field_to_str(filter, arg); + + case FILTER_ARG_EXP: + return exp_to_str(filter, arg); + + default: + /* ?? */ + return NULL; + } + +} + +/** + * pevent_filter_make_string - return a string showing the filter + * @filter: filter struct with filter information + * @event_id: the event id to return the filter string with + * + * Returns a string that displays the filter contents. + * This string must be freed with free(str). + * NULL is returned if no filter is found. + */ +char * +pevent_filter_make_string(struct event_filter *filter, int event_id) +{ + struct filter_type *filter_type; + + if (!filter->filters) + return NULL; + + filter_type = find_filter_type(filter, event_id); + + if (!filter_type) + return NULL; + + return arg_to_str(filter, filter_type->filter); +} + +/** + * pevent_filter_compare - compare two filters and return if they are the same + * @filter1: Filter to compare with @filter2 + * @filter2: Filter to compare with @filter1 + * + * Returns: + * 1 if the two filters hold the same content. + * 0 if they do not. + */ +int pevent_filter_compare(struct event_filter *filter1, struct event_filter *filter2) +{ + struct filter_type *filter_type1; + struct filter_type *filter_type2; + char *str1, *str2; + int result; + int i; + + /* Do the easy checks first */ + if (filter1->filters != filter2->filters) + return 0; + if (!filter1->filters && !filter2->filters) + return 1; + + /* + * Now take a look at each of the events to see if they have the same + * filters to them. + */ + for (i = 0; i < filter1->filters; i++) { + filter_type1 = &filter1->event_filters[i]; + filter_type2 = find_filter_type(filter2, filter_type1->event_id); + if (!filter_type2) + break; + if (filter_type1->filter->type != filter_type2->filter->type) + break; + switch (filter_type1->filter->type) { + case FILTER_TRIVIAL_FALSE: + case FILTER_TRIVIAL_TRUE: + /* trivial types just need the type compared */ + continue; + default: + break; + } + /* The best way to compare complex filters is with strings */ + str1 = arg_to_str(filter1, filter_type1->filter); + str2 = arg_to_str(filter2, filter_type2->filter); + result = strcmp(str1, str2) != 0; + free(str1); + free(str2); + if (result) + break; + } + + if (i < filter1->filters) + return 0; + return 1; +} + diff --git a/tools/lib/traceevent/parse-utils.c b/tools/lib/traceevent/parse-utils.c new file mode 100644 index 000000000000..f023a133abb6 --- /dev/null +++ b/tools/lib/traceevent/parse-utils.c @@ -0,0 +1,110 @@ +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <stdarg.h> +#include <errno.h> + +#define __weak __attribute__((weak)) + +void __vdie(const char *fmt, va_list ap) +{ + int ret = errno; + + if (errno) + perror("trace-cmd"); + else + ret = -1; + + fprintf(stderr, " "); + vfprintf(stderr, fmt, ap); + + fprintf(stderr, "\n"); + exit(ret); +} + +void __die(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + __vdie(fmt, ap); + va_end(ap); +} + +void __weak die(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + __vdie(fmt, ap); + va_end(ap); +} + +void __vwarning(const char *fmt, va_list ap) +{ + if (errno) + perror("trace-cmd"); + errno = 0; + + fprintf(stderr, " "); + vfprintf(stderr, fmt, ap); + + fprintf(stderr, "\n"); +} + +void __warning(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + __vwarning(fmt, ap); + va_end(ap); +} + +void __weak warning(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + __vwarning(fmt, ap); + va_end(ap); +} + +void __vpr_stat(const char *fmt, va_list ap) +{ + vprintf(fmt, ap); + printf("\n"); +} + +void __pr_stat(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + __vpr_stat(fmt, ap); + va_end(ap); +} + +void __weak vpr_stat(const char *fmt, va_list ap) +{ + __vpr_stat(fmt, ap); +} + +void __weak pr_stat(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + __vpr_stat(fmt, ap); + va_end(ap); +} + +void __weak *malloc_or_die(unsigned int size) +{ + void *data; + + data = malloc(size); + if (!data) + die("malloc"); + return data; +} diff --git a/tools/lib/traceevent/trace-seq.c b/tools/lib/traceevent/trace-seq.c new file mode 100644 index 000000000000..b1ccc923e8a5 --- /dev/null +++ b/tools/lib/traceevent/trace-seq.c @@ -0,0 +1,200 @@ +/* + * Copyright (C) 2009 Red Hat Inc, Steven Rostedt <srostedt@redhat.com> + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License (not later!) + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + * + * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + */ +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <stdarg.h> + +#include "event-parse.h" +#include "event-utils.h" + +/* + * The TRACE_SEQ_POISON is to catch the use of using + * a trace_seq structure after it was destroyed. + */ +#define TRACE_SEQ_POISON ((void *)0xdeadbeef) +#define TRACE_SEQ_CHECK(s) \ +do { \ + if ((s)->buffer == TRACE_SEQ_POISON) \ + die("Usage of trace_seq after it was destroyed"); \ +} while (0) + +/** + * trace_seq_init - initialize the trace_seq structure + * @s: a pointer to the trace_seq structure to initialize + */ +void trace_seq_init(struct trace_seq *s) +{ + s->len = 0; + s->readpos = 0; + s->buffer_size = TRACE_SEQ_BUF_SIZE; + s->buffer = malloc_or_die(s->buffer_size); +} + +/** + * trace_seq_destroy - free up memory of a trace_seq + * @s: a pointer to the trace_seq to free the buffer + * + * Only frees the buffer, not the trace_seq struct itself. + */ +void trace_seq_destroy(struct trace_seq *s) +{ + if (!s) + return; + TRACE_SEQ_CHECK(s); + free(s->buffer); + s->buffer = TRACE_SEQ_POISON; +} + +static void expand_buffer(struct trace_seq *s) +{ + s->buffer_size += TRACE_SEQ_BUF_SIZE; + s->buffer = realloc(s->buffer, s->buffer_size); + if (!s->buffer) + die("Can't allocate trace_seq buffer memory"); +} + +/** + * trace_seq_printf - sequence printing of trace information + * @s: trace sequence descriptor + * @fmt: printf format string + * + * It returns 0 if the trace oversizes the buffer's free + * space, 1 otherwise. + * + * The tracer may use either sequence operations or its own + * copy to user routines. To simplify formating of a trace + * trace_seq_printf is used to store strings into a special + * buffer (@s). Then the output may be either used by + * the sequencer or pulled into another buffer. + */ +int +trace_seq_printf(struct trace_seq *s, const char *fmt, ...) +{ + va_list ap; + int len; + int ret; + + TRACE_SEQ_CHECK(s); + + try_again: + len = (s->buffer_size - 1) - s->len; + + va_start(ap, fmt); + ret = vsnprintf(s->buffer + s->len, len, fmt, ap); + va_end(ap); + + if (ret >= len) { + expand_buffer(s); + goto try_again; + } + + s->len += ret; + + return 1; +} + +/** + * trace_seq_vprintf - sequence printing of trace information + * @s: trace sequence descriptor + * @fmt: printf format string + * + * The tracer may use either sequence operations or its own + * copy to user routines. To simplify formating of a trace + * trace_seq_printf is used to store strings into a special + * buffer (@s). Then the output may be either used by + * the sequencer or pulled into another buffer. + */ +int +trace_seq_vprintf(struct trace_seq *s, const char *fmt, va_list args) +{ + int len; + int ret; + + TRACE_SEQ_CHECK(s); + + try_again: + len = (s->buffer_size - 1) - s->len; + + ret = vsnprintf(s->buffer + s->len, len, fmt, args); + + if (ret >= len) { + expand_buffer(s); + goto try_again; + } + + s->len += ret; + + return len; +} + +/** + * trace_seq_puts - trace sequence printing of simple string + * @s: trace sequence descriptor + * @str: simple string to record + * + * The tracer may use either the sequence operations or its own + * copy to user routines. This function records a simple string + * into a special buffer (@s) for later retrieval by a sequencer + * or other mechanism. + */ +int trace_seq_puts(struct trace_seq *s, const char *str) +{ + int len; + + TRACE_SEQ_CHECK(s); + + len = strlen(str); + + while (len > ((s->buffer_size - 1) - s->len)) + expand_buffer(s); + + memcpy(s->buffer + s->len, str, len); + s->len += len; + + return len; +} + +int trace_seq_putc(struct trace_seq *s, unsigned char c) +{ + TRACE_SEQ_CHECK(s); + + while (s->len >= (s->buffer_size - 1)) + expand_buffer(s); + + s->buffer[s->len++] = c; + + return 1; +} + +void trace_seq_terminate(struct trace_seq *s) +{ + TRACE_SEQ_CHECK(s); + + /* There's always one character left on the buffer */ + s->buffer[s->len] = 0; +} + +int trace_seq_do_printf(struct trace_seq *s) +{ + TRACE_SEQ_CHECK(s); + return printf("%.*s", s->len, s->buffer); +} diff --git a/tools/perf/.gitignore b/tools/perf/.gitignore index 416684be0ad3..26b823b61aa1 100644 --- a/tools/perf/.gitignore +++ b/tools/perf/.gitignore @@ -19,3 +19,5 @@ TAGS cscope* config.mak config.mak.autogen +*-bison.* +*-flex.* diff --git a/tools/perf/Documentation/perf-evlist.txt b/tools/perf/Documentation/perf-evlist.txt index 0507ec7bad71..15217345c2fa 100644 --- a/tools/perf/Documentation/perf-evlist.txt +++ b/tools/perf/Documentation/perf-evlist.txt @@ -20,6 +20,14 @@ OPTIONS --input=:: Input file name. (default: perf.data unless stdin is a fifo) +-F:: +--freq=:: + Show just the sample frequency used for each event. + +-v:: +--verbose=:: + Show all fields. + SEE ALSO -------- linkperf:perf-record[1], linkperf:perf-list[1], diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt index 2780d9ce48bf..b715cb71592b 100644 --- a/tools/perf/Documentation/perf-probe.txt +++ b/tools/perf/Documentation/perf-probe.txt @@ -77,7 +77,8 @@ OPTIONS -F:: --funcs:: - Show available functions in given module or kernel. + Show available functions in given module or kernel. With -x/--exec, + can also list functions in a user space executable / shared library. --filter=FILTER:: (Only for --vars and --funcs) Set filter. FILTER is a combination of glob @@ -98,6 +99,15 @@ OPTIONS --max-probes:: Set the maximum number of probe points for an event. Default is 128. +-x:: +--exec=PATH:: + Specify path to the executable or shared library file for user + space tracing. Can also be used with --funcs option. + +In absence of -m/-x options, perf probe checks if the first argument after +the options is an absolute path name. If its an absolute path, perf probe +uses it as a target module/target user space binary to probe. + PROBE SYNTAX ------------ Probe points are defined by following syntax. @@ -182,6 +192,13 @@ Delete all probes on schedule(). ./perf probe --del='schedule*' +Add probes at zfree() function on /bin/zsh + + ./perf probe -x /bin/zsh zfree or ./perf probe /bin/zsh zfree + +Add probes at malloc() function on libc + + ./perf probe -x /lib/libc.so.6 malloc or ./perf probe /lib/libc.so.6 malloc SEE ALSO -------- diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt index a1386b2fff00..b38a1f9ad460 100644 --- a/tools/perf/Documentation/perf-record.txt +++ b/tools/perf/Documentation/perf-record.txt @@ -168,7 +168,7 @@ following filters are defined: - any: any type of branches - any_call: any function call or system call - any_ret: any function return or system call return - - any_ind: any indirect branch + - ind_call: any indirect branch - u: only when the branch target is at the user level - k: only when the branch target is in the kernel - hv: only when the target is at the hypervisor level diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt index 87feeee8b90c..2d89f02719b5 100644 --- a/tools/perf/Documentation/perf-report.txt +++ b/tools/perf/Documentation/perf-report.txt @@ -48,6 +48,9 @@ OPTIONS Only consider these symbols. CSV that understands file://filename entries. +--symbol-filter=:: + Only show symbols that match (partially) with this filter. + -U:: --hide-unresolved:: Only display entries resolved to a symbol. @@ -110,6 +113,8 @@ OPTIONS requires a tty, if one is not present, as when piping to other commands, the stdio interface is used. +--gtk:: Use the GTK2 interface. + -k:: --vmlinux=<file>:: vmlinux pathname diff --git a/tools/perf/Documentation/perfconfig.example b/tools/perf/Documentation/perfconfig.example index d1448668f4d4..767ea2436e1c 100644 --- a/tools/perf/Documentation/perfconfig.example +++ b/tools/perf/Documentation/perfconfig.example @@ -6,6 +6,7 @@ normal = black, lightgray selected = lightgray, magenta code = blue, lightgray + addr = magenta, lightgray [tui] @@ -18,3 +19,11 @@ # Default, disable using /dev/null dir = /root/.debug + +[annotate] + + # Defaults + hide_src_code = false + use_offset = true + jump_arrows = true + show_nr_jumps = false diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST index 5476bc0a1eac..b4b572e8c100 100644 --- a/tools/perf/MANIFEST +++ b/tools/perf/MANIFEST @@ -1,4 +1,6 @@ tools/perf +tools/scripts +tools/lib/traceevent include/linux/const.h include/linux/perf_event.h include/linux/rbtree.h diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 74fd7f89208a..0eee64cfe9a0 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -1,18 +1,10 @@ -ifeq ("$(origin O)", "command line") - OUTPUT := $(O)/ -endif +include ../scripts/Makefile.include # The default target of this Makefile is... all: include config/utilities.mak -ifneq ($(OUTPUT),) -# check that the output directory actually exists -OUTDIR := $(shell cd $(OUTPUT) && /bin/pwd) -$(if $(OUTDIR),, $(error output directory "$(OUTPUT)" does not exist)) -endif - # Define V to have a more verbose compile. # # Define O to save output files in a separate directory. @@ -84,39 +76,20 @@ ifneq ($(WERROR),0) CFLAGS_WERROR := -Werror endif -# -# Include saner warnings here, which can catch bugs: -# - -EXTRA_WARNINGS := -Wformat -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-security -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-y2k -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wshadow -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Winit-self -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wpacked -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wredundant-decls -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-aliasing=3 -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wswitch-default -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wswitch-enum -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wno-system-headers -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wundef -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wwrite-strings -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wbad-function-cast -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wmissing-declarations -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wmissing-prototypes -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wnested-externs -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wold-style-definition -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-prototypes -EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wdeclaration-after-statement - ifeq ("$(origin DEBUG)", "command line") PERF_DEBUG = $(DEBUG) endif ifndef PERF_DEBUG - CFLAGS_OPTIMIZE = -O6 + CFLAGS_OPTIMIZE = -O6 -D_FORTIFY_SOURCE=2 endif -CFLAGS = -fno-omit-frame-pointer -ggdb3 -Wall -Wextra -std=gnu99 $(CFLAGS_WERROR) $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS) $(EXTRA_CFLAGS) +ifdef PARSER_DEBUG + PARSER_DEBUG_BISON := -t + PARSER_DEBUG_FLEX := -d + PARSER_DEBUG_CFLAGS := -DPARSER_DEBUG +endif + +CFLAGS = -fno-omit-frame-pointer -ggdb3 -Wall -Wextra -std=gnu99 $(CFLAGS_WERROR) $(CFLAGS_OPTIMIZE) $(EXTRA_WARNINGS) $(EXTRA_CFLAGS) $(PARSER_DEBUG_CFLAGS) EXTLIBS = -lpthread -lrt -lelf -lm ALL_CFLAGS = $(CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE ALL_LDFLAGS = $(LDFLAGS) @@ -182,7 +155,7 @@ endif ### --- END CONFIGURATION SECTION --- -BASIC_CFLAGS = -Iutil/include -Iarch/$(ARCH)/include -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE +BASIC_CFLAGS = -Iutil/include -Iarch/$(ARCH)/include -I$(OUTPUT)/util -I$(TRACE_EVENT_DIR) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE BASIC_LDFLAGS = # Guard against environment variables @@ -211,6 +184,17 @@ $(OUTPUT)python/perf.so: $(PYRF_OBJS) $(PYTHON_EXT_SRCS) $(PYTHON_EXT_DEPS) SCRIPTS = $(patsubst %.sh,%,$(SCRIPT_SH)) +TRACE_EVENT_DIR = ../lib/traceevent/ + +ifeq ("$(origin O)", "command line") + TE_PATH=$(OUTPUT)/ +else + TE_PATH=$(TRACE_EVENT_DIR)/ +endif + +LIBTRACEEVENT = $(TE_PATH)libtraceevent.a +TE_LIB := -L$(TE_PATH) -ltraceevent + # # Single 'perf' binary right now: # @@ -234,6 +218,24 @@ endif export PERL_PATH +FLEX = flex +BISON= bison + +$(OUTPUT)util/parse-events-flex.c: util/parse-events.l + $(QUIET_FLEX)$(FLEX) --header-file=$(OUTPUT)util/parse-events-flex.h $(PARSER_DEBUG_FLEX) -t util/parse-events.l > $(OUTPUT)util/parse-events-flex.c + +$(OUTPUT)util/parse-events-bison.c: util/parse-events.y + $(QUIET_BISON)$(BISON) -v util/parse-events.y -d $(PARSER_DEBUG_BISON) -o $(OUTPUT)util/parse-events-bison.c + +$(OUTPUT)util/pmu-flex.c: util/pmu.l + $(QUIET_FLEX)$(FLEX) --header-file=$(OUTPUT)util/pmu-flex.h -t util/pmu.l > $(OUTPUT)util/pmu-flex.c + +$(OUTPUT)util/pmu-bison.c: util/pmu.y + $(QUIET_BISON)$(BISON) -v util/pmu.y -d -o $(OUTPUT)util/pmu-bison.c + +$(OUTPUT)util/parse-events.o: $(OUTPUT)util/parse-events-flex.c $(OUTPUT)util/parse-events-bison.c +$(OUTPUT)util/pmu.o: $(OUTPUT)util/pmu-flex.c $(OUTPUT)util/pmu-bison.c + LIB_FILE=$(OUTPUT)libperf.a LIB_H += ../../include/linux/perf_event.h @@ -249,7 +251,7 @@ LIB_H += util/include/linux/const.h LIB_H += util/include/linux/ctype.h LIB_H += util/include/linux/kernel.h LIB_H += util/include/linux/list.h -LIB_H += util/include/linux/module.h +LIB_H += util/include/linux/export.h LIB_H += util/include/linux/poison.h LIB_H += util/include/linux/prefetch.h LIB_H += util/include/linux/rbtree.h @@ -276,6 +278,7 @@ LIB_H += util/build-id.h LIB_H += util/debug.h LIB_H += util/debugfs.h LIB_H += util/sysfs.h +LIB_H += util/pmu.h LIB_H += util/event.h LIB_H += util/evsel.h LIB_H += util/evlist.h @@ -314,6 +317,8 @@ LIB_H += util/cpumap.h LIB_H += util/top.h LIB_H += $(ARCH_INCLUDE) LIB_H += util/cgroup.h +LIB_H += $(TRACE_EVENT_DIR)event-parse.h +LIB_H += util/target.h LIB_OBJS += $(OUTPUT)util/abspath.o LIB_OBJS += $(OUTPUT)util/alias.o @@ -323,6 +328,7 @@ LIB_OBJS += $(OUTPUT)util/config.o LIB_OBJS += $(OUTPUT)util/ctype.o LIB_OBJS += $(OUTPUT)util/debugfs.o LIB_OBJS += $(OUTPUT)util/sysfs.o +LIB_OBJS += $(OUTPUT)util/pmu.o LIB_OBJS += $(OUTPUT)util/environment.o LIB_OBJS += $(OUTPUT)util/event.o LIB_OBJS += $(OUTPUT)util/evlist.o @@ -332,6 +338,7 @@ LIB_OBJS += $(OUTPUT)util/help.o LIB_OBJS += $(OUTPUT)util/levenshtein.o LIB_OBJS += $(OUTPUT)util/parse-options.o LIB_OBJS += $(OUTPUT)util/parse-events.o +LIB_OBJS += $(OUTPUT)util/parse-events-test.o LIB_OBJS += $(OUTPUT)util/path.o LIB_OBJS += $(OUTPUT)util/rbtree.o LIB_OBJS += $(OUTPUT)util/bitmap.o @@ -359,6 +366,10 @@ LIB_OBJS += $(OUTPUT)util/session.o LIB_OBJS += $(OUTPUT)util/thread.o LIB_OBJS += $(OUTPUT)util/thread_map.o LIB_OBJS += $(OUTPUT)util/trace-event-parse.o +LIB_OBJS += $(OUTPUT)util/parse-events-flex.o +LIB_OBJS += $(OUTPUT)util/parse-events-bison.o +LIB_OBJS += $(OUTPUT)util/pmu-flex.o +LIB_OBJS += $(OUTPUT)util/pmu-bison.o LIB_OBJS += $(OUTPUT)util/trace-event-read.o LIB_OBJS += $(OUTPUT)util/trace-event-info.o LIB_OBJS += $(OUTPUT)util/trace-event-scripting.o @@ -370,6 +381,7 @@ LIB_OBJS += $(OUTPUT)util/util.o LIB_OBJS += $(OUTPUT)util/xyarray.o LIB_OBJS += $(OUTPUT)util/cpumap.o LIB_OBJS += $(OUTPUT)util/cgroup.o +LIB_OBJS += $(OUTPUT)util/target.o BUILTIN_OBJS += $(OUTPUT)builtin-annotate.o @@ -405,7 +417,7 @@ BUILTIN_OBJS += $(OUTPUT)builtin-kvm.o BUILTIN_OBJS += $(OUTPUT)builtin-test.o BUILTIN_OBJS += $(OUTPUT)builtin-inject.o -PERFLIBS = $(LIB_FILE) +PERFLIBS = $(LIB_FILE) $(LIBTRACEEVENT) # Files needed for the python binding, perf.so # pyrf is just an internal name needed for all those wrappers. @@ -482,22 +494,42 @@ else # Fedora has /usr/include/slang/slang.h, but ubuntu /usr/include/slang.h BASIC_CFLAGS += -I/usr/include/slang EXTLIBS += -lnewt -lslang - LIB_OBJS += $(OUTPUT)util/ui/setup.o - LIB_OBJS += $(OUTPUT)util/ui/browser.o - LIB_OBJS += $(OUTPUT)util/ui/browsers/annotate.o - LIB_OBJS += $(OUTPUT)util/ui/browsers/hists.o - LIB_OBJS += $(OUTPUT)util/ui/browsers/map.o - LIB_OBJS += $(OUTPUT)util/ui/helpline.o - LIB_OBJS += $(OUTPUT)util/ui/progress.o - LIB_OBJS += $(OUTPUT)util/ui/util.o - LIB_H += util/ui/browser.h - LIB_H += util/ui/browsers/map.h - LIB_H += util/ui/helpline.h - LIB_H += util/ui/keysyms.h - LIB_H += util/ui/libslang.h - LIB_H += util/ui/progress.h - LIB_H += util/ui/util.h - LIB_H += util/ui/ui.h + LIB_OBJS += $(OUTPUT)ui/setup.o + LIB_OBJS += $(OUTPUT)ui/browser.o + LIB_OBJS += $(OUTPUT)ui/browsers/annotate.o + LIB_OBJS += $(OUTPUT)ui/browsers/hists.o + LIB_OBJS += $(OUTPUT)ui/browsers/map.o + LIB_OBJS += $(OUTPUT)ui/helpline.o + LIB_OBJS += $(OUTPUT)ui/progress.o + LIB_OBJS += $(OUTPUT)ui/util.o + LIB_OBJS += $(OUTPUT)ui/tui/setup.o + LIB_H += ui/browser.h + LIB_H += ui/browsers/map.h + LIB_H += ui/helpline.h + LIB_H += ui/keysyms.h + LIB_H += ui/libslang.h + LIB_H += ui/progress.h + LIB_H += ui/util.h + LIB_H += ui/ui.h + endif +endif + +ifdef NO_GTK2 + BASIC_CFLAGS += -DNO_GTK2_SUPPORT +else + FLAGS_GTK2=$(ALL_CFLAGS) $(ALL_LDFLAGS) $(EXTLIBS) $(shell pkg-config --libs --cflags gtk+-2.0) + ifneq ($(call try-cc,$(SOURCE_GTK2),$(FLAGS_GTK2)),y) + msg := $(warning GTK2 not found, disables GTK2 support. Please install gtk2-devel or libgtk2.0-dev); + BASIC_CFLAGS += -DNO_GTK2_SUPPORT + else + BASIC_CFLAGS += $(shell pkg-config --cflags gtk+-2.0) + EXTLIBS += $(shell pkg-config --libs gtk+-2.0) + LIB_OBJS += $(OUTPUT)ui/gtk/browser.o + LIB_OBJS += $(OUTPUT)ui/gtk/setup.o + # Make sure that it'd be included only once. + ifneq ($(findstring -DNO_NEWT_SUPPORT,$(BASIC_CFLAGS)),) + LIB_OBJS += $(OUTPUT)ui/setup.o + endif endif endif @@ -640,16 +672,6 @@ else endif endif -ifneq ($(findstring $(MAKEFLAGS),s),s) -ifndef V - QUIET_CC = @echo ' ' CC $@; - QUIET_AR = @echo ' ' AR $@; - QUIET_LINK = @echo ' ' LINK $@; - QUIET_MKDIR = @echo ' ' MKDIR $@; - QUIET_GEN = @echo ' ' GEN $@; -endif -endif - ifdef ASCIIDOC8 export ASCIIDOC8 endif @@ -727,12 +749,28 @@ $(OUTPUT)perf.o perf.spec \ $(SCRIPTS) \ : $(OUTPUT)PERF-VERSION-FILE +.SUFFIXES: +.SUFFIXES: .o .c .S .s + +# These two need to be here so that when O= is not used they take precedence +# over the general rule for .o + +$(OUTPUT)util/%-flex.o: $(OUTPUT)util/%-flex.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -Iutil/ -w $< + +$(OUTPUT)util/%-bison.o: $(OUTPUT)util/%-bison.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DYYENABLE_NLS=0 -DYYLTYPE_IS_TRIVIAL=0 -Iutil/ -w $< + $(OUTPUT)%.o: %.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $< +$(OUTPUT)%.i: %.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -E $(ALL_CFLAGS) $< $(OUTPUT)%.s: %.c $(OUTPUT)PERF-CFLAGS - $(QUIET_CC)$(CC) -S $(ALL_CFLAGS) $< + $(QUIET_CC)$(CC) -o $@ -S $(ALL_CFLAGS) $< $(OUTPUT)%.o: %.S $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $< +$(OUTPUT)%.s: %.S + $(QUIET_CC)$(CC) -o $@ -E $(ALL_CFLAGS) $< $(OUTPUT)util/exec_cmd.o: util/exec_cmd.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) \ @@ -744,16 +782,16 @@ $(OUTPUT)util/exec_cmd.o: util/exec_cmd.c $(OUTPUT)PERF-CFLAGS $(OUTPUT)util/config.o: util/config.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< -$(OUTPUT)util/ui/browser.o: util/ui/browser.c $(OUTPUT)PERF-CFLAGS +$(OUTPUT)ui/browser.o: ui/browser.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $< -$(OUTPUT)util/ui/browsers/annotate.o: util/ui/browsers/annotate.c $(OUTPUT)PERF-CFLAGS +$(OUTPUT)ui/browsers/annotate.o: ui/browsers/annotate.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $< -$(OUTPUT)util/ui/browsers/hists.o: util/ui/browsers/hists.c $(OUTPUT)PERF-CFLAGS +$(OUTPUT)ui/browsers/hists.o: ui/browsers/hists.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $< -$(OUTPUT)util/ui/browsers/map.o: util/ui/browsers/map.c $(OUTPUT)PERF-CFLAGS +$(OUTPUT)ui/browsers/map.o: ui/browsers/map.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DENABLE_SLFUTURE_CONST $< $(OUTPUT)util/rbtree.o: ../../lib/rbtree.c $(OUTPUT)PERF-CFLAGS @@ -788,6 +826,10 @@ $(sort $(dir $(DIRECTORY_DEPS))): $(LIB_FILE): $(LIB_OBJS) $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIB_OBJS) +# libtraceevent.a +$(LIBTRACEEVENT): + $(QUIET_SUBDIR0)$(TRACE_EVENT_DIR) $(QUIET_SUBDIR1) $(COMMAND_O) libtraceevent.a + help: @echo 'Perf make targets:' @echo ' doc - make *all* documentation (see below)' @@ -931,8 +973,9 @@ clean: $(RM) *.spec *.pyc *.pyo */*.pyc */*.pyo $(OUTPUT)common-cmds.h TAGS tags cscope* $(MAKE) -C Documentation/ clean $(RM) $(OUTPUT)PERF-VERSION-FILE $(OUTPUT)PERF-CFLAGS + $(RM) $(OUTPUT)util/*-{bison,flex}* $(python-clean) -.PHONY: all install clean strip +.PHONY: all install clean strip $(LIBTRACEEVENT) .PHONY: shell_compatibility_test please_set_SHELL_PATH_to_a_more_modern_shell .PHONY: .FORCE-PERF-VERSION-FILE TAGS tags cscope .FORCE-PERF-CFLAGS diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 806e0a286634..67522cf87405 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -215,7 +215,7 @@ static int __cmd_annotate(struct perf_annotate *ann) } if (total_nr_samples == 0) { - ui__warning("The %s file has no samples!\n", session->filename); + ui__error("The %s file has no samples!\n", session->filename); goto out_delete; } out_delete: diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c index 52480467e9ff..6b2bcfbde150 100644 --- a/tools/perf/builtin-buildid-list.c +++ b/tools/perf/builtin-buildid-list.c @@ -84,7 +84,11 @@ static int perf_session__list_build_ids(void) if (filename__fprintf_build_id(session->filename, stdout)) goto out; - if (with_hits) + /* + * in pipe-mode, the only way to get the buildids is to parse + * the record stream. Buildids are stored as RECORD_HEADER_BUILD_ID + */ + if (with_hits || session->fd_pipe) perf_session__process_events(session, &build_id__mark_dso_hit_ops); perf_session__fprintf_dsos_buildid(session, stdout, with_hits); diff --git a/tools/perf/builtin-diff.c b/tools/perf/builtin-diff.c index 4f19513d7dda..d29d350fb2b7 100644 --- a/tools/perf/builtin-diff.c +++ b/tools/perf/builtin-diff.c @@ -24,6 +24,11 @@ static char diff__default_sort_order[] = "dso,symbol"; static bool force; static bool show_displacement; +struct perf_diff { + struct perf_tool tool; + struct perf_session *session; +}; + static int hists__add_entry(struct hists *self, struct addr_location *al, u64 period) { @@ -32,12 +37,14 @@ static int hists__add_entry(struct hists *self, return -ENOMEM; } -static int diff__process_sample_event(struct perf_tool *tool __used, +static int diff__process_sample_event(struct perf_tool *tool, union perf_event *event, struct perf_sample *sample, struct perf_evsel *evsel __used, struct machine *machine) { + struct perf_diff *_diff = container_of(tool, struct perf_diff, tool); + struct perf_session *session = _diff->session; struct addr_location al; if (perf_event__preprocess_sample(event, machine, &al, sample, NULL) < 0) { @@ -49,24 +56,26 @@ static int diff__process_sample_event(struct perf_tool *tool __used, if (al.filtered || al.sym == NULL) return 0; - if (hists__add_entry(&evsel->hists, &al, sample->period)) { + if (hists__add_entry(&session->hists, &al, sample->period)) { pr_warning("problem incrementing symbol period, skipping event\n"); return -1; } - evsel->hists.stats.total_period += sample->period; + session->hists.stats.total_period += sample->period; return 0; } -static struct perf_tool perf_diff = { - .sample = diff__process_sample_event, - .mmap = perf_event__process_mmap, - .comm = perf_event__process_comm, - .exit = perf_event__process_task, - .fork = perf_event__process_task, - .lost = perf_event__process_lost, - .ordered_samples = true, - .ordering_requires_timestamps = true, +static struct perf_diff diff = { + .tool = { + .sample = diff__process_sample_event, + .mmap = perf_event__process_mmap, + .comm = perf_event__process_comm, + .exit = perf_event__process_task, + .fork = perf_event__process_task, + .lost = perf_event__process_lost, + .ordered_samples = true, + .ordering_requires_timestamps = true, + }, }; static void perf_session__insert_hist_entry_by_name(struct rb_root *root, @@ -107,12 +116,6 @@ static void hists__resort_entries(struct hists *self) self->entries = tmp; } -static void hists__set_positions(struct hists *self) -{ - hists__output_resort(self); - hists__resort_entries(self); -} - static struct hist_entry *hists__find_entry(struct hists *self, struct hist_entry *he) { @@ -146,30 +149,37 @@ static void hists__match(struct hists *older, struct hists *newer) static int __cmd_diff(void) { int ret, i; +#define older (session[0]) +#define newer (session[1]) struct perf_session *session[2]; - session[0] = perf_session__new(input_old, O_RDONLY, force, false, &perf_diff); - session[1] = perf_session__new(input_new, O_RDONLY, force, false, &perf_diff); + older = perf_session__new(input_old, O_RDONLY, force, false, + &diff.tool); + newer = perf_session__new(input_new, O_RDONLY, force, false, + &diff.tool); if (session[0] == NULL || session[1] == NULL) return -ENOMEM; for (i = 0; i < 2; ++i) { - ret = perf_session__process_events(session[i], &perf_diff); + diff.session = session[i]; + ret = perf_session__process_events(session[i], &diff.tool); if (ret) goto out_delete; + hists__output_resort(&session[i]->hists); } - hists__output_resort(&session[1]->hists); if (show_displacement) - hists__set_positions(&session[0]->hists); + hists__resort_entries(&older->hists); - hists__match(&session[0]->hists, &session[1]->hists); - hists__fprintf(&session[1]->hists, &session[0]->hists, + hists__match(&older->hists, &newer->hists); + hists__fprintf(&newer->hists, &older->hists, show_displacement, true, 0, 0, stdout); out_delete: for (i = 0; i < 2; ++i) perf_session__delete(session[i]); return ret; +#undef older +#undef newer } static const char * const diff_usage[] = { diff --git a/tools/perf/builtin-evlist.c b/tools/perf/builtin-evlist.c index 26760322c4f4..acd78dc28341 100644 --- a/tools/perf/builtin-evlist.c +++ b/tools/perf/builtin-evlist.c @@ -15,9 +15,40 @@ #include "util/parse-options.h" #include "util/session.h" -static const char *input_name; +struct perf_attr_details { + bool freq; + bool verbose; +}; + +static int comma_printf(bool *first, const char *fmt, ...) +{ + va_list args; + int ret = 0; + + if (!*first) { + ret += printf(","); + } else { + ret += printf(":"); + *first = false; + } + + va_start(args, fmt); + ret += vprintf(fmt, args); + va_end(args); + return ret; +} + +static int __if_print(bool *first, const char *field, u64 value) +{ + if (value == 0) + return 0; + + return comma_printf(first, " %s: %" PRIu64, field, value); +} + +#define if_print(field) __if_print(&first, #field, pos->attr.field) -static int __cmd_evlist(void) +static int __cmd_evlist(const char *input_name, struct perf_attr_details *details) { struct perf_session *session; struct perf_evsel *pos; @@ -26,8 +57,52 @@ static int __cmd_evlist(void) if (session == NULL) return -ENOMEM; - list_for_each_entry(pos, &session->evlist->entries, node) - printf("%s\n", event_name(pos)); + list_for_each_entry(pos, &session->evlist->entries, node) { + bool first = true; + + printf("%s", event_name(pos)); + + if (details->verbose || details->freq) { + comma_printf(&first, " sample_freq=%" PRIu64, + (u64)pos->attr.sample_freq); + } + + if (details->verbose) { + if_print(type); + if_print(config); + if_print(config1); + if_print(config2); + if_print(size); + if_print(sample_type); + if_print(read_format); + if_print(disabled); + if_print(inherit); + if_print(pinned); + if_print(exclusive); + if_print(exclude_user); + if_print(exclude_kernel); + if_print(exclude_hv); + if_print(exclude_idle); + if_print(mmap); + if_print(comm); + if_print(freq); + if_print(inherit_stat); + if_print(enable_on_exec); + if_print(task); + if_print(watermark); + if_print(precise_ip); + if_print(mmap_data); + if_print(sample_id_all); + if_print(exclude_host); + if_print(exclude_guest); + if_print(__reserved_1); + if_print(wakeup_events); + if_print(bp_type); + if_print(branch_sample_type); + } + + putchar('\n'); + } perf_session__delete(session); return 0; @@ -38,17 +113,23 @@ static const char * const evlist_usage[] = { NULL }; -static const struct option options[] = { - OPT_STRING('i', "input", &input_name, "file", - "input file name"), - OPT_END() -}; - int cmd_evlist(int argc, const char **argv, const char *prefix __used) { + struct perf_attr_details details = { .verbose = false, }; + const char *input_name = NULL; + const struct option options[] = { + OPT_STRING('i', "input", &input_name, "file", + "Input file name"), + OPT_BOOLEAN('F', "freq", &details.freq, + "Show the sample frequency"), + OPT_BOOLEAN('v', "verbose", &details.verbose, + "Show all event attr details"), + OPT_END() + }; + argc = parse_options(argc, argv, options, evlist_usage, 0); if (argc) usage_with_options(evlist_usage, options); - return __cmd_evlist(); + return __cmd_evlist(input_name, &details); } diff --git a/tools/perf/builtin-inject.c b/tools/perf/builtin-inject.c index 09c106193e65..3beab489afc5 100644 --- a/tools/perf/builtin-inject.c +++ b/tools/perf/builtin-inject.c @@ -60,6 +60,11 @@ static int perf_event__repipe_tracing_data_synth(union perf_event *event, static int perf_event__repipe_attr(union perf_event *event, struct perf_evlist **pevlist __used) { + int ret; + ret = perf_event__process_attr(event, pevlist); + if (ret) + return ret; + return perf_event__repipe_synth(NULL, event, NULL); } diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 39104c0beea3..547af48deb4f 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -192,7 +192,7 @@ static void insert_caller_stat(unsigned long call_site, } static void process_alloc_event(void *data, - struct event *event, + struct event_format *event, int cpu, u64 timestamp __used, struct thread *thread __used, @@ -253,7 +253,7 @@ static struct alloc_stat *search_alloc_stat(unsigned long ptr, } static void process_free_event(void *data, - struct event *event, + struct event_format *event, int cpu, u64 timestamp __used, struct thread *thread __used) @@ -281,7 +281,7 @@ static void process_free_event(void *data, static void process_raw_event(union perf_event *raw_event __used, void *data, int cpu, u64 timestamp, struct thread *thread) { - struct event *event; + struct event_format *event; int type; type = trace_parse_common_type(data); diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c index 12c814838993..fd53319de20d 100644 --- a/tools/perf/builtin-lock.c +++ b/tools/perf/builtin-lock.c @@ -356,25 +356,25 @@ struct trace_release_event { struct trace_lock_handler { void (*acquire_event)(struct trace_acquire_event *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*acquired_event)(struct trace_acquired_event *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*contended_event)(struct trace_contended_event *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*release_event)(struct trace_release_event *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); @@ -416,7 +416,7 @@ enum acquire_flags { static void report_lock_acquire_event(struct trace_acquire_event *acquire_event, - struct event *__event __used, + struct event_format *__event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -480,7 +480,7 @@ end: static void report_lock_acquired_event(struct trace_acquired_event *acquired_event, - struct event *__event __used, + struct event_format *__event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -536,7 +536,7 @@ end: static void report_lock_contended_event(struct trace_contended_event *contended_event, - struct event *__event __used, + struct event_format *__event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -583,7 +583,7 @@ end: static void report_lock_release_event(struct trace_release_event *release_event, - struct event *__event __used, + struct event_format *__event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -647,7 +647,7 @@ static struct trace_lock_handler *trace_handler; static void process_lock_acquire_event(void *data, - struct event *event __used, + struct event_format *event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -666,7 +666,7 @@ process_lock_acquire_event(void *data, static void process_lock_acquired_event(void *data, - struct event *event __used, + struct event_format *event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -684,7 +684,7 @@ process_lock_acquired_event(void *data, static void process_lock_contended_event(void *data, - struct event *event __used, + struct event_format *event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -702,7 +702,7 @@ process_lock_contended_event(void *data, static void process_lock_release_event(void *data, - struct event *event __used, + struct event_format *event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -721,7 +721,7 @@ process_lock_release_event(void *data, static void process_raw_event(void *data, int cpu, u64 timestamp, struct thread *thread) { - struct event *event; + struct event_format *event; int type; type = trace_parse_common_type(data); diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index 4935c09dd5b5..e215ae61b2ae 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -54,6 +54,7 @@ static struct { bool show_ext_vars; bool show_funcs; bool mod_events; + bool uprobes; int nevents; struct perf_probe_event events[MAX_PROBES]; struct strlist *dellist; @@ -75,6 +76,8 @@ static int parse_probe_event(const char *str) return -1; } + pev->uprobes = params.uprobes; + /* Parse a perf-probe command into event */ ret = parse_perf_probe_command(str, pev); pr_debug("%d arguments\n", pev->nargs); @@ -82,21 +85,58 @@ static int parse_probe_event(const char *str) return ret; } +static int set_target(const char *ptr) +{ + int found = 0; + const char *buf; + + /* + * The first argument after options can be an absolute path + * to an executable / library or kernel module. + * + * TODO: Support relative path, and $PATH, $LD_LIBRARY_PATH, + * short module name. + */ + if (!params.target && ptr && *ptr == '/') { + params.target = ptr; + found = 1; + buf = ptr + (strlen(ptr) - 3); + + if (strcmp(buf, ".ko")) + params.uprobes = true; + + } + + return found; +} + static int parse_probe_event_argv(int argc, const char **argv) { - int i, len, ret; + int i, len, ret, found_target; char *buf; + found_target = set_target(argv[0]); + if (found_target && argc == 1) + return 0; + /* Bind up rest arguments */ len = 0; - for (i = 0; i < argc; i++) + for (i = 0; i < argc; i++) { + if (i == 0 && found_target) + continue; + len += strlen(argv[i]) + 1; + } buf = zalloc(len + 1); if (buf == NULL) return -ENOMEM; len = 0; - for (i = 0; i < argc; i++) + for (i = 0; i < argc; i++) { + if (i == 0 && found_target) + continue; + len += sprintf(&buf[len], "%s ", argv[i]); + } params.mod_events = true; ret = parse_probe_event(buf); free(buf); @@ -125,6 +165,28 @@ static int opt_del_probe_event(const struct option *opt __used, return 0; } +static int opt_set_target(const struct option *opt, const char *str, + int unset __used) +{ + int ret = -ENOENT; + + if (str && !params.target) { + if (!strcmp(opt->long_name, "exec")) + params.uprobes = true; +#ifdef DWARF_SUPPORT + else if (!strcmp(opt->long_name, "module")) + params.uprobes = false; +#endif + else + return ret; + + params.target = str; + ret = 0; + } + + return ret; +} + #ifdef DWARF_SUPPORT static int opt_show_lines(const struct option *opt __used, const char *str, int unset __used) @@ -246,9 +308,9 @@ static const struct option options[] = { "file", "vmlinux pathname"), OPT_STRING('s', "source", &symbol_conf.source_prefix, "directory", "path to kernel source"), - OPT_STRING('m', "module", ¶ms.target, - "modname|path", - "target module name (for online) or path (for offline)"), + OPT_CALLBACK('m', "module", NULL, "modname|path", + "target module name (for online) or path (for offline)", + opt_set_target), #endif OPT__DRY_RUN(&probe_event_dry_run), OPT_INTEGER('\0', "max-probes", ¶ms.max_probe_points, @@ -260,6 +322,8 @@ static const struct option options[] = { "\t\t\t(default: \"" DEFAULT_VAR_FILTER "\" for --vars,\n" "\t\t\t \"" DEFAULT_FUNC_FILTER "\" for --funcs)", opt_set_filter), + OPT_CALLBACK('x', "exec", NULL, "executable|path", + "target executable name or path", opt_set_target), OPT_END() }; @@ -310,6 +374,10 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) pr_err(" Error: Don't use --list with --funcs.\n"); usage_with_options(probe_usage, options); } + if (params.uprobes) { + pr_warning(" Error: Don't use --list with --exec.\n"); + usage_with_options(probe_usage, options); + } ret = show_perf_probe_events(); if (ret < 0) pr_err(" Error: Failed to show event list. (%d)\n", @@ -333,8 +401,8 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) if (!params.filter) params.filter = strfilter__new(DEFAULT_FUNC_FILTER, NULL); - ret = show_available_funcs(params.target, - params.filter); + ret = show_available_funcs(params.target, params.filter, + params.uprobes); strfilter__delete(params.filter); if (ret < 0) pr_err(" Error: Failed to show functions." @@ -343,7 +411,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) } #ifdef DWARF_SUPPORT - if (params.show_lines) { + if (params.show_lines && !params.uprobes) { if (params.mod_events) { pr_err(" Error: Don't use --line with" " --add/--del.\n"); diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index be4e1eee782e..f95840d04e4c 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -44,7 +44,6 @@ struct perf_record { struct perf_evlist *evlist; struct perf_session *session; const char *progname; - const char *uid_str; int output; unsigned int page_size; int realtime_prio; @@ -218,7 +217,7 @@ try_again: if (err == EPERM || err == EACCES) { ui__error_paranoid(); exit(EXIT_FAILURE); - } else if (err == ENODEV && opts->cpu_list) { + } else if (err == ENODEV && opts->target.cpu_list) { die("No such device - did you specify" " an out-of-range profile CPU?\n"); } else if (err == EINVAL) { @@ -243,9 +242,13 @@ try_again: /* * If it's cycles then fall back to hrtimer * based cpu-clock-tick sw counter, which - * is always available even if no PMU support: + * is always available even if no PMU support. + * + * PPC returns ENXIO until 2.6.37 (behavior changed + * with commit b0a873e). */ - if (attr->type == PERF_TYPE_HARDWARE + if ((err == ENOENT || err == ENXIO) + && attr->type == PERF_TYPE_HARDWARE && attr->config == PERF_COUNT_HW_CPU_CYCLES) { if (verbose) @@ -253,11 +256,15 @@ try_again: "trying to fall back to cpu-clock-ticks\n"); attr->type = PERF_TYPE_SOFTWARE; attr->config = PERF_COUNT_SW_CPU_CLOCK; + if (pos->name) { + free(pos->name); + pos->name = NULL; + } goto try_again; } if (err == ENOENT) { - ui__warning("The %s event is not supported.\n", + ui__error("The %s event is not supported.\n", event_name(pos)); exit(EXIT_FAILURE); } @@ -389,7 +396,7 @@ static void perf_record__mmap_read_all(struct perf_record *rec) perf_record__mmap_read(rec, &rec->evlist->mmap[i]); } - if (perf_header__has_feat(&rec->session->header, HEADER_TRACE_INFO)) + if (perf_header__has_feat(&rec->session->header, HEADER_TRACING_DATA)) write_output(rec, &finished_round_event, sizeof(finished_round_event)); } @@ -471,7 +478,7 @@ static int __cmd_record(struct perf_record *rec, int argc, const char **argv) perf_header__clear_feat(&session->header, HEADER_BUILD_ID); if (!have_tracepoints(&evsel_list->entries)) - perf_header__clear_feat(&session->header, HEADER_TRACE_INFO); + perf_header__clear_feat(&session->header, HEADER_TRACING_DATA); if (!rec->opts.branch_stack) perf_header__clear_feat(&session->header, HEADER_BRANCH_STACK); @@ -578,7 +585,7 @@ static int __cmd_record(struct perf_record *rec, int argc, const char **argv) perf_session__process_machines(session, tool, perf_event__synthesize_guest_os); - if (!opts->system_wide) + if (!opts->target.system_wide) perf_event__synthesize_thread_map(tool, evsel_list->threads, process_synthesized_event, machine); @@ -746,7 +753,10 @@ static struct perf_record record = { .mmap_pages = UINT_MAX, .user_freq = UINT_MAX, .user_interval = ULLONG_MAX, - .freq = 1000, + .freq = 4000, + .target = { + .uses_mmap = true, + }, }, .write_mode = WRITE_FORCE, .file_new = true, @@ -765,9 +775,9 @@ const struct option record_options[] = { parse_events_option), OPT_CALLBACK(0, "filter", &record.evlist, "filter", "event filter", parse_filter), - OPT_STRING('p', "pid", &record.opts.target_pid, "pid", + OPT_STRING('p', "pid", &record.opts.target.pid, "pid", "record events on existing process id"), - OPT_STRING('t', "tid", &record.opts.target_tid, "tid", + OPT_STRING('t', "tid", &record.opts.target.tid, "tid", "record events on existing thread id"), OPT_INTEGER('r', "realtime", &record.realtime_prio, "collect data with this RT SCHED_FIFO priority"), @@ -775,11 +785,11 @@ const struct option record_options[] = { "collect data without buffering"), OPT_BOOLEAN('R', "raw-samples", &record.opts.raw_samples, "collect raw sample records from all opened counters"), - OPT_BOOLEAN('a', "all-cpus", &record.opts.system_wide, + OPT_BOOLEAN('a', "all-cpus", &record.opts.target.system_wide, "system-wide collection from all CPUs"), OPT_BOOLEAN('A', "append", &record.append_file, "append to the output file to do incremental profiling"), - OPT_STRING('C', "cpu", &record.opts.cpu_list, "cpu", + OPT_STRING('C', "cpu", &record.opts.target.cpu_list, "cpu", "list of cpus to monitor"), OPT_BOOLEAN('f', "force", &record.force, "overwrite existing data file (deprecated)"), @@ -813,7 +823,8 @@ const struct option record_options[] = { OPT_CALLBACK('G', "cgroup", &record.evlist, "name", "monitor event in cgroup name only", parse_cgroups), - OPT_STRING('u', "uid", &record.uid_str, "user", "user to profile"), + OPT_STRING('u', "uid", &record.opts.target.uid_str, "user", + "user to profile"), OPT_CALLBACK_NOOPT('b', "branch-any", &record.opts.branch_stack, "branch any", "sample any taken branches", @@ -831,6 +842,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) struct perf_evsel *pos; struct perf_evlist *evsel_list; struct perf_record *rec = &record; + char errbuf[BUFSIZ]; perf_header__set_cmdline(argc, argv); @@ -842,13 +854,12 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) argc = parse_options(argc, argv, record_options, record_usage, PARSE_OPT_STOP_AT_NON_OPTION); - if (!argc && !rec->opts.target_pid && !rec->opts.target_tid && - !rec->opts.system_wide && !rec->opts.cpu_list && !rec->uid_str) + if (!argc && perf_target__none(&rec->opts.target)) usage_with_options(record_usage, record_options); if (rec->force && rec->append_file) { - fprintf(stderr, "Can't overwrite and append at the same time." - " You need to choose between -f and -A"); + ui__error("Can't overwrite and append at the same time." + " You need to choose between -f and -A"); usage_with_options(record_usage, record_options); } else if (rec->append_file) { rec->write_mode = WRITE_APPEND; @@ -856,9 +867,9 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) rec->write_mode = WRITE_FORCE; } - if (nr_cgroups && !rec->opts.system_wide) { - fprintf(stderr, "cgroup monitoring only available in" - " system-wide mode\n"); + if (nr_cgroups && !rec->opts.target.system_wide) { + ui__error("cgroup monitoring only available in" + " system-wide mode\n"); usage_with_options(record_usage, record_options); } @@ -883,17 +894,25 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) goto out_symbol_exit; } - rec->opts.uid = parse_target_uid(rec->uid_str, rec->opts.target_tid, - rec->opts.target_pid); - if (rec->uid_str != NULL && rec->opts.uid == UINT_MAX - 1) - goto out_free_fd; + err = perf_target__validate(&rec->opts.target); + if (err) { + perf_target__strerror(&rec->opts.target, err, errbuf, BUFSIZ); + ui__warning("%s", errbuf); + } + + err = perf_target__parse_uid(&rec->opts.target); + if (err) { + int saved_errno = errno; - if (rec->opts.target_pid) - rec->opts.target_tid = rec->opts.target_pid; + perf_target__strerror(&rec->opts.target, err, errbuf, BUFSIZ); + ui__error("%s", errbuf); + + err = -saved_errno; + goto out_free_fd; + } - if (perf_evlist__create_maps(evsel_list, rec->opts.target_pid, - rec->opts.target_tid, rec->opts.uid, - rec->opts.cpu_list) < 0) + err = -ENOMEM; + if (perf_evlist__create_maps(evsel_list, &rec->opts.target) < 0) usage_with_options(record_usage, record_options); list_for_each_entry(pos, &evsel_list->entries, node) { @@ -914,7 +933,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) else if (rec->opts.freq) { rec->opts.default_interval = rec->opts.freq; } else { - fprintf(stderr, "frequency and count are zero, aborting\n"); + ui__error("frequency and count are zero, aborting\n"); err = -EINVAL; goto out_free_fd; } diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 8e91c6eba18a..25249f76329d 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -40,7 +40,7 @@ struct perf_report { struct perf_tool tool; struct perf_session *session; char const *input_name; - bool force, use_tui, use_stdio; + bool force, use_tui, use_gtk, use_stdio; bool hide_unresolved; bool dont_use_callchains; bool show_full_info; @@ -50,6 +50,7 @@ struct perf_report { const char *pretty_printing_style; symbol_filter_t annotate_init; const char *cpu_list; + const char *symbol_filter_str; DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS); }; @@ -151,7 +152,7 @@ static int perf_evsel__add_hist_entry(struct perf_evsel *evsel, if (symbol_conf.use_callchain) { err = callchain_append(he->callchain, - &evsel->hists.callchain_cursor, + &callchain_cursor, sample->period); if (err) return err; @@ -161,7 +162,7 @@ static int perf_evsel__add_hist_entry(struct perf_evsel *evsel, * so we don't allocated the extra space needed because the stdio * code will not use it. */ - if (al->sym != NULL && use_browser > 0) { + if (he->ms.sym != NULL && use_browser > 0) { struct annotation *notes = symbol__annotation(he->ms.sym); assert(evsel != NULL); @@ -250,13 +251,13 @@ static int perf_report__setup_sample_type(struct perf_report *rep) if (!(self->sample_type & PERF_SAMPLE_CALLCHAIN)) { if (sort__has_parent) { - ui__warning("Selected --sort parent, but no " + ui__error("Selected --sort parent, but no " "callchain data. Did you call " "'perf record' without -g?\n"); return -EINVAL; } if (symbol_conf.use_callchain) { - ui__warning("Selected -g but no callchain data. Did " + ui__error("Selected -g but no callchain data. Did " "you call 'perf record' without -g?\n"); return -1; } @@ -265,17 +266,15 @@ static int perf_report__setup_sample_type(struct perf_report *rep) !symbol_conf.use_callchain) { symbol_conf.use_callchain = true; if (callchain_register_param(&callchain_param) < 0) { - ui__warning("Can't register callchain " - "params.\n"); + ui__error("Can't register callchain params.\n"); return -EINVAL; } } if (sort__branch_mode == 1) { if (!(self->sample_type & PERF_SAMPLE_BRANCH_STACK)) { - fprintf(stderr, "selected -b but no branch data." - " Did you call perf record without" - " -b?\n"); + ui__error("Selected -b but no branch data. " + "Did you call perf record without -b?\n"); return -1; } } @@ -295,12 +294,15 @@ static size_t hists__fprintf_nr_sample_events(struct hists *self, { size_t ret; char unit; - unsigned long nr_events = self->stats.nr_events[PERF_RECORD_SAMPLE]; + unsigned long nr_samples = self->stats.nr_events[PERF_RECORD_SAMPLE]; + u64 nr_events = self->stats.total_period; - nr_events = convert_unit(nr_events, &unit); - ret = fprintf(fp, "# Events: %lu%c", nr_events, unit); + nr_samples = convert_unit(nr_samples, &unit); + ret = fprintf(fp, "# Samples: %lu%c", nr_samples, unit); if (evname != NULL) - ret += fprintf(fp, " %s", evname); + ret += fprintf(fp, " of event '%s'", evname); + + ret += fprintf(fp, "\n# Event count (approx.): %" PRIu64, nr_events); return ret + fprintf(fp, "\n#\n"); } @@ -373,16 +375,23 @@ static int __cmd_report(struct perf_report *rep) (kernel_map->dso->hit && (kernel_kmap->ref_reloc_sym == NULL || kernel_kmap->ref_reloc_sym->addr == 0))) { - const struct dso *kdso = kernel_map->dso; + const char *desc = + "As no suitable kallsyms nor vmlinux was found, kernel samples\n" + "can't be resolved."; + + if (kernel_map) { + const struct dso *kdso = kernel_map->dso; + if (!RB_EMPTY_ROOT(&kdso->symbols[MAP__FUNCTION])) { + desc = "If some relocation was applied (e.g. " + "kexec) symbols may be misresolved."; + } + } ui__warning( "Kernel address maps (/proc/{kallsyms,modules}) were restricted.\n\n" "Check /proc/sys/kernel/kptr_restrict before running 'perf record'.\n\n%s\n\n" "Samples in kernel modules can't be resolved as well.\n\n", - RB_EMPTY_ROOT(&kdso->symbols[MAP__FUNCTION]) ? -"As no suitable kallsyms nor vmlinux was found, kernel samples\n" -"can't be resolved." : -"If some relocation was applied (e.g. kexec) symbols may be misresolved."); + desc); } if (dump_trace) { @@ -400,19 +409,27 @@ static int __cmd_report(struct perf_report *rep) list_for_each_entry(pos, &session->evlist->entries, node) { struct hists *hists = &pos->hists; + if (pos->idx == 0) + hists->symbol_filter_str = rep->symbol_filter_str; + hists__collapse_resort(hists); hists__output_resort(hists); nr_samples += hists->stats.nr_events[PERF_RECORD_SAMPLE]; } if (nr_samples == 0) { - ui__warning("The %s file has no samples!\n", session->filename); + ui__error("The %s file has no samples!\n", session->filename); goto out_delete; } if (use_browser > 0) { - perf_evlist__tui_browse_hists(session->evlist, help, - NULL, NULL, 0); + if (use_browser == 1) { + perf_evlist__tui_browse_hists(session->evlist, help, + NULL, NULL, 0); + } else if (use_browser == 2) { + perf_evlist__gtk_browse_hists(session->evlist, help, + NULL, NULL, 0); + } } else perf_evlist__tty_browse_hists(session->evlist, rep, help); @@ -569,6 +586,7 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) OPT_STRING(0, "pretty", &report.pretty_printing_style, "key", "pretty printing style key: normal raw"), OPT_BOOLEAN(0, "tui", &report.use_tui, "Use the TUI interface"), + OPT_BOOLEAN(0, "gtk", &report.use_gtk, "Use the GTK2 interface"), OPT_BOOLEAN(0, "stdio", &report.use_stdio, "Use the stdio interface"), OPT_STRING('s', "sort", &sort_order, "key[,key2...]", @@ -591,6 +609,8 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) "only consider symbols in these comms"), OPT_STRING('S', "symbols", &symbol_conf.sym_list_str, "symbol[,symbol...]", "only consider these symbols"), + OPT_STRING(0, "symbol-filter", &report.symbol_filter_str, "filter", + "only show symbols that (partially) match with this filter"), OPT_STRING('w', "column-widths", &symbol_conf.col_width_list_str, "width[,width...]", "don't try to adjust column width, use these fixed values"), @@ -624,6 +644,8 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) use_browser = 0; else if (report.use_tui) use_browser = 1; + else if (report.use_gtk) + use_browser = 2; if (report.inverted_callchain) callchain_param.order = ORDER_CALLER; @@ -659,11 +681,10 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) } - if (strcmp(report.input_name, "-") != 0) { + if (strcmp(report.input_name, "-") != 0) setup_browser(true); - } else { + else use_browser = 0; - } /* * Only in the newt browser we are doing integrated annotation, @@ -709,11 +730,16 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) } else symbol_conf.exclude_other = false; - /* - * Any (unrecognized) arguments left? - */ - if (argc) - usage_with_options(report_usage, options); + if (argc) { + /* + * Special case: if there's an argument left then assume that + * it's a symbol filter: + */ + if (argc > 1) + usage_with_options(report_usage, options); + + report.symbol_filter_str = argv[0]; + } sort_entry__setup_elide(&sort_comm, symbol_conf.comm_list, "comm", stdout); diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c index fb8b5f83b4a0..b125e07eb399 100644 --- a/tools/perf/builtin-sched.c +++ b/tools/perf/builtin-sched.c @@ -17,6 +17,7 @@ #include "util/debug.h" #include <sys/prctl.h> +#include <sys/resource.h> #include <semaphore.h> #include <pthread.h> @@ -727,34 +728,34 @@ struct trace_migrate_task_event { struct trace_sched_handler { void (*switch_event)(struct trace_switch_event *, struct machine *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*runtime_event)(struct trace_runtime_event *, struct machine *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*wakeup_event)(struct trace_wakeup_event *, struct machine *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*fork_event)(struct trace_fork_event *, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); void (*migrate_task_event)(struct trace_migrate_task_event *, struct machine *machine, - struct event *, + struct event_format *, int cpu, u64 timestamp, struct thread *thread); @@ -764,7 +765,7 @@ struct trace_sched_handler { static void replay_wakeup_event(struct trace_wakeup_event *wakeup_event, struct machine *machine __used, - struct event *event, + struct event_format *event, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -791,7 +792,7 @@ static u64 cpu_last_switched[MAX_CPUS]; static void replay_switch_event(struct trace_switch_event *switch_event, struct machine *machine __used, - struct event *event, + struct event_format *event, int cpu, u64 timestamp, struct thread *thread __used) @@ -834,7 +835,7 @@ replay_switch_event(struct trace_switch_event *switch_event, static void replay_fork_event(struct trace_fork_event *fork_event, - struct event *event, + struct event_format *event, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -943,7 +944,7 @@ static void thread_atoms_insert(struct thread *thread) static void latency_fork_event(struct trace_fork_event *fork_event __used, - struct event *event __used, + struct event_format *event __used, int cpu __used, u64 timestamp __used, struct thread *thread __used) @@ -1025,7 +1026,7 @@ add_sched_in_event(struct work_atoms *atoms, u64 timestamp) static void latency_switch_event(struct trace_switch_event *switch_event, struct machine *machine, - struct event *event __used, + struct event_format *event __used, int cpu, u64 timestamp, struct thread *thread __used) @@ -1078,7 +1079,7 @@ latency_switch_event(struct trace_switch_event *switch_event, static void latency_runtime_event(struct trace_runtime_event *runtime_event, struct machine *machine, - struct event *event __used, + struct event_format *event __used, int cpu, u64 timestamp, struct thread *this_thread __used) @@ -1101,7 +1102,7 @@ latency_runtime_event(struct trace_runtime_event *runtime_event, static void latency_wakeup_event(struct trace_wakeup_event *wakeup_event, struct machine *machine, - struct event *__event __used, + struct event_format *__event __used, int cpu __used, u64 timestamp, struct thread *thread __used) @@ -1149,7 +1150,7 @@ latency_wakeup_event(struct trace_wakeup_event *wakeup_event, static void latency_migrate_task_event(struct trace_migrate_task_event *migrate_task_event, struct machine *machine, - struct event *__event __used, + struct event_format *__event __used, int cpu __used, u64 timestamp, struct thread *thread __used) @@ -1360,7 +1361,7 @@ static struct trace_sched_handler *trace_handler; static void process_sched_wakeup_event(struct perf_tool *tool __used, - struct event *event, + struct event_format *event, struct perf_sample *sample, struct machine *machine, struct thread *thread) @@ -1397,7 +1398,7 @@ static char next_shortname2 = '0'; static void map_switch_event(struct trace_switch_event *switch_event, struct machine *machine, - struct event *event __used, + struct event_format *event __used, int this_cpu, u64 timestamp, struct thread *thread __used) @@ -1475,7 +1476,7 @@ map_switch_event(struct trace_switch_event *switch_event, static void process_sched_switch_event(struct perf_tool *tool __used, - struct event *event, + struct event_format *event, struct perf_sample *sample, struct machine *machine, struct thread *thread) @@ -1511,7 +1512,7 @@ process_sched_switch_event(struct perf_tool *tool __used, static void process_sched_runtime_event(struct perf_tool *tool __used, - struct event *event, + struct event_format *event, struct perf_sample *sample, struct machine *machine, struct thread *thread) @@ -1531,7 +1532,7 @@ process_sched_runtime_event(struct perf_tool *tool __used, static void process_sched_fork_event(struct perf_tool *tool __used, - struct event *event, + struct event_format *event, struct perf_sample *sample, struct machine *machine __used, struct thread *thread) @@ -1553,7 +1554,7 @@ process_sched_fork_event(struct perf_tool *tool __used, static void process_sched_exit_event(struct perf_tool *tool __used, - struct event *event, + struct event_format *event, struct perf_sample *sample __used, struct machine *machine __used, struct thread *thread __used) @@ -1564,7 +1565,7 @@ process_sched_exit_event(struct perf_tool *tool __used, static void process_sched_migrate_task_event(struct perf_tool *tool __used, - struct event *event, + struct event_format *event, struct perf_sample *sample, struct machine *machine, struct thread *thread) @@ -1585,7 +1586,7 @@ process_sched_migrate_task_event(struct perf_tool *tool __used, sample->time, thread); } -typedef void (*tracepoint_handler)(struct perf_tool *tool, struct event *event, +typedef void (*tracepoint_handler)(struct perf_tool *tool, struct event_format *event, struct perf_sample *sample, struct machine *machine, struct thread *thread); diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index d4ce733b9eba..8e395a538eb9 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -261,7 +261,7 @@ static void print_sample_start(struct perf_sample *sample, struct perf_event_attr *attr) { int type; - struct event *event; + struct event_format *event; const char *evname = NULL; unsigned long secs; unsigned long usecs; diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index ea40e4e8b227..262589991ea4 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -173,24 +173,23 @@ static struct perf_event_attr very_very_detailed_attrs[] = { -struct perf_evlist *evsel_list; +static struct perf_evlist *evsel_list; -static bool system_wide = false; -static int run_idx = 0; +static struct perf_target target = { + .uid = UINT_MAX, +}; +static int run_idx = 0; static int run_count = 1; static bool no_inherit = false; static bool scale = true; static bool no_aggr = false; -static const char *target_pid; -static const char *target_tid; static pid_t child_pid = -1; static bool null_run = false; static int detailed_run = 0; static bool sync_run = false; static bool big_num = true; static int big_num_opt = -1; -static const char *cpu_list; static const char *csv_sep = NULL; static bool csv_output = false; static bool group = false; @@ -265,24 +264,26 @@ static double stddev_stats(struct stats *stats) return sqrt(variance_mean); } -struct stats runtime_nsecs_stats[MAX_NR_CPUS]; -struct stats runtime_cycles_stats[MAX_NR_CPUS]; -struct stats runtime_stalled_cycles_front_stats[MAX_NR_CPUS]; -struct stats runtime_stalled_cycles_back_stats[MAX_NR_CPUS]; -struct stats runtime_branches_stats[MAX_NR_CPUS]; -struct stats runtime_cacherefs_stats[MAX_NR_CPUS]; -struct stats runtime_l1_dcache_stats[MAX_NR_CPUS]; -struct stats runtime_l1_icache_stats[MAX_NR_CPUS]; -struct stats runtime_ll_cache_stats[MAX_NR_CPUS]; -struct stats runtime_itlb_cache_stats[MAX_NR_CPUS]; -struct stats runtime_dtlb_cache_stats[MAX_NR_CPUS]; -struct stats walltime_nsecs_stats; +static struct stats runtime_nsecs_stats[MAX_NR_CPUS]; +static struct stats runtime_cycles_stats[MAX_NR_CPUS]; +static struct stats runtime_stalled_cycles_front_stats[MAX_NR_CPUS]; +static struct stats runtime_stalled_cycles_back_stats[MAX_NR_CPUS]; +static struct stats runtime_branches_stats[MAX_NR_CPUS]; +static struct stats runtime_cacherefs_stats[MAX_NR_CPUS]; +static struct stats runtime_l1_dcache_stats[MAX_NR_CPUS]; +static struct stats runtime_l1_icache_stats[MAX_NR_CPUS]; +static struct stats runtime_ll_cache_stats[MAX_NR_CPUS]; +static struct stats runtime_itlb_cache_stats[MAX_NR_CPUS]; +static struct stats runtime_dtlb_cache_stats[MAX_NR_CPUS]; +static struct stats walltime_nsecs_stats; static int create_perf_stat_counter(struct perf_evsel *evsel, struct perf_evsel *first) { struct perf_event_attr *attr = &evsel->attr; struct xyarray *group_fd = NULL; + bool exclude_guest_missing = false; + int ret; if (group && evsel != first) group_fd = first->fd; @@ -293,16 +294,39 @@ static int create_perf_stat_counter(struct perf_evsel *evsel, attr->inherit = !no_inherit; - if (system_wide) - return perf_evsel__open_per_cpu(evsel, evsel_list->cpus, - group, group_fd); - if (!target_pid && !target_tid) { +retry: + if (exclude_guest_missing) + evsel->attr.exclude_guest = evsel->attr.exclude_host = 0; + + if (perf_target__has_cpu(&target)) { + ret = perf_evsel__open_per_cpu(evsel, evsel_list->cpus, + group, group_fd); + if (ret) + goto check_ret; + return 0; + } + + if (!perf_target__has_task(&target) && (!group || evsel == first)) { attr->disabled = 1; attr->enable_on_exec = 1; } - return perf_evsel__open_per_thread(evsel, evsel_list->threads, - group, group_fd); + ret = perf_evsel__open_per_thread(evsel, evsel_list->threads, + group, group_fd); + if (!ret) + return 0; + /* fall through */ +check_ret: + if (ret && errno == EINVAL) { + if (!exclude_guest_missing && + (evsel->attr.exclude_guest || evsel->attr.exclude_host)) { + pr_debug("Old kernel, cannot exclude " + "guest or host samples.\n"); + exclude_guest_missing = true; + goto retry; + } + } + return ret; } /* @@ -446,7 +470,7 @@ static int run_perf_stat(int argc __used, const char **argv) exit(-1); } - if (!target_tid && !target_pid && !system_wide) + if (perf_target__none(&target)) evsel_list->threads->map[0] = child_pid; /* @@ -463,8 +487,13 @@ static int run_perf_stat(int argc __used, const char **argv) list_for_each_entry(counter, &evsel_list->entries, node) { if (create_perf_stat_counter(counter, first) < 0) { + /* + * PPC returns ENXIO for HW counters until 2.6.37 + * (behavior changed with commit b0a873e). + */ if (errno == EINVAL || errno == ENOSYS || - errno == ENOENT || errno == EOPNOTSUPP) { + errno == ENOENT || errno == EOPNOTSUPP || + errno == ENXIO) { if (verbose) ui__warning("%s event is not supported by the kernel.\n", event_name(counter)); @@ -476,7 +505,7 @@ static int run_perf_stat(int argc __used, const char **argv) error("You may not have permission to collect %sstats.\n" "\t Consider tweaking" " /proc/sys/kernel/perf_event_paranoid or running as root.", - system_wide ? "system-wide " : ""); + target.system_wide ? "system-wide " : ""); } else { error("open_counter returned with %d (%s). " "/bin/dmesg may provide additional information.\n", @@ -968,14 +997,14 @@ static void print_stat(int argc, const char **argv) if (!csv_output) { fprintf(output, "\n"); fprintf(output, " Performance counter stats for "); - if (!target_pid && !target_tid) { + if (!perf_target__has_task(&target)) { fprintf(output, "\'%s", argv[0]); for (i = 1; i < argc; i++) fprintf(output, " %s", argv[i]); - } else if (target_pid) - fprintf(output, "process id \'%s", target_pid); + } else if (target.pid) + fprintf(output, "process id \'%s", target.pid); else - fprintf(output, "thread id \'%s", target_tid); + fprintf(output, "thread id \'%s", target.tid); fprintf(output, "\'"); if (run_count > 1) @@ -1049,11 +1078,11 @@ static const struct option options[] = { "event filter", parse_filter), OPT_BOOLEAN('i', "no-inherit", &no_inherit, "child tasks do not inherit counters"), - OPT_STRING('p', "pid", &target_pid, "pid", + OPT_STRING('p', "pid", &target.pid, "pid", "stat events on existing process id"), - OPT_STRING('t', "tid", &target_tid, "tid", + OPT_STRING('t', "tid", &target.tid, "tid", "stat events on existing thread id"), - OPT_BOOLEAN('a', "all-cpus", &system_wide, + OPT_BOOLEAN('a', "all-cpus", &target.system_wide, "system-wide collection from all CPUs"), OPT_BOOLEAN('g', "group", &group, "put the counters into a counter group"), @@ -1072,7 +1101,7 @@ static const struct option options[] = { OPT_CALLBACK_NOOPT('B', "big-num", NULL, NULL, "print large numbers with thousands\' separators", stat__set_big_num), - OPT_STRING('C', "cpu", &cpu_list, "cpu", + OPT_STRING('C', "cpu", &target.cpu_list, "cpu", "list of cpus to monitor in system-wide"), OPT_BOOLEAN('A', "no-aggr", &no_aggr, "disable CPU count aggregation"), @@ -1100,7 +1129,7 @@ static int add_default_attributes(void) return 0; if (!evsel_list->nr_entries) { - if (perf_evlist__add_attrs_array(evsel_list, default_attrs) < 0) + if (perf_evlist__add_default_attrs(evsel_list, default_attrs) < 0) return -1; } @@ -1110,21 +1139,21 @@ static int add_default_attributes(void) return 0; /* Append detailed run extra attributes: */ - if (perf_evlist__add_attrs_array(evsel_list, detailed_attrs) < 0) + if (perf_evlist__add_default_attrs(evsel_list, detailed_attrs) < 0) return -1; if (detailed_run < 2) return 0; /* Append very detailed run extra attributes: */ - if (perf_evlist__add_attrs_array(evsel_list, very_detailed_attrs) < 0) + if (perf_evlist__add_default_attrs(evsel_list, very_detailed_attrs) < 0) return -1; if (detailed_run < 3) return 0; /* Append very, very detailed run extra attributes: */ - return perf_evlist__add_attrs_array(evsel_list, very_very_detailed_attrs); + return perf_evlist__add_default_attrs(evsel_list, very_very_detailed_attrs); } int cmd_stat(int argc, const char **argv, const char *prefix __used) @@ -1190,13 +1219,13 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used) } else if (big_num_opt == 0) /* User passed --no-big-num */ big_num = false; - if (!argc && !target_pid && !target_tid) + if (!argc && !perf_target__has_task(&target)) usage_with_options(stat_usage, options); if (run_count <= 0) usage_with_options(stat_usage, options); /* no_aggr, cgroup are for system-wide only */ - if ((no_aggr || nr_cgroups) && !system_wide) { + if ((no_aggr || nr_cgroups) && !perf_target__has_cpu(&target)) { fprintf(stderr, "both cgroup and no-aggregation " "modes only available in system-wide mode\n"); @@ -1206,23 +1235,14 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used) if (add_default_attributes()) goto out; - if (target_pid) - target_tid = target_pid; + perf_target__validate(&target); - evsel_list->threads = thread_map__new_str(target_pid, - target_tid, UINT_MAX); - if (evsel_list->threads == NULL) { - pr_err("Problems finding threads of monitor\n"); - usage_with_options(stat_usage, options); - } - - if (system_wide) - evsel_list->cpus = cpu_map__new(cpu_list); - else - evsel_list->cpus = cpu_map__dummy_new(); + if (perf_evlist__create_maps(evsel_list, &target) < 0) { + if (perf_target__has_task(&target)) + pr_err("Problems finding threads of monitor\n"); + if (perf_target__has_cpu(&target)) + perror("failed to parse CPUs map"); - if (evsel_list->cpus == NULL) { - perror("failed to parse CPUs map"); usage_with_options(stat_usage, options); return -1; } diff --git a/tools/perf/builtin-test.c b/tools/perf/builtin-test.c index 3e087ce8daa6..5a8727c08757 100644 --- a/tools/perf/builtin-test.c +++ b/tools/perf/builtin-test.c @@ -13,6 +13,7 @@ #include "util/parse-events.h" #include "util/symbol.h" #include "util/thread_map.h" +#include "util/pmu.h" #include "../../include/linux/hw_breakpoint.h" #include <sys/mman.h> @@ -603,372 +604,6 @@ out_free_threads: #undef nsyscalls } -#define TEST_ASSERT_VAL(text, cond) \ -do { \ - if (!(cond)) { \ - pr_debug("FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ - return -1; \ - } \ -} while (0) - -static int test__checkevent_tracepoint(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_TRACEPOINT == evsel->attr.type); - TEST_ASSERT_VAL("wrong sample_type", - (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME | PERF_SAMPLE_CPU) == - evsel->attr.sample_type); - TEST_ASSERT_VAL("wrong sample_period", 1 == evsel->attr.sample_period); - return 0; -} - -static int test__checkevent_tracepoint_multi(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - TEST_ASSERT_VAL("wrong number of entries", evlist->nr_entries > 1); - - list_for_each_entry(evsel, &evlist->entries, node) { - TEST_ASSERT_VAL("wrong type", - PERF_TYPE_TRACEPOINT == evsel->attr.type); - TEST_ASSERT_VAL("wrong sample_type", - (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME | PERF_SAMPLE_CPU) - == evsel->attr.sample_type); - TEST_ASSERT_VAL("wrong sample_period", - 1 == evsel->attr.sample_period); - } - return 0; -} - -static int test__checkevent_raw(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", 1 == evsel->attr.config); - return 0; -} - -static int test__checkevent_numeric(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", 1 == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", 1 == evsel->attr.config); - return 0; -} - -static int test__checkevent_symbolic_name(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_HARDWARE == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", - PERF_COUNT_HW_INSTRUCTIONS == evsel->attr.config); - return 0; -} - -static int test__checkevent_symbolic_alias(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_SOFTWARE == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", - PERF_COUNT_SW_PAGE_FAULTS == evsel->attr.config); - return 0; -} - -static int test__checkevent_genhw(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_HW_CACHE == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", (1 << 16) == evsel->attr.config); - return 0; -} - -static int test__checkevent_breakpoint(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_BREAKPOINT == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); - TEST_ASSERT_VAL("wrong bp_type", (HW_BREAKPOINT_R | HW_BREAKPOINT_W) == - evsel->attr.bp_type); - TEST_ASSERT_VAL("wrong bp_len", HW_BREAKPOINT_LEN_4 == - evsel->attr.bp_len); - return 0; -} - -static int test__checkevent_breakpoint_x(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", PERF_TYPE_BREAKPOINT == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); - TEST_ASSERT_VAL("wrong bp_type", - HW_BREAKPOINT_X == evsel->attr.bp_type); - TEST_ASSERT_VAL("wrong bp_len", sizeof(long) == evsel->attr.bp_len); - return 0; -} - -static int test__checkevent_breakpoint_r(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", - PERF_TYPE_BREAKPOINT == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); - TEST_ASSERT_VAL("wrong bp_type", - HW_BREAKPOINT_R == evsel->attr.bp_type); - TEST_ASSERT_VAL("wrong bp_len", - HW_BREAKPOINT_LEN_4 == evsel->attr.bp_len); - return 0; -} - -static int test__checkevent_breakpoint_w(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); - TEST_ASSERT_VAL("wrong type", - PERF_TYPE_BREAKPOINT == evsel->attr.type); - TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); - TEST_ASSERT_VAL("wrong bp_type", - HW_BREAKPOINT_W == evsel->attr.bp_type); - TEST_ASSERT_VAL("wrong bp_len", - HW_BREAKPOINT_LEN_4 == evsel->attr.bp_len); - return 0; -} - -static int test__checkevent_tracepoint_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); - - return test__checkevent_tracepoint(evlist); -} - -static int -test__checkevent_tracepoint_multi_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - TEST_ASSERT_VAL("wrong number of entries", evlist->nr_entries > 1); - - list_for_each_entry(evsel, &evlist->entries, node) { - TEST_ASSERT_VAL("wrong exclude_user", - !evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", - evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); - } - - return test__checkevent_tracepoint_multi(evlist); -} - -static int test__checkevent_raw_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); - - return test__checkevent_raw(evlist); -} - -static int test__checkevent_numeric_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); - - return test__checkevent_numeric(evlist); -} - -static int test__checkevent_symbolic_name_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); - - return test__checkevent_symbolic_name(evlist); -} - -static int test__checkevent_symbolic_alias_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); - - return test__checkevent_symbolic_alias(evlist); -} - -static int test__checkevent_genhw_modifier(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel = list_entry(evlist->entries.next, - struct perf_evsel, node); - - TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); - TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); - TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); - TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); - - return test__checkevent_genhw(evlist); -} - -static struct test__event_st { - const char *name; - __u32 type; - int (*check)(struct perf_evlist *evlist); -} test__events[] = { - { - .name = "syscalls:sys_enter_open", - .check = test__checkevent_tracepoint, - }, - { - .name = "syscalls:*", - .check = test__checkevent_tracepoint_multi, - }, - { - .name = "r1", - .check = test__checkevent_raw, - }, - { - .name = "1:1", - .check = test__checkevent_numeric, - }, - { - .name = "instructions", - .check = test__checkevent_symbolic_name, - }, - { - .name = "faults", - .check = test__checkevent_symbolic_alias, - }, - { - .name = "L1-dcache-load-miss", - .check = test__checkevent_genhw, - }, - { - .name = "mem:0", - .check = test__checkevent_breakpoint, - }, - { - .name = "mem:0:x", - .check = test__checkevent_breakpoint_x, - }, - { - .name = "mem:0:r", - .check = test__checkevent_breakpoint_r, - }, - { - .name = "mem:0:w", - .check = test__checkevent_breakpoint_w, - }, - { - .name = "syscalls:sys_enter_open:k", - .check = test__checkevent_tracepoint_modifier, - }, - { - .name = "syscalls:*:u", - .check = test__checkevent_tracepoint_multi_modifier, - }, - { - .name = "r1:kp", - .check = test__checkevent_raw_modifier, - }, - { - .name = "1:1:hp", - .check = test__checkevent_numeric_modifier, - }, - { - .name = "instructions:h", - .check = test__checkevent_symbolic_name_modifier, - }, - { - .name = "faults:u", - .check = test__checkevent_symbolic_alias_modifier, - }, - { - .name = "L1-dcache-load-miss:kp", - .check = test__checkevent_genhw_modifier, - }, -}; - -#define TEST__EVENTS_CNT (sizeof(test__events) / sizeof(struct test__event_st)) - -static int test__parse_events(void) -{ - struct perf_evlist *evlist; - u_int i; - int ret = 0; - - for (i = 0; i < TEST__EVENTS_CNT; i++) { - struct test__event_st *e = &test__events[i]; - - evlist = perf_evlist__new(NULL, NULL); - if (evlist == NULL) - break; - - ret = parse_events(evlist, e->name, 0); - if (ret) { - pr_debug("failed to parse event '%s', err %d\n", - e->name, ret); - break; - } - - ret = e->check(evlist); - if (ret) - break; - - perf_evlist__delete(evlist); - } - - return ret; -} - static int sched__get_first_possible_cpu(pid_t pid, cpu_set_t **maskp, size_t *sizep) { @@ -1010,6 +645,10 @@ realloc: static int test__PERF_RECORD(void) { struct perf_record_opts opts = { + .target = { + .uid = UINT_MAX, + .uses_mmap = true, + }, .no_delay = true, .freq = 10, .mmap_pages = 256, @@ -1052,8 +691,7 @@ static int test__PERF_RECORD(void) * perf_evlist__prepare_workload we'll fill in the only thread * we're monitoring, the one forked there. */ - err = perf_evlist__create_maps(evlist, opts.target_pid, - opts.target_tid, UINT_MAX, opts.cpu_list); + err = perf_evlist__create_maps(evlist, &opts.target); if (err < 0) { pr_debug("Not enough memory to create thread/cpu maps\n"); goto out_delete_evlist; @@ -1394,8 +1032,6 @@ static int __test__rdpmc(void) sa.sa_sigaction = segfault_handler; sigaction(SIGSEGV, &sa, NULL); - fprintf(stderr, "\n\n"); - fd = sys_perf_event_open(&attr, 0, -1, -1, 0); if (fd < 0) { die("Error: sys_perf_event_open() syscall returned " @@ -1420,7 +1056,7 @@ static int __test__rdpmc(void) loops *= 10; delta = now - stamp; - fprintf(stderr, "%14d: %14Lu\n", n, (long long)delta); + pr_debug("%14d: %14Lu\n", n, (long long)delta); delta_sum += delta; } @@ -1428,7 +1064,7 @@ static int __test__rdpmc(void) munmap(addr, page_size); close(fd); - fprintf(stderr, " "); + pr_debug(" "); if (!delta_sum) return -1; @@ -1462,6 +1098,11 @@ static int test__rdpmc(void) #endif +static int test__perf_pmu(void) +{ + return perf_pmu__test(); +} + static struct test { const char *desc; int (*func)(void); @@ -1484,7 +1125,7 @@ static struct test { }, { .desc = "parse events tests", - .func = test__parse_events, + .func = parse_events__test, }, #if defined(__x86_64__) || defined(__i386__) { @@ -1497,6 +1138,10 @@ static struct test { .func = test__PERF_RECORD, }, { + .desc = "Test perf pmu format parsing", + .func = test__perf_pmu, + }, + { .func = NULL, }, }; diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index e3c63aef8efc..6bb0277b7dfe 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -42,6 +42,7 @@ #include "util/debug.h" #include <assert.h> +#include <elf.h> #include <fcntl.h> #include <stdio.h> @@ -59,6 +60,7 @@ #include <sys/prctl.h> #include <sys/wait.h> #include <sys/uio.h> +#include <sys/utsname.h> #include <sys/mman.h> #include <linux/unistd.h> @@ -162,12 +164,40 @@ static void __zero_source_counters(struct hist_entry *he) symbol__annotate_zero_histograms(sym); } +static void ui__warn_map_erange(struct map *map, struct symbol *sym, u64 ip) +{ + struct utsname uts; + int err = uname(&uts); + + ui__warning("Out of bounds address found:\n\n" + "Addr: %" PRIx64 "\n" + "DSO: %s %c\n" + "Map: %" PRIx64 "-%" PRIx64 "\n" + "Symbol: %" PRIx64 "-%" PRIx64 " %c %s\n" + "Arch: %s\n" + "Kernel: %s\n" + "Tools: %s\n\n" + "Not all samples will be on the annotation output.\n\n" + "Please report to linux-kernel@vger.kernel.org\n", + ip, map->dso->long_name, dso__symtab_origin(map->dso), + map->start, map->end, sym->start, sym->end, + sym->binding == STB_GLOBAL ? 'g' : + sym->binding == STB_LOCAL ? 'l' : 'w', sym->name, + err ? "[unknown]" : uts.machine, + err ? "[unknown]" : uts.release, perf_version_string); + if (use_browser <= 0) + sleep(5); + + map->erange_warned = true; +} + static void perf_top__record_precise_ip(struct perf_top *top, struct hist_entry *he, int counter, u64 ip) { struct annotation *notes; struct symbol *sym; + int err; if (he == NULL || he->ms.sym == NULL || ((top->sym_filter_entry == NULL || @@ -189,9 +219,12 @@ static void perf_top__record_precise_ip(struct perf_top *top, } ip = he->ms.map->map_ip(he->ms.map, ip); - symbol__inc_addr_samples(sym, he->ms.map, counter, ip); + err = symbol__inc_addr_samples(sym, he->ms.map, counter, ip); pthread_mutex_unlock(¬es->lock); + + if (err == -ERANGE && !he->ms.map->erange_warned) + ui__warn_map_erange(he->ms.map, sym, ip); } static void perf_top__show_details(struct perf_top *top) @@ -555,7 +588,7 @@ static void *display_thread_tui(void *arg) * via --uid. */ list_for_each_entry(pos, &top->evlist->entries, node) - pos->hists.uid_filter_str = top->uid_str; + pos->hists.uid_filter_str = top->target.uid_str; perf_evlist__tui_browse_hists(top->evlist, help, perf_top__sort_new_samples, @@ -615,6 +648,7 @@ process_hotkey: /* Tag samples to be skipped. */ static const char *skip_symbols[] = { + "intel_idle", "default_idle", "native_safe_halt", "cpu_idle", @@ -753,7 +787,7 @@ static void perf_event__process_sample(struct perf_tool *tool, } if (symbol_conf.use_callchain) { - err = callchain_append(he->callchain, &evsel->hists.callchain_cursor, + err = callchain_append(he->callchain, &callchain_cursor, sample->period); if (err) return; @@ -866,6 +900,9 @@ static void perf_top__start_counters(struct perf_top *top) attr->read_format |= PERF_FORMAT_ID; } + if (perf_target__has_cpu(&top->target)) + attr->sample_type |= PERF_SAMPLE_CPU; + if (symbol_conf.use_callchain) attr->sample_type |= PERF_SAMPLE_CALLCHAIN; @@ -914,20 +951,24 @@ try_again: attr->type = PERF_TYPE_SOFTWARE; attr->config = PERF_COUNT_SW_CPU_CLOCK; + if (counter->name) { + free(counter->name); + counter->name = NULL; + } goto try_again; } if (err == ENOENT) { - ui__warning("The %s event is not supported.\n", + ui__error("The %s event is not supported.\n", event_name(counter)); goto out_err; } else if (err == EMFILE) { - ui__warning("Too many events are opened.\n" + ui__error("Too many events are opened.\n" "Try again after reducing the number of events\n"); goto out_err; } - ui__warning("The sys_perf_event_open() syscall " + ui__error("The sys_perf_event_open() syscall " "returned with %d (%s). /bin/dmesg " "may provide additional information.\n" "No CONFIG_PERF_EVENTS=y kernel support " @@ -937,7 +978,7 @@ try_again: } if (perf_evlist__mmap(evlist, top->mmap_pages, false) < 0) { - ui__warning("Failed to mmap with %d (%s)\n", + ui__error("Failed to mmap with %d (%s)\n", errno, strerror(errno)); goto out_err; } @@ -953,12 +994,12 @@ static int perf_top__setup_sample_type(struct perf_top *top) { if (!top->sort_has_symbols) { if (symbol_conf.use_callchain) { - ui__warning("Selected -g but \"sym\" not present in --sort/-s."); + ui__error("Selected -g but \"sym\" not present in --sort/-s."); return -EINVAL; } } else if (!top->dont_use_callchains && callchain_param.mode != CHAIN_NONE) { if (callchain_register_param(&callchain_param) < 0) { - ui__warning("Can't register callchain params.\n"); + ui__error("Can't register callchain params.\n"); return -EINVAL; } } @@ -982,7 +1023,7 @@ static int __cmd_top(struct perf_top *top) if (ret) goto out_delete; - if (top->target_tid || top->uid != UINT_MAX) + if (perf_target__has_task(&top->target)) perf_event__synthesize_thread_map(&top->tool, top->evlist->threads, perf_event__process, &top->session->host_machine); @@ -1000,7 +1041,7 @@ static int __cmd_top(struct perf_top *top) if (pthread_create(&thread, NULL, (use_browser > 0 ? display_thread_tui : display_thread), top)) { - printf("Could not create display thread.\n"); + ui__error("Could not create display thread.\n"); exit(-1); } @@ -1009,7 +1050,7 @@ static int __cmd_top(struct perf_top *top) param.sched_priority = top->realtime_prio; if (sched_setscheduler(0, SCHED_FIFO, ¶m)) { - printf("Could not set realtime priority.\n"); + ui__error("Could not set realtime priority.\n"); exit(-1); } } @@ -1116,14 +1157,17 @@ static const char * const top_usage[] = { int cmd_top(int argc, const char **argv, const char *prefix __used) { struct perf_evsel *pos; - int status = -ENOMEM; + int status; + char errbuf[BUFSIZ]; struct perf_top top = { .count_filter = 5, .delay_secs = 2, - .uid = UINT_MAX, - .freq = 1000, /* 1 KHz */ + .freq = 4000, /* 4 KHz */ .mmap_pages = 128, .sym_pcnt_filter = 5, + .target = { + .uses_mmap = true, + }, }; char callchain_default_opt[] = "fractal,0.5,callee"; const struct option options[] = { @@ -1132,13 +1176,13 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) parse_events_option), OPT_INTEGER('c', "count", &top.default_interval, "event period to sample"), - OPT_STRING('p', "pid", &top.target_pid, "pid", + OPT_STRING('p', "pid", &top.target.pid, "pid", "profile events on existing process id"), - OPT_STRING('t', "tid", &top.target_tid, "tid", + OPT_STRING('t', "tid", &top.target.tid, "tid", "profile events on existing thread id"), - OPT_BOOLEAN('a', "all-cpus", &top.system_wide, + OPT_BOOLEAN('a', "all-cpus", &top.target.system_wide, "system-wide collection from all CPUs"), - OPT_STRING('C', "cpu", &top.cpu_list, "cpu", + OPT_STRING('C', "cpu", &top.target.cpu_list, "cpu", "list of cpus to monitor"), OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name, "file", "vmlinux pathname"), @@ -1193,7 +1237,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) "Display raw encoding of assembly instructions (default)"), OPT_STRING('M', "disassembler-style", &disassembler_style, "disassembler style", "Specify disassembler style (e.g. -M intel for intel syntax)"), - OPT_STRING('u', "uid", &top.uid_str, "user", "user to profile"), + OPT_STRING('u', "uid", &top.target.uid_str, "user", "user to profile"), OPT_END() }; @@ -1219,27 +1263,32 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) setup_browser(false); - top.uid = parse_target_uid(top.uid_str, top.target_tid, top.target_pid); - if (top.uid_str != NULL && top.uid == UINT_MAX - 1) - goto out_delete_evlist; + status = perf_target__validate(&top.target); + if (status) { + perf_target__strerror(&top.target, status, errbuf, BUFSIZ); + ui__warning("%s", errbuf); + } - /* CPU and PID are mutually exclusive */ - if (top.target_tid && top.cpu_list) { - printf("WARNING: PID switch overriding CPU\n"); - sleep(1); - top.cpu_list = NULL; + status = perf_target__parse_uid(&top.target); + if (status) { + int saved_errno = errno; + + perf_target__strerror(&top.target, status, errbuf, BUFSIZ); + ui__error("%s", errbuf); + + status = -saved_errno; + goto out_delete_evlist; } - if (top.target_pid) - top.target_tid = top.target_pid; + if (perf_target__none(&top.target)) + top.target.system_wide = true; - if (perf_evlist__create_maps(top.evlist, top.target_pid, - top.target_tid, top.uid, top.cpu_list) < 0) + if (perf_evlist__create_maps(top.evlist, &top.target) < 0) usage_with_options(top_usage, options); if (!top.evlist->nr_entries && perf_evlist__add_default(top.evlist) < 0) { - pr_err("Not enough memory for event selector list\n"); + ui__error("Not enough memory for event selector list\n"); return -ENOMEM; } @@ -1256,7 +1305,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) else if (top.freq) { top.default_interval = top.freq; } else { - fprintf(stderr, "frequency and count are zero, aborting\n"); + ui__error("frequency and count are zero, aborting\n"); exit(EXIT_FAILURE); } diff --git a/tools/perf/config/feature-tests.mak b/tools/perf/config/feature-tests.mak index 6170fd2531b5..d9084e03ce56 100644 --- a/tools/perf/config/feature-tests.mak +++ b/tools/perf/config/feature-tests.mak @@ -65,6 +65,21 @@ int main(void) endef endif +ifndef NO_GTK2 +define SOURCE_GTK2 +#pragma GCC diagnostic ignored \"-Wstrict-prototypes\" +#include <gtk/gtk.h> +#pragma GCC diagnostic error \"-Wstrict-prototypes\" + +int main(int argc, char *argv[]) +{ + gtk_init(&argc, &argv); + + return 0; +} +endef +endif + ifndef NO_LIBPERL define SOURCE_PERL_EMBED #include <EXTERN.h> diff --git a/tools/perf/design.txt b/tools/perf/design.txt index bd0bb1b1279b..67e5d0cace85 100644 --- a/tools/perf/design.txt +++ b/tools/perf/design.txt @@ -409,14 +409,15 @@ Counters can be enabled and disabled in two ways: via ioctl and via prctl. When a counter is disabled, it doesn't count or generate events but does continue to exist and maintain its count value. -An individual counter or counter group can be enabled with +An individual counter can be enabled with - ioctl(fd, PERF_EVENT_IOC_ENABLE); + ioctl(fd, PERF_EVENT_IOC_ENABLE, 0); or disabled with - ioctl(fd, PERF_EVENT_IOC_DISABLE); + ioctl(fd, PERF_EVENT_IOC_DISABLE, 0); +For a counter group, pass PERF_IOC_FLAG_GROUP as the third argument. Enabling or disabling the leader of a group enables or disables the whole group; that is, while the group leader is disabled, none of the counters in the group will count. Enabling or disabling a member of a diff --git a/tools/perf/perf-archive.sh b/tools/perf/perf-archive.sh index 677e59d62a8d..95b6f8b6177a 100644 --- a/tools/perf/perf-archive.sh +++ b/tools/perf/perf-archive.sh @@ -29,13 +29,14 @@ if [ ! -s $BUILDIDS ] ; then fi MANIFEST=$(mktemp /tmp/perf-archive-manifest.XXXXXX) +PERF_BUILDID_LINKDIR=$(readlink -f $PERF_BUILDID_DIR)/ cut -d ' ' -f 1 $BUILDIDS | \ while read build_id ; do linkname=$PERF_BUILDID_DIR.build-id/${build_id:0:2}/${build_id:2} filename=$(readlink -f $linkname) echo ${linkname#$PERF_BUILDID_DIR} >> $MANIFEST - echo ${filename#$PERF_BUILDID_DIR} >> $MANIFEST + echo ${filename#$PERF_BUILDID_LINKDIR} >> $MANIFEST done tar cfj $PERF_DATA.tar.bz2 -C $PERF_BUILDID_DIR -T $MANIFEST diff --git a/tools/perf/perf.h b/tools/perf/perf.h index 89e3355ab173..f960ccb2edc6 100644 --- a/tools/perf/perf.h +++ b/tools/perf/perf.h @@ -207,10 +207,10 @@ extern const char perf_version_string[]; void pthread__unblock_sigwinch(void); +#include "util/target.h" + struct perf_record_opts { - const char *target_pid; - const char *target_tid; - uid_t uid; + struct perf_target target; bool call_graph; bool group; bool inherit_stat; @@ -223,15 +223,13 @@ struct perf_record_opts { bool sample_time; bool sample_id_all_missing; bool exclude_guest_missing; - bool system_wide; bool period; unsigned int freq; unsigned int mmap_pages; unsigned int user_freq; - int branch_stack; + u64 branch_stack; u64 default_interval; u64 user_interval; - const char *cpu_list; }; #endif diff --git a/tools/perf/util/ui/browser.c b/tools/perf/ui/browser.c index 556829124b02..1818a531f1d3 100644 --- a/tools/perf/util/ui/browser.c +++ b/tools/perf/ui/browser.c @@ -27,21 +27,24 @@ static int ui_browser__percent_color(struct ui_browser *browser, return HE_COLORSET_NORMAL; } -void ui_browser__set_color(struct ui_browser *self __used, int color) +int ui_browser__set_color(struct ui_browser *browser, int color) { + int ret = browser->current_color; + browser->current_color = color; SLsmg_set_color(color); + return ret; } -void ui_browser__set_percent_color(struct ui_browser *self, +void ui_browser__set_percent_color(struct ui_browser *browser, double percent, bool current) { - int color = ui_browser__percent_color(self, percent, current); - ui_browser__set_color(self, color); + int color = ui_browser__percent_color(browser, percent, current); + ui_browser__set_color(browser, color); } -void ui_browser__gotorc(struct ui_browser *self, int y, int x) +void ui_browser__gotorc(struct ui_browser *browser, int y, int x) { - SLsmg_gotorc(self->y + y, self->x + x); + SLsmg_gotorc(browser->y + y, browser->x + x); } static struct list_head * @@ -70,23 +73,23 @@ ui_browser__list_head_filter_prev_entries(struct ui_browser *browser, return NULL; } -void ui_browser__list_head_seek(struct ui_browser *self, off_t offset, int whence) +void ui_browser__list_head_seek(struct ui_browser *browser, off_t offset, int whence) { - struct list_head *head = self->entries; + struct list_head *head = browser->entries; struct list_head *pos; - if (self->nr_entries == 0) + if (browser->nr_entries == 0) return; switch (whence) { case SEEK_SET: - pos = ui_browser__list_head_filter_entries(self, head->next); + pos = ui_browser__list_head_filter_entries(browser, head->next); break; case SEEK_CUR: - pos = self->top; + pos = browser->top; break; case SEEK_END: - pos = ui_browser__list_head_filter_prev_entries(self, head->prev); + pos = ui_browser__list_head_filter_prev_entries(browser, head->prev); break; default: return; @@ -96,18 +99,18 @@ void ui_browser__list_head_seek(struct ui_browser *self, off_t offset, int whenc if (offset > 0) { while (offset-- != 0) - pos = ui_browser__list_head_filter_entries(self, pos->next); + pos = ui_browser__list_head_filter_entries(browser, pos->next); } else { while (offset++ != 0) - pos = ui_browser__list_head_filter_prev_entries(self, pos->prev); + pos = ui_browser__list_head_filter_prev_entries(browser, pos->prev); } - self->top = pos; + browser->top = pos; } -void ui_browser__rb_tree_seek(struct ui_browser *self, off_t offset, int whence) +void ui_browser__rb_tree_seek(struct ui_browser *browser, off_t offset, int whence) { - struct rb_root *root = self->entries; + struct rb_root *root = browser->entries; struct rb_node *nd; switch (whence) { @@ -115,7 +118,7 @@ void ui_browser__rb_tree_seek(struct ui_browser *self, off_t offset, int whence) nd = rb_first(root); break; case SEEK_CUR: - nd = self->top; + nd = browser->top; break; case SEEK_END: nd = rb_last(root); @@ -132,23 +135,23 @@ void ui_browser__rb_tree_seek(struct ui_browser *self, off_t offset, int whence) nd = rb_prev(nd); } - self->top = nd; + browser->top = nd; } -unsigned int ui_browser__rb_tree_refresh(struct ui_browser *self) +unsigned int ui_browser__rb_tree_refresh(struct ui_browser *browser) { struct rb_node *nd; int row = 0; - if (self->top == NULL) - self->top = rb_first(self->entries); + if (browser->top == NULL) + browser->top = rb_first(browser->entries); - nd = self->top; + nd = browser->top; while (nd != NULL) { - ui_browser__gotorc(self, row, 0); - self->write(self, nd, row); - if (++row == self->height) + ui_browser__gotorc(browser, row, 0); + browser->write(browser, nd, row); + if (++row == browser->height) break; nd = rb_next(nd); } @@ -156,17 +159,17 @@ unsigned int ui_browser__rb_tree_refresh(struct ui_browser *self) return row; } -bool ui_browser__is_current_entry(struct ui_browser *self, unsigned row) +bool ui_browser__is_current_entry(struct ui_browser *browser, unsigned row) { - return self->top_idx + row == self->index; + return browser->top_idx + row == browser->index; } -void ui_browser__refresh_dimensions(struct ui_browser *self) +void ui_browser__refresh_dimensions(struct ui_browser *browser) { - self->width = SLtt_Screen_Cols - 1; - self->height = SLtt_Screen_Rows - 2; - self->y = 1; - self->x = 0; + browser->width = SLtt_Screen_Cols - 1; + browser->height = SLtt_Screen_Rows - 2; + browser->y = 1; + browser->x = 0; } void ui_browser__handle_resize(struct ui_browser *browser) @@ -222,10 +225,10 @@ bool ui_browser__dialog_yesno(struct ui_browser *browser, const char *text) return key == K_ENTER || toupper(key) == 'Y'; } -void ui_browser__reset_index(struct ui_browser *self) +void ui_browser__reset_index(struct ui_browser *browser) { - self->index = self->top_idx = 0; - self->seek(self, 0, SEEK_SET); + browser->index = browser->top_idx = 0; + browser->seek(browser, 0, SEEK_SET); } void __ui_browser__show_title(struct ui_browser *browser, const char *title) @@ -242,26 +245,26 @@ void ui_browser__show_title(struct ui_browser *browser, const char *title) pthread_mutex_unlock(&ui__lock); } -int ui_browser__show(struct ui_browser *self, const char *title, +int ui_browser__show(struct ui_browser *browser, const char *title, const char *helpline, ...) { int err; va_list ap; - ui_browser__refresh_dimensions(self); + ui_browser__refresh_dimensions(browser); pthread_mutex_lock(&ui__lock); - __ui_browser__show_title(self, title); + __ui_browser__show_title(browser, title); - self->title = title; - free(self->helpline); - self->helpline = NULL; + browser->title = title; + free(browser->helpline); + browser->helpline = NULL; va_start(ap, helpline); - err = vasprintf(&self->helpline, helpline, ap); + err = vasprintf(&browser->helpline, helpline, ap); va_end(ap); if (err > 0) - ui_helpline__push(self->helpline); + ui_helpline__push(browser->helpline); pthread_mutex_unlock(&ui__lock); return err ? 0 : -1; } @@ -347,7 +350,7 @@ void ui_browser__update_nr_entries(struct ui_browser *browser, u32 nr_entries) browser->seek(browser, browser->top_idx, SEEK_SET); } -int ui_browser__run(struct ui_browser *self, int delay_secs) +int ui_browser__run(struct ui_browser *browser, int delay_secs) { int err, key; @@ -355,7 +358,7 @@ int ui_browser__run(struct ui_browser *self, int delay_secs) off_t offset; pthread_mutex_lock(&ui__lock); - err = __ui_browser__refresh(self); + err = __ui_browser__refresh(browser); SLsmg_refresh(); pthread_mutex_unlock(&ui__lock); if (err < 0) @@ -365,18 +368,18 @@ int ui_browser__run(struct ui_browser *self, int delay_secs) if (key == K_RESIZE) { ui__refresh_dimensions(false); - ui_browser__refresh_dimensions(self); - __ui_browser__show_title(self, self->title); - ui_helpline__puts(self->helpline); + ui_browser__refresh_dimensions(browser); + __ui_browser__show_title(browser, browser->title); + ui_helpline__puts(browser->helpline); continue; } - if (self->use_navkeypressed && !self->navkeypressed) { + if (browser->use_navkeypressed && !browser->navkeypressed) { if (key == K_DOWN || key == K_UP || key == K_PGDN || key == K_PGUP || key == K_HOME || key == K_END || key == ' ') { - self->navkeypressed = true; + browser->navkeypressed = true; continue; } else return key; @@ -384,59 +387,59 @@ int ui_browser__run(struct ui_browser *self, int delay_secs) switch (key) { case K_DOWN: - if (self->index == self->nr_entries - 1) + if (browser->index == browser->nr_entries - 1) break; - ++self->index; - if (self->index == self->top_idx + self->height) { - ++self->top_idx; - self->seek(self, +1, SEEK_CUR); + ++browser->index; + if (browser->index == browser->top_idx + browser->height) { + ++browser->top_idx; + browser->seek(browser, +1, SEEK_CUR); } break; case K_UP: - if (self->index == 0) + if (browser->index == 0) break; - --self->index; - if (self->index < self->top_idx) { - --self->top_idx; - self->seek(self, -1, SEEK_CUR); + --browser->index; + if (browser->index < browser->top_idx) { + --browser->top_idx; + browser->seek(browser, -1, SEEK_CUR); } break; case K_PGDN: case ' ': - if (self->top_idx + self->height > self->nr_entries - 1) + if (browser->top_idx + browser->height > browser->nr_entries - 1) break; - offset = self->height; - if (self->index + offset > self->nr_entries - 1) - offset = self->nr_entries - 1 - self->index; - self->index += offset; - self->top_idx += offset; - self->seek(self, +offset, SEEK_CUR); + offset = browser->height; + if (browser->index + offset > browser->nr_entries - 1) + offset = browser->nr_entries - 1 - browser->index; + browser->index += offset; + browser->top_idx += offset; + browser->seek(browser, +offset, SEEK_CUR); break; case K_PGUP: - if (self->top_idx == 0) + if (browser->top_idx == 0) break; - if (self->top_idx < self->height) - offset = self->top_idx; + if (browser->top_idx < browser->height) + offset = browser->top_idx; else - offset = self->height; + offset = browser->height; - self->index -= offset; - self->top_idx -= offset; - self->seek(self, -offset, SEEK_CUR); + browser->index -= offset; + browser->top_idx -= offset; + browser->seek(browser, -offset, SEEK_CUR); break; case K_HOME: - ui_browser__reset_index(self); + ui_browser__reset_index(browser); break; case K_END: - offset = self->height - 1; - if (offset >= self->nr_entries) - offset = self->nr_entries - 1; + offset = browser->height - 1; + if (offset >= browser->nr_entries) + offset = browser->nr_entries - 1; - self->index = self->nr_entries - 1; - self->top_idx = self->index - offset; - self->seek(self, -offset, SEEK_END); + browser->index = browser->nr_entries - 1; + browser->top_idx = browser->index - offset; + browser->seek(browser, -offset, SEEK_END); break; default: return key; @@ -445,22 +448,22 @@ int ui_browser__run(struct ui_browser *self, int delay_secs) return -1; } -unsigned int ui_browser__list_head_refresh(struct ui_browser *self) +unsigned int ui_browser__list_head_refresh(struct ui_browser *browser) { struct list_head *pos; - struct list_head *head = self->entries; + struct list_head *head = browser->entries; int row = 0; - if (self->top == NULL || self->top == self->entries) - self->top = ui_browser__list_head_filter_entries(self, head->next); + if (browser->top == NULL || browser->top == browser->entries) + browser->top = ui_browser__list_head_filter_entries(browser, head->next); - pos = self->top; + pos = browser->top; list_for_each_from(pos, head) { - if (!self->filter || !self->filter(self, pos)) { - ui_browser__gotorc(self, row, 0); - self->write(self, pos, row); - if (++row == self->height) + if (!browser->filter || !browser->filter(browser, pos)) { + ui_browser__gotorc(browser, row, 0); + browser->write(browser, pos, row); + if (++row == browser->height) break; } } @@ -503,6 +506,12 @@ static struct ui_browser__colorset { .bg = "default", }, { + .colorset = HE_COLORSET_ADDR, + .name = "addr", + .fg = "magenta", + .bg = "default", + }, + { .name = NULL, } }; @@ -584,6 +593,111 @@ unsigned int ui_browser__argv_refresh(struct ui_browser *browser) return row; } +void __ui_browser__vline(struct ui_browser *browser, unsigned int column, + u16 start, u16 end) +{ + SLsmg_set_char_set(1); + ui_browser__gotorc(browser, start, column); + SLsmg_draw_vline(end - start + 1); + SLsmg_set_char_set(0); +} + +void ui_browser__write_graph(struct ui_browser *browser __used, int graph) +{ + SLsmg_set_char_set(1); + SLsmg_write_char(graph); + SLsmg_set_char_set(0); +} + +static void __ui_browser__line_arrow_up(struct ui_browser *browser, + unsigned int column, + u64 start, u64 end) +{ + unsigned int row, end_row; + + SLsmg_set_char_set(1); + + if (start < browser->top_idx + browser->height) { + row = start - browser->top_idx; + ui_browser__gotorc(browser, row, column); + SLsmg_write_char(SLSMG_LLCORN_CHAR); + ui_browser__gotorc(browser, row, column + 1); + SLsmg_draw_hline(2); + + if (row-- == 0) + goto out; + } else + row = browser->height - 1; + + if (end > browser->top_idx) + end_row = end - browser->top_idx; + else + end_row = 0; + + ui_browser__gotorc(browser, end_row, column); + SLsmg_draw_vline(row - end_row + 1); + + ui_browser__gotorc(browser, end_row, column); + if (end >= browser->top_idx) { + SLsmg_write_char(SLSMG_ULCORN_CHAR); + ui_browser__gotorc(browser, end_row, column + 1); + SLsmg_write_char(SLSMG_HLINE_CHAR); + ui_browser__gotorc(browser, end_row, column + 2); + SLsmg_write_char(SLSMG_RARROW_CHAR); + } +out: + SLsmg_set_char_set(0); +} + +static void __ui_browser__line_arrow_down(struct ui_browser *browser, + unsigned int column, + u64 start, u64 end) +{ + unsigned int row, end_row; + + SLsmg_set_char_set(1); + + if (start >= browser->top_idx) { + row = start - browser->top_idx; + ui_browser__gotorc(browser, row, column); + SLsmg_write_char(SLSMG_ULCORN_CHAR); + ui_browser__gotorc(browser, row, column + 1); + SLsmg_draw_hline(2); + + if (row++ == 0) + goto out; + } else + row = 0; + + if (end >= browser->top_idx + browser->height) + end_row = browser->height - 1; + else + end_row = end - browser->top_idx;; + + ui_browser__gotorc(browser, row, column); + SLsmg_draw_vline(end_row - row + 1); + + ui_browser__gotorc(browser, end_row, column); + if (end < browser->top_idx + browser->height) { + SLsmg_write_char(SLSMG_LLCORN_CHAR); + ui_browser__gotorc(browser, end_row, column + 1); + SLsmg_write_char(SLSMG_HLINE_CHAR); + ui_browser__gotorc(browser, end_row, column + 2); + SLsmg_write_char(SLSMG_RARROW_CHAR); + } +out: + SLsmg_set_char_set(0); +} + +void __ui_browser__line_arrow(struct ui_browser *browser, unsigned int column, + u64 start, u64 end) +{ + if (start > end) + __ui_browser__line_arrow_up(browser, column, start, end); + else + __ui_browser__line_arrow_down(browser, column, start, end); +} + void ui_browser__init(void) { int i = 0; @@ -594,4 +708,6 @@ void ui_browser__init(void) struct ui_browser__colorset *c = &ui_browser__colorsets[i++]; sltt_set_color(c->colorset, c->name, c->fg, c->bg); } + + annotate_browser__init(); } diff --git a/tools/perf/util/ui/browser.h b/tools/perf/ui/browser.h index 84d761b730c1..af70314605e5 100644 --- a/tools/perf/util/ui/browser.h +++ b/tools/perf/ui/browser.h @@ -10,11 +10,13 @@ #define HE_COLORSET_NORMAL 52 #define HE_COLORSET_SELECTED 53 #define HE_COLORSET_CODE 54 +#define HE_COLORSET_ADDR 55 struct ui_browser { u64 index, top_idx; void *top, *entries; u16 y, x, width, height; + int current_color; void *priv; const char *title; char *helpline; @@ -27,7 +29,7 @@ struct ui_browser { bool use_navkeypressed; }; -void ui_browser__set_color(struct ui_browser *self, int color); +int ui_browser__set_color(struct ui_browser *browser, int color); void ui_browser__set_percent_color(struct ui_browser *self, double percent, bool current); bool ui_browser__is_current_entry(struct ui_browser *self, unsigned row); @@ -35,6 +37,9 @@ void ui_browser__refresh_dimensions(struct ui_browser *self); void ui_browser__reset_index(struct ui_browser *self); void ui_browser__gotorc(struct ui_browser *self, int y, int x); +void ui_browser__write_graph(struct ui_browser *browser, int graph); +void __ui_browser__line_arrow(struct ui_browser *browser, unsigned int column, + u64 start, u64 end); void __ui_browser__show_title(struct ui_browser *browser, const char *title); void ui_browser__show_title(struct ui_browser *browser, const char *title); int ui_browser__show(struct ui_browser *self, const char *title, @@ -44,11 +49,15 @@ int ui_browser__refresh(struct ui_browser *self); int ui_browser__run(struct ui_browser *browser, int delay_secs); void ui_browser__update_nr_entries(struct ui_browser *browser, u32 nr_entries); void ui_browser__handle_resize(struct ui_browser *browser); +void __ui_browser__vline(struct ui_browser *browser, unsigned int column, + u16 start, u16 end); int ui_browser__warning(struct ui_browser *browser, int timeout, const char *format, ...); int ui_browser__help_window(struct ui_browser *browser, const char *text); bool ui_browser__dialog_yesno(struct ui_browser *browser, const char *text); +int ui_browser__input_window(const char *title, const char *text, char *input, + const char *exit_msg, int delay_sec); void ui_browser__argv_seek(struct ui_browser *browser, off_t offset, int whence); unsigned int ui_browser__argv_refresh(struct ui_browser *browser); @@ -60,4 +69,5 @@ void ui_browser__list_head_seek(struct ui_browser *self, off_t offset, int whenc unsigned int ui_browser__list_head_refresh(struct ui_browser *self); void ui_browser__init(void); +void annotate_browser__init(void); #endif /* _PERF_UI_BROWSER_H_ */ diff --git a/tools/perf/ui/browsers/annotate.c b/tools/perf/ui/browsers/annotate.c new file mode 100644 index 000000000000..34b1c46eaf42 --- /dev/null +++ b/tools/perf/ui/browsers/annotate.c @@ -0,0 +1,951 @@ +#include "../../util/util.h" +#include "../browser.h" +#include "../helpline.h" +#include "../libslang.h" +#include "../ui.h" +#include "../util.h" +#include "../../util/annotate.h" +#include "../../util/hist.h" +#include "../../util/sort.h" +#include "../../util/symbol.h" +#include <pthread.h> +#include <newt.h> + +struct browser_disasm_line { + struct rb_node rb_node; + double percent; + u32 idx; + int idx_asm; + int jump_sources; +}; + +static struct annotate_browser_opt { + bool hide_src_code, + use_offset, + jump_arrows, + show_nr_jumps; +} annotate_browser__opts = { + .use_offset = true, + .jump_arrows = true, +}; + +struct annotate_browser { + struct ui_browser b; + struct rb_root entries; + struct rb_node *curr_hot; + struct disasm_line *selection; + struct disasm_line **offsets; + u64 start; + int nr_asm_entries; + int nr_entries; + int max_jump_sources; + int nr_jumps; + bool searching_backwards; + u8 addr_width; + u8 jumps_width; + u8 target_width; + u8 min_addr_width; + u8 max_addr_width; + char search_bf[128]; +}; + +static inline struct browser_disasm_line *disasm_line__browser(struct disasm_line *dl) +{ + return (struct browser_disasm_line *)(dl + 1); +} + +static bool disasm_line__filter(struct ui_browser *browser __used, void *entry) +{ + if (annotate_browser__opts.hide_src_code) { + struct disasm_line *dl = list_entry(entry, struct disasm_line, node); + return dl->offset == -1; + } + + return false; +} + +static int annotate_browser__jumps_percent_color(struct annotate_browser *browser, + int nr, bool current) +{ + if (current && (!browser->b.use_navkeypressed || browser->b.navkeypressed)) + return HE_COLORSET_SELECTED; + if (nr == browser->max_jump_sources) + return HE_COLORSET_TOP; + if (nr > 1) + return HE_COLORSET_MEDIUM; + return HE_COLORSET_NORMAL; +} + +static int annotate_browser__set_jumps_percent_color(struct annotate_browser *browser, + int nr, bool current) +{ + int color = annotate_browser__jumps_percent_color(browser, nr, current); + return ui_browser__set_color(&browser->b, color); +} + +static void annotate_browser__write(struct ui_browser *browser, void *entry, int row) +{ + struct annotate_browser *ab = container_of(browser, struct annotate_browser, b); + struct disasm_line *dl = list_entry(entry, struct disasm_line, node); + struct browser_disasm_line *bdl = disasm_line__browser(dl); + bool current_entry = ui_browser__is_current_entry(browser, row); + bool change_color = (!annotate_browser__opts.hide_src_code && + (!current_entry || (browser->use_navkeypressed && + !browser->navkeypressed))); + int width = browser->width, printed; + char bf[256]; + + if (dl->offset != -1 && bdl->percent != 0.0) { + ui_browser__set_percent_color(browser, bdl->percent, current_entry); + slsmg_printf("%6.2f ", bdl->percent); + } else { + ui_browser__set_percent_color(browser, 0, current_entry); + slsmg_write_nstring(" ", 7); + } + + SLsmg_write_char(' '); + + /* The scroll bar isn't being used */ + if (!browser->navkeypressed) + width += 1; + + if (!*dl->line) + slsmg_write_nstring(" ", width - 7); + else if (dl->offset == -1) { + printed = scnprintf(bf, sizeof(bf), "%*s ", + ab->addr_width, " "); + slsmg_write_nstring(bf, printed); + slsmg_write_nstring(dl->line, width - printed - 6); + } else { + u64 addr = dl->offset; + int color = -1; + + if (!annotate_browser__opts.use_offset) + addr += ab->start; + + if (!annotate_browser__opts.use_offset) { + printed = scnprintf(bf, sizeof(bf), "%" PRIx64 ": ", addr); + } else { + if (bdl->jump_sources) { + if (annotate_browser__opts.show_nr_jumps) { + int prev; + printed = scnprintf(bf, sizeof(bf), "%*d ", + ab->jumps_width, + bdl->jump_sources); + prev = annotate_browser__set_jumps_percent_color(ab, bdl->jump_sources, + current_entry); + slsmg_write_nstring(bf, printed); + ui_browser__set_color(browser, prev); + } + + printed = scnprintf(bf, sizeof(bf), "%*" PRIx64 ": ", + ab->target_width, addr); + } else { + printed = scnprintf(bf, sizeof(bf), "%*s ", + ab->addr_width, " "); + } + } + + if (change_color) + color = ui_browser__set_color(browser, HE_COLORSET_ADDR); + slsmg_write_nstring(bf, printed); + if (change_color) + ui_browser__set_color(browser, color); + if (dl->ins && dl->ins->ops->scnprintf) { + if (ins__is_jump(dl->ins)) { + bool fwd = dl->ops.target.offset > (u64)dl->offset; + + ui_browser__write_graph(browser, fwd ? SLSMG_DARROW_CHAR : + SLSMG_UARROW_CHAR); + SLsmg_write_char(' '); + } else if (ins__is_call(dl->ins)) { + ui_browser__write_graph(browser, SLSMG_RARROW_CHAR); + SLsmg_write_char(' '); + } else { + slsmg_write_nstring(" ", 2); + } + } else { + if (strcmp(dl->name, "retq")) { + slsmg_write_nstring(" ", 2); + } else { + ui_browser__write_graph(browser, SLSMG_LARROW_CHAR); + SLsmg_write_char(' '); + } + } + + disasm_line__scnprintf(dl, bf, sizeof(bf), !annotate_browser__opts.use_offset); + slsmg_write_nstring(bf, width - 10 - printed); + } + + if (current_entry) + ab->selection = dl; +} + +static void annotate_browser__draw_current_jump(struct ui_browser *browser) +{ + struct annotate_browser *ab = container_of(browser, struct annotate_browser, b); + struct disasm_line *cursor = ab->selection, *target; + struct browser_disasm_line *btarget, *bcursor; + unsigned int from, to; + + if (!cursor || !cursor->ins || !ins__is_jump(cursor->ins) || + !disasm_line__has_offset(cursor)) + return; + + target = ab->offsets[cursor->ops.target.offset]; + if (!target) + return; + + bcursor = disasm_line__browser(cursor); + btarget = disasm_line__browser(target); + + if (annotate_browser__opts.hide_src_code) { + from = bcursor->idx_asm; + to = btarget->idx_asm; + } else { + from = (u64)bcursor->idx; + to = (u64)btarget->idx; + } + + ui_browser__set_color(browser, HE_COLORSET_CODE); + __ui_browser__line_arrow(browser, 9 + ab->addr_width, from, to); +} + +static unsigned int annotate_browser__refresh(struct ui_browser *browser) +{ + int ret = ui_browser__list_head_refresh(browser); + + if (annotate_browser__opts.jump_arrows) + annotate_browser__draw_current_jump(browser); + + ui_browser__set_color(browser, HE_COLORSET_NORMAL); + __ui_browser__vline(browser, 7, 0, browser->height - 1); + return ret; +} + +static double disasm_line__calc_percent(struct disasm_line *dl, struct symbol *sym, int evidx) +{ + double percent = 0.0; + + if (dl->offset != -1) { + int len = sym->end - sym->start; + unsigned int hits = 0; + struct annotation *notes = symbol__annotation(sym); + struct source_line *src_line = notes->src->lines; + struct sym_hist *h = annotation__histogram(notes, evidx); + s64 offset = dl->offset; + struct disasm_line *next; + + next = disasm__get_next_ip_line(¬es->src->source, dl); + while (offset < (s64)len && + (next == NULL || offset < next->offset)) { + if (src_line) { + percent += src_line[offset].percent; + } else + hits += h->addr[offset]; + + ++offset; + } + /* + * If the percentage wasn't already calculated in + * symbol__get_source_line, do it now: + */ + if (src_line == NULL && h->sum) + percent = 100.0 * hits / h->sum; + } + + return percent; +} + +static void disasm_rb_tree__insert(struct rb_root *root, struct browser_disasm_line *bdl) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + struct browser_disasm_line *l; + + while (*p != NULL) { + parent = *p; + l = rb_entry(parent, struct browser_disasm_line, rb_node); + if (bdl->percent < l->percent) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + rb_link_node(&bdl->rb_node, parent, p); + rb_insert_color(&bdl->rb_node, root); +} + +static void annotate_browser__set_top(struct annotate_browser *browser, + struct disasm_line *pos, u32 idx) +{ + unsigned back; + + ui_browser__refresh_dimensions(&browser->b); + back = browser->b.height / 2; + browser->b.top_idx = browser->b.index = idx; + + while (browser->b.top_idx != 0 && back != 0) { + pos = list_entry(pos->node.prev, struct disasm_line, node); + + if (disasm_line__filter(&browser->b, &pos->node)) + continue; + + --browser->b.top_idx; + --back; + } + + browser->b.top = pos; + browser->b.navkeypressed = true; +} + +static void annotate_browser__set_rb_top(struct annotate_browser *browser, + struct rb_node *nd) +{ + struct browser_disasm_line *bpos; + struct disasm_line *pos; + u32 idx; + + bpos = rb_entry(nd, struct browser_disasm_line, rb_node); + pos = ((struct disasm_line *)bpos) - 1; + idx = bpos->idx; + if (annotate_browser__opts.hide_src_code) + idx = bpos->idx_asm; + annotate_browser__set_top(browser, pos, idx); + browser->curr_hot = nd; +} + +static void annotate_browser__calc_percent(struct annotate_browser *browser, + int evidx) +{ + struct map_symbol *ms = browser->b.priv; + struct symbol *sym = ms->sym; + struct annotation *notes = symbol__annotation(sym); + struct disasm_line *pos; + + browser->entries = RB_ROOT; + + pthread_mutex_lock(¬es->lock); + + list_for_each_entry(pos, ¬es->src->source, node) { + struct browser_disasm_line *bpos = disasm_line__browser(pos); + bpos->percent = disasm_line__calc_percent(pos, sym, evidx); + if (bpos->percent < 0.01) { + RB_CLEAR_NODE(&bpos->rb_node); + continue; + } + disasm_rb_tree__insert(&browser->entries, bpos); + } + pthread_mutex_unlock(¬es->lock); + + browser->curr_hot = rb_last(&browser->entries); +} + +static bool annotate_browser__toggle_source(struct annotate_browser *browser) +{ + struct disasm_line *dl; + struct browser_disasm_line *bdl; + off_t offset = browser->b.index - browser->b.top_idx; + + browser->b.seek(&browser->b, offset, SEEK_CUR); + dl = list_entry(browser->b.top, struct disasm_line, node); + bdl = disasm_line__browser(dl); + + if (annotate_browser__opts.hide_src_code) { + if (bdl->idx_asm < offset) + offset = bdl->idx; + + browser->b.nr_entries = browser->nr_entries; + annotate_browser__opts.hide_src_code = false; + browser->b.seek(&browser->b, -offset, SEEK_CUR); + browser->b.top_idx = bdl->idx - offset; + browser->b.index = bdl->idx; + } else { + if (bdl->idx_asm < 0) { + ui_helpline__puts("Only available for assembly lines."); + browser->b.seek(&browser->b, -offset, SEEK_CUR); + return false; + } + + if (bdl->idx_asm < offset) + offset = bdl->idx_asm; + + browser->b.nr_entries = browser->nr_asm_entries; + annotate_browser__opts.hide_src_code = true; + browser->b.seek(&browser->b, -offset, SEEK_CUR); + browser->b.top_idx = bdl->idx_asm - offset; + browser->b.index = bdl->idx_asm; + } + + return true; +} + +static void annotate_browser__init_asm_mode(struct annotate_browser *browser) +{ + ui_browser__reset_index(&browser->b); + browser->b.nr_entries = browser->nr_asm_entries; +} + +static bool annotate_browser__callq(struct annotate_browser *browser, + int evidx, void (*timer)(void *arg), + void *arg, int delay_secs) +{ + struct map_symbol *ms = browser->b.priv; + struct disasm_line *dl = browser->selection; + struct symbol *sym = ms->sym; + struct annotation *notes; + struct symbol *target; + u64 ip; + + if (!ins__is_call(dl->ins)) + return false; + + ip = ms->map->map_ip(ms->map, dl->ops.target.addr); + target = map__find_symbol(ms->map, ip, NULL); + if (target == NULL) { + ui_helpline__puts("The called function was not found."); + return true; + } + + notes = symbol__annotation(target); + pthread_mutex_lock(¬es->lock); + + if (notes->src == NULL && symbol__alloc_hist(target) < 0) { + pthread_mutex_unlock(¬es->lock); + ui__warning("Not enough memory for annotating '%s' symbol!\n", + target->name); + return true; + } + + pthread_mutex_unlock(¬es->lock); + symbol__tui_annotate(target, ms->map, evidx, timer, arg, delay_secs); + ui_browser__show_title(&browser->b, sym->name); + return true; +} + +static +struct disasm_line *annotate_browser__find_offset(struct annotate_browser *browser, + s64 offset, s64 *idx) +{ + struct map_symbol *ms = browser->b.priv; + struct symbol *sym = ms->sym; + struct annotation *notes = symbol__annotation(sym); + struct disasm_line *pos; + + *idx = 0; + list_for_each_entry(pos, ¬es->src->source, node) { + if (pos->offset == offset) + return pos; + if (!disasm_line__filter(&browser->b, &pos->node)) + ++*idx; + } + + return NULL; +} + +static bool annotate_browser__jump(struct annotate_browser *browser) +{ + struct disasm_line *dl = browser->selection; + s64 idx; + + if (!ins__is_jump(dl->ins)) + return false; + + dl = annotate_browser__find_offset(browser, dl->ops.target.offset, &idx); + if (dl == NULL) { + ui_helpline__puts("Invallid jump offset"); + return true; + } + + annotate_browser__set_top(browser, dl, idx); + + return true; +} + +static +struct disasm_line *annotate_browser__find_string(struct annotate_browser *browser, + char *s, s64 *idx) +{ + struct map_symbol *ms = browser->b.priv; + struct symbol *sym = ms->sym; + struct annotation *notes = symbol__annotation(sym); + struct disasm_line *pos = browser->selection; + + *idx = browser->b.index; + list_for_each_entry_continue(pos, ¬es->src->source, node) { + if (disasm_line__filter(&browser->b, &pos->node)) + continue; + + ++*idx; + + if (pos->line && strstr(pos->line, s) != NULL) + return pos; + } + + return NULL; +} + +static bool __annotate_browser__search(struct annotate_browser *browser) +{ + struct disasm_line *dl; + s64 idx; + + dl = annotate_browser__find_string(browser, browser->search_bf, &idx); + if (dl == NULL) { + ui_helpline__puts("String not found!"); + return false; + } + + annotate_browser__set_top(browser, dl, idx); + browser->searching_backwards = false; + return true; +} + +static +struct disasm_line *annotate_browser__find_string_reverse(struct annotate_browser *browser, + char *s, s64 *idx) +{ + struct map_symbol *ms = browser->b.priv; + struct symbol *sym = ms->sym; + struct annotation *notes = symbol__annotation(sym); + struct disasm_line *pos = browser->selection; + + *idx = browser->b.index; + list_for_each_entry_continue_reverse(pos, ¬es->src->source, node) { + if (disasm_line__filter(&browser->b, &pos->node)) + continue; + + --*idx; + + if (pos->line && strstr(pos->line, s) != NULL) + return pos; + } + + return NULL; +} + +static bool __annotate_browser__search_reverse(struct annotate_browser *browser) +{ + struct disasm_line *dl; + s64 idx; + + dl = annotate_browser__find_string_reverse(browser, browser->search_bf, &idx); + if (dl == NULL) { + ui_helpline__puts("String not found!"); + return false; + } + + annotate_browser__set_top(browser, dl, idx); + browser->searching_backwards = true; + return true; +} + +static bool annotate_browser__search_window(struct annotate_browser *browser, + int delay_secs) +{ + if (ui_browser__input_window("Search", "String: ", browser->search_bf, + "ENTER: OK, ESC: Cancel", + delay_secs * 2) != K_ENTER || + !*browser->search_bf) + return false; + + return true; +} + +static bool annotate_browser__search(struct annotate_browser *browser, int delay_secs) +{ + if (annotate_browser__search_window(browser, delay_secs)) + return __annotate_browser__search(browser); + + return false; +} + +static bool annotate_browser__continue_search(struct annotate_browser *browser, + int delay_secs) +{ + if (!*browser->search_bf) + return annotate_browser__search(browser, delay_secs); + + return __annotate_browser__search(browser); +} + +static bool annotate_browser__search_reverse(struct annotate_browser *browser, + int delay_secs) +{ + if (annotate_browser__search_window(browser, delay_secs)) + return __annotate_browser__search_reverse(browser); + + return false; +} + +static +bool annotate_browser__continue_search_reverse(struct annotate_browser *browser, + int delay_secs) +{ + if (!*browser->search_bf) + return annotate_browser__search_reverse(browser, delay_secs); + + return __annotate_browser__search_reverse(browser); +} + +static void annotate_browser__update_addr_width(struct annotate_browser *browser) +{ + if (annotate_browser__opts.use_offset) + browser->target_width = browser->min_addr_width; + else + browser->target_width = browser->max_addr_width; + + browser->addr_width = browser->target_width; + + if (annotate_browser__opts.show_nr_jumps) + browser->addr_width += browser->jumps_width + 1; +} + +static int annotate_browser__run(struct annotate_browser *browser, int evidx, + void(*timer)(void *arg), + void *arg, int delay_secs) +{ + struct rb_node *nd = NULL; + struct map_symbol *ms = browser->b.priv; + struct symbol *sym = ms->sym; + const char *help = "Press 'h' for help on key bindings"; + int key; + + if (ui_browser__show(&browser->b, sym->name, help) < 0) + return -1; + + annotate_browser__calc_percent(browser, evidx); + + if (browser->curr_hot) { + annotate_browser__set_rb_top(browser, browser->curr_hot); + browser->b.navkeypressed = false; + } + + nd = browser->curr_hot; + + while (1) { + key = ui_browser__run(&browser->b, delay_secs); + + if (delay_secs != 0) { + annotate_browser__calc_percent(browser, evidx); + /* + * Current line focus got out of the list of most active + * lines, NULL it so that if TAB|UNTAB is pressed, we + * move to curr_hot (current hottest line). + */ + if (nd != NULL && RB_EMPTY_NODE(nd)) + nd = NULL; + } + + switch (key) { + case K_TIMER: + if (timer != NULL) + timer(arg); + + if (delay_secs != 0) + symbol__annotate_decay_histogram(sym, evidx); + continue; + case K_TAB: + if (nd != NULL) { + nd = rb_prev(nd); + if (nd == NULL) + nd = rb_last(&browser->entries); + } else + nd = browser->curr_hot; + break; + case K_UNTAB: + if (nd != NULL) + nd = rb_next(nd); + if (nd == NULL) + nd = rb_first(&browser->entries); + else + nd = browser->curr_hot; + break; + case K_F1: + case 'h': + ui_browser__help_window(&browser->b, + "UP/DOWN/PGUP\n" + "PGDN/SPACE Navigate\n" + "q/ESC/CTRL+C Exit\n\n" + "-> Go to target\n" + "<- Exit\n" + "H Cycle thru hottest instructions\n" + "j Toggle showing jump to target arrows\n" + "J Toggle showing number of jump sources on targets\n" + "n Search next string\n" + "o Toggle disassembler output/simplified view\n" + "s Toggle source code view\n" + "/ Search string\n" + "? Search previous string\n"); + continue; + case 'H': + nd = browser->curr_hot; + break; + case 's': + if (annotate_browser__toggle_source(browser)) + ui_helpline__puts(help); + continue; + case 'o': + annotate_browser__opts.use_offset = !annotate_browser__opts.use_offset; + annotate_browser__update_addr_width(browser); + continue; + case 'j': + annotate_browser__opts.jump_arrows = !annotate_browser__opts.jump_arrows; + continue; + case 'J': + annotate_browser__opts.show_nr_jumps = !annotate_browser__opts.show_nr_jumps; + annotate_browser__update_addr_width(browser); + continue; + case '/': + if (annotate_browser__search(browser, delay_secs)) { +show_help: + ui_helpline__puts(help); + } + continue; + case 'n': + if (browser->searching_backwards ? + annotate_browser__continue_search_reverse(browser, delay_secs) : + annotate_browser__continue_search(browser, delay_secs)) + goto show_help; + continue; + case '?': + if (annotate_browser__search_reverse(browser, delay_secs)) + goto show_help; + continue; + case 'D': { + static int seq; + ui_helpline__pop(); + ui_helpline__fpush("%d: nr_ent=%d, height=%d, idx=%d, top_idx=%d, nr_asm_entries=%d", + seq++, browser->b.nr_entries, + browser->b.height, + browser->b.index, + browser->b.top_idx, + browser->nr_asm_entries); + } + continue; + case K_ENTER: + case K_RIGHT: + if (browser->selection == NULL) + ui_helpline__puts("Huh? No selection. Report to linux-kernel@vger.kernel.org"); + else if (browser->selection->offset == -1) + ui_helpline__puts("Actions are only available for assembly lines."); + else if (!browser->selection->ins) { + if (strcmp(browser->selection->name, "retq")) + goto show_sup_ins; + goto out; + } else if (!(annotate_browser__jump(browser) || + annotate_browser__callq(browser, evidx, timer, arg, delay_secs))) { +show_sup_ins: + ui_helpline__puts("Actions are only available for 'callq', 'retq' & jump instructions."); + } + continue; + case K_LEFT: + case K_ESC: + case 'q': + case CTRL('c'): + goto out; + default: + continue; + } + + if (nd != NULL) + annotate_browser__set_rb_top(browser, nd); + } +out: + ui_browser__hide(&browser->b); + return key; +} + +int hist_entry__tui_annotate(struct hist_entry *he, int evidx, + void(*timer)(void *arg), void *arg, int delay_secs) +{ + return symbol__tui_annotate(he->ms.sym, he->ms.map, evidx, + timer, arg, delay_secs); +} + +static void annotate_browser__mark_jump_targets(struct annotate_browser *browser, + size_t size) +{ + u64 offset; + + for (offset = 0; offset < size; ++offset) { + struct disasm_line *dl = browser->offsets[offset], *dlt; + struct browser_disasm_line *bdlt; + + if (!dl || !dl->ins || !ins__is_jump(dl->ins) || + !disasm_line__has_offset(dl)) + continue; + + if (dl->ops.target.offset >= size) { + ui__error("jump to after symbol!\n" + "size: %zx, jump target: %" PRIx64, + size, dl->ops.target.offset); + continue; + } + + dlt = browser->offsets[dl->ops.target.offset]; + /* + * FIXME: Oops, no jump target? Buggy disassembler? Or do we + * have to adjust to the previous offset? + */ + if (dlt == NULL) + continue; + + bdlt = disasm_line__browser(dlt); + if (++bdlt->jump_sources > browser->max_jump_sources) + browser->max_jump_sources = bdlt->jump_sources; + + ++browser->nr_jumps; + } + +} + +static inline int width_jumps(int n) +{ + if (n >= 100) + return 5; + if (n / 10) + return 2; + return 1; +} + +int symbol__tui_annotate(struct symbol *sym, struct map *map, int evidx, + void(*timer)(void *arg), void *arg, + int delay_secs) +{ + struct disasm_line *pos, *n; + struct annotation *notes; + const size_t size = symbol__size(sym); + struct map_symbol ms = { + .map = map, + .sym = sym, + }; + struct annotate_browser browser = { + .b = { + .refresh = annotate_browser__refresh, + .seek = ui_browser__list_head_seek, + .write = annotate_browser__write, + .filter = disasm_line__filter, + .priv = &ms, + .use_navkeypressed = true, + }, + }; + int ret = -1; + + if (sym == NULL) + return -1; + + if (map->dso->annotate_warned) + return -1; + + browser.offsets = zalloc(size * sizeof(struct disasm_line *)); + if (browser.offsets == NULL) { + ui__error("Not enough memory!"); + return -1; + } + + if (symbol__annotate(sym, map, sizeof(struct browser_disasm_line)) < 0) { + ui__error("%s", ui_helpline__last_msg); + goto out_free_offsets; + } + + ui_helpline__push("Press <- or ESC to exit"); + + notes = symbol__annotation(sym); + browser.start = map__rip_2objdump(map, sym->start); + + list_for_each_entry(pos, ¬es->src->source, node) { + struct browser_disasm_line *bpos; + size_t line_len = strlen(pos->line); + + if (browser.b.width < line_len) + browser.b.width = line_len; + bpos = disasm_line__browser(pos); + bpos->idx = browser.nr_entries++; + if (pos->offset != -1) { + bpos->idx_asm = browser.nr_asm_entries++; + /* + * FIXME: short term bandaid to cope with assembly + * routines that comes with labels in the same column + * as the address in objdump, sigh. + * + * E.g. copy_user_generic_unrolled + */ + if (pos->offset < (s64)size) + browser.offsets[pos->offset] = pos; + } else + bpos->idx_asm = -1; + } + + annotate_browser__mark_jump_targets(&browser, size); + + browser.addr_width = browser.target_width = browser.min_addr_width = hex_width(size); + browser.max_addr_width = hex_width(sym->end); + browser.jumps_width = width_jumps(browser.max_jump_sources); + browser.b.nr_entries = browser.nr_entries; + browser.b.entries = ¬es->src->source, + browser.b.width += 18; /* Percentage */ + + if (annotate_browser__opts.hide_src_code) + annotate_browser__init_asm_mode(&browser); + + annotate_browser__update_addr_width(&browser); + + ret = annotate_browser__run(&browser, evidx, timer, arg, delay_secs); + list_for_each_entry_safe(pos, n, ¬es->src->source, node) { + list_del(&pos->node); + disasm_line__free(pos); + } + +out_free_offsets: + free(browser.offsets); + return ret; +} + +#define ANNOTATE_CFG(n) \ + { .name = #n, .value = &annotate_browser__opts.n, } + +/* + * Keep the entries sorted, they are bsearch'ed + */ +static struct annotate__config { + const char *name; + bool *value; +} annotate__configs[] = { + ANNOTATE_CFG(hide_src_code), + ANNOTATE_CFG(jump_arrows), + ANNOTATE_CFG(show_nr_jumps), + ANNOTATE_CFG(use_offset), +}; + +#undef ANNOTATE_CFG + +static int annotate_config__cmp(const void *name, const void *cfgp) +{ + const struct annotate__config *cfg = cfgp; + + return strcmp(name, cfg->name); +} + +static int annotate__config(const char *var, const char *value, void *data __used) +{ + struct annotate__config *cfg; + const char *name; + + if (prefixcmp(var, "annotate.") != 0) + return 0; + + name = var + 9; + cfg = bsearch(name, annotate__configs, ARRAY_SIZE(annotate__configs), + sizeof(struct annotate__config), annotate_config__cmp); + + if (cfg == NULL) + return -1; + + *cfg->value = perf_config_bool(name, value); + return 0; +} + +void annotate_browser__init(void) +{ + perf_config(annotate__config, NULL); +} diff --git a/tools/perf/util/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index fa530fcc764a..53f6697d014e 100644 --- a/tools/perf/util/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -5,12 +5,12 @@ #include <newt.h> #include <linux/rbtree.h> -#include "../../evsel.h" -#include "../../evlist.h" -#include "../../hist.h" -#include "../../pstack.h" -#include "../../sort.h" -#include "../../util.h" +#include "../../util/evsel.h" +#include "../../util/evlist.h" +#include "../../util/hist.h" +#include "../../util/pstack.h" +#include "../../util/sort.h" +#include "../../util/util.h" #include "../browser.h" #include "../helpline.h" @@ -26,21 +26,21 @@ struct hist_browser { bool has_symbols; }; -static int hists__browser_title(struct hists *self, char *bf, size_t size, +static int hists__browser_title(struct hists *hists, char *bf, size_t size, const char *ev_name); -static void hist_browser__refresh_dimensions(struct hist_browser *self) +static void hist_browser__refresh_dimensions(struct hist_browser *browser) { /* 3 == +/- toggle symbol before actual hist_entry rendering */ - self->b.width = 3 + (hists__sort_list_width(self->hists) + + browser->b.width = 3 + (hists__sort_list_width(browser->hists) + sizeof("[k]")); } -static void hist_browser__reset(struct hist_browser *self) +static void hist_browser__reset(struct hist_browser *browser) { - self->b.nr_entries = self->hists->nr_entries; - hist_browser__refresh_dimensions(self); - ui_browser__reset_index(&self->b); + browser->b.nr_entries = browser->hists->nr_entries; + hist_browser__refresh_dimensions(browser); + ui_browser__reset_index(&browser->b); } static char tree__folded_sign(bool unfolded) @@ -48,32 +48,32 @@ static char tree__folded_sign(bool unfolded) return unfolded ? '-' : '+'; } -static char map_symbol__folded(const struct map_symbol *self) +static char map_symbol__folded(const struct map_symbol *ms) { - return self->has_children ? tree__folded_sign(self->unfolded) : ' '; + return ms->has_children ? tree__folded_sign(ms->unfolded) : ' '; } -static char hist_entry__folded(const struct hist_entry *self) +static char hist_entry__folded(const struct hist_entry *he) { - return map_symbol__folded(&self->ms); + return map_symbol__folded(&he->ms); } -static char callchain_list__folded(const struct callchain_list *self) +static char callchain_list__folded(const struct callchain_list *cl) { - return map_symbol__folded(&self->ms); + return map_symbol__folded(&cl->ms); } -static void map_symbol__set_folding(struct map_symbol *self, bool unfold) +static void map_symbol__set_folding(struct map_symbol *ms, bool unfold) { - self->unfolded = unfold ? self->has_children : false; + ms->unfolded = unfold ? ms->has_children : false; } -static int callchain_node__count_rows_rb_tree(struct callchain_node *self) +static int callchain_node__count_rows_rb_tree(struct callchain_node *node) { int n = 0; struct rb_node *nd; - for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) { + for (nd = rb_first(&node->rb_root); nd; nd = rb_next(nd)) { struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node); struct callchain_list *chain; char folded_sign = ' '; /* No children */ @@ -123,20 +123,23 @@ static int callchain__count_rows(struct rb_root *chain) return n; } -static bool map_symbol__toggle_fold(struct map_symbol *self) +static bool map_symbol__toggle_fold(struct map_symbol *ms) { - if (!self->has_children) + if (!ms) return false; - self->unfolded = !self->unfolded; + if (!ms->has_children) + return false; + + ms->unfolded = !ms->unfolded; return true; } -static void callchain_node__init_have_children_rb_tree(struct callchain_node *self) +static void callchain_node__init_have_children_rb_tree(struct callchain_node *node) { - struct rb_node *nd = rb_first(&self->rb_root); + struct rb_node *nd = rb_first(&node->rb_root); - for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) { + for (nd = rb_first(&node->rb_root); nd; nd = rb_next(nd)) { struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node); struct callchain_list *chain; bool first = true; @@ -155,49 +158,49 @@ static void callchain_node__init_have_children_rb_tree(struct callchain_node *se } } -static void callchain_node__init_have_children(struct callchain_node *self) +static void callchain_node__init_have_children(struct callchain_node *node) { struct callchain_list *chain; - list_for_each_entry(chain, &self->val, list) - chain->ms.has_children = !RB_EMPTY_ROOT(&self->rb_root); + list_for_each_entry(chain, &node->val, list) + chain->ms.has_children = !RB_EMPTY_ROOT(&node->rb_root); - callchain_node__init_have_children_rb_tree(self); + callchain_node__init_have_children_rb_tree(node); } -static void callchain__init_have_children(struct rb_root *self) +static void callchain__init_have_children(struct rb_root *root) { struct rb_node *nd; - for (nd = rb_first(self); nd; nd = rb_next(nd)) { + for (nd = rb_first(root); nd; nd = rb_next(nd)) { struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node); callchain_node__init_have_children(node); } } -static void hist_entry__init_have_children(struct hist_entry *self) +static void hist_entry__init_have_children(struct hist_entry *he) { - if (!self->init_have_children) { - self->ms.has_children = !RB_EMPTY_ROOT(&self->sorted_chain); - callchain__init_have_children(&self->sorted_chain); - self->init_have_children = true; + if (!he->init_have_children) { + he->ms.has_children = !RB_EMPTY_ROOT(&he->sorted_chain); + callchain__init_have_children(&he->sorted_chain); + he->init_have_children = true; } } -static bool hist_browser__toggle_fold(struct hist_browser *self) +static bool hist_browser__toggle_fold(struct hist_browser *browser) { - if (map_symbol__toggle_fold(self->selection)) { - struct hist_entry *he = self->he_selection; + if (map_symbol__toggle_fold(browser->selection)) { + struct hist_entry *he = browser->he_selection; hist_entry__init_have_children(he); - self->hists->nr_entries -= he->nr_rows; + browser->hists->nr_entries -= he->nr_rows; if (he->ms.unfolded) he->nr_rows = callchain__count_rows(&he->sorted_chain); else he->nr_rows = 0; - self->hists->nr_entries += he->nr_rows; - self->b.nr_entries = self->hists->nr_entries; + browser->hists->nr_entries += he->nr_rows; + browser->b.nr_entries = browser->hists->nr_entries; return true; } @@ -206,12 +209,12 @@ static bool hist_browser__toggle_fold(struct hist_browser *self) return false; } -static int callchain_node__set_folding_rb_tree(struct callchain_node *self, bool unfold) +static int callchain_node__set_folding_rb_tree(struct callchain_node *node, bool unfold) { int n = 0; struct rb_node *nd; - for (nd = rb_first(&self->rb_root); nd; nd = rb_next(nd)) { + for (nd = rb_first(&node->rb_root); nd; nd = rb_next(nd)) { struct callchain_node *child = rb_entry(nd, struct callchain_node, rb_node); struct callchain_list *chain; bool has_children = false; @@ -260,37 +263,37 @@ static int callchain__set_folding(struct rb_root *chain, bool unfold) return n; } -static void hist_entry__set_folding(struct hist_entry *self, bool unfold) +static void hist_entry__set_folding(struct hist_entry *he, bool unfold) { - hist_entry__init_have_children(self); - map_symbol__set_folding(&self->ms, unfold); + hist_entry__init_have_children(he); + map_symbol__set_folding(&he->ms, unfold); - if (self->ms.has_children) { - int n = callchain__set_folding(&self->sorted_chain, unfold); - self->nr_rows = unfold ? n : 0; + if (he->ms.has_children) { + int n = callchain__set_folding(&he->sorted_chain, unfold); + he->nr_rows = unfold ? n : 0; } else - self->nr_rows = 0; + he->nr_rows = 0; } -static void hists__set_folding(struct hists *self, bool unfold) +static void hists__set_folding(struct hists *hists, bool unfold) { struct rb_node *nd; - self->nr_entries = 0; + hists->nr_entries = 0; - for (nd = rb_first(&self->entries); nd; nd = rb_next(nd)) { + for (nd = rb_first(&hists->entries); nd; nd = rb_next(nd)) { struct hist_entry *he = rb_entry(nd, struct hist_entry, rb_node); hist_entry__set_folding(he, unfold); - self->nr_entries += 1 + he->nr_rows; + hists->nr_entries += 1 + he->nr_rows; } } -static void hist_browser__set_folding(struct hist_browser *self, bool unfold) +static void hist_browser__set_folding(struct hist_browser *browser, bool unfold) { - hists__set_folding(self->hists, unfold); - self->b.nr_entries = self->hists->nr_entries; + hists__set_folding(browser->hists, unfold); + browser->b.nr_entries = browser->hists->nr_entries; /* Go to the start, we may be way after valid entries after a collapse */ - ui_browser__reset_index(&self->b); + ui_browser__reset_index(&browser->b); } static void ui_browser__warn_lost_events(struct ui_browser *browser) @@ -302,64 +305,64 @@ static void ui_browser__warn_lost_events(struct ui_browser *browser) "Or reduce the sampling frequency."); } -static int hist_browser__run(struct hist_browser *self, const char *ev_name, +static int hist_browser__run(struct hist_browser *browser, const char *ev_name, void(*timer)(void *arg), void *arg, int delay_secs) { int key; char title[160]; - self->b.entries = &self->hists->entries; - self->b.nr_entries = self->hists->nr_entries; + browser->b.entries = &browser->hists->entries; + browser->b.nr_entries = browser->hists->nr_entries; - hist_browser__refresh_dimensions(self); - hists__browser_title(self->hists, title, sizeof(title), ev_name); + hist_browser__refresh_dimensions(browser); + hists__browser_title(browser->hists, title, sizeof(title), ev_name); - if (ui_browser__show(&self->b, title, + if (ui_browser__show(&browser->b, title, "Press '?' for help on key bindings") < 0) return -1; while (1) { - key = ui_browser__run(&self->b, delay_secs); + key = ui_browser__run(&browser->b, delay_secs); switch (key) { case K_TIMER: timer(arg); - ui_browser__update_nr_entries(&self->b, self->hists->nr_entries); + ui_browser__update_nr_entries(&browser->b, browser->hists->nr_entries); - if (self->hists->stats.nr_lost_warned != - self->hists->stats.nr_events[PERF_RECORD_LOST]) { - self->hists->stats.nr_lost_warned = - self->hists->stats.nr_events[PERF_RECORD_LOST]; - ui_browser__warn_lost_events(&self->b); + if (browser->hists->stats.nr_lost_warned != + browser->hists->stats.nr_events[PERF_RECORD_LOST]) { + browser->hists->stats.nr_lost_warned = + browser->hists->stats.nr_events[PERF_RECORD_LOST]; + ui_browser__warn_lost_events(&browser->b); } - hists__browser_title(self->hists, title, sizeof(title), ev_name); - ui_browser__show_title(&self->b, title); + hists__browser_title(browser->hists, title, sizeof(title), ev_name); + ui_browser__show_title(&browser->b, title); continue; case 'D': { /* Debug */ static int seq; - struct hist_entry *h = rb_entry(self->b.top, + struct hist_entry *h = rb_entry(browser->b.top, struct hist_entry, rb_node); ui_helpline__pop(); ui_helpline__fpush("%d: nr_ent=(%d,%d), height=%d, idx=%d, fve: idx=%d, row_off=%d, nrows=%d", - seq++, self->b.nr_entries, - self->hists->nr_entries, - self->b.height, - self->b.index, - self->b.top_idx, + seq++, browser->b.nr_entries, + browser->hists->nr_entries, + browser->b.height, + browser->b.index, + browser->b.top_idx, h->row_offset, h->nr_rows); } break; case 'C': /* Collapse the whole world. */ - hist_browser__set_folding(self, false); + hist_browser__set_folding(browser, false); break; case 'E': /* Expand the whole world. */ - hist_browser__set_folding(self, true); + hist_browser__set_folding(browser, true); break; case K_ENTER: - if (hist_browser__toggle_fold(self)) + if (hist_browser__toggle_fold(browser)) break; /* fall thru */ default: @@ -367,23 +370,23 @@ static int hist_browser__run(struct hist_browser *self, const char *ev_name, } } out: - ui_browser__hide(&self->b); + ui_browser__hide(&browser->b); return key; } -static char *callchain_list__sym_name(struct callchain_list *self, +static char *callchain_list__sym_name(struct callchain_list *cl, char *bf, size_t bfsize) { - if (self->ms.sym) - return self->ms.sym->name; + if (cl->ms.sym) + return cl->ms.sym->name; - snprintf(bf, bfsize, "%#" PRIx64, self->ip); + snprintf(bf, bfsize, "%#" PRIx64, cl->ip); return bf; } #define LEVEL_OFFSET_STEP 3 -static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *self, +static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *browser, struct callchain_node *chain_node, u64 total, int level, unsigned short row, @@ -441,21 +444,21 @@ static int hist_browser__show_callchain_node_rb_tree(struct hist_browser *self, } color = HE_COLORSET_NORMAL; - width = self->b.width - (offset + extra_offset + 2); - if (ui_browser__is_current_entry(&self->b, row)) { - self->selection = &chain->ms; + width = browser->b.width - (offset + extra_offset + 2); + if (ui_browser__is_current_entry(&browser->b, row)) { + browser->selection = &chain->ms; color = HE_COLORSET_SELECTED; *is_current_entry = true; } - ui_browser__set_color(&self->b, color); - ui_browser__gotorc(&self->b, row, 0); + ui_browser__set_color(&browser->b, color); + ui_browser__gotorc(&browser->b, row, 0); slsmg_write_nstring(" ", offset + extra_offset); slsmg_printf("%c ", folded_sign); slsmg_write_nstring(str, width); free(alloc_str); - if (++row == self->b.height) + if (++row == browser->b.height) goto out; do_next: if (folded_sign == '+') @@ -464,11 +467,11 @@ do_next: if (folded_sign == '-') { const int new_level = level + (extra_offset ? 2 : 1); - row += hist_browser__show_callchain_node_rb_tree(self, child, new_total, + row += hist_browser__show_callchain_node_rb_tree(browser, child, new_total, new_level, row, row_offset, is_current_entry); } - if (row == self->b.height) + if (row == browser->b.height) goto out; node = next; } @@ -476,7 +479,7 @@ out: return row - first_row; } -static int hist_browser__show_callchain_node(struct hist_browser *self, +static int hist_browser__show_callchain_node(struct hist_browser *browser, struct callchain_node *node, int level, unsigned short row, off_t *row_offset, @@ -485,7 +488,7 @@ static int hist_browser__show_callchain_node(struct hist_browser *self, struct callchain_list *chain; int first_row = row, offset = level * LEVEL_OFFSET_STEP, - width = self->b.width - offset; + width = browser->b.width - offset; char folded_sign = ' '; list_for_each_entry(chain, &node->val, list) { @@ -500,26 +503,26 @@ static int hist_browser__show_callchain_node(struct hist_browser *self, } color = HE_COLORSET_NORMAL; - if (ui_browser__is_current_entry(&self->b, row)) { - self->selection = &chain->ms; + if (ui_browser__is_current_entry(&browser->b, row)) { + browser->selection = &chain->ms; color = HE_COLORSET_SELECTED; *is_current_entry = true; } s = callchain_list__sym_name(chain, ipstr, sizeof(ipstr)); - ui_browser__gotorc(&self->b, row, 0); - ui_browser__set_color(&self->b, color); + ui_browser__gotorc(&browser->b, row, 0); + ui_browser__set_color(&browser->b, color); slsmg_write_nstring(" ", offset); slsmg_printf("%c ", folded_sign); slsmg_write_nstring(s, width - 2); - if (++row == self->b.height) + if (++row == browser->b.height) goto out; } if (folded_sign == '-') - row += hist_browser__show_callchain_node_rb_tree(self, node, - self->hists->stats.total_period, + row += hist_browser__show_callchain_node_rb_tree(browser, node, + browser->hists->stats.total_period, level + 1, row, row_offset, is_current_entry); @@ -527,7 +530,7 @@ out: return row - first_row; } -static int hist_browser__show_callchain(struct hist_browser *self, +static int hist_browser__show_callchain(struct hist_browser *browser, struct rb_root *chain, int level, unsigned short row, off_t *row_offset, @@ -539,31 +542,31 @@ static int hist_browser__show_callchain(struct hist_browser *self, for (nd = rb_first(chain); nd; nd = rb_next(nd)) { struct callchain_node *node = rb_entry(nd, struct callchain_node, rb_node); - row += hist_browser__show_callchain_node(self, node, level, + row += hist_browser__show_callchain_node(browser, node, level, row, row_offset, is_current_entry); - if (row == self->b.height) + if (row == browser->b.height) break; } return row - first_row; } -static int hist_browser__show_entry(struct hist_browser *self, +static int hist_browser__show_entry(struct hist_browser *browser, struct hist_entry *entry, unsigned short row) { char s[256]; double percent; int printed = 0; - int width = self->b.width - 6; /* The percentage */ + int width = browser->b.width - 6; /* The percentage */ char folded_sign = ' '; - bool current_entry = ui_browser__is_current_entry(&self->b, row); + bool current_entry = ui_browser__is_current_entry(&browser->b, row); off_t row_offset = entry->row_offset; if (current_entry) { - self->he_selection = entry; - self->selection = &entry->ms; + browser->he_selection = entry; + browser->selection = &entry->ms; } if (symbol_conf.use_callchain) { @@ -572,11 +575,11 @@ static int hist_browser__show_entry(struct hist_browser *self, } if (row_offset == 0) { - hist_entry__snprintf(entry, s, sizeof(s), self->hists); - percent = (entry->period * 100.0) / self->hists->stats.total_period; + hist_entry__snprintf(entry, s, sizeof(s), browser->hists); + percent = (entry->period * 100.0) / browser->hists->stats.total_period; - ui_browser__set_percent_color(&self->b, percent, current_entry); - ui_browser__gotorc(&self->b, row, 0); + ui_browser__set_percent_color(&browser->b, percent, current_entry); + ui_browser__gotorc(&browser->b, row, 0); if (symbol_conf.use_callchain) { slsmg_printf("%c ", folded_sign); width -= 2; @@ -585,11 +588,11 @@ static int hist_browser__show_entry(struct hist_browser *self, slsmg_printf(" %5.2f%%", percent); /* The scroll bar isn't being used */ - if (!self->b.navkeypressed) + if (!browser->b.navkeypressed) width += 1; - if (!current_entry || !self->b.navkeypressed) - ui_browser__set_color(&self->b, HE_COLORSET_NORMAL); + if (!current_entry || !browser->b.navkeypressed) + ui_browser__set_color(&browser->b, HE_COLORSET_NORMAL); if (symbol_conf.show_nr_samples) { slsmg_printf(" %11u", entry->nr_events); @@ -607,12 +610,12 @@ static int hist_browser__show_entry(struct hist_browser *self, } else --row_offset; - if (folded_sign == '-' && row != self->b.height) { - printed += hist_browser__show_callchain(self, &entry->sorted_chain, + if (folded_sign == '-' && row != browser->b.height) { + printed += hist_browser__show_callchain(browser, &entry->sorted_chain, 1, row, &row_offset, ¤t_entry); if (current_entry) - self->he_selection = entry; + browser->he_selection = entry; } return printed; @@ -628,22 +631,22 @@ static void ui_browser__hists_init_top(struct ui_browser *browser) } } -static unsigned int hist_browser__refresh(struct ui_browser *self) +static unsigned int hist_browser__refresh(struct ui_browser *browser) { unsigned row = 0; struct rb_node *nd; - struct hist_browser *hb = container_of(self, struct hist_browser, b); + struct hist_browser *hb = container_of(browser, struct hist_browser, b); - ui_browser__hists_init_top(self); + ui_browser__hists_init_top(browser); - for (nd = self->top; nd; nd = rb_next(nd)) { + for (nd = browser->top; nd; nd = rb_next(nd)) { struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); if (h->filtered) continue; row += hist_browser__show_entry(hb, h, row); - if (row == self->height) + if (row == browser->height) break; } @@ -676,27 +679,27 @@ static struct rb_node *hists__filter_prev_entries(struct rb_node *nd) return NULL; } -static void ui_browser__hists_seek(struct ui_browser *self, +static void ui_browser__hists_seek(struct ui_browser *browser, off_t offset, int whence) { struct hist_entry *h; struct rb_node *nd; bool first = true; - if (self->nr_entries == 0) + if (browser->nr_entries == 0) return; - ui_browser__hists_init_top(self); + ui_browser__hists_init_top(browser); switch (whence) { case SEEK_SET: - nd = hists__filter_entries(rb_first(self->entries)); + nd = hists__filter_entries(rb_first(browser->entries)); break; case SEEK_CUR: - nd = self->top; + nd = browser->top; goto do_offset; case SEEK_END: - nd = hists__filter_prev_entries(rb_last(self->entries)); + nd = hists__filter_prev_entries(rb_last(browser->entries)); first = false; break; default: @@ -707,7 +710,7 @@ static void ui_browser__hists_seek(struct ui_browser *self, * Moves not relative to the first visible entry invalidates its * row_offset: */ - h = rb_entry(self->top, struct hist_entry, rb_node); + h = rb_entry(browser->top, struct hist_entry, rb_node); h->row_offset = 0; /* @@ -735,7 +738,7 @@ do_offset: } else { h->row_offset += offset; offset = 0; - self->top = nd; + browser->top = nd; break; } } @@ -743,7 +746,7 @@ do_offset: if (nd == NULL) break; --offset; - self->top = nd; + browser->top = nd; } while (offset != 0); } else if (offset < 0) { while (1) { @@ -756,7 +759,7 @@ do_offset: } else { h->row_offset += offset; offset = 0; - self->top = nd; + browser->top = nd; break; } } else { @@ -766,7 +769,7 @@ do_offset: } else { h->row_offset = h->nr_rows + offset; offset = 0; - self->top = nd; + browser->top = nd; break; } } @@ -776,7 +779,7 @@ do_offset: if (nd == NULL) break; ++offset; - self->top = nd; + browser->top = nd; if (offset == 0) { /* * Last unfiltered hist_entry, check if it is @@ -791,7 +794,7 @@ do_offset: first = false; } } else { - self->top = nd; + browser->top = nd; h = rb_entry(nd, struct hist_entry, rb_node); h->row_offset = 0; } @@ -799,52 +802,56 @@ do_offset: static struct hist_browser *hist_browser__new(struct hists *hists) { - struct hist_browser *self = zalloc(sizeof(*self)); + struct hist_browser *browser = zalloc(sizeof(*browser)); - if (self) { - self->hists = hists; - self->b.refresh = hist_browser__refresh; - self->b.seek = ui_browser__hists_seek; - self->b.use_navkeypressed = true; + if (browser) { + browser->hists = hists; + browser->b.refresh = hist_browser__refresh; + browser->b.seek = ui_browser__hists_seek; + browser->b.use_navkeypressed = true; if (sort__branch_mode == 1) - self->has_symbols = sort_sym_from.list.next != NULL; + browser->has_symbols = sort_sym_from.list.next != NULL; else - self->has_symbols = sort_sym.list.next != NULL; + browser->has_symbols = sort_sym.list.next != NULL; } - return self; + return browser; } -static void hist_browser__delete(struct hist_browser *self) +static void hist_browser__delete(struct hist_browser *browser) { - free(self); + free(browser); } -static struct hist_entry *hist_browser__selected_entry(struct hist_browser *self) +static struct hist_entry *hist_browser__selected_entry(struct hist_browser *browser) { - return self->he_selection; + return browser->he_selection; } -static struct thread *hist_browser__selected_thread(struct hist_browser *self) +static struct thread *hist_browser__selected_thread(struct hist_browser *browser) { - return self->he_selection->thread; + return browser->he_selection->thread; } -static int hists__browser_title(struct hists *self, char *bf, size_t size, +static int hists__browser_title(struct hists *hists, char *bf, size_t size, const char *ev_name) { char unit; int printed; - const struct dso *dso = self->dso_filter; - const struct thread *thread = self->thread_filter; - unsigned long nr_events = self->stats.nr_events[PERF_RECORD_SAMPLE]; + const struct dso *dso = hists->dso_filter; + const struct thread *thread = hists->thread_filter; + unsigned long nr_samples = hists->stats.nr_events[PERF_RECORD_SAMPLE]; + u64 nr_events = hists->stats.total_period; - nr_events = convert_unit(nr_events, &unit); - printed = scnprintf(bf, size, "Events: %lu%c %s", nr_events, unit, ev_name); + nr_samples = convert_unit(nr_samples, &unit); + printed = scnprintf(bf, size, + "Samples: %lu%c of event '%s', Event count (approx.): %lu", + nr_samples, unit, ev_name, nr_events); - if (self->uid_filter_str) + + if (hists->uid_filter_str) printed += snprintf(bf + printed, size - printed, - ", UID: %s", self->uid_filter_str); + ", UID: %s", hists->uid_filter_str); if (thread) printed += scnprintf(bf + printed, size - printed, ", Thread: %s(%d)", @@ -872,13 +879,14 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, void(*timer)(void *arg), void *arg, int delay_secs) { - struct hists *self = &evsel->hists; - struct hist_browser *browser = hist_browser__new(self); + struct hists *hists = &evsel->hists; + struct hist_browser *browser = hist_browser__new(hists); struct branch_info *bi; struct pstack *fstack; char *options[16]; int nr_options = 0; int key = -1; + char buf[64]; if (browser == NULL) return -1; @@ -933,6 +941,16 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, goto zoom_dso; case 't': goto zoom_thread; + case '/': + if (ui_browser__input_window("Symbol to show", + "Please enter the name of symbol you want to see", + buf, "ENTER: OK, ESC: Cancel", + delay_secs * 2) == K_ENTER) { + hists->symbol_filter_str = *buf ? buf : NULL; + hists__filter_by_symbol(hists); + hist_browser__reset(browser); + } + continue; case K_F1: case 'h': case '?': @@ -950,7 +968,8 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, "C Collapse all callchains\n" "E Expand all callchains\n" "d Zoom into current DSO\n" - "t Zoom into current Thread"); + "t Zoom into current Thread\n" + "/ Filter symbol by name"); continue; case K_ENTER: case K_RIGHT: @@ -1109,7 +1128,7 @@ zoom_out_dso: sort_dso.elide = true; pstack__push(fstack, &browser->hists->dso_filter); } - hists__filter_by_dso(self); + hists__filter_by_dso(hists); hist_browser__reset(browser); } else if (choice == zoom_thread) { zoom_thread: @@ -1127,7 +1146,7 @@ zoom_out_thread: sort_thread.elide = true; pstack__push(fstack, &browser->hists->thread_filter); } - hists__filter_by_thread(self); + hists__filter_by_thread(hists); hist_browser__reset(browser); } } diff --git a/tools/perf/util/ui/browsers/map.c b/tools/perf/ui/browsers/map.c index eca6575abfd0..98851d55a53e 100644 --- a/tools/perf/util/ui/browsers/map.c +++ b/tools/perf/ui/browsers/map.c @@ -5,9 +5,9 @@ #include <sys/ttydefaults.h> #include <string.h> #include <linux/bitops.h> -#include "../../util.h" -#include "../../debug.h" -#include "../../symbol.h" +#include "../../util/util.h" +#include "../../util/debug.h" +#include "../../util/symbol.h" #include "../browser.h" #include "../helpline.h" #include "map.h" diff --git a/tools/perf/util/ui/browsers/map.h b/tools/perf/ui/browsers/map.h index df8581a43e17..df8581a43e17 100644 --- a/tools/perf/util/ui/browsers/map.h +++ b/tools/perf/ui/browsers/map.h diff --git a/tools/perf/ui/gtk/browser.c b/tools/perf/ui/gtk/browser.c new file mode 100644 index 000000000000..0656c381a89c --- /dev/null +++ b/tools/perf/ui/gtk/browser.c @@ -0,0 +1,178 @@ +#include "../evlist.h" +#include "../cache.h" +#include "../evsel.h" +#include "../sort.h" +#include "../hist.h" +#include "gtk.h" + +#include <signal.h> + +#define MAX_COLUMNS 32 + +static void perf_gtk__signal(int sig) +{ + psignal(sig, "perf"); + gtk_main_quit(); +} + +static void perf_gtk__resize_window(GtkWidget *window) +{ + GdkRectangle rect; + GdkScreen *screen; + int monitor; + int height; + int width; + + screen = gtk_widget_get_screen(window); + + monitor = gdk_screen_get_monitor_at_window(screen, window->window); + + gdk_screen_get_monitor_geometry(screen, monitor, &rect); + + width = rect.width * 3 / 4; + height = rect.height * 3 / 4; + + gtk_window_resize(GTK_WINDOW(window), width, height); +} + +static void perf_gtk__show_hists(GtkWidget *window, struct hists *hists) +{ + GType col_types[MAX_COLUMNS]; + GtkCellRenderer *renderer; + struct sort_entry *se; + GtkListStore *store; + struct rb_node *nd; + u64 total_period; + GtkWidget *view; + int col_idx; + int nr_cols; + + nr_cols = 0; + + /* The percentage column */ + col_types[nr_cols++] = G_TYPE_STRING; + + list_for_each_entry(se, &hist_entry__sort_list, list) { + if (se->elide) + continue; + + col_types[nr_cols++] = G_TYPE_STRING; + } + + store = gtk_list_store_newv(nr_cols, col_types); + + view = gtk_tree_view_new(); + + renderer = gtk_cell_renderer_text_new(); + + col_idx = 0; + + /* The percentage column */ + gtk_tree_view_insert_column_with_attributes(GTK_TREE_VIEW(view), + -1, "Overhead (%)", + renderer, "text", + col_idx++, NULL); + + list_for_each_entry(se, &hist_entry__sort_list, list) { + if (se->elide) + continue; + + gtk_tree_view_insert_column_with_attributes(GTK_TREE_VIEW(view), + -1, se->se_header, + renderer, "text", + col_idx++, NULL); + } + + gtk_tree_view_set_model(GTK_TREE_VIEW(view), GTK_TREE_MODEL(store)); + + g_object_unref(GTK_TREE_MODEL(store)); + + total_period = hists->stats.total_period; + + for (nd = rb_first(&hists->entries); nd; nd = rb_next(nd)) { + struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); + GtkTreeIter iter; + double percent; + char s[512]; + + if (h->filtered) + continue; + + gtk_list_store_append(store, &iter); + + col_idx = 0; + + percent = (h->period * 100.0) / total_period; + + snprintf(s, ARRAY_SIZE(s), "%.2f", percent); + + gtk_list_store_set(store, &iter, col_idx++, s, -1); + + list_for_each_entry(se, &hist_entry__sort_list, list) { + if (se->elide) + continue; + + se->se_snprintf(h, s, ARRAY_SIZE(s), + hists__col_len(hists, se->se_width_idx)); + + gtk_list_store_set(store, &iter, col_idx++, s, -1); + } + } + + gtk_container_add(GTK_CONTAINER(window), view); +} + +int perf_evlist__gtk_browse_hists(struct perf_evlist *evlist, + const char *help __used, + void (*timer) (void *arg)__used, + void *arg __used, int delay_secs __used) +{ + struct perf_evsel *pos; + GtkWidget *notebook; + GtkWidget *window; + + signal(SIGSEGV, perf_gtk__signal); + signal(SIGFPE, perf_gtk__signal); + signal(SIGINT, perf_gtk__signal); + signal(SIGQUIT, perf_gtk__signal); + signal(SIGTERM, perf_gtk__signal); + + window = gtk_window_new(GTK_WINDOW_TOPLEVEL); + + gtk_window_set_title(GTK_WINDOW(window), "perf report"); + + g_signal_connect(window, "delete_event", gtk_main_quit, NULL); + + notebook = gtk_notebook_new(); + + list_for_each_entry(pos, &evlist->entries, node) { + struct hists *hists = &pos->hists; + const char *evname = event_name(pos); + GtkWidget *scrolled_window; + GtkWidget *tab_label; + + scrolled_window = gtk_scrolled_window_new(NULL, NULL); + + gtk_scrolled_window_set_policy(GTK_SCROLLED_WINDOW(scrolled_window), + GTK_POLICY_AUTOMATIC, + GTK_POLICY_AUTOMATIC); + + perf_gtk__show_hists(scrolled_window, hists); + + tab_label = gtk_label_new(evname); + + gtk_notebook_append_page(GTK_NOTEBOOK(notebook), scrolled_window, tab_label); + } + + gtk_container_add(GTK_CONTAINER(window), notebook); + + gtk_widget_show_all(window); + + perf_gtk__resize_window(window); + + gtk_window_set_position(GTK_WINDOW(window), GTK_WIN_POS_CENTER); + + gtk_main(); + + return 0; +} diff --git a/tools/perf/ui/gtk/gtk.h b/tools/perf/ui/gtk/gtk.h new file mode 100644 index 000000000000..75177ee04032 --- /dev/null +++ b/tools/perf/ui/gtk/gtk.h @@ -0,0 +1,8 @@ +#ifndef _PERF_GTK_H_ +#define _PERF_GTK_H_ 1 + +#pragma GCC diagnostic ignored "-Wstrict-prototypes" +#include <gtk/gtk.h> +#pragma GCC diagnostic error "-Wstrict-prototypes" + +#endif /* _PERF_GTK_H_ */ diff --git a/tools/perf/ui/gtk/setup.c b/tools/perf/ui/gtk/setup.c new file mode 100644 index 000000000000..829529957766 --- /dev/null +++ b/tools/perf/ui/gtk/setup.c @@ -0,0 +1,12 @@ +#include "gtk.h" +#include "../../util/cache.h" + +int perf_gtk__init(void) +{ + return gtk_init_check(NULL, NULL) ? 0 : -1; +} + +void perf_gtk__exit(bool wait_for_ok __used) +{ + gtk_main_quit(); +} diff --git a/tools/perf/util/ui/helpline.c b/tools/perf/ui/helpline.c index 2f950c2641c8..2f950c2641c8 100644 --- a/tools/perf/util/ui/helpline.c +++ b/tools/perf/ui/helpline.c diff --git a/tools/perf/util/ui/helpline.h b/tools/perf/ui/helpline.h index 7bab6b34e35e..7bab6b34e35e 100644 --- a/tools/perf/util/ui/helpline.h +++ b/tools/perf/ui/helpline.h diff --git a/tools/perf/util/ui/keysyms.h b/tools/perf/ui/keysyms.h index 3458b1985761..809eca5707fa 100644 --- a/tools/perf/util/ui/keysyms.h +++ b/tools/perf/ui/keysyms.h @@ -16,6 +16,8 @@ #define K_TAB '\t' #define K_UNTAB SL_KEY_UNTAB #define K_UP SL_KEY_UP +#define K_BKSPC 0x7f +#define K_DEL SL_KEY_DELETE /* Not really keys */ #define K_TIMER -1 diff --git a/tools/perf/util/ui/libslang.h b/tools/perf/ui/libslang.h index 4d54b6450f5b..4d54b6450f5b 100644 --- a/tools/perf/util/ui/libslang.h +++ b/tools/perf/ui/libslang.h diff --git a/tools/perf/util/ui/progress.c b/tools/perf/ui/progress.c index 13aa64e50e11..13aa64e50e11 100644 --- a/tools/perf/util/ui/progress.c +++ b/tools/perf/ui/progress.c diff --git a/tools/perf/util/ui/progress.h b/tools/perf/ui/progress.h index d9c205b59aa1..d9c205b59aa1 100644 --- a/tools/perf/util/ui/progress.h +++ b/tools/perf/ui/progress.h diff --git a/tools/perf/ui/setup.c b/tools/perf/ui/setup.c new file mode 100644 index 000000000000..791fb15ce350 --- /dev/null +++ b/tools/perf/ui/setup.c @@ -0,0 +1,46 @@ +#include "../cache.h" +#include "../debug.h" + + +void setup_browser(bool fallback_to_pager) +{ + if (!isatty(1) || dump_trace) + use_browser = 0; + + /* default to TUI */ + if (use_browser < 0) + use_browser = 1; + + switch (use_browser) { + case 2: + if (perf_gtk__init() == 0) + break; + /* fall through */ + case 1: + use_browser = 1; + if (ui__init() == 0) + break; + /* fall through */ + default: + use_browser = 0; + if (fallback_to_pager) + setup_pager(); + break; + } +} + +void exit_browser(bool wait_for_ok) +{ + switch (use_browser) { + case 2: + perf_gtk__exit(wait_for_ok); + break; + + case 1: + ui__exit(wait_for_ok); + break; + + default: + break; + } +} diff --git a/tools/perf/util/ui/setup.c b/tools/perf/ui/tui/setup.c index 85a69faa09aa..d33e943ac434 100644 --- a/tools/perf/util/ui/setup.c +++ b/tools/perf/ui/tui/setup.c @@ -2,14 +2,14 @@ #include <signal.h> #include <stdbool.h> -#include "../cache.h" -#include "../debug.h" -#include "browser.h" -#include "helpline.h" -#include "ui.h" -#include "util.h" -#include "libslang.h" -#include "keysyms.h" +#include "../../util/cache.h" +#include "../../util/debug.h" +#include "../browser.h" +#include "../helpline.h" +#include "../ui.h" +#include "../util.h" +#include "../libslang.h" +#include "../keysyms.h" pthread_mutex_t ui__lock = PTHREAD_MUTEX_INITIALIZER; @@ -93,45 +93,26 @@ static void newt_suspend(void *d __used) newtResume(); } -static int ui__init(void) -{ - int err = SLkp_init(); - - if (err < 0) - goto out; - - SLkp_define_keysym((char *)"^(kB)", SL_KEY_UNTAB); -out: - return err; -} - -static void ui__exit(void) -{ - SLtt_set_cursor_visibility(1); - SLsmg_refresh(); - SLsmg_reset_smg(); - SLang_reset_tty(); -} - static void ui__signal(int sig) { - ui__exit(); + ui__exit(false); psignal(sig, "perf"); exit(0); } -void setup_browser(bool fallback_to_pager) +int ui__init(void) { - if (!isatty(1) || !use_browser || dump_trace) { - use_browser = 0; - if (fallback_to_pager) - setup_pager(); - return; - } + int err; - use_browser = 1; newtInit(); - ui__init(); + err = SLkp_init(); + if (err < 0) { + pr_err("TUI initialization failed.\n"); + goto out; + } + + SLkp_define_keysym((char *)"^(kB)", SL_KEY_UNTAB); + newtSetSuspendCallback(newt_suspend, NULL); ui_helpline__init(); ui_browser__init(); @@ -141,15 +122,19 @@ void setup_browser(bool fallback_to_pager) signal(SIGINT, ui__signal); signal(SIGQUIT, ui__signal); signal(SIGTERM, ui__signal); +out: + return err; } -void exit_browser(bool wait_for_ok) +void ui__exit(bool wait_for_ok) { - if (use_browser > 0) { - if (wait_for_ok) - ui__question_window("Fatal Error", - ui_helpline__last_msg, - "Press any key...", 0); - ui__exit(); - } + if (wait_for_ok) + ui__question_window("Fatal Error", + ui_helpline__last_msg, + "Press any key...", 0); + + SLtt_set_cursor_visibility(1); + SLsmg_refresh(); + SLsmg_reset_smg(); + SLang_reset_tty(); } diff --git a/tools/perf/util/ui/ui.h b/tools/perf/ui/ui.h index 7b67045479f6..7b67045479f6 100644 --- a/tools/perf/util/ui/ui.h +++ b/tools/perf/ui/ui.h diff --git a/tools/perf/util/ui/util.c b/tools/perf/ui/util.c index 45daa7c41dad..ad4374a16bb0 100644 --- a/tools/perf/util/ui/util.c +++ b/tools/perf/ui/util.c @@ -69,6 +69,88 @@ int ui__popup_menu(int argc, char * const argv[]) return popup_menu__run(&menu); } +int ui_browser__input_window(const char *title, const char *text, char *input, + const char *exit_msg, int delay_secs) +{ + int x, y, len, key; + int max_len = 60, nr_lines = 0; + static char buf[50]; + const char *t; + + t = text; + while (1) { + const char *sep = strchr(t, '\n'); + + if (sep == NULL) + sep = strchr(t, '\0'); + len = sep - t; + if (max_len < len) + max_len = len; + ++nr_lines; + if (*sep == '\0') + break; + t = sep + 1; + } + + max_len += 2; + nr_lines += 8; + y = SLtt_Screen_Rows / 2 - nr_lines / 2; + x = SLtt_Screen_Cols / 2 - max_len / 2; + + SLsmg_set_color(0); + SLsmg_draw_box(y, x++, nr_lines, max_len); + if (title) { + SLsmg_gotorc(y, x + 1); + SLsmg_write_string((char *)title); + } + SLsmg_gotorc(++y, x); + nr_lines -= 7; + max_len -= 2; + SLsmg_write_wrapped_string((unsigned char *)text, y, x, + nr_lines, max_len, 1); + y += nr_lines; + len = 5; + while (len--) { + SLsmg_gotorc(y + len - 1, x); + SLsmg_write_nstring((char *)" ", max_len); + } + SLsmg_draw_box(y++, x + 1, 3, max_len - 2); + + SLsmg_gotorc(y + 3, x); + SLsmg_write_nstring((char *)exit_msg, max_len); + SLsmg_refresh(); + + x += 2; + len = 0; + key = ui__getch(delay_secs); + while (key != K_TIMER && key != K_ENTER && key != K_ESC) { + if (key == K_BKSPC) { + if (len == 0) + goto next_key; + SLsmg_gotorc(y, x + --len); + SLsmg_write_char(' '); + } else { + buf[len] = key; + SLsmg_gotorc(y, x + len++); + SLsmg_write_char(key); + } + SLsmg_refresh(); + + /* XXX more graceful overflow handling needed */ + if (len == sizeof(buf) - 1) { + ui_helpline__push("maximum size of symbol name reached!"); + key = K_ENTER; + break; + } +next_key: + key = ui__getch(delay_secs); + } + + buf[len] = '\0'; + strncpy(input, buf, len+1); + return key; +} + int ui__question_window(const char *title, const char *text, const char *exit_msg, int delay_secs) { diff --git a/tools/perf/util/ui/util.h b/tools/perf/ui/util.h index 2d1738bd71c8..2d1738bd71c8 100644 --- a/tools/perf/util/ui/util.h +++ b/tools/perf/ui/util.h diff --git a/tools/perf/util/PERF-VERSION-GEN b/tools/perf/util/PERF-VERSION-GEN index ad73300f7bac..95264f304179 100755 --- a/tools/perf/util/PERF-VERSION-GEN +++ b/tools/perf/util/PERF-VERSION-GEN @@ -12,7 +12,7 @@ LF=' # First check if there is a .git to get the version from git describe # otherwise try to get the version from the kernel makefile if test -d ../../.git -o -f ../../.git && - VN=$(git describe --abbrev=4 HEAD 2>/dev/null) && + VN=$(git describe --match 'v[0-9].[0-9]*' --abbrev=4 HEAD 2>/dev/null) && case "$VN" in *$LF*) (exit 1) ;; v[0-9]*) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index e5a462f1d07c..8069dfb5ba77 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -18,6 +18,403 @@ const char *disassembler_style; +static struct ins *ins__find(const char *name); +static int disasm_line__parse(char *line, char **namep, char **rawp); + +static void ins__delete(struct ins_operands *ops) +{ + free(ops->source.raw); + free(ops->source.name); + free(ops->target.raw); + free(ops->target.name); +} + +static int ins__raw_scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + return scnprintf(bf, size, "%-6.6s %s", ins->name, ops->raw); +} + +int ins__scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + if (ins->ops->scnprintf) + return ins->ops->scnprintf(ins, bf, size, ops); + + return ins__raw_scnprintf(ins, bf, size, ops); +} + +static int call__parse(struct ins_operands *ops) +{ + char *endptr, *tok, *name; + + ops->target.addr = strtoull(ops->raw, &endptr, 16); + + name = strchr(endptr, '<'); + if (name == NULL) + goto indirect_call; + + name++; + + tok = strchr(name, '>'); + if (tok == NULL) + return -1; + + *tok = '\0'; + ops->target.name = strdup(name); + *tok = '>'; + + return ops->target.name == NULL ? -1 : 0; + +indirect_call: + tok = strchr(endptr, '('); + if (tok != NULL) { + ops->target.addr = 0; + return 0; + } + + tok = strchr(endptr, '*'); + if (tok == NULL) + return -1; + + ops->target.addr = strtoull(tok + 1, NULL, 16); + return 0; +} + +static int call__scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + if (ops->target.name) + return scnprintf(bf, size, "%-6.6s %s", ins->name, ops->target.name); + + if (ops->target.addr == 0) + return ins__raw_scnprintf(ins, bf, size, ops); + + return scnprintf(bf, size, "%-6.6s *%" PRIx64, ins->name, ops->target.addr); +} + +static struct ins_ops call_ops = { + .parse = call__parse, + .scnprintf = call__scnprintf, +}; + +bool ins__is_call(const struct ins *ins) +{ + return ins->ops == &call_ops; +} + +static int jump__parse(struct ins_operands *ops) +{ + const char *s = strchr(ops->raw, '+'); + + ops->target.addr = strtoll(ops->raw, NULL, 16); + + if (s++ != NULL) + ops->target.offset = strtoll(s, NULL, 16); + else + ops->target.offset = UINT64_MAX; + + return 0; +} + +static int jump__scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + return scnprintf(bf, size, "%-6.6s %" PRIx64, ins->name, ops->target.offset); +} + +static struct ins_ops jump_ops = { + .parse = jump__parse, + .scnprintf = jump__scnprintf, +}; + +bool ins__is_jump(const struct ins *ins) +{ + return ins->ops == &jump_ops; +} + +static int comment__symbol(char *raw, char *comment, u64 *addrp, char **namep) +{ + char *endptr, *name, *t; + + if (strstr(raw, "(%rip)") == NULL) + return 0; + + *addrp = strtoull(comment, &endptr, 16); + name = strchr(endptr, '<'); + if (name == NULL) + return -1; + + name++; + + t = strchr(name, '>'); + if (t == NULL) + return 0; + + *t = '\0'; + *namep = strdup(name); + *t = '>'; + + return 0; +} + +static int lock__parse(struct ins_operands *ops) +{ + char *name; + + ops->locked.ops = zalloc(sizeof(*ops->locked.ops)); + if (ops->locked.ops == NULL) + return 0; + + if (disasm_line__parse(ops->raw, &name, &ops->locked.ops->raw) < 0) + goto out_free_ops; + + ops->locked.ins = ins__find(name); + if (ops->locked.ins == NULL) + goto out_free_ops; + + if (!ops->locked.ins->ops) + return 0; + + if (ops->locked.ins->ops->parse) + ops->locked.ins->ops->parse(ops->locked.ops); + + return 0; + +out_free_ops: + free(ops->locked.ops); + ops->locked.ops = NULL; + return 0; +} + +static int lock__scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + int printed; + + if (ops->locked.ins == NULL) + return ins__raw_scnprintf(ins, bf, size, ops); + + printed = scnprintf(bf, size, "%-6.6s ", ins->name); + return printed + ins__scnprintf(ops->locked.ins, bf + printed, + size - printed, ops->locked.ops); +} + +static void lock__delete(struct ins_operands *ops) +{ + free(ops->locked.ops); + free(ops->target.raw); + free(ops->target.name); +} + +static struct ins_ops lock_ops = { + .free = lock__delete, + .parse = lock__parse, + .scnprintf = lock__scnprintf, +}; + +static int mov__parse(struct ins_operands *ops) +{ + char *s = strchr(ops->raw, ','), *target, *comment, prev; + + if (s == NULL) + return -1; + + *s = '\0'; + ops->source.raw = strdup(ops->raw); + *s = ','; + + if (ops->source.raw == NULL) + return -1; + + target = ++s; + + while (s[0] != '\0' && !isspace(s[0])) + ++s; + prev = *s; + *s = '\0'; + + ops->target.raw = strdup(target); + *s = prev; + + if (ops->target.raw == NULL) + goto out_free_source; + + comment = strchr(s, '#'); + if (comment == NULL) + return 0; + + while (comment[0] != '\0' && isspace(comment[0])) + ++comment; + + comment__symbol(ops->source.raw, comment, &ops->source.addr, &ops->source.name); + comment__symbol(ops->target.raw, comment, &ops->target.addr, &ops->target.name); + + return 0; + +out_free_source: + free(ops->source.raw); + ops->source.raw = NULL; + return -1; +} + +static int mov__scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + return scnprintf(bf, size, "%-6.6s %s,%s", ins->name, + ops->source.name ?: ops->source.raw, + ops->target.name ?: ops->target.raw); +} + +static struct ins_ops mov_ops = { + .parse = mov__parse, + .scnprintf = mov__scnprintf, +}; + +static int dec__parse(struct ins_operands *ops) +{ + char *target, *comment, *s, prev; + + target = s = ops->raw; + + while (s[0] != '\0' && !isspace(s[0])) + ++s; + prev = *s; + *s = '\0'; + + ops->target.raw = strdup(target); + *s = prev; + + if (ops->target.raw == NULL) + return -1; + + comment = strchr(s, '#'); + if (comment == NULL) + return 0; + + while (comment[0] != '\0' && isspace(comment[0])) + ++comment; + + comment__symbol(ops->target.raw, comment, &ops->target.addr, &ops->target.name); + + return 0; +} + +static int dec__scnprintf(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops) +{ + return scnprintf(bf, size, "%-6.6s %s", ins->name, + ops->target.name ?: ops->target.raw); +} + +static struct ins_ops dec_ops = { + .parse = dec__parse, + .scnprintf = dec__scnprintf, +}; + +static int nop__scnprintf(struct ins *ins __used, char *bf, size_t size, + struct ins_operands *ops __used) +{ + return scnprintf(bf, size, "%-6.6s", "nop"); +} + +static struct ins_ops nop_ops = { + .scnprintf = nop__scnprintf, +}; + +/* + * Must be sorted by name! + */ +static struct ins instructions[] = { + { .name = "add", .ops = &mov_ops, }, + { .name = "addl", .ops = &mov_ops, }, + { .name = "addq", .ops = &mov_ops, }, + { .name = "addw", .ops = &mov_ops, }, + { .name = "and", .ops = &mov_ops, }, + { .name = "bts", .ops = &mov_ops, }, + { .name = "call", .ops = &call_ops, }, + { .name = "callq", .ops = &call_ops, }, + { .name = "cmp", .ops = &mov_ops, }, + { .name = "cmpb", .ops = &mov_ops, }, + { .name = "cmpl", .ops = &mov_ops, }, + { .name = "cmpq", .ops = &mov_ops, }, + { .name = "cmpw", .ops = &mov_ops, }, + { .name = "cmpxch", .ops = &mov_ops, }, + { .name = "dec", .ops = &dec_ops, }, + { .name = "decl", .ops = &dec_ops, }, + { .name = "imul", .ops = &mov_ops, }, + { .name = "inc", .ops = &dec_ops, }, + { .name = "incl", .ops = &dec_ops, }, + { .name = "ja", .ops = &jump_ops, }, + { .name = "jae", .ops = &jump_ops, }, + { .name = "jb", .ops = &jump_ops, }, + { .name = "jbe", .ops = &jump_ops, }, + { .name = "jc", .ops = &jump_ops, }, + { .name = "jcxz", .ops = &jump_ops, }, + { .name = "je", .ops = &jump_ops, }, + { .name = "jecxz", .ops = &jump_ops, }, + { .name = "jg", .ops = &jump_ops, }, + { .name = "jge", .ops = &jump_ops, }, + { .name = "jl", .ops = &jump_ops, }, + { .name = "jle", .ops = &jump_ops, }, + { .name = "jmp", .ops = &jump_ops, }, + { .name = "jmpq", .ops = &jump_ops, }, + { .name = "jna", .ops = &jump_ops, }, + { .name = "jnae", .ops = &jump_ops, }, + { .name = "jnb", .ops = &jump_ops, }, + { .name = "jnbe", .ops = &jump_ops, }, + { .name = "jnc", .ops = &jump_ops, }, + { .name = "jne", .ops = &jump_ops, }, + { .name = "jng", .ops = &jump_ops, }, + { .name = "jnge", .ops = &jump_ops, }, + { .name = "jnl", .ops = &jump_ops, }, + { .name = "jnle", .ops = &jump_ops, }, + { .name = "jno", .ops = &jump_ops, }, + { .name = "jnp", .ops = &jump_ops, }, + { .name = "jns", .ops = &jump_ops, }, + { .name = "jnz", .ops = &jump_ops, }, + { .name = "jo", .ops = &jump_ops, }, + { .name = "jp", .ops = &jump_ops, }, + { .name = "jpe", .ops = &jump_ops, }, + { .name = "jpo", .ops = &jump_ops, }, + { .name = "jrcxz", .ops = &jump_ops, }, + { .name = "js", .ops = &jump_ops, }, + { .name = "jz", .ops = &jump_ops, }, + { .name = "lea", .ops = &mov_ops, }, + { .name = "lock", .ops = &lock_ops, }, + { .name = "mov", .ops = &mov_ops, }, + { .name = "movb", .ops = &mov_ops, }, + { .name = "movdqa",.ops = &mov_ops, }, + { .name = "movl", .ops = &mov_ops, }, + { .name = "movq", .ops = &mov_ops, }, + { .name = "movslq", .ops = &mov_ops, }, + { .name = "movzbl", .ops = &mov_ops, }, + { .name = "movzwl", .ops = &mov_ops, }, + { .name = "nop", .ops = &nop_ops, }, + { .name = "nopl", .ops = &nop_ops, }, + { .name = "nopw", .ops = &nop_ops, }, + { .name = "or", .ops = &mov_ops, }, + { .name = "orl", .ops = &mov_ops, }, + { .name = "test", .ops = &mov_ops, }, + { .name = "testb", .ops = &mov_ops, }, + { .name = "testl", .ops = &mov_ops, }, + { .name = "xadd", .ops = &mov_ops, }, +}; + +static int ins__cmp(const void *name, const void *insp) +{ + const struct ins *ins = insp; + + return strcmp(name, ins->name); +} + +static struct ins *ins__find(const char *name) +{ + const int nmemb = ARRAY_SIZE(instructions); + + return bsearch(name, instructions, nmemb, sizeof(struct ins), ins__cmp); +} + int symbol__annotate_init(struct map *map __used, struct symbol *sym) { struct annotation *notes = symbol__annotation(sym); @@ -28,8 +425,8 @@ int symbol__annotate_init(struct map *map __used, struct symbol *sym) int symbol__alloc_hist(struct symbol *sym) { struct annotation *notes = symbol__annotation(sym); - size_t sizeof_sym_hist = (sizeof(struct sym_hist) + - (sym->end - sym->start) * sizeof(u64)); + const size_t size = symbol__size(sym); + size_t sizeof_sym_hist = (sizeof(struct sym_hist) + size * sizeof(u64)); notes->src = zalloc(sizeof(*notes->src) + symbol_conf.nr_events * sizeof_sym_hist); if (notes->src == NULL) @@ -64,8 +461,8 @@ int symbol__inc_addr_samples(struct symbol *sym, struct map *map, pr_debug3("%s: addr=%#" PRIx64 "\n", __func__, map->unmap_ip(map, addr)); - if (addr >= sym->end) - return 0; + if (addr < sym->start || addr > sym->end) + return -ERANGE; offset = addr - sym->start; h = annotation__histogram(notes, evidx); @@ -78,31 +475,110 @@ int symbol__inc_addr_samples(struct symbol *sym, struct map *map, return 0; } -static struct objdump_line *objdump_line__new(s64 offset, char *line, size_t privsize) +static void disasm_line__init_ins(struct disasm_line *dl) +{ + dl->ins = ins__find(dl->name); + + if (dl->ins == NULL) + return; + + if (!dl->ins->ops) + return; + + if (dl->ins->ops->parse) + dl->ins->ops->parse(&dl->ops); +} + +static int disasm_line__parse(char *line, char **namep, char **rawp) +{ + char *name = line, tmp; + + while (isspace(name[0])) + ++name; + + if (name[0] == '\0') + return -1; + + *rawp = name + 1; + + while ((*rawp)[0] != '\0' && !isspace((*rawp)[0])) + ++*rawp; + + tmp = (*rawp)[0]; + (*rawp)[0] = '\0'; + *namep = strdup(name); + + if (*namep == NULL) + goto out_free_name; + + (*rawp)[0] = tmp; + + if ((*rawp)[0] != '\0') { + (*rawp)++; + while (isspace((*rawp)[0])) + ++(*rawp); + } + + return 0; + +out_free_name: + free(*namep); + *namep = NULL; + return -1; +} + +static struct disasm_line *disasm_line__new(s64 offset, char *line, size_t privsize) { - struct objdump_line *self = malloc(sizeof(*self) + privsize); + struct disasm_line *dl = zalloc(sizeof(*dl) + privsize); - if (self != NULL) { - self->offset = offset; - self->line = line; + if (dl != NULL) { + dl->offset = offset; + dl->line = strdup(line); + if (dl->line == NULL) + goto out_delete; + + if (offset != -1) { + if (disasm_line__parse(dl->line, &dl->name, &dl->ops.raw) < 0) + goto out_free_line; + + disasm_line__init_ins(dl); + } } - return self; + return dl; + +out_free_line: + free(dl->line); +out_delete: + free(dl); + return NULL; } -void objdump_line__free(struct objdump_line *self) +void disasm_line__free(struct disasm_line *dl) { - free(self->line); - free(self); + free(dl->line); + free(dl->name); + if (dl->ins && dl->ins->ops->free) + dl->ins->ops->free(&dl->ops); + else + ins__delete(&dl->ops); + free(dl); } -static void objdump__add_line(struct list_head *head, struct objdump_line *line) +int disasm_line__scnprintf(struct disasm_line *dl, char *bf, size_t size, bool raw) +{ + if (raw || !dl->ins) + return scnprintf(bf, size, "%-6.6s %s", dl->name, dl->ops.raw); + + return ins__scnprintf(dl->ins, bf, size, &dl->ops); +} + +static void disasm__add(struct list_head *head, struct disasm_line *line) { list_add_tail(&line->node, head); } -struct objdump_line *objdump__get_next_ip_line(struct list_head *head, - struct objdump_line *pos) +struct disasm_line *disasm__get_next_ip_line(struct list_head *head, struct disasm_line *pos) { list_for_each_entry_continue(pos, head, node) if (pos->offset >= 0) @@ -111,15 +587,14 @@ struct objdump_line *objdump__get_next_ip_line(struct list_head *head, return NULL; } -static int objdump_line__print(struct objdump_line *oline, struct symbol *sym, - int evidx, u64 len, int min_pcnt, - int printed, int max_lines, - struct objdump_line *queue) +static int disasm_line__print(struct disasm_line *dl, struct symbol *sym, u64 start, + int evidx, u64 len, int min_pcnt, int printed, + int max_lines, struct disasm_line *queue) { static const char *prev_line; static const char *prev_color; - if (oline->offset != -1) { + if (dl->offset != -1) { const char *path = NULL; unsigned int hits = 0; double percent = 0.0; @@ -127,10 +602,11 @@ static int objdump_line__print(struct objdump_line *oline, struct symbol *sym, struct annotation *notes = symbol__annotation(sym); struct source_line *src_line = notes->src->lines; struct sym_hist *h = annotation__histogram(notes, evidx); - s64 offset = oline->offset; - struct objdump_line *next; + s64 offset = dl->offset; + const u64 addr = start + offset; + struct disasm_line *next; - next = objdump__get_next_ip_line(¬es->src->source, oline); + next = disasm__get_next_ip_line(¬es->src->source, dl); while (offset < (s64)len && (next == NULL || offset < next->offset)) { @@ -155,9 +631,9 @@ static int objdump_line__print(struct objdump_line *oline, struct symbol *sym, if (queue != NULL) { list_for_each_entry_from(queue, ¬es->src->source, node) { - if (queue == oline) + if (queue == dl) break; - objdump_line__print(queue, sym, evidx, len, + disasm_line__print(queue, sym, start, evidx, len, 0, 0, 1, NULL); } } @@ -180,17 +656,18 @@ static int objdump_line__print(struct objdump_line *oline, struct symbol *sym, color_fprintf(stdout, color, " %7.2f", percent); printf(" : "); - color_fprintf(stdout, PERF_COLOR_BLUE, "%s\n", oline->line); + color_fprintf(stdout, PERF_COLOR_MAGENTA, " %" PRIx64 ":", addr); + color_fprintf(stdout, PERF_COLOR_BLUE, "%s\n", dl->line); } else if (max_lines && printed >= max_lines) return 1; else { if (queue) return -1; - if (!*oline->line) + if (!*dl->line) printf(" :\n"); else - printf(" : %s\n", oline->line); + printf(" : %s\n", dl->line); } return 0; @@ -200,8 +677,8 @@ static int symbol__parse_objdump_line(struct symbol *sym, struct map *map, FILE *file, size_t privsize) { struct annotation *notes = symbol__annotation(sym); - struct objdump_line *objdump_line; - char *line = NULL, *tmp, *tmp2, *c; + struct disasm_line *dl; + char *line = NULL, *parsed_line, *tmp, *tmp2, *c; size_t line_len; s64 line_ip, offset = -1; @@ -219,6 +696,7 @@ static int symbol__parse_objdump_line(struct symbol *sym, struct map *map, *c = 0; line_ip = -1; + parsed_line = line; /* * Strip leading spaces: @@ -246,14 +724,17 @@ static int symbol__parse_objdump_line(struct symbol *sym, struct map *map, offset = line_ip - start; if (offset < 0 || (u64)line_ip > end) offset = -1; + else + parsed_line = tmp2 + 1; } - objdump_line = objdump_line__new(offset, line, privsize); - if (objdump_line == NULL) { - free(line); + dl = disasm_line__new(offset, parsed_line, privsize); + free(line); + + if (dl == NULL) return -1; - } - objdump__add_line(¬es->src->source, objdump_line); + + disasm__add(¬es->src->source, dl); return 0; } @@ -408,7 +889,7 @@ static int symbol__get_source_line(struct symbol *sym, struct map *map, if (!notes->src->lines) return -1; - start = map->unmap_ip(map, sym->start); + start = map__rip_2objdump(map, sym->start); for (i = 0; i < len; i++) { char *path = NULL; @@ -476,7 +957,7 @@ static void symbol__annotate_hits(struct symbol *sym, int evidx) { struct annotation *notes = symbol__annotation(sym); struct sym_hist *h = annotation__histogram(notes, evidx); - u64 len = sym->end - sym->start, offset; + u64 len = symbol__size(sym), offset; for (offset = 0; offset < len; ++offset) if (h->addr[offset] != 0) @@ -492,7 +973,8 @@ int symbol__annotate_printf(struct symbol *sym, struct map *map, int evidx, struct dso *dso = map->dso; const char *filename = dso->long_name, *d_filename; struct annotation *notes = symbol__annotation(sym); - struct objdump_line *pos, *queue = NULL; + struct disasm_line *pos, *queue = NULL; + u64 start = map__rip_2objdump(map, sym->start); int printed = 2, queue_len = 0; int more = 0; u64 len; @@ -502,7 +984,7 @@ int symbol__annotate_printf(struct symbol *sym, struct map *map, int evidx, else d_filename = basename(filename); - len = sym->end - sym->start; + len = symbol__size(sym); printf(" Percent | Source code & Disassembly of %s\n", d_filename); printf("------------------------------------------------\n"); @@ -516,8 +998,9 @@ int symbol__annotate_printf(struct symbol *sym, struct map *map, int evidx, queue_len = 0; } - switch (objdump_line__print(pos, sym, evidx, len, min_pcnt, - printed, max_lines, queue)) { + switch (disasm_line__print(pos, sym, start, evidx, len, + min_pcnt, printed, max_lines, + queue)) { case 0: ++printed; if (context) { @@ -561,27 +1044,51 @@ void symbol__annotate_decay_histogram(struct symbol *sym, int evidx) { struct annotation *notes = symbol__annotation(sym); struct sym_hist *h = annotation__histogram(notes, evidx); - struct objdump_line *pos; - int len = sym->end - sym->start; + int len = symbol__size(sym), offset; h->sum = 0; - - list_for_each_entry(pos, ¬es->src->source, node) { - if (pos->offset != -1 && pos->offset < len) { - h->addr[pos->offset] = h->addr[pos->offset] * 7 / 8; - h->sum += h->addr[pos->offset]; - } + for (offset = 0; offset < len; ++offset) { + h->addr[offset] = h->addr[offset] * 7 / 8; + h->sum += h->addr[offset]; } } -void objdump_line_list__purge(struct list_head *head) +void disasm__purge(struct list_head *head) { - struct objdump_line *pos, *n; + struct disasm_line *pos, *n; list_for_each_entry_safe(pos, n, head, node) { list_del(&pos->node); - objdump_line__free(pos); + disasm_line__free(pos); + } +} + +static size_t disasm_line__fprintf(struct disasm_line *dl, FILE *fp) +{ + size_t printed; + + if (dl->offset == -1) + return fprintf(fp, "%s\n", dl->line); + + printed = fprintf(fp, "%#" PRIx64 " %s", dl->offset, dl->name); + + if (dl->ops.raw[0] != '\0') { + printed += fprintf(fp, "%.*s %s\n", 6 - (int)printed, " ", + dl->ops.raw); } + + return printed + fprintf(fp, "\n"); +} + +size_t disasm__fprintf(struct list_head *head, FILE *fp) +{ + struct disasm_line *pos; + size_t printed = 0; + + list_for_each_entry(pos, head, node) + printed += disasm_line__fprintf(pos, fp); + + return printed; } int symbol__tty_annotate(struct symbol *sym, struct map *map, int evidx, @@ -596,7 +1103,7 @@ int symbol__tty_annotate(struct symbol *sym, struct map *map, int evidx, if (symbol__annotate(sym, map, 0) < 0) return -1; - len = sym->end - sym->start; + len = symbol__size(sym); if (print_lines) { symbol__get_source_line(sym, map, evidx, &source_line, @@ -609,7 +1116,7 @@ int symbol__tty_annotate(struct symbol *sym, struct map *map, int evidx, if (print_lines) symbol__free_source_line(sym, len); - objdump_line_list__purge(&symbol__annotation(sym)->src->source); + disasm__purge(&symbol__annotation(sym)->src->source); return 0; } diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h index efa5dc82bfae..78a5692dd718 100644 --- a/tools/perf/util/annotate.h +++ b/tools/perf/util/annotate.h @@ -2,20 +2,69 @@ #define __PERF_ANNOTATE_H #include <stdbool.h> +#include <stdint.h> #include "types.h" #include "symbol.h" #include <linux/list.h> #include <linux/rbtree.h> -struct objdump_line { - struct list_head node; - s64 offset; - char *line; +struct ins; + +struct ins_operands { + char *raw; + struct { + char *raw; + char *name; + u64 addr; + u64 offset; + } target; + union { + struct { + char *raw; + char *name; + u64 addr; + } source; + struct { + struct ins *ins; + struct ins_operands *ops; + } locked; + }; }; -void objdump_line__free(struct objdump_line *self); -struct objdump_line *objdump__get_next_ip_line(struct list_head *head, - struct objdump_line *pos); +struct ins_ops { + void (*free)(struct ins_operands *ops); + int (*parse)(struct ins_operands *ops); + int (*scnprintf)(struct ins *ins, char *bf, size_t size, + struct ins_operands *ops); +}; + +struct ins { + const char *name; + struct ins_ops *ops; +}; + +bool ins__is_jump(const struct ins *ins); +bool ins__is_call(const struct ins *ins); +int ins__scnprintf(struct ins *ins, char *bf, size_t size, struct ins_operands *ops); + +struct disasm_line { + struct list_head node; + s64 offset; + char *line; + char *name; + struct ins *ins; + struct ins_operands ops; +}; + +static inline bool disasm_line__has_offset(const struct disasm_line *dl) +{ + return dl->ops.target.offset != UINT64_MAX; +} + +void disasm_line__free(struct disasm_line *dl); +struct disasm_line *disasm__get_next_ip_line(struct list_head *head, struct disasm_line *pos); +int disasm_line__scnprintf(struct disasm_line *dl, char *bf, size_t size, bool raw); +size_t disasm__fprintf(struct list_head *head, FILE *fp); struct sym_hist { u64 sum; @@ -32,7 +81,7 @@ struct source_line { * * @histogram: Array of addr hit histograms per event being monitored * @lines: If 'print_lines' is specified, per source code line percentages - * @source: source parsed from objdump -dS + * @source: source parsed from a disassembler like objdump -dS * * lines is allocated, percentages calculated and all sorted by percentage * when the annotation is about to be presented, so the percentages are for @@ -82,7 +131,7 @@ int symbol__annotate_printf(struct symbol *sym, struct map *map, int evidx, int context); void symbol__annotate_zero_histogram(struct symbol *sym, int evidx); void symbol__annotate_decay_histogram(struct symbol *sym, int evidx); -void objdump_line_list__purge(struct list_head *head); +void disasm__purge(struct list_head *head); int symbol__tty_annotate(struct symbol *sym, struct map *map, int evidx, bool print_lines, bool full_paths, int min_pcnt, diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c index dff9c7a725f4..fd9a5944b627 100644 --- a/tools/perf/util/build-id.c +++ b/tools/perf/util/build-id.c @@ -65,6 +65,8 @@ struct perf_tool build_id__mark_dso_hit_ops = { .mmap = perf_event__process_mmap, .fork = perf_event__process_task, .exit = perf_event__exit_del_thread, + .attr = perf_event__process_attr, + .build_id = perf_event__process_build_id, }; char *dso__build_id_filename(struct dso *self, char *bf, size_t size) diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h index fc5e5a09d5b9..cff18c617d13 100644 --- a/tools/perf/util/cache.h +++ b/tools/perf/util/cache.h @@ -33,7 +33,7 @@ extern int pager_use_color; extern int use_browser; -#ifdef NO_NEWT_SUPPORT +#if defined(NO_NEWT_SUPPORT) && defined(NO_GTK2_SUPPORT) static inline void setup_browser(bool fallback_to_pager) { if (fallback_to_pager) @@ -43,7 +43,29 @@ static inline void exit_browser(bool wait_for_ok __used) {} #else void setup_browser(bool fallback_to_pager); void exit_browser(bool wait_for_ok); + +#ifdef NO_NEWT_SUPPORT +static inline int ui__init(void) +{ + return -1; +} +static inline void ui__exit(bool wait_for_ok __used) {} +#else +int ui__init(void); +void ui__exit(bool wait_for_ok); +#endif + +#ifdef NO_GTK2_SUPPORT +static inline int perf_gtk__init(void) +{ + return -1; +} +static inline void perf_gtk__exit(bool wait_for_ok __used) {} +#else +int perf_gtk__init(void); +void perf_gtk__exit(bool wait_for_ok); #endif +#endif /* NO_NEWT_SUPPORT && NO_GTK2_SUPPORT */ char *alias_lookup(const char *alias); int split_cmdline(char *cmdline, const char ***argv); diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c index 9f7106a8d9a4..3a6bff47614f 100644 --- a/tools/perf/util/callchain.c +++ b/tools/perf/util/callchain.c @@ -18,6 +18,8 @@ #include "util.h" #include "callchain.h" +__thread struct callchain_cursor callchain_cursor; + bool ip_callchain__valid(struct ip_callchain *chain, const union perf_event *event) { diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h index 7f9c0f1ae3a9..3bdb407f9cd9 100644 --- a/tools/perf/util/callchain.h +++ b/tools/perf/util/callchain.h @@ -76,6 +76,8 @@ struct callchain_cursor { struct callchain_cursor_node *curr; }; +extern __thread struct callchain_cursor callchain_cursor; + static inline void callchain_init(struct callchain_root *root) { INIT_LIST_HEAD(&root->node.siblings); diff --git a/tools/perf/util/config.c b/tools/perf/util/config.c index 0deac6a14b65..6faa3a18bfbd 100644 --- a/tools/perf/util/config.c +++ b/tools/perf/util/config.c @@ -120,7 +120,7 @@ static char *parse_value(void) static inline int iskeychar(int c) { - return isalnum(c) || c == '-'; + return isalnum(c) || c == '-' || c == '_'; } static int get_value(config_fn_t fn, void *data, char *name, unsigned int len) diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c index 26817daa2961..efb1fce259a4 100644 --- a/tools/perf/util/debug.c +++ b/tools/perf/util/debug.c @@ -11,6 +11,7 @@ #include "event.h" #include "debug.h" #include "util.h" +#include "target.h" int verbose; bool dump_trace = false, quiet = false; diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h index f2ce88d04f54..6bebe7f0a20c 100644 --- a/tools/perf/util/debug.h +++ b/tools/perf/util/debug.h @@ -26,7 +26,7 @@ static inline void ui_progress__update(u64 curr __used, u64 total __used, #else extern char ui_helpline__last_msg[]; int ui_helpline__show_help(const char *format, va_list ap); -#include "ui/progress.h" +#include "../ui/progress.h" int ui__error(const char *format, ...) __attribute__((format(printf, 1, 2))); #endif diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index 159263d17c2d..7400fb3fc50c 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -11,6 +11,7 @@ #include <poll.h> #include "cpumap.h" #include "thread_map.h" +#include "target.h" #include "evlist.h" #include "evsel.h" #include <unistd.h> @@ -51,13 +52,15 @@ struct perf_evlist *perf_evlist__new(struct cpu_map *cpus, void perf_evlist__config_attrs(struct perf_evlist *evlist, struct perf_record_opts *opts) { - struct perf_evsel *evsel; + struct perf_evsel *evsel, *first; if (evlist->cpus->map[0] < 0) opts->no_inherit = true; + first = list_entry(evlist->entries.next, struct perf_evsel, node); + list_for_each_entry(evsel, &evlist->entries, node) { - perf_evsel__config(evsel, opts); + perf_evsel__config(evsel, opts, first); if (evlist->nr_entries > 1) evsel->attr.sample_type |= PERF_SAMPLE_ID; @@ -156,6 +159,17 @@ out_delete_partial_list: return -1; } +int __perf_evlist__add_default_attrs(struct perf_evlist *evlist, + struct perf_event_attr *attrs, size_t nr_attrs) +{ + size_t i; + + for (i = 0; i < nr_attrs; i++) + event_attr_init(attrs + i); + + return perf_evlist__add_attrs(evlist, attrs, nr_attrs); +} + static int trace_event__id(const char *evname) { char *filename, *colon; @@ -260,7 +274,8 @@ void perf_evlist__disable(struct perf_evlist *evlist) for (cpu = 0; cpu < evlist->cpus->nr; cpu++) { list_for_each_entry(pos, &evlist->entries, node) { for (thread = 0; thread < evlist->threads->nr; thread++) - ioctl(FD(pos, cpu, thread), PERF_EVENT_IOC_DISABLE); + ioctl(FD(pos, cpu, thread), + PERF_EVENT_IOC_DISABLE, 0); } } } @@ -273,7 +288,8 @@ void perf_evlist__enable(struct perf_evlist *evlist) for (cpu = 0; cpu < evlist->cpus->nr; cpu++) { list_for_each_entry(pos, &evlist->entries, node) { for (thread = 0; thread < evlist->threads->nr; thread++) - ioctl(FD(pos, cpu, thread), PERF_EVENT_IOC_ENABLE); + ioctl(FD(pos, cpu, thread), + PERF_EVENT_IOC_ENABLE, 0); } } } @@ -597,18 +613,21 @@ int perf_evlist__mmap(struct perf_evlist *evlist, unsigned int pages, return perf_evlist__mmap_per_cpu(evlist, prot, mask); } -int perf_evlist__create_maps(struct perf_evlist *evlist, const char *target_pid, - const char *target_tid, uid_t uid, const char *cpu_list) +int perf_evlist__create_maps(struct perf_evlist *evlist, + struct perf_target *target) { - evlist->threads = thread_map__new_str(target_pid, target_tid, uid); + evlist->threads = thread_map__new_str(target->pid, target->tid, + target->uid); if (evlist->threads == NULL) return -1; - if (uid != UINT_MAX || (cpu_list == NULL && target_tid)) + if (perf_target__has_task(target)) + evlist->cpus = cpu_map__dummy_new(); + else if (!perf_target__has_cpu(target) && !target->uses_mmap) evlist->cpus = cpu_map__dummy_new(); else - evlist->cpus = cpu_map__new(cpu_list); + evlist->cpus = cpu_map__new(target->cpu_list); if (evlist->cpus == NULL) goto out_delete_threads; @@ -825,7 +844,7 @@ int perf_evlist__prepare_workload(struct perf_evlist *evlist, exit(-1); } - if (!opts->system_wide && !opts->target_tid && !opts->target_pid) + if (perf_target__none(&opts->target)) evlist->threads->map[0] = evlist->workload.pid; close(child_ready_pipe[1]); diff --git a/tools/perf/util/evlist.h b/tools/perf/util/evlist.h index 21f1c9e57f13..989bee9624c2 100644 --- a/tools/perf/util/evlist.h +++ b/tools/perf/util/evlist.h @@ -54,6 +54,8 @@ void perf_evlist__add(struct perf_evlist *evlist, struct perf_evsel *entry); int perf_evlist__add_default(struct perf_evlist *evlist); int perf_evlist__add_attrs(struct perf_evlist *evlist, struct perf_event_attr *attrs, size_t nr_attrs); +int __perf_evlist__add_default_attrs(struct perf_evlist *evlist, + struct perf_event_attr *attrs, size_t nr_attrs); int perf_evlist__add_tracepoints(struct perf_evlist *evlist, const char *tracepoints[], size_t nr_tracepoints); int perf_evlist__set_tracepoints_handlers(struct perf_evlist *evlist, @@ -62,6 +64,8 @@ int perf_evlist__set_tracepoints_handlers(struct perf_evlist *evlist, #define perf_evlist__add_attrs_array(evlist, array) \ perf_evlist__add_attrs(evlist, array, ARRAY_SIZE(array)) +#define perf_evlist__add_default_attrs(evlist, array) \ + __perf_evlist__add_default_attrs(evlist, array, ARRAY_SIZE(array)) #define perf_evlist__add_tracepoints_array(evlist, array) \ perf_evlist__add_tracepoints(evlist, array, ARRAY_SIZE(array)) @@ -106,8 +110,8 @@ static inline void perf_evlist__set_maps(struct perf_evlist *evlist, evlist->threads = threads; } -int perf_evlist__create_maps(struct perf_evlist *evlist, const char *target_pid, - const char *tid, uid_t uid, const char *cpu_list); +int perf_evlist__create_maps(struct perf_evlist *evlist, + struct perf_target *target); void perf_evlist__delete_maps(struct perf_evlist *evlist); int perf_evlist__set_filters(struct perf_evlist *evlist); diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index f421f7cbc0d3..9f6cebd798ee 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -14,6 +14,8 @@ #include "util.h" #include "cpumap.h" #include "thread_map.h" +#include "target.h" +#include "../../include/linux/perf_event.h" #define FD(e, x, y) (*(int *)xyarray__entry(e->fd, x, y)) #define GROUP_FD(group_fd, cpu) (*(int *)xyarray__entry(group_fd, cpu, 0)) @@ -34,7 +36,7 @@ int __perf_evsel__sample_size(u64 sample_type) return size; } -static void hists__init(struct hists *hists) +void hists__init(struct hists *hists) { memset(hists, 0, sizeof(*hists)); hists->entries_in_array[0] = hists->entries_in_array[1] = RB_ROOT; @@ -63,11 +65,102 @@ struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr, int idx) return evsel; } -void perf_evsel__config(struct perf_evsel *evsel, struct perf_record_opts *opts) +static const char *perf_evsel__hw_names[PERF_COUNT_HW_MAX] = { + "cycles", + "instructions", + "cache-references", + "cache-misses", + "branches", + "branch-misses", + "bus-cycles", + "stalled-cycles-frontend", + "stalled-cycles-backend", + "ref-cycles", +}; + +const char *__perf_evsel__hw_name(u64 config) +{ + if (config < PERF_COUNT_HW_MAX && perf_evsel__hw_names[config]) + return perf_evsel__hw_names[config]; + + return "unknown-hardware"; +} + +static int perf_evsel__hw_name(struct perf_evsel *evsel, char *bf, size_t size) +{ + int colon = 0; + struct perf_event_attr *attr = &evsel->attr; + int r = scnprintf(bf, size, "%s", __perf_evsel__hw_name(attr->config)); + bool exclude_guest_default = false; + +#define MOD_PRINT(context, mod) do { \ + if (!attr->exclude_##context) { \ + if (!colon) colon = r++; \ + r += scnprintf(bf + r, size - r, "%c", mod); \ + } } while(0) + + if (attr->exclude_kernel || attr->exclude_user || attr->exclude_hv) { + MOD_PRINT(kernel, 'k'); + MOD_PRINT(user, 'u'); + MOD_PRINT(hv, 'h'); + exclude_guest_default = true; + } + + if (attr->precise_ip) { + if (!colon) + colon = r++; + r += scnprintf(bf + r, size - r, "%.*s", attr->precise_ip, "ppp"); + exclude_guest_default = true; + } + + if (attr->exclude_host || attr->exclude_guest == exclude_guest_default) { + MOD_PRINT(host, 'H'); + MOD_PRINT(guest, 'G'); + } +#undef MOD_PRINT + if (colon) + bf[colon] = ':'; + return r; +} + +int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size) +{ + int ret; + + switch (evsel->attr.type) { + case PERF_TYPE_RAW: + ret = scnprintf(bf, size, "raw 0x%" PRIx64, evsel->attr.config); + break; + + case PERF_TYPE_HARDWARE: + ret = perf_evsel__hw_name(evsel, bf, size); + break; + default: + /* + * FIXME + * + * This is the minimal perf_evsel__name so that we can + * reconstruct event names taking into account event modifiers. + * + * The old event_name uses it now for raw anr hw events, so that + * we don't drag all the parsing stuff into the python binding. + * + * On the next devel cycle the rest of the event naming will be + * brought here. + */ + return 0; + } + + return ret; +} + +void perf_evsel__config(struct perf_evsel *evsel, struct perf_record_opts *opts, + struct perf_evsel *first) { struct perf_event_attr *attr = &evsel->attr; int track = !evsel->idx; /* only the first counter needs these */ + attr->disabled = 1; attr->sample_id_all = opts->sample_id_all_missing ? 0 : 1; attr->inherit = !opts->no_inherit; attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED | @@ -105,15 +198,15 @@ void perf_evsel__config(struct perf_evsel *evsel, struct perf_record_opts *opts) if (opts->call_graph) attr->sample_type |= PERF_SAMPLE_CALLCHAIN; - if (opts->system_wide) + if (perf_target__has_cpu(&opts->target)) attr->sample_type |= PERF_SAMPLE_CPU; if (opts->period) attr->sample_type |= PERF_SAMPLE_PERIOD; if (!opts->sample_id_all_missing && - (opts->sample_time || opts->system_wide || - !opts->no_inherit || opts->cpu_list)) + (opts->sample_time || !opts->no_inherit || + perf_target__has_cpu(&opts->target))) attr->sample_type |= PERF_SAMPLE_TIME; if (opts->raw_samples) { @@ -134,8 +227,8 @@ void perf_evsel__config(struct perf_evsel *evsel, struct perf_record_opts *opts) attr->mmap = track; attr->comm = track; - if (!opts->target_pid && !opts->target_tid && !opts->system_wide) { - attr->disabled = 1; + if (perf_target__none(&opts->target) && + (!opts->group || evsel == first)) { attr->enable_on_exec = 1; } } @@ -401,16 +494,24 @@ int perf_evsel__open_per_thread(struct perf_evsel *evsel, } static int perf_event__parse_id_sample(const union perf_event *event, u64 type, - struct perf_sample *sample) + struct perf_sample *sample, + bool swapped) { const u64 *array = event->sample.array; + union u64_swap u; array += ((event->header.size - sizeof(event->header)) / sizeof(u64)) - 1; if (type & PERF_SAMPLE_CPU) { - u32 *p = (u32 *)array; - sample->cpu = *p; + u.val64 = *array; + if (swapped) { + /* undo swap of u64, then swap on individual u32s */ + u.val64 = bswap_64(u.val64); + u.val32[0] = bswap_32(u.val32[0]); + } + + sample->cpu = u.val32[0]; array--; } @@ -430,9 +531,16 @@ static int perf_event__parse_id_sample(const union perf_event *event, u64 type, } if (type & PERF_SAMPLE_TID) { - u32 *p = (u32 *)array; - sample->pid = p[0]; - sample->tid = p[1]; + u.val64 = *array; + if (swapped) { + /* undo swap of u64, then swap on individual u32s */ + u.val64 = bswap_64(u.val64); + u.val32[0] = bswap_32(u.val32[0]); + u.val32[1] = bswap_32(u.val32[1]); + } + + sample->pid = u.val32[0]; + sample->tid = u.val32[1]; } return 0; @@ -459,10 +567,7 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, * used for cross-endian analysis. See git commit 65014ab3 * for why this goofiness is needed. */ - union { - u64 val64; - u32 val32[2]; - } u; + union u64_swap u; memset(data, 0, sizeof(*data)); data->cpu = data->pid = data->tid = -1; @@ -472,7 +577,7 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, if (event->header.type != PERF_RECORD_SAMPLE) { if (!sample_id_all) return 0; - return perf_event__parse_id_sample(event, type, data); + return perf_event__parse_id_sample(event, type, data, swapped); } array = event->sample.array; @@ -578,6 +683,8 @@ int perf_event__parse_sample(const union perf_event *event, u64 type, return -EFAULT; data->raw_data = (void *) pdata; + + array = (void *)array + data->raw_size + sizeof(u32); } if (type & PERF_SAMPLE_BRANCH_STACK) { @@ -603,10 +710,7 @@ int perf_event__synthesize_sample(union perf_event *event, u64 type, * used for cross-endian analysis. See git commit 65014ab3 * for why this goofiness is needed. */ - union { - u64 val64; - u32 val32[2]; - } u; + union u64_swap u; array = event->sample.array; diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h index 326b8e4d5035..4ba8b564e6f4 100644 --- a/tools/perf/util/evsel.h +++ b/tools/perf/util/evsel.h @@ -80,7 +80,11 @@ void perf_evsel__exit(struct perf_evsel *evsel); void perf_evsel__delete(struct perf_evsel *evsel); void perf_evsel__config(struct perf_evsel *evsel, - struct perf_record_opts *opts); + struct perf_record_opts *opts, + struct perf_evsel *first); + +const char* __perf_evsel__hw_name(u64 config); +int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size); int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads); int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads); @@ -169,4 +173,6 @@ static inline int perf_evsel__sample_size(struct perf_evsel *evsel) return __perf_evsel__sample_size(evsel->attr.sample_type); } +void hists__init(struct hists *hists); + #endif /* __PERF_EVSEL_H */ diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index fcd9cf3ea63e..2dd5edf161b7 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -31,21 +31,16 @@ static const char **header_argv; int perf_header__push_event(u64 id, const char *name) { + struct perf_trace_event_type *nevents; + if (strlen(name) > MAX_EVENT_NAME) pr_warning("Event %s will be truncated\n", name); - if (!events) { - events = malloc(sizeof(struct perf_trace_event_type)); - if (events == NULL) - return -ENOMEM; - } else { - struct perf_trace_event_type *nevents; + nevents = realloc(events, (event_count + 1) * sizeof(*events)); + if (nevents == NULL) + return -ENOMEM; + events = nevents; - nevents = realloc(events, (event_count + 1) * sizeof(*events)); - if (nevents == NULL) - return -ENOMEM; - events = nevents; - } memset(&events[event_count], 0, sizeof(struct perf_trace_event_type)); events[event_count].event_id = id; strncpy(events[event_count].name, name, MAX_EVENT_NAME - 1); @@ -296,7 +291,7 @@ int build_id_cache__add_s(const char *sbuild_id, const char *debugdir, if (mkdir_p(filename, 0755)) goto out_free; - snprintf(filename + len, sizeof(filename) - len, "/%s", sbuild_id); + snprintf(filename + len, size - len, "/%s", sbuild_id); if (access(filename, F_OK)) { if (is_kallsyms) { @@ -442,7 +437,7 @@ static bool perf_session__read_build_ids(struct perf_session *session, bool with return ret; } -static int write_trace_info(int fd, struct perf_header *h __used, +static int write_tracing_data(int fd, struct perf_header *h __used, struct perf_evlist *evlist) { return read_tracing_data(fd, &evlist->entries); @@ -1177,7 +1172,7 @@ static void print_event_desc(struct perf_header *ph, int fd, FILE *fp) goto error; msz = sizeof(attr); - if (sz < (ssize_t)msz) + if (sz < msz) msz = sz; for (i = 0 ; i < nre; i++) { @@ -1477,7 +1472,7 @@ out: return err; } -static int process_trace_info(struct perf_file_section *section __unused, +static int process_tracing_data(struct perf_file_section *section __unused, struct perf_header *ph __unused, int feat __unused, int fd) { @@ -1513,11 +1508,11 @@ struct feature_ops { .full_only = true } /* feature_ops not implemented: */ -#define print_trace_info NULL -#define print_build_id NULL +#define print_tracing_data NULL +#define print_build_id NULL static const struct feature_ops feat_ops[HEADER_LAST_FEATURE] = { - FEAT_OPP(HEADER_TRACE_INFO, trace_info), + FEAT_OPP(HEADER_TRACING_DATA, tracing_data), FEAT_OPP(HEADER_BUILD_ID, build_id), FEAT_OPA(HEADER_HOSTNAME, hostname), FEAT_OPA(HEADER_OSRELEASE, osrelease), diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index 21a6be09c129..2d42b3e1826f 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -12,7 +12,7 @@ enum { HEADER_RESERVED = 0, /* always cleared */ HEADER_FIRST_FEATURE = 1, - HEADER_TRACE_INFO = 1, + HEADER_TRACING_DATA = 1, HEADER_BUILD_ID, HEADER_HOSTNAME, diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index 3dc99a9b71f5..514e2a4b367d 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -10,11 +10,14 @@ static bool hists__filter_entry_by_dso(struct hists *hists, struct hist_entry *he); static bool hists__filter_entry_by_thread(struct hists *hists, struct hist_entry *he); +static bool hists__filter_entry_by_symbol(struct hists *hists, + struct hist_entry *he); enum hist_filter { HIST_FILTER__DSO, HIST_FILTER__THREAD, HIST_FILTER__PARENT, + HIST_FILTER__SYMBOL, }; struct callchain_param callchain_param = { @@ -253,6 +256,18 @@ static struct hist_entry *add_hist_entry(struct hists *hists, if (!cmp) { he->period += period; ++he->nr_events; + + /* If the map of an existing hist_entry has + * become out-of-date due to an exec() or + * similar, update it. Otherwise we will + * mis-adjust symbol addresses when computing + * the history counter to increment. + */ + if (he->ms.map != entry->ms.map) { + he->ms.map = entry->ms.map; + if (he->ms.map) + he->ms.map->referenced = true; + } goto out; } @@ -363,7 +378,7 @@ void hist_entry__free(struct hist_entry *he) * collapse the histogram */ -static bool hists__collapse_insert_entry(struct hists *hists, +static bool hists__collapse_insert_entry(struct hists *hists __used, struct rb_root *root, struct hist_entry *he) { @@ -382,8 +397,9 @@ static bool hists__collapse_insert_entry(struct hists *hists, iter->period += he->period; iter->nr_events += he->nr_events; if (symbol_conf.use_callchain) { - callchain_cursor_reset(&hists->callchain_cursor); - callchain_merge(&hists->callchain_cursor, iter->callchain, + callchain_cursor_reset(&callchain_cursor); + callchain_merge(&callchain_cursor, + iter->callchain, he->callchain); } hist_entry__free(he); @@ -420,6 +436,7 @@ static void hists__apply_filters(struct hists *hists, struct hist_entry *he) { hists__filter_entry_by_dso(hists, he); hists__filter_entry_by_thread(hists, he); + hists__filter_entry_by_symbol(hists, he); } static void __hists__collapse_resort(struct hists *hists, bool threaded) @@ -583,7 +600,7 @@ static size_t ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, if (chain->ms.sym) ret += fprintf(fp, "%s\n", chain->ms.sym->name); else - ret += fprintf(fp, "%p\n", (void *)(long)chain->ip); + ret += fprintf(fp, "0x%0" PRIx64 "\n", chain->ip); return ret; } @@ -603,7 +620,7 @@ static void init_rem_hits(void) rem_hits.ms.sym = rem_sq_bracket; } -static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, +static size_t __callchain__fprintf_graph(FILE *fp, struct rb_root *root, u64 total_samples, int depth, int depth_mask, int left_margin) { @@ -611,21 +628,16 @@ static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, struct callchain_node *child; struct callchain_list *chain; int new_depth_mask = depth_mask; - u64 new_total; u64 remaining; size_t ret = 0; int i; uint entries_printed = 0; - if (callchain_param.mode == CHAIN_GRAPH_REL) - new_total = self->children_hit; - else - new_total = total_samples; - - remaining = new_total; + remaining = total_samples; - node = rb_first(&self->rb_root); + node = rb_first(root); while (node) { + u64 new_total; u64 cumul; child = rb_entry(node, struct callchain_node, rb_node); @@ -653,11 +665,17 @@ static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, list_for_each_entry(chain, &child->val, list) { ret += ipchain__fprintf_graph(fp, chain, depth, new_depth_mask, i++, - new_total, + total_samples, cumul, left_margin); } - ret += __callchain__fprintf_graph(fp, child, new_total, + + if (callchain_param.mode == CHAIN_GRAPH_REL) + new_total = child->children_hit; + else + new_total = total_samples; + + ret += __callchain__fprintf_graph(fp, &child->rb_root, new_total, depth + 1, new_depth_mask | (1 << depth), left_margin); @@ -667,61 +685,75 @@ static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, } if (callchain_param.mode == CHAIN_GRAPH_REL && - remaining && remaining != new_total) { + remaining && remaining != total_samples) { if (!rem_sq_bracket) return ret; new_depth_mask &= ~(1 << (depth - 1)); - ret += ipchain__fprintf_graph(fp, &rem_hits, depth, - new_depth_mask, 0, new_total, + new_depth_mask, 0, total_samples, remaining, left_margin); } return ret; } -static size_t callchain__fprintf_graph(FILE *fp, struct callchain_node *self, +static size_t callchain__fprintf_graph(FILE *fp, struct rb_root *root, u64 total_samples, int left_margin) { + struct callchain_node *cnode; struct callchain_list *chain; + u32 entries_printed = 0; bool printed = false; + struct rb_node *node; int i = 0; - int ret = 0; - u32 entries_printed = 0; - - list_for_each_entry(chain, &self->val, list) { - if (!i++ && sort__first_dimension == SORT_SYM) - continue; - - if (!printed) { - ret += callchain__fprintf_left_margin(fp, left_margin); - ret += fprintf(fp, "|\n"); - ret += callchain__fprintf_left_margin(fp, left_margin); - ret += fprintf(fp, "---"); - - left_margin += 3; - printed = true; - } else - ret += callchain__fprintf_left_margin(fp, left_margin); + int ret; - if (chain->ms.sym) - ret += fprintf(fp, " %s\n", chain->ms.sym->name); - else - ret += fprintf(fp, " %p\n", (void *)(long)chain->ip); + /* + * If have one single callchain root, don't bother printing + * its percentage (100 % in fractal mode and the same percentage + * than the hist in graph mode). This also avoid one level of column. + */ + node = rb_first(root); + if (node && !rb_next(node)) { + cnode = rb_entry(node, struct callchain_node, rb_node); + list_for_each_entry(chain, &cnode->val, list) { + /* + * If we sort by symbol, the first entry is the same than + * the symbol. No need to print it otherwise it appears as + * displayed twice. + */ + if (!i++ && sort__first_dimension == SORT_SYM) + continue; + if (!printed) { + ret += callchain__fprintf_left_margin(fp, left_margin); + ret += fprintf(fp, "|\n"); + ret += callchain__fprintf_left_margin(fp, left_margin); + ret += fprintf(fp, "---"); + left_margin += 3; + printed = true; + } else + ret += callchain__fprintf_left_margin(fp, left_margin); + + if (chain->ms.sym) + ret += fprintf(fp, " %s\n", chain->ms.sym->name); + else + ret += fprintf(fp, " %p\n", (void *)(long)chain->ip); - if (++entries_printed == callchain_param.print_limit) - break; + if (++entries_printed == callchain_param.print_limit) + break; + } + root = &cnode->rb_root; } - ret += __callchain__fprintf_graph(fp, self, total_samples, 1, 1, left_margin); - - return ret; + return __callchain__fprintf_graph(fp, root, total_samples, + 1, 1, left_margin); } -static size_t callchain__fprintf_flat(FILE *fp, struct callchain_node *self, - u64 total_samples) +static size_t __callchain__fprintf_flat(FILE *fp, + struct callchain_node *self, + u64 total_samples) { struct callchain_list *chain; size_t ret = 0; @@ -729,7 +761,7 @@ static size_t callchain__fprintf_flat(FILE *fp, struct callchain_node *self, if (!self) return 0; - ret += callchain__fprintf_flat(fp, self->parent, total_samples); + ret += __callchain__fprintf_flat(fp, self->parent, total_samples); list_for_each_entry(chain, &self->val, list) { @@ -745,44 +777,58 @@ static size_t callchain__fprintf_flat(FILE *fp, struct callchain_node *self, return ret; } -static size_t hist_entry_callchain__fprintf(struct hist_entry *he, - u64 total_samples, int left_margin, - FILE *fp) +static size_t callchain__fprintf_flat(FILE *fp, struct rb_root *self, + u64 total_samples) { - struct rb_node *rb_node; - struct callchain_node *chain; size_t ret = 0; u32 entries_printed = 0; + struct rb_node *rb_node; + struct callchain_node *chain; - rb_node = rb_first(&he->sorted_chain); + rb_node = rb_first(self); while (rb_node) { double percent; chain = rb_entry(rb_node, struct callchain_node, rb_node); percent = chain->hit * 100.0 / total_samples; - switch (callchain_param.mode) { - case CHAIN_FLAT: - ret += percent_color_fprintf(fp, " %6.2f%%\n", - percent); - ret += callchain__fprintf_flat(fp, chain, total_samples); - break; - case CHAIN_GRAPH_ABS: /* Falldown */ - case CHAIN_GRAPH_REL: - ret += callchain__fprintf_graph(fp, chain, total_samples, - left_margin); - case CHAIN_NONE: - default: - break; - } + + ret = percent_color_fprintf(fp, " %6.2f%%\n", percent); + ret += __callchain__fprintf_flat(fp, chain, total_samples); ret += fprintf(fp, "\n"); if (++entries_printed == callchain_param.print_limit) break; + rb_node = rb_next(rb_node); } return ret; } +static size_t hist_entry_callchain__fprintf(struct hist_entry *he, + u64 total_samples, int left_margin, + FILE *fp) +{ + switch (callchain_param.mode) { + case CHAIN_GRAPH_REL: + return callchain__fprintf_graph(fp, &he->sorted_chain, he->period, + left_margin); + break; + case CHAIN_GRAPH_ABS: + return callchain__fprintf_graph(fp, &he->sorted_chain, total_samples, + left_margin); + break; + case CHAIN_FLAT: + return callchain__fprintf_flat(fp, &he->sorted_chain, total_samples); + break; + case CHAIN_NONE: + break; + default: + pr_err("Bad callchain mode\n"); + } + + return 0; +} + void hists__output_recalc_col_len(struct hists *hists, int max_rows) { struct rb_node *next = rb_first(&hists->entries); @@ -887,9 +933,9 @@ static int hist_entry__pcnt_snprintf(struct hist_entry *he, char *s, diff = new_percent - old_percent; if (fabs(diff) >= 0.01) - ret += scnprintf(bf, sizeof(bf), "%+4.2F%%", diff); + scnprintf(bf, sizeof(bf), "%+4.2F%%", diff); else - ret += scnprintf(bf, sizeof(bf), " "); + scnprintf(bf, sizeof(bf), " "); if (sep) ret += scnprintf(s + ret, size - ret, "%c%s", *sep, bf); @@ -898,9 +944,9 @@ static int hist_entry__pcnt_snprintf(struct hist_entry *he, char *s, if (show_displacement) { if (displacement) - ret += scnprintf(bf, sizeof(bf), "%+4ld", displacement); + scnprintf(bf, sizeof(bf), "%+4ld", displacement); else - ret += scnprintf(bf, sizeof(bf), " "); + scnprintf(bf, sizeof(bf), " "); if (sep) ret += scnprintf(s + ret, size - ret, "%c%s", *sep, bf); @@ -1247,6 +1293,37 @@ void hists__filter_by_thread(struct hists *hists) } } +static bool hists__filter_entry_by_symbol(struct hists *hists, + struct hist_entry *he) +{ + if (hists->symbol_filter_str != NULL && + (!he->ms.sym || strstr(he->ms.sym->name, + hists->symbol_filter_str) == NULL)) { + he->filtered |= (1 << HIST_FILTER__SYMBOL); + return true; + } + + return false; +} + +void hists__filter_by_symbol(struct hists *hists) +{ + struct rb_node *nd; + + hists->nr_entries = hists->stats.total_period = 0; + hists->stats.nr_events[PERF_RECORD_SAMPLE] = 0; + hists__reset_col_len(hists); + + for (nd = rb_first(&hists->entries); nd; nd = rb_next(nd)) { + struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); + + if (hists__filter_entry_by_symbol(hists, h)) + continue; + + hists__remove_entry_filter(hists, h, HIST_FILTER__SYMBOL); + } +} + int hist_entry__inc_addr_samples(struct hist_entry *he, int evidx, u64 ip) { return symbol__inc_addr_samples(he->ms.sym, he->ms.map, evidx, ip); diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h index 9413f3e31fea..34bb556d6219 100644 --- a/tools/perf/util/hist.h +++ b/tools/perf/util/hist.h @@ -62,12 +62,11 @@ struct hists { const struct thread *thread_filter; const struct dso *dso_filter; const char *uid_filter_str; + const char *symbol_filter_str; pthread_mutex_t lock; struct events_stats stats; u64 event_stream; u16 col_len[HISTC_NR_COLS]; - /* Best would be to reuse the session callchain cursor */ - struct callchain_cursor callchain_cursor; }; struct hist_entry *__hists__add_entry(struct hists *self, @@ -107,6 +106,7 @@ int hist_entry__annotate(struct hist_entry *self, size_t privsize); void hists__filter_by_dso(struct hists *hists); void hists__filter_by_thread(struct hists *hists); +void hists__filter_by_symbol(struct hists *hists); u16 hists__col_len(struct hists *self, enum hist_column col); void hists__set_col_len(struct hists *self, enum hist_column col, u16 len); @@ -136,7 +136,7 @@ static inline int hist_entry__tui_annotate(struct hist_entry *self __used, #define K_LEFT -1 #define K_RIGHT -2 #else -#include "ui/keysyms.h" +#include "../ui/keysyms.h" int hist_entry__tui_annotate(struct hist_entry *he, int evidx, void(*timer)(void *arg), void *arg, int delay_secs); @@ -145,6 +145,23 @@ int perf_evlist__tui_browse_hists(struct perf_evlist *evlist, const char *help, int refresh); #endif +#ifdef NO_GTK2_SUPPORT +static inline +int perf_evlist__gtk_browse_hists(struct perf_evlist *evlist __used, + const char *help __used, + void(*timer)(void *arg) __used, + void *arg __used, + int refresh __used) +{ + return 0; +} + +#else +int perf_evlist__gtk_browse_hists(struct perf_evlist *evlist, const char *help, + void(*timer)(void *arg), void *arg, + int refresh); +#endif + unsigned int hists__sort_list_width(struct hists *self); #endif /* __PERF_HIST_H */ diff --git a/tools/perf/util/include/linux/module.h b/tools/perf/util/include/linux/export.h index b43e2dc21e04..b43e2dc21e04 100644 --- a/tools/perf/util/include/linux/module.h +++ b/tools/perf/util/include/linux/export.h diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index dea6d1c1a954..35ae56864e4f 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -38,6 +38,7 @@ void map__init(struct map *self, enum map_type type, RB_CLEAR_NODE(&self->rb_node); self->groups = NULL; self->referenced = false; + self->erange_warned = false; } struct map *map__new(struct list_head *dsos__list, u64 start, u64 len, diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index b100c20b7f94..81371bad4ef0 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -33,6 +33,7 @@ struct map { u64 end; u8 /* enum map_type */ type; bool referenced; + bool erange_warned; u32 priv; u64 pgoff; diff --git a/tools/perf/util/pager.c b/tools/perf/util/pager.c index 1915de20dcac..3322b8446e89 100644 --- a/tools/perf/util/pager.c +++ b/tools/perf/util/pager.c @@ -57,6 +57,10 @@ void setup_pager(void) } if (!pager) pager = getenv("PAGER"); + if (!pager) { + if (!access("/usr/bin/pager", X_OK)) + pager = "/usr/bin/pager"; + } if (!pager) pager = "less"; else if (!*pager || !strcmp(pager, "cat")) diff --git a/tools/perf/util/parse-events-test.c b/tools/perf/util/parse-events-test.c new file mode 100644 index 000000000000..76b98e2a587d --- /dev/null +++ b/tools/perf/util/parse-events-test.c @@ -0,0 +1,625 @@ + +#include "parse-events.h" +#include "evsel.h" +#include "evlist.h" +#include "sysfs.h" +#include "../../../include/linux/hw_breakpoint.h" + +#define TEST_ASSERT_VAL(text, cond) \ +do { \ + if (!(cond)) { \ + pr_debug("FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ + return -1; \ + } \ +} while (0) + +static int test__checkevent_tracepoint(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_TRACEPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong sample_type", + (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME | PERF_SAMPLE_CPU) == + evsel->attr.sample_type); + TEST_ASSERT_VAL("wrong sample_period", 1 == evsel->attr.sample_period); + return 0; +} + +static int test__checkevent_tracepoint_multi(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + TEST_ASSERT_VAL("wrong number of entries", evlist->nr_entries > 1); + + list_for_each_entry(evsel, &evlist->entries, node) { + TEST_ASSERT_VAL("wrong type", + PERF_TYPE_TRACEPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong sample_type", + (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME | PERF_SAMPLE_CPU) + == evsel->attr.sample_type); + TEST_ASSERT_VAL("wrong sample_period", + 1 == evsel->attr.sample_period); + } + return 0; +} + +static int test__checkevent_raw(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 0x1a == evsel->attr.config); + return 0; +} + +static int test__checkevent_numeric(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", 1 == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 1 == evsel->attr.config); + return 0; +} + +static int test__checkevent_symbolic_name(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_HARDWARE == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", + PERF_COUNT_HW_INSTRUCTIONS == evsel->attr.config); + return 0; +} + +static int test__checkevent_symbolic_name_config(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_HARDWARE == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", + PERF_COUNT_HW_CPU_CYCLES == evsel->attr.config); + TEST_ASSERT_VAL("wrong period", + 100000 == evsel->attr.sample_period); + TEST_ASSERT_VAL("wrong config1", + 0 == evsel->attr.config1); + TEST_ASSERT_VAL("wrong config2", + 1 == evsel->attr.config2); + return 0; +} + +static int test__checkevent_symbolic_alias(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_SOFTWARE == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", + PERF_COUNT_SW_PAGE_FAULTS == evsel->attr.config); + return 0; +} + +static int test__checkevent_genhw(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_HW_CACHE == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", (1 << 16) == evsel->attr.config); + return 0; +} + +static int test__checkevent_breakpoint(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_BREAKPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); + TEST_ASSERT_VAL("wrong bp_type", (HW_BREAKPOINT_R | HW_BREAKPOINT_W) == + evsel->attr.bp_type); + TEST_ASSERT_VAL("wrong bp_len", HW_BREAKPOINT_LEN_4 == + evsel->attr.bp_len); + return 0; +} + +static int test__checkevent_breakpoint_x(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_BREAKPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); + TEST_ASSERT_VAL("wrong bp_type", + HW_BREAKPOINT_X == evsel->attr.bp_type); + TEST_ASSERT_VAL("wrong bp_len", sizeof(long) == evsel->attr.bp_len); + return 0; +} + +static int test__checkevent_breakpoint_r(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", + PERF_TYPE_BREAKPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); + TEST_ASSERT_VAL("wrong bp_type", + HW_BREAKPOINT_R == evsel->attr.bp_type); + TEST_ASSERT_VAL("wrong bp_len", + HW_BREAKPOINT_LEN_4 == evsel->attr.bp_len); + return 0; +} + +static int test__checkevent_breakpoint_w(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", + PERF_TYPE_BREAKPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 0 == evsel->attr.config); + TEST_ASSERT_VAL("wrong bp_type", + HW_BREAKPOINT_W == evsel->attr.bp_type); + TEST_ASSERT_VAL("wrong bp_len", + HW_BREAKPOINT_LEN_4 == evsel->attr.bp_len); + return 0; +} + +static int test__checkevent_tracepoint_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + return test__checkevent_tracepoint(evlist); +} + +static int +test__checkevent_tracepoint_multi_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + TEST_ASSERT_VAL("wrong number of entries", evlist->nr_entries > 1); + + list_for_each_entry(evsel, &evlist->entries, node) { + TEST_ASSERT_VAL("wrong exclude_user", + !evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", + evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + } + + return test__checkevent_tracepoint_multi(evlist); +} + +static int test__checkevent_raw_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); + + return test__checkevent_raw(evlist); +} + +static int test__checkevent_numeric_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); + + return test__checkevent_numeric(evlist); +} + +static int test__checkevent_symbolic_name_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + return test__checkevent_symbolic_name(evlist); +} + +static int test__checkevent_exclude_host_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude guest", !evsel->attr.exclude_guest); + TEST_ASSERT_VAL("wrong exclude host", evsel->attr.exclude_host); + + return test__checkevent_symbolic_name(evlist); +} + +static int test__checkevent_exclude_guest_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude guest", evsel->attr.exclude_guest); + TEST_ASSERT_VAL("wrong exclude host", !evsel->attr.exclude_host); + + return test__checkevent_symbolic_name(evlist); +} + +static int test__checkevent_symbolic_alias_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + return test__checkevent_symbolic_alias(evlist); +} + +static int test__checkevent_genhw_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); + + return test__checkevent_genhw(evlist); +} + +static int test__checkevent_breakpoint_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + return test__checkevent_breakpoint(evlist); +} + +static int test__checkevent_breakpoint_x_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + return test__checkevent_breakpoint_x(evlist); +} + +static int test__checkevent_breakpoint_r_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); + + return test__checkevent_breakpoint_r(evlist); +} + +static int test__checkevent_breakpoint_w_modifier(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); + + return test__checkevent_breakpoint_w(evlist); +} + +static int test__checkevent_pmu(struct perf_evlist *evlist) +{ + + struct perf_evsel *evsel = list_entry(evlist->entries.next, + struct perf_evsel, node); + + TEST_ASSERT_VAL("wrong number of entries", 1 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 10 == evsel->attr.config); + TEST_ASSERT_VAL("wrong config1", 1 == evsel->attr.config1); + TEST_ASSERT_VAL("wrong config2", 3 == evsel->attr.config2); + TEST_ASSERT_VAL("wrong period", 1000 == evsel->attr.sample_period); + + return 0; +} + +static int test__checkevent_list(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + TEST_ASSERT_VAL("wrong number of entries", 3 == evlist->nr_entries); + + /* r1 */ + evsel = list_entry(evlist->entries.next, struct perf_evsel, node); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 1 == evsel->attr.config); + TEST_ASSERT_VAL("wrong config1", 0 == evsel->attr.config1); + TEST_ASSERT_VAL("wrong config2", 0 == evsel->attr.config2); + TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + /* syscalls:sys_enter_open:k */ + evsel = list_entry(evsel->node.next, struct perf_evsel, node); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_TRACEPOINT == evsel->attr.type); + TEST_ASSERT_VAL("wrong sample_type", + (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME | PERF_SAMPLE_CPU) == + evsel->attr.sample_type); + TEST_ASSERT_VAL("wrong sample_period", 1 == evsel->attr.sample_period); + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip); + + /* 1:1:hp */ + evsel = list_entry(evsel->node.next, struct perf_evsel, node); + TEST_ASSERT_VAL("wrong type", 1 == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 1 == evsel->attr.config); + TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user); + TEST_ASSERT_VAL("wrong exclude_kernel", evsel->attr.exclude_kernel); + TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv); + TEST_ASSERT_VAL("wrong precise_ip", evsel->attr.precise_ip); + + return 0; +} + +static int test__checkevent_pmu_name(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + /* cpu/config=1,name=krava1/u */ + evsel = list_entry(evlist->entries.next, struct perf_evsel, node); + TEST_ASSERT_VAL("wrong number of entries", 2 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 1 == evsel->attr.config); + TEST_ASSERT_VAL("wrong name", !strcmp(evsel->name, "krava")); + + /* cpu/config=2/" */ + evsel = list_entry(evsel->node.next, struct perf_evsel, node); + TEST_ASSERT_VAL("wrong number of entries", 2 == evlist->nr_entries); + TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type); + TEST_ASSERT_VAL("wrong config", 2 == evsel->attr.config); + TEST_ASSERT_VAL("wrong name", !strcmp(evsel->name, "raw 0x2")); + + return 0; +} + +struct test__event_st { + const char *name; + __u32 type; + int (*check)(struct perf_evlist *evlist); +}; + +static struct test__event_st test__events[] = { + [0] = { + .name = "syscalls:sys_enter_open", + .check = test__checkevent_tracepoint, + }, + [1] = { + .name = "syscalls:*", + .check = test__checkevent_tracepoint_multi, + }, + [2] = { + .name = "r1a", + .check = test__checkevent_raw, + }, + [3] = { + .name = "1:1", + .check = test__checkevent_numeric, + }, + [4] = { + .name = "instructions", + .check = test__checkevent_symbolic_name, + }, + [5] = { + .name = "cycles/period=100000,config2/", + .check = test__checkevent_symbolic_name_config, + }, + [6] = { + .name = "faults", + .check = test__checkevent_symbolic_alias, + }, + [7] = { + .name = "L1-dcache-load-miss", + .check = test__checkevent_genhw, + }, + [8] = { + .name = "mem:0", + .check = test__checkevent_breakpoint, + }, + [9] = { + .name = "mem:0:x", + .check = test__checkevent_breakpoint_x, + }, + [10] = { + .name = "mem:0:r", + .check = test__checkevent_breakpoint_r, + }, + [11] = { + .name = "mem:0:w", + .check = test__checkevent_breakpoint_w, + }, + [12] = { + .name = "syscalls:sys_enter_open:k", + .check = test__checkevent_tracepoint_modifier, + }, + [13] = { + .name = "syscalls:*:u", + .check = test__checkevent_tracepoint_multi_modifier, + }, + [14] = { + .name = "r1a:kp", + .check = test__checkevent_raw_modifier, + }, + [15] = { + .name = "1:1:hp", + .check = test__checkevent_numeric_modifier, + }, + [16] = { + .name = "instructions:h", + .check = test__checkevent_symbolic_name_modifier, + }, + [17] = { + .name = "faults:u", + .check = test__checkevent_symbolic_alias_modifier, + }, + [18] = { + .name = "L1-dcache-load-miss:kp", + .check = test__checkevent_genhw_modifier, + }, + [19] = { + .name = "mem:0:u", + .check = test__checkevent_breakpoint_modifier, + }, + [20] = { + .name = "mem:0:x:k", + .check = test__checkevent_breakpoint_x_modifier, + }, + [21] = { + .name = "mem:0:r:hp", + .check = test__checkevent_breakpoint_r_modifier, + }, + [22] = { + .name = "mem:0:w:up", + .check = test__checkevent_breakpoint_w_modifier, + }, + [23] = { + .name = "r1,syscalls:sys_enter_open:k,1:1:hp", + .check = test__checkevent_list, + }, + [24] = { + .name = "instructions:G", + .check = test__checkevent_exclude_host_modifier, + }, + [25] = { + .name = "instructions:H", + .check = test__checkevent_exclude_guest_modifier, + }, +}; + +#define TEST__EVENTS_CNT (sizeof(test__events) / sizeof(struct test__event_st)) + +static struct test__event_st test__events_pmu[] = { + [0] = { + .name = "cpu/config=10,config1,config2=3,period=1000/u", + .check = test__checkevent_pmu, + }, + [1] = { + .name = "cpu/config=1,name=krava/u,cpu/config=2/u", + .check = test__checkevent_pmu_name, + }, +}; + +#define TEST__EVENTS_PMU_CNT (sizeof(test__events_pmu) / \ + sizeof(struct test__event_st)) + +static int test(struct test__event_st *e) +{ + struct perf_evlist *evlist; + int ret; + + evlist = perf_evlist__new(NULL, NULL); + if (evlist == NULL) + return -ENOMEM; + + ret = parse_events(evlist, e->name, 0); + if (ret) { + pr_debug("failed to parse event '%s', err %d\n", + e->name, ret); + return ret; + } + + ret = e->check(evlist); + perf_evlist__delete(evlist); + + return ret; +} + +static int test_events(struct test__event_st *events, unsigned cnt) +{ + int ret = 0; + unsigned i; + + for (i = 0; i < cnt; i++) { + struct test__event_st *e = &events[i]; + + pr_debug("running test %d '%s'\n", i, e->name); + ret = test(e); + if (ret) + break; + } + + return ret; +} + +static int test_pmu(void) +{ + struct stat st; + char path[PATH_MAX]; + int ret; + + snprintf(path, PATH_MAX, "%s/bus/event_source/devices/cpu/format/", + sysfs_find_mountpoint()); + + ret = stat(path, &st); + if (ret) + pr_debug("ommiting PMU cpu tests\n"); + return !ret; +} + +int parse_events__test(void) +{ + int ret; + + ret = test_events(test__events, TEST__EVENTS_CNT); + if (!ret && test_pmu()) + ret = test_events(test__events_pmu, TEST__EVENTS_PMU_CNT); + + return ret; +} diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index c7a6f6faf91e..05dbc8b3c767 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -11,6 +11,10 @@ #include "cache.h" #include "header.h" #include "debugfs.h" +#include "parse-events-flex.h" +#include "pmu.h" + +#define MAX_NAME_LEN 100 struct event_symbol { u8 type; @@ -19,11 +23,10 @@ struct event_symbol { const char *alias; }; -enum event_result { - EVT_FAILED, - EVT_HANDLED, - EVT_HANDLED_ALL -}; +#ifdef PARSER_DEBUG +extern int parse_events_debug; +#endif +int parse_events_parse(struct list_head *list, int *idx); #define CHW(x) .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_##x #define CSW(x) .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_##x @@ -59,19 +62,6 @@ static struct event_symbol event_symbols[] = { #define PERF_EVENT_TYPE(config) __PERF_EVENT_FIELD(config, TYPE) #define PERF_EVENT_ID(config) __PERF_EVENT_FIELD(config, EVENT) -static const char *hw_event_names[PERF_COUNT_HW_MAX] = { - "cycles", - "instructions", - "cache-references", - "cache-misses", - "branches", - "branch-misses", - "bus-cycles", - "stalled-cycles-frontend", - "stalled-cycles-backend", - "ref-cycles", -}; - static const char *sw_event_names[PERF_COUNT_SW_MAX] = { "cpu-clock", "task-clock", @@ -297,6 +287,16 @@ const char *event_name(struct perf_evsel *evsel) u64 config = evsel->attr.config; int type = evsel->attr.type; + if (type == PERF_TYPE_RAW || type == PERF_TYPE_HARDWARE) { + /* + * XXX minimal fix, see comment on perf_evsen__name, this static buffer + * will go away together with event_name in the next devel cycle. + */ + static char bf[128]; + perf_evsel__name(evsel, bf, sizeof(bf)); + return bf; + } + if (evsel->name) return evsel->name; @@ -314,9 +314,7 @@ const char *__event_name(int type, u64 config) switch (type) { case PERF_TYPE_HARDWARE: - if (config < PERF_COUNT_HW_MAX && hw_event_names[config]) - return hw_event_names[config]; - return "unknown-hardware"; + return __perf_evsel__hw_name(config); case PERF_TYPE_HW_CACHE: { u8 cache_type, cache_op, cache_result; @@ -354,7 +352,34 @@ const char *__event_name(int type, u64 config) return "unknown"; } -static int parse_aliases(const char **str, const char *names[][MAX_ALIASES], int size) +static int add_event(struct list_head **_list, int *idx, + struct perf_event_attr *attr, char *name) +{ + struct perf_evsel *evsel; + struct list_head *list = *_list; + + if (!list) { + list = malloc(sizeof(*list)); + if (!list) + return -ENOMEM; + INIT_LIST_HEAD(list); + } + + event_attr_init(attr); + + evsel = perf_evsel__new(attr, (*idx)++); + if (!evsel) { + free(list); + return -ENOMEM; + } + + evsel->name = strdup(name); + list_add_tail(&evsel->node, list); + *_list = list; + return 0; +} + +static int parse_aliases(char *str, const char *names[][MAX_ALIASES], int size) { int i, j; int n, longest = -1; @@ -362,58 +387,57 @@ static int parse_aliases(const char **str, const char *names[][MAX_ALIASES], int for (i = 0; i < size; i++) { for (j = 0; j < MAX_ALIASES && names[i][j]; j++) { n = strlen(names[i][j]); - if (n > longest && !strncasecmp(*str, names[i][j], n)) + if (n > longest && !strncasecmp(str, names[i][j], n)) longest = n; } - if (longest > 0) { - *str += longest; + if (longest > 0) return i; - } } return -1; } -static enum event_result -parse_generic_hw_event(const char **str, struct perf_event_attr *attr) +int parse_events_add_cache(struct list_head **list, int *idx, + char *type, char *op_result1, char *op_result2) { - const char *s = *str; + struct perf_event_attr attr; + char name[MAX_NAME_LEN]; int cache_type = -1, cache_op = -1, cache_result = -1; + char *op_result[2] = { op_result1, op_result2 }; + int i, n; - cache_type = parse_aliases(&s, hw_cache, PERF_COUNT_HW_CACHE_MAX); /* * No fallback - if we cannot get a clear cache type * then bail out: */ + cache_type = parse_aliases(type, hw_cache, + PERF_COUNT_HW_CACHE_MAX); if (cache_type == -1) - return EVT_FAILED; + return -EINVAL; + + n = snprintf(name, MAX_NAME_LEN, "%s", type); - while ((cache_op == -1 || cache_result == -1) && *s == '-') { - ++s; + for (i = 0; (i < 2) && (op_result[i]); i++) { + char *str = op_result[i]; + + snprintf(name + n, MAX_NAME_LEN - n, "-%s\n", str); if (cache_op == -1) { - cache_op = parse_aliases(&s, hw_cache_op, - PERF_COUNT_HW_CACHE_OP_MAX); + cache_op = parse_aliases(str, hw_cache_op, + PERF_COUNT_HW_CACHE_OP_MAX); if (cache_op >= 0) { if (!is_cache_op_valid(cache_type, cache_op)) - return EVT_FAILED; + return -EINVAL; continue; } } if (cache_result == -1) { - cache_result = parse_aliases(&s, hw_cache_result, + cache_result = parse_aliases(str, hw_cache_result, PERF_COUNT_HW_CACHE_RESULT_MAX); if (cache_result >= 0) continue; } - - /* - * Can't parse this as a cache op or result, so back up - * to the '-'. - */ - --s; - break; } /* @@ -428,20 +452,17 @@ parse_generic_hw_event(const char **str, struct perf_event_attr *attr) if (cache_result == -1) cache_result = PERF_COUNT_HW_CACHE_RESULT_ACCESS; - attr->config = cache_type | (cache_op << 8) | (cache_result << 16); - attr->type = PERF_TYPE_HW_CACHE; - - *str = s; - return EVT_HANDLED; + memset(&attr, 0, sizeof(attr)); + attr.config = cache_type | (cache_op << 8) | (cache_result << 16); + attr.type = PERF_TYPE_HW_CACHE; + return add_event(list, idx, &attr, name); } -static enum event_result -parse_single_tracepoint_event(char *sys_name, - const char *evt_name, - unsigned int evt_length, - struct perf_event_attr *attr, - const char **strp) +static int add_tracepoint(struct list_head **list, int *idx, + char *sys_name, char *evt_name) { + struct perf_event_attr attr; + char name[MAX_NAME_LEN]; char evt_path[MAXPATHLEN]; char id_buf[4]; u64 id; @@ -452,130 +473,80 @@ parse_single_tracepoint_event(char *sys_name, fd = open(evt_path, O_RDONLY); if (fd < 0) - return EVT_FAILED; + return -1; if (read(fd, id_buf, sizeof(id_buf)) < 0) { close(fd); - return EVT_FAILED; + return -1; } close(fd); id = atoll(id_buf); - attr->config = id; - attr->type = PERF_TYPE_TRACEPOINT; - *strp += strlen(sys_name) + evt_length + 1; /* + 1 for the ':' */ - - attr->sample_type |= PERF_SAMPLE_RAW; - attr->sample_type |= PERF_SAMPLE_TIME; - attr->sample_type |= PERF_SAMPLE_CPU; - attr->sample_period = 1; + memset(&attr, 0, sizeof(attr)); + attr.config = id; + attr.type = PERF_TYPE_TRACEPOINT; + attr.sample_type |= PERF_SAMPLE_RAW; + attr.sample_type |= PERF_SAMPLE_TIME; + attr.sample_type |= PERF_SAMPLE_CPU; + attr.sample_period = 1; - - return EVT_HANDLED; + snprintf(name, MAX_NAME_LEN, "%s:%s", sys_name, evt_name); + return add_event(list, idx, &attr, name); } -/* sys + ':' + event + ':' + flags*/ -#define MAX_EVOPT_LEN (MAX_EVENT_LENGTH * 2 + 2 + 128) -static enum event_result -parse_multiple_tracepoint_event(struct perf_evlist *evlist, char *sys_name, - const char *evt_exp, char *flags) +static int add_tracepoint_multi(struct list_head **list, int *idx, + char *sys_name, char *evt_name) { char evt_path[MAXPATHLEN]; struct dirent *evt_ent; DIR *evt_dir; + int ret = 0; snprintf(evt_path, MAXPATHLEN, "%s/%s", tracing_events_path, sys_name); evt_dir = opendir(evt_path); - if (!evt_dir) { perror("Can't open event dir"); - return EVT_FAILED; + return -1; } - while ((evt_ent = readdir(evt_dir))) { - char event_opt[MAX_EVOPT_LEN + 1]; - int len; - + while (!ret && (evt_ent = readdir(evt_dir))) { if (!strcmp(evt_ent->d_name, ".") || !strcmp(evt_ent->d_name, "..") || !strcmp(evt_ent->d_name, "enable") || !strcmp(evt_ent->d_name, "filter")) continue; - if (!strglobmatch(evt_ent->d_name, evt_exp)) + if (!strglobmatch(evt_ent->d_name, evt_name)) continue; - len = snprintf(event_opt, MAX_EVOPT_LEN, "%s:%s%s%s", sys_name, - evt_ent->d_name, flags ? ":" : "", - flags ?: ""); - if (len < 0) - return EVT_FAILED; - - if (parse_events(evlist, event_opt, 0)) - return EVT_FAILED; + ret = add_tracepoint(list, idx, sys_name, evt_ent->d_name); } - return EVT_HANDLED_ALL; + return ret; } -static enum event_result -parse_tracepoint_event(struct perf_evlist *evlist, const char **strp, - struct perf_event_attr *attr) +int parse_events_add_tracepoint(struct list_head **list, int *idx, + char *sys, char *event) { - const char *evt_name; - char *flags = NULL, *comma_loc; - char sys_name[MAX_EVENT_LENGTH]; - unsigned int sys_length, evt_length; - - if (debugfs_valid_mountpoint(tracing_events_path)) - return 0; - - evt_name = strchr(*strp, ':'); - if (!evt_name) - return EVT_FAILED; - - sys_length = evt_name - *strp; - if (sys_length >= MAX_EVENT_LENGTH) - return 0; - - strncpy(sys_name, *strp, sys_length); - sys_name[sys_length] = '\0'; - evt_name = evt_name + 1; + int ret; - comma_loc = strchr(evt_name, ','); - if (comma_loc) { - /* take the event name up to the comma */ - evt_name = strndup(evt_name, comma_loc - evt_name); - } - flags = strchr(evt_name, ':'); - if (flags) { - /* split it out: */ - evt_name = strndup(evt_name, flags - evt_name); - flags++; - } + ret = debugfs_valid_mountpoint(tracing_events_path); + if (ret) + return ret; - evt_length = strlen(evt_name); - if (evt_length >= MAX_EVENT_LENGTH) - return EVT_FAILED; - if (strpbrk(evt_name, "*?")) { - *strp += strlen(sys_name) + evt_length + 1; /* 1 == the ':' */ - return parse_multiple_tracepoint_event(evlist, sys_name, - evt_name, flags); - } else { - return parse_single_tracepoint_event(sys_name, evt_name, - evt_length, attr, strp); - } + return strpbrk(event, "*?") ? + add_tracepoint_multi(list, idx, sys, event) : + add_tracepoint(list, idx, sys, event); } -static enum event_result -parse_breakpoint_type(const char *type, const char **strp, - struct perf_event_attr *attr) +static int +parse_breakpoint_type(const char *type, struct perf_event_attr *attr) { int i; for (i = 0; i < 3; i++) { - if (!type[i]) + if (!type || !type[i]) break; switch (type[i]) { @@ -589,164 +560,179 @@ parse_breakpoint_type(const char *type, const char **strp, attr->bp_type |= HW_BREAKPOINT_X; break; default: - return EVT_FAILED; + return -EINVAL; } } + if (!attr->bp_type) /* Default */ attr->bp_type = HW_BREAKPOINT_R | HW_BREAKPOINT_W; - *strp = type + i; - - return EVT_HANDLED; + return 0; } -static enum event_result -parse_breakpoint_event(const char **strp, struct perf_event_attr *attr) +int parse_events_add_breakpoint(struct list_head **list, int *idx, + void *ptr, char *type) { - const char *target; - const char *type; - char *endaddr; - u64 addr; - enum event_result err; - - target = strchr(*strp, ':'); - if (!target) - return EVT_FAILED; - - if (strncmp(*strp, "mem", target - *strp) != 0) - return EVT_FAILED; - - target++; - - addr = strtoull(target, &endaddr, 0); - if (target == endaddr) - return EVT_FAILED; - - attr->bp_addr = addr; - *strp = endaddr; + struct perf_event_attr attr; + char name[MAX_NAME_LEN]; - type = strchr(target, ':'); + memset(&attr, 0, sizeof(attr)); + attr.bp_addr = (unsigned long) ptr; - /* If no type is defined, just rw as default */ - if (!type) { - attr->bp_type = HW_BREAKPOINT_R | HW_BREAKPOINT_W; - } else { - err = parse_breakpoint_type(++type, strp, attr); - if (err == EVT_FAILED) - return EVT_FAILED; - } + if (parse_breakpoint_type(type, &attr)) + return -EINVAL; /* * We should find a nice way to override the access length * Provide some defaults for now */ - if (attr->bp_type == HW_BREAKPOINT_X) - attr->bp_len = sizeof(long); + if (attr.bp_type == HW_BREAKPOINT_X) + attr.bp_len = sizeof(long); else - attr->bp_len = HW_BREAKPOINT_LEN_4; + attr.bp_len = HW_BREAKPOINT_LEN_4; - attr->type = PERF_TYPE_BREAKPOINT; + attr.type = PERF_TYPE_BREAKPOINT; - return EVT_HANDLED; + snprintf(name, MAX_NAME_LEN, "mem:%p:%s", ptr, type ? type : "rw"); + return add_event(list, idx, &attr, name); } -static int check_events(const char *str, unsigned int i) +static int config_term(struct perf_event_attr *attr, + struct parse_events__term *term) { - int n; +#define CHECK_TYPE_VAL(type) \ +do { \ + if (PARSE_EVENTS__TERM_TYPE_ ## type != term->type_val) \ + return -EINVAL; \ +} while (0) + + switch (term->type_term) { + case PARSE_EVENTS__TERM_TYPE_CONFIG: + CHECK_TYPE_VAL(NUM); + attr->config = term->val.num; + break; + case PARSE_EVENTS__TERM_TYPE_CONFIG1: + CHECK_TYPE_VAL(NUM); + attr->config1 = term->val.num; + break; + case PARSE_EVENTS__TERM_TYPE_CONFIG2: + CHECK_TYPE_VAL(NUM); + attr->config2 = term->val.num; + break; + case PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD: + CHECK_TYPE_VAL(NUM); + attr->sample_period = term->val.num; + break; + case PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE: + /* + * TODO uncomment when the field is available + * attr->branch_sample_type = term->val.num; + */ + break; + case PARSE_EVENTS__TERM_TYPE_NAME: + CHECK_TYPE_VAL(STR); + break; + default: + return -EINVAL; + } + + return 0; +#undef CHECK_TYPE_VAL +} - n = strlen(event_symbols[i].symbol); - if (!strncasecmp(str, event_symbols[i].symbol, n)) - return n; +static int config_attr(struct perf_event_attr *attr, + struct list_head *head, int fail) +{ + struct parse_events__term *term; - n = strlen(event_symbols[i].alias); - if (n) { - if (!strncasecmp(str, event_symbols[i].alias, n)) - return n; - } + list_for_each_entry(term, head, list) + if (config_term(attr, term) && fail) + return -EINVAL; return 0; } -static enum event_result -parse_symbolic_event(const char **strp, struct perf_event_attr *attr) +int parse_events_add_numeric(struct list_head **list, int *idx, + unsigned long type, unsigned long config, + struct list_head *head_config) { - const char *str = *strp; - unsigned int i; - int n; - - for (i = 0; i < ARRAY_SIZE(event_symbols); i++) { - n = check_events(str, i); - if (n > 0) { - attr->type = event_symbols[i].type; - attr->config = event_symbols[i].config; - *strp = str + n; - return EVT_HANDLED; - } - } - return EVT_FAILED; + struct perf_event_attr attr; + + memset(&attr, 0, sizeof(attr)); + attr.type = type; + attr.config = config; + + if (head_config && + config_attr(&attr, head_config, 1)) + return -EINVAL; + + return add_event(list, idx, &attr, + (char *) __event_name(type, config)); } -static enum event_result -parse_raw_event(const char **strp, struct perf_event_attr *attr) +static int parse_events__is_name_term(struct parse_events__term *term) { - const char *str = *strp; - u64 config; - int n; - - if (*str != 'r') - return EVT_FAILED; - n = hex2u64(str + 1, &config); - if (n > 0) { - const char *end = str + n + 1; - if (*end != '\0' && *end != ',' && *end != ':') - return EVT_FAILED; - - *strp = end; - attr->type = PERF_TYPE_RAW; - attr->config = config; - return EVT_HANDLED; - } - return EVT_FAILED; + return term->type_term == PARSE_EVENTS__TERM_TYPE_NAME; } -static enum event_result -parse_numeric_event(const char **strp, struct perf_event_attr *attr) +static char *pmu_event_name(struct perf_event_attr *attr, + struct list_head *head_terms) { - const char *str = *strp; - char *endp; - unsigned long type; - u64 config; - - type = strtoul(str, &endp, 0); - if (endp > str && type < PERF_TYPE_MAX && *endp == ':') { - str = endp + 1; - config = strtoul(str, &endp, 0); - if (endp > str) { - attr->type = type; - attr->config = config; - *strp = endp; - return EVT_HANDLED; - } - } - return EVT_FAILED; + struct parse_events__term *term; + + list_for_each_entry(term, head_terms, list) + if (parse_events__is_name_term(term)) + return term->val.str; + + return (char *) __event_name(PERF_TYPE_RAW, attr->config); } -static int -parse_event_modifier(const char **strp, struct perf_event_attr *attr) +int parse_events_add_pmu(struct list_head **list, int *idx, + char *name, struct list_head *head_config) +{ + struct perf_event_attr attr; + struct perf_pmu *pmu; + + pmu = perf_pmu__find(name); + if (!pmu) + return -EINVAL; + + memset(&attr, 0, sizeof(attr)); + + /* + * Configure hardcoded terms first, no need to check + * return value when called with fail == 0 ;) + */ + config_attr(&attr, head_config, 0); + + if (perf_pmu__config(pmu, &attr, head_config)) + return -EINVAL; + + return add_event(list, idx, &attr, + pmu_event_name(&attr, head_config)); +} + +void parse_events_update_lists(struct list_head *list_event, + struct list_head *list_all) +{ + /* + * Called for single event definition. Update the + * 'all event' list, and reinit the 'signle event' + * list, for next event definition. + */ + list_splice_tail(list_event, list_all); + free(list_event); +} + +int parse_events_modifier(struct list_head *list, char *str) { - const char *str = *strp; + struct perf_evsel *evsel; int exclude = 0, exclude_GH = 0; int eu = 0, ek = 0, eh = 0, eH = 0, eG = 0, precise = 0; - if (!*str) + if (str == NULL) return 0; - if (*str == ',') - return 0; - - if (*str++ != ':') - return -1; - while (*str) { if (*str == 'u') { if (!exclude) @@ -775,111 +761,66 @@ parse_event_modifier(const char **strp, struct perf_event_attr *attr) ++str; } - if (str < *strp + 2) - return -1; - *strp = str; + /* + * precise ip: + * + * 0 - SAMPLE_IP can have arbitrary skid + * 1 - SAMPLE_IP must have constant skid + * 2 - SAMPLE_IP requested to have 0 skid + * 3 - SAMPLE_IP must have 0 skid + * + * See also PERF_RECORD_MISC_EXACT_IP + */ + if (precise > 3) + return -EINVAL; - attr->exclude_user = eu; - attr->exclude_kernel = ek; - attr->exclude_hv = eh; - attr->precise_ip = precise; - attr->exclude_host = eH; - attr->exclude_guest = eG; + list_for_each_entry(evsel, list, node) { + evsel->attr.exclude_user = eu; + evsel->attr.exclude_kernel = ek; + evsel->attr.exclude_hv = eh; + evsel->attr.precise_ip = precise; + evsel->attr.exclude_host = eH; + evsel->attr.exclude_guest = eG; + } return 0; } -/* - * Each event can have multiple symbolic names. - * Symbolic names are (almost) exactly matched. - */ -static enum event_result -parse_event_symbols(struct perf_evlist *evlist, const char **str, - struct perf_event_attr *attr) +int parse_events(struct perf_evlist *evlist, const char *str, int unset __used) { - enum event_result ret; - - ret = parse_tracepoint_event(evlist, str, attr); - if (ret != EVT_FAILED) - goto modifier; - - ret = parse_raw_event(str, attr); - if (ret != EVT_FAILED) - goto modifier; - - ret = parse_numeric_event(str, attr); - if (ret != EVT_FAILED) - goto modifier; + LIST_HEAD(list); + LIST_HEAD(list_tmp); + YY_BUFFER_STATE buffer; + int ret, idx = evlist->nr_entries; - ret = parse_symbolic_event(str, attr); - if (ret != EVT_FAILED) - goto modifier; + buffer = parse_events__scan_string(str); - ret = parse_generic_hw_event(str, attr); - if (ret != EVT_FAILED) - goto modifier; +#ifdef PARSER_DEBUG + parse_events_debug = 1; +#endif + ret = parse_events_parse(&list, &idx); - ret = parse_breakpoint_event(str, attr); - if (ret != EVT_FAILED) - goto modifier; + parse_events__flush_buffer(buffer); + parse_events__delete_buffer(buffer); + parse_events_lex_destroy(); - fprintf(stderr, "invalid or unsupported event: '%s'\n", *str); - fprintf(stderr, "Run 'perf list' for a list of valid events\n"); - return EVT_FAILED; - -modifier: - if (parse_event_modifier(str, attr) < 0) { - fprintf(stderr, "invalid event modifier: '%s'\n", *str); - fprintf(stderr, "Run 'perf list' for a list of valid events and modifiers\n"); - - return EVT_FAILED; + if (!ret) { + int entries = idx - evlist->nr_entries; + perf_evlist__splice_list_tail(evlist, &list, entries); + return 0; } + /* + * There are 2 users - builtin-record and builtin-test objects. + * Both call perf_evlist__delete in case of error, so we dont + * need to bother. + */ + fprintf(stderr, "invalid or unsupported event: '%s'\n", str); + fprintf(stderr, "Run 'perf list' for a list of valid events\n"); return ret; } -int parse_events(struct perf_evlist *evlist , const char *str, int unset __used) -{ - struct perf_event_attr attr; - enum event_result ret; - const char *ostr; - - for (;;) { - ostr = str; - memset(&attr, 0, sizeof(attr)); - event_attr_init(&attr); - ret = parse_event_symbols(evlist, &str, &attr); - if (ret == EVT_FAILED) - return -1; - - if (!(*str == 0 || *str == ',' || isspace(*str))) - return -1; - - if (ret != EVT_HANDLED_ALL) { - struct perf_evsel *evsel; - evsel = perf_evsel__new(&attr, evlist->nr_entries); - if (evsel == NULL) - return -1; - perf_evlist__add(evlist, evsel); - - evsel->name = calloc(str - ostr + 1, 1); - if (!evsel->name) - return -1; - strncpy(evsel->name, ostr, str - ostr); - } - - if (*str == 0) - break; - if (*str == ',') - ++str; - while (isspace(*str)) - ++str; - } - - return 0; -} - int parse_events_option(const struct option *opt, const char *str, int unset __used) { @@ -1052,8 +993,6 @@ int print_hwcache_events(const char *event_glob) return printed; } -#define MAX_NAME_LEN 100 - /* * Print the help text for the event symbols: */ @@ -1102,8 +1041,12 @@ void print_events(const char *event_glob) printf("\n"); printf(" %-50s [%s]\n", - "rNNN (see 'perf list --help' on how to encode it)", + "rNNN", event_type_descriptors[PERF_TYPE_RAW]); + printf(" %-50s [%s]\n", + "cpu/t1=v1[,t2=v2,t3 ...]/modifier", + event_type_descriptors[PERF_TYPE_RAW]); + printf(" (see 'perf list --help' on how to encode it)\n"); printf("\n"); printf(" %-50s [%s]\n", @@ -1113,3 +1056,62 @@ void print_events(const char *event_glob) print_tracepoint_events(NULL, NULL); } + +int parse_events__is_hardcoded_term(struct parse_events__term *term) +{ + return term->type_term != PARSE_EVENTS__TERM_TYPE_USER; +} + +static int new_term(struct parse_events__term **_term, int type_val, + int type_term, char *config, + char *str, long num) +{ + struct parse_events__term *term; + + term = zalloc(sizeof(*term)); + if (!term) + return -ENOMEM; + + INIT_LIST_HEAD(&term->list); + term->type_val = type_val; + term->type_term = type_term; + term->config = config; + + switch (type_val) { + case PARSE_EVENTS__TERM_TYPE_NUM: + term->val.num = num; + break; + case PARSE_EVENTS__TERM_TYPE_STR: + term->val.str = str; + break; + default: + return -EINVAL; + } + + *_term = term; + return 0; +} + +int parse_events__term_num(struct parse_events__term **term, + int type_term, char *config, long num) +{ + return new_term(term, PARSE_EVENTS__TERM_TYPE_NUM, type_term, + config, NULL, num); +} + +int parse_events__term_str(struct parse_events__term **term, + int type_term, char *config, char *str) +{ + return new_term(term, PARSE_EVENTS__TERM_TYPE_STR, type_term, + config, str, 0); +} + +void parse_events__free_terms(struct list_head *terms) +{ + struct parse_events__term *term, *h; + + list_for_each_entry_safe(term, h, terms, list) + free(term); + + free(terms); +} diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h index 7e0cbe75d5f1..8cac57ab4ee6 100644 --- a/tools/perf/util/parse-events.h +++ b/tools/perf/util/parse-events.h @@ -4,7 +4,11 @@ * Parse symbolic events/counts passed in as options: */ +#include <linux/list.h> +#include <stdbool.h> +#include "types.h" #include "../../../include/linux/perf_event.h" +#include "types.h" struct list_head; struct perf_evsel; @@ -33,6 +37,56 @@ extern int parse_filter(const struct option *opt, const char *str, int unset); #define EVENTS_HELP_MAX (128*1024) +enum { + PARSE_EVENTS__TERM_TYPE_NUM, + PARSE_EVENTS__TERM_TYPE_STR, +}; + +enum { + PARSE_EVENTS__TERM_TYPE_USER, + PARSE_EVENTS__TERM_TYPE_CONFIG, + PARSE_EVENTS__TERM_TYPE_CONFIG1, + PARSE_EVENTS__TERM_TYPE_CONFIG2, + PARSE_EVENTS__TERM_TYPE_NAME, + PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD, + PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE, +}; + +struct parse_events__term { + char *config; + union { + char *str; + long num; + } val; + int type_val; + int type_term; + struct list_head list; +}; + +int parse_events__is_hardcoded_term(struct parse_events__term *term); +int parse_events__term_num(struct parse_events__term **_term, + int type_term, char *config, long num); +int parse_events__term_str(struct parse_events__term **_term, + int type_term, char *config, char *str); +void parse_events__free_terms(struct list_head *terms); +int parse_events_modifier(struct list_head *list, char *str); +int parse_events_add_tracepoint(struct list_head **list, int *idx, + char *sys, char *event); +int parse_events_add_numeric(struct list_head **list, int *idx, + unsigned long type, unsigned long config, + struct list_head *head_config); +int parse_events_add_cache(struct list_head **list, int *idx, + char *type, char *op_result1, char *op_result2); +int parse_events_add_breakpoint(struct list_head **list, int *idx, + void *ptr, char *type); +int parse_events_add_pmu(struct list_head **list, int *idx, + char *pmu , struct list_head *head_config); +void parse_events_update_lists(struct list_head *list_event, + struct list_head *list_all); +void parse_events_error(struct list_head *list_all, + int *idx, char const *msg); +int parse_events__test(void); + void print_events(const char *event_glob); void print_events_type(u8 type); void print_tracepoint_events(const char *subsys_glob, const char *event_glob); diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l new file mode 100644 index 000000000000..618a8e788399 --- /dev/null +++ b/tools/perf/util/parse-events.l @@ -0,0 +1,149 @@ + +%option prefix="parse_events_" +%option stack + +%{ +#include <errno.h> +#include "../perf.h" +#include "parse-events-bison.h" +#include "parse-events.h" + +static int __value(char *str, int base, int token) +{ + long num; + + errno = 0; + num = strtoul(str, NULL, base); + if (errno) + return PE_ERROR; + + parse_events_lval.num = num; + return token; +} + +static int value(int base) +{ + return __value(parse_events_text, base, PE_VALUE); +} + +static int raw(void) +{ + return __value(parse_events_text + 1, 16, PE_RAW); +} + +static int str(int token) +{ + parse_events_lval.str = strdup(parse_events_text); + return token; +} + +static int sym(int type, int config) +{ + parse_events_lval.num = (type << 16) + config; + return PE_VALUE_SYM; +} + +static int term(int type) +{ + parse_events_lval.num = type; + return PE_TERM; +} + +%} + +%x mem + +num_dec [0-9]+ +num_hex 0x[a-fA-F0-9]+ +num_raw_hex [a-fA-F0-9]+ +name [a-zA-Z_*?][a-zA-Z0-9_*?]* +modifier_event [ukhpGH]{1,8} +modifier_bp [rwx] + +%% +cpu-cycles|cycles { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_CPU_CYCLES); } +stalled-cycles-frontend|idle-cycles-frontend { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_STALLED_CYCLES_FRONTEND); } +stalled-cycles-backend|idle-cycles-backend { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_STALLED_CYCLES_BACKEND); } +instructions { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_INSTRUCTIONS); } +cache-references { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_CACHE_REFERENCES); } +cache-misses { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_CACHE_MISSES); } +branch-instructions|branches { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_BRANCH_INSTRUCTIONS); } +branch-misses { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_BRANCH_MISSES); } +bus-cycles { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_BUS_CYCLES); } +ref-cycles { return sym(PERF_TYPE_HARDWARE, PERF_COUNT_HW_REF_CPU_CYCLES); } +cpu-clock { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_CPU_CLOCK); } +task-clock { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_TASK_CLOCK); } +page-faults|faults { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_PAGE_FAULTS); } +minor-faults { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_PAGE_FAULTS_MIN); } +major-faults { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_PAGE_FAULTS_MAJ); } +context-switches|cs { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_CONTEXT_SWITCHES); } +cpu-migrations|migrations { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_CPU_MIGRATIONS); } +alignment-faults { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_ALIGNMENT_FAULTS); } +emulation-faults { return sym(PERF_TYPE_SOFTWARE, PERF_COUNT_SW_EMULATION_FAULTS); } + +L1-dcache|l1-d|l1d|L1-data | +L1-icache|l1-i|l1i|L1-instruction | +LLC|L2 | +dTLB|d-tlb|Data-TLB | +iTLB|i-tlb|Instruction-TLB | +branch|branches|bpu|btb|bpc | +node { return str(PE_NAME_CACHE_TYPE); } + +load|loads|read | +store|stores|write | +prefetch|prefetches | +speculative-read|speculative-load | +refs|Reference|ops|access | +misses|miss { return str(PE_NAME_CACHE_OP_RESULT); } + + /* + * These are event config hardcoded term names to be specified + * within xxx/.../ syntax. So far we dont clash with other names, + * so we can put them here directly. In case the we have a conflict + * in future, this needs to go into '//' condition block. + */ +config { return term(PARSE_EVENTS__TERM_TYPE_CONFIG); } +config1 { return term(PARSE_EVENTS__TERM_TYPE_CONFIG1); } +config2 { return term(PARSE_EVENTS__TERM_TYPE_CONFIG2); } +name { return term(PARSE_EVENTS__TERM_TYPE_NAME); } +period { return term(PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD); } +branch_type { return term(PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE); } + +mem: { BEGIN(mem); return PE_PREFIX_MEM; } +r{num_raw_hex} { return raw(); } +{num_dec} { return value(10); } +{num_hex} { return value(16); } + +{modifier_event} { return str(PE_MODIFIER_EVENT); } +{name} { return str(PE_NAME); } +"/" { return '/'; } +- { return '-'; } +, { return ','; } +: { return ':'; } += { return '='; } + +<mem>{ +{modifier_bp} { return str(PE_MODIFIER_BP); } +: { return ':'; } +{num_dec} { return value(10); } +{num_hex} { return value(16); } + /* + * We need to separate 'mem:' scanner part, in order to get specific + * modifier bits parsed out. Otherwise we would need to handle PE_NAME + * and we'd need to parse it manually. During the escape from <mem> + * state we need to put the escaping char back, so we dont miss it. + */ +. { unput(*parse_events_text); BEGIN(INITIAL); } + /* + * We destroy the scanner after reaching EOF, + * but anyway just to be sure get back to INIT state. + */ +<<EOF>> { BEGIN(INITIAL); } +} + +%% + +int parse_events_wrap(void) +{ + return 1; +} diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y new file mode 100644 index 000000000000..362cc59332ae --- /dev/null +++ b/tools/perf/util/parse-events.y @@ -0,0 +1,274 @@ + +%name-prefix "parse_events_" +%parse-param {struct list_head *list_all} +%parse-param {int *idx} + +%{ + +#define YYDEBUG 1 + +#include <linux/compiler.h> +#include <linux/list.h> +#include "types.h" +#include "util.h" +#include "parse-events.h" + +extern int parse_events_lex (void); + +#define ABORT_ON(val) \ +do { \ + if (val) \ + YYABORT; \ +} while (0) + +%} + +%token PE_VALUE PE_VALUE_SYM PE_RAW PE_TERM +%token PE_NAME +%token PE_MODIFIER_EVENT PE_MODIFIER_BP +%token PE_NAME_CACHE_TYPE PE_NAME_CACHE_OP_RESULT +%token PE_PREFIX_MEM PE_PREFIX_RAW +%token PE_ERROR +%type <num> PE_VALUE +%type <num> PE_VALUE_SYM +%type <num> PE_RAW +%type <num> PE_TERM +%type <str> PE_NAME +%type <str> PE_NAME_CACHE_TYPE +%type <str> PE_NAME_CACHE_OP_RESULT +%type <str> PE_MODIFIER_EVENT +%type <str> PE_MODIFIER_BP +%type <head> event_config +%type <term> event_term +%type <head> event_pmu +%type <head> event_legacy_symbol +%type <head> event_legacy_cache +%type <head> event_legacy_mem +%type <head> event_legacy_tracepoint +%type <head> event_legacy_numeric +%type <head> event_legacy_raw +%type <head> event_def + +%union +{ + char *str; + unsigned long num; + struct list_head *head; + struct parse_events__term *term; +} +%% + +events: +events ',' event | event + +event: +event_def PE_MODIFIER_EVENT +{ + /* + * Apply modifier on all events added by single event definition + * (there could be more events added for multiple tracepoint + * definitions via '*?'. + */ + ABORT_ON(parse_events_modifier($1, $2)); + parse_events_update_lists($1, list_all); +} +| +event_def +{ + parse_events_update_lists($1, list_all); +} + +event_def: event_pmu | + event_legacy_symbol | + event_legacy_cache sep_dc | + event_legacy_mem | + event_legacy_tracepoint sep_dc | + event_legacy_numeric sep_dc | + event_legacy_raw sep_dc + +event_pmu: +PE_NAME '/' event_config '/' +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_pmu(&list, idx, $1, $3)); + parse_events__free_terms($3); + $$ = list; +} + +event_legacy_symbol: +PE_VALUE_SYM '/' event_config '/' +{ + struct list_head *list = NULL; + int type = $1 >> 16; + int config = $1 & 255; + + ABORT_ON(parse_events_add_numeric(&list, idx, type, config, $3)); + parse_events__free_terms($3); + $$ = list; +} +| +PE_VALUE_SYM sep_slash_dc +{ + struct list_head *list = NULL; + int type = $1 >> 16; + int config = $1 & 255; + + ABORT_ON(parse_events_add_numeric(&list, idx, type, config, NULL)); + $$ = list; +} + +event_legacy_cache: +PE_NAME_CACHE_TYPE '-' PE_NAME_CACHE_OP_RESULT '-' PE_NAME_CACHE_OP_RESULT +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_cache(&list, idx, $1, $3, $5)); + $$ = list; +} +| +PE_NAME_CACHE_TYPE '-' PE_NAME_CACHE_OP_RESULT +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_cache(&list, idx, $1, $3, NULL)); + $$ = list; +} +| +PE_NAME_CACHE_TYPE +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_cache(&list, idx, $1, NULL, NULL)); + $$ = list; +} + +event_legacy_mem: +PE_PREFIX_MEM PE_VALUE ':' PE_MODIFIER_BP sep_dc +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_breakpoint(&list, idx, (void *) $2, $4)); + $$ = list; +} +| +PE_PREFIX_MEM PE_VALUE sep_dc +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_breakpoint(&list, idx, (void *) $2, NULL)); + $$ = list; +} + +event_legacy_tracepoint: +PE_NAME ':' PE_NAME +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_tracepoint(&list, idx, $1, $3)); + $$ = list; +} + +event_legacy_numeric: +PE_VALUE ':' PE_VALUE +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_numeric(&list, idx, $1, $3, NULL)); + $$ = list; +} + +event_legacy_raw: +PE_RAW +{ + struct list_head *list = NULL; + + ABORT_ON(parse_events_add_numeric(&list, idx, PERF_TYPE_RAW, $1, NULL)); + $$ = list; +} + +event_config: +event_config ',' event_term +{ + struct list_head *head = $1; + struct parse_events__term *term = $3; + + ABORT_ON(!head); + list_add_tail(&term->list, head); + $$ = $1; +} +| +event_term +{ + struct list_head *head = malloc(sizeof(*head)); + struct parse_events__term *term = $1; + + ABORT_ON(!head); + INIT_LIST_HEAD(head); + list_add_tail(&term->list, head); + $$ = head; +} + +event_term: +PE_NAME '=' PE_NAME +{ + struct parse_events__term *term; + + ABORT_ON(parse_events__term_str(&term, PARSE_EVENTS__TERM_TYPE_USER, + $1, $3)); + $$ = term; +} +| +PE_NAME '=' PE_VALUE +{ + struct parse_events__term *term; + + ABORT_ON(parse_events__term_num(&term, PARSE_EVENTS__TERM_TYPE_USER, + $1, $3)); + $$ = term; +} +| +PE_NAME +{ + struct parse_events__term *term; + + ABORT_ON(parse_events__term_num(&term, PARSE_EVENTS__TERM_TYPE_USER, + $1, 1)); + $$ = term; +} +| +PE_TERM '=' PE_NAME +{ + struct parse_events__term *term; + + ABORT_ON(parse_events__term_str(&term, $1, NULL, $3)); + $$ = term; +} +| +PE_TERM '=' PE_VALUE +{ + struct parse_events__term *term; + + ABORT_ON(parse_events__term_num(&term, $1, NULL, $3)); + $$ = term; +} +| +PE_TERM +{ + struct parse_events__term *term; + + ABORT_ON(parse_events__term_num(&term, $1, NULL, 1)); + $$ = term; +} + +sep_dc: ':' | + +sep_slash_dc: '/' | ':' | + +%% + +void parse_events_error(struct list_head *list_all __used, + int *idx __used, + char const *msg __used) +{ +} diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c new file mode 100644 index 000000000000..a119a5371699 --- /dev/null +++ b/tools/perf/util/pmu.c @@ -0,0 +1,483 @@ + +#include <linux/list.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <unistd.h> +#include <stdio.h> +#include <dirent.h> +#include "sysfs.h" +#include "util.h" +#include "pmu.h" +#include "parse-events.h" + +int perf_pmu_parse(struct list_head *list, char *name); +extern FILE *perf_pmu_in; + +static LIST_HEAD(pmus); + +/* + * Parse & process all the sysfs attributes located under + * the directory specified in 'dir' parameter. + */ +static int pmu_format_parse(char *dir, struct list_head *head) +{ + struct dirent *evt_ent; + DIR *format_dir; + int ret = 0; + + format_dir = opendir(dir); + if (!format_dir) + return -EINVAL; + + while (!ret && (evt_ent = readdir(format_dir))) { + char path[PATH_MAX]; + char *name = evt_ent->d_name; + FILE *file; + + if (!strcmp(name, ".") || !strcmp(name, "..")) + continue; + + snprintf(path, PATH_MAX, "%s/%s", dir, name); + + ret = -EINVAL; + file = fopen(path, "r"); + if (!file) + break; + + perf_pmu_in = file; + ret = perf_pmu_parse(head, name); + fclose(file); + } + + closedir(format_dir); + return ret; +} + +/* + * Reading/parsing the default pmu format definition, which should be + * located at: + * /sys/bus/event_source/devices/<dev>/format as sysfs group attributes. + */ +static int pmu_format(char *name, struct list_head *format) +{ + struct stat st; + char path[PATH_MAX]; + const char *sysfs; + + sysfs = sysfs_find_mountpoint(); + if (!sysfs) + return -1; + + snprintf(path, PATH_MAX, + "%s/bus/event_source/devices/%s/format", sysfs, name); + + if (stat(path, &st) < 0) + return -1; + + if (pmu_format_parse(path, format)) + return -1; + + return 0; +} + +/* + * Reading/parsing the default pmu type value, which should be + * located at: + * /sys/bus/event_source/devices/<dev>/type as sysfs attribute. + */ +static int pmu_type(char *name, __u32 *type) +{ + struct stat st; + char path[PATH_MAX]; + const char *sysfs; + FILE *file; + int ret = 0; + + sysfs = sysfs_find_mountpoint(); + if (!sysfs) + return -1; + + snprintf(path, PATH_MAX, + "%s/bus/event_source/devices/%s/type", sysfs, name); + + if (stat(path, &st) < 0) + return -1; + + file = fopen(path, "r"); + if (!file) + return -EINVAL; + + if (1 != fscanf(file, "%u", type)) + ret = -1; + + fclose(file); + return ret; +} + +static struct perf_pmu *pmu_lookup(char *name) +{ + struct perf_pmu *pmu; + LIST_HEAD(format); + __u32 type; + + /* + * The pmu data we store & need consists of the pmu + * type value and format definitions. Load both right + * now. + */ + if (pmu_format(name, &format)) + return NULL; + + if (pmu_type(name, &type)) + return NULL; + + pmu = zalloc(sizeof(*pmu)); + if (!pmu) + return NULL; + + INIT_LIST_HEAD(&pmu->format); + list_splice(&format, &pmu->format); + pmu->name = strdup(name); + pmu->type = type; + return pmu; +} + +static struct perf_pmu *pmu_find(char *name) +{ + struct perf_pmu *pmu; + + list_for_each_entry(pmu, &pmus, list) + if (!strcmp(pmu->name, name)) + return pmu; + + return NULL; +} + +struct perf_pmu *perf_pmu__find(char *name) +{ + struct perf_pmu *pmu; + + /* + * Once PMU is loaded it stays in the list, + * so we keep us from multiple reading/parsing + * the pmu format definitions. + */ + pmu = pmu_find(name); + if (pmu) + return pmu; + + return pmu_lookup(name); +} + +static struct perf_pmu__format* +pmu_find_format(struct list_head *formats, char *name) +{ + struct perf_pmu__format *format; + + list_for_each_entry(format, formats, list) + if (!strcmp(format->name, name)) + return format; + + return NULL; +} + +/* + * Returns value based on the format definition (format parameter) + * and unformated value (value parameter). + * + * TODO maybe optimize a little ;) + */ +static __u64 pmu_format_value(unsigned long *format, __u64 value) +{ + unsigned long fbit, vbit; + __u64 v = 0; + + for (fbit = 0, vbit = 0; fbit < PERF_PMU_FORMAT_BITS; fbit++) { + + if (!test_bit(fbit, format)) + continue; + + if (!(value & (1llu << vbit++))) + continue; + + v |= (1llu << fbit); + } + + return v; +} + +/* + * Setup one of config[12] attr members based on the + * user input data - temr parameter. + */ +static int pmu_config_term(struct list_head *formats, + struct perf_event_attr *attr, + struct parse_events__term *term) +{ + struct perf_pmu__format *format; + __u64 *vp; + + /* + * Support only for hardcoded and numnerial terms. + * Hardcoded terms should be already in, so nothing + * to be done for them. + */ + if (parse_events__is_hardcoded_term(term)) + return 0; + + if (term->type_val != PARSE_EVENTS__TERM_TYPE_NUM) + return -EINVAL; + + format = pmu_find_format(formats, term->config); + if (!format) + return -EINVAL; + + switch (format->value) { + case PERF_PMU_FORMAT_VALUE_CONFIG: + vp = &attr->config; + break; + case PERF_PMU_FORMAT_VALUE_CONFIG1: + vp = &attr->config1; + break; + case PERF_PMU_FORMAT_VALUE_CONFIG2: + vp = &attr->config2; + break; + default: + return -EINVAL; + } + + /* + * XXX If we ever decide to go with string values for + * non-hardcoded terms, here's the place to translate + * them into value. + */ + *vp |= pmu_format_value(format->bits, term->val.num); + return 0; +} + +static int pmu_config(struct list_head *formats, struct perf_event_attr *attr, + struct list_head *head_terms) +{ + struct parse_events__term *term; + + list_for_each_entry(term, head_terms, list) + if (pmu_config_term(formats, attr, term)) + return -EINVAL; + + return 0; +} + +/* + * Configures event's 'attr' parameter based on the: + * 1) users input - specified in terms parameter + * 2) pmu format definitions - specified by pmu parameter + */ +int perf_pmu__config(struct perf_pmu *pmu, struct perf_event_attr *attr, + struct list_head *head_terms) +{ + attr->type = pmu->type; + return pmu_config(&pmu->format, attr, head_terms); +} + +int perf_pmu__new_format(struct list_head *list, char *name, + int config, unsigned long *bits) +{ + struct perf_pmu__format *format; + + format = zalloc(sizeof(*format)); + if (!format) + return -ENOMEM; + + format->name = strdup(name); + format->value = config; + memcpy(format->bits, bits, sizeof(format->bits)); + + list_add_tail(&format->list, list); + return 0; +} + +void perf_pmu__set_format(unsigned long *bits, long from, long to) +{ + long b; + + if (!to) + to = from; + + memset(bits, 0, BITS_TO_LONGS(PERF_PMU_FORMAT_BITS)); + for (b = from; b <= to; b++) + set_bit(b, bits); +} + +/* Simulated format definitions. */ +static struct test_format { + const char *name; + const char *value; +} test_formats[] = { + { "krava01", "config:0-1,62-63\n", }, + { "krava02", "config:10-17\n", }, + { "krava03", "config:5\n", }, + { "krava11", "config1:0,2,4,6,8,20-28\n", }, + { "krava12", "config1:63\n", }, + { "krava13", "config1:45-47\n", }, + { "krava21", "config2:0-3,10-13,20-23,30-33,40-43,50-53,60-63\n", }, + { "krava22", "config2:8,18,48,58\n", }, + { "krava23", "config2:28-29,38\n", }, +}; + +#define TEST_FORMATS_CNT (sizeof(test_formats) / sizeof(struct test_format)) + +/* Simulated users input. */ +static struct parse_events__term test_terms[] = { + { + .config = (char *) "krava01", + .val.num = 15, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava02", + .val.num = 170, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava03", + .val.num = 1, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava11", + .val.num = 27, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava12", + .val.num = 1, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava13", + .val.num = 2, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava21", + .val.num = 119, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava22", + .val.num = 11, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, + { + .config = (char *) "krava23", + .val.num = 2, + .type_val = PARSE_EVENTS__TERM_TYPE_NUM, + .type_term = PARSE_EVENTS__TERM_TYPE_USER, + }, +}; +#define TERMS_CNT (sizeof(test_terms) / sizeof(struct parse_events__term)) + +/* + * Prepare format directory data, exported by kernel + * at /sys/bus/event_source/devices/<dev>/format. + */ +static char *test_format_dir_get(void) +{ + static char dir[PATH_MAX]; + unsigned int i; + + snprintf(dir, PATH_MAX, "/tmp/perf-pmu-test-format-XXXXXX"); + if (!mkdtemp(dir)) + return NULL; + + for (i = 0; i < TEST_FORMATS_CNT; i++) { + static char name[PATH_MAX]; + struct test_format *format = &test_formats[i]; + FILE *file; + + snprintf(name, PATH_MAX, "%s/%s", dir, format->name); + + file = fopen(name, "w"); + if (!file) + return NULL; + + if (1 != fwrite(format->value, strlen(format->value), 1, file)) + break; + + fclose(file); + } + + return dir; +} + +/* Cleanup format directory. */ +static int test_format_dir_put(char *dir) +{ + char buf[PATH_MAX]; + snprintf(buf, PATH_MAX, "rm -f %s/*\n", dir); + if (system(buf)) + return -1; + + snprintf(buf, PATH_MAX, "rmdir %s\n", dir); + return system(buf); +} + +static struct list_head *test_terms_list(void) +{ + static LIST_HEAD(terms); + unsigned int i; + + for (i = 0; i < TERMS_CNT; i++) + list_add_tail(&test_terms[i].list, &terms); + + return &terms; +} + +#undef TERMS_CNT + +int perf_pmu__test(void) +{ + char *format = test_format_dir_get(); + LIST_HEAD(formats); + struct list_head *terms = test_terms_list(); + int ret; + + if (!format) + return -EINVAL; + + do { + struct perf_event_attr attr; + + memset(&attr, 0, sizeof(attr)); + + ret = pmu_format_parse(format, &formats); + if (ret) + break; + + ret = pmu_config(&formats, &attr, terms); + if (ret) + break; + + ret = -EINVAL; + + if (attr.config != 0xc00000000002a823) + break; + if (attr.config1 != 0x8000400000000145) + break; + if (attr.config2 != 0x0400000020041d07) + break; + + ret = 0; + } while (0); + + test_format_dir_put(format); + return ret; +} diff --git a/tools/perf/util/pmu.h b/tools/perf/util/pmu.h new file mode 100644 index 000000000000..68c0db965e1f --- /dev/null +++ b/tools/perf/util/pmu.h @@ -0,0 +1,41 @@ +#ifndef __PMU_H +#define __PMU_H + +#include <linux/bitops.h> +#include "../../../include/linux/perf_event.h" + +enum { + PERF_PMU_FORMAT_VALUE_CONFIG, + PERF_PMU_FORMAT_VALUE_CONFIG1, + PERF_PMU_FORMAT_VALUE_CONFIG2, +}; + +#define PERF_PMU_FORMAT_BITS 64 + +struct perf_pmu__format { + char *name; + int value; + DECLARE_BITMAP(bits, PERF_PMU_FORMAT_BITS); + struct list_head list; +}; + +struct perf_pmu { + char *name; + __u32 type; + struct list_head format; + struct list_head list; +}; + +struct perf_pmu *perf_pmu__find(char *name); +int perf_pmu__config(struct perf_pmu *pmu, struct perf_event_attr *attr, + struct list_head *head_terms); + +int perf_pmu_wrap(void); +void perf_pmu_error(struct list_head *list, char *name, char const *msg); + +int perf_pmu__new_format(struct list_head *list, char *name, + int config, unsigned long *bits); +void perf_pmu__set_format(unsigned long *bits, long from, long to); + +int perf_pmu__test(void); +#endif /* __PMU_H */ diff --git a/tools/perf/util/pmu.l b/tools/perf/util/pmu.l new file mode 100644 index 000000000000..a15d9fbd7c0e --- /dev/null +++ b/tools/perf/util/pmu.l @@ -0,0 +1,43 @@ +%option prefix="perf_pmu_" + +%{ +#include <stdlib.h> +#include <linux/bitops.h> +#include "pmu.h" +#include "pmu-bison.h" + +static int value(int base) +{ + long num; + + errno = 0; + num = strtoul(perf_pmu_text, NULL, base); + if (errno) + return PP_ERROR; + + perf_pmu_lval.num = num; + return PP_VALUE; +} + +%} + +num_dec [0-9]+ + +%% + +{num_dec} { return value(10); } +config { return PP_CONFIG; } +config1 { return PP_CONFIG1; } +config2 { return PP_CONFIG2; } +- { return '-'; } +: { return ':'; } +, { return ','; } +. { ; } +\n { ; } + +%% + +int perf_pmu_wrap(void) +{ + return 1; +} diff --git a/tools/perf/util/pmu.y b/tools/perf/util/pmu.y new file mode 100644 index 000000000000..20ea77e93169 --- /dev/null +++ b/tools/perf/util/pmu.y @@ -0,0 +1,93 @@ + +%name-prefix "perf_pmu_" +%parse-param {struct list_head *format} +%parse-param {char *name} + +%{ + +#include <linux/compiler.h> +#include <linux/list.h> +#include <linux/bitmap.h> +#include <string.h> +#include "pmu.h" + +extern int perf_pmu_lex (void); + +#define ABORT_ON(val) \ +do { \ + if (val) \ + YYABORT; \ +} while (0) + +%} + +%token PP_CONFIG PP_CONFIG1 PP_CONFIG2 +%token PP_VALUE PP_ERROR +%type <num> PP_VALUE +%type <bits> bit_term +%type <bits> bits + +%union +{ + unsigned long num; + DECLARE_BITMAP(bits, PERF_PMU_FORMAT_BITS); +} + +%% + +format: +format format_term +| +format_term + +format_term: +PP_CONFIG ':' bits +{ + ABORT_ON(perf_pmu__new_format(format, name, + PERF_PMU_FORMAT_VALUE_CONFIG, + $3)); +} +| +PP_CONFIG1 ':' bits +{ + ABORT_ON(perf_pmu__new_format(format, name, + PERF_PMU_FORMAT_VALUE_CONFIG1, + $3)); +} +| +PP_CONFIG2 ':' bits +{ + ABORT_ON(perf_pmu__new_format(format, name, + PERF_PMU_FORMAT_VALUE_CONFIG2, + $3)); +} + +bits: +bits ',' bit_term +{ + bitmap_or($$, $1, $3, 64); +} +| +bit_term +{ + memcpy($$, $1, sizeof($1)); +} + +bit_term: +PP_VALUE '-' PP_VALUE +{ + perf_pmu__set_format($$, $1, $3); +} +| +PP_VALUE +{ + perf_pmu__set_format($$, $1, 0); +} + +%% + +void perf_pmu_error(struct list_head *list __used, + char *name __used, + char const *msg __used) +{ +} diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index 8a8ee64e72d1..0dda25d82d06 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -44,6 +44,7 @@ #include "trace-event.h" /* For __unused */ #include "probe-event.h" #include "probe-finder.h" +#include "session.h" #define MAX_CMDLEN 256 #define MAX_PROBE_ARGS 128 @@ -70,6 +71,8 @@ static int e_snprintf(char *str, size_t size, const char *format, ...) } static char *synthesize_perf_probe_point(struct perf_probe_point *pp); +static int convert_name_to_addr(struct perf_probe_event *pev, + const char *exec); static struct machine machine; /* Initialize symbol maps and path of vmlinux/modules */ @@ -170,6 +173,34 @@ const char *kernel_get_module_path(const char *module) return (dso) ? dso->long_name : NULL; } +static int init_user_exec(void) +{ + int ret = 0; + + symbol_conf.try_vmlinux_path = false; + symbol_conf.sort_by_name = true; + ret = symbol__init(); + + if (ret < 0) + pr_debug("Failed to init symbol map.\n"); + + return ret; +} + +static int convert_to_perf_probe_point(struct probe_trace_point *tp, + struct perf_probe_point *pp) +{ + pp->function = strdup(tp->symbol); + + if (pp->function == NULL) + return -ENOMEM; + + pp->offset = tp->offset; + pp->retprobe = tp->retprobe; + + return 0; +} + #ifdef DWARF_SUPPORT /* Open new debuginfo of given module */ static struct debuginfo *open_debuginfo(const char *module) @@ -224,10 +255,7 @@ static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp, if (ret <= 0) { pr_debug("Failed to find corresponding probes from " "debuginfo. Use kprobe event information.\n"); - pp->function = strdup(tp->symbol); - if (pp->function == NULL) - return -ENOMEM; - pp->offset = tp->offset; + return convert_to_perf_probe_point(tp, pp); } pp->retprobe = tp->retprobe; @@ -275,9 +303,20 @@ static int try_to_find_probe_trace_events(struct perf_probe_event *pev, int max_tevs, const char *target) { bool need_dwarf = perf_probe_event_need_dwarf(pev); - struct debuginfo *dinfo = open_debuginfo(target); + struct debuginfo *dinfo; int ntevs, ret = 0; + if (pev->uprobes) { + if (need_dwarf) { + pr_warning("Debuginfo-analysis is not yet supported" + " with -x/--exec option.\n"); + return -ENOSYS; + } + return convert_name_to_addr(pev, target); + } + + dinfo = open_debuginfo(target); + if (!dinfo) { if (need_dwarf) { pr_warning("Failed to open debuginfo file.\n"); @@ -603,23 +642,22 @@ static int kprobe_convert_to_perf_probe(struct probe_trace_point *tp, pr_err("Failed to find symbol %s in kernel.\n", tp->symbol); return -ENOENT; } - pp->function = strdup(tp->symbol); - if (pp->function == NULL) - return -ENOMEM; - pp->offset = tp->offset; - pp->retprobe = tp->retprobe; - return 0; + return convert_to_perf_probe_point(tp, pp); } static int try_to_find_probe_trace_events(struct perf_probe_event *pev, struct probe_trace_event **tevs __unused, - int max_tevs __unused, const char *mod __unused) + int max_tevs __unused, const char *target) { if (perf_probe_event_need_dwarf(pev)) { pr_warning("Debuginfo-analysis is not supported.\n"); return -ENOSYS; } + + if (pev->uprobes) + return convert_name_to_addr(pev, target); + return 0; } @@ -1341,11 +1379,18 @@ char *synthesize_probe_trace_command(struct probe_trace_event *tev) if (buf == NULL) return NULL; - len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s%s%s+%lu", - tp->retprobe ? 'r' : 'p', - tev->group, tev->event, - tp->module ?: "", tp->module ? ":" : "", - tp->symbol, tp->offset); + if (tev->uprobes) + len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s:%s", + tp->retprobe ? 'r' : 'p', + tev->group, tev->event, + tp->module, tp->symbol); + else + len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s%s%s+%lu", + tp->retprobe ? 'r' : 'p', + tev->group, tev->event, + tp->module ?: "", tp->module ? ":" : "", + tp->symbol, tp->offset); + if (len <= 0) goto error; @@ -1364,7 +1409,7 @@ error: } static int convert_to_perf_probe_event(struct probe_trace_event *tev, - struct perf_probe_event *pev) + struct perf_probe_event *pev, bool is_kprobe) { char buf[64] = ""; int i, ret; @@ -1376,7 +1421,11 @@ static int convert_to_perf_probe_event(struct probe_trace_event *tev, return -ENOMEM; /* Convert trace_point to probe_point */ - ret = kprobe_convert_to_perf_probe(&tev->point, &pev->point); + if (is_kprobe) + ret = kprobe_convert_to_perf_probe(&tev->point, &pev->point); + else + ret = convert_to_perf_probe_point(&tev->point, &pev->point); + if (ret < 0) return ret; @@ -1472,7 +1521,26 @@ static void clear_probe_trace_event(struct probe_trace_event *tev) memset(tev, 0, sizeof(*tev)); } -static int open_kprobe_events(bool readwrite) +static void print_warn_msg(const char *file, bool is_kprobe) +{ + + if (errno == ENOENT) { + const char *config; + + if (!is_kprobe) + config = "CONFIG_UPROBE_EVENTS"; + else + config = "CONFIG_KPROBE_EVENTS"; + + pr_warning("%s file does not exist - please rebuild kernel" + " with %s.\n", file, config); + } else + pr_warning("Failed to open %s file: %s\n", file, + strerror(errno)); +} + +static int open_probe_events(const char *trace_file, bool readwrite, + bool is_kprobe) { char buf[PATH_MAX]; const char *__debugfs; @@ -1484,27 +1552,31 @@ static int open_kprobe_events(bool readwrite) return -ENOENT; } - ret = e_snprintf(buf, PATH_MAX, "%stracing/kprobe_events", __debugfs); + ret = e_snprintf(buf, PATH_MAX, "%s/%s", __debugfs, trace_file); if (ret >= 0) { pr_debug("Opening %s write=%d\n", buf, readwrite); if (readwrite && !probe_event_dry_run) ret = open(buf, O_RDWR, O_APPEND); else ret = open(buf, O_RDONLY, 0); - } - if (ret < 0) { - if (errno == ENOENT) - pr_warning("kprobe_events file does not exist - please" - " rebuild kernel with CONFIG_KPROBE_EVENT.\n"); - else - pr_warning("Failed to open kprobe_events file: %s\n", - strerror(errno)); + if (ret < 0) + print_warn_msg(buf, is_kprobe); } return ret; } -/* Get raw string list of current kprobe_events */ +static int open_kprobe_events(bool readwrite) +{ + return open_probe_events("tracing/kprobe_events", readwrite, true); +} + +static int open_uprobe_events(bool readwrite) +{ + return open_probe_events("tracing/uprobe_events", readwrite, false); +} + +/* Get raw string list of current kprobe_events or uprobe_events */ static struct strlist *get_probe_trace_command_rawlist(int fd) { int ret, idx; @@ -1569,36 +1641,26 @@ static int show_perf_probe_event(struct perf_probe_event *pev) return ret; } -/* List up current perf-probe events */ -int show_perf_probe_events(void) +static int __show_perf_probe_events(int fd, bool is_kprobe) { - int fd, ret; + int ret = 0; struct probe_trace_event tev; struct perf_probe_event pev; struct strlist *rawlist; struct str_node *ent; - setup_pager(); - ret = init_vmlinux(); - if (ret < 0) - return ret; - memset(&tev, 0, sizeof(tev)); memset(&pev, 0, sizeof(pev)); - fd = open_kprobe_events(false); - if (fd < 0) - return fd; - rawlist = get_probe_trace_command_rawlist(fd); - close(fd); if (!rawlist) return -ENOENT; strlist__for_each(ent, rawlist) { ret = parse_probe_trace_command(ent->s, &tev); if (ret >= 0) { - ret = convert_to_perf_probe_event(&tev, &pev); + ret = convert_to_perf_probe_event(&tev, &pev, + is_kprobe); if (ret >= 0) ret = show_perf_probe_event(&pev); } @@ -1612,6 +1674,33 @@ int show_perf_probe_events(void) return ret; } +/* List up current perf-probe events */ +int show_perf_probe_events(void) +{ + int fd, ret; + + setup_pager(); + fd = open_kprobe_events(false); + + if (fd < 0) + return fd; + + ret = init_vmlinux(); + if (ret < 0) + return ret; + + ret = __show_perf_probe_events(fd, true); + close(fd); + + fd = open_uprobe_events(false); + if (fd >= 0) { + ret = __show_perf_probe_events(fd, false); + close(fd); + } + + return ret; +} + /* Get current perf-probe event names */ static struct strlist *get_probe_trace_event_names(int fd, bool include_group) { @@ -1717,7 +1806,11 @@ static int __add_probe_trace_events(struct perf_probe_event *pev, const char *event, *group; struct strlist *namelist; - fd = open_kprobe_events(true); + if (pev->uprobes) + fd = open_uprobe_events(true); + else + fd = open_kprobe_events(true); + if (fd < 0) return fd; /* Get current event names */ @@ -1829,6 +1922,8 @@ static int convert_to_probe_trace_events(struct perf_probe_event *pev, tev->point.offset = pev->point.offset; tev->point.retprobe = pev->point.retprobe; tev->nargs = pev->nargs; + tev->uprobes = pev->uprobes; + if (tev->nargs) { tev->args = zalloc(sizeof(struct probe_trace_arg) * tev->nargs); @@ -1859,6 +1954,9 @@ static int convert_to_probe_trace_events(struct perf_probe_event *pev, } } + if (pev->uprobes) + return 1; + /* Currently just checking function name from symbol map */ sym = __find_kernel_function_by_name(tev->point.symbol, NULL); if (!sym) { @@ -1894,12 +1992,18 @@ int add_perf_probe_events(struct perf_probe_event *pevs, int npevs, int i, j, ret; struct __event_package *pkgs; + ret = 0; pkgs = zalloc(sizeof(struct __event_package) * npevs); + if (pkgs == NULL) return -ENOMEM; - /* Init vmlinux path */ - ret = init_vmlinux(); + if (!pevs->uprobes) + /* Init vmlinux path */ + ret = init_vmlinux(); + else + ret = init_user_exec(); + if (ret < 0) { free(pkgs); return ret; @@ -1971,23 +2075,15 @@ error: return ret; } -static int del_trace_probe_event(int fd, const char *group, - const char *event, struct strlist *namelist) +static int del_trace_probe_event(int fd, const char *buf, + struct strlist *namelist) { - char buf[128]; struct str_node *ent, *n; - int found = 0, ret = 0; - - ret = e_snprintf(buf, 128, "%s:%s", group, event); - if (ret < 0) { - pr_err("Failed to copy event.\n"); - return ret; - } + int ret = -1; if (strpbrk(buf, "*?")) { /* Glob-exp */ strlist__for_each_safe(ent, n, namelist) if (strglobmatch(ent->s, buf)) { - found++; ret = __del_trace_probe_event(fd, ent); if (ret < 0) break; @@ -1996,40 +2092,43 @@ static int del_trace_probe_event(int fd, const char *group, } else { ent = strlist__find(namelist, buf); if (ent) { - found++; ret = __del_trace_probe_event(fd, ent); if (ret >= 0) strlist__remove(namelist, ent); } } - if (found == 0 && ret >= 0) - pr_info("Info: Event \"%s\" does not exist.\n", buf); return ret; } int del_perf_probe_events(struct strlist *dellist) { - int fd, ret = 0; + int ret = -1, ufd = -1, kfd = -1; + char buf[128]; const char *group, *event; char *p, *str; struct str_node *ent; - struct strlist *namelist; - - fd = open_kprobe_events(true); - if (fd < 0) - return fd; + struct strlist *namelist = NULL, *unamelist = NULL; /* Get current event names */ - namelist = get_probe_trace_event_names(fd, true); - if (namelist == NULL) - return -EINVAL; + kfd = open_kprobe_events(true); + if (kfd < 0) + return kfd; + + namelist = get_probe_trace_event_names(kfd, true); + ufd = open_uprobe_events(true); + + if (ufd >= 0) + unamelist = get_probe_trace_event_names(ufd, true); + + if (namelist == NULL && unamelist == NULL) + goto error; strlist__for_each(ent, dellist) { str = strdup(ent->s); if (str == NULL) { ret = -ENOMEM; - break; + goto error; } pr_debug("Parsing: %s\n", str); p = strchr(str, ':'); @@ -2041,17 +2140,42 @@ int del_perf_probe_events(struct strlist *dellist) group = "*"; event = str; } + + ret = e_snprintf(buf, 128, "%s:%s", group, event); + if (ret < 0) { + pr_err("Failed to copy event."); + free(str); + goto error; + } + pr_debug("Group: %s, Event: %s\n", group, event); - ret = del_trace_probe_event(fd, group, event, namelist); + + if (namelist) + ret = del_trace_probe_event(kfd, buf, namelist); + + if (unamelist && ret != 0) + ret = del_trace_probe_event(ufd, buf, unamelist); + + if (ret != 0) + pr_info("Info: Event \"%s\" does not exist.\n", buf); + free(str); - if (ret < 0) - break; } - strlist__delete(namelist); - close(fd); + +error: + if (kfd >= 0) { + strlist__delete(namelist); + close(kfd); + } + + if (ufd >= 0) { + strlist__delete(unamelist); + close(ufd); + } return ret; } + /* TODO: don't use a global variable for filter ... */ static struct strfilter *available_func_filter; @@ -2068,30 +2192,152 @@ static int filter_available_functions(struct map *map __unused, return 1; } -int show_available_funcs(const char *target, struct strfilter *_filter) +static int __show_available_funcs(struct map *map) +{ + if (map__load(map, filter_available_functions)) { + pr_err("Failed to load map.\n"); + return -EINVAL; + } + if (!dso__sorted_by_name(map->dso, map->type)) + dso__sort_by_name(map->dso, map->type); + + dso__fprintf_symbols_by_name(map->dso, map->type, stdout); + return 0; +} + +static int available_kernel_funcs(const char *module) { struct map *map; int ret; - setup_pager(); - ret = init_vmlinux(); if (ret < 0) return ret; - map = kernel_get_module_map(target); + map = kernel_get_module_map(module); if (!map) { - pr_err("Failed to find %s map.\n", (target) ? : "kernel"); + pr_err("Failed to find %s map.\n", (module) ? : "kernel"); return -EINVAL; } + return __show_available_funcs(map); +} + +static int available_user_funcs(const char *target) +{ + struct map *map; + int ret; + + ret = init_user_exec(); + if (ret < 0) + return ret; + + map = dso__new_map(target); + ret = __show_available_funcs(map); + dso__delete(map->dso); + map__delete(map); + return ret; +} + +int show_available_funcs(const char *target, struct strfilter *_filter, + bool user) +{ + setup_pager(); available_func_filter = _filter; + + if (!user) + return available_kernel_funcs(target); + + return available_user_funcs(target); +} + +/* + * uprobe_events only accepts address: + * Convert function and any offset to address + */ +static int convert_name_to_addr(struct perf_probe_event *pev, const char *exec) +{ + struct perf_probe_point *pp = &pev->point; + struct symbol *sym; + struct map *map = NULL; + char *function = NULL, *name = NULL; + int ret = -EINVAL; + unsigned long long vaddr = 0; + + if (!pp->function) { + pr_warning("No function specified for uprobes"); + goto out; + } + + function = strdup(pp->function); + if (!function) { + pr_warning("Failed to allocate memory by strdup.\n"); + ret = -ENOMEM; + goto out; + } + + name = realpath(exec, NULL); + if (!name) { + pr_warning("Cannot find realpath for %s.\n", exec); + goto out; + } + map = dso__new_map(name); + if (!map) { + pr_warning("Cannot find appropriate DSO for %s.\n", exec); + goto out; + } + available_func_filter = strfilter__new(function, NULL); if (map__load(map, filter_available_functions)) { pr_err("Failed to load map.\n"); - return -EINVAL; + goto out; } - if (!dso__sorted_by_name(map->dso, map->type)) - dso__sort_by_name(map->dso, map->type); - dso__fprintf_symbols_by_name(map->dso, map->type, stdout); - return 0; + sym = map__find_symbol_by_name(map, function, NULL); + if (!sym) { + pr_warning("Cannot find %s in DSO %s\n", function, exec); + goto out; + } + + if (map->start > sym->start) + vaddr = map->start; + vaddr += sym->start + pp->offset + map->pgoff; + pp->offset = 0; + + if (!pev->event) { + pev->event = function; + function = NULL; + } + if (!pev->group) { + char *ptr1, *ptr2; + + pev->group = zalloc(sizeof(char *) * 64); + ptr1 = strdup(basename(exec)); + if (ptr1) { + ptr2 = strpbrk(ptr1, "-._"); + if (ptr2) + *ptr2 = '\0'; + e_snprintf(pev->group, 64, "%s_%s", PERFPROBE_GROUP, + ptr1); + free(ptr1); + } + } + free(pp->function); + pp->function = zalloc(sizeof(char *) * MAX_PROBE_ARGS); + if (!pp->function) { + ret = -ENOMEM; + pr_warning("Failed to allocate memory by zalloc.\n"); + goto out; + } + e_snprintf(pp->function, MAX_PROBE_ARGS, "0x%llx", vaddr); + ret = 0; + +out: + if (map) { + dso__delete(map->dso); + map__delete(map); + } + if (function) + free(function); + if (name) + free(name); + return ret; } diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index a7dee835f49c..f9f3de8b4220 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -7,7 +7,7 @@ extern bool probe_event_dry_run; -/* kprobe-tracer tracing point */ +/* kprobe-tracer and uprobe-tracer tracing point */ struct probe_trace_point { char *symbol; /* Base symbol */ char *module; /* Module name */ @@ -21,7 +21,7 @@ struct probe_trace_arg_ref { long offset; /* Offset value */ }; -/* kprobe-tracer tracing argument */ +/* kprobe-tracer and uprobe-tracer tracing argument */ struct probe_trace_arg { char *name; /* Argument name */ char *value; /* Base value */ @@ -29,12 +29,13 @@ struct probe_trace_arg { struct probe_trace_arg_ref *ref; /* Referencing offset */ }; -/* kprobe-tracer tracing event (point + arg) */ +/* kprobe-tracer and uprobe-tracer tracing event (point + arg) */ struct probe_trace_event { char *event; /* Event name */ char *group; /* Group name */ struct probe_trace_point point; /* Trace point */ int nargs; /* Number of args */ + bool uprobes; /* uprobes only */ struct probe_trace_arg *args; /* Arguments */ }; @@ -70,6 +71,7 @@ struct perf_probe_event { char *group; /* Group name */ struct perf_probe_point point; /* Probe point */ int nargs; /* Number of arguments */ + bool uprobes; struct perf_probe_arg *args; /* Arguments */ }; @@ -129,8 +131,8 @@ extern int show_line_range(struct line_range *lr, const char *module); extern int show_available_vars(struct perf_probe_event *pevs, int npevs, int max_probe_points, const char *module, struct strfilter *filter, bool externs); -extern int show_available_funcs(const char *module, struct strfilter *filter); - +extern int show_available_funcs(const char *module, struct strfilter *filter, + bool user); /* Maximum index number of event-name postfix */ #define MAX_EVENT_INDEX 1024 diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 2cc162d3b78c..d448984ed789 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -972,10 +972,12 @@ static int probe_point_search_cb(Dwarf_Die *sp_die, void *data) struct dwarf_callback_param *param = data; struct probe_finder *pf = param->data; struct perf_probe_point *pp = &pf->pev->point; + Dwarf_Attribute attr; /* Check tag and diename */ if (dwarf_tag(sp_die) != DW_TAG_subprogram || - !die_compare_name(sp_die, pp->function)) + !die_compare_name(sp_die, pp->function) || + dwarf_attr(sp_die, DW_AT_declaration, &attr)) return DWARF_CB_OK; /* Check declared file */ diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c index e30749e38a9b..4c1b3d72a1d2 100644 --- a/tools/perf/util/scripting-engines/trace-event-perl.c +++ b/tools/perf/util/scripting-engines/trace-event-perl.c @@ -56,7 +56,7 @@ INTERP my_perl; #define FTRACE_MAX_EVENT \ ((1 << (sizeof(unsigned short) * 8)) - 1) -struct event *events[FTRACE_MAX_EVENT]; +struct event_format *events[FTRACE_MAX_EVENT]; extern struct scripting_context *scripting_context; @@ -181,7 +181,7 @@ static void define_flag_field(const char *ev_name, LEAVE; } -static void define_event_symbols(struct event *event, +static void define_event_symbols(struct event_format *event, const char *ev_name, struct print_arg *args) { @@ -209,6 +209,8 @@ static void define_event_symbols(struct event *event, define_symbolic_values(args->symbol.symbols, ev_name, cur_field_name); break; + case PRINT_BSTRING: + case PRINT_DYNAMIC_ARRAY: case PRINT_STRING: break; case PRINT_TYPE: @@ -220,7 +222,9 @@ static void define_event_symbols(struct event *event, define_event_symbols(event, ev_name, args->op.left); define_event_symbols(event, ev_name, args->op.right); break; + case PRINT_FUNC: default: + pr_err("Unsupported print arg type\n"); /* we should warn... */ return; } @@ -229,10 +233,10 @@ static void define_event_symbols(struct event *event, define_event_symbols(event, ev_name, args->next); } -static inline struct event *find_cache_event(int type) +static inline struct event_format *find_cache_event(int type) { static char ev_name[256]; - struct event *event; + struct event_format *event; if (events[type]) return events[type]; @@ -258,7 +262,7 @@ static void perl_process_tracepoint(union perf_event *pevent __unused, static char handler[256]; unsigned long long val; unsigned long s, ns; - struct event *event; + struct event_format *event; int type; int pid; int cpu = sample->cpu; @@ -446,7 +450,7 @@ static int perl_stop_script(void) static int perl_generate_script(const char *outfile) { - struct event *event = NULL; + struct event_format *event = NULL; struct format_field *f; char fname[PATH_MAX]; int not_first, count; diff --git a/tools/perf/util/scripting-engines/trace-event-python.c b/tools/perf/util/scripting-engines/trace-event-python.c index c2623c6f9b51..acb9795286c4 100644 --- a/tools/perf/util/scripting-engines/trace-event-python.c +++ b/tools/perf/util/scripting-engines/trace-event-python.c @@ -37,7 +37,7 @@ PyMODINIT_FUNC initperf_trace_context(void); #define FTRACE_MAX_EVENT \ ((1 << (sizeof(unsigned short) * 8)) - 1) -struct event *events[FTRACE_MAX_EVENT]; +struct event_format *events[FTRACE_MAX_EVENT]; #define MAX_FIELDS 64 #define N_COMMON_FIELDS 7 @@ -136,7 +136,7 @@ static void define_field(enum print_arg_type field_type, Py_DECREF(t); } -static void define_event_symbols(struct event *event, +static void define_event_symbols(struct event_format *event, const char *ev_name, struct print_arg *args) { @@ -178,6 +178,10 @@ static void define_event_symbols(struct event *event, define_event_symbols(event, ev_name, args->op.right); break; default: + /* gcc warns for these? */ + case PRINT_BSTRING: + case PRINT_DYNAMIC_ARRAY: + case PRINT_FUNC: /* we should warn... */ return; } @@ -186,10 +190,10 @@ static void define_event_symbols(struct event *event, define_event_symbols(event, ev_name, args->next); } -static inline struct event *find_cache_event(int type) +static inline struct event_format *find_cache_event(int type) { static char ev_name[256]; - struct event *event; + struct event_format *event; if (events[type]) return events[type]; @@ -216,7 +220,7 @@ static void python_process_event(union perf_event *pevent __unused, struct format_field *field; unsigned long long val; unsigned long s, ns; - struct event *event; + struct event_format *event; unsigned n = 0; int type; int pid; @@ -436,7 +440,7 @@ out: static int python_generate_script(const char *outfile) { - struct event *event = NULL; + struct event_format *event = NULL; struct format_field *f; char fname[PATH_MAX]; int not_first, count; diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 002ebbf59f48..2600916efa83 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -140,6 +140,7 @@ struct perf_session *perf_session__new(const char *filename, int mode, INIT_LIST_HEAD(&self->ordered_samples.sample_cache); INIT_LIST_HEAD(&self->ordered_samples.to_free); machine__init(&self->host_machine, "", HOST_KERNEL_ID); + hists__init(&self->hists); if (mode == O_RDONLY) { if (perf_session__open(self, force) < 0) @@ -287,7 +288,8 @@ struct branch_info *machine__resolve_bstack(struct machine *self, return bi; } -int machine__resolve_callchain(struct machine *self, struct perf_evsel *evsel, +int machine__resolve_callchain(struct machine *self, + struct perf_evsel *evsel __used, struct thread *thread, struct ip_callchain *chain, struct symbol **parent) @@ -296,7 +298,12 @@ int machine__resolve_callchain(struct machine *self, struct perf_evsel *evsel, unsigned int i; int err; - callchain_cursor_reset(&evsel->hists.callchain_cursor); + callchain_cursor_reset(&callchain_cursor); + + if (chain->nr > PERF_MAX_STACK_DEPTH) { + pr_warning("corrupted callchain. skipping...\n"); + return 0; + } for (i = 0; i < chain->nr; i++) { u64 ip; @@ -316,7 +323,14 @@ int machine__resolve_callchain(struct machine *self, struct perf_evsel *evsel, case PERF_CONTEXT_USER: cpumode = PERF_RECORD_MISC_USER; break; default: - break; + pr_debug("invalid callchain context: " + "%"PRId64"\n", (s64) ip); + /* + * It seems the callchain is corrupted. + * Discard all. + */ + callchain_cursor_reset(&callchain_cursor); + return 0; } continue; } @@ -332,7 +346,7 @@ int machine__resolve_callchain(struct machine *self, struct perf_evsel *evsel, break; } - err = callchain_cursor_append(&evsel->hists.callchain_cursor, + err = callchain_cursor_append(&callchain_cursor, ip, al.map, al.sym); if (err) return err; @@ -440,37 +454,65 @@ void mem_bswap_64(void *src, int byte_size) } } -static void perf_event__all64_swap(union perf_event *event) +static void swap_sample_id_all(union perf_event *event, void *data) +{ + void *end = (void *) event + event->header.size; + int size = end - data; + + BUG_ON(size % sizeof(u64)); + mem_bswap_64(data, size); +} + +static void perf_event__all64_swap(union perf_event *event, + bool sample_id_all __used) { struct perf_event_header *hdr = &event->header; mem_bswap_64(hdr + 1, event->header.size - sizeof(*hdr)); } -static void perf_event__comm_swap(union perf_event *event) +static void perf_event__comm_swap(union perf_event *event, bool sample_id_all) { event->comm.pid = bswap_32(event->comm.pid); event->comm.tid = bswap_32(event->comm.tid); + + if (sample_id_all) { + void *data = &event->comm.comm; + + data += ALIGN(strlen(data) + 1, sizeof(u64)); + swap_sample_id_all(event, data); + } } -static void perf_event__mmap_swap(union perf_event *event) +static void perf_event__mmap_swap(union perf_event *event, + bool sample_id_all) { event->mmap.pid = bswap_32(event->mmap.pid); event->mmap.tid = bswap_32(event->mmap.tid); event->mmap.start = bswap_64(event->mmap.start); event->mmap.len = bswap_64(event->mmap.len); event->mmap.pgoff = bswap_64(event->mmap.pgoff); + + if (sample_id_all) { + void *data = &event->mmap.filename; + + data += ALIGN(strlen(data) + 1, sizeof(u64)); + swap_sample_id_all(event, data); + } } -static void perf_event__task_swap(union perf_event *event) +static void perf_event__task_swap(union perf_event *event, bool sample_id_all) { event->fork.pid = bswap_32(event->fork.pid); event->fork.tid = bswap_32(event->fork.tid); event->fork.ppid = bswap_32(event->fork.ppid); event->fork.ptid = bswap_32(event->fork.ptid); event->fork.time = bswap_64(event->fork.time); + + if (sample_id_all) + swap_sample_id_all(event, &event->fork + 1); } -static void perf_event__read_swap(union perf_event *event) +static void perf_event__read_swap(union perf_event *event, bool sample_id_all) { event->read.pid = bswap_32(event->read.pid); event->read.tid = bswap_32(event->read.tid); @@ -478,6 +520,41 @@ static void perf_event__read_swap(union perf_event *event) event->read.time_enabled = bswap_64(event->read.time_enabled); event->read.time_running = bswap_64(event->read.time_running); event->read.id = bswap_64(event->read.id); + + if (sample_id_all) + swap_sample_id_all(event, &event->read + 1); +} + +static u8 revbyte(u8 b) +{ + int rev = (b >> 4) | ((b & 0xf) << 4); + rev = ((rev & 0xcc) >> 2) | ((rev & 0x33) << 2); + rev = ((rev & 0xaa) >> 1) | ((rev & 0x55) << 1); + return (u8) rev; +} + +/* + * XXX this is hack in attempt to carry flags bitfield + * throught endian village. ABI says: + * + * Bit-fields are allocated from right to left (least to most significant) + * on little-endian implementations and from left to right (most to least + * significant) on big-endian implementations. + * + * The above seems to be byte specific, so we need to reverse each + * byte of the bitfield. 'Internet' also says this might be implementation + * specific and we probably need proper fix and carry perf_event_attr + * bitfield flags in separate data file FEAT_ section. Thought this seems + * to work for now. + */ +static void swap_bitfield(u8 *p, unsigned len) +{ + unsigned i; + + for (i = 0; i < len; i++) { + *p = revbyte(*p); + p++; + } } /* exported for swapping attributes in file header */ @@ -493,9 +570,12 @@ void perf_event__attr_swap(struct perf_event_attr *attr) attr->bp_type = bswap_32(attr->bp_type); attr->bp_addr = bswap_64(attr->bp_addr); attr->bp_len = bswap_64(attr->bp_len); + + swap_bitfield((u8 *) (&attr->read_format + 1), sizeof(u64)); } -static void perf_event__hdr_attr_swap(union perf_event *event) +static void perf_event__hdr_attr_swap(union perf_event *event, + bool sample_id_all __used) { size_t size; @@ -506,18 +586,21 @@ static void perf_event__hdr_attr_swap(union perf_event *event) mem_bswap_64(event->attr.id, size); } -static void perf_event__event_type_swap(union perf_event *event) +static void perf_event__event_type_swap(union perf_event *event, + bool sample_id_all __used) { event->event_type.event_type.event_id = bswap_64(event->event_type.event_type.event_id); } -static void perf_event__tracing_data_swap(union perf_event *event) +static void perf_event__tracing_data_swap(union perf_event *event, + bool sample_id_all __used) { event->tracing_data.size = bswap_32(event->tracing_data.size); } -typedef void (*perf_event__swap_op)(union perf_event *event); +typedef void (*perf_event__swap_op)(union perf_event *event, + bool sample_id_all); static perf_event__swap_op perf_event__swap_ops[] = { [PERF_RECORD_MMAP] = perf_event__mmap_swap, @@ -825,8 +908,16 @@ static struct machine * { const u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; - if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL && perf_guest) - return perf_session__find_machine(session, event->ip.pid); + if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL && perf_guest) { + u32 pid; + + if (event->header.type == PERF_RECORD_MMAP) + pid = event->mmap.pid; + else + pid = event->ip.pid; + + return perf_session__find_machine(session, pid); + } return perf_session__find_host_machine(session); } @@ -867,11 +958,11 @@ static int perf_session_deliver_event(struct perf_session *session, dump_sample(session, event, sample); if (evsel == NULL) { ++session->hists.stats.nr_unknown_id; - return -1; + return 0; } if (machine == NULL) { ++session->hists.stats.nr_unprocessable_samples; - return -1; + return 0; } return tool->sample(tool, event, sample, evsel, machine); case PERF_RECORD_MMAP: @@ -943,6 +1034,15 @@ static int perf_session__process_user_event(struct perf_session *session, union } } +static void event_swap(union perf_event *event, bool sample_id_all) +{ + perf_event__swap_op swap; + + swap = perf_event__swap_ops[event->header.type]; + if (swap) + swap(event, sample_id_all); +} + static int perf_session__process_event(struct perf_session *session, union perf_event *event, struct perf_tool *tool, @@ -951,9 +1051,8 @@ static int perf_session__process_event(struct perf_session *session, struct perf_sample sample; int ret; - if (session->header.needs_swap && - perf_event__swap_ops[event->header.type]) - perf_event__swap_ops[event->header.type](event); + if (session->header.needs_swap) + event_swap(event, session->sample_id_all); if (event->header.type >= PERF_RECORD_HEADER_MAX) return -EINVAL; @@ -1055,8 +1154,9 @@ volatile int session_done; static int __perf_session__process_pipe_events(struct perf_session *self, struct perf_tool *tool) { - union perf_event event; - uint32_t size; + union perf_event *event; + uint32_t size, cur_size = 0; + void *buf = NULL; int skip = 0; u64 head; int err; @@ -1065,8 +1165,14 @@ static int __perf_session__process_pipe_events(struct perf_session *self, perf_tool__fill_defaults(tool); head = 0; + cur_size = sizeof(union perf_event); + + buf = malloc(cur_size); + if (!buf) + return -errno; more: - err = readn(self->fd, &event, sizeof(struct perf_event_header)); + event = buf; + err = readn(self->fd, event, sizeof(struct perf_event_header)); if (err <= 0) { if (err == 0) goto done; @@ -1076,13 +1182,23 @@ more: } if (self->header.needs_swap) - perf_event_header__bswap(&event.header); + perf_event_header__bswap(&event->header); - size = event.header.size; + size = event->header.size; if (size == 0) size = 8; - p = &event; + if (size > cur_size) { + void *new = realloc(buf, size); + if (!new) { + pr_err("failed to allocate memory to read event\n"); + goto out_err; + } + buf = new; + cur_size = size; + event = buf; + } + p = event; p += sizeof(struct perf_event_header); if (size - sizeof(struct perf_event_header)) { @@ -1098,17 +1214,11 @@ more: } } - if ((skip = perf_session__process_event(self, &event, tool, head)) < 0) { - dump_printf("%#" PRIx64 " [%#x]: skipping unknown header type: %d\n", - head, event.header.size, event.header.type); - /* - * assume we lost track of the stream, check alignment, and - * increment a single u64 in the hope to catch on again 'soon'. - */ - if (unlikely(head & 7)) - head &= ~7ULL; - - size = 8; + if ((skip = perf_session__process_event(self, event, tool, head)) < 0) { + pr_err("%#" PRIx64 " [%#x]: failed to process type: %d\n", + head, event->header.size, event->header.type); + err = -EINVAL; + goto out_err; } head += size; @@ -1121,6 +1231,7 @@ more: done: err = 0; out_err: + free(buf); perf_session__warn_about_errors(self, tool); perf_session_free_sample_buffers(self); return err; @@ -1217,17 +1328,11 @@ more: if (size == 0 || perf_session__process_event(session, event, tool, file_pos) < 0) { - dump_printf("%#" PRIx64 " [%#x]: skipping unknown header type: %d\n", - file_offset + head, event->header.size, - event->header.type); - /* - * assume we lost track of the stream, check alignment, and - * increment a single u64 in the hope to catch on again 'soon'. - */ - if (unlikely(head & 7)) - head &= ~7ULL; - - size = 8; + pr_err("%#" PRIx64 " [%#x]: failed to process type: %d\n", + file_offset + head, event->header.size, + event->header.type); + err = -EINVAL; + goto out_err; } head += size; @@ -1379,7 +1484,6 @@ void perf_event__print_ip(union perf_event *event, struct perf_sample *sample, int print_sym, int print_dso, int print_symoffset) { struct addr_location al; - struct callchain_cursor *cursor = &evsel->hists.callchain_cursor; struct callchain_cursor_node *node; if (perf_event__preprocess_sample(event, machine, &al, sample, @@ -1397,10 +1501,10 @@ void perf_event__print_ip(union perf_event *event, struct perf_sample *sample, error("Failed to resolve callchain. Skipping\n"); return; } - callchain_cursor_commit(cursor); + callchain_cursor_commit(&callchain_cursor); while (1) { - node = callchain_cursor_current(cursor); + node = callchain_cursor_current(&callchain_cursor); if (!node) break; @@ -1411,12 +1515,12 @@ void perf_event__print_ip(union perf_event *event, struct perf_sample *sample, } if (print_dso) { printf(" ("); - map__fprintf_dsoname(al.map, stdout); + map__fprintf_dsoname(node->map, stdout); printf(")"); } printf("\n"); - callchain_cursor_advance(cursor); + callchain_cursor_advance(&callchain_cursor); } } else { diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 5dd83c3e2c0c..3e2e5ea0f03f 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1,6 +1,5 @@ #include <dirent.h> #include <errno.h> -#include <libgen.h> #include <stdlib.h> #include <stdio.h> #include <string.h> @@ -51,6 +50,8 @@ struct symbol_conf symbol_conf = { int dso__name_len(const struct dso *dso) { + if (!dso) + return strlen("[unknown]"); if (verbose) return dso->long_name_len; @@ -322,6 +323,7 @@ struct dso *dso__new(const char *name) dso->sorted_by_name = 0; dso->has_build_id = 0; dso->kernel = DSO_TYPE_USER; + dso->needs_swap = DSO_SWAP__UNSET; INIT_LIST_HEAD(&dso->node); } @@ -976,8 +978,9 @@ static Elf_Scn *elf_section_by_name(Elf *elf, GElf_Ehdr *ep, * And always look at the original dso, not at debuginfo packages, that * have the PLT data stripped out (shdr_rel_plt.sh_type == SHT_NOBITS). */ -static int dso__synthesize_plt_symbols(struct dso *dso, struct map *map, - symbol_filter_t filter) +static int +dso__synthesize_plt_symbols(struct dso *dso, char *name, struct map *map, + symbol_filter_t filter) { uint32_t nr_rel_entries, idx; GElf_Sym sym; @@ -992,10 +995,7 @@ static int dso__synthesize_plt_symbols(struct dso *dso, struct map *map, char sympltname[1024]; Elf *elf; int nr = 0, symidx, fd, err = 0; - char name[PATH_MAX]; - snprintf(name, sizeof(name), "%s%s", - symbol_conf.symfs, dso->long_name); fd = open(name, O_RDONLY); if (fd < 0) goto out; @@ -1157,6 +1157,33 @@ static size_t elf_addr_to_index(Elf *elf, GElf_Addr addr) return -1; } +static int dso__swap_init(struct dso *dso, unsigned char eidata) +{ + static unsigned int const endian = 1; + + dso->needs_swap = DSO_SWAP__NO; + + switch (eidata) { + case ELFDATA2LSB: + /* We are big endian, DSO is little endian. */ + if (*(unsigned char const *)&endian != 1) + dso->needs_swap = DSO_SWAP__YES; + break; + + case ELFDATA2MSB: + /* We are little endian, DSO is big endian. */ + if (*(unsigned char const *)&endian != 0) + dso->needs_swap = DSO_SWAP__YES; + break; + + default: + pr_err("unrecognized DSO data encoding %d\n", eidata); + return -EINVAL; + } + + return 0; +} + static int dso__load_sym(struct dso *dso, struct map *map, const char *name, int fd, symbol_filter_t filter, int kmodule, int want_symtab) @@ -1188,6 +1215,9 @@ static int dso__load_sym(struct dso *dso, struct map *map, const char *name, goto out_elf_end; } + if (dso__swap_init(dso, ehdr.e_ident[EI_DATA])) + goto out_elf_end; + /* Always reject images with a mismatched build-id: */ if (dso->has_build_id) { u8 build_id[BUILD_ID_SIZE]; @@ -1273,7 +1303,7 @@ static int dso__load_sym(struct dso *dso, struct map *map, const char *name, if (opdsec && sym.st_shndx == opdidx) { u32 offset = sym.st_value - opdshdr.sh_addr; u64 *opd = opddata->d_buf + offset; - sym.st_value = *opd; + sym.st_value = DSO__SWAP(dso, u64, *opd); sym.st_shndx = elf_addr_to_index(elf, sym.st_value); } @@ -1702,8 +1732,9 @@ restart: continue; if (ret > 0) { - int nr_plt = dso__synthesize_plt_symbols(dso, map, - filter); + int nr_plt; + + nr_plt = dso__synthesize_plt_symbols(dso, name, map, filter); if (nr_plt > 0) ret += nr_plt; break; @@ -2783,3 +2814,14 @@ int machine__load_vmlinux_path(struct machine *machine, enum map_type type, return ret; } + +struct map *dso__new_map(const char *name) +{ + struct map *map = NULL; + struct dso *dso = dso__new(name); + + if (dso) + map = map__new2(0, dso, MAP__FUNCTION); + + return map; +} diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index ac49ef208a5f..af0752b1aca1 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -9,6 +9,7 @@ #include <linux/list.h> #include <linux/rbtree.h> #include <stdio.h> +#include <byteswap.h> #ifdef HAVE_CPLUS_DEMANGLE extern char *cplus_demangle(const char *, int); @@ -65,6 +66,11 @@ struct symbol { void symbol__delete(struct symbol *sym); +static inline size_t symbol__size(const struct symbol *sym) +{ + return sym->end - sym->start + 1; +} + struct strlist; struct symbol_conf { @@ -155,11 +161,18 @@ enum dso_kernel_type { DSO_TYPE_GUEST_KERNEL }; +enum dso_swap_type { + DSO_SWAP__UNSET, + DSO_SWAP__NO, + DSO_SWAP__YES, +}; + struct dso { struct list_head node; struct rb_root symbols[MAP__NR_TYPES]; struct rb_root symbol_names[MAP__NR_TYPES]; enum dso_kernel_type kernel; + enum dso_swap_type needs_swap; u8 adjust_symbols:1; u8 has_build_id:1; u8 hit:1; @@ -177,6 +190,28 @@ struct dso { char name[0]; }; +#define DSO__SWAP(dso, type, val) \ +({ \ + type ____r = val; \ + BUG_ON(dso->needs_swap == DSO_SWAP__UNSET); \ + if (dso->needs_swap == DSO_SWAP__YES) { \ + switch (sizeof(____r)) { \ + case 2: \ + ____r = bswap_16(val); \ + break; \ + case 4: \ + ____r = bswap_32(val); \ + break; \ + case 8: \ + ____r = bswap_64(val); \ + break; \ + default: \ + BUG_ON(1); \ + } \ + } \ + ____r; \ +}) + struct dso *dso__new(const char *name); void dso__delete(struct dso *dso); @@ -237,6 +272,7 @@ void dso__set_long_name(struct dso *dso, char *name); void dso__set_build_id(struct dso *dso, void *build_id); void dso__read_running_kernel_build_id(struct dso *dso, struct machine *machine); +struct map *dso__new_map(const char *name); struct symbol *dso__find_symbol(struct dso *dso, enum map_type type, u64 addr); struct symbol *dso__find_symbol_by_name(struct dso *dso, enum map_type type, diff --git a/tools/perf/util/target.c b/tools/perf/util/target.c new file mode 100644 index 000000000000..1064d5b148ad --- /dev/null +++ b/tools/perf/util/target.c @@ -0,0 +1,142 @@ +/* + * Helper functions for handling target threads/cpus + * + * Copyright (C) 2012, LG Electronics, Namhyung Kim <namhyung.kim@lge.com> + * + * Released under the GPL v2. + */ + +#include "target.h" +#include "debug.h" + +#include <pwd.h> +#include <string.h> + + +enum perf_target_errno perf_target__validate(struct perf_target *target) +{ + enum perf_target_errno ret = PERF_ERRNO_TARGET__SUCCESS; + + if (target->pid) + target->tid = target->pid; + + /* CPU and PID are mutually exclusive */ + if (target->tid && target->cpu_list) { + target->cpu_list = NULL; + if (ret == PERF_ERRNO_TARGET__SUCCESS) + ret = PERF_ERRNO_TARGET__PID_OVERRIDE_CPU; + } + + /* UID and PID are mutually exclusive */ + if (target->tid && target->uid_str) { + target->uid_str = NULL; + if (ret == PERF_ERRNO_TARGET__SUCCESS) + ret = PERF_ERRNO_TARGET__PID_OVERRIDE_UID; + } + + /* UID and CPU are mutually exclusive */ + if (target->uid_str && target->cpu_list) { + target->cpu_list = NULL; + if (ret == PERF_ERRNO_TARGET__SUCCESS) + ret = PERF_ERRNO_TARGET__UID_OVERRIDE_CPU; + } + + /* PID and SYSTEM are mutually exclusive */ + if (target->tid && target->system_wide) { + target->system_wide = false; + if (ret == PERF_ERRNO_TARGET__SUCCESS) + ret = PERF_ERRNO_TARGET__PID_OVERRIDE_SYSTEM; + } + + /* UID and SYSTEM are mutually exclusive */ + if (target->uid_str && target->system_wide) { + target->system_wide = false; + if (ret == PERF_ERRNO_TARGET__SUCCESS) + ret = PERF_ERRNO_TARGET__UID_OVERRIDE_SYSTEM; + } + + return ret; +} + +enum perf_target_errno perf_target__parse_uid(struct perf_target *target) +{ + struct passwd pwd, *result; + char buf[1024]; + const char *str = target->uid_str; + + target->uid = UINT_MAX; + if (str == NULL) + return PERF_ERRNO_TARGET__SUCCESS; + + /* Try user name first */ + getpwnam_r(str, &pwd, buf, sizeof(buf), &result); + + if (result == NULL) { + /* + * The user name not found. Maybe it's a UID number. + */ + char *endptr; + int uid = strtol(str, &endptr, 10); + + if (*endptr != '\0') + return PERF_ERRNO_TARGET__INVALID_UID; + + getpwuid_r(uid, &pwd, buf, sizeof(buf), &result); + + if (result == NULL) + return PERF_ERRNO_TARGET__USER_NOT_FOUND; + } + + target->uid = result->pw_uid; + return PERF_ERRNO_TARGET__SUCCESS; +} + +/* + * This must have a same ordering as the enum perf_target_errno. + */ +static const char *perf_target__error_str[] = { + "PID/TID switch overriding CPU", + "PID/TID switch overriding UID", + "UID switch overriding CPU", + "PID/TID switch overriding SYSTEM", + "UID switch overriding SYSTEM", + "Invalid User: %s", + "Problems obtaining information for user %s", +}; + +int perf_target__strerror(struct perf_target *target, int errnum, + char *buf, size_t buflen) +{ + int idx; + const char *msg; + + if (errnum >= 0) { + strerror_r(errnum, buf, buflen); + return 0; + } + + if (errnum < __PERF_ERRNO_TARGET__START || + errnum >= __PERF_ERRNO_TARGET__END) + return -1; + + idx = errnum - __PERF_ERRNO_TARGET__START; + msg = perf_target__error_str[idx]; + + switch (errnum) { + case PERF_ERRNO_TARGET__PID_OVERRIDE_CPU + ... PERF_ERRNO_TARGET__UID_OVERRIDE_SYSTEM: + snprintf(buf, buflen, "%s", msg); + break; + + case PERF_ERRNO_TARGET__INVALID_UID: + case PERF_ERRNO_TARGET__USER_NOT_FOUND: + snprintf(buf, buflen, msg, target->uid_str); + break; + + default: + /* cannot reach here */ + break; + } + + return 0; +} diff --git a/tools/perf/util/target.h b/tools/perf/util/target.h new file mode 100644 index 000000000000..a4be8575fda5 --- /dev/null +++ b/tools/perf/util/target.h @@ -0,0 +1,65 @@ +#ifndef _PERF_TARGET_H +#define _PERF_TARGET_H + +#include <stdbool.h> +#include <sys/types.h> + +struct perf_target { + const char *pid; + const char *tid; + const char *cpu_list; + const char *uid_str; + uid_t uid; + bool system_wide; + bool uses_mmap; +}; + +enum perf_target_errno { + PERF_ERRNO_TARGET__SUCCESS = 0, + + /* + * Choose an arbitrary negative big number not to clash with standard + * errno since SUS requires the errno has distinct positive values. + * See 'Issue 6' in the link below. + * + * http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/errno.h.html + */ + __PERF_ERRNO_TARGET__START = -10000, + + + /* for perf_target__validate() */ + PERF_ERRNO_TARGET__PID_OVERRIDE_CPU = __PERF_ERRNO_TARGET__START, + PERF_ERRNO_TARGET__PID_OVERRIDE_UID, + PERF_ERRNO_TARGET__UID_OVERRIDE_CPU, + PERF_ERRNO_TARGET__PID_OVERRIDE_SYSTEM, + PERF_ERRNO_TARGET__UID_OVERRIDE_SYSTEM, + + /* for perf_target__parse_uid() */ + PERF_ERRNO_TARGET__INVALID_UID, + PERF_ERRNO_TARGET__USER_NOT_FOUND, + + __PERF_ERRNO_TARGET__END, +}; + +enum perf_target_errno perf_target__validate(struct perf_target *target); +enum perf_target_errno perf_target__parse_uid(struct perf_target *target); + +int perf_target__strerror(struct perf_target *target, int errnum, char *buf, + size_t buflen); + +static inline bool perf_target__has_task(struct perf_target *target) +{ + return target->tid || target->pid || target->uid_str; +} + +static inline bool perf_target__has_cpu(struct perf_target *target) +{ + return target->system_wide || target->cpu_list; +} + +static inline bool perf_target__none(struct perf_target *target) +{ + return !perf_target__has_task(target) && !perf_target__has_cpu(target); +} + +#endif /* _PERF_TARGET_H */ diff --git a/tools/perf/util/thread_map.c b/tools/perf/util/thread_map.c index 84d9bd782004..9b5f856cc280 100644 --- a/tools/perf/util/thread_map.c +++ b/tools/perf/util/thread_map.c @@ -188,28 +188,27 @@ static struct thread_map *thread_map__new_by_pid_str(const char *pid_str) nt = realloc(threads, (sizeof(*threads) + sizeof(pid_t) * total_tasks)); if (nt == NULL) - goto out_free_threads; + goto out_free_namelist; threads = nt; - if (threads) { - for (i = 0; i < items; i++) - threads->map[j++] = atoi(namelist[i]->d_name); - threads->nr = total_tasks; - } - - for (i = 0; i < items; i++) + for (i = 0; i < items; i++) { + threads->map[j++] = atoi(namelist[i]->d_name); free(namelist[i]); + } + threads->nr = total_tasks; free(namelist); - - if (!threads) - break; } out: strlist__delete(slist); return threads; +out_free_namelist: + for (i = 0; i < items; i++) + free(namelist[i]); + free(namelist); + out_free_threads: free(threads); threads = NULL; diff --git a/tools/perf/util/thread_map.h b/tools/perf/util/thread_map.h index 7da80f14418b..f718df8a3c59 100644 --- a/tools/perf/util/thread_map.h +++ b/tools/perf/util/thread_map.h @@ -6,7 +6,7 @@ struct thread_map { int nr; - int map[]; + pid_t map[]; }; struct thread_map *thread_map__new_by_pid(pid_t pid); diff --git a/tools/perf/util/top.c b/tools/perf/util/top.c index 09fe579ccafb..abe0e8e95068 100644 --- a/tools/perf/util/top.c +++ b/tools/perf/util/top.c @@ -69,23 +69,24 @@ size_t perf_top__header_snprintf(struct perf_top *top, char *bf, size_t size) ret += SNPRINTF(bf + ret, size - ret, "], "); - if (top->target_pid) + if (top->target.pid) ret += SNPRINTF(bf + ret, size - ret, " (target_pid: %s", - top->target_pid); - else if (top->target_tid) + top->target.pid); + else if (top->target.tid) ret += SNPRINTF(bf + ret, size - ret, " (target_tid: %s", - top->target_tid); - else if (top->uid_str != NULL) + top->target.tid); + else if (top->target.uid_str != NULL) ret += SNPRINTF(bf + ret, size - ret, " (uid: %s", - top->uid_str); + top->target.uid_str); else ret += SNPRINTF(bf + ret, size - ret, " (all"); - if (top->cpu_list) + if (top->target.cpu_list) ret += SNPRINTF(bf + ret, size - ret, ", CPU%s: %s)", - top->evlist->cpus->nr > 1 ? "s" : "", top->cpu_list); + top->evlist->cpus->nr > 1 ? "s" : "", + top->target.cpu_list); else { - if (top->target_tid) + if (top->target.tid) ret += SNPRINTF(bf + ret, size - ret, ")"); else ret += SNPRINTF(bf + ret, size - ret, ", %d CPU%s)", diff --git a/tools/perf/util/top.h b/tools/perf/util/top.h index ce61cb2d1acf..33347ca89ee4 100644 --- a/tools/perf/util/top.h +++ b/tools/perf/util/top.h @@ -13,6 +13,7 @@ struct perf_session; struct perf_top { struct perf_tool tool; struct perf_evlist *evlist; + struct perf_target target; /* * Symbols will be added here in perf_event__process_sample and will * get out after decayed. @@ -23,10 +24,7 @@ struct perf_top { u64 guest_us_samples, guest_kernel_samples; int print_entries, count_filter, delay_secs; int freq; - const char *target_pid, *target_tid; - uid_t uid; bool hide_kernel_symbols, hide_user_symbols, zero; - bool system_wide; bool use_tui, use_stdio; bool sort_has_symbols; bool dont_use_callchains; @@ -37,7 +35,6 @@ struct perf_top { bool sample_id_all_missing; bool exclude_guest_missing; bool dump_symtab; - const char *cpu_list; struct hist_entry *sym_filter_entry; struct perf_evsel *sym_evsel; struct perf_session *session; @@ -47,7 +44,6 @@ struct perf_top { int realtime_prio; int sym_pcnt_filter; const char *sym_filter; - const char *uid_str; }; size_t perf_top__header_snprintf(struct perf_top *top, char *bf, size_t size); diff --git a/tools/perf/util/trace-event-info.c b/tools/perf/util/trace-event-info.c index fc22cf5c605f..a8d81c35ef66 100644 --- a/tools/perf/util/trace-event-info.c +++ b/tools/perf/util/trace-event-info.c @@ -68,7 +68,7 @@ struct events { }; -void *malloc_or_die(unsigned int size) +static void *malloc_or_die(unsigned int size) { void *data; @@ -448,6 +448,8 @@ static void tracing_data_header(void) else buf[0] = 0; + read_trace_init(buf[0], buf[0]); + write_or_die(buf, 1); /* save size of long */ diff --git a/tools/perf/util/trace-event-parse.c b/tools/perf/util/trace-event-parse.c index a4088ced1e64..df2fddbf0cd2 100644 --- a/tools/perf/util/trace-event-parse.c +++ b/tools/perf/util/trace-event-parse.c @@ -17,2171 +17,305 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - * - * The parts for function graph printing was taken and modified from the - * Linux Kernel that were written by Frederic Weisbecker. */ - #include <stdio.h> #include <stdlib.h> #include <string.h> +#include <ctype.h> #include <errno.h> #include "../perf.h" #include "util.h" #include "trace-event.h" -int header_page_ts_offset; -int header_page_ts_size; -int header_page_size_offset; int header_page_size_size; -int header_page_overwrite_offset; -int header_page_overwrite_size; +int header_page_ts_size; int header_page_data_offset; -int header_page_data_size; - -bool latency_format; - -static char *input_buf; -static unsigned long long input_buf_ptr; -static unsigned long long input_buf_siz; - -static int cpus; -static int long_size; -static int is_flag_field; -static int is_symbolic_field; - -static struct format_field * -find_any_field(struct event *event, const char *name); - -static void init_input_buf(char *buf, unsigned long long size) -{ - input_buf = buf; - input_buf_siz = size; - input_buf_ptr = 0; -} - -struct cmdline { - char *comm; - int pid; -}; - -static struct cmdline *cmdlines; -static int cmdline_count; - -static int cmdline_cmp(const void *a, const void *b) -{ - const struct cmdline *ca = a; - const struct cmdline *cb = b; - - if (ca->pid < cb->pid) - return -1; - if (ca->pid > cb->pid) - return 1; - - return 0; -} -void parse_cmdlines(char *file, int size __unused) -{ - struct cmdline_list { - struct cmdline_list *next; - char *comm; - int pid; - } *list = NULL, *item; - char *line; - char *next = NULL; - int i; +struct pevent *perf_pevent; +static struct pevent *pevent; - line = strtok_r(file, "\n", &next); - while (line) { - item = malloc_or_die(sizeof(*item)); - sscanf(line, "%d %as", &item->pid, - (float *)(void *)&item->comm); /* workaround gcc warning */ - item->next = list; - list = item; - line = strtok_r(NULL, "\n", &next); - cmdline_count++; - } - - cmdlines = malloc_or_die(sizeof(*cmdlines) * cmdline_count); - - i = 0; - while (list) { - cmdlines[i].pid = list->pid; - cmdlines[i].comm = list->comm; - i++; - item = list; - list = list->next; - free(item); - } - - qsort(cmdlines, cmdline_count, sizeof(*cmdlines), cmdline_cmp); -} - -static struct func_map { - unsigned long long addr; - char *func; - char *mod; -} *func_list; -static unsigned int func_count; - -static int func_cmp(const void *a, const void *b) -{ - const struct func_map *fa = a; - const struct func_map *fb = b; - - if (fa->addr < fb->addr) - return -1; - if (fa->addr > fb->addr) - return 1; - - return 0; -} - -void parse_proc_kallsyms(char *file, unsigned int size __unused) -{ - struct func_list { - struct func_list *next; - unsigned long long addr; - char *func; - char *mod; - } *list = NULL, *item; - char *line; - char *next = NULL; - char *addr_str; - char ch; - int ret __used; - int i; - - line = strtok_r(file, "\n", &next); - while (line) { - item = malloc_or_die(sizeof(*item)); - item->mod = NULL; - ret = sscanf(line, "%as %c %as\t[%as", - (float *)(void *)&addr_str, /* workaround gcc warning */ - &ch, - (float *)(void *)&item->func, - (float *)(void *)&item->mod); - item->addr = strtoull(addr_str, NULL, 16); - free(addr_str); - - /* truncate the extra ']' */ - if (item->mod) - item->mod[strlen(item->mod) - 1] = 0; - - - item->next = list; - list = item; - line = strtok_r(NULL, "\n", &next); - func_count++; - } - - func_list = malloc_or_die(sizeof(*func_list) * (func_count + 1)); - - i = 0; - while (list) { - func_list[i].func = list->func; - func_list[i].addr = list->addr; - func_list[i].mod = list->mod; - i++; - item = list; - list = list->next; - free(item); - } - - qsort(func_list, func_count, sizeof(*func_list), func_cmp); - - /* - * Add a special record at the end. - */ - func_list[func_count].func = NULL; - func_list[func_count].addr = 0; - func_list[func_count].mod = NULL; -} +bool latency_format; -/* - * We are searching for a record in between, not an exact - * match. - */ -static int func_bcmp(const void *a, const void *b) +int read_trace_init(int file_bigendian, int host_bigendian) { - const struct func_map *fa = a; - const struct func_map *fb = b; - - if ((fa->addr == fb->addr) || - - (fa->addr > fb->addr && - fa->addr < (fb+1)->addr)) + if (pevent) return 0; - if (fa->addr < fb->addr) - return -1; - - return 1; -} - -static struct func_map *find_func(unsigned long long addr) -{ - struct func_map *func; - struct func_map key; - - key.addr = addr; - - func = bsearch(&key, func_list, func_count, sizeof(*func_list), - func_bcmp); - - return func; -} - -void print_funcs(void) -{ - int i; - - for (i = 0; i < (int)func_count; i++) { - printf("%016llx %s", - func_list[i].addr, - func_list[i].func); - if (func_list[i].mod) - printf(" [%s]\n", func_list[i].mod); - else - printf("\n"); - } -} + perf_pevent = pevent_alloc(); + pevent = perf_pevent; -static struct printk_map { - unsigned long long addr; - char *printk; -} *printk_list; -static unsigned int printk_count; - -static int printk_cmp(const void *a, const void *b) -{ - const struct func_map *fa = a; - const struct func_map *fb = b; - - if (fa->addr < fb->addr) - return -1; - if (fa->addr > fb->addr) - return 1; + pevent_set_flag(pevent, PEVENT_NSEC_OUTPUT); + pevent_set_file_bigendian(pevent, file_bigendian); + pevent_set_host_bigendian(pevent, host_bigendian); return 0; } -static struct printk_map *find_printk(unsigned long long addr) -{ - struct printk_map *printk; - struct printk_map key; - - key.addr = addr; - - printk = bsearch(&key, printk_list, printk_count, sizeof(*printk_list), - printk_cmp); - - return printk; -} - -void parse_ftrace_printk(char *file, unsigned int size __unused) -{ - struct printk_list { - struct printk_list *next; - unsigned long long addr; - char *printk; - } *list = NULL, *item; - char *line; - char *next = NULL; - char *addr_str; - int i; - - line = strtok_r(file, "\n", &next); - while (line) { - addr_str = strsep(&line, ":"); - if (!line) { - warning("error parsing print strings"); - break; - } - item = malloc_or_die(sizeof(*item)); - item->addr = strtoull(addr_str, NULL, 16); - /* fmt still has a space, skip it */ - item->printk = strdup(line+1); - item->next = list; - list = item; - line = strtok_r(NULL, "\n", &next); - printk_count++; - } - - printk_list = malloc_or_die(sizeof(*printk_list) * printk_count + 1); - - i = 0; - while (list) { - printk_list[i].printk = list->printk; - printk_list[i].addr = list->addr; - i++; - item = list; - list = list->next; - free(item); - } - - qsort(printk_list, printk_count, sizeof(*printk_list), printk_cmp); -} - -void print_printk(void) -{ - int i; - - for (i = 0; i < (int)printk_count; i++) { - printf("%016llx %s\n", - printk_list[i].addr, - printk_list[i].printk); - } -} - -static struct event *alloc_event(void) -{ - struct event *event; - - event = malloc_or_die(sizeof(*event)); - memset(event, 0, sizeof(*event)); - - return event; -} - -enum event_type { - EVENT_ERROR, - EVENT_NONE, - EVENT_SPACE, - EVENT_NEWLINE, - EVENT_OP, - EVENT_DELIM, - EVENT_ITEM, - EVENT_DQUOTE, - EVENT_SQUOTE, -}; - -static struct event *event_list; - -static void add_event(struct event *event) +static int get_common_field(struct scripting_context *context, + int *offset, int *size, const char *type) { - event->next = event_list; - event_list = event; -} - -static int event_item_type(enum event_type type) -{ - switch (type) { - case EVENT_ITEM ... EVENT_SQUOTE: - return 1; - case EVENT_ERROR ... EVENT_DELIM: - default: - return 0; - } -} - -static void free_arg(struct print_arg *arg) -{ - if (!arg) - return; - - switch (arg->type) { - case PRINT_ATOM: - if (arg->atom.atom) - free(arg->atom.atom); - break; - case PRINT_NULL: - case PRINT_FIELD ... PRINT_OP: - default: - /* todo */ - break; - } - - free(arg); -} - -static enum event_type get_type(int ch) -{ - if (ch == '\n') - return EVENT_NEWLINE; - if (isspace(ch)) - return EVENT_SPACE; - if (isalnum(ch) || ch == '_') - return EVENT_ITEM; - if (ch == '\'') - return EVENT_SQUOTE; - if (ch == '"') - return EVENT_DQUOTE; - if (!isprint(ch)) - return EVENT_NONE; - if (ch == '(' || ch == ')' || ch == ',') - return EVENT_DELIM; - - return EVENT_OP; -} - -static int __read_char(void) -{ - if (input_buf_ptr >= input_buf_siz) - return -1; - - return input_buf[input_buf_ptr++]; -} - -static int __peek_char(void) -{ - if (input_buf_ptr >= input_buf_siz) - return -1; - - return input_buf[input_buf_ptr]; -} - -static enum event_type __read_token(char **tok) -{ - char buf[BUFSIZ]; - int ch, last_ch, quote_ch, next_ch; - int i = 0; - int tok_size = 0; - enum event_type type; - - *tok = NULL; - - - ch = __read_char(); - if (ch < 0) - return EVENT_NONE; - - type = get_type(ch); - if (type == EVENT_NONE) - return type; - - buf[i++] = ch; - - switch (type) { - case EVENT_NEWLINE: - case EVENT_DELIM: - *tok = malloc_or_die(2); - (*tok)[0] = ch; - (*tok)[1] = 0; - return type; - - case EVENT_OP: - switch (ch) { - case '-': - next_ch = __peek_char(); - if (next_ch == '>') { - buf[i++] = __read_char(); - break; - } - /* fall through */ - case '+': - case '|': - case '&': - case '>': - case '<': - last_ch = ch; - ch = __peek_char(); - if (ch != last_ch) - goto test_equal; - buf[i++] = __read_char(); - switch (last_ch) { - case '>': - case '<': - goto test_equal; - default: - break; - } - break; - case '!': - case '=': - goto test_equal; - default: /* what should we do instead? */ - break; - } - buf[i] = 0; - *tok = strdup(buf); - return type; - - test_equal: - ch = __peek_char(); - if (ch == '=') - buf[i++] = __read_char(); - break; - - case EVENT_DQUOTE: - case EVENT_SQUOTE: - /* don't keep quotes */ - i--; - quote_ch = ch; - last_ch = 0; - do { - if (i == (BUFSIZ - 1)) { - buf[i] = 0; - if (*tok) { - *tok = realloc(*tok, tok_size + BUFSIZ); - if (!*tok) - return EVENT_NONE; - strcat(*tok, buf); - } else - *tok = strdup(buf); - - if (!*tok) - return EVENT_NONE; - tok_size += BUFSIZ; - i = 0; - } - last_ch = ch; - ch = __read_char(); - buf[i++] = ch; - /* the '\' '\' will cancel itself */ - if (ch == '\\' && last_ch == '\\') - last_ch = 0; - } while (ch != quote_ch || last_ch == '\\'); - /* remove the last quote */ - i--; - goto out; - - case EVENT_ERROR ... EVENT_SPACE: - case EVENT_ITEM: - default: - break; - } - - while (get_type(__peek_char()) == type) { - if (i == (BUFSIZ - 1)) { - buf[i] = 0; - if (*tok) { - *tok = realloc(*tok, tok_size + BUFSIZ); - if (!*tok) - return EVENT_NONE; - strcat(*tok, buf); - } else - *tok = strdup(buf); - - if (!*tok) - return EVENT_NONE; - tok_size += BUFSIZ; - i = 0; - } - ch = __read_char(); - buf[i++] = ch; - } - - out: - buf[i] = 0; - if (*tok) { - *tok = realloc(*tok, tok_size + i); - if (!*tok) - return EVENT_NONE; - strcat(*tok, buf); - } else - *tok = strdup(buf); - if (!*tok) - return EVENT_NONE; - - return type; -} - -static void free_token(char *tok) -{ - if (tok) - free(tok); -} - -static enum event_type read_token(char **tok) -{ - enum event_type type; - - for (;;) { - type = __read_token(tok); - if (type != EVENT_SPACE) - return type; - - free_token(*tok); - } - - /* not reached */ - return EVENT_NONE; -} - -/* no newline */ -static enum event_type read_token_item(char **tok) -{ - enum event_type type; - - for (;;) { - type = __read_token(tok); - if (type != EVENT_SPACE && type != EVENT_NEWLINE) - return type; - - free_token(*tok); - } - - /* not reached */ - return EVENT_NONE; -} - -static int test_type(enum event_type type, enum event_type expect) -{ - if (type != expect) { - warning("Error: expected type %d but read %d", - expect, type); - return -1; - } - return 0; -} + struct event_format *event; + struct format_field *field; -static int __test_type_token(enum event_type type, char *token, - enum event_type expect, const char *expect_tok, - bool warn) -{ - if (type != expect) { - if (warn) - warning("Error: expected type %d but read %d", - expect, type); - return -1; - } + if (!*size) { + if (!pevent->events) + return 0; - if (strcmp(token, expect_tok) != 0) { - if (warn) - warning("Error: expected '%s' but read '%s'", - expect_tok, token); - return -1; + event = pevent->events[0]; + field = pevent_find_common_field(event, type); + if (!field) + return 0; + *offset = field->offset; + *size = field->size; } - return 0; -} -static int test_type_token(enum event_type type, char *token, - enum event_type expect, const char *expect_tok) -{ - return __test_type_token(type, token, expect, expect_tok, true); + return pevent_read_number(pevent, context->event_data + *offset, *size); } -static int __read_expect_type(enum event_type expect, char **tok, int newline_ok) -{ - enum event_type type; - - if (newline_ok) - type = read_token(tok); - else - type = read_token_item(tok); - return test_type(type, expect); -} - -static int read_expect_type(enum event_type expect, char **tok) -{ - return __read_expect_type(expect, tok, 1); -} - -static int __read_expected(enum event_type expect, const char *str, - int newline_ok, bool warn) +int common_lock_depth(struct scripting_context *context) { - enum event_type type; - char *token; + static int offset; + static int size; int ret; - if (newline_ok) - type = read_token(&token); - else - type = read_token_item(&token); - - ret = __test_type_token(type, token, expect, str, warn); - - free_token(token); + ret = get_common_field(context, &size, &offset, + "common_lock_depth"); + if (ret < 0) + return -1; return ret; } -static int read_expected(enum event_type expect, const char *str) -{ - return __read_expected(expect, str, 1, true); -} - -static int read_expected_item(enum event_type expect, const char *str) -{ - return __read_expected(expect, str, 0, true); -} - -static char *event_read_name(void) -{ - char *token; - - if (read_expected(EVENT_ITEM, "name") < 0) - return NULL; - - if (read_expected(EVENT_OP, ":") < 0) - return NULL; - - if (read_expect_type(EVENT_ITEM, &token) < 0) - goto fail; - - return token; - - fail: - free_token(token); - return NULL; -} - -static int event_read_id(void) +int common_flags(struct scripting_context *context) { - char *token; - int id; - - if (read_expected_item(EVENT_ITEM, "ID") < 0) - return -1; + static int offset; + static int size; + int ret; - if (read_expected(EVENT_OP, ":") < 0) + ret = get_common_field(context, &size, &offset, + "common_flags"); + if (ret < 0) return -1; - if (read_expect_type(EVENT_ITEM, &token) < 0) - goto fail; - - id = strtoul(token, NULL, 0); - free_token(token); - return id; - - fail: - free_token(token); - return -1; -} - -static int field_is_string(struct format_field *field) -{ - if ((field->flags & FIELD_IS_ARRAY) && - (!strstr(field->type, "char") || !strstr(field->type, "u8") || - !strstr(field->type, "s8"))) - return 1; - - return 0; -} - -static int field_is_dynamic(struct format_field *field) -{ - if (!strncmp(field->type, "__data_loc", 10)) - return 1; - - return 0; -} - -static int event_read_fields(struct event *event, struct format_field **fields) -{ - struct format_field *field = NULL; - enum event_type type; - char *token; - char *last_token; - int count = 0; - - do { - type = read_token(&token); - if (type == EVENT_NEWLINE) { - free_token(token); - return count; - } - - count++; - - if (test_type_token(type, token, EVENT_ITEM, "field")) - goto fail; - free_token(token); - - type = read_token(&token); - /* - * The ftrace fields may still use the "special" name. - * Just ignore it. - */ - if (event->flags & EVENT_FL_ISFTRACE && - type == EVENT_ITEM && strcmp(token, "special") == 0) { - free_token(token); - type = read_token(&token); - } - - if (test_type_token(type, token, EVENT_OP, ":") < 0) - return -1; - - if (read_expect_type(EVENT_ITEM, &token) < 0) - goto fail; - - last_token = token; - - field = malloc_or_die(sizeof(*field)); - memset(field, 0, sizeof(*field)); - - /* read the rest of the type */ - for (;;) { - type = read_token(&token); - if (type == EVENT_ITEM || - (type == EVENT_OP && strcmp(token, "*") == 0) || - /* - * Some of the ftrace fields are broken and have - * an illegal "." in them. - */ - (event->flags & EVENT_FL_ISFTRACE && - type == EVENT_OP && strcmp(token, ".") == 0)) { - - if (strcmp(token, "*") == 0) - field->flags |= FIELD_IS_POINTER; - - if (field->type) { - field->type = realloc(field->type, - strlen(field->type) + - strlen(last_token) + 2); - strcat(field->type, " "); - strcat(field->type, last_token); - } else - field->type = last_token; - last_token = token; - continue; - } - - break; - } - - if (!field->type) { - die("no type found"); - goto fail; - } - field->name = last_token; - - if (test_type(type, EVENT_OP)) - goto fail; - - if (strcmp(token, "[") == 0) { - enum event_type last_type = type; - char *brackets = token; - int len; - - field->flags |= FIELD_IS_ARRAY; - - type = read_token(&token); - while (strcmp(token, "]") != 0) { - if (last_type == EVENT_ITEM && - type == EVENT_ITEM) - len = 2; - else - len = 1; - last_type = type; - - brackets = realloc(brackets, - strlen(brackets) + - strlen(token) + len); - if (len == 2) - strcat(brackets, " "); - strcat(brackets, token); - free_token(token); - type = read_token(&token); - if (type == EVENT_NONE) { - die("failed to find token"); - goto fail; - } - } - - free_token(token); - - brackets = realloc(brackets, strlen(brackets) + 2); - strcat(brackets, "]"); - - /* add brackets to type */ - - type = read_token(&token); - /* - * If the next token is not an OP, then it is of - * the format: type [] item; - */ - if (type == EVENT_ITEM) { - field->type = realloc(field->type, - strlen(field->type) + - strlen(field->name) + - strlen(brackets) + 2); - strcat(field->type, " "); - strcat(field->type, field->name); - free_token(field->name); - strcat(field->type, brackets); - field->name = token; - type = read_token(&token); - } else { - field->type = realloc(field->type, - strlen(field->type) + - strlen(brackets) + 1); - strcat(field->type, brackets); - } - free(brackets); - } - - if (field_is_string(field)) { - field->flags |= FIELD_IS_STRING; - if (field_is_dynamic(field)) - field->flags |= FIELD_IS_DYNAMIC; - } - - if (test_type_token(type, token, EVENT_OP, ";")) - goto fail; - free_token(token); - - if (read_expected(EVENT_ITEM, "offset") < 0) - goto fail_expect; - - if (read_expected(EVENT_OP, ":") < 0) - goto fail_expect; - - if (read_expect_type(EVENT_ITEM, &token)) - goto fail; - field->offset = strtoul(token, NULL, 0); - free_token(token); - - if (read_expected(EVENT_OP, ";") < 0) - goto fail_expect; - - if (read_expected(EVENT_ITEM, "size") < 0) - goto fail_expect; - - if (read_expected(EVENT_OP, ":") < 0) - goto fail_expect; - - if (read_expect_type(EVENT_ITEM, &token)) - goto fail; - field->size = strtoul(token, NULL, 0); - free_token(token); - - if (read_expected(EVENT_OP, ";") < 0) - goto fail_expect; - - type = read_token(&token); - if (type != EVENT_NEWLINE) { - /* newer versions of the kernel have a "signed" type */ - if (test_type_token(type, token, EVENT_ITEM, "signed")) - goto fail; - - free_token(token); - - if (read_expected(EVENT_OP, ":") < 0) - goto fail_expect; - - if (read_expect_type(EVENT_ITEM, &token)) - goto fail; - - if (strtoul(token, NULL, 0)) - field->flags |= FIELD_IS_SIGNED; - - free_token(token); - if (read_expected(EVENT_OP, ";") < 0) - goto fail_expect; - - if (read_expect_type(EVENT_NEWLINE, &token)) - goto fail; - } - - free_token(token); - - *fields = field; - fields = &field->next; - - } while (1); - - return 0; - -fail: - free_token(token); -fail_expect: - if (field) - free(field); - return -1; + return ret; } -static int event_read_format(struct event *event) +int common_pc(struct scripting_context *context) { - char *token; + static int offset; + static int size; int ret; - if (read_expected_item(EVENT_ITEM, "format") < 0) - return -1; - - if (read_expected(EVENT_OP, ":") < 0) - return -1; - - if (read_expect_type(EVENT_NEWLINE, &token)) - goto fail; - free_token(token); - - ret = event_read_fields(event, &event->format.common_fields); + ret = get_common_field(context, &size, &offset, + "common_preempt_count"); if (ret < 0) - return ret; - event->format.nr_common = ret; - - ret = event_read_fields(event, &event->format.fields); - if (ret < 0) - return ret; - event->format.nr_fields = ret; - - return 0; - - fail: - free_token(token); - return -1; -} - -enum event_type -process_arg_token(struct event *event, struct print_arg *arg, - char **tok, enum event_type type); - -static enum event_type -process_arg(struct event *event, struct print_arg *arg, char **tok) -{ - enum event_type type; - char *token; - - type = read_token(&token); - *tok = token; + return -1; - return process_arg_token(event, arg, tok, type); + return ret; } -static enum event_type -process_cond(struct event *event, struct print_arg *top, char **tok) +unsigned long long +raw_field_value(struct event_format *event, const char *name, void *data) { - struct print_arg *arg, *left, *right; - enum event_type type; - char *token = NULL; - - arg = malloc_or_die(sizeof(*arg)); - memset(arg, 0, sizeof(*arg)); - - left = malloc_or_die(sizeof(*left)); - - right = malloc_or_die(sizeof(*right)); - - arg->type = PRINT_OP; - arg->op.left = left; - arg->op.right = right; - - *tok = NULL; - type = process_arg(event, left, &token); - if (test_type_token(type, token, EVENT_OP, ":")) - goto out_free; - - arg->op.op = token; - - type = process_arg(event, right, &token); + struct format_field *field; + unsigned long long val; - top->op.right = arg; + field = pevent_find_any_field(event, name); + if (!field) + return 0ULL; - *tok = token; - return type; + pevent_read_number_field(field, data, &val); -out_free: - free_token(*tok); - free(right); - free(left); - free_arg(arg); - return EVENT_ERROR; + return val; } -static enum event_type -process_array(struct event *event, struct print_arg *top, char **tok) +void *raw_field_ptr(struct event_format *event, const char *name, void *data) { - struct print_arg *arg; - enum event_type type; - char *token = NULL; - - arg = malloc_or_die(sizeof(*arg)); - memset(arg, 0, sizeof(*arg)); - - *tok = NULL; - type = process_arg(event, arg, &token); - if (test_type_token(type, token, EVENT_OP, "]")) - goto out_free; - - top->op.right = arg; - - free_token(token); - type = read_token_item(&token); - *tok = token; - - return type; + struct format_field *field; -out_free: - free_token(*tok); - free_arg(arg); - return EVENT_ERROR; -} + field = pevent_find_any_field(event, name); + if (!field) + return NULL; -static int get_op_prio(char *op) -{ - if (!op[1]) { - switch (op[0]) { - case '*': - case '/': - case '%': - return 6; - case '+': - case '-': - return 7; - /* '>>' and '<<' are 8 */ - case '<': - case '>': - return 9; - /* '==' and '!=' are 10 */ - case '&': - return 11; - case '^': - return 12; - case '|': - return 13; - case '?': - return 16; - default: - die("unknown op '%c'", op[0]); - return -1; - } - } else { - if (strcmp(op, "++") == 0 || - strcmp(op, "--") == 0) { - return 3; - } else if (strcmp(op, ">>") == 0 || - strcmp(op, "<<") == 0) { - return 8; - } else if (strcmp(op, ">=") == 0 || - strcmp(op, "<=") == 0) { - return 9; - } else if (strcmp(op, "==") == 0 || - strcmp(op, "!=") == 0) { - return 10; - } else if (strcmp(op, "&&") == 0) { - return 14; - } else if (strcmp(op, "||") == 0) { - return 15; - } else { - die("unknown op '%s'", op); - return -1; - } - } -} + if (field->flags & FIELD_IS_DYNAMIC) { + int offset; -static void set_op_prio(struct print_arg *arg) -{ + offset = *(int *)(data + field->offset); + offset &= 0xffff; - /* single ops are the greatest */ - if (!arg->op.left || arg->op.left->type == PRINT_NULL) { - arg->op.prio = 0; - return; + return data + offset; } - arg->op.prio = get_op_prio(arg->op.op); + return data + field->offset; } -static enum event_type -process_op(struct event *event, struct print_arg *arg, char **tok) +int trace_parse_common_type(void *data) { - struct print_arg *left, *right = NULL; - enum event_type type; - char *token; - - /* the op is passed in via tok */ - token = *tok; - - if (arg->type == PRINT_OP && !arg->op.left) { - /* handle single op */ - if (token[1]) { - die("bad op token %s", token); - return EVENT_ERROR; - } - switch (token[0]) { - case '!': - case '+': - case '-': - break; - default: - die("bad op token %s", token); - return EVENT_ERROR; - } - - /* make an empty left */ - left = malloc_or_die(sizeof(*left)); - left->type = PRINT_NULL; - arg->op.left = left; - - right = malloc_or_die(sizeof(*right)); - arg->op.right = right; - - type = process_arg(event, right, tok); - - } else if (strcmp(token, "?") == 0) { - - left = malloc_or_die(sizeof(*left)); - /* copy the top arg to the left */ - *left = *arg; - - arg->type = PRINT_OP; - arg->op.op = token; - arg->op.left = left; - arg->op.prio = 0; - - type = process_cond(event, arg, tok); - - } else if (strcmp(token, ">>") == 0 || - strcmp(token, "<<") == 0 || - strcmp(token, "&") == 0 || - strcmp(token, "|") == 0 || - strcmp(token, "&&") == 0 || - strcmp(token, "||") == 0 || - strcmp(token, "-") == 0 || - strcmp(token, "+") == 0 || - strcmp(token, "*") == 0 || - strcmp(token, "^") == 0 || - strcmp(token, "/") == 0 || - strcmp(token, "<") == 0 || - strcmp(token, ">") == 0 || - strcmp(token, "==") == 0 || - strcmp(token, "!=") == 0) { - - left = malloc_or_die(sizeof(*left)); - - /* copy the top arg to the left */ - *left = *arg; - - arg->type = PRINT_OP; - arg->op.op = token; - arg->op.left = left; - - set_op_prio(arg); - - right = malloc_or_die(sizeof(*right)); - - type = read_token_item(&token); - *tok = token; - - /* could just be a type pointer */ - if ((strcmp(arg->op.op, "*") == 0) && - type == EVENT_DELIM && (strcmp(token, ")") == 0)) { - if (left->type != PRINT_ATOM) - die("bad pointer type"); - left->atom.atom = realloc(left->atom.atom, - sizeof(left->atom.atom) + 3); - strcat(left->atom.atom, " *"); - *arg = *left; - free(arg); - - return type; - } - - type = process_arg_token(event, right, tok, type); - - arg->op.right = right; - - } else if (strcmp(token, "[") == 0) { - - left = malloc_or_die(sizeof(*left)); - *left = *arg; - - arg->type = PRINT_OP; - arg->op.op = token; - arg->op.left = left; - - arg->op.prio = 0; - type = process_array(event, arg, tok); - - } else { - warning("unknown op '%s'", token); - event->flags |= EVENT_FL_FAILED; - /* the arg is now the left side */ - return EVENT_NONE; - } - - if (type == EVENT_OP) { - int prio; - - /* higher prios need to be closer to the root */ - prio = get_op_prio(*tok); - - if (prio > arg->op.prio) - return process_op(event, arg, tok); + struct pevent_record record; - return process_op(event, right, tok); - } - - return type; + record.data = data; + return pevent_data_type(pevent, &record); } -static enum event_type -process_entry(struct event *event __unused, struct print_arg *arg, - char **tok) +int trace_parse_common_pid(void *data) { - enum event_type type; - char *field; - char *token; - - if (read_expected(EVENT_OP, "->") < 0) - return EVENT_ERROR; - - if (read_expect_type(EVENT_ITEM, &token) < 0) - goto fail; - field = token; - - arg->type = PRINT_FIELD; - arg->field.name = field; - - if (is_flag_field) { - arg->field.field = find_any_field(event, arg->field.name); - arg->field.field->flags |= FIELD_IS_FLAG; - is_flag_field = 0; - } else if (is_symbolic_field) { - arg->field.field = find_any_field(event, arg->field.name); - arg->field.field->flags |= FIELD_IS_SYMBOLIC; - is_symbolic_field = 0; - } - - type = read_token(&token); - *tok = token; + struct pevent_record record; - return type; - -fail: - free_token(token); - return EVENT_ERROR; + record.data = data; + return pevent_data_pid(pevent, &record); } -static char *arg_eval (struct print_arg *arg); - -static long long arg_num_eval(struct print_arg *arg) +unsigned long long read_size(void *ptr, int size) { - long long left, right; - long long val = 0; - - switch (arg->type) { - case PRINT_ATOM: - val = strtoll(arg->atom.atom, NULL, 0); - break; - case PRINT_TYPE: - val = arg_num_eval(arg->typecast.item); - break; - case PRINT_OP: - switch (arg->op.op[0]) { - case '|': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - if (arg->op.op[1]) - val = left || right; - else - val = left | right; - break; - case '&': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - if (arg->op.op[1]) - val = left && right; - else - val = left & right; - break; - case '<': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - switch (arg->op.op[1]) { - case 0: - val = left < right; - break; - case '<': - val = left << right; - break; - case '=': - val = left <= right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - case '>': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - switch (arg->op.op[1]) { - case 0: - val = left > right; - break; - case '>': - val = left >> right; - break; - case '=': - val = left >= right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - case '=': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - - if (arg->op.op[1] != '=') - die("unknown op '%s'", arg->op.op); - - val = left == right; - break; - case '!': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - - switch (arg->op.op[1]) { - case '=': - val = left != right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - case '+': - left = arg_num_eval(arg->op.left); - right = arg_num_eval(arg->op.right); - val = left + right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - - case PRINT_NULL: - case PRINT_FIELD ... PRINT_SYMBOL: - case PRINT_STRING: - default: - die("invalid eval type %d", arg->type); - - } - return val; + return pevent_read_number(pevent, ptr, size); } -static char *arg_eval (struct print_arg *arg) +struct event_format *trace_find_event(int type) { - long long val; - static char buf[20]; - - switch (arg->type) { - case PRINT_ATOM: - return arg->atom.atom; - case PRINT_TYPE: - return arg_eval(arg->typecast.item); - case PRINT_OP: - val = arg_num_eval(arg); - sprintf(buf, "%lld", val); - return buf; - - case PRINT_NULL: - case PRINT_FIELD ... PRINT_SYMBOL: - case PRINT_STRING: - default: - die("invalid eval type %d", arg->type); - break; - } - - return NULL; + return pevent_find_event(pevent, type); } -static enum event_type -process_fields(struct event *event, struct print_flag_sym **list, char **tok) -{ - enum event_type type; - struct print_arg *arg = NULL; - struct print_flag_sym *field; - char *token = NULL; - char *value; - - do { - free_token(token); - type = read_token_item(&token); - if (test_type_token(type, token, EVENT_OP, "{")) - break; - - arg = malloc_or_die(sizeof(*arg)); - - free_token(token); - type = process_arg(event, arg, &token); - - if (type == EVENT_OP) - type = process_op(event, arg, &token); - - if (type == EVENT_ERROR) - goto out_free; - - if (test_type_token(type, token, EVENT_DELIM, ",")) - goto out_free; - - field = malloc_or_die(sizeof(*field)); - memset(field, 0, sizeof(*field)); - - value = arg_eval(arg); - field->value = strdup(value); - - free_token(token); - type = process_arg(event, arg, &token); - if (test_type_token(type, token, EVENT_OP, "}")) - goto out_free; - - value = arg_eval(arg); - field->str = strdup(value); - free_arg(arg); - arg = NULL; - - *list = field; - list = &field->next; - free_token(token); - type = read_token_item(&token); - } while (type == EVENT_DELIM && strcmp(token, ",") == 0); - - *tok = token; - return type; - -out_free: - free_arg(arg); - free_token(token); - - return EVENT_ERROR; -} - -static enum event_type -process_flags(struct event *event, struct print_arg *arg, char **tok) +void print_trace_event(int cpu, void *data, int size) { - struct print_arg *field; - enum event_type type; - char *token; - - memset(arg, 0, sizeof(*arg)); - arg->type = PRINT_FLAGS; - - if (read_expected_item(EVENT_DELIM, "(") < 0) - return EVENT_ERROR; - - field = malloc_or_die(sizeof(*field)); - - type = process_arg(event, field, &token); - while (type == EVENT_OP) - type = process_op(event, field, &token); - if (test_type_token(type, token, EVENT_DELIM, ",")) - goto out_free; + struct event_format *event; + struct pevent_record record; + struct trace_seq s; + int type; - arg->flags.field = field; + type = trace_parse_common_type(data); - type = read_token_item(&token); - if (event_item_type(type)) { - arg->flags.delim = token; - type = read_token_item(&token); + event = trace_find_event(type); + if (!event) { + warning("ug! no event found for type %d", type); + return; } - if (test_type_token(type, token, EVENT_DELIM, ",")) - goto out_free; - - type = process_fields(event, &arg->flags.flags, &token); - if (test_type_token(type, token, EVENT_DELIM, ")")) - goto out_free; + memset(&record, 0, sizeof(record)); + record.cpu = cpu; + record.size = size; + record.data = data; - free_token(token); - type = read_token_item(tok); - return type; - -out_free: - free_token(token); - return EVENT_ERROR; + trace_seq_init(&s); + pevent_print_event(pevent, &s, &record); + trace_seq_do_printf(&s); + printf("\n"); } -static enum event_type -process_symbols(struct event *event, struct print_arg *arg, char **tok) +void print_event(int cpu, void *data, int size, unsigned long long nsecs, + char *comm) { - struct print_arg *field; - enum event_type type; - char *token; - - memset(arg, 0, sizeof(*arg)); - arg->type = PRINT_SYMBOL; - - if (read_expected_item(EVENT_DELIM, "(") < 0) - return EVENT_ERROR; - - field = malloc_or_die(sizeof(*field)); - - type = process_arg(event, field, &token); - if (test_type_token(type, token, EVENT_DELIM, ",")) - goto out_free; + struct pevent_record record; + struct trace_seq s; + int pid; - arg->symbol.field = field; + pevent->latency_format = latency_format; - type = process_fields(event, &arg->symbol.symbols, &token); - if (test_type_token(type, token, EVENT_DELIM, ")")) - goto out_free; + record.ts = nsecs; + record.cpu = cpu; + record.size = size; + record.data = data; + pid = pevent_data_pid(pevent, &record); - free_token(token); - type = read_token_item(tok); - return type; + if (!pevent_pid_is_registered(pevent, pid)) + pevent_register_comm(pevent, comm, pid); -out_free: - free_token(token); - return EVENT_ERROR; + trace_seq_init(&s); + pevent_print_event(pevent, &s, &record); + trace_seq_do_printf(&s); + printf("\n"); } -static enum event_type -process_paren(struct event *event, struct print_arg *arg, char **tok) +void parse_proc_kallsyms(char *file, unsigned int size __unused) { - struct print_arg *item_arg; - enum event_type type; - char *token; - - type = process_arg(event, arg, &token); - - if (type == EVENT_ERROR) - return EVENT_ERROR; - - if (type == EVENT_OP) - type = process_op(event, arg, &token); - - if (type == EVENT_ERROR) - return EVENT_ERROR; - - if (test_type_token(type, token, EVENT_DELIM, ")")) { - free_token(token); - return EVENT_ERROR; - } - - free_token(token); - type = read_token_item(&token); - - /* - * If the next token is an item or another open paren, then - * this was a typecast. - */ - if (event_item_type(type) || - (type == EVENT_DELIM && strcmp(token, "(") == 0)) { - - /* make this a typecast and contine */ + unsigned long long addr; + char *func; + char *line; + char *next = NULL; + char *addr_str; + char *mod; + char ch; - /* prevous must be an atom */ - if (arg->type != PRINT_ATOM) - die("previous needed to be PRINT_ATOM"); + line = strtok_r(file, "\n", &next); + while (line) { + mod = NULL; + sscanf(line, "%as %c %as\t[%as", + (float *)(void *)&addr_str, /* workaround gcc warning */ + &ch, (float *)(void *)&func, (float *)(void *)&mod); + addr = strtoull(addr_str, NULL, 16); + free(addr_str); - item_arg = malloc_or_die(sizeof(*item_arg)); + /* truncate the extra ']' */ + if (mod) + mod[strlen(mod) - 1] = 0; - arg->type = PRINT_TYPE; - arg->typecast.type = arg->atom.atom; - arg->typecast.item = item_arg; - type = process_arg_token(event, item_arg, &token, type); + pevent_register_function(pevent, func, addr, mod); + free(func); + free(mod); + line = strtok_r(NULL, "\n", &next); } - - *tok = token; - return type; } - -static enum event_type -process_str(struct event *event __unused, struct print_arg *arg, char **tok) +void parse_ftrace_printk(char *file, unsigned int size __unused) { - enum event_type type; - char *token; - - if (read_expected(EVENT_DELIM, "(") < 0) - return EVENT_ERROR; - - if (read_expect_type(EVENT_ITEM, &token) < 0) - goto fail; - - arg->type = PRINT_STRING; - arg->string.string = token; - arg->string.offset = -1; - - if (read_expected(EVENT_DELIM, ")") < 0) - return EVENT_ERROR; - - type = read_token(&token); - *tok = token; - - return type; -fail: - free_token(token); - return EVENT_ERROR; -} + unsigned long long addr; + char *printk; + char *line; + char *next = NULL; + char *addr_str; + char *fmt; -enum event_type -process_arg_token(struct event *event, struct print_arg *arg, - char **tok, enum event_type type) -{ - char *token; - char *atom; - - token = *tok; - - switch (type) { - case EVENT_ITEM: - if (strcmp(token, "REC") == 0) { - free_token(token); - type = process_entry(event, arg, &token); - } else if (strcmp(token, "__print_flags") == 0) { - free_token(token); - is_flag_field = 1; - type = process_flags(event, arg, &token); - } else if (strcmp(token, "__print_symbolic") == 0) { - free_token(token); - is_symbolic_field = 1; - type = process_symbols(event, arg, &token); - } else if (strcmp(token, "__get_str") == 0) { - free_token(token); - type = process_str(event, arg, &token); - } else { - atom = token; - /* test the next token */ - type = read_token_item(&token); - - /* atoms can be more than one token long */ - while (type == EVENT_ITEM) { - atom = realloc(atom, strlen(atom) + strlen(token) + 2); - strcat(atom, " "); - strcat(atom, token); - free_token(token); - type = read_token_item(&token); - } - - /* todo, test for function */ - - arg->type = PRINT_ATOM; - arg->atom.atom = atom; - } - break; - case EVENT_DQUOTE: - case EVENT_SQUOTE: - arg->type = PRINT_ATOM; - arg->atom.atom = token; - type = read_token_item(&token); - break; - case EVENT_DELIM: - if (strcmp(token, "(") == 0) { - free_token(token); - type = process_paren(event, arg, &token); + line = strtok_r(file, "\n", &next); + while (line) { + addr_str = strtok_r(line, ":", &fmt); + if (!addr_str) { + warning("printk format with empty entry"); break; } - case EVENT_OP: - /* handle single ops */ - arg->type = PRINT_OP; - arg->op.op = token; - arg->op.left = NULL; - type = process_op(event, arg, &token); - - break; - - case EVENT_ERROR ... EVENT_NEWLINE: - default: - die("unexpected type %d", type); - } - *tok = token; - - return type; -} - -static int event_read_print_args(struct event *event, struct print_arg **list) -{ - enum event_type type = EVENT_ERROR; - struct print_arg *arg; - char *token; - int args = 0; - - do { - if (type == EVENT_NEWLINE) { - free_token(token); - type = read_token_item(&token); - continue; - } - - arg = malloc_or_die(sizeof(*arg)); - memset(arg, 0, sizeof(*arg)); - - type = process_arg(event, arg, &token); - - if (type == EVENT_ERROR) { - free_arg(arg); - return -1; - } - - *list = arg; - args++; - - if (type == EVENT_OP) { - type = process_op(event, arg, &token); - list = &arg->next; - continue; - } - - if (type == EVENT_DELIM && strcmp(token, ",") == 0) { - free_token(token); - *list = arg; - list = &arg->next; - continue; - } - break; - } while (type != EVENT_NONE); - - if (type != EVENT_NONE) - free_token(token); - - return args; -} - -static int event_read_print(struct event *event) -{ - enum event_type type; - char *token; - int ret; - - if (read_expected_item(EVENT_ITEM, "print") < 0) - return -1; - - if (read_expected(EVENT_ITEM, "fmt") < 0) - return -1; - - if (read_expected(EVENT_OP, ":") < 0) - return -1; - - if (read_expect_type(EVENT_DQUOTE, &token) < 0) - goto fail; - - concat: - event->print_fmt.format = token; - event->print_fmt.args = NULL; - - /* ok to have no arg */ - type = read_token_item(&token); - - if (type == EVENT_NONE) - return 0; - - /* Handle concatination of print lines */ - if (type == EVENT_DQUOTE) { - char *cat; - - cat = malloc_or_die(strlen(event->print_fmt.format) + - strlen(token) + 1); - strcpy(cat, event->print_fmt.format); - strcat(cat, token); - free_token(token); - free_token(event->print_fmt.format); - event->print_fmt.format = NULL; - token = cat; - goto concat; - } - - if (test_type_token(type, token, EVENT_DELIM, ",")) - goto fail; - - free_token(token); - - ret = event_read_print_args(event, &event->print_fmt.args); - if (ret < 0) - return -1; - - return ret; - - fail: - free_token(token); - return -1; -} - -static struct format_field * -find_common_field(struct event *event, const char *name) -{ - struct format_field *format; - - for (format = event->format.common_fields; - format; format = format->next) { - if (strcmp(format->name, name) == 0) - break; - } - - return format; -} - -static struct format_field * -find_field(struct event *event, const char *name) -{ - struct format_field *format; - - for (format = event->format.fields; - format; format = format->next) { - if (strcmp(format->name, name) == 0) - break; + addr = strtoull(addr_str, NULL, 16); + /* fmt still has a space, skip it */ + printk = strdup(fmt+1); + line = strtok_r(NULL, "\n", &next); + pevent_register_print_string(pevent, printk, addr); } - - return format; } -static struct format_field * -find_any_field(struct event *event, const char *name) +int parse_ftrace_file(char *buf, unsigned long size) { - struct format_field *format; - - format = find_common_field(event, name); - if (format) - return format; - return find_field(event, name); + return pevent_parse_event(pevent, buf, size, "ftrace"); } -unsigned long long read_size(void *ptr, int size) +int parse_event_file(char *buf, unsigned long size, char *sys) { - switch (size) { - case 1: - return *(unsigned char *)ptr; - case 2: - return data2host2(ptr); - case 4: - return data2host4(ptr); - case 8: - return data2host8(ptr); - default: - /* BUG! */ - return 0; - } + return pevent_parse_event(pevent, buf, size, sys); } -unsigned long long -raw_field_value(struct event *event, const char *name, void *data) +struct event_format *trace_find_next_event(struct event_format *event) { - struct format_field *field; - - field = find_any_field(event, name); - if (!field) - return 0ULL; + static int idx; - return read_size(data + field->offset, field->size); -} - -void *raw_field_ptr(struct event *event, const char *name, void *data) -{ - struct format_field *field; - - field = find_any_field(event, name); - if (!field) + if (!pevent->events) return NULL; - if (field->flags & FIELD_IS_DYNAMIC) { - int offset; - - offset = *(int *)(data + field->offset); - offset &= 0xffff; - - return data + offset; - } - - return data + field->offset; -} - -static int get_common_info(const char *type, int *offset, int *size) -{ - struct event *event; - struct format_field *field; - - /* - * All events should have the same common elements. - * Pick any event to find where the type is; - */ - if (!event_list) - die("no event_list!"); - - event = event_list; - field = find_common_field(event, type); - if (!field) - die("field '%s' not found", type); - - *offset = field->offset; - *size = field->size; - - return 0; -} - -static int __parse_common(void *data, int *size, int *offset, - const char *name) -{ - int ret; - - if (!*size) { - ret = get_common_info(name, offset, size); - if (ret < 0) - return ret; + if (!event) { + idx = 0; + return pevent->events[0]; } - return read_size(data + *offset, *size); -} - -int trace_parse_common_type(void *data) -{ - static int type_offset; - static int type_size; - - return __parse_common(data, &type_size, &type_offset, - "common_type"); -} - -int trace_parse_common_pid(void *data) -{ - static int pid_offset; - static int pid_size; - - return __parse_common(data, &pid_size, &pid_offset, - "common_pid"); -} - -int parse_common_pc(void *data) -{ - static int pc_offset; - static int pc_size; - - return __parse_common(data, &pc_size, &pc_offset, - "common_preempt_count"); -} - -int parse_common_flags(void *data) -{ - static int flags_offset; - static int flags_size; - - return __parse_common(data, &flags_size, &flags_offset, - "common_flags"); -} - -int parse_common_lock_depth(void *data) -{ - static int ld_offset; - static int ld_size; - int ret; - ret = __parse_common(data, &ld_size, &ld_offset, - "common_lock_depth"); - if (ret < 0) - return -1; - - return ret; -} - -struct event *trace_find_event(int id) -{ - struct event *event; - - for (event = event_list; event; event = event->next) { - if (event->id == id) - break; + if (idx < pevent->nr_events && event == pevent->events[idx]) { + idx++; + if (idx == pevent->nr_events) + return NULL; + return pevent->events[idx]; } - return event; -} - -struct event *trace_find_next_event(struct event *event) -{ - if (!event) - return event_list; - - return event->next; -} - -static unsigned long long eval_num_arg(void *data, int size, - struct event *event, struct print_arg *arg) -{ - unsigned long long val = 0; - unsigned long long left, right; - struct print_arg *larg; - switch (arg->type) { - case PRINT_NULL: - /* ?? */ - return 0; - case PRINT_ATOM: - return strtoull(arg->atom.atom, NULL, 0); - case PRINT_FIELD: - if (!arg->field.field) { - arg->field.field = find_any_field(event, arg->field.name); - if (!arg->field.field) - die("field %s not found", arg->field.name); - } - /* must be a number */ - val = read_size(data + arg->field.field->offset, - arg->field.field->size); - break; - case PRINT_FLAGS: - case PRINT_SYMBOL: - break; - case PRINT_TYPE: - return eval_num_arg(data, size, event, arg->typecast.item); - case PRINT_STRING: - return 0; - break; - case PRINT_OP: - if (strcmp(arg->op.op, "[") == 0) { - /* - * Arrays are special, since we don't want - * to read the arg as is. - */ - if (arg->op.left->type != PRINT_FIELD) - goto default_op; /* oops, all bets off */ - larg = arg->op.left; - if (!larg->field.field) { - larg->field.field = - find_any_field(event, larg->field.name); - if (!larg->field.field) - die("field %s not found", larg->field.name); - } - right = eval_num_arg(data, size, event, arg->op.right); - val = read_size(data + larg->field.field->offset + - right * long_size, long_size); - break; - } - default_op: - left = eval_num_arg(data, size, event, arg->op.left); - right = eval_num_arg(data, size, event, arg->op.right); - switch (arg->op.op[0]) { - case '|': - if (arg->op.op[1]) - val = left || right; - else - val = left | right; - break; - case '&': - if (arg->op.op[1]) - val = left && right; - else - val = left & right; - break; - case '<': - switch (arg->op.op[1]) { - case 0: - val = left < right; - break; - case '<': - val = left << right; - break; - case '=': - val = left <= right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - case '>': - switch (arg->op.op[1]) { - case 0: - val = left > right; - break; - case '>': - val = left >> right; - break; - case '=': - val = left >= right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - case '=': - if (arg->op.op[1] != '=') - die("unknown op '%s'", arg->op.op); - val = left == right; - break; - case '-': - val = left - right; - break; - case '+': - val = left + right; - break; - default: - die("unknown op '%s'", arg->op.op); - } - break; - default: /* not sure what to do there */ - return 0; + for (idx = 1; idx < pevent->nr_events; idx++) { + if (event == pevent->events[idx - 1]) + return pevent->events[idx]; } - return val; + return NULL; } struct flag { @@ -2223,933 +357,3 @@ unsigned long long eval_flag(const char *flag) return 0; } - -static void print_str_arg(void *data, int size, - struct event *event, struct print_arg *arg) -{ - struct print_flag_sym *flag; - unsigned long long val, fval; - char *str; - int print; - - switch (arg->type) { - case PRINT_NULL: - /* ?? */ - return; - case PRINT_ATOM: - printf("%s", arg->atom.atom); - return; - case PRINT_FIELD: - if (!arg->field.field) { - arg->field.field = find_any_field(event, arg->field.name); - if (!arg->field.field) - die("field %s not found", arg->field.name); - } - str = malloc_or_die(arg->field.field->size + 1); - memcpy(str, data + arg->field.field->offset, - arg->field.field->size); - str[arg->field.field->size] = 0; - printf("%s", str); - free(str); - break; - case PRINT_FLAGS: - val = eval_num_arg(data, size, event, arg->flags.field); - print = 0; - for (flag = arg->flags.flags; flag; flag = flag->next) { - fval = eval_flag(flag->value); - if (!val && !fval) { - printf("%s", flag->str); - break; - } - if (fval && (val & fval) == fval) { - if (print && arg->flags.delim) - printf("%s", arg->flags.delim); - printf("%s", flag->str); - print = 1; - val &= ~fval; - } - } - break; - case PRINT_SYMBOL: - val = eval_num_arg(data, size, event, arg->symbol.field); - for (flag = arg->symbol.symbols; flag; flag = flag->next) { - fval = eval_flag(flag->value); - if (val == fval) { - printf("%s", flag->str); - break; - } - } - break; - - case PRINT_TYPE: - break; - case PRINT_STRING: { - int str_offset; - - if (arg->string.offset == -1) { - struct format_field *f; - - f = find_any_field(event, arg->string.string); - arg->string.offset = f->offset; - } - str_offset = *(int *)(data + arg->string.offset); - str_offset &= 0xffff; - printf("%s", ((char *)data) + str_offset); - break; - } - case PRINT_OP: - /* - * The only op for string should be ? : - */ - if (arg->op.op[0] != '?') - return; - val = eval_num_arg(data, size, event, arg->op.left); - if (val) - print_str_arg(data, size, event, arg->op.right->op.left); - else - print_str_arg(data, size, event, arg->op.right->op.right); - break; - default: - /* well... */ - break; - } -} - -static struct print_arg *make_bprint_args(char *fmt, void *data, int size, struct event *event) -{ - static struct format_field *field, *ip_field; - struct print_arg *args, *arg, **next; - unsigned long long ip, val; - char *ptr; - void *bptr; - - if (!field) { - field = find_field(event, "buf"); - if (!field) - die("can't find buffer field for binary printk"); - ip_field = find_field(event, "ip"); - if (!ip_field) - die("can't find ip field for binary printk"); - } - - ip = read_size(data + ip_field->offset, ip_field->size); - - /* - * The first arg is the IP pointer. - */ - args = malloc_or_die(sizeof(*args)); - arg = args; - arg->next = NULL; - next = &arg->next; - - arg->type = PRINT_ATOM; - arg->atom.atom = malloc_or_die(32); - sprintf(arg->atom.atom, "%lld", ip); - - /* skip the first "%pf : " */ - for (ptr = fmt + 6, bptr = data + field->offset; - bptr < data + size && *ptr; ptr++) { - int ls = 0; - - if (*ptr == '%') { - process_again: - ptr++; - switch (*ptr) { - case '%': - break; - case 'l': - ls++; - goto process_again; - case 'L': - ls = 2; - goto process_again; - case '0' ... '9': - goto process_again; - case 'p': - ls = 1; - /* fall through */ - case 'd': - case 'u': - case 'x': - case 'i': - /* the pointers are always 4 bytes aligned */ - bptr = (void *)(((unsigned long)bptr + 3) & - ~3); - switch (ls) { - case 0: - case 1: - ls = long_size; - break; - case 2: - ls = 8; - default: - break; - } - val = read_size(bptr, ls); - bptr += ls; - arg = malloc_or_die(sizeof(*arg)); - arg->next = NULL; - arg->type = PRINT_ATOM; - arg->atom.atom = malloc_or_die(32); - sprintf(arg->atom.atom, "%lld", val); - *next = arg; - next = &arg->next; - break; - case 's': - arg = malloc_or_die(sizeof(*arg)); - arg->next = NULL; - arg->type = PRINT_STRING; - arg->string.string = strdup(bptr); - bptr += strlen(bptr) + 1; - *next = arg; - next = &arg->next; - default: - break; - } - } - } - - return args; -} - -static void free_args(struct print_arg *args) -{ - struct print_arg *next; - - while (args) { - next = args->next; - - if (args->type == PRINT_ATOM) - free(args->atom.atom); - else - free(args->string.string); - free(args); - args = next; - } -} - -static char *get_bprint_format(void *data, int size __unused, struct event *event) -{ - unsigned long long addr; - static struct format_field *field; - struct printk_map *printk; - char *format; - char *p; - - if (!field) { - field = find_field(event, "fmt"); - if (!field) - die("can't find format field for binary printk"); - printf("field->offset = %d size=%d\n", field->offset, field->size); - } - - addr = read_size(data + field->offset, field->size); - - printk = find_printk(addr); - if (!printk) { - format = malloc_or_die(45); - sprintf(format, "%%pf : (NO FORMAT FOUND at %llx)\n", - addr); - return format; - } - - p = printk->printk; - /* Remove any quotes. */ - if (*p == '"') - p++; - format = malloc_or_die(strlen(p) + 10); - sprintf(format, "%s : %s", "%pf", p); - /* remove ending quotes and new line since we will add one too */ - p = format + strlen(format) - 1; - if (*p == '"') - *p = 0; - - p -= 2; - if (strcmp(p, "\\n") == 0) - *p = 0; - - return format; -} - -static void pretty_print(void *data, int size, struct event *event) -{ - struct print_fmt *print_fmt = &event->print_fmt; - struct print_arg *arg = print_fmt->args; - struct print_arg *args = NULL; - const char *ptr = print_fmt->format; - unsigned long long val; - struct func_map *func; - const char *saveptr; - char *bprint_fmt = NULL; - char format[32]; - int show_func; - int len; - int ls; - - if (event->flags & EVENT_FL_ISFUNC) - ptr = " %pF <-- %pF"; - - if (event->flags & EVENT_FL_ISBPRINT) { - bprint_fmt = get_bprint_format(data, size, event); - args = make_bprint_args(bprint_fmt, data, size, event); - arg = args; - ptr = bprint_fmt; - } - - for (; *ptr; ptr++) { - ls = 0; - if (*ptr == '\\') { - ptr++; - switch (*ptr) { - case 'n': - printf("\n"); - break; - case 't': - printf("\t"); - break; - case 'r': - printf("\r"); - break; - case '\\': - printf("\\"); - break; - default: - printf("%c", *ptr); - break; - } - - } else if (*ptr == '%') { - saveptr = ptr; - show_func = 0; - cont_process: - ptr++; - switch (*ptr) { - case '%': - printf("%%"); - break; - case 'l': - ls++; - goto cont_process; - case 'L': - ls = 2; - goto cont_process; - case 'z': - case 'Z': - case '0' ... '9': - goto cont_process; - case 'p': - if (long_size == 4) - ls = 1; - else - ls = 2; - - if (*(ptr+1) == 'F' || - *(ptr+1) == 'f') { - ptr++; - show_func = *ptr; - } - - /* fall through */ - case 'd': - case 'i': - case 'x': - case 'X': - case 'u': - if (!arg) - die("no argument match"); - - len = ((unsigned long)ptr + 1) - - (unsigned long)saveptr; - - /* should never happen */ - if (len > 32) - die("bad format!"); - - memcpy(format, saveptr, len); - format[len] = 0; - - val = eval_num_arg(data, size, event, arg); - arg = arg->next; - - if (show_func) { - func = find_func(val); - if (func) { - printf("%s", func->func); - if (show_func == 'F') - printf("+0x%llx", - val - func->addr); - break; - } - } - switch (ls) { - case 0: - printf(format, (int)val); - break; - case 1: - printf(format, (long)val); - break; - case 2: - printf(format, (long long)val); - break; - default: - die("bad count (%d)", ls); - } - break; - case 's': - if (!arg) - die("no matching argument"); - - print_str_arg(data, size, event, arg); - arg = arg->next; - break; - default: - printf(">%c<", *ptr); - - } - } else - printf("%c", *ptr); - } - - if (args) { - free_args(args); - free(bprint_fmt); - } -} - -static inline int log10_cpu(int nb) -{ - if (nb / 100) - return 3; - if (nb / 10) - return 2; - return 1; -} - -static void print_lat_fmt(void *data, int size __unused) -{ - unsigned int lat_flags; - unsigned int pc; - int lock_depth; - int hardirq; - int softirq; - - lat_flags = parse_common_flags(data); - pc = parse_common_pc(data); - lock_depth = parse_common_lock_depth(data); - - hardirq = lat_flags & TRACE_FLAG_HARDIRQ; - softirq = lat_flags & TRACE_FLAG_SOFTIRQ; - - printf("%c%c%c", - (lat_flags & TRACE_FLAG_IRQS_OFF) ? 'd' : - (lat_flags & TRACE_FLAG_IRQS_NOSUPPORT) ? - 'X' : '.', - (lat_flags & TRACE_FLAG_NEED_RESCHED) ? - 'N' : '.', - (hardirq && softirq) ? 'H' : - hardirq ? 'h' : softirq ? 's' : '.'); - - if (pc) - printf("%x", pc); - else - printf("."); - - if (lock_depth < 0) - printf(". "); - else - printf("%d ", lock_depth); -} - -#define TRACE_GRAPH_INDENT 2 - -static struct record * -get_return_for_leaf(int cpu, int cur_pid, unsigned long long cur_func, - struct record *next) -{ - struct format_field *field; - struct event *event; - unsigned long val; - int type; - int pid; - - type = trace_parse_common_type(next->data); - event = trace_find_event(type); - if (!event) - return NULL; - - if (!(event->flags & EVENT_FL_ISFUNCRET)) - return NULL; - - pid = trace_parse_common_pid(next->data); - field = find_field(event, "func"); - if (!field) - die("function return does not have field func"); - - val = read_size(next->data + field->offset, field->size); - - if (cur_pid != pid || cur_func != val) - return NULL; - - /* this is a leaf, now advance the iterator */ - return trace_read_data(cpu); -} - -/* Signal a overhead of time execution to the output */ -static void print_graph_overhead(unsigned long long duration) -{ - /* Non nested entry or return */ - if (duration == ~0ULL) - return (void)printf(" "); - - /* Duration exceeded 100 msecs */ - if (duration > 100000ULL) - return (void)printf("! "); - - /* Duration exceeded 10 msecs */ - if (duration > 10000ULL) - return (void)printf("+ "); - - printf(" "); -} - -static void print_graph_duration(unsigned long long duration) -{ - unsigned long usecs = duration / 1000; - unsigned long nsecs_rem = duration % 1000; - /* log10(ULONG_MAX) + '\0' */ - char msecs_str[21]; - char nsecs_str[5]; - int len; - int i; - - sprintf(msecs_str, "%lu", usecs); - - /* Print msecs */ - len = printf("%lu", usecs); - - /* Print nsecs (we don't want to exceed 7 numbers) */ - if (len < 7) { - snprintf(nsecs_str, 8 - len, "%03lu", nsecs_rem); - len += printf(".%s", nsecs_str); - } - - printf(" us "); - - /* Print remaining spaces to fit the row's width */ - for (i = len; i < 7; i++) - printf(" "); - - printf("| "); -} - -static void -print_graph_entry_leaf(struct event *event, void *data, struct record *ret_rec) -{ - unsigned long long rettime, calltime; - unsigned long long duration, depth; - unsigned long long val; - struct format_field *field; - struct func_map *func; - struct event *ret_event; - int type; - int i; - - type = trace_parse_common_type(ret_rec->data); - ret_event = trace_find_event(type); - - field = find_field(ret_event, "rettime"); - if (!field) - die("can't find rettime in return graph"); - rettime = read_size(ret_rec->data + field->offset, field->size); - - field = find_field(ret_event, "calltime"); - if (!field) - die("can't find rettime in return graph"); - calltime = read_size(ret_rec->data + field->offset, field->size); - - duration = rettime - calltime; - - /* Overhead */ - print_graph_overhead(duration); - - /* Duration */ - print_graph_duration(duration); - - field = find_field(event, "depth"); - if (!field) - die("can't find depth in entry graph"); - depth = read_size(data + field->offset, field->size); - - /* Function */ - for (i = 0; i < (int)(depth * TRACE_GRAPH_INDENT); i++) - printf(" "); - - field = find_field(event, "func"); - if (!field) - die("can't find func in entry graph"); - val = read_size(data + field->offset, field->size); - func = find_func(val); - - if (func) - printf("%s();", func->func); - else - printf("%llx();", val); -} - -static void print_graph_nested(struct event *event, void *data) -{ - struct format_field *field; - unsigned long long depth; - unsigned long long val; - struct func_map *func; - int i; - - /* No overhead */ - print_graph_overhead(-1); - - /* No time */ - printf(" | "); - - field = find_field(event, "depth"); - if (!field) - die("can't find depth in entry graph"); - depth = read_size(data + field->offset, field->size); - - /* Function */ - for (i = 0; i < (int)(depth * TRACE_GRAPH_INDENT); i++) - printf(" "); - - field = find_field(event, "func"); - if (!field) - die("can't find func in entry graph"); - val = read_size(data + field->offset, field->size); - func = find_func(val); - - if (func) - printf("%s() {", func->func); - else - printf("%llx() {", val); -} - -static void -pretty_print_func_ent(void *data, int size, struct event *event, - int cpu, int pid) -{ - struct format_field *field; - struct record *rec; - void *copy_data; - unsigned long val; - - if (latency_format) { - print_lat_fmt(data, size); - printf(" | "); - } - - field = find_field(event, "func"); - if (!field) - die("function entry does not have func field"); - - val = read_size(data + field->offset, field->size); - - /* - * peek_data may unmap the data pointer. Copy it first. - */ - copy_data = malloc_or_die(size); - memcpy(copy_data, data, size); - data = copy_data; - - rec = trace_peek_data(cpu); - if (rec) { - rec = get_return_for_leaf(cpu, pid, val, rec); - if (rec) { - print_graph_entry_leaf(event, data, rec); - goto out_free; - } - } - print_graph_nested(event, data); -out_free: - free(data); -} - -static void -pretty_print_func_ret(void *data, int size __unused, struct event *event) -{ - unsigned long long rettime, calltime; - unsigned long long duration, depth; - struct format_field *field; - int i; - - if (latency_format) { - print_lat_fmt(data, size); - printf(" | "); - } - - field = find_field(event, "rettime"); - if (!field) - die("can't find rettime in return graph"); - rettime = read_size(data + field->offset, field->size); - - field = find_field(event, "calltime"); - if (!field) - die("can't find calltime in return graph"); - calltime = read_size(data + field->offset, field->size); - - duration = rettime - calltime; - - /* Overhead */ - print_graph_overhead(duration); - - /* Duration */ - print_graph_duration(duration); - - field = find_field(event, "depth"); - if (!field) - die("can't find depth in entry graph"); - depth = read_size(data + field->offset, field->size); - - /* Function */ - for (i = 0; i < (int)(depth * TRACE_GRAPH_INDENT); i++) - printf(" "); - - printf("}"); -} - -static void -pretty_print_func_graph(void *data, int size, struct event *event, - int cpu, int pid) -{ - if (event->flags & EVENT_FL_ISFUNCENT) - pretty_print_func_ent(data, size, event, cpu, pid); - else if (event->flags & EVENT_FL_ISFUNCRET) - pretty_print_func_ret(data, size, event); - printf("\n"); -} - -void print_trace_event(int cpu, void *data, int size) -{ - struct event *event; - int type; - int pid; - - type = trace_parse_common_type(data); - - event = trace_find_event(type); - if (!event) { - warning("ug! no event found for type %d", type); - return; - } - - pid = trace_parse_common_pid(data); - - if (event->flags & (EVENT_FL_ISFUNCENT | EVENT_FL_ISFUNCRET)) - return pretty_print_func_graph(data, size, event, cpu, pid); - - if (latency_format) - print_lat_fmt(data, size); - - if (event->flags & EVENT_FL_FAILED) { - printf("EVENT '%s' FAILED TO PARSE\n", - event->name); - return; - } - - pretty_print(data, size, event); -} - -static void print_fields(struct print_flag_sym *field) -{ - printf("{ %s, %s }", field->value, field->str); - if (field->next) { - printf(", "); - print_fields(field->next); - } -} - -static void print_args(struct print_arg *args) -{ - int print_paren = 1; - - switch (args->type) { - case PRINT_NULL: - printf("null"); - break; - case PRINT_ATOM: - printf("%s", args->atom.atom); - break; - case PRINT_FIELD: - printf("REC->%s", args->field.name); - break; - case PRINT_FLAGS: - printf("__print_flags("); - print_args(args->flags.field); - printf(", %s, ", args->flags.delim); - print_fields(args->flags.flags); - printf(")"); - break; - case PRINT_SYMBOL: - printf("__print_symbolic("); - print_args(args->symbol.field); - printf(", "); - print_fields(args->symbol.symbols); - printf(")"); - break; - case PRINT_STRING: - printf("__get_str(%s)", args->string.string); - break; - case PRINT_TYPE: - printf("(%s)", args->typecast.type); - print_args(args->typecast.item); - break; - case PRINT_OP: - if (strcmp(args->op.op, ":") == 0) - print_paren = 0; - if (print_paren) - printf("("); - print_args(args->op.left); - printf(" %s ", args->op.op); - print_args(args->op.right); - if (print_paren) - printf(")"); - break; - default: - /* we should warn... */ - return; - } - if (args->next) { - printf("\n"); - print_args(args->next); - } -} - -int parse_ftrace_file(char *buf, unsigned long size) -{ - struct format_field *field; - struct print_arg *arg, **list; - struct event *event; - int ret; - - init_input_buf(buf, size); - - event = alloc_event(); - if (!event) - return -ENOMEM; - - event->flags |= EVENT_FL_ISFTRACE; - - event->name = event_read_name(); - if (!event->name) - die("failed to read ftrace event name"); - - if (strcmp(event->name, "function") == 0) - event->flags |= EVENT_FL_ISFUNC; - - else if (strcmp(event->name, "funcgraph_entry") == 0) - event->flags |= EVENT_FL_ISFUNCENT; - - else if (strcmp(event->name, "funcgraph_exit") == 0) - event->flags |= EVENT_FL_ISFUNCRET; - - else if (strcmp(event->name, "bprint") == 0) - event->flags |= EVENT_FL_ISBPRINT; - - event->id = event_read_id(); - if (event->id < 0) - die("failed to read ftrace event id"); - - add_event(event); - - ret = event_read_format(event); - if (ret < 0) - die("failed to read ftrace event format"); - - ret = event_read_print(event); - if (ret < 0) - die("failed to read ftrace event print fmt"); - - /* New ftrace handles args */ - if (ret > 0) - return 0; - /* - * The arguments for ftrace files are parsed by the fields. - * Set up the fields as their arguments. - */ - list = &event->print_fmt.args; - for (field = event->format.fields; field; field = field->next) { - arg = malloc_or_die(sizeof(*arg)); - memset(arg, 0, sizeof(*arg)); - *list = arg; - list = &arg->next; - arg->type = PRINT_FIELD; - arg->field.name = field->name; - arg->field.field = field; - } - return 0; -} - -int parse_event_file(char *buf, unsigned long size, char *sys) -{ - struct event *event; - int ret; - - init_input_buf(buf, size); - - event = alloc_event(); - if (!event) - return -ENOMEM; - - event->name = event_read_name(); - if (!event->name) - die("failed to read event name"); - - event->id = event_read_id(); - if (event->id < 0) - die("failed to read event id"); - - ret = event_read_format(event); - if (ret < 0) { - warning("failed to read event format for %s", event->name); - goto event_failed; - } - - ret = event_read_print(event); - if (ret < 0) { - warning("failed to read event print fmt for %s", event->name); - goto event_failed; - } - - event->system = strdup(sys); - -#define PRINT_ARGS 0 - if (PRINT_ARGS && event->print_fmt.args) - print_args(event->print_fmt.args); - - add_event(event); - return 0; - - event_failed: - event->flags |= EVENT_FL_FAILED; - /* still add it even if it failed */ - add_event(event); - return -1; -} - -void parse_set_info(int nr_cpus, int long_sz) -{ - cpus = nr_cpus; - long_size = long_sz; -} - -int common_pc(struct scripting_context *context) -{ - return parse_common_pc(context->event_data); -} - -int common_flags(struct scripting_context *context) -{ - return parse_common_flags(context->event_data); -} - -int common_lock_depth(struct scripting_context *context) -{ - return parse_common_lock_depth(context->event_data); -} diff --git a/tools/perf/util/trace-event-read.c b/tools/perf/util/trace-event-read.c index b9592e0de8d7..f097e0dd6c5c 100644 --- a/tools/perf/util/trace-event-read.c +++ b/tools/perf/util/trace-event-read.c @@ -52,6 +52,16 @@ static unsigned long page_size; static ssize_t calc_data_size; static bool repipe; +static void *malloc_or_die(int size) +{ + void *ret; + + ret = malloc(size); + if (!ret) + die("malloc"); + return ret; +} + static int do_read(int fd, void *buf, int size) { int rsize = size; @@ -109,7 +119,7 @@ static unsigned int read4(void) unsigned int data; read_or_die(&data, 4); - return __data2host4(data); + return __data2host4(perf_pevent, data); } static unsigned long long read8(void) @@ -117,7 +127,7 @@ static unsigned long long read8(void) unsigned long long data; read_or_die(&data, 8); - return __data2host8(data); + return __data2host8(perf_pevent, data); } static char *read_string(void) @@ -282,7 +292,7 @@ struct cpu_data { unsigned long long offset; unsigned long long size; unsigned long long timestamp; - struct record *next; + struct pevent_record *next; char *page; int cpu; int index; @@ -367,9 +377,9 @@ static int calc_index(void *ptr, int cpu) return (unsigned long)ptr - (unsigned long)cpu_data[cpu].page; } -struct record *trace_peek_data(int cpu) +struct pevent_record *trace_peek_data(int cpu) { - struct record *data; + struct pevent_record *data; void *page = cpu_data[cpu].page; int idx = cpu_data[cpu].index; void *ptr = page + idx; @@ -389,15 +399,15 @@ struct record *trace_peek_data(int cpu) /* FIXME: handle header page */ if (header_page_ts_size != 8) die("expected a long long type for timestamp"); - cpu_data[cpu].timestamp = data2host8(ptr); + cpu_data[cpu].timestamp = data2host8(perf_pevent, ptr); ptr += 8; switch (header_page_size_size) { case 4: - cpu_data[cpu].page_size = data2host4(ptr); + cpu_data[cpu].page_size = data2host4(perf_pevent, ptr); ptr += 4; break; case 8: - cpu_data[cpu].page_size = data2host8(ptr); + cpu_data[cpu].page_size = data2host8(perf_pevent, ptr); ptr += 8; break; default: @@ -414,7 +424,7 @@ read_again: return trace_peek_data(cpu); } - type_len_ts = data2host4(ptr); + type_len_ts = data2host4(perf_pevent, ptr); ptr += 4; type_len = type_len4host(type_len_ts); @@ -424,14 +434,14 @@ read_again: case RINGBUF_TYPE_PADDING: if (!delta) die("error, hit unexpected end of page"); - length = data2host4(ptr); + length = data2host4(perf_pevent, ptr); ptr += 4; length *= 4; ptr += length; goto read_again; case RINGBUF_TYPE_TIME_EXTEND: - extend = data2host4(ptr); + extend = data2host4(perf_pevent, ptr); ptr += 4; extend <<= TS_SHIFT; extend += delta; @@ -442,7 +452,7 @@ read_again: ptr += 12; break; case 0: - length = data2host4(ptr); + length = data2host4(perf_pevent, ptr); ptr += 4; die("here! length=%d", length); break; @@ -467,9 +477,9 @@ read_again: return data; } -struct record *trace_read_data(int cpu) +struct pevent_record *trace_read_data(int cpu) { - struct record *data; + struct pevent_record *data; data = trace_peek_data(cpu); cpu_data[cpu].next = NULL; @@ -509,6 +519,8 @@ ssize_t trace_report(int fd, bool __repipe) file_bigendian = buf[0]; host_bigendian = bigendian(); + read_trace_init(file_bigendian, host_bigendian); + read_or_die(buf, 1); long_size = buf[0]; @@ -526,11 +538,11 @@ ssize_t trace_report(int fd, bool __repipe) repipe = false; if (show_funcs) { - print_funcs(); + pevent_print_funcs(perf_pevent); return size; } if (show_printk) { - print_printk(); + pevent_print_printk(perf_pevent); return size; } diff --git a/tools/perf/util/trace-event.h b/tools/perf/util/trace-event.h index 58ae14c5baac..639852ac1117 100644 --- a/tools/perf/util/trace-event.h +++ b/tools/perf/util/trace-event.h @@ -1,20 +1,21 @@ -#ifndef __PERF_TRACE_EVENTS_H -#define __PERF_TRACE_EVENTS_H +#ifndef _PERF_UTIL_TRACE_EVENT_H +#define _PERF_UTIL_TRACE_EVENT_H -#include <stdbool.h> #include "parse-events.h" +#include "event-parse.h" +#include "session.h" struct machine; struct perf_sample; union perf_event; struct thread; -#define __unused __attribute__((unused)) - +extern int header_page_size_size; +extern int header_page_ts_size; +extern int header_page_data_offset; -#ifndef PAGE_MASK -#define PAGE_MASK (page_size - 1) -#endif +extern bool latency_format; +extern struct pevent *perf_pevent; enum { RINGBUF_TYPE_PADDING = 29, @@ -26,246 +27,37 @@ enum { #define TS_SHIFT 27 #endif -#define NSECS_PER_SEC 1000000000ULL -#define NSECS_PER_USEC 1000ULL - -enum format_flags { - FIELD_IS_ARRAY = 1, - FIELD_IS_POINTER = 2, - FIELD_IS_SIGNED = 4, - FIELD_IS_STRING = 8, - FIELD_IS_DYNAMIC = 16, - FIELD_IS_FLAG = 32, - FIELD_IS_SYMBOLIC = 64, -}; - -struct format_field { - struct format_field *next; - char *type; - char *name; - int offset; - int size; - unsigned long flags; -}; - -struct format { - int nr_common; - int nr_fields; - struct format_field *common_fields; - struct format_field *fields; -}; - -struct print_arg_atom { - char *atom; -}; - -struct print_arg_string { - char *string; - int offset; -}; - -struct print_arg_field { - char *name; - struct format_field *field; -}; - -struct print_flag_sym { - struct print_flag_sym *next; - char *value; - char *str; -}; - -struct print_arg_typecast { - char *type; - struct print_arg *item; -}; - -struct print_arg_flags { - struct print_arg *field; - char *delim; - struct print_flag_sym *flags; -}; - -struct print_arg_symbol { - struct print_arg *field; - struct print_flag_sym *symbols; -}; - -struct print_arg; - -struct print_arg_op { - char *op; - int prio; - struct print_arg *left; - struct print_arg *right; -}; - -struct print_arg_func { - char *name; - struct print_arg *args; -}; - -enum print_arg_type { - PRINT_NULL, - PRINT_ATOM, - PRINT_FIELD, - PRINT_FLAGS, - PRINT_SYMBOL, - PRINT_TYPE, - PRINT_STRING, - PRINT_OP, -}; - -struct print_arg { - struct print_arg *next; - enum print_arg_type type; - union { - struct print_arg_atom atom; - struct print_arg_field field; - struct print_arg_typecast typecast; - struct print_arg_flags flags; - struct print_arg_symbol symbol; - struct print_arg_func func; - struct print_arg_string string; - struct print_arg_op op; - }; -}; - -struct print_fmt { - char *format; - struct print_arg *args; -}; - -struct event { - struct event *next; - char *name; - int id; - int flags; - struct format format; - struct print_fmt print_fmt; - char *system; -}; - -enum { - EVENT_FL_ISFTRACE = 0x01, - EVENT_FL_ISPRINT = 0x02, - EVENT_FL_ISBPRINT = 0x04, - EVENT_FL_ISFUNC = 0x08, - EVENT_FL_ISFUNCENT = 0x10, - EVENT_FL_ISFUNCRET = 0x20, - - EVENT_FL_FAILED = 0x80000000 -}; - -struct record { - unsigned long long ts; - int size; - void *data; -}; - -struct record *trace_peek_data(int cpu); -struct record *trace_read_data(int cpu); - -void parse_set_info(int nr_cpus, int long_sz); - -ssize_t trace_report(int fd, bool repipe); - -void *malloc_or_die(unsigned int size); +int bigendian(void); -void parse_cmdlines(char *file, int size); -void parse_proc_kallsyms(char *file, unsigned int size); -void parse_ftrace_printk(char *file, unsigned int size); +int read_trace_init(int file_bigendian, int host_bigendian); +void print_trace_event(int cpu, void *data, int size); -void print_funcs(void); -void print_printk(void); +void print_event(int cpu, void *data, int size, unsigned long long nsecs, + char *comm); int parse_ftrace_file(char *buf, unsigned long size); int parse_event_file(char *buf, unsigned long size, char *sys); -void print_trace_event(int cpu, void *data, int size); - -extern int file_bigendian; -extern int host_bigendian; - -int bigendian(void); - -static inline unsigned short __data2host2(unsigned short data) -{ - unsigned short swap; - - if (host_bigendian == file_bigendian) - return data; - swap = ((data & 0xffULL) << 8) | - ((data & (0xffULL << 8)) >> 8); +struct pevent_record *trace_peek_data(int cpu); +struct event_format *trace_find_event(int type); - return swap; -} - -static inline unsigned int __data2host4(unsigned int data) -{ - unsigned int swap; - - if (host_bigendian == file_bigendian) - return data; - - swap = ((data & 0xffULL) << 24) | - ((data & (0xffULL << 8)) << 8) | - ((data & (0xffULL << 16)) >> 8) | - ((data & (0xffULL << 24)) >> 24); - - return swap; -} - -static inline unsigned long long __data2host8(unsigned long long data) -{ - unsigned long long swap; - - if (host_bigendian == file_bigendian) - return data; - - swap = ((data & 0xffULL) << 56) | - ((data & (0xffULL << 8)) << 40) | - ((data & (0xffULL << 16)) << 24) | - ((data & (0xffULL << 24)) << 8) | - ((data & (0xffULL << 32)) >> 8) | - ((data & (0xffULL << 40)) >> 24) | - ((data & (0xffULL << 48)) >> 40) | - ((data & (0xffULL << 56)) >> 56); - - return swap; -} +unsigned long long +raw_field_value(struct event_format *event, const char *name, void *data); +void *raw_field_ptr(struct event_format *event, const char *name, void *data); -#define data2host2(ptr) __data2host2(*(unsigned short *)ptr) -#define data2host4(ptr) __data2host4(*(unsigned int *)ptr) -#define data2host8(ptr) ({ \ - unsigned long long __val; \ - \ - memcpy(&__val, (ptr), sizeof(unsigned long long)); \ - __data2host8(__val); \ -}) +void parse_proc_kallsyms(char *file, unsigned int size __unused); +void parse_ftrace_printk(char *file, unsigned int size __unused); -extern int header_page_ts_offset; -extern int header_page_ts_size; -extern int header_page_size_offset; -extern int header_page_size_size; -extern int header_page_data_offset; -extern int header_page_data_size; - -extern bool latency_format; +ssize_t trace_report(int fd, bool repipe); int trace_parse_common_type(void *data); int trace_parse_common_pid(void *data); -int parse_common_pc(void *data); -int parse_common_flags(void *data); -int parse_common_lock_depth(void *data); -struct event *trace_find_event(int id); -struct event *trace_find_next_event(struct event *event); + +struct event_format *trace_find_next_event(struct event_format *event); unsigned long long read_size(void *ptr, int size); -unsigned long long -raw_field_value(struct event *event, const char *name, void *data); -void *raw_field_ptr(struct event *event, const char *name, void *data); unsigned long long eval_flag(const char *flag); +struct pevent_record *trace_read_data(int cpu); int read_tracing_data(int fd, struct list_head *pattrs); struct tracing_data { @@ -280,15 +72,6 @@ struct tracing_data *tracing_data_get(struct list_head *pattrs, void tracing_data_put(struct tracing_data *tdata); -/* taken from kernel/trace/trace.h */ -enum trace_flag_type { - TRACE_FLAG_IRQS_OFF = 0x01, - TRACE_FLAG_IRQS_NOSUPPORT = 0x02, - TRACE_FLAG_NEED_RESCHED = 0x04, - TRACE_FLAG_HARDIRQ = 0x08, - TRACE_FLAG_SOFTIRQ = 0x10, -}; - struct scripting_ops { const char *name; int (*start_script) (const char *script, int argc, const char **argv); @@ -314,4 +97,4 @@ int common_pc(struct scripting_context *context); int common_flags(struct scripting_context *context); int common_lock_depth(struct scripting_context *context); -#endif /* __PERF_TRACE_EVENTS_H */ +#endif /* _PERF_UTIL_TRACE_EVENT_H */ diff --git a/tools/perf/util/types.h b/tools/perf/util/types.h index 5f3689a3d085..c51fa6b70a28 100644 --- a/tools/perf/util/types.h +++ b/tools/perf/util/types.h @@ -16,4 +16,9 @@ typedef signed short s16; typedef unsigned char u8; typedef signed char s8; +union u64_swap { + u64 val64; + u32 val32[2]; +}; + #endif /* __PERF_TYPES_H */ diff --git a/tools/perf/util/ui/browsers/annotate.c b/tools/perf/util/ui/browsers/annotate.c deleted file mode 100644 index 57a4c6ef3fd2..000000000000 --- a/tools/perf/util/ui/browsers/annotate.c +++ /dev/null @@ -1,433 +0,0 @@ -#include "../../util.h" -#include "../browser.h" -#include "../helpline.h" -#include "../libslang.h" -#include "../ui.h" -#include "../util.h" -#include "../../annotate.h" -#include "../../hist.h" -#include "../../sort.h" -#include "../../symbol.h" -#include <pthread.h> -#include <newt.h> - -struct annotate_browser { - struct ui_browser b; - struct rb_root entries; - struct rb_node *curr_hot; - struct objdump_line *selection; - int nr_asm_entries; - int nr_entries; - bool hide_src_code; -}; - -struct objdump_line_rb_node { - struct rb_node rb_node; - double percent; - u32 idx; - int idx_asm; -}; - -static inline -struct objdump_line_rb_node *objdump_line__rb(struct objdump_line *self) -{ - return (struct objdump_line_rb_node *)(self + 1); -} - -static bool objdump_line__filter(struct ui_browser *browser, void *entry) -{ - struct annotate_browser *ab = container_of(browser, struct annotate_browser, b); - - if (ab->hide_src_code) { - struct objdump_line *ol = list_entry(entry, struct objdump_line, node); - return ol->offset == -1; - } - - return false; -} - -static void annotate_browser__write(struct ui_browser *self, void *entry, int row) -{ - struct annotate_browser *ab = container_of(self, struct annotate_browser, b); - struct objdump_line *ol = list_entry(entry, struct objdump_line, node); - bool current_entry = ui_browser__is_current_entry(self, row); - int width = self->width; - - if (ol->offset != -1) { - struct objdump_line_rb_node *olrb = objdump_line__rb(ol); - ui_browser__set_percent_color(self, olrb->percent, current_entry); - slsmg_printf(" %7.2f ", olrb->percent); - } else { - ui_browser__set_percent_color(self, 0, current_entry); - slsmg_write_nstring(" ", 9); - } - - SLsmg_write_char(':'); - slsmg_write_nstring(" ", 8); - - /* The scroll bar isn't being used */ - if (!self->navkeypressed) - width += 1; - - if (!ab->hide_src_code && ol->offset != -1) - if (!current_entry || (self->use_navkeypressed && - !self->navkeypressed)) - ui_browser__set_color(self, HE_COLORSET_CODE); - - if (!*ol->line) - slsmg_write_nstring(" ", width - 18); - else - slsmg_write_nstring(ol->line, width - 18); - - if (current_entry) - ab->selection = ol; -} - -static double objdump_line__calc_percent(struct objdump_line *self, - struct symbol *sym, int evidx) -{ - double percent = 0.0; - - if (self->offset != -1) { - int len = sym->end - sym->start; - unsigned int hits = 0; - struct annotation *notes = symbol__annotation(sym); - struct source_line *src_line = notes->src->lines; - struct sym_hist *h = annotation__histogram(notes, evidx); - s64 offset = self->offset; - struct objdump_line *next; - - next = objdump__get_next_ip_line(¬es->src->source, self); - while (offset < (s64)len && - (next == NULL || offset < next->offset)) { - if (src_line) { - percent += src_line[offset].percent; - } else - hits += h->addr[offset]; - - ++offset; - } - /* - * If the percentage wasn't already calculated in - * symbol__get_source_line, do it now: - */ - if (src_line == NULL && h->sum) - percent = 100.0 * hits / h->sum; - } - - return percent; -} - -static void objdump__insert_line(struct rb_root *self, - struct objdump_line_rb_node *line) -{ - struct rb_node **p = &self->rb_node; - struct rb_node *parent = NULL; - struct objdump_line_rb_node *l; - - while (*p != NULL) { - parent = *p; - l = rb_entry(parent, struct objdump_line_rb_node, rb_node); - if (line->percent < l->percent) - p = &(*p)->rb_left; - else - p = &(*p)->rb_right; - } - rb_link_node(&line->rb_node, parent, p); - rb_insert_color(&line->rb_node, self); -} - -static void annotate_browser__set_top(struct annotate_browser *self, - struct rb_node *nd) -{ - struct objdump_line_rb_node *rbpos; - struct objdump_line *pos; - unsigned back; - - ui_browser__refresh_dimensions(&self->b); - back = self->b.height / 2; - rbpos = rb_entry(nd, struct objdump_line_rb_node, rb_node); - pos = ((struct objdump_line *)rbpos) - 1; - self->b.top_idx = self->b.index = rbpos->idx; - - while (self->b.top_idx != 0 && back != 0) { - pos = list_entry(pos->node.prev, struct objdump_line, node); - - --self->b.top_idx; - --back; - } - - self->b.top = pos; - self->curr_hot = nd; -} - -static void annotate_browser__calc_percent(struct annotate_browser *browser, - int evidx) -{ - struct map_symbol *ms = browser->b.priv; - struct symbol *sym = ms->sym; - struct annotation *notes = symbol__annotation(sym); - struct objdump_line *pos; - - browser->entries = RB_ROOT; - - pthread_mutex_lock(¬es->lock); - - list_for_each_entry(pos, ¬es->src->source, node) { - struct objdump_line_rb_node *rbpos = objdump_line__rb(pos); - rbpos->percent = objdump_line__calc_percent(pos, sym, evidx); - if (rbpos->percent < 0.01) { - RB_CLEAR_NODE(&rbpos->rb_node); - continue; - } - objdump__insert_line(&browser->entries, rbpos); - } - pthread_mutex_unlock(¬es->lock); - - browser->curr_hot = rb_last(&browser->entries); -} - -static bool annotate_browser__toggle_source(struct annotate_browser *browser) -{ - struct objdump_line *ol; - struct objdump_line_rb_node *olrb; - off_t offset = browser->b.index - browser->b.top_idx; - - browser->b.seek(&browser->b, offset, SEEK_CUR); - ol = list_entry(browser->b.top, struct objdump_line, node); - olrb = objdump_line__rb(ol); - - if (browser->hide_src_code) { - if (olrb->idx_asm < offset) - offset = olrb->idx; - - browser->b.nr_entries = browser->nr_entries; - browser->hide_src_code = false; - browser->b.seek(&browser->b, -offset, SEEK_CUR); - browser->b.top_idx = olrb->idx - offset; - browser->b.index = olrb->idx; - } else { - if (olrb->idx_asm < 0) { - ui_helpline__puts("Only available for assembly lines."); - browser->b.seek(&browser->b, -offset, SEEK_CUR); - return false; - } - - if (olrb->idx_asm < offset) - offset = olrb->idx_asm; - - browser->b.nr_entries = browser->nr_asm_entries; - browser->hide_src_code = true; - browser->b.seek(&browser->b, -offset, SEEK_CUR); - browser->b.top_idx = olrb->idx_asm - offset; - browser->b.index = olrb->idx_asm; - } - - return true; -} - -static int annotate_browser__run(struct annotate_browser *self, int evidx, - void(*timer)(void *arg), - void *arg, int delay_secs) -{ - struct rb_node *nd = NULL; - struct map_symbol *ms = self->b.priv; - struct symbol *sym = ms->sym; - const char *help = "<-/ESC: Exit, TAB/shift+TAB: Cycle hot lines, " - "H: Go to hottest line, ->/ENTER: Line action, " - "S: Toggle source code view"; - int key; - - if (ui_browser__show(&self->b, sym->name, help) < 0) - return -1; - - annotate_browser__calc_percent(self, evidx); - - if (self->curr_hot) - annotate_browser__set_top(self, self->curr_hot); - - nd = self->curr_hot; - - while (1) { - key = ui_browser__run(&self->b, delay_secs); - - if (delay_secs != 0) { - annotate_browser__calc_percent(self, evidx); - /* - * Current line focus got out of the list of most active - * lines, NULL it so that if TAB|UNTAB is pressed, we - * move to curr_hot (current hottest line). - */ - if (nd != NULL && RB_EMPTY_NODE(nd)) - nd = NULL; - } - - switch (key) { - case K_TIMER: - if (timer != NULL) - timer(arg); - - if (delay_secs != 0) - symbol__annotate_decay_histogram(sym, evidx); - continue; - case K_TAB: - if (nd != NULL) { - nd = rb_prev(nd); - if (nd == NULL) - nd = rb_last(&self->entries); - } else - nd = self->curr_hot; - break; - case K_UNTAB: - if (nd != NULL) - nd = rb_next(nd); - if (nd == NULL) - nd = rb_first(&self->entries); - else - nd = self->curr_hot; - break; - case 'H': - case 'h': - nd = self->curr_hot; - break; - case 'S': - case 's': - if (annotate_browser__toggle_source(self)) - ui_helpline__puts(help); - continue; - case K_ENTER: - case K_RIGHT: - if (self->selection == NULL) { - ui_helpline__puts("Huh? No selection. Report to linux-kernel@vger.kernel.org"); - continue; - } - - if (self->selection->offset == -1) { - ui_helpline__puts("Actions are only available for assembly lines."); - continue; - } else { - char *s = strstr(self->selection->line, "callq "); - struct annotation *notes; - struct symbol *target; - u64 ip; - - if (s == NULL) { - ui_helpline__puts("Actions are only available for the 'callq' instruction."); - continue; - } - - s = strchr(s, ' '); - if (s++ == NULL) { - ui_helpline__puts("Invallid callq instruction."); - continue; - } - - ip = strtoull(s, NULL, 16); - ip = ms->map->map_ip(ms->map, ip); - target = map__find_symbol(ms->map, ip, NULL); - if (target == NULL) { - ui_helpline__puts("The called function was not found."); - continue; - } - - notes = symbol__annotation(target); - pthread_mutex_lock(¬es->lock); - - if (notes->src == NULL && symbol__alloc_hist(target) < 0) { - pthread_mutex_unlock(¬es->lock); - ui__warning("Not enough memory for annotating '%s' symbol!\n", - target->name); - continue; - } - - pthread_mutex_unlock(¬es->lock); - symbol__tui_annotate(target, ms->map, evidx, - timer, arg, delay_secs); - ui_browser__show_title(&self->b, sym->name); - } - continue; - case K_LEFT: - case K_ESC: - case 'q': - case CTRL('c'): - goto out; - default: - continue; - } - - if (nd != NULL) - annotate_browser__set_top(self, nd); - } -out: - ui_browser__hide(&self->b); - return key; -} - -int hist_entry__tui_annotate(struct hist_entry *he, int evidx, - void(*timer)(void *arg), void *arg, int delay_secs) -{ - return symbol__tui_annotate(he->ms.sym, he->ms.map, evidx, - timer, arg, delay_secs); -} - -int symbol__tui_annotate(struct symbol *sym, struct map *map, int evidx, - void(*timer)(void *arg), void *arg, - int delay_secs) -{ - struct objdump_line *pos, *n; - struct annotation *notes; - struct map_symbol ms = { - .map = map, - .sym = sym, - }; - struct annotate_browser browser = { - .b = { - .refresh = ui_browser__list_head_refresh, - .seek = ui_browser__list_head_seek, - .write = annotate_browser__write, - .filter = objdump_line__filter, - .priv = &ms, - .use_navkeypressed = true, - }, - }; - int ret; - - if (sym == NULL) - return -1; - - if (map->dso->annotate_warned) - return -1; - - if (symbol__annotate(sym, map, sizeof(struct objdump_line_rb_node)) < 0) { - ui__error("%s", ui_helpline__last_msg); - return -1; - } - - ui_helpline__push("Press <- or ESC to exit"); - - notes = symbol__annotation(sym); - - list_for_each_entry(pos, ¬es->src->source, node) { - struct objdump_line_rb_node *rbpos; - size_t line_len = strlen(pos->line); - - if (browser.b.width < line_len) - browser.b.width = line_len; - rbpos = objdump_line__rb(pos); - rbpos->idx = browser.nr_entries++; - if (pos->offset != -1) - rbpos->idx_asm = browser.nr_asm_entries++; - else - rbpos->idx_asm = -1; - } - - browser.b.nr_entries = browser.nr_entries; - browser.b.entries = ¬es->src->source, - browser.b.width += 18; /* Percentage */ - ret = annotate_browser__run(&browser, evidx, timer, arg, delay_secs); - list_for_each_entry_safe(pos, n, ¬es->src->source, node) { - list_del(&pos->node); - objdump_line__free(pos); - } - return ret; -} diff --git a/tools/perf/util/usage.c b/tools/perf/util/usage.c index 52bb07c6442a..4007aca8e0ca 100644 --- a/tools/perf/util/usage.c +++ b/tools/perf/util/usage.c @@ -82,41 +82,3 @@ void warning(const char *warn, ...) warn_routine(warn, params); va_end(params); } - -uid_t parse_target_uid(const char *str, const char *tid, const char *pid) -{ - struct passwd pwd, *result; - char buf[1024]; - - if (str == NULL) - return UINT_MAX; - - /* UID and PID are mutually exclusive */ - if (tid || pid) { - ui__warning("PID/TID switch overriding UID\n"); - sleep(1); - return UINT_MAX; - } - - getpwnam_r(str, &pwd, buf, sizeof(buf), &result); - - if (result == NULL) { - char *endptr; - int uid = strtol(str, &endptr, 10); - - if (*endptr != '\0') { - ui__error("Invalid user %s\n", str); - return UINT_MAX - 1; - } - - getpwuid_r(uid, &pwd, buf, sizeof(buf), &result); - - if (result == NULL) { - ui__error("Problems obtaining information for user %s\n", - str); - return UINT_MAX - 1; - } - } - - return result->pw_uid; -} diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c index 8109a907841e..d03599fbe78b 100644 --- a/tools/perf/util/util.c +++ b/tools/perf/util/util.c @@ -148,3 +148,13 @@ int readn(int fd, void *buf, size_t n) return buf - buf_start; } + +size_t hex_width(u64 v) +{ + size_t n = 1; + + while ((v >>= 4)) + ++n; + + return n; +} diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 0f99f394d8e0..2daaedb83d84 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -74,7 +74,6 @@ #include <netinet/tcp.h> #include <arpa/inet.h> #include <netdb.h> -#include <pwd.h> #include <inttypes.h> #include "../../../include/linux/magic.h" #include "types.h" @@ -249,8 +248,6 @@ struct perf_event_attr; void event_attr_init(struct perf_event_attr *attr); -uid_t parse_target_uid(const char *str, const char *tid, const char *pid); - #define _STR(x) #x #define STR(x) _STR(x) @@ -265,4 +262,6 @@ bool is_power_of_2(unsigned long n) return (n != 0 && ((n & (n - 1)) == 0)); } +size_t hex_width(u64 v); + #endif diff --git a/tools/power/cpupower/Makefile b/tools/power/cpupower/Makefile index e8a03aceceb1..a93e06cfcc2a 100644 --- a/tools/power/cpupower/Makefile +++ b/tools/power/cpupower/Makefile @@ -19,6 +19,16 @@ # along with this program; if not, write to the Free Software # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # +OUTPUT=./ +ifeq ("$(origin O)", "command line") + OUTPUT := $(O)/ +endif + +ifneq ($(OUTPUT),) +# check that the output directory actually exists +OUTDIR := $(shell cd $(OUTPUT) && /bin/pwd) +$(if $(OUTDIR),, $(error output directory "$(OUTPUT)" does not exist)) +endif # --- CONFIGURATION BEGIN --- @@ -87,6 +97,7 @@ AR = $(CROSS)ar STRIP = $(CROSS)strip RANLIB = $(CROSS)ranlib HOSTCC = gcc +MKDIR = mkdir # Now we set up the build system @@ -95,7 +106,7 @@ HOSTCC = gcc # set up PWD so that older versions of make will work with our build. PWD = $(shell pwd) -GMO_FILES = ${shell for HLANG in ${LANGUAGES}; do echo po/$$HLANG.gmo; done;} +GMO_FILES = ${shell for HLANG in ${LANGUAGES}; do echo $(OUTPUT)po/$$HLANG.gmo; done;} export CROSS CC AR STRIP RANLIB CFLAGS LDFLAGS LIB_OBJS @@ -122,15 +133,18 @@ UTIL_OBJS = utils/helpers/amd.o utils/helpers/topology.o utils/helpers/msr.o \ utils/cpupower.o utils/cpufreq-info.o utils/cpufreq-set.o \ utils/cpupower-set.o utils/cpupower-info.o utils/cpuidle-info.o +UTIL_SRC := $(UTIL_OBJS:.o=.c) + +UTIL_OBJS := $(addprefix $(OUTPUT),$(UTIL_OBJS)) + UTIL_HEADERS = utils/helpers/helpers.h utils/idle_monitor/cpupower-monitor.h \ utils/helpers/bitmask.h \ utils/idle_monitor/idle_monitors.h utils/idle_monitor/idle_monitors.def -UTIL_SRC := $(UTIL_OBJS:.o=.c) - LIB_HEADERS = lib/cpufreq.h lib/sysfs.h LIB_SRC = lib/cpufreq.c lib/sysfs.c LIB_OBJS = lib/cpufreq.o lib/sysfs.o +LIB_OBJS := $(addprefix $(OUTPUT),$(LIB_OBJS)) CFLAGS += -pipe @@ -168,83 +182,90 @@ endif # the actual make rules -all: libcpupower cpupower $(COMPILE_NLS) $(COMPILE_BENCH) +all: libcpupower $(OUTPUT)cpupower $(COMPILE_NLS) $(COMPILE_BENCH) -lib/%.o: $(LIB_SRC) $(LIB_HEADERS) +$(OUTPUT)lib/%.o: $(LIB_SRC) $(LIB_HEADERS) $(ECHO) " CC " $@ $(QUIET) $(CC) $(CFLAGS) -fPIC -o $@ -c lib/$*.c -libcpupower.so.$(LIB_MAJ): $(LIB_OBJS) +$(OUTPUT)libcpupower.so.$(LIB_MAJ): $(LIB_OBJS) $(ECHO) " LD " $@ $(QUIET) $(CC) -shared $(CFLAGS) $(LDFLAGS) -o $@ \ -Wl,-soname,libcpupower.so.$(LIB_MIN) $(LIB_OBJS) - @ln -sf $@ libcpupower.so - @ln -sf $@ libcpupower.so.$(LIB_MIN) + @ln -sf $(@F) $(OUTPUT)libcpupower.so + @ln -sf $(@F) $(OUTPUT)libcpupower.so.$(LIB_MIN) -libcpupower: libcpupower.so.$(LIB_MAJ) +libcpupower: $(OUTPUT)libcpupower.so.$(LIB_MAJ) # Let all .o files depend on its .c file and all headers # Might be worth to put this into utils/Makefile at some point of time $(UTIL_OBJS): $(UTIL_HEADERS) -.c.o: +$(OUTPUT)%.o: %.c $(ECHO) " CC " $@ $(QUIET) $(CC) $(CFLAGS) -I./lib -I ./utils -o $@ -c $*.c -cpupower: $(UTIL_OBJS) libcpupower.so.$(LIB_MAJ) +$(OUTPUT)cpupower: $(UTIL_OBJS) $(OUTPUT)libcpupower.so.$(LIB_MAJ) $(ECHO) " CC " $@ - $(QUIET) $(CC) $(CFLAGS) $(LDFLAGS) -lcpupower -lrt -lpci -L. -o $@ $(UTIL_OBJS) + $(QUIET) $(CC) $(CFLAGS) $(LDFLAGS) $(UTIL_OBJS) -lcpupower -lrt -lpci -L$(OUTPUT) -o $@ $(QUIET) $(STRIPCMD) $@ -po/$(PACKAGE).pot: $(UTIL_SRC) +$(OUTPUT)po/$(PACKAGE).pot: $(UTIL_SRC) $(ECHO) " GETTEXT " $@ $(QUIET) xgettext --default-domain=$(PACKAGE) --add-comments \ - --keyword=_ --keyword=N_ $(UTIL_SRC) && \ - test -f $(PACKAGE).po && \ - mv -f $(PACKAGE).po po/$(PACKAGE).pot + --keyword=_ --keyword=N_ $(UTIL_SRC) -p $(@D) -o $(@F) -po/%.gmo: po/%.po +$(OUTPUT)po/%.gmo: po/%.po $(ECHO) " MSGFMT " $@ $(QUIET) msgfmt -o $@ po/$*.po create-gmo: ${GMO_FILES} -update-po: po/$(PACKAGE).pot +update-po: $(OUTPUT)po/$(PACKAGE).pot $(ECHO) " MSGMRG " $@ $(QUIET) @for HLANG in $(LANGUAGES); do \ echo -n "Updating $$HLANG "; \ - if msgmerge po/$$HLANG.po po/$(PACKAGE).pot -o \ - po/$$HLANG.new.po; then \ - mv -f po/$$HLANG.new.po po/$$HLANG.po; \ + if msgmerge po/$$HLANG.po $< -o \ + $(OUTPUT)po/$$HLANG.new.po; then \ + mv -f $(OUTPUT)po/$$HLANG.new.po $(OUTPUT)po/$$HLANG.po; \ else \ echo "msgmerge for $$HLANG failed!"; \ - rm -f po/$$HLANG.new.po; \ + rm -f $(OUTPUT)po/$$HLANG.new.po; \ fi; \ done; -compile-bench: libcpupower.so.$(LIB_MAJ) - @V=$(V) confdir=$(confdir) $(MAKE) -C bench +compile-bench: $(OUTPUT)libcpupower.so.$(LIB_MAJ) + @V=$(V) confdir=$(confdir) $(MAKE) -C bench O=$(OUTPUT) + +# we compile into subdirectories. if the target directory is not the +# source directory, they might not exists. So we depend the various +# files onto their directories. +DIRECTORY_DEPS = $(LIB_OBJS) $(UTIL_OBJS) $(GMO_FILES) +$(DIRECTORY_DEPS): | $(sort $(dir $(DIRECTORY_DEPS))) + +# In the second step, we make a rule to actually create these directories +$(sort $(dir $(DIRECTORY_DEPS))): + $(ECHO) " MKDIR " $@ + $(QUIET) $(MKDIR) -p $@ 2>/dev/null clean: - -find . \( -not -type d \) -and \( -name '*~' -o -name '*.[oas]' \) -type f -print \ + -find $(OUTPUT) \( -not -type d \) -and \( -name '*~' -o -name '*.[oas]' \) -type f -print \ | xargs rm -f - -rm -f $(UTIL_BINS) - -rm -f $(IDLE_OBJS) - -rm -f cpupower - -rm -f libcpupower.so* - -rm -rf po/*.gmo po/*.pot - $(MAKE) -C bench clean + -rm -f $(OUTPUT)cpupower + -rm -f $(OUTPUT)libcpupower.so* + -rm -rf $(OUTPUT)po/*.{gmo,pot} + $(MAKE) -C bench O=$(OUTPUT) clean install-lib: $(INSTALL) -d $(DESTDIR)${libdir} - $(CP) libcpupower.so* $(DESTDIR)${libdir}/ + $(CP) $(OUTPUT)libcpupower.so* $(DESTDIR)${libdir}/ $(INSTALL) -d $(DESTDIR)${includedir} $(INSTALL_DATA) lib/cpufreq.h $(DESTDIR)${includedir}/cpufreq.h install-tools: $(INSTALL) -d $(DESTDIR)${bindir} - $(INSTALL_PROGRAM) cpupower $(DESTDIR)${bindir} + $(INSTALL_PROGRAM) $(OUTPUT)cpupower $(DESTDIR)${bindir} install-man: $(INSTALL_DATA) -D man/cpupower.1 $(DESTDIR)${mandir}/man1/cpupower.1 @@ -257,13 +278,13 @@ install-man: install-gmo: $(INSTALL) -d $(DESTDIR)${localedir} for HLANG in $(LANGUAGES); do \ - echo '$(INSTALL_DATA) -D po/$$HLANG.gmo $(DESTDIR)${localedir}/$$HLANG/LC_MESSAGES/cpupower.mo'; \ - $(INSTALL_DATA) -D po/$$HLANG.gmo $(DESTDIR)${localedir}/$$HLANG/LC_MESSAGES/cpupower.mo; \ + echo '$(INSTALL_DATA) -D $(OUTPUT)po/$$HLANG.gmo $(DESTDIR)${localedir}/$$HLANG/LC_MESSAGES/cpupower.mo'; \ + $(INSTALL_DATA) -D $(OUTPUT)po/$$HLANG.gmo $(DESTDIR)${localedir}/$$HLANG/LC_MESSAGES/cpupower.mo; \ done; install-bench: @#DESTDIR must be set from outside to survive - @sbindir=$(sbindir) bindir=$(bindir) docdir=$(docdir) confdir=$(confdir) $(MAKE) -C bench install + @sbindir=$(sbindir) bindir=$(bindir) docdir=$(docdir) confdir=$(confdir) $(MAKE) -C bench O=$(OUTPUT) install install: all install-lib install-tools install-man $(INSTALL_NLS) $(INSTALL_BENCH) diff --git a/tools/power/cpupower/bench/Makefile b/tools/power/cpupower/bench/Makefile index 2b67606fc3e3..7ec7021a29cd 100644 --- a/tools/power/cpupower/bench/Makefile +++ b/tools/power/cpupower/bench/Makefile @@ -1,29 +1,36 @@ -LIBS = -L../ -lm -lcpupower +OUTPUT := ./ +ifeq ("$(origin O)", "command line") +ifneq ($(O),) + OUTPUT := $(O)/ +endif +endif -OBJS = main.o parse.o system.o benchmark.o +LIBS = -L../ -L$(OUTPUT) -lm -lcpupower + +OBJS = $(OUTPUT)main.o $(OUTPUT)parse.o $(OUTPUT)system.o $(OUTPUT)benchmark.o CFLAGS += -D_GNU_SOURCE -I../lib -DDEFAULT_CONFIG_FILE=\"$(confdir)/cpufreq-bench.conf\" -%.o : %.c +$(OUTPUT)%.o : %.c $(ECHO) " CC " $@ $(QUIET) $(CC) -c $(CFLAGS) $< -o $@ -cpufreq-bench: $(OBJS) +$(OUTPUT)cpufreq-bench: $(OBJS) $(ECHO) " CC " $@ $(QUIET) $(CC) -o $@ $(CFLAGS) $(OBJS) $(LIBS) -all: cpufreq-bench +all: $(OUTPUT)cpufreq-bench install: mkdir -p $(DESTDIR)/$(sbindir) mkdir -p $(DESTDIR)/$(bindir) mkdir -p $(DESTDIR)/$(docdir) mkdir -p $(DESTDIR)/$(confdir) - install -m 755 cpufreq-bench $(DESTDIR)/$(sbindir)/cpufreq-bench + install -m 755 $(OUTPUT)cpufreq-bench $(DESTDIR)/$(sbindir)/cpufreq-bench install -m 755 cpufreq-bench_plot.sh $(DESTDIR)/$(bindir)/cpufreq-bench_plot.sh install -m 644 README-BENCH $(DESTDIR)/$(docdir)/README-BENCH install -m 755 cpufreq-bench_script.sh $(DESTDIR)/$(docdir)/cpufreq-bench_script.sh install -m 644 example.cfg $(DESTDIR)/$(confdir)/cpufreq-bench.conf clean: - rm -f *.o - rm -f cpufreq-bench + rm -f $(OUTPUT)*.o + rm -f $(OUTPUT)cpufreq-bench diff --git a/tools/power/cpupower/debug/i386/Makefile b/tools/power/cpupower/debug/i386/Makefile index d08cc1ead9bc..3ba158f0e287 100644 --- a/tools/power/cpupower/debug/i386/Makefile +++ b/tools/power/cpupower/debug/i386/Makefile @@ -1,20 +1,38 @@ +OUTPUT=./ +ifeq ("$(origin O)", "command line") + OUTPUT := $(O)/ +endif + +DESTDIR = +bindir = /usr/bin + +INSTALL = /usr/bin/install + + default: all -centrino-decode: centrino-decode.c - $(CC) $(CFLAGS) -o centrino-decode centrino-decode.c +$(OUTPUT)centrino-decode: centrino-decode.c + $(CC) $(CFLAGS) -o $@ centrino-decode.c -dump_psb: dump_psb.c - $(CC) $(CFLAGS) -o dump_psb dump_psb.c +$(OUTPUT)dump_psb: dump_psb.c + $(CC) $(CFLAGS) -o $@ dump_psb.c -intel_gsic: intel_gsic.c - $(CC) $(CFLAGS) -o intel_gsic -llrmi intel_gsic.c +$(OUTPUT)intel_gsic: intel_gsic.c + $(CC) $(CFLAGS) -o $@ -llrmi intel_gsic.c -powernow-k8-decode: powernow-k8-decode.c - $(CC) $(CFLAGS) -o powernow-k8-decode powernow-k8-decode.c +$(OUTPUT)powernow-k8-decode: powernow-k8-decode.c + $(CC) $(CFLAGS) -o $@ powernow-k8-decode.c -all: centrino-decode dump_psb intel_gsic powernow-k8-decode +all: $(OUTPUT)centrino-decode $(OUTPUT)dump_psb $(OUTPUT)intel_gsic $(OUTPUT)powernow-k8-decode clean: - rm -rf centrino-decode dump_psb intel_gsic powernow-k8-decode + rm -rf $(OUTPUT){centrino-decode,dump_psb,intel_gsic,powernow-k8-decode} + +install: + $(INSTALL) -d $(DESTDIR)${bindir} + $(INSTALL) $(OUTPUT)centrino-decode $(DESTDIR)${bindir} + $(INSTALL) $(OUTPUT)powernow-k8-decode $(DESTDIR)${bindir} + $(INSTALL) $(OUTPUT)dump_psb $(DESTDIR)${bindir} + $(INSTALL) $(OUTPUT)intel_gsic $(DESTDIR)${bindir} -.PHONY: all default clean +.PHONY: all default clean install diff --git a/tools/power/cpupower/debug/x86_64/Makefile b/tools/power/cpupower/debug/x86_64/Makefile index 3326217dd311..1c5214526716 100644 --- a/tools/power/cpupower/debug/x86_64/Makefile +++ b/tools/power/cpupower/debug/x86_64/Makefile @@ -1,14 +1,30 @@ +OUTPUT=./ +ifeq ("$(origin O)", "command line") + OUTPUT := $(O)/ +endif + +DESTDIR = +bindir = /usr/bin + +INSTALL = /usr/bin/install + + default: all -centrino-decode: ../i386/centrino-decode.c +$(OUTPUT)centrino-decode: ../i386/centrino-decode.c $(CC) $(CFLAGS) -o $@ $< -powernow-k8-decode: ../i386/powernow-k8-decode.c +$(OUTPUT)powernow-k8-decode: ../i386/powernow-k8-decode.c $(CC) $(CFLAGS) -o $@ $< -all: centrino-decode powernow-k8-decode +all: $(OUTPUT)centrino-decode $(OUTPUT)powernow-k8-decode clean: - rm -rf centrino-decode powernow-k8-decode + rm -rf $(OUTPUT)centrino-decode $(OUTPUT)powernow-k8-decode + +install: + $(INSTALL) -d $(DESTDIR)${bindir} + $(INSTALL) $(OUTPUT)centrino-decode $(DESTDIR)${bindir} + $(INSTALL) $(OUTPUT)powernow-k8-decode $(DESTDIR)${bindir} -.PHONY: all default clean +.PHONY: all default clean install diff --git a/tools/power/cpupower/man/cpupower-frequency-info.1 b/tools/power/cpupower/man/cpupower-frequency-info.1 index bb60a8d1e45a..4a1918ea8f9c 100644 --- a/tools/power/cpupower/man/cpupower-frequency-info.1 +++ b/tools/power/cpupower/man/cpupower-frequency-info.1 @@ -1,4 +1,4 @@ -.TH "cpupower-frequency-info" "1" "0.1" "Mattia Dongili" "" +.TH "CPUPOWER\-FREQUENCY\-INFO" "1" "0.1" "" "cpupower Manual" .SH "NAME" .LP cpupower frequency\-info \- Utility to retrieve cpufreq kernel information @@ -50,8 +50,6 @@ Prints out information like provided by the /proc/cpufreq interface in 2.4. and \fB\-m\fR \fB\-\-human\fR human\-readable output for the \-f, \-w, \-s and \-y parameters. .TP -\fB\-h\fR \fB\-\-help\fR -Prints out the help screen. .SH "REMARKS" .LP By default only values of core zero are displayed. How to display settings of diff --git a/tools/power/cpupower/man/cpupower-frequency-set.1 b/tools/power/cpupower/man/cpupower-frequency-set.1 index 685f469093ad..3eacc8d03d1a 100644 --- a/tools/power/cpupower/man/cpupower-frequency-set.1 +++ b/tools/power/cpupower/man/cpupower-frequency-set.1 @@ -1,4 +1,4 @@ -.TH "cpupower-freqency-set" "1" "0.1" "Mattia Dongili" "" +.TH "CPUPOWER\-FREQUENCY\-SET" "1" "0.1" "" "cpupower Manual" .SH "NAME" .LP cpupower frequency\-set \- A small tool which allows to modify cpufreq settings. @@ -26,8 +26,6 @@ specific frequency to be set. Requires userspace governor to be available and lo \fB\-r\fR \fB\-\-related\fR modify all hardware-related CPUs at the same time .TP -\fB\-h\fR \fB\-\-help\fR -Prints out the help screen. .SH "REMARKS" .LP By default values are applied on all cores. How to modify single core diff --git a/tools/power/cpupower/man/cpupower-idle-info.1 b/tools/power/cpupower/man/cpupower-idle-info.1 new file mode 100644 index 000000000000..4178effd9e99 --- /dev/null +++ b/tools/power/cpupower/man/cpupower-idle-info.1 @@ -0,0 +1,90 @@ +.TH "CPUPOWER-IDLE-INFO" "1" "0.1" "" "cpupower Manual" +.SH "NAME" +.LP +cpupower idle\-info \- Utility to retrieve cpu idle kernel information +.SH "SYNTAX" +.LP +cpupower [ \-c cpulist ] idle\-info [\fIoptions\fP] +.SH "DESCRIPTION" +.LP +A tool which prints out per cpu idle information helpful to developers and interested users. +.SH "OPTIONS" +.LP +.TP +\fB\-f\fR \fB\-\-silent\fR +Only print a summary of all available C-states in the system. +.TP +\fB\-e\fR \fB\-\-proc\fR +deprecated. +Prints out idle information in old /proc/acpi/processor/*/power format. This +interface has been removed from the kernel for quite some time, do not let +further code depend on this option, best do not use it. + +.SH IDLE\-INFO DESCRIPTIONS +CPU sleep state statistics and descriptions are retrieved from sysfs files, +exported by the cpuidle kernel subsystem. The kernel only updates these +statistics when it enters or leaves an idle state, therefore on a very idle or +a very busy system, these statistics may not be accurate. They still provide a +good overview about the usage and availability of processor sleep states on +the platform. + +Be aware that the sleep states as exported by the hardware or BIOS and used by +the Linux kernel may not exactly reflect the capabilities of the +processor. This often is the case on the X86 architecture when the acpi_idle +driver is used. It is also possible that the hardware overrules the kernel +requests, due to internal activity monitors or other reasons. +On recent X86 platforms it is often possible to read out hardware registers +which monitor the duration of sleep states the processor resided in. The +cpupower monitor tool (cpupower\-monitor(1)) can be used to show real sleep +state residencies. Please refer to the architecture specific description +section below. + +.SH IDLE\-INFO ARCHITECTURE SPECIFIC DESCRIPTIONS +.SS "X86" +POLL idle state + +If cpuidle is active, X86 platforms have one special idle state. +The POLL idle state is not a real idle state, it does not save any +power. Instead, a busy\-loop is executed doing nothing for a short period of +time. This state is used if the kernel knows that work has to be processed +very soon and entering any real hardware idle state may result in a slight +performance penalty. + +There exist two different cpuidle drivers on the X86 architecture platform: + +"acpi_idle" cpuidle driver + +The acpi_idle cpuidle driver retrieves available sleep states (C\-states) from +the ACPI BIOS tables (from the _CST ACPI function on recent platforms or from +the FADT BIOS table on older ones). +The C1 state is not retrieved from ACPI tables. If the C1 state is entered, +the kernel will call the hlt instruction (or mwait on Intel). + +"intel_idle" cpuidle driver + +In kernel 2.6.36 the intel_idle driver was introduced. +It only serves recent Intel CPUs (Nehalem, Westmere, Sandybridge, Atoms or +newer). On older Intel CPUs the acpi_idle driver is still used (if the BIOS +provides C\-state ACPI tables). +The intel_idle driver knows the sleep state capabilities of the processor and +ignores ACPI BIOS exported processor sleep states tables. + +.SH "REMARKS" +.LP +By default only values of core zero are displayed. How to display settings of +other cores is described in the cpupower(1) manpage in the \-\-cpu option +section. +.SH REFERENCES +http://www.acpi.info/spec.htm +.SH "FILES" +.nf +\fI/sys/devices/system/cpu/cpu*/cpuidle/state*\fP +\fI/sys/devices/system/cpu/cpuidle/*\fP +.fi +.SH "AUTHORS" +.nf +Thomas Renninger <trenn@suse.de> +.fi +.SH "SEE ALSO" +.LP +cpupower(1), cpupower\-monitor(1), cpupower\-info(1), cpupower\-set(1) diff --git a/tools/power/cpupower/man/cpupower-monitor.1 b/tools/power/cpupower/man/cpupower-monitor.1 index d5cfa265c3d3..1141c2073719 100644 --- a/tools/power/cpupower/man/cpupower-monitor.1 +++ b/tools/power/cpupower/man/cpupower-monitor.1 @@ -107,7 +107,7 @@ Deepest package sleep states may in reality show up as machine/platform wide sleep states and can only be entered if all cores are idle. Look up Intel manuals (some are provided in the References section) for further details. -.SS "Ontario" "Liano" +.SS "Fam_12h" "Fam_14h" AMD laptop and desktop processor (family 12h and 14h) sleep state counters. The registers are accessed via PCI and therefore can still be read out while cores have been offlined. diff --git a/tools/power/cpupower/man/cpupower-set.1 b/tools/power/cpupower/man/cpupower-set.1 index c4954a9fe4e7..9dbd536518ab 100644 --- a/tools/power/cpupower/man/cpupower-set.1 +++ b/tools/power/cpupower/man/cpupower-set.1 @@ -85,15 +85,6 @@ Possible values are: savings .RE -sched_mc_power_savings is dependent upon SCHED_MC, which is -itself architecture dependent. - -sched_smt_power_savings is dependent upon SCHED_SMT, which -is itself architecture dependent. - -The two files are independent of each other. It is possible -that one file may be present without the other. - .SH "SEE ALSO" cpupower-info(1), cpupower-monitor(1), powertop(1) .PP diff --git a/tools/power/cpupower/utils/cpuidle-info.c b/tools/power/cpupower/utils/cpuidle-info.c index b028267c1376..8145af5f93a6 100644 --- a/tools/power/cpupower/utils/cpuidle-info.c +++ b/tools/power/cpupower/utils/cpuidle-info.c @@ -35,17 +35,9 @@ static void cpuidle_cpu_output(unsigned int cpu, int verbose) printf(_("CPU %u: Can't read idle state info\n"), cpu); return; } - tmp = sysfs_get_idlestate_name(cpu, idlestates - 1); - if (!tmp) { - printf(_("Could not determine max idle state %u\n"), - idlestates - 1); - return; - } - printf(_("Number of idle states: %d\n"), idlestates); - printf(_("Available idle states:")); - for (idlestate = 1; idlestate < idlestates; idlestate++) { + for (idlestate = 0; idlestate < idlestates; idlestate++) { tmp = sysfs_get_idlestate_name(cpu, idlestate); if (!tmp) continue; @@ -57,7 +49,7 @@ static void cpuidle_cpu_output(unsigned int cpu, int verbose) if (!verbose) return; - for (idlestate = 1; idlestate < idlestates; idlestate++) { + for (idlestate = 0; idlestate < idlestates; idlestate++) { tmp = sysfs_get_idlestate_name(cpu, idlestate); if (!tmp) continue; diff --git a/tools/power/cpupower/utils/helpers/amd.c b/tools/power/cpupower/utils/helpers/amd.c index 87d5605bdda8..6437ef39aeea 100644 --- a/tools/power/cpupower/utils/helpers/amd.c +++ b/tools/power/cpupower/utils/helpers/amd.c @@ -112,14 +112,12 @@ int decode_pstates(unsigned int cpu, unsigned int cpu_family, int amd_pci_get_num_boost_states(int *active, int *states) { struct pci_access *pci_acc; - int vendor_id = 0x1022; - int boost_dev_ids[4] = {0x1204, 0x1604, 0x1704, 0}; struct pci_dev *device; uint8_t val = 0; *active = *states = 0; - device = pci_acc_init(&pci_acc, vendor_id, boost_dev_ids); + device = pci_slot_func_init(&pci_acc, 0x18, 4); if (device == NULL) return -ENODEV; diff --git a/tools/power/cpupower/utils/helpers/helpers.h b/tools/power/cpupower/utils/helpers/helpers.h index 2747e738efb0..2eb584cf2f55 100644 --- a/tools/power/cpupower/utils/helpers/helpers.h +++ b/tools/power/cpupower/utils/helpers/helpers.h @@ -66,8 +66,8 @@ enum cpupower_cpu_vendor {X86_VENDOR_UNKNOWN = 0, X86_VENDOR_INTEL, #define CPUPOWER_CAP_AMD_CBP 0x00000004 #define CPUPOWER_CAP_PERF_BIAS 0x00000008 #define CPUPOWER_CAP_HAS_TURBO_RATIO 0x00000010 -#define CPUPOWER_CAP_IS_SNB 0x00000011 -#define CPUPOWER_CAP_INTEL_IDA 0x00000012 +#define CPUPOWER_CAP_IS_SNB 0x00000020 +#define CPUPOWER_CAP_INTEL_IDA 0x00000040 #define MAX_HW_PSTATES 10 @@ -132,8 +132,11 @@ extern unsigned long long msr_intel_get_turbo_ratio(unsigned int cpu); /* PCI stuff ****************************/ extern int amd_pci_get_num_boost_states(int *active, int *states); -extern struct pci_dev *pci_acc_init(struct pci_access **pacc, int vendor_id, - int *dev_ids); +extern struct pci_dev *pci_acc_init(struct pci_access **pacc, int domain, + int bus, int slot, int func, int vendor, + int dev); +extern struct pci_dev *pci_slot_func_init(struct pci_access **pacc, + int slot, int func); /* PCI stuff ****************************/ diff --git a/tools/power/cpupower/utils/helpers/pci.c b/tools/power/cpupower/utils/helpers/pci.c index cd2eb6fe41c4..9690798e6446 100644 --- a/tools/power/cpupower/utils/helpers/pci.c +++ b/tools/power/cpupower/utils/helpers/pci.c @@ -10,19 +10,24 @@ * **pacc : if a valid pci_dev is returned * *pacc must be passed to pci_acc_cleanup to free it * - * vendor_id : the pci vendor id matching the pci device to access - * dev_ids : device ids matching the pci device to access + * domain: domain + * bus: bus + * slot: slot + * func: func + * vendor: vendor + * device: device + * Pass -1 for one of the six above to match any * * Returns : * struct pci_dev which can be used with pci_{read,write}_* functions * to access the PCI config space of matching pci devices */ -struct pci_dev *pci_acc_init(struct pci_access **pacc, int vendor_id, - int *dev_ids) +struct pci_dev *pci_acc_init(struct pci_access **pacc, int domain, int bus, + int slot, int func, int vendor, int dev) { - struct pci_filter filter_nb_link = { -1, -1, -1, -1, vendor_id, 0}; + struct pci_filter filter_nb_link = { domain, bus, slot, func, + vendor, dev }; struct pci_dev *device; - unsigned int i; *pacc = pci_alloc(); if (*pacc == NULL) @@ -31,14 +36,20 @@ struct pci_dev *pci_acc_init(struct pci_access **pacc, int vendor_id, pci_init(*pacc); pci_scan_bus(*pacc); - for (i = 0; dev_ids[i] != 0; i++) { - filter_nb_link.device = dev_ids[i]; - for (device = (*pacc)->devices; device; device = device->next) { - if (pci_filter_match(&filter_nb_link, device)) - return device; - } + for (device = (*pacc)->devices; device; device = device->next) { + if (pci_filter_match(&filter_nb_link, device)) + return device; } pci_cleanup(*pacc); return NULL; } + +/* Typically one wants to get a specific slot(device)/func of the root domain + and bus */ +struct pci_dev *pci_slot_func_init(struct pci_access **pacc, int slot, + int func) +{ + return pci_acc_init(pacc, 0, 0, slot, func, -1, -1); +} + #endif /* defined(__i386__) || defined(__x86_64__) */ diff --git a/tools/power/cpupower/utils/helpers/sysfs.c b/tools/power/cpupower/utils/helpers/sysfs.c index c6343024a611..96e28c124b5c 100644 --- a/tools/power/cpupower/utils/helpers/sysfs.c +++ b/tools/power/cpupower/utils/helpers/sysfs.c @@ -362,22 +362,7 @@ char *sysfs_get_cpuidle_driver(void) */ int sysfs_get_sched(const char *smt_mc) { - unsigned long value; - char linebuf[MAX_LINE_LEN]; - char *endp; - char path[SYSFS_PATH_MAX]; - - if (strcmp("mc", smt_mc) && strcmp("smt", smt_mc)) - return -EINVAL; - - snprintf(path, sizeof(path), - PATH_TO_CPU "sched_%s_power_savings", smt_mc); - if (sysfs_read_file(path, linebuf, MAX_LINE_LEN) == 0) - return -1; - value = strtoul(linebuf, &endp, 0); - if (endp == linebuf || errno == ERANGE) - return -1; - return value; + return -ENODEV; } /* @@ -388,21 +373,5 @@ int sysfs_get_sched(const char *smt_mc) */ int sysfs_set_sched(const char *smt_mc, int val) { - char linebuf[MAX_LINE_LEN]; - char path[SYSFS_PATH_MAX]; - struct stat statbuf; - - if (strcmp("mc", smt_mc) && strcmp("smt", smt_mc)) - return -EINVAL; - - snprintf(path, sizeof(path), - PATH_TO_CPU "sched_%s_power_savings", smt_mc); - sprintf(linebuf, "%d", val); - - if (stat(path, &statbuf) != 0) - return -ENODEV; - - if (sysfs_write_file(path, linebuf, MAX_LINE_LEN) == 0) - return -1; - return 0; + return -ENODEV; } diff --git a/tools/power/cpupower/utils/idle_monitor/amd_fam14h_idle.c b/tools/power/cpupower/utils/idle_monitor/amd_fam14h_idle.c index 202e555988be..2116df9ad832 100644 --- a/tools/power/cpupower/utils/idle_monitor/amd_fam14h_idle.c +++ b/tools/power/cpupower/utils/idle_monitor/amd_fam14h_idle.c @@ -20,8 +20,6 @@ #include "idle_monitor/cpupower-monitor.h" #include "helpers/helpers.h" -/******** PCI parts could go into own file and get shared ***************/ - #define PCI_NON_PC0_OFFSET 0xb0 #define PCI_PC1_OFFSET 0xb4 #define PCI_PC6_OFFSET 0xb8 @@ -82,10 +80,7 @@ static cstate_t amd_fam14h_cstates[AMD_FAM14H_STATE_NUM] = { }; static struct pci_access *pci_acc; -static int pci_vendor_id = 0x1022; -static int pci_dev_ids[2] = {0x1716, 0}; static struct pci_dev *amd_fam14h_pci_dev; - static int nbp1_entered; struct timespec start_time; @@ -286,13 +281,13 @@ struct cpuidle_monitor *amd_fam14h_register(void) if (cpupower_cpu_info.vendor != X86_VENDOR_AMD) return NULL; - if (cpupower_cpu_info.family == 0x14) { - if (cpu_count <= 0 || cpu_count > 2) { - fprintf(stderr, "AMD fam14h: Invalid cpu count: %d\n", - cpu_count); - return NULL; - } - } else + if (cpupower_cpu_info.family == 0x14) + strncpy(amd_fam14h_monitor.name, "Fam_14h", + MONITOR_NAME_LEN - 1); + else if (cpupower_cpu_info.family == 0x12) + strncpy(amd_fam14h_monitor.name, "Fam_12h", + MONITOR_NAME_LEN - 1); + else return NULL; /* We do not alloc for nbp1 machine wide counter */ @@ -303,7 +298,9 @@ struct cpuidle_monitor *amd_fam14h_register(void) sizeof(unsigned long long)); } - amd_fam14h_pci_dev = pci_acc_init(&pci_acc, pci_vendor_id, pci_dev_ids); + /* We need PCI device: Slot 18, Func 6, compare with BKDG + for fam 12h/14h */ + amd_fam14h_pci_dev = pci_slot_func_init(&pci_acc, 0x18, 6); if (amd_fam14h_pci_dev == NULL || pci_acc == NULL) return NULL; @@ -325,7 +322,7 @@ static void amd_fam14h_unregister(void) } struct cpuidle_monitor amd_fam14h_monitor = { - .name = "Ontario", + .name = "", .hw_states = amd_fam14h_cstates, .hw_states_num = AMD_FAM14H_STATE_NUM, .start = amd_fam14h_start, diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index 555c69a5592a..adf175f61496 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -4,11 +4,13 @@ turbostat \- Report processor frequency and idle statistics .SH SYNOPSIS .ft B .B turbostat +.RB [ "\-s" ] .RB [ "\-v" ] .RB [ "\-M MSR#" ] .RB command .br .B turbostat +.RB [ "\-s" ] .RB [ "\-v" ] .RB [ "\-M MSR#" ] .RB [ "\-i interval_sec" ] @@ -25,6 +27,8 @@ supports an "invariant" TSC, plus the APERF and MPERF MSRs. on processors that additionally support C-state residency counters. .SS Options +The \fB-s\fP option prints only a 1-line summary for each sample interval. +.PP The \fB-v\fP option increases verbosity. .PP The \fB-M MSR#\fP option dumps the specified MSR, @@ -39,13 +43,14 @@ displays the statistics gathered since it was forked. .SH FIELD DESCRIPTIONS .nf \fBpk\fP processor package number. -\fBcr\fP processor core number. +\fBcor\fP processor core number. \fBCPU\fP Linux CPU (logical processor) number. +Note that multiple CPUs per core indicate support for Intel(R) Hyper-Threading Technology. \fB%c0\fP percent of the interval that the CPU retired instructions. \fBGHz\fP average clock rate while the CPU was in c0 state. \fBTSC\fP average GHz that the TSC ran during the entire interval. -\fB%c1, %c3, %c6\fP show the percentage residency in hardware core idle states. -\fB%pc3, %pc6\fP percentage residency in hardware package idle states. +\fB%c1, %c3, %c6, %c7\fP show the percentage residency in hardware core idle states. +\fB%pc2, %pc3, %pc6, %pc7\fP percentage residency in hardware package idle states. .fi .PP .SH EXAMPLE @@ -53,25 +58,37 @@ Without any parameters, turbostat prints out counters ever 5 seconds. (override interval with "-i sec" option, or specify a command for turbostat to fork). -The first row of statistics reflect the average for the entire system. +The first row of statistics is a summary for the entire system. +Note that the summary is a weighted average. Subsequent rows show per-CPU statistics. .nf [root@x980]# ./turbostat -cr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 - 0.04 1.62 3.38 0.11 0.00 99.85 0.00 95.07 - 0 0 0.04 1.62 3.38 0.06 0.00 99.90 0.00 95.07 - 0 6 0.02 1.62 3.38 0.08 0.00 99.90 0.00 95.07 - 1 2 0.10 1.62 3.38 0.29 0.00 99.61 0.00 95.07 - 1 8 0.11 1.62 3.38 0.28 0.00 99.61 0.00 95.07 - 2 4 0.01 1.62 3.38 0.01 0.00 99.98 0.00 95.07 - 2 10 0.01 1.61 3.38 0.02 0.00 99.98 0.00 95.07 - 8 1 0.07 1.62 3.38 0.15 0.00 99.78 0.00 95.07 - 8 7 0.03 1.62 3.38 0.19 0.00 99.78 0.00 95.07 - 9 3 0.01 1.62 3.38 0.02 0.00 99.98 0.00 95.07 - 9 9 0.01 1.62 3.38 0.02 0.00 99.98 0.00 95.07 - 10 5 0.01 1.62 3.38 0.13 0.00 99.86 0.00 95.07 - 10 11 0.08 1.62 3.38 0.05 0.00 99.86 0.00 95.07 +cor CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 + 0.60 1.63 3.38 2.91 0.00 96.49 0.00 76.64 + 0 0 0.59 1.62 3.38 4.51 0.00 94.90 0.00 76.64 + 0 6 1.13 1.64 3.38 3.97 0.00 94.90 0.00 76.64 + 1 2 0.08 1.62 3.38 0.07 0.00 99.85 0.00 76.64 + 1 8 0.03 1.62 3.38 0.12 0.00 99.85 0.00 76.64 + 2 4 0.01 1.62 3.38 0.06 0.00 99.93 0.00 76.64 + 2 10 0.04 1.62 3.38 0.02 0.00 99.93 0.00 76.64 + 8 1 2.85 1.62 3.38 11.71 0.00 85.44 0.00 76.64 + 8 7 1.98 1.62 3.38 12.58 0.00 85.44 0.00 76.64 + 9 3 0.36 1.62 3.38 0.71 0.00 98.93 0.00 76.64 + 9 9 0.09 1.62 3.38 0.98 0.00 98.93 0.00 76.64 + 10 5 0.03 1.62 3.38 0.09 0.00 99.87 0.00 76.64 + 10 11 0.07 1.62 3.38 0.06 0.00 99.87 0.00 76.64 +.fi +.SH SUMMARY EXAMPLE +The "-s" option prints the column headers just once, +and then the one line system summary for each sample interval. + +.nf +[root@x980]# ./turbostat -s + %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 + 0.61 1.89 3.38 5.95 0.00 93.44 0.00 66.33 + 0.52 1.62 3.38 6.83 0.00 92.65 0.00 61.11 + 0.62 1.92 3.38 5.47 0.00 93.91 0.00 67.31 .fi .SH VERBOSE EXAMPLE The "-v" option adds verbosity to the output: @@ -101,33 +118,33 @@ until ^C while the other CPUs are mostly idle: .nf [root@x980 lenb]# ./turbostat cat /dev/zero > /dev/null - -^Ccr CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 - 8.49 3.63 3.38 16.23 0.66 74.63 0.00 0.00 - 0 0 1.22 3.62 3.38 32.18 0.00 66.60 0.00 0.00 - 0 6 0.40 3.61 3.38 33.00 0.00 66.60 0.00 0.00 - 1 2 0.11 3.14 3.38 0.19 3.95 95.75 0.00 0.00 - 1 8 0.05 2.88 3.38 0.25 3.95 95.75 0.00 0.00 - 2 4 0.00 3.13 3.38 0.02 0.00 99.98 0.00 0.00 - 2 10 0.00 3.09 3.38 0.02 0.00 99.98 0.00 0.00 - 8 1 0.04 3.50 3.38 14.43 0.00 85.54 0.00 0.00 - 8 7 0.03 2.98 3.38 14.43 0.00 85.54 0.00 0.00 - 9 3 0.00 3.16 3.38 100.00 0.00 0.00 0.00 0.00 - 9 9 99.93 3.63 3.38 0.06 0.00 0.00 0.00 0.00 - 10 5 0.01 2.82 3.38 0.08 0.00 99.91 0.00 0.00 - 10 11 0.02 3.36 3.38 0.06 0.00 99.91 0.00 0.00 -6.950866 sec +^C +cor CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 + 8.63 3.64 3.38 14.46 0.49 76.42 0.00 0.00 + 0 0 0.34 3.36 3.38 99.66 0.00 0.00 0.00 0.00 + 0 6 99.96 3.64 3.38 0.04 0.00 0.00 0.00 0.00 + 1 2 0.14 3.50 3.38 1.75 2.04 96.07 0.00 0.00 + 1 8 0.38 3.57 3.38 1.51 2.04 96.07 0.00 0.00 + 2 4 0.01 2.65 3.38 0.06 0.00 99.93 0.00 0.00 + 2 10 0.03 2.12 3.38 0.04 0.00 99.93 0.00 0.00 + 8 1 0.91 3.59 3.38 35.27 0.92 62.90 0.00 0.00 + 8 7 1.61 3.63 3.38 34.57 0.92 62.90 0.00 0.00 + 9 3 0.04 3.38 3.38 0.20 0.00 99.76 0.00 0.00 + 9 9 0.04 3.29 3.38 0.20 0.00 99.76 0.00 0.00 + 10 5 0.03 3.08 3.38 0.12 0.00 99.85 0.00 0.00 + 10 11 0.05 3.07 3.38 0.10 0.00 99.85 0.00 0.00 +4.907015 sec .fi -Above the cycle soaker drives cpu9 up 3.6 Ghz turbo limit +Above the cycle soaker drives cpu6 up 3.6 Ghz turbo limit while the other processors are generally in various states of idle. -Note that cpu3 is an HT sibling sharing core9 -with cpu9, and thus it is unable to get to an idle state -deeper than c1 while cpu9 is busy. +Note that cpu0 is an HT sibling sharing core0 +with cpu6, and thus it is unable to get to an idle state +deeper than c1 while cpu6 is busy. -Note that turbostat reports average GHz of 3.61, while -the arithmetic average of the GHz column above is 3.24. +Note that turbostat reports average GHz of 3.64, while +the arithmetic average of the GHz column above is lower. This is a weighted average, where the weight is %c0. ie. it is the total number of un-halted cycles elapsed per time divided by the number of CPUs. .SH NOTES @@ -167,6 +184,6 @@ http://www.intel.com/products/processor/manuals/ .SH "SEE ALSO" msr(4), vmstat(8) .PP -.SH AUTHORS +.SH AUTHOR .nf Written by Len Brown <len.brown@intel.com> diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 310d3dd5e547..16de7ad4850f 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -2,7 +2,7 @@ * turbostat -- show CPU frequency and C-state residency * on modern Intel turbo-capable processors. * - * Copyright (c) 2010, Intel Corporation. + * Copyright (c) 2012 Intel Corporation. * Len Brown <len.brown@intel.com> * * This program is free software; you can redistribute it and/or modify it @@ -19,6 +19,7 @@ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. */ +#define _GNU_SOURCE #include <stdio.h> #include <unistd.h> #include <sys/types.h> @@ -32,6 +33,7 @@ #include <dirent.h> #include <string.h> #include <ctype.h> +#include <sched.h> #define MSR_TSC 0x10 #define MSR_NEHALEM_PLATFORM_INFO 0xCE @@ -49,6 +51,7 @@ char *proc_stat = "/proc/stat"; unsigned int interval_sec = 5; /* set with -i interval_sec */ unsigned int verbose; /* set with -v */ +unsigned int summary_only; /* set with -s */ unsigned int skip_c0; unsigned int skip_c1; unsigned int do_nhm_cstates; @@ -68,9 +71,10 @@ unsigned int show_cpu; int aperf_mperf_unstable; int backwards_count; char *progname; -int need_reinitialize; int num_cpus; +cpu_set_t *cpu_present_set, *cpu_mask; +size_t cpu_present_setsize, cpu_mask_size; struct counters { unsigned long long tsc; /* per thread */ @@ -99,44 +103,97 @@ struct timeval tv_even; struct timeval tv_odd; struct timeval tv_delta; -unsigned long long get_msr(int cpu, off_t offset) +int mark_cpu_present(int pkg, int core, int cpu) +{ + CPU_SET_S(cpu, cpu_present_setsize, cpu_present_set); + return 0; +} + +/* + * cpu_mask_init(ncpus) + * + * allocate and clear cpu_mask + * set cpu_mask_size + */ +void cpu_mask_init(int ncpus) +{ + cpu_mask = CPU_ALLOC(ncpus); + if (cpu_mask == NULL) { + perror("CPU_ALLOC"); + exit(3); + } + cpu_mask_size = CPU_ALLOC_SIZE(ncpus); + CPU_ZERO_S(cpu_mask_size, cpu_mask); + + /* + * Allocate and initialize cpu_present_set + */ + cpu_present_set = CPU_ALLOC(ncpus); + if (cpu_present_set == NULL) { + perror("CPU_ALLOC"); + exit(3); + } + cpu_present_setsize = CPU_ALLOC_SIZE(ncpus); + CPU_ZERO_S(cpu_present_setsize, cpu_present_set); + for_all_cpus(mark_cpu_present); +} + +void cpu_mask_uninit() +{ + CPU_FREE(cpu_mask); + cpu_mask = NULL; + cpu_mask_size = 0; + CPU_FREE(cpu_present_set); + cpu_present_set = NULL; + cpu_present_setsize = 0; +} + +int cpu_migrate(int cpu) +{ + CPU_ZERO_S(cpu_mask_size, cpu_mask); + CPU_SET_S(cpu, cpu_mask_size, cpu_mask); + if (sched_setaffinity(0, cpu_mask_size, cpu_mask) == -1) + return -1; + else + return 0; +} + +int get_msr(int cpu, off_t offset, unsigned long long *msr) { ssize_t retval; - unsigned long long msr; char pathname[32]; int fd; sprintf(pathname, "/dev/cpu/%d/msr", cpu); fd = open(pathname, O_RDONLY); - if (fd < 0) { - perror(pathname); - need_reinitialize = 1; - return 0; - } - - retval = pread(fd, &msr, sizeof msr, offset); - if (retval != sizeof msr) { - fprintf(stderr, "cpu%d pread(..., 0x%zx) = %jd\n", - cpu, offset, retval); - exit(-2); - } + if (fd < 0) + return -1; + retval = pread(fd, msr, sizeof *msr, offset); close(fd); - return msr; + + if (retval != sizeof *msr) + return -1; + + return 0; } void print_header(void) { if (show_pkg) fprintf(stderr, "pk"); + if (show_pkg) + fprintf(stderr, " "); if (show_core) - fprintf(stderr, " cr"); + fprintf(stderr, "cor"); if (show_cpu) fprintf(stderr, " CPU"); + if (show_pkg || show_core || show_cpu) + fprintf(stderr, " "); if (do_nhm_cstates) - fprintf(stderr, " %%c0 "); + fprintf(stderr, " %%c0"); if (has_aperf) - fprintf(stderr, " GHz"); + fprintf(stderr, " GHz"); fprintf(stderr, " TSC"); if (do_nhm_cstates) fprintf(stderr, " %%c1"); @@ -147,13 +204,13 @@ void print_header(void) if (do_snb_cstates) fprintf(stderr, " %%c7"); if (do_snb_cstates) - fprintf(stderr, " %%pc2"); + fprintf(stderr, " %%pc2"); if (do_nhm_cstates) - fprintf(stderr, " %%pc3"); + fprintf(stderr, " %%pc3"); if (do_nhm_cstates) - fprintf(stderr, " %%pc6"); + fprintf(stderr, " %%pc6"); if (do_snb_cstates) - fprintf(stderr, " %%pc7"); + fprintf(stderr, " %%pc7"); if (extra_msr_offset) fprintf(stderr, " MSR 0x%x ", extra_msr_offset); @@ -187,6 +244,15 @@ void dump_list(struct counters *cnt) dump_cnt(cnt); } +/* + * column formatting convention & formats + * package: "pk" 2 columns %2d + * core: "cor" 3 columns %3d + * CPU: "CPU" 3 columns %3d + * GHz: "GHz" 3 columns %3.2 + * TSC: "TSC" 3 columns %3.2 + * percentage " %pc3" %6.2 + */ void print_cnt(struct counters *p) { double interval_float; @@ -196,39 +262,45 @@ void print_cnt(struct counters *p) /* topology columns, print blanks on 1st (average) line */ if (p == cnt_average) { if (show_pkg) + fprintf(stderr, " "); + if (show_pkg && show_core) fprintf(stderr, " "); if (show_core) - fprintf(stderr, " "); + fprintf(stderr, " "); if (show_cpu) - fprintf(stderr, " "); + fprintf(stderr, " " " "); } else { if (show_pkg) - fprintf(stderr, "%d", p->pkg); + fprintf(stderr, "%2d", p->pkg); + if (show_pkg && show_core) + fprintf(stderr, " "); if (show_core) - fprintf(stderr, "%4d", p->core); + fprintf(stderr, "%3d", p->core); if (show_cpu) - fprintf(stderr, "%4d", p->cpu); + fprintf(stderr, " %3d", p->cpu); } /* %c0 */ if (do_nhm_cstates) { + if (show_pkg || show_core || show_cpu) + fprintf(stderr, " "); if (!skip_c0) - fprintf(stderr, "%7.2f", 100.0 * p->mperf/p->tsc); + fprintf(stderr, "%6.2f", 100.0 * p->mperf/p->tsc); else - fprintf(stderr, " ****"); + fprintf(stderr, " ****"); } /* GHz */ if (has_aperf) { if (!aperf_mperf_unstable) { - fprintf(stderr, "%5.2f", + fprintf(stderr, " %3.2f", 1.0 * p->tsc / units * p->aperf / p->mperf / interval_float); } else { if (p->aperf > p->tsc || p->mperf > p->tsc) { - fprintf(stderr, " ****"); + fprintf(stderr, " ***"); } else { - fprintf(stderr, "%4.1f*", + fprintf(stderr, "%3.1f*", 1.0 * p->tsc / units * p->aperf / p->mperf / interval_float); @@ -241,7 +313,7 @@ void print_cnt(struct counters *p) if (do_nhm_cstates) { if (!skip_c1) - fprintf(stderr, "%7.2f", 100.0 * p->c1/p->tsc); + fprintf(stderr, " %6.2f", 100.0 * p->c1/p->tsc); else fprintf(stderr, " ****"); } @@ -252,13 +324,13 @@ void print_cnt(struct counters *p) if (do_snb_cstates) fprintf(stderr, " %6.2f", 100.0 * p->c7/p->tsc); if (do_snb_cstates) - fprintf(stderr, " %5.2f", 100.0 * p->pc2/p->tsc); + fprintf(stderr, " %6.2f", 100.0 * p->pc2/p->tsc); if (do_nhm_cstates) - fprintf(stderr, " %5.2f", 100.0 * p->pc3/p->tsc); + fprintf(stderr, " %6.2f", 100.0 * p->pc3/p->tsc); if (do_nhm_cstates) - fprintf(stderr, " %5.2f", 100.0 * p->pc6/p->tsc); + fprintf(stderr, " %6.2f", 100.0 * p->pc6/p->tsc); if (do_snb_cstates) - fprintf(stderr, " %5.2f", 100.0 * p->pc7/p->tsc); + fprintf(stderr, " %6.2f", 100.0 * p->pc7/p->tsc); if (extra_msr_offset) fprintf(stderr, " 0x%016llx", p->extra_msr); putc('\n', stderr); @@ -267,12 +339,20 @@ void print_cnt(struct counters *p) void print_counters(struct counters *counters) { struct counters *cnt; + static int printed; + - print_header(); + if (!printed || !summary_only) + print_header(); if (num_cpus > 1) print_cnt(cnt_average); + printed = 1; + + if (summary_only) + return; + for (cnt = counters; cnt != NULL; cnt = cnt->next) print_cnt(cnt); @@ -440,31 +520,51 @@ void compute_average(struct counters *delta, struct counters *avg) free(sum); } -void get_counters(struct counters *cnt) +int get_counters(struct counters *cnt) { for ( ; cnt; cnt = cnt->next) { - cnt->tsc = get_msr(cnt->cpu, MSR_TSC); - if (do_nhm_cstates) - cnt->c3 = get_msr(cnt->cpu, MSR_CORE_C3_RESIDENCY); - if (do_nhm_cstates) - cnt->c6 = get_msr(cnt->cpu, MSR_CORE_C6_RESIDENCY); - if (do_snb_cstates) - cnt->c7 = get_msr(cnt->cpu, MSR_CORE_C7_RESIDENCY); - if (has_aperf) - cnt->aperf = get_msr(cnt->cpu, MSR_APERF); - if (has_aperf) - cnt->mperf = get_msr(cnt->cpu, MSR_MPERF); - if (do_snb_cstates) - cnt->pc2 = get_msr(cnt->cpu, MSR_PKG_C2_RESIDENCY); - if (do_nhm_cstates) - cnt->pc3 = get_msr(cnt->cpu, MSR_PKG_C3_RESIDENCY); - if (do_nhm_cstates) - cnt->pc6 = get_msr(cnt->cpu, MSR_PKG_C6_RESIDENCY); + + if (cpu_migrate(cnt->cpu)) + return -1; + + if (get_msr(cnt->cpu, MSR_TSC, &cnt->tsc)) + return -1; + + if (has_aperf) { + if (get_msr(cnt->cpu, MSR_APERF, &cnt->aperf)) + return -1; + if (get_msr(cnt->cpu, MSR_MPERF, &cnt->mperf)) + return -1; + } + + if (do_nhm_cstates) { + if (get_msr(cnt->cpu, MSR_CORE_C3_RESIDENCY, &cnt->c3)) + return -1; + if (get_msr(cnt->cpu, MSR_CORE_C6_RESIDENCY, &cnt->c6)) + return -1; + } + if (do_snb_cstates) - cnt->pc7 = get_msr(cnt->cpu, MSR_PKG_C7_RESIDENCY); + if (get_msr(cnt->cpu, MSR_CORE_C7_RESIDENCY, &cnt->c7)) + return -1; + + if (do_nhm_cstates) { + if (get_msr(cnt->cpu, MSR_PKG_C3_RESIDENCY, &cnt->pc3)) + return -1; + if (get_msr(cnt->cpu, MSR_PKG_C6_RESIDENCY, &cnt->pc6)) + return -1; + } + if (do_snb_cstates) { + if (get_msr(cnt->cpu, MSR_PKG_C2_RESIDENCY, &cnt->pc2)) + return -1; + if (get_msr(cnt->cpu, MSR_PKG_C7_RESIDENCY, &cnt->pc7)) + return -1; + } if (extra_msr_offset) - cnt->extra_msr = get_msr(cnt->cpu, extra_msr_offset); + if (get_msr(cnt->cpu, extra_msr_offset, &cnt->extra_msr)) + return -1; } + return 0; } void print_nehalem_info(void) @@ -475,7 +575,7 @@ void print_nehalem_info(void) if (!do_nehalem_platform_info) return; - msr = get_msr(0, MSR_NEHALEM_PLATFORM_INFO); + get_msr(0, MSR_NEHALEM_PLATFORM_INFO, &msr); ratio = (msr >> 40) & 0xFF; fprintf(stderr, "%d * %.0f = %.0f MHz max efficiency\n", @@ -491,7 +591,7 @@ void print_nehalem_info(void) if (!do_nehalem_turbo_ratio_limit) return; - msr = get_msr(0, MSR_NEHALEM_TURBO_RATIO_LIMIT); + get_msr(0, MSR_NEHALEM_TURBO_RATIO_LIMIT, &msr); ratio = (msr >> 24) & 0xFF; if (ratio) @@ -557,7 +657,8 @@ void insert_counters(struct counters **list, return; } - show_cpu = 1; /* there is more than one CPU */ + if (!summary_only) + show_cpu = 1; /* there is more than one CPU */ /* * insert on front of list. @@ -575,13 +676,15 @@ void insert_counters(struct counters **list, while (prev->next && (prev->next->pkg < new->pkg)) { prev = prev->next; - show_pkg = 1; /* there is more than 1 package */ + if (!summary_only) + show_pkg = 1; /* there is more than 1 package */ } while (prev->next && (prev->next->pkg == new->pkg) && (prev->next->core < new->core)) { prev = prev->next; - show_core = 1; /* there is more than 1 core */ + if (!summary_only) + show_core = 1; /* there is more than 1 core */ } while (prev->next && (prev->next->pkg == new->pkg) @@ -681,7 +784,7 @@ int get_core_id(int cpu) } /* - * run func(index, cpu) on every cpu in /proc/stat + * run func(pkg, core, cpu) on every cpu in /proc/stat */ int for_all_cpus(void (func)(int, int, int)) @@ -717,18 +820,18 @@ int for_all_cpus(void (func)(int, int, int)) void re_initialize(void) { - printf("turbostat: topology changed, re-initializing.\n"); free_all_counters(); num_cpus = for_all_cpus(alloc_new_counters); - need_reinitialize = 0; - printf("num_cpus is now %d\n", num_cpus); + cpu_mask_uninit(); + cpu_mask_init(num_cpus); + printf("turbostat: re-initialized with num_cpus %d\n", num_cpus); } void dummy(int pkg, int core, int cpu) { return; } /* * check to see if a cpu came on-line */ -void verify_num_cpus(void) +int verify_num_cpus(void) { int new_num_cpus; @@ -738,8 +841,9 @@ void verify_num_cpus(void) if (verbose) printf("num_cpus was %d, is now %d\n", num_cpus, new_num_cpus); - need_reinitialize = 1; + return -1; } + return 0; } void turbostat_loop() @@ -749,25 +853,25 @@ restart: gettimeofday(&tv_even, (struct timezone *)NULL); while (1) { - verify_num_cpus(); - if (need_reinitialize) { + if (verify_num_cpus()) { re_initialize(); goto restart; } sleep(interval_sec); - get_counters(cnt_odd); + if (get_counters(cnt_odd)) { + re_initialize(); + goto restart; + } gettimeofday(&tv_odd, (struct timezone *)NULL); - compute_delta(cnt_odd, cnt_even, cnt_delta); timersub(&tv_odd, &tv_even, &tv_delta); compute_average(cnt_delta, cnt_average); print_counters(cnt_delta); - if (need_reinitialize) { + sleep(interval_sec); + if (get_counters(cnt_even)) { re_initialize(); goto restart; } - sleep(interval_sec); - get_counters(cnt_even); gettimeofday(&tv_even, (struct timezone *)NULL); compute_delta(cnt_even, cnt_odd, cnt_delta); timersub(&tv_even, &tv_odd, &tv_delta); @@ -829,6 +933,8 @@ int is_snb(unsigned int family, unsigned int model) switch (model) { case 0x2A: case 0x2D: + case 0x3A: /* IVB */ + case 0x3D: /* IVB Xeon */ return 1; } return 0; @@ -953,6 +1059,7 @@ void turbostat_init() check_super_user(); num_cpus = for_all_cpus(alloc_new_counters); + cpu_mask_init(num_cpus); if (verbose) print_nehalem_info(); @@ -963,6 +1070,9 @@ int fork_it(char **argv) int retval; pid_t child_pid; get_counters(cnt_even); + + /* clear affinity side-effect of get_counters() */ + sched_setaffinity(0, cpu_present_setsize, cpu_present_set); gettimeofday(&tv_even, (struct timezone *)NULL); child_pid = fork(); @@ -1005,8 +1115,11 @@ void cmdline(int argc, char **argv) progname = argv[0]; - while ((opt = getopt(argc, argv, "+vi:M:")) != -1) { + while ((opt = getopt(argc, argv, "+svi:M:")) != -1) { switch (opt) { + case 's': + summary_only++; + break; case 'v': verbose++; break; diff --git a/tools/scripts/Makefile.include b/tools/scripts/Makefile.include new file mode 100644 index 000000000000..bde8521d56bb --- /dev/null +++ b/tools/scripts/Makefile.include @@ -0,0 +1,58 @@ +ifeq ("$(origin O)", "command line") + OUTPUT := $(O)/ + COMMAND_O := O=$(O) +endif + +ifneq ($(OUTPUT),) +# check that the output directory actually exists +OUTDIR := $(shell cd $(OUTPUT) && /bin/pwd) +$(if $(OUTDIR),, $(error output directory "$(OUTPUT)" does not exist)) +endif + +# +# Include saner warnings here, which can catch bugs: +# +EXTRA_WARNINGS := -Wbad-function-cast +EXTRA_WARNINGS += -Wdeclaration-after-statement +EXTRA_WARNINGS += -Wformat-security +EXTRA_WARNINGS += -Wformat-y2k +EXTRA_WARNINGS += -Winit-self +EXTRA_WARNINGS += -Wmissing-declarations +EXTRA_WARNINGS += -Wmissing-prototypes +EXTRA_WARNINGS += -Wnested-externs +EXTRA_WARNINGS += -Wno-system-headers +EXTRA_WARNINGS += -Wold-style-definition +EXTRA_WARNINGS += -Wpacked +EXTRA_WARNINGS += -Wredundant-decls +EXTRA_WARNINGS += -Wshadow +EXTRA_WARNINGS += -Wstrict-aliasing=3 +EXTRA_WARNINGS += -Wstrict-prototypes +EXTRA_WARNINGS += -Wswitch-default +EXTRA_WARNINGS += -Wswitch-enum +EXTRA_WARNINGS += -Wundef +EXTRA_WARNINGS += -Wwrite-strings +EXTRA_WARNINGS += -Wformat + +ifneq ($(findstring $(MAKEFLAGS), w),w) +PRINT_DIR = --no-print-directory +else +NO_SUBDIR = : +endif + +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),s),s) +ifndef V + QUIET_CC = @echo ' ' CC $@; + QUIET_AR = @echo ' ' AR $@; + QUIET_LINK = @echo ' ' LINK $@; + QUIET_MKDIR = @echo ' ' MKDIR $@; + QUIET_GEN = @echo ' ' GEN $@; + QUIET_SUBDIR0 = +@subdir= + QUIET_SUBDIR1 = ;$(NO_SUBDIR) echo ' ' SUBDIR $$subdir; \ + $(MAKE) $(PRINT_DIR) -C $$subdir + QUIET_FLEX = @echo ' ' FLEX $@; + QUIET_BISON = @echo ' ' BISON $@; +endif +endif diff --git a/tools/testing/ktest/examples/README b/tools/testing/ktest/examples/README new file mode 100644 index 000000000000..a12d295a09d8 --- /dev/null +++ b/tools/testing/ktest/examples/README @@ -0,0 +1,32 @@ +This directory contains example configs to use ktest for various tasks. +The configs still need to be customized for your environment, but it +is broken up by task which makes it easier to understand how to set up +ktest. + +The configs are based off of real working configs but have been modified +and commented to show more generic use cases that are more helpful for +developers. + +crosstests.conf - this config shows an example of testing a git repo against + lots of different architectures. It only does build tests, but makes + it easy to compile test different archs. You can download the arch + cross compilers from: + http://kernel.org/pub/tools/crosstool/files/bin/x86_64/ + +test.conf - A generic example of a config. This is based on an actual config + used to perform real testing. + +kvm.conf - A example of a config that is used to test a virtual guest running + on a host. + +snowball.conf - An example config that was used to demo ktest.pl against + a snowball ARM board. + +include/ - The include directory holds default configs that can be + included into other configs. This is a real use example that shows how + to reuse configs for various machines or set ups. The files here + are included by other config files, where the other config files define + options and variables that will make the included config work for the + given environment. + + diff --git a/tools/testing/ktest/examples/crosstests.conf b/tools/testing/ktest/examples/crosstests.conf new file mode 100644 index 000000000000..46736604c26c --- /dev/null +++ b/tools/testing/ktest/examples/crosstests.conf @@ -0,0 +1,260 @@ +# +# Example config for cross compiling +# +# In this config, it is expected that the tool chains from: +# +# http://kernel.org/pub/tools/crosstool/files/bin/x86_64/ +# +# running on a x86_64 system have been downloaded and installed into: +# +# /usr/local/ +# +# such that the compiler binaries are something like: +# +# /usr/local/gcc-4.5.2-nolibc/mips-linux/bin/mips-linux-gcc +# +# Some of the archs will use gcc-4.5.1 instead of gcc-4.5.2 +# this config uses variables to differentiate them. +# +# Comments describe some of the options, but full descriptions of +# options are described in the samples.conf file. + +# ${PWD} is defined by ktest.pl to be the directory that the user +# was in when they executed ktest.pl. It may be better to hardcode the +# path name here. THIS_DIR is the variable used through out the config file +# in case you want to change it. + +THIS_DIR := ${PWD} + +# Update the BUILD_DIR option to the location of your git repo you want to test. +BUILD_DIR = ${THIS_DIR}/linux.git + +# The build will go into this directory. It will be created when you run the test. +OUTPUT_DIR = ${THIS_DIR}/cross-compile + +# The build will be compiled with -j8 +BUILD_OPTIONS = -j8 + +# The test will not stop when it hits a failure. +DIE_ON_FAILURE = 0 + +# If you want to have ktest.pl store the failure somewhere, uncomment this option +# and change the directory where ktest should store the failures. +#STORE_FAILURES = ${THIS_DIR}/failures + +# The log file is stored in the OUTPUT_DIR called cross.log +# If you enable this, you need to create the OUTPUT_DIR. It wont be created for you. +LOG_FILE = ${OUTPUT_DIR}/cross.log + +# The log file will be cleared each time you run ktest. +CLEAR_LOG = 1 + +# As some archs do not build with the defconfig, they have been marked +# to be ignored. If you want to test them anyway, change DO_FAILED to 1. +# If a test that has been marked as DO_FAILED passes, then you should change +# that test to be DO_DEFAULT + +DO_FAILED := 0 +DO_DEFAULT := 1 + +# By setting both DO_FAILED and DO_DEFAULT to zero, you can pick a single +# arch that you want to test. (uncomment RUN and chose your arch) +#RUN := m32r + +# At the bottom of the config file exists a bisect test. You can update that +# test and set DO_FAILED and DO_DEFAULT to zero, and uncomment this variable +# to run the bisect on the arch. +#RUN := bisect + +# By default all tests will be running gcc 4.5.2. Some tests are using 4.5.1 +# and they select that in the test. +# Note: GCC_VER is declared as on option and not a variable ('=' instead of ':=') +# This is important. A variable is used only in the config file and if it is set +# it stays that way for the rest of the config file until it is change again. +# Here we want GCC_VER to remain persistent and change for each test, as it is used in +# the MAKE_CMD. By using '=' instead of ':=' we achieve our goal. + +GCC_VER = 4.5.2 +MAKE_CMD = PATH=/usr/local/gcc-${GCC_VER}-nolibc/${CROSS}/bin:$PATH CROSS_COMPILE=${CROSS}- make ARCH=${ARCH} + +# all tests are only doing builds. +TEST_TYPE = build + +# If you want to add configs on top of the defconfig, you can add those configs into +# the add-config file and uncomment this option. This is useful if you want to test +# all cross compiles with PREEMPT set, or TRACING on, etc. +#ADD_CONFIG = ${THIS_DIR}/add-config + +# All tests are using defconfig +BUILD_TYPE = defconfig + +# The test names will have the arch and cross compiler used. This will be shown in +# the results. +TEST_NAME = ${ARCH} ${CROSS} + +# alpha +TEST_START IF ${RUN} == alpha || ${DO_DEFAULT} +# Notice that CROSS and ARCH are also options and not variables (again '=' instead +# of ':='). This is because TEST_NAME and MAKE_CMD wil use them for each test. +# Only options are available during runs. Variables are only present in parsing the +# config file. +CROSS = alpha-linux +ARCH = alpha + +# arm +TEST_START IF ${RUN} == arm || ${DO_DEFAULT} +CROSS = arm-unknown-linux-gnueabi +ARCH = arm + +# black fin +TEST_START IF ${RUN} == bfin || ${DO_DEFAULT} +CROSS = bfin-uclinux +ARCH = blackfin +BUILD_OPTIONS = -j8 vmlinux + +# cris - FAILS? +TEST_START IF ${RUN} == cris || ${RUN} == cris64 || ${DO_FAILED} +CROSS = cris-linux +ARCH = cris + +# cris32 - not right arch? +TEST_START IF ${RUN} == cris || ${RUN} == cris32 || ${DO_FAILED} +CROSS = crisv32-linux +ARCH = cris + +# ia64 +TEST_START IF ${RUN} == ia64 || ${DO_DEFAULT} +CROSS = ia64-linux +ARCH = ia64 + +# frv +TEST_START IF ${RUN} == frv || ${DO_FAILED} +CROSS = frv-linux +ARCH = frv +GCC_VER = 4.5.1 + +# h8300 - failed make defconfig?? +TEST_START IF ${RUN} == h8300 || ${DO_FAILED} +CROSS = h8300-elf +ARCH = h8300 +GCC_VER = 4.5.1 + +# m68k fails with error? +TEST_START IF ${RUN} == m68k || ${DO_DEFAULT} +CROSS = m68k-linux +ARCH = m68k + +# mips64 +TEST_START IF ${RUN} == mips || ${RUN} == mips64 || ${DO_DEFAULT} +CROSS = mips64-linux +ARCH = mips + +# mips32 +TEST_START IF ${RUN} == mips || ${RUN} == mips32 || ${DO_DEFAULT} +CROSS = mips-linux +ARCH = mips + +# m32r +TEST_START IF ${RUN} == m32r || ${DO_FAILED} +CROSS = m32r-linux +ARCH = m32r +GCC_VER = 4.5.1 +BUILD_OPTIONS = -j8 vmlinux + +# parisc64 failed? +TEST_START IF ${RUN} == hppa || ${RUN} == hppa64 || ${DO_FAILED} +CROSS = hppa64-linux +ARCH = parisc + +# parisc +TEST_START IF ${RUN} == hppa || ${RUN} == hppa32 || ${DO_FAILED} +CROSS = hppa-linux +ARCH = parisc + +# ppc +TEST_START IF ${RUN} == ppc || ${RUN} == ppc32 || ${DO_DEFAULT} +CROSS = powerpc-linux +ARCH = powerpc + +# ppc64 +TEST_START IF ${RUN} == ppc || ${RUN} == ppc64 || ${DO_DEFAULT} +CROSS = powerpc64-linux +ARCH = powerpc + +# s390 +TEST_START IF ${RUN} == s390 || ${DO_DEFAULT} +CROSS = s390x-linux +ARCH = s390 + +# sh +TEST_START IF ${RUN} == sh || ${DO_DEFAULT} +CROSS = sh4-linux +ARCH = sh + +# sparc64 +TEST_START IF ${RUN} == sparc || ${RUN} == sparc64 || ${DO_DEFAULT} +CROSS = sparc64-linux +ARCH = sparc64 + +# sparc +TEST_START IF ${RUN} == sparc || ${RUN} == sparc32 || ${DO_DEFAULT} +CROSS = sparc-linux +ARCH = sparc + +# xtensa failed +TEST_START IF ${RUN} == xtensa || ${DO_FAILED} +CROSS = xtensa-linux +ARCH = xtensa + +# UML +TEST_START IF ${RUN} == uml || ${DO_DEFAULT} +MAKE_CMD = make ARCH=um SUBARCH=x86_64 +ARCH = uml +CROSS = + +TEST_START IF ${RUN} == x86 || ${RUN} == i386 || ${DO_DEFAULT} +MAKE_CMD = make ARCH=i386 +ARCH = i386 +CROSS = + +TEST_START IF ${RUN} == x86 || ${RUN} == x86_64 || ${DO_DEFAULT} +MAKE_CMD = make ARCH=x86_64 +ARCH = x86_64 +CROSS = + +################################# + +# This is a bisect if needed. You need to give it a MIN_CONFIG that +# will be the config file it uses. Basically, just copy the created defconfig +# for the arch someplace and point MIN_CONFIG to it. +TEST_START IF ${RUN} == bisect +MIN_CONFIG = ${THIS_DIR}/min-config +CROSS = s390x-linux +ARCH = s390 +TEST_TYPE = bisect +BISECT_TYPE = build +BISECT_GOOD = v3.1 +BISECT_BAD = v3.2 +CHECKOUT = v3.2 + +################################# + +# These defaults are needed to keep ktest.pl from complaining. They are +# ignored because the test does not go pass the build. No install or +# booting of the target images. + +DEFAULTS +MACHINE = crosstest +SSH_USER = root +BUILD_TARGET = cross +TARGET_IMAGE = image +POWER_CYCLE = cycle +CONSOLE = console +LOCALVERSION = version +GRUB_MENU = grub + +REBOOT_ON_ERROR = 0 +POWEROFF_ON_ERROR = 0 +POWEROFF_ON_SUCCESS = 0 +REBOOT_ON_SUCCESS = 0 + diff --git a/tools/testing/ktest/examples/include/bisect.conf b/tools/testing/ktest/examples/include/bisect.conf new file mode 100644 index 000000000000..009bea65bfb6 --- /dev/null +++ b/tools/testing/ktest/examples/include/bisect.conf @@ -0,0 +1,90 @@ +# +# This example shows the bisect tests (git bisect and config bisect) +# + + +# The config that includes this file may define a RUN_TEST +# variable that will tell this config what test to run. +# (what to set the TEST option to). +# +DEFAULTS IF NOT DEFINED RUN_TEST +# Requires that hackbench is in the PATH +RUN_TEST := ${SSH} hackbench 50 + + +# Set TEST to 'bisect' to do a normal git bisect. You need +# to modify the options below to make it bisect the exact +# commits you are interested in. +# +TEST_START IF ${TEST} == bisect +TEST_TYPE = bisect +# You must set the commit that was considered good (git bisect good) +BISECT_GOOD = v3.3 +# You must set the commit that was considered bad (git bisect bad) +BISECT_BAD = HEAD +# It's best to specify the branch to checkout before starting the bisect. +CHECKOUT = origin/master +# This can be build, boot, or test. Here we are doing a bisect +# that requires to run a test to know if the bisect was good or bad. +# The test should exit with 0 on good, non-zero for bad. But see +# the BISECT_RET_* options in samples.conf to override this. +BISECT_TYPE = test +TEST = ${RUN_TEST} +# It is usually a good idea to confirm that the GOOD and the BAD +# commits are truly good and bad respectively. Having BISECT_CHECK +# set to 1 will check both that the good commit works and the bad +# commit fails. If you only want to check one or the other, +# set BISECT_CHECK to 'good' or to 'bad'. +BISECT_CHECK = 1 +#BISECT_CHECK = good +#BISECT_CHECK = bad + +# Usually it's a good idea to specify the exact config you +# want to use throughout the entire bisect. Here we placed +# it in the directory we called ktest.pl from and named it +# 'config-bisect'. +MIN_CONFIG = ${THIS_DIR}/config-bisect +# By default, if we are doing a BISECT_TYPE = test run but the +# build or boot fails, ktest.pl will do a 'git bisect skip'. +# Uncomment the below option to make ktest stop testing on such +# an error. +#BISECT_SKIP = 0 +# Now if you had BISECT_SKIP = 0 and the test fails, you can +# examine what happened and then do 'git bisect log > /tmp/replay' +# Set BISECT_REPLAY to /tmp/replay and ktest.pl will run the +# 'git bisect replay /tmp/replay' before continuing the bisect test. +#BISECT_REPLAY = /tmp/replay +# If you used BISECT_REPLAY after the bisect test failed, you may +# not want to continue the bisect on that commit that failed. +# By setting BISECT_START to a new commit. ktest.pl will checkout +# that commit after it has performed the 'git bisect replay' but +# before it continues running the bisect test. +#BISECT_START = 2545eb6198e7e1ec50daa0cfc64a4cdfecf24ec9 + +# Now if you don't trust ktest.pl to make the decisions for you, then +# set BISECT_MANUAL to 1. This will cause ktest.pl not to decide +# if the commit was good or bad. Instead, it will ask you to tell +# it if the current commit was good. In the mean time, you could +# take the result, load it on any machine you want. Run several tests, +# or whatever you feel like. Then, when you are happy, you can tell +# ktest if you think it was good or not and ktest.pl will continue +# the git bisect. You can even change what commit it is currently at. +#BISECT_MANUAL = 1 + + +# One of the unique tests that ktest does is the config bisect. +# Currently (which hopefully will be fixed soon), the bad config +# must be a superset of the good config. This is because it only +# searches for a config that causes the target to fail. If the +# good config is not a subset of the bad config, or if the target +# fails because of a lack of a config, then it will not find +# the config for you. +TEST_START IF ${TEST} == config-bisect +TEST_TYPE = config_bisect +# set to build, boot, test +CONFIG_BISECT_TYPE = boot +# Set the config that is considered bad. +CONFIG_BISECT = ${THIS_DIR}/config-bad +# This config is optional. By default it uses the +# MIN_CONFIG as the good config. +CONFIG_BISECT_GOOD = ${THIS_DIR}/config-good diff --git a/tools/testing/ktest/examples/include/defaults.conf b/tools/testing/ktest/examples/include/defaults.conf new file mode 100644 index 000000000000..323a552ce642 --- /dev/null +++ b/tools/testing/ktest/examples/include/defaults.conf @@ -0,0 +1,157 @@ +# This file holds defaults for most the tests. It defines the options that +# are most common to tests that are likely to be shared. +# +# Note, after including this file, a config file may override any option +# with a DEFAULTS OVERRIDE section. +# + +# For those cases that use the same machine to boot a 64 bit +# and a 32 bit version. The MACHINE is the DNS name to get to the +# box (usually different if it was 64 bit or 32 bit) but the +# BOX here is defined as a variable that will be the name of the box +# itself. It is useful for calling scripts that will power cycle +# the box, as only one script needs to be created to power cycle +# even though the box itself has multiple operating systems on it. +# By default, BOX and MACHINE are the same. + +DEFAULTS IF NOT DEFINED BOX +BOX := ${MACHINE} + + +# Consider each box as 64 bit box, unless the config including this file +# has defined BITS = 32 + +DEFAULTS IF NOT DEFINED BITS +BITS := 64 + + +DEFAULTS + +# THIS_DIR is used through out the configs and defaults to ${PWD} which +# is the directory that ktest.pl was called from. + +THIS_DIR := ${PWD} + + +# to orginize your configs, having each machine save their configs +# into a separate directly is useful. +CONFIG_DIR := ${THIS_DIR}/configs/${MACHINE} + +# Reset the log before running each test. +CLEAR_LOG = 1 + +# As installing kernels usually requires root privilege, default the +# user on the target as root. It is also required that the target +# allows ssh to root from the host without asking for a password. + +SSH_USER = root + +# For accesing the machine, we will ssh to root@machine. +SSH := ssh ${SSH_USER}@${MACHINE} + +# Update this. The default here is ktest will ssh to the target box +# and run a script called 'run-test' located on that box. +TEST = ${SSH} run-test + +# Point build dir to the git repo you use +BUILD_DIR = ${THIS_DIR}/linux.git + +# Each machine will have its own output build directory. +OUTPUT_DIR = ${THIS_DIR}/build/${MACHINE} + +# Yes this config is focused on x86 (but ktest works for other archs too) +BUILD_TARGET = arch/x86/boot/bzImage +TARGET_IMAGE = /boot/vmlinuz-test + +# have directory for the scripts to reboot and power cycle the boxes +SCRIPTS_DIR := ${THIS_DIR}/scripts + +# You can have each box/machine have a script to power cycle it. +# Name your script <box>-cycle. +POWER_CYCLE = ${SCRIPTS_DIR}/${BOX}-cycle + +# This script is used to power off the box. +POWER_OFF = ${SCRIPTS_DIR}/${BOX}-poweroff + +# Keep your test kernels separate from your other kernels. +LOCALVERSION = -test + +# The /boot/grub/menu.lst is searched for the line: +# title Test Kernel +# and ktest will use that kernel to reboot into. +# For grub2 or other boot loaders, you need to set BOOT_TYPE +# to 'script' and define other ways to load the kernel. +# See snowball.conf example. +# +GRUB_MENU = Test Kernel + +# The kernel build will use this option. +BUILD_OPTIONS = -j8 + +# Keeping the log file with the output dir is convenient. +LOG_FILE = ${OUTPUT_DIR}/${MACHINE}.log + +# Each box should have their own minum configuration +# See min-config.conf +MIN_CONFIG = ${CONFIG_DIR}/config-min + +# For things like randconfigs, there may be configs you find that +# are already broken, or there may be some configs that you always +# want set. Uncomment ADD_CONFIG and point it to the make config files +# that set the configs you want to keep on (or off) in your build. +# ADD_CONFIG is usually something to add configs to all machines, +# where as, MIN_CONFIG is specific per machine. +#ADD_CONFIG = ${THIS_DIR}/config-broken ${THIS_DIR}/config-general + +# To speed up reboots for bisects and patchcheck, instead of +# waiting 60 seconds for the console to be idle, if this line is +# seen in the console output, ktest will know the good kernel has +# finished rebooting and it will be able to continue the tests. +REBOOT_SUCCESS_LINE = ${MACHINE} login: + +# The following is different ways to end the test. +# by setting the variable REBOOT to: none, error, fail or +# something else, ktest will power cycle or reboot the target box +# at the end of the tests. +# +# REBOOT := none +# Don't do anything at the end of the test. +# +# REBOOT := error +# Reboot the box if ktest detects an error +# +# REBOOT := fail +# Do not stop on failure, and after all tests are complete +# power off the box (for both success and error) +# This is good to run over a weekend and you don't want to waste +# electricity. +# + +DEFAULTS IF ${REBOOT} == none +REBOOT_ON_SUCCESS = 0 +REBOOT_ON_ERROR = 0 +POWEROFF_ON_ERROR = 0 +POWEROFF_ON_SUCCESS = 0 + +DEFAULTS ELSE IF ${REBOOT} == error +REBOOT_ON_SUCCESS = 0 +REBOOT_ON_ERROR = 1 +POWEROFF_ON_ERROR = 0 +POWEROFF_ON_SUCCESS = 0 + +DEFAULTS ELSE IF ${REBOOT} == fail +REBOOT_ON_SUCCESS = 0 +POWEROFF_ON_ERROR = 1 +POWEROFF_ON_SUCCESS = 1 +POWEROFF_AFTER_HALT = 120 +DIE_ON_FAILURE = 0 + +# Store the failure information into this directory +# such as the .config, dmesg, and build log. +STORE_FAILURES = ${THIS_DIR}/failures + +DEFAULTS ELSE +REBOOT_ON_SUCCESS = 1 +REBOOT_ON_ERROR = 1 +POWEROFF_ON_ERROR = 0 +POWEROFF_ON_SUCCESS = 0 diff --git a/tools/testing/ktest/examples/include/min-config.conf b/tools/testing/ktest/examples/include/min-config.conf new file mode 100644 index 000000000000..c703cc46d151 --- /dev/null +++ b/tools/testing/ktest/examples/include/min-config.conf @@ -0,0 +1,60 @@ +# +# This file has some examples for creating a MIN_CONFIG. +# (A .config file that is the minimum for a machine to boot, or +# to boot and make a network connection.) +# +# A MIN_CONFIG is very useful as it is the minimum configuration +# needed to boot a given machine. You can debug someone else's +# .config by only setting the configs in your MIN_CONFIG. The closer +# your MIN_CONFIG is to the true minimum set of configs needed to +# boot your machine, the closer the config you test with will be +# to the users config that had the failure. +# +# The make_min_config test allows you to create a MIN_CONFIG that +# is truly the minimum set of configs needed to boot a box. +# +# In this example, the final config will reside in +# ${CONFIG_DIR}/config-new-min and ${CONFIG_DIR}/config-new-min-net. +# Just move one to the location you have set for MIN_CONFIG. +# +# The first test creates a MIN_CONFIG that will be the minimum +# configuration to boot ${MACHINE} and be able to ssh to it. +# +# The second test creates a MIN_CONFIG that will only boot +# the target and most likely will not let you ssh to it. (Notice +# how the second test uses the first test's result to continue with. +# This is because the second test config is a subset of the first). +# +# The ${CONFIG_DIR}/config-skip (and -net) will hold the configs +# that ktest.pl found would not boot the target without them set. +# The config-new-min holds configs that ktest.pl could not test +# directly because another config that was needed to boot the box +# selected them. Sometimes it is possible that this file will hold +# the true minimum configuration. You can test to see if this is +# the case by running the boot test with BOOT_TYPE = allnoconfig and +# setting setting the MIN_CONFIG to ${CONFIG_DIR}/config-skip. If the +# machine still boots, then you can use the config-skip as your MIN_CONFIG. +# +# These tests can run for several hours (and perhaps days). +# It's OK to kill the test with a Ctrl^C. By restarting without +# modifying this config, ktest.pl will notice that the config-new-min(-net) +# exists, and will use that instead as the starting point. +# The USE_OUTPUT_MIN_CONFIG is set to 1 to keep ktest.pl from asking +# you if you want to use the OUTPUT_MIN_CONFIG as the starting point. +# By using the OUTPUT_MIN_CONFIG as the starting point will allow ktest.pl to +# start almost where it left off. +# +TEST_START IF ${TEST} == min-config +TEST_TYPE = make_min_config +OUTPUT_MIN_CONFIG = ${CONFIG_DIR}/config-new-min-net +IGNORE_CONFIG = ${CONFIG_DIR}/config-skip-net +MIN_CONFIG_TYPE = test +TEST = ${SSH} echo hi +USE_OUTPUT_MIN_CONFIG = 1 + +TEST_START IF ${TEST} == min-config && ${MULTI} +TEST_TYPE = make_min_config +OUTPUT_MIN_CONFIG = ${CONFIG_DIR}/config-new-min +IGNORE_CONFIG = ${CONFIG_DIR}/config-skip +MIN_CONFIG = ${CONFIG_DIR}/config-new-min-net +USE_OUTPUT_MIN_CONFIG = 1 diff --git a/tools/testing/ktest/examples/include/patchcheck.conf b/tools/testing/ktest/examples/include/patchcheck.conf new file mode 100644 index 000000000000..339d3e1700ff --- /dev/null +++ b/tools/testing/ktest/examples/include/patchcheck.conf @@ -0,0 +1,74 @@ +# patchcheck.conf +# +# This contains a test that takes two git commits and will test each +# commit between the two. The build test will look at what files the +# commit has touched, and if any of those files produce a warning, then +# the build will fail. + + +# PATCH_START is the commit to begin with and PATCH_END is the commit +# to end with (inclusive). This is similar to doing a git rebase -i PATCH_START~1 +# and then testing each commit and doing a git rebase --continue. +# You can use a SHA1, a git tag, or anything that git will accept for a checkout + +PATCH_START := HEAD~3 +PATCH_END := HEAD + +# Change PATCH_CHECKOUT to be the branch you want to test. The test will +# do a git checkout of this branch before starting. Obviously both +# PATCH_START and PATCH_END must be in this branch (and PATCH_START must +# be contained by PATCH_END). + +PATCH_CHECKOUT := test/branch + +# Usually it's a good idea to have a set config to use for testing individual +# patches. +PATCH_CONFIG := ${CONFIG_DIR}/config-patchcheck + +# Change PATCH_TEST to run some test for each patch. Each commit that is +# tested, after it is built and installed on the test machine, this command +# will be executed. Usually what is done is to ssh to the target box and +# run some test scripts. If you just want to boot test your patches +# comment PATCH_TEST out. +PATCH_TEST := ${SSH} "/usr/local/bin/ktest-test-script" + +DEFAULTS IF DEFINED PATCH_TEST +PATCH_TEST_TYPE := test + +DEFAULTS ELSE +PATCH_TEST_TYPE := boot + +# If for some reason a file has a warning that one of your patches touch +# but you do not care about it, set IGNORE_WARNINGS to that commit(s) +# (space delimited) +#IGNORE_WARNINGS = 39eaf7ef884dcc44f7ff1bac803ca2a1dcf43544 6edb2a8a385f0cdef51dae37ff23e74d76d8a6ce + +# If you are running a multi test, and the test failed on the first +# test but on, say the 5th patch. If you want to restart on the +# fifth patch, set PATCH_START1. This will make the first test start +# from this commit instead of the PATCH_START commit. +# Note, do not change this option. Just define PATCH_START1 in the +# top config (the one you pass to ktest.pl), and this will use it, +# otherwise it will just use PATCH_START if PATCH_START1 is not defined. +DEFAULTS IF NOT DEFINED PATCH_START1 +PATCH_START1 := ${PATCH_START} + +TEST_START IF ${TEST} == patchcheck +TEST_TYPE = patchcheck +MIN_CONFIG = ${PATCH_CONFIG} +TEST = ${PATCH_TEST} +PATCHCHECK_TYPE = ${PATCH_TEST_TYPE} +PATCHCHECK_START = ${PATCH_START1} +PATCHCHECK_END = ${PATCH_END} +CHECKOUT = ${PATCH_CHECKOUT} + +TEST_START IF ${TEST} == patchcheck && ${MULTI} +TEST_TYPE = patchcheck +MIN_CONFIG = ${PATCH_CONFIG} +TEST = ${PATCH_TEST} +PATCHCHECK_TYPE = ${PATCH_TEST_TYPE} +PATCHCHECK_START = ${PATCH_START} +PATCHCHECK_END = ${PATCH_END} +CHECKOUT = ${PATCH_CHECKOUT} +# Use multi to test different compilers? +MAKE_CMD = CC=gcc-4.5.1 make diff --git a/tools/testing/ktest/examples/include/tests.conf b/tools/testing/ktest/examples/include/tests.conf new file mode 100644 index 000000000000..4fdb811bd810 --- /dev/null +++ b/tools/testing/ktest/examples/include/tests.conf @@ -0,0 +1,74 @@ +# +# This is an example of various tests that you can run +# +# The variable TEST can be of boot, build, randconfig, or test. +# +# Note that TEST is a variable created with ':=' and only exists +# throughout the config processing (not during the tests itself). +# +# The TEST option (defined with '=') is used to tell ktest.pl +# what test to run after a successful boot. The TEST option is +# persistent into the test runs. +# + +# The config that includes this file may define a BOOT_TYPE +# variable that tells this config what type of boot test to run. +# If it's not defined, the below DEFAULTS will set the default +# to 'oldconfig'. +# +DEFAULTS IF NOT DEFINED BOOT_TYPE +BOOT_TYPE := oldconfig + +# The config that includes this file may define a RUN_TEST +# variable that will tell this config what test to run. +# (what to set the TEST option to). +# +DEFAULTS IF NOT DEFINED RUN_TEST +# Requires that hackbench is in the PATH +RUN_TEST := ${SSH} hackbench 50 + + +# If TEST is set to 'boot' then just build a kernel and boot +# the target. +TEST_START IF ${TEST} == boot +TEST_TYPE = boot +# Notice how we set the BUILD_TYPE option to the BOOT_TYPE variable. +BUILD_TYPE = ${BOOT_TYPE} +# Do not do a make mrproper. +BUILD_NOCLEAN = 1 + +# If you only want to build the kernel, and perhaps install +# and test it yourself, then just set TEST to build. +TEST_START IF ${TEST} == build +TEST_TYPE = build +BUILD_TYPE = ${BOOT_TYPE} +BUILD_NOCLEAN = 1 + +# Build, install, boot and test with a randconfg 10 times. +# It is important that you have set MIN_CONFIG in the config +# that includes this file otherwise it is likely that the +# randconfig will not have the neccessary configs needed to +# boot your box. This version of the test requires a min +# config that has enough to make sure the target has network +# working. +TEST_START ITERATE 10 IF ${TEST} == randconfig +MIN_CONFIG = ${CONFIG_DIR}/config-min-net +TEST_TYPE = test +BUILD_TYPE = randconfig +TEST = ${RUN_TEST} + +# This is the same as above, but only tests to a boot prompt. +# The MIN_CONFIG used here does not need to have networking +# working. +TEST_START ITERATE 10 IF ${TEST} == randconfig && ${MULTI} +TEST_TYPE = boot +BUILD_TYPE = randconfig +MIN_CONFIG = ${CONFIG_DIR}/config-min +MAKE_CMD = make + +# This builds, installs, boots and tests the target. +TEST_START IF ${TEST} == test +TEST_TYPE = test +BUILD_TYPE = ${BOOT_TYPE} +TEST = ${RUN_TEST} +BUILD_NOCLEAN = 1 diff --git a/tools/testing/ktest/examples/kvm.conf b/tools/testing/ktest/examples/kvm.conf new file mode 100644 index 000000000000..831c7c5395f1 --- /dev/null +++ b/tools/testing/ktest/examples/kvm.conf @@ -0,0 +1,88 @@ +# +# This config is an example usage of ktest.pl with a kvm guest +# +# The guest is called 'Guest' and this would be something that +# could be run on the host to test a virtual machine target. + +MACHINE = Guest + + +# Use virsh to read the serial console of the guest +CONSOLE = virsh console ${MACHINE} + +#*************************************# +# This part is the same as test.conf # +#*************************************# + +# The include files will set up the type of test to run. Just set TEST to +# which test you want to run. +# +# TESTS = patchcheck, randconfig, boot, test, config-bisect, bisect, min-config +# +# See the include/*.conf files that define these tests +# +TEST := patchcheck + +# Some tests may have more than one test to run. Define MULTI := 1 to run +# the extra tests. +MULTI := 0 + +# In case you want to differentiate which type of system you are testing +BITS := 64 + +# REBOOT = none, error, fail, empty +# See include/defaults.conf +REBOOT := empty + + +# The defaults file will set up various settings that can be used by all +# machine configs. +INCLUDE include/defaults.conf + + +#*************************************# +# Now we are different from test.conf # +#*************************************# + + +# The example here assumes that Guest is running a Fedora release +# that uses dracut for its initfs. The POST_INSTALL will be executed +# after the install of the kernel and modules are complete. +# +POST_INSTALL = ${SSH} /sbin/dracut -f /boot/initramfs-test.img $KERNEL_VERSION + +# Guests sometimes get stuck on reboot. We wait 3 seconds after running +# the reboot command and then do a full power-cycle of the guest. +# This forces the guest to restart. +# +POWERCYCLE_AFTER_REBOOT = 3 + +# We do the same after the halt command, but this time we wait 20 seconds. +POWEROFF_AFTER_HALT = 20 + + +# As the defaults.conf file has a POWER_CYCLE option already defined, +# and options can not be defined in the same section more than once +# (all DEFAULTS sections are considered the same). We use the +# DEFAULTS OVERRIDE to tell ktest.pl to ignore the previous defined +# options, for the options set in the OVERRIDE section. +# +DEFAULTS OVERRIDE + +# Instead of using the default POWER_CYCLE option defined in +# defaults.conf, we use virsh to cycle it. To do so, we destroy +# the guest, wait 5 seconds, and then start it up again. +# Crude, but effective. +# +POWER_CYCLE = virsh destroy ${MACHINE}; sleep 5; virsh start ${MACHINE} + + +DEFAULTS + +# The following files each handle a different test case. +# Having them included allows you to set up more than one machine and share +# the same tests. +INCLUDE include/patchcheck.conf +INCLUDE include/tests.conf +INCLUDE include/bisect.conf +INCLUDE include/min-config.conf diff --git a/tools/testing/ktest/examples/snowball.conf b/tools/testing/ktest/examples/snowball.conf new file mode 100644 index 000000000000..a82a3c5bc2b7 --- /dev/null +++ b/tools/testing/ktest/examples/snowball.conf @@ -0,0 +1,53 @@ +# This example was used to boot the snowball ARM board. +# See http://people.redhat.com/srostedt/ktest-embedded-2012/ + +# PWD is a ktest.pl variable that will result in the process working +# directory that ktest.pl is executed in. + +# THIS_DIR is automatically assigned the PWD of the path that generated +# the config file. It is best to use this variable when assigning other +# directory paths within this directory. This allows you to easily +# move the test cases to other locations or to other machines. +# +THIS_DIR := /home/rostedt/work/demo/ktest-embed +LOG_FILE = ${OUTPUT_DIR}/snowball.log +CLEAR_LOG = 1 +MAKE_CMD = PATH=/usr/local/gcc-4.5.2-nolibc/arm-unknown-linux-gnueabi/bin:$PATH CROSS_COMPILE=arm-unknown-linux-gnueabi- make ARCH=arm +ADD_CONFIG = ${THIS_DIR}/addconfig + +SCP_TO_TARGET = echo "don't do scp" + +TFTPBOOT := /var/lib/tftpboot +TFTPDEF := ${TFTPBOOT}/snowball-default +TFTPTEST := ${OUTPUT_DIR}/${BUILD_TARGET} + +SWITCH_TO_GOOD = cp ${TFTPDEF} ${TARGET_IMAGE} +SWITCH_TO_TEST = cp ${TFTPTEST} ${TARGET_IMAGE} + +# Define each test with TEST_START +# The config options below it will override the defaults +TEST_START SKIP +TEST_TYPE = boot +BUILD_TYPE = u8500_defconfig +BUILD_NOCLEAN = 1 + +TEST_START +TEST_TYPE = make_min_config +OUTPUT_MIN_CONFIG = ${THIS_DIR}/config.newmin +START_MIN_CONFIG = ${THIS_DIR}/config.orig +IGNORE_CONFIG = ${THIS_DIR}/config.ignore +BUILD_NOCLEAN = 1 + + +DEFAULTS +LOCALVERSION = -test +POWER_CYCLE = echo use the thumb luke; read a +CONSOLE = cat ${THIS_DIR}/snowball-cat +REBOOT_TYPE = script +SSH_USER = root +BUILD_OPTIONS = -j8 uImage +BUILD_DIR = ${THIS_DIR}/linux.git +OUTPUT_DIR = ${THIS_DIR}/snowball-build +MACHINE = snowball +TARGET_IMAGE = /var/lib/tftpboot/snowball-image +BUILD_TARGET = arch/arm/boot/uImage diff --git a/tools/testing/ktest/examples/test.conf b/tools/testing/ktest/examples/test.conf new file mode 100644 index 000000000000..b725210efb7f --- /dev/null +++ b/tools/testing/ktest/examples/test.conf @@ -0,0 +1,62 @@ +# +# Generic config for a machine +# + +# Name your machine (the DNS name, what you ssh to) +MACHINE = foo + +# BOX can be different than foo, if the machine BOX has +# multiple partitions with different systems installed. For example, +# you may have a i386 and x86_64 installation on a test box. +# If this is the case, MACHINE defines the way to connect to the +# machine, which may be different between which system the machine +# is booting into. BOX is used for the scripts to reboot and power cycle +# the machine, where it does not matter which system the machine boots into. +# +#BOX := bar + +# Define a way to read the console +CONSOLE = stty -F /dev/ttyS0 115200 parodd; cat /dev/ttyS0 + +# The include files will set up the type of test to run. Just set TEST to +# which test you want to run. +# +# TESTS = patchcheck, randconfig, boot, test, config-bisect, bisect, min-config +# +# See the include/*.conf files that define these tests +# +TEST := patchcheck + +# Some tests may have more than one test to run. Define MULTI := 1 to run +# the extra tests. +MULTI := 0 + +# In case you want to differentiate which type of system you are testing +BITS := 64 + +# REBOOT = none, error, fail, empty +# See include/defaults.conf +REBOOT := empty + +# The defaults file will set up various settings that can be used by all +# machine configs. +INCLUDE include/defaults.conf + +# In case you need to add a patch for a bisect or something +#PRE_BUILD = patch -p1 < ${THIS_DIR}/fix.patch + +# Reset the repo after the build and remove all 'test' modules from the target +# Notice that DO_POST_BUILD is a variable (defined by ':=') and POST_BUILD +# is the option (defined by '=') + +DO_POST_BUILD := git reset --hard +POST_BUILD = ${SSH} 'rm -rf /lib/modules/*-test*'; ${DO_POST_BUILD} + +# The following files each handle a different test case. +# Having them included allows you to set up more than one machine and share +# the same tests. +INCLUDE include/patchcheck.conf +INCLUDE include/tests.conf +INCLUDE include/bisect.conf +INCLUDE include/min-config.conf + diff --git a/tools/testing/ktest/ktest.pl b/tools/testing/ktest/ktest.pl index 95d6a6f7c33a..292b13ad03f5 100755 --- a/tools/testing/ktest/ktest.pl +++ b/tools/testing/ktest/ktest.pl @@ -39,6 +39,7 @@ my %default = ( "CLEAR_LOG" => 0, "BISECT_MANUAL" => 0, "BISECT_SKIP" => 1, + "MIN_CONFIG_TYPE" => "boot", "SUCCESS_LINE" => "login:", "DETECT_TRIPLE_FAULT" => 1, "NO_INSTALL" => 0, @@ -66,6 +67,7 @@ my %default = ( my $ktest_config; my $version; +my $have_version = 0; my $machine; my $ssh_user; my $tmpdir; @@ -106,6 +108,8 @@ my $minconfig; my $start_minconfig; my $start_minconfig_defined; my $output_minconfig; +my $minconfig_type; +my $use_output_minconfig; my $ignore_config; my $ignore_errors; my $addconfig; @@ -183,6 +187,9 @@ my %force_config; # do not force reboots on config problems my $no_reboot = 1; +# reboot on success +my $reboot_success = 0; + my %option_map = ( "MACHINE" => \$machine, "SSH_USER" => \$ssh_user, @@ -202,6 +209,8 @@ my %option_map = ( "MIN_CONFIG" => \$minconfig, "OUTPUT_MIN_CONFIG" => \$output_minconfig, "START_MIN_CONFIG" => \$start_minconfig, + "MIN_CONFIG_TYPE" => \$minconfig_type, + "USE_OUTPUT_MIN_CONFIG" => \$use_output_minconfig, "IGNORE_CONFIG" => \$ignore_config, "TEST" => \$run_test, "ADD_CONFIG" => \$addconfig, @@ -1699,10 +1708,12 @@ sub install { sub get_version { # get the release name + return if ($have_version); doprint "$make kernelrelease ... "; $version = `$make kernelrelease | tail -1`; chomp($version); doprint "$version\n"; + $have_version = 1; } sub start_monitor_and_boot { @@ -1825,6 +1836,9 @@ sub build { my $save_no_reboot = $no_reboot; $no_reboot = 1; + # Calculate a new version from here. + $have_version = 0; + if (defined($pre_build)) { my $ret = run_command $pre_build; if (!$ret && defined($pre_build_die) && @@ -1884,6 +1898,9 @@ sub build { undef $redirect; if (defined($post_build)) { + # Because a post build may change the kernel version + # do it now. + get_version; my $ret = run_command $post_build; if (!$ret && defined($post_build_die) && $post_build_die) { @@ -2192,7 +2209,7 @@ sub run_bisect { } # Are we looking for where it worked, not failed? - if ($reverse_bisect) { + if ($reverse_bisect && $ret >= 0) { $ret = !$ret; } @@ -3116,6 +3133,12 @@ sub test_this_config { sub make_min_config { my ($i) = @_; + my $type = $minconfig_type; + if ($type ne "boot" && $type ne "test") { + fail "Invalid MIN_CONFIG_TYPE '$minconfig_type'\n" . + " make_min_config works only with 'boot' and 'test'\n" and return; + } + if (!defined($output_minconfig)) { fail "OUTPUT_MIN_CONFIG not defined" and return; } @@ -3125,8 +3148,15 @@ sub make_min_config { # that instead. if (-f $output_minconfig && !$start_minconfig_defined) { print "$output_minconfig exists\n"; - if (read_yn " Use it as minconfig?") { + if (!defined($use_output_minconfig)) { + if (read_yn " Use it as minconfig?") { + $start_minconfig = $output_minconfig; + } + } elsif ($use_output_minconfig > 0) { + doprint "Using $output_minconfig as MIN_CONFIG\n"; $start_minconfig = $output_minconfig; + } else { + doprint "Set to still use MIN_CONFIG as starting point\n"; } } @@ -3275,6 +3305,11 @@ sub make_min_config { build "oldconfig" or $failed = 1; if (!$failed) { start_monitor_and_boot or $failed = 1; + + if ($type eq "test" && !$failed) { + do_run_test or $failed = 1; + } + end_monitor; } @@ -3469,6 +3504,9 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { # Do not reboot on failing test options $no_reboot = 1; + $reboot_success = 0; + + $have_version = 0; $iteration = $i; @@ -3554,9 +3592,11 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { die "failed to checkout $checkout"; } + $no_reboot = 0; + # A test may opt to not reboot the box if ($reboot_on_success) { - $no_reboot = 0; + $reboot_success = 1; } if ($test_type eq "bisect") { @@ -3600,7 +3640,7 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { if ($opt{"POWEROFF_ON_SUCCESS"}) { halt; -} elsif ($opt{"REBOOT_ON_SUCCESS"} && !do_not_reboot) { +} elsif ($opt{"REBOOT_ON_SUCCESS"} && !do_not_reboot && $reboot_success) { reboot_to_good; } elsif (defined($switch_to_good)) { # still need to get to the good kernel diff --git a/tools/testing/ktest/sample.conf b/tools/testing/ktest/sample.conf index b682456afda8..cf362b3d1ec9 100644 --- a/tools/testing/ktest/sample.conf +++ b/tools/testing/ktest/sample.conf @@ -158,7 +158,7 @@ # # TEST_START IF (DEFINED ALL_TESTS || ${MYTEST} == boottest) && ${MACHINE} == gandalf # -# Notice the use of paranthesis. Without any paranthesis the above would be +# Notice the use of parentheses. Without any parentheses the above would be # processed the same as: # # TEST_START IF DEFINED ALL_TESTS || (${MYTEST} == boottest && ${MACHINE} == gandalf) @@ -1105,10 +1105,26 @@ # and will not be tested again in later runs. # (optional) # +# MIN_CONFIG_TYPE can be either 'boot' or 'test'. With 'boot' it will +# test if the created config can just boot the machine. If this is +# set to 'test', then the TEST option must be defined and the created +# config will not only boot the target, but also make sure that the +# config lets the test succeed. This is useful to make sure the final +# config that is generated allows network activity (ssh). +# (optional) +# +# USE_OUTPUT_MIN_CONFIG set this to 1 if you do not want to be prompted +# about using the OUTPUT_MIN_CONFIG as the MIN_CONFIG as the starting +# point. Set it to 0 if you want to always just use the given MIN_CONFIG. +# If it is not defined, it will prompt you to pick which config +# to start with (MIN_CONFIG or OUTPUT_MIN_CONFIG). +# # Example: # # TEST_TYPE = make_min_config # OUTPUT_MIN_CONFIG = /path/to/config-new-min # START_MIN_CONFIG = /path/to/config-min # IGNORE_CONFIG = /path/to/config-tested +# MIN_CONFIG_TYPE = test +# TEST = ssh ${USER}@${MACHINE} echo hi # diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index 4ec84018cc13..a4162e15c25f 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -1,10 +1,15 @@ -TARGETS = breakpoints +TARGETS = breakpoints kcmp mqueue vm all: for TARGET in $(TARGETS); do \ make -C $$TARGET; \ done; +run_tests: all + for TARGET in $(TARGETS); do \ + make -C $$TARGET run_tests; \ + done; + clean: for TARGET in $(TARGETS); do \ make -C $$TARGET clean; \ diff --git a/tools/testing/selftests/breakpoints/Makefile b/tools/testing/selftests/breakpoints/Makefile index f362722cdce7..931278035f5c 100644 --- a/tools/testing/selftests/breakpoints/Makefile +++ b/tools/testing/selftests/breakpoints/Makefile @@ -11,10 +11,13 @@ endif all: ifeq ($(ARCH),x86) - gcc breakpoint_test.c -o run_test + gcc breakpoint_test.c -o breakpoint_test else echo "Not an x86 target, can't build breakpoints selftests" endif +run_tests: + ./breakpoint_test + clean: - rm -fr run_test + rm -fr breakpoint_test diff --git a/tools/testing/selftests/kcmp/Makefile b/tools/testing/selftests/kcmp/Makefile new file mode 100644 index 000000000000..dc79b86ea65c --- /dev/null +++ b/tools/testing/selftests/kcmp/Makefile @@ -0,0 +1,29 @@ +uname_M := $(shell uname -m 2>/dev/null || echo not) +ARCH ?= $(shell echo $(uname_M) | sed -e s/i.86/i386/) +ifeq ($(ARCH),i386) + ARCH := X86 + CFLAGS := -DCONFIG_X86_32 -D__i386__ +endif +ifeq ($(ARCH),x86_64) + ARCH := X86 + CFLAGS := -DCONFIG_X86_64 -D__x86_64__ +endif + +CFLAGS += -I../../../../arch/x86/include/generated/ +CFLAGS += -I../../../../include/ +CFLAGS += -I../../../../usr/include/ +CFLAGS += -I../../../../arch/x86/include/ + +all: +ifeq ($(ARCH),X86) + gcc $(CFLAGS) kcmp_test.c -o run_test +else + echo "Not an x86 target, can't build kcmp selftest" +endif + +run-tests: all + ./kcmp_test + +clean: + rm -fr ./run_test + rm -fr ./test-file diff --git a/tools/testing/selftests/kcmp/kcmp_test.c b/tools/testing/selftests/kcmp/kcmp_test.c new file mode 100644 index 000000000000..358cc6bfa35d --- /dev/null +++ b/tools/testing/selftests/kcmp/kcmp_test.c @@ -0,0 +1,94 @@ +#define _GNU_SOURCE + +#include <stdio.h> +#include <stdlib.h> +#include <signal.h> +#include <limits.h> +#include <unistd.h> +#include <errno.h> +#include <string.h> +#include <fcntl.h> + +#include <linux/unistd.h> +#include <linux/kcmp.h> + +#include <sys/syscall.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/wait.h> + +static long sys_kcmp(int pid1, int pid2, int type, int fd1, int fd2) +{ + return syscall(__NR_kcmp, pid1, pid2, type, fd1, fd2); +} + +int main(int argc, char **argv) +{ + const char kpath[] = "kcmp-test-file"; + int pid1, pid2; + int fd1, fd2; + int status; + + fd1 = open(kpath, O_RDWR | O_CREAT | O_TRUNC, 0644); + pid1 = getpid(); + + if (fd1 < 0) { + perror("Can't create file"); + exit(1); + } + + pid2 = fork(); + if (pid2 < 0) { + perror("fork failed"); + exit(1); + } + + if (!pid2) { + int pid2 = getpid(); + int ret; + + fd2 = open(kpath, O_RDWR, 0644); + if (fd2 < 0) { + perror("Can't open file"); + exit(1); + } + + /* An example of output and arguments */ + printf("pid1: %6d pid2: %6d FD: %2ld FILES: %2ld VM: %2ld " + "FS: %2ld SIGHAND: %2ld IO: %2ld SYSVSEM: %2ld " + "INV: %2ld\n", + pid1, pid2, + sys_kcmp(pid1, pid2, KCMP_FILE, fd1, fd2), + sys_kcmp(pid1, pid2, KCMP_FILES, 0, 0), + sys_kcmp(pid1, pid2, KCMP_VM, 0, 0), + sys_kcmp(pid1, pid2, KCMP_FS, 0, 0), + sys_kcmp(pid1, pid2, KCMP_SIGHAND, 0, 0), + sys_kcmp(pid1, pid2, KCMP_IO, 0, 0), + sys_kcmp(pid1, pid2, KCMP_SYSVSEM, 0, 0), + + /* This one should fail */ + sys_kcmp(pid1, pid2, KCMP_TYPES + 1, 0, 0)); + + /* This one should return same fd */ + ret = sys_kcmp(pid1, pid2, KCMP_FILE, fd1, fd1); + if (ret) { + printf("FAIL: 0 expected but %d returned\n", ret); + ret = -1; + } else + printf("PASS: 0 returned as expected\n"); + + /* Compare with self */ + ret = sys_kcmp(pid1, pid1, KCMP_VM, 0, 0); + if (ret) { + printf("FAIL: 0 expected but %li returned\n", ret); + ret = -1; + } else + printf("PASS: 0 returned as expected\n"); + + exit(ret); + } + + waitpid(pid2, &status, P_ALL); + + return 0; +} diff --git a/tools/testing/selftests/mqueue/.gitignore b/tools/testing/selftests/mqueue/.gitignore new file mode 100644 index 000000000000..d8d42377205a --- /dev/null +++ b/tools/testing/selftests/mqueue/.gitignore @@ -0,0 +1,2 @@ +mq_open_tests +mq_perf_tests diff --git a/tools/testing/selftests/mqueue/Makefile b/tools/testing/selftests/mqueue/Makefile new file mode 100644 index 000000000000..54c0aad2b47c --- /dev/null +++ b/tools/testing/selftests/mqueue/Makefile @@ -0,0 +1,10 @@ +all: + gcc -O2 -lrt mq_open_tests.c -o mq_open_tests + gcc -O2 -lrt -lpthread -lpopt -o mq_perf_tests mq_perf_tests.c + +run_tests: + ./mq_open_tests /test1 + ./mq_perf_tests + +clean: + rm -f mq_open_tests mq_perf_tests diff --git a/tools/testing/selftests/mqueue/mq_open_tests.c b/tools/testing/selftests/mqueue/mq_open_tests.c new file mode 100644 index 000000000000..711cc2923047 --- /dev/null +++ b/tools/testing/selftests/mqueue/mq_open_tests.c @@ -0,0 +1,492 @@ +/* + * This application is Copyright 2012 Red Hat, Inc. + * Doug Ledford <dledford@redhat.com> + * + * mq_open_tests is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, version 3. + * + * mq_open_tests is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * For the full text of the license, see <http://www.gnu.org/licenses/>. + * + * mq_open_tests.c + * Tests the various situations that should either succeed or fail to + * open a posix message queue and then reports whether or not they + * did as they were supposed to. + * + */ +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> +#include <fcntl.h> +#include <string.h> +#include <limits.h> +#include <errno.h> +#include <sys/types.h> +#include <sys/time.h> +#include <sys/resource.h> +#include <sys/stat.h> +#include <mqueue.h> + +static char *usage = +"Usage:\n" +" %s path\n" +"\n" +" path Path name of the message queue to create\n" +"\n" +" Note: this program must be run as root in order to enable all tests\n" +"\n"; + +char *DEF_MSGS = "/proc/sys/fs/mqueue/msg_default"; +char *DEF_MSGSIZE = "/proc/sys/fs/mqueue/msgsize_default"; +char *MAX_MSGS = "/proc/sys/fs/mqueue/msg_max"; +char *MAX_MSGSIZE = "/proc/sys/fs/mqueue/msgsize_max"; + +int default_settings; +struct rlimit saved_limits, cur_limits; +int saved_def_msgs, saved_def_msgsize, saved_max_msgs, saved_max_msgsize; +int cur_def_msgs, cur_def_msgsize, cur_max_msgs, cur_max_msgsize; +FILE *def_msgs, *def_msgsize, *max_msgs, *max_msgsize; +char *queue_path; +mqd_t queue = -1; + +static inline void __set(FILE *stream, int value, char *err_msg); +void shutdown(int exit_val, char *err_cause, int line_no); +static inline int get(FILE *stream); +static inline void set(FILE *stream, int value); +static inline void getr(int type, struct rlimit *rlim); +static inline void setr(int type, struct rlimit *rlim); +void validate_current_settings(); +static inline void test_queue(struct mq_attr *attr, struct mq_attr *result); +static inline int test_queue_fail(struct mq_attr *attr, struct mq_attr *result); + +static inline void __set(FILE *stream, int value, char *err_msg) +{ + rewind(stream); + if (fprintf(stream, "%d", value) < 0) + perror(err_msg); +} + + +void shutdown(int exit_val, char *err_cause, int line_no) +{ + static int in_shutdown = 0; + + /* In case we get called recursively by a set() call below */ + if (in_shutdown++) + return; + + seteuid(0); + + if (queue != -1) + if (mq_close(queue)) + perror("mq_close() during shutdown"); + if (queue_path) + /* + * Be silent if this fails, if we cleaned up already it's + * expected to fail + */ + mq_unlink(queue_path); + if (default_settings) { + if (saved_def_msgs) + __set(def_msgs, saved_def_msgs, + "failed to restore saved_def_msgs"); + if (saved_def_msgsize) + __set(def_msgsize, saved_def_msgsize, + "failed to restore saved_def_msgsize"); + } + if (saved_max_msgs) + __set(max_msgs, saved_max_msgs, + "failed to restore saved_max_msgs"); + if (saved_max_msgsize) + __set(max_msgsize, saved_max_msgsize, + "failed to restore saved_max_msgsize"); + if (exit_val) + error(exit_val, errno, "%s at %d", err_cause, line_no); + exit(0); +} + +static inline int get(FILE *stream) +{ + int value; + rewind(stream); + if (fscanf(stream, "%d", &value) != 1) + shutdown(4, "Error reading /proc entry", __LINE__ - 1); + return value; +} + +static inline void set(FILE *stream, int value) +{ + int new_value; + + rewind(stream); + if (fprintf(stream, "%d", value) < 0) + return shutdown(5, "Failed writing to /proc file", + __LINE__ - 1); + new_value = get(stream); + if (new_value != value) + return shutdown(5, "We didn't get what we wrote to /proc back", + __LINE__ - 1); +} + +static inline void getr(int type, struct rlimit *rlim) +{ + if (getrlimit(type, rlim)) + shutdown(6, "getrlimit()", __LINE__ - 1); +} + +static inline void setr(int type, struct rlimit *rlim) +{ + if (setrlimit(type, rlim)) + shutdown(7, "setrlimit()", __LINE__ - 1); +} + +void validate_current_settings() +{ + int rlim_needed; + + if (cur_limits.rlim_cur < 4096) { + printf("Current rlimit value for POSIX message queue bytes is " + "unreasonably low,\nincreasing.\n\n"); + cur_limits.rlim_cur = 8192; + cur_limits.rlim_max = 16384; + setr(RLIMIT_MSGQUEUE, &cur_limits); + } + + if (default_settings) { + rlim_needed = (cur_def_msgs + 1) * (cur_def_msgsize + 1 + + 2 * sizeof(void *)); + if (rlim_needed > cur_limits.rlim_cur) { + printf("Temporarily lowering default queue parameters " + "to something that will work\n" + "with the current rlimit values.\n\n"); + set(def_msgs, 10); + cur_def_msgs = 10; + set(def_msgsize, 128); + cur_def_msgsize = 128; + } + } else { + rlim_needed = (cur_max_msgs + 1) * (cur_max_msgsize + 1 + + 2 * sizeof(void *)); + if (rlim_needed > cur_limits.rlim_cur) { + printf("Temporarily lowering maximum queue parameters " + "to something that will work\n" + "with the current rlimit values in case this is " + "a kernel that ties the default\n" + "queue parameters to the maximum queue " + "parameters.\n\n"); + set(max_msgs, 10); + cur_max_msgs = 10; + set(max_msgsize, 128); + cur_max_msgsize = 128; + } + } +} + +/* + * test_queue - Test opening a queue, shutdown if we fail. This should + * only be called in situations that should never fail. We clean up + * after ourselves and return the queue attributes in *result. + */ +static inline void test_queue(struct mq_attr *attr, struct mq_attr *result) +{ + int flags = O_RDWR | O_EXCL | O_CREAT; + int perms = DEFFILEMODE; + + if ((queue = mq_open(queue_path, flags, perms, attr)) == -1) + shutdown(1, "mq_open()", __LINE__); + if (mq_getattr(queue, result)) + shutdown(1, "mq_getattr()", __LINE__); + if (mq_close(queue)) + shutdown(1, "mq_close()", __LINE__); + queue = -1; + if (mq_unlink(queue_path)) + shutdown(1, "mq_unlink()", __LINE__); +} + +/* + * Same as test_queue above, but failure is not fatal. + * Returns: + * 0 - Failed to create a queue + * 1 - Created a queue, attributes in *result + */ +static inline int test_queue_fail(struct mq_attr *attr, struct mq_attr *result) +{ + int flags = O_RDWR | O_EXCL | O_CREAT; + int perms = DEFFILEMODE; + + if ((queue = mq_open(queue_path, flags, perms, attr)) == -1) + return 0; + if (mq_getattr(queue, result)) + shutdown(1, "mq_getattr()", __LINE__); + if (mq_close(queue)) + shutdown(1, "mq_close()", __LINE__); + queue = -1; + if (mq_unlink(queue_path)) + shutdown(1, "mq_unlink()", __LINE__); + return 1; +} + +int main(int argc, char *argv[]) +{ + struct mq_attr attr, result; + + if (argc != 2) { + fprintf(stderr, "Must pass a valid queue name\n\n"); + fprintf(stderr, usage, argv[0]); + exit(1); + } + + /* + * Although we can create a msg queue with a non-absolute path name, + * unlink will fail. So, if the name doesn't start with a /, add one + * when we save it. + */ + if (*argv[1] == '/') + queue_path = strdup(argv[1]); + else { + queue_path = malloc(strlen(argv[1]) + 2); + if (!queue_path) { + perror("malloc()"); + exit(1); + } + queue_path[0] = '/'; + queue_path[1] = 0; + strcat(queue_path, argv[1]); + } + + if (getuid() != 0) { + fprintf(stderr, "Not running as root, but almost all tests " + "require root in order to modify\nsystem settings. " + "Exiting.\n"); + exit(1); + } + + /* Find out what files there are for us to make tweaks in */ + def_msgs = fopen(DEF_MSGS, "r+"); + def_msgsize = fopen(DEF_MSGSIZE, "r+"); + max_msgs = fopen(MAX_MSGS, "r+"); + max_msgsize = fopen(MAX_MSGSIZE, "r+"); + + if (!max_msgs) + shutdown(2, "Failed to open msg_max", __LINE__); + if (!max_msgsize) + shutdown(2, "Failed to open msgsize_max", __LINE__); + if (def_msgs || def_msgsize) + default_settings = 1; + + /* Load up the current system values for everything we can */ + getr(RLIMIT_MSGQUEUE, &saved_limits); + cur_limits = saved_limits; + if (default_settings) { + saved_def_msgs = cur_def_msgs = get(def_msgs); + saved_def_msgsize = cur_def_msgsize = get(def_msgsize); + } + saved_max_msgs = cur_max_msgs = get(max_msgs); + saved_max_msgsize = cur_max_msgsize = get(max_msgsize); + + /* Tell the user our initial state */ + printf("\nInitial system state:\n"); + printf("\tUsing queue path:\t\t%s\n", queue_path); + printf("\tRLIMIT_MSGQUEUE(soft):\t\t%d\n", saved_limits.rlim_cur); + printf("\tRLIMIT_MSGQUEUE(hard):\t\t%d\n", saved_limits.rlim_max); + printf("\tMaximum Message Size:\t\t%d\n", saved_max_msgsize); + printf("\tMaximum Queue Size:\t\t%d\n", saved_max_msgs); + if (default_settings) { + printf("\tDefault Message Size:\t\t%d\n", saved_def_msgsize); + printf("\tDefault Queue Size:\t\t%d\n", saved_def_msgs); + } else { + printf("\tDefault Message Size:\t\tNot Supported\n"); + printf("\tDefault Queue Size:\t\tNot Supported\n"); + } + printf("\n"); + + validate_current_settings(); + + printf("Adjusted system state for testing:\n"); + printf("\tRLIMIT_MSGQUEUE(soft):\t\t%d\n", cur_limits.rlim_cur); + printf("\tRLIMIT_MSGQUEUE(hard):\t\t%d\n", cur_limits.rlim_max); + printf("\tMaximum Message Size:\t\t%d\n", cur_max_msgsize); + printf("\tMaximum Queue Size:\t\t%d\n", cur_max_msgs); + if (default_settings) { + printf("\tDefault Message Size:\t\t%d\n", cur_def_msgsize); + printf("\tDefault Queue Size:\t\t%d\n", cur_def_msgs); + } + + printf("\n\nTest series 1, behavior when no attr struct " + "passed to mq_open:\n"); + if (!default_settings) { + test_queue(NULL, &result); + printf("Given sane system settings, mq_open without an attr " + "struct succeeds:\tPASS\n"); + if (result.mq_maxmsg != cur_max_msgs || + result.mq_msgsize != cur_max_msgsize) { + printf("Kernel does not support setting the default " + "mq attributes,\nbut also doesn't tie the " + "defaults to the maximums:\t\t\tPASS\n"); + } else { + set(max_msgs, ++cur_max_msgs); + set(max_msgsize, ++cur_max_msgsize); + test_queue(NULL, &result); + if (result.mq_maxmsg == cur_max_msgs && + result.mq_msgsize == cur_max_msgsize) + printf("Kernel does not support setting the " + "default mq attributes and\n" + "also ties system wide defaults to " + "the system wide maximums:\t\t" + "FAIL\n"); + else + printf("Kernel does not support setting the " + "default mq attributes,\n" + "but also doesn't tie the defaults to " + "the maximums:\t\t\tPASS\n"); + } + } else { + printf("Kernel supports setting defaults separately from " + "maximums:\t\tPASS\n"); + /* + * While we are here, go ahead and test that the kernel + * properly follows the default settings + */ + test_queue(NULL, &result); + printf("Given sane values, mq_open without an attr struct " + "succeeds:\t\tPASS\n"); + if (result.mq_maxmsg != cur_def_msgs || + result.mq_msgsize != cur_def_msgsize) + printf("Kernel supports setting defaults, but does " + "not actually honor them:\tFAIL\n\n"); + else { + set(def_msgs, ++cur_def_msgs); + set(def_msgsize, ++cur_def_msgsize); + /* In case max was the same as the default */ + set(max_msgs, ++cur_max_msgs); + set(max_msgsize, ++cur_max_msgsize); + test_queue(NULL, &result); + if (result.mq_maxmsg != cur_def_msgs || + result.mq_msgsize != cur_def_msgsize) + printf("Kernel supports setting defaults, but " + "does not actually honor them:\t" + "FAIL\n"); + else + printf("Kernel properly honors default setting " + "knobs:\t\t\t\tPASS\n"); + } + set(def_msgs, cur_max_msgs + 1); + cur_def_msgs = cur_max_msgs + 1; + set(def_msgsize, cur_max_msgsize + 1); + cur_def_msgsize = cur_max_msgsize + 1; + if (cur_def_msgs * (cur_def_msgsize + 2 * sizeof(void *)) >= + cur_limits.rlim_cur) { + cur_limits.rlim_cur = (cur_def_msgs + 2) * + (cur_def_msgsize + 2 * sizeof(void *)); + cur_limits.rlim_max = 2 * cur_limits.rlim_cur; + setr(RLIMIT_MSGQUEUE, &cur_limits); + } + if (test_queue_fail(NULL, &result)) { + if (result.mq_maxmsg == cur_max_msgs && + result.mq_msgsize == cur_max_msgsize) + printf("Kernel properly limits default values " + "to lesser of default/max:\t\tPASS\n"); + else + printf("Kernel does not properly set default " + "queue parameters when\ndefaults > " + "max:\t\t\t\t\t\t\t\tFAIL\n"); + } else + printf("Kernel fails to open mq because defaults are " + "greater than maximums:\tFAIL\n"); + set(def_msgs, --cur_def_msgs); + set(def_msgsize, --cur_def_msgsize); + cur_limits.rlim_cur = cur_limits.rlim_max = cur_def_msgs * + cur_def_msgsize; + setr(RLIMIT_MSGQUEUE, &cur_limits); + if (test_queue_fail(NULL, &result)) + printf("Kernel creates queue even though defaults " + "would exceed\nrlimit setting:" + "\t\t\t\t\t\t\t\tFAIL\n"); + else + printf("Kernel properly fails to create queue when " + "defaults would\nexceed rlimit:" + "\t\t\t\t\t\t\t\tPASS\n"); + } + + /* + * Test #2 - open with an attr struct that exceeds rlimit + */ + printf("\n\nTest series 2, behavior when attr struct is " + "passed to mq_open:\n"); + cur_max_msgs = 32; + cur_max_msgsize = cur_limits.rlim_max >> 4; + set(max_msgs, cur_max_msgs); + set(max_msgsize, cur_max_msgsize); + attr.mq_maxmsg = cur_max_msgs; + attr.mq_msgsize = cur_max_msgsize; + if (test_queue_fail(&attr, &result)) + printf("Queue open in excess of rlimit max when euid = 0 " + "succeeded:\t\tFAIL\n"); + else + printf("Queue open in excess of rlimit max when euid = 0 " + "failed:\t\tPASS\n"); + attr.mq_maxmsg = cur_max_msgs + 1; + attr.mq_msgsize = 10; + if (test_queue_fail(&attr, &result)) + printf("Queue open with mq_maxmsg > limit when euid = 0 " + "succeeded:\t\tPASS\n"); + else + printf("Queue open with mq_maxmsg > limit when euid = 0 " + "failed:\t\tFAIL\n"); + attr.mq_maxmsg = 1; + attr.mq_msgsize = cur_max_msgsize + 1; + if (test_queue_fail(&attr, &result)) + printf("Queue open with mq_msgsize > limit when euid = 0 " + "succeeded:\t\tPASS\n"); + else + printf("Queue open with mq_msgsize > limit when euid = 0 " + "failed:\t\tFAIL\n"); + attr.mq_maxmsg = 65536; + attr.mq_msgsize = 65536; + if (test_queue_fail(&attr, &result)) + printf("Queue open with total size > 2GB when euid = 0 " + "succeeded:\t\tFAIL\n"); + else + printf("Queue open with total size > 2GB when euid = 0 " + "failed:\t\t\tPASS\n"); + seteuid(99); + attr.mq_maxmsg = cur_max_msgs; + attr.mq_msgsize = cur_max_msgsize; + if (test_queue_fail(&attr, &result)) + printf("Queue open in excess of rlimit max when euid = 99 " + "succeeded:\t\tFAIL\n"); + else + printf("Queue open in excess of rlimit max when euid = 99 " + "failed:\t\tPASS\n"); + attr.mq_maxmsg = cur_max_msgs + 1; + attr.mq_msgsize = 10; + if (test_queue_fail(&attr, &result)) + printf("Queue open with mq_maxmsg > limit when euid = 99 " + "succeeded:\t\tFAIL\n"); + else + printf("Queue open with mq_maxmsg > limit when euid = 99 " + "failed:\t\tPASS\n"); + attr.mq_maxmsg = 1; + attr.mq_msgsize = cur_max_msgsize + 1; + if (test_queue_fail(&attr, &result)) + printf("Queue open with mq_msgsize > limit when euid = 99 " + "succeeded:\t\tFAIL\n"); + else + printf("Queue open with mq_msgsize > limit when euid = 99 " + "failed:\t\tPASS\n"); + attr.mq_maxmsg = 65536; + attr.mq_msgsize = 65536; + if (test_queue_fail(&attr, &result)) + printf("Queue open with total size > 2GB when euid = 99 " + "succeeded:\t\tFAIL\n"); + else + printf("Queue open with total size > 2GB when euid = 99 " + "failed:\t\t\tPASS\n"); + + shutdown(0,"",0); +} diff --git a/tools/testing/selftests/mqueue/mq_perf_tests.c b/tools/testing/selftests/mqueue/mq_perf_tests.c new file mode 100644 index 000000000000..2fadd4b97045 --- /dev/null +++ b/tools/testing/selftests/mqueue/mq_perf_tests.c @@ -0,0 +1,741 @@ +/* + * This application is Copyright 2012 Red Hat, Inc. + * Doug Ledford <dledford@redhat.com> + * + * mq_perf_tests is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, version 3. + * + * mq_perf_tests is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * For the full text of the license, see <http://www.gnu.org/licenses/>. + * + * mq_perf_tests.c + * Tests various types of message queue workloads, concentrating on those + * situations that invole large message sizes, large message queue depths, + * or both, and reports back useful metrics about kernel message queue + * performance. + * + */ +#define _GNU_SOURCE +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> +#include <fcntl.h> +#include <string.h> +#include <limits.h> +#include <errno.h> +#include <signal.h> +#include <pthread.h> +#include <sched.h> +#include <sys/types.h> +#include <sys/time.h> +#include <sys/resource.h> +#include <sys/stat.h> +#include <mqueue.h> +#include <popt.h> + +static char *usage = +"Usage:\n" +" %s [-c #[,#..] -f] path\n" +"\n" +" -c # Skip most tests and go straight to a high queue depth test\n" +" and then run that test continuously (useful for running at\n" +" the same time as some other workload to see how much the\n" +" cache thrashing caused by adding messages to a very deep\n" +" queue impacts the performance of other programs). The number\n" +" indicates which CPU core we should bind the process to during\n" +" the run. If you have more than one physical CPU, then you\n" +" will need one copy per physical CPU package, and you should\n" +" specify the CPU cores to pin ourself to via a comma separated\n" +" list of CPU values.\n" +" -f Only usable with continuous mode. Pin ourself to the CPUs\n" +" as requested, then instead of looping doing a high mq\n" +" workload, just busy loop. This will allow us to lock up a\n" +" single CPU just like we normally would, but without actually\n" +" thrashing the CPU cache. This is to make it easier to get\n" +" comparable numbers from some other workload running on the\n" +" other CPUs. One set of numbers with # CPUs locked up running\n" +" an mq workload, and another set of numbers with those same\n" +" CPUs locked away from the test workload, but not doing\n" +" anything to trash the cache like the mq workload might.\n" +" path Path name of the message queue to create\n" +"\n" +" Note: this program must be run as root in order to enable all tests\n" +"\n"; + +char *MAX_MSGS = "/proc/sys/fs/mqueue/msg_max"; +char *MAX_MSGSIZE = "/proc/sys/fs/mqueue/msgsize_max"; + +#define min(a, b) ((a) < (b) ? (a) : (b)) +#define MAX_CPUS 64 +char *cpu_option_string; +int cpus_to_pin[MAX_CPUS]; +int num_cpus_to_pin; +pthread_t cpu_threads[MAX_CPUS]; +pthread_t main_thread; +cpu_set_t *cpu_set; +int cpu_set_size; +int cpus_online; + +#define MSG_SIZE 16 +#define TEST1_LOOPS 10000000 +#define TEST2_LOOPS 100000 +int continuous_mode; +int continuous_mode_fake; + +struct rlimit saved_limits, cur_limits; +int saved_max_msgs, saved_max_msgsize; +int cur_max_msgs, cur_max_msgsize; +FILE *max_msgs, *max_msgsize; +int cur_nice; +char *queue_path = "/mq_perf_tests"; +mqd_t queue = -1; +struct mq_attr result; +int mq_prio_max; + +const struct poptOption options[] = { + { + .longName = "continuous", + .shortName = 'c', + .argInfo = POPT_ARG_STRING, + .arg = &cpu_option_string, + .val = 'c', + .descrip = "Run continuous tests at a high queue depth in " + "order to test the effects of cache thrashing on " + "other tasks on the system. This test is intended " + "to be run on one core of each physical CPU while " + "some other CPU intensive task is run on all the other " + "cores of that same physical CPU and the other task " + "is timed. It is assumed that the process of adding " + "messages to the message queue in a tight loop will " + "impact that other task to some degree. Once the " + "tests are performed in this way, you should then " + "re-run the tests using fake mode in order to check " + "the difference in time required to perform the CPU " + "intensive task", + .argDescrip = "cpu[,cpu]", + }, + { + .longName = "fake", + .shortName = 'f', + .argInfo = POPT_ARG_NONE, + .arg = &continuous_mode_fake, + .val = 0, + .descrip = "Tie up the CPUs that we would normally tie up in" + "continuous mode, but don't actually do any mq stuff, " + "just keep the CPU busy so it can't be used to process " + "system level tasks as this would free up resources on " + "the other CPU cores and skew the comparison between " + "the no-mqueue work and mqueue work tests", + .argDescrip = NULL, + }, + { + .longName = "path", + .shortName = 'p', + .argInfo = POPT_ARG_STRING | POPT_ARGFLAG_SHOW_DEFAULT, + .arg = &queue_path, + .val = 'p', + .descrip = "The name of the path to use in the mqueue " + "filesystem for our tests", + .argDescrip = "pathname", + }, + POPT_AUTOHELP + POPT_TABLEEND +}; + +static inline void __set(FILE *stream, int value, char *err_msg); +void shutdown(int exit_val, char *err_cause, int line_no); +void sig_action_SIGUSR1(int signum, siginfo_t *info, void *context); +void sig_action(int signum, siginfo_t *info, void *context); +static inline int get(FILE *stream); +static inline void set(FILE *stream, int value); +static inline int try_set(FILE *stream, int value); +static inline void getr(int type, struct rlimit *rlim); +static inline void setr(int type, struct rlimit *rlim); +static inline void open_queue(struct mq_attr *attr); +void increase_limits(void); + +static inline void __set(FILE *stream, int value, char *err_msg) +{ + rewind(stream); + if (fprintf(stream, "%d", value) < 0) + perror(err_msg); +} + + +void shutdown(int exit_val, char *err_cause, int line_no) +{ + static int in_shutdown = 0; + int errno_at_shutdown = errno; + int i; + + /* In case we get called by multiple threads or from an sighandler */ + if (in_shutdown++) + return; + + for (i = 0; i < num_cpus_to_pin; i++) + if (cpu_threads[i]) { + pthread_kill(cpu_threads[i], SIGUSR1); + pthread_join(cpu_threads[i], NULL); + } + + if (queue != -1) + if (mq_close(queue)) + perror("mq_close() during shutdown"); + if (queue_path) + /* + * Be silent if this fails, if we cleaned up already it's + * expected to fail + */ + mq_unlink(queue_path); + if (saved_max_msgs) + __set(max_msgs, saved_max_msgs, + "failed to restore saved_max_msgs"); + if (saved_max_msgsize) + __set(max_msgsize, saved_max_msgsize, + "failed to restore saved_max_msgsize"); + if (exit_val) + error(exit_val, errno_at_shutdown, "%s at %d", + err_cause, line_no); + exit(0); +} + +void sig_action_SIGUSR1(int signum, siginfo_t *info, void *context) +{ + if (pthread_self() != main_thread) + pthread_exit(0); + else { + fprintf(stderr, "Caught signal %d in SIGUSR1 handler, " + "exiting\n", signum); + shutdown(0, "", 0); + fprintf(stderr, "\n\nReturned from shutdown?!?!\n\n"); + exit(0); + } +} + +void sig_action(int signum, siginfo_t *info, void *context) +{ + if (pthread_self() != main_thread) + pthread_kill(main_thread, signum); + else { + fprintf(stderr, "Caught signal %d, exiting\n", signum); + shutdown(0, "", 0); + fprintf(stderr, "\n\nReturned from shutdown?!?!\n\n"); + exit(0); + } +} + +static inline int get(FILE *stream) +{ + int value; + rewind(stream); + if (fscanf(stream, "%d", &value) != 1) + shutdown(4, "Error reading /proc entry", __LINE__); + return value; +} + +static inline void set(FILE *stream, int value) +{ + int new_value; + + rewind(stream); + if (fprintf(stream, "%d", value) < 0) + return shutdown(5, "Failed writing to /proc file", __LINE__); + new_value = get(stream); + if (new_value != value) + return shutdown(5, "We didn't get what we wrote to /proc back", + __LINE__); +} + +static inline int try_set(FILE *stream, int value) +{ + int new_value; + + rewind(stream); + fprintf(stream, "%d", value); + new_value = get(stream); + return new_value == value; +} + +static inline void getr(int type, struct rlimit *rlim) +{ + if (getrlimit(type, rlim)) + shutdown(6, "getrlimit()", __LINE__); +} + +static inline void setr(int type, struct rlimit *rlim) +{ + if (setrlimit(type, rlim)) + shutdown(7, "setrlimit()", __LINE__); +} + +/** + * open_queue - open the global queue for testing + * @attr - An attr struct specifying the desired queue traits + * @result - An attr struct that lists the actual traits the queue has + * + * This open is not allowed to fail, failure will result in an orderly + * shutdown of the program. The global queue_path is used to set what + * queue to open, the queue descriptor is saved in the global queue + * variable. + */ +static inline void open_queue(struct mq_attr *attr) +{ + int flags = O_RDWR | O_EXCL | O_CREAT | O_NONBLOCK; + int perms = DEFFILEMODE; + + queue = mq_open(queue_path, flags, perms, attr); + if (queue == -1) + shutdown(1, "mq_open()", __LINE__); + if (mq_getattr(queue, &result)) + shutdown(1, "mq_getattr()", __LINE__); + printf("\n\tQueue %s created:\n", queue_path); + printf("\t\tmq_flags:\t\t\t%s\n", result.mq_flags & O_NONBLOCK ? + "O_NONBLOCK" : "(null)"); + printf("\t\tmq_maxmsg:\t\t\t%d\n", result.mq_maxmsg); + printf("\t\tmq_msgsize:\t\t\t%d\n", result.mq_msgsize); + printf("\t\tmq_curmsgs:\t\t\t%d\n", result.mq_curmsgs); +} + +void *fake_cont_thread(void *arg) +{ + int i; + + for (i = 0; i < num_cpus_to_pin; i++) + if (cpu_threads[i] == pthread_self()) + break; + printf("\tStarted fake continuous mode thread %d on CPU %d\n", i, + cpus_to_pin[i]); + while (1) + ; +} + +void *cont_thread(void *arg) +{ + char buff[MSG_SIZE]; + int i, priority; + + for (i = 0; i < num_cpus_to_pin; i++) + if (cpu_threads[i] == pthread_self()) + break; + printf("\tStarted continuous mode thread %d on CPU %d\n", i, + cpus_to_pin[i]); + while (1) { + while (mq_send(queue, buff, sizeof(buff), 0) == 0) + ; + mq_receive(queue, buff, sizeof(buff), &priority); + } +} + +#define drain_queue() \ + while (mq_receive(queue, buff, MSG_SIZE, &prio_in) == MSG_SIZE) + +#define do_untimed_send() \ + do { \ + if (mq_send(queue, buff, MSG_SIZE, prio_out)) \ + shutdown(3, "Test send failure", __LINE__); \ + } while (0) + +#define do_send_recv() \ + do { \ + clock_gettime(clock, &start); \ + if (mq_send(queue, buff, MSG_SIZE, prio_out)) \ + shutdown(3, "Test send failure", __LINE__); \ + clock_gettime(clock, &middle); \ + if (mq_receive(queue, buff, MSG_SIZE, &prio_in) != MSG_SIZE) \ + shutdown(3, "Test receive failure", __LINE__); \ + clock_gettime(clock, &end); \ + nsec = ((middle.tv_sec - start.tv_sec) * 1000000000) + \ + (middle.tv_nsec - start.tv_nsec); \ + send_total.tv_nsec += nsec; \ + if (send_total.tv_nsec >= 1000000000) { \ + send_total.tv_sec++; \ + send_total.tv_nsec -= 1000000000; \ + } \ + nsec = ((end.tv_sec - middle.tv_sec) * 1000000000) + \ + (end.tv_nsec - middle.tv_nsec); \ + recv_total.tv_nsec += nsec; \ + if (recv_total.tv_nsec >= 1000000000) { \ + recv_total.tv_sec++; \ + recv_total.tv_nsec -= 1000000000; \ + } \ + } while (0) + +struct test { + char *desc; + void (*func)(int *); +}; + +void const_prio(int *prio) +{ + return; +} + +void inc_prio(int *prio) +{ + if (++*prio == mq_prio_max) + *prio = 0; +} + +void dec_prio(int *prio) +{ + if (--*prio < 0) + *prio = mq_prio_max - 1; +} + +void random_prio(int *prio) +{ + *prio = random() % mq_prio_max; +} + +struct test test2[] = { + {"\n\tTest #2a: Time send/recv message, queue full, constant prio\n", + const_prio}, + {"\n\tTest #2b: Time send/recv message, queue full, increasing prio\n", + inc_prio}, + {"\n\tTest #2c: Time send/recv message, queue full, decreasing prio\n", + dec_prio}, + {"\n\tTest #2d: Time send/recv message, queue full, random prio\n", + random_prio}, + {NULL, NULL} +}; + +/** + * Tests to perform (all done with MSG_SIZE messages): + * + * 1) Time to add/remove message with 0 messages on queue + * 1a) with constant prio + * 2) Time to add/remove message when queue close to capacity: + * 2a) with constant prio + * 2b) with increasing prio + * 2c) with decreasing prio + * 2d) with random prio + * 3) Test limits of priorities honored (double check _SC_MQ_PRIO_MAX) + */ +void *perf_test_thread(void *arg) +{ + char buff[MSG_SIZE]; + int prio_out, prio_in; + int i; + clockid_t clock; + pthread_t *t; + struct timespec res, start, middle, end, send_total, recv_total; + unsigned long long nsec; + struct test *cur_test; + + t = &cpu_threads[0]; + printf("\n\tStarted mqueue performance test thread on CPU %d\n", + cpus_to_pin[0]); + mq_prio_max = sysconf(_SC_MQ_PRIO_MAX); + if (mq_prio_max == -1) + shutdown(2, "sysconf(_SC_MQ_PRIO_MAX)", __LINE__); + if (pthread_getcpuclockid(cpu_threads[0], &clock) != 0) + shutdown(2, "pthread_getcpuclockid", __LINE__); + + if (clock_getres(clock, &res)) + shutdown(2, "clock_getres()", __LINE__); + + printf("\t\tMax priorities:\t\t\t%d\n", mq_prio_max); + printf("\t\tClock resolution:\t\t%d nsec%s\n", res.tv_nsec, + res.tv_nsec > 1 ? "s" : ""); + + + + printf("\n\tTest #1: Time send/recv message, queue empty\n"); + printf("\t\t(%d iterations)\n", TEST1_LOOPS); + prio_out = 0; + send_total.tv_sec = 0; + send_total.tv_nsec = 0; + recv_total.tv_sec = 0; + recv_total.tv_nsec = 0; + for (i = 0; i < TEST1_LOOPS; i++) + do_send_recv(); + printf("\t\tSend msg:\t\t\t%d.%ds total time\n", + send_total.tv_sec, send_total.tv_nsec); + nsec = ((unsigned long long)send_total.tv_sec * 1000000000 + + send_total.tv_nsec) / TEST1_LOOPS; + printf("\t\t\t\t\t\t%d nsec/msg\n", nsec); + printf("\t\tRecv msg:\t\t\t%d.%ds total time\n", + recv_total.tv_sec, recv_total.tv_nsec); + nsec = ((unsigned long long)recv_total.tv_sec * 1000000000 + + recv_total.tv_nsec) / TEST1_LOOPS; + printf("\t\t\t\t\t\t%d nsec/msg\n", nsec); + + + for (cur_test = test2; cur_test->desc != NULL; cur_test++) { + printf(cur_test->desc); + printf("\t\t(%d iterations)\n", TEST2_LOOPS); + prio_out = 0; + send_total.tv_sec = 0; + send_total.tv_nsec = 0; + recv_total.tv_sec = 0; + recv_total.tv_nsec = 0; + printf("\t\tFilling queue..."); + fflush(stdout); + clock_gettime(clock, &start); + for (i = 0; i < result.mq_maxmsg - 1; i++) { + do_untimed_send(); + cur_test->func(&prio_out); + } + clock_gettime(clock, &end); + nsec = ((unsigned long long)(end.tv_sec - start.tv_sec) * + 1000000000) + (end.tv_nsec - start.tv_nsec); + printf("done.\t\t%lld.%llds\n", nsec / 1000000000, + nsec % 1000000000); + printf("\t\tTesting..."); + fflush(stdout); + for (i = 0; i < TEST2_LOOPS; i++) { + do_send_recv(); + cur_test->func(&prio_out); + } + printf("done.\n"); + printf("\t\tSend msg:\t\t\t%d.%ds total time\n", + send_total.tv_sec, send_total.tv_nsec); + nsec = ((unsigned long long)send_total.tv_sec * 1000000000 + + send_total.tv_nsec) / TEST2_LOOPS; + printf("\t\t\t\t\t\t%d nsec/msg\n", nsec); + printf("\t\tRecv msg:\t\t\t%d.%ds total time\n", + recv_total.tv_sec, recv_total.tv_nsec); + nsec = ((unsigned long long)recv_total.tv_sec * 1000000000 + + recv_total.tv_nsec) / TEST2_LOOPS; + printf("\t\t\t\t\t\t%d nsec/msg\n", nsec); + printf("\t\tDraining queue..."); + fflush(stdout); + clock_gettime(clock, &start); + drain_queue(); + clock_gettime(clock, &end); + nsec = ((unsigned long long)(end.tv_sec - start.tv_sec) * + 1000000000) + (end.tv_nsec - start.tv_nsec); + printf("done.\t\t%lld.%llds\n", nsec / 1000000000, + nsec % 1000000000); + } + return 0; +} + +void increase_limits(void) +{ + cur_limits.rlim_cur = RLIM_INFINITY; + cur_limits.rlim_max = RLIM_INFINITY; + setr(RLIMIT_MSGQUEUE, &cur_limits); + while (try_set(max_msgs, cur_max_msgs += 10)) + ; + cur_max_msgs = get(max_msgs); + while (try_set(max_msgsize, cur_max_msgsize += 1024)) + ; + cur_max_msgsize = get(max_msgsize); + if (setpriority(PRIO_PROCESS, 0, -20) != 0) + shutdown(2, "setpriority()", __LINE__); + cur_nice = -20; +} + +int main(int argc, char *argv[]) +{ + struct mq_attr attr; + char *option, *next_option; + int i, cpu; + struct sigaction sa; + poptContext popt_context; + char rc; + void *retval; + + main_thread = pthread_self(); + num_cpus_to_pin = 0; + + if (sysconf(_SC_NPROCESSORS_ONLN) == -1) { + perror("sysconf(_SC_NPROCESSORS_ONLN)"); + exit(1); + } + cpus_online = min(MAX_CPUS, sysconf(_SC_NPROCESSORS_ONLN)); + cpu_set = CPU_ALLOC(cpus_online); + if (cpu_set == NULL) { + perror("CPU_ALLOC()"); + exit(1); + } + cpu_set_size = CPU_ALLOC_SIZE(cpus_online); + CPU_ZERO_S(cpu_set_size, cpu_set); + + popt_context = poptGetContext(NULL, argc, (const char **)argv, + options, 0); + + while ((rc = poptGetNextOpt(popt_context)) > 0) { + switch (rc) { + case 'c': + continuous_mode = 1; + option = cpu_option_string; + do { + next_option = strchr(option, ','); + if (next_option) + *next_option = '\0'; + cpu = atoi(option); + if (cpu >= cpus_online) + fprintf(stderr, "CPU %d exceeds " + "cpus online, ignoring.\n", + cpu); + else + cpus_to_pin[num_cpus_to_pin++] = cpu; + if (next_option) + option = ++next_option; + } while (next_option && num_cpus_to_pin < MAX_CPUS); + /* Double check that they didn't give us the same CPU + * more than once */ + for (cpu = 0; cpu < num_cpus_to_pin; cpu++) { + if (CPU_ISSET_S(cpus_to_pin[cpu], cpu_set_size, + cpu_set)) { + fprintf(stderr, "Any given CPU may " + "only be given once.\n"); + exit(1); + } else + CPU_SET_S(cpus_to_pin[cpu], + cpu_set_size, cpu_set); + } + break; + case 'p': + /* + * Although we can create a msg queue with a + * non-absolute path name, unlink will fail. So, + * if the name doesn't start with a /, add one + * when we save it. + */ + option = queue_path; + if (*option != '/') { + queue_path = malloc(strlen(option) + 2); + if (!queue_path) { + perror("malloc()"); + exit(1); + } + queue_path[0] = '/'; + queue_path[1] = 0; + strcat(queue_path, option); + free(option); + } + break; + } + } + + if (continuous_mode && num_cpus_to_pin == 0) { + fprintf(stderr, "Must pass at least one CPU to continuous " + "mode.\n"); + poptPrintUsage(popt_context, stderr, 0); + exit(1); + } else if (!continuous_mode) { + num_cpus_to_pin = 1; + cpus_to_pin[0] = cpus_online - 1; + } + + if (getuid() != 0) { + fprintf(stderr, "Not running as root, but almost all tests " + "require root in order to modify\nsystem settings. " + "Exiting.\n"); + exit(1); + } + + max_msgs = fopen(MAX_MSGS, "r+"); + max_msgsize = fopen(MAX_MSGSIZE, "r+"); + if (!max_msgs) + shutdown(2, "Failed to open msg_max", __LINE__); + if (!max_msgsize) + shutdown(2, "Failed to open msgsize_max", __LINE__); + + /* Load up the current system values for everything we can */ + getr(RLIMIT_MSGQUEUE, &saved_limits); + cur_limits = saved_limits; + saved_max_msgs = cur_max_msgs = get(max_msgs); + saved_max_msgsize = cur_max_msgsize = get(max_msgsize); + errno = 0; + cur_nice = getpriority(PRIO_PROCESS, 0); + if (errno) + shutdown(2, "getpriority()", __LINE__); + + /* Tell the user our initial state */ + printf("\nInitial system state:\n"); + printf("\tUsing queue path:\t\t\t%s\n", queue_path); + printf("\tRLIMIT_MSGQUEUE(soft):\t\t\t%d\n", saved_limits.rlim_cur); + printf("\tRLIMIT_MSGQUEUE(hard):\t\t\t%d\n", saved_limits.rlim_max); + printf("\tMaximum Message Size:\t\t\t%d\n", saved_max_msgsize); + printf("\tMaximum Queue Size:\t\t\t%d\n", saved_max_msgs); + printf("\tNice value:\t\t\t\t%d\n", cur_nice); + printf("\n"); + + increase_limits(); + + printf("Adjusted system state for testing:\n"); + if (cur_limits.rlim_cur == RLIM_INFINITY) { + printf("\tRLIMIT_MSGQUEUE(soft):\t\t\t(unlimited)\n"); + printf("\tRLIMIT_MSGQUEUE(hard):\t\t\t(unlimited)\n"); + } else { + printf("\tRLIMIT_MSGQUEUE(soft):\t\t\t%d\n", + cur_limits.rlim_cur); + printf("\tRLIMIT_MSGQUEUE(hard):\t\t\t%d\n", + cur_limits.rlim_max); + } + printf("\tMaximum Message Size:\t\t\t%d\n", cur_max_msgsize); + printf("\tMaximum Queue Size:\t\t\t%d\n", cur_max_msgs); + printf("\tNice value:\t\t\t\t%d\n", cur_nice); + printf("\tContinuous mode:\t\t\t(%s)\n", continuous_mode ? + (continuous_mode_fake ? "fake mode" : "enabled") : + "disabled"); + printf("\tCPUs to pin:\t\t\t\t%d", cpus_to_pin[0]); + for (cpu = 1; cpu < num_cpus_to_pin; cpu++) + printf(",%d", cpus_to_pin[cpu]); + printf("\n"); + + sa.sa_sigaction = sig_action_SIGUSR1; + sigemptyset(&sa.sa_mask); + sigaddset(&sa.sa_mask, SIGHUP); + sigaddset(&sa.sa_mask, SIGINT); + sigaddset(&sa.sa_mask, SIGQUIT); + sigaddset(&sa.sa_mask, SIGTERM); + sa.sa_flags = SA_SIGINFO; + if (sigaction(SIGUSR1, &sa, NULL) == -1) + shutdown(1, "sigaction(SIGUSR1)", __LINE__); + sa.sa_sigaction = sig_action; + if (sigaction(SIGHUP, &sa, NULL) == -1) + shutdown(1, "sigaction(SIGHUP)", __LINE__); + if (sigaction(SIGINT, &sa, NULL) == -1) + shutdown(1, "sigaction(SIGINT)", __LINE__); + if (sigaction(SIGQUIT, &sa, NULL) == -1) + shutdown(1, "sigaction(SIGQUIT)", __LINE__); + if (sigaction(SIGTERM, &sa, NULL) == -1) + shutdown(1, "sigaction(SIGTERM)", __LINE__); + + if (!continuous_mode_fake) { + attr.mq_flags = O_NONBLOCK; + attr.mq_maxmsg = cur_max_msgs; + attr.mq_msgsize = MSG_SIZE; + open_queue(&attr); + } + for (i = 0; i < num_cpus_to_pin; i++) { + pthread_attr_t thread_attr; + void *thread_func; + + if (continuous_mode_fake) + thread_func = &fake_cont_thread; + else if (continuous_mode) + thread_func = &cont_thread; + else + thread_func = &perf_test_thread; + + CPU_ZERO_S(cpu_set_size, cpu_set); + CPU_SET_S(cpus_to_pin[i], cpu_set_size, cpu_set); + pthread_attr_init(&thread_attr); + pthread_attr_setaffinity_np(&thread_attr, cpu_set_size, + cpu_set); + if (pthread_create(&cpu_threads[i], &thread_attr, thread_func, + NULL)) + shutdown(1, "pthread_create()", __LINE__); + pthread_attr_destroy(&thread_attr); + } + + if (!continuous_mode) { + pthread_join(cpu_threads[0], &retval); + shutdown((long)retval, "perf_test_thread()", __LINE__); + } else { + while (1) + sleep(1); + } + shutdown(0, "", 0); +} diff --git a/tools/testing/selftests/run_tests b/tools/testing/selftests/run_tests deleted file mode 100644 index 320718a4e6bf..000000000000 --- a/tools/testing/selftests/run_tests +++ /dev/null @@ -1,8 +0,0 @@ -#!/bin/bash - -TARGETS=breakpoints - -for TARGET in $TARGETS -do - $TARGET/run_test -done diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile new file mode 100644 index 000000000000..b336b24aa6c0 --- /dev/null +++ b/tools/testing/selftests/vm/Makefile @@ -0,0 +1,14 @@ +# Makefile for vm selftests + +CC = $(CROSS_COMPILE)gcc +CFLAGS = -Wall -Wextra + +all: hugepage-mmap hugepage-shm map_hugetlb +%: %.c + $(CC) $(CFLAGS) -o $@ $^ + +run_tests: all + /bin/sh ./run_vmtests + +clean: + $(RM) hugepage-mmap hugepage-shm map_hugetlb diff --git a/tools/testing/selftests/vm/hugepage-mmap.c b/tools/testing/selftests/vm/hugepage-mmap.c new file mode 100644 index 000000000000..a10f310d2362 --- /dev/null +++ b/tools/testing/selftests/vm/hugepage-mmap.c @@ -0,0 +1,92 @@ +/* + * hugepage-mmap: + * + * Example of using huge page memory in a user application using the mmap + * system call. Before running this application, make sure that the + * administrator has mounted the hugetlbfs filesystem (on some directory + * like /mnt) using the command mount -t hugetlbfs nodev /mnt. In this + * example, the app is requesting memory of size 256MB that is backed by + * huge pages. + * + * For the ia64 architecture, the Linux kernel reserves Region number 4 for + * huge pages. That means that if one requires a fixed address, a huge page + * aligned address starting with 0x800000... will be required. If a fixed + * address is not required, the kernel will select an address in the proper + * range. + * Other architectures, such as ppc64, i386 or x86_64 are not so constrained. + */ + +#include <stdlib.h> +#include <stdio.h> +#include <unistd.h> +#include <sys/mman.h> +#include <fcntl.h> + +#define FILE_NAME "huge/hugepagefile" +#define LENGTH (256UL*1024*1024) +#define PROTECTION (PROT_READ | PROT_WRITE) + +/* Only ia64 requires this */ +#ifdef __ia64__ +#define ADDR (void *)(0x8000000000000000UL) +#define FLAGS (MAP_SHARED | MAP_FIXED) +#else +#define ADDR (void *)(0x0UL) +#define FLAGS (MAP_SHARED) +#endif + +static void check_bytes(char *addr) +{ + printf("First hex is %x\n", *((unsigned int *)addr)); +} + +static void write_bytes(char *addr) +{ + unsigned long i; + + for (i = 0; i < LENGTH; i++) + *(addr + i) = (char)i; +} + +static int read_bytes(char *addr) +{ + unsigned long i; + + check_bytes(addr); + for (i = 0; i < LENGTH; i++) + if (*(addr + i) != (char)i) { + printf("Mismatch at %lu\n", i); + return 1; + } + return 0; +} + +int main(void) +{ + void *addr; + int fd, ret; + + fd = open(FILE_NAME, O_CREAT | O_RDWR, 0755); + if (fd < 0) { + perror("Open failed"); + exit(1); + } + + addr = mmap(ADDR, LENGTH, PROTECTION, FLAGS, fd, 0); + if (addr == MAP_FAILED) { + perror("mmap"); + unlink(FILE_NAME); + exit(1); + } + + printf("Returned address is %p\n", addr); + check_bytes(addr); + write_bytes(addr); + ret = read_bytes(addr); + + munmap(addr, LENGTH); + close(fd); + unlink(FILE_NAME); + + return ret; +} diff --git a/tools/testing/selftests/vm/hugepage-shm.c b/tools/testing/selftests/vm/hugepage-shm.c new file mode 100644 index 000000000000..0d0ef4fc0c04 --- /dev/null +++ b/tools/testing/selftests/vm/hugepage-shm.c @@ -0,0 +1,100 @@ +/* + * hugepage-shm: + * + * Example of using huge page memory in a user application using Sys V shared + * memory system calls. In this example the app is requesting 256MB of + * memory that is backed by huge pages. The application uses the flag + * SHM_HUGETLB in the shmget system call to inform the kernel that it is + * requesting huge pages. + * + * For the ia64 architecture, the Linux kernel reserves Region number 4 for + * huge pages. That means that if one requires a fixed address, a huge page + * aligned address starting with 0x800000... will be required. If a fixed + * address is not required, the kernel will select an address in the proper + * range. + * Other architectures, such as ppc64, i386 or x86_64 are not so constrained. + * + * Note: The default shared memory limit is quite low on many kernels, + * you may need to increase it via: + * + * echo 268435456 > /proc/sys/kernel/shmmax + * + * This will increase the maximum size per shared memory segment to 256MB. + * The other limit that you will hit eventually is shmall which is the + * total amount of shared memory in pages. To set it to 16GB on a system + * with a 4kB pagesize do: + * + * echo 4194304 > /proc/sys/kernel/shmall + */ + +#include <stdlib.h> +#include <stdio.h> +#include <sys/types.h> +#include <sys/ipc.h> +#include <sys/shm.h> +#include <sys/mman.h> + +#ifndef SHM_HUGETLB +#define SHM_HUGETLB 04000 +#endif + +#define LENGTH (256UL*1024*1024) + +#define dprintf(x) printf(x) + +/* Only ia64 requires this */ +#ifdef __ia64__ +#define ADDR (void *)(0x8000000000000000UL) +#define SHMAT_FLAGS (SHM_RND) +#else +#define ADDR (void *)(0x0UL) +#define SHMAT_FLAGS (0) +#endif + +int main(void) +{ + int shmid; + unsigned long i; + char *shmaddr; + + shmid = shmget(2, LENGTH, SHM_HUGETLB | IPC_CREAT | SHM_R | SHM_W); + if (shmid < 0) { + perror("shmget"); + exit(1); + } + printf("shmid: 0x%x\n", shmid); + + shmaddr = shmat(shmid, ADDR, SHMAT_FLAGS); + if (shmaddr == (char *)-1) { + perror("Shared memory attach failure"); + shmctl(shmid, IPC_RMID, NULL); + exit(2); + } + printf("shmaddr: %p\n", shmaddr); + + dprintf("Starting the writes:\n"); + for (i = 0; i < LENGTH; i++) { + shmaddr[i] = (char)(i); + if (!(i % (1024 * 1024))) + dprintf("."); + } + dprintf("\n"); + + dprintf("Starting the Check..."); + for (i = 0; i < LENGTH; i++) + if (shmaddr[i] != (char)i) { + printf("\nIndex %lu mismatched\n", i); + exit(3); + } + dprintf("Done.\n"); + + if (shmdt((const void *)shmaddr) != 0) { + perror("Detach failure"); + shmctl(shmid, IPC_RMID, NULL); + exit(4); + } + + shmctl(shmid, IPC_RMID, NULL); + + return 0; +} diff --git a/tools/testing/selftests/vm/map_hugetlb.c b/tools/testing/selftests/vm/map_hugetlb.c new file mode 100644 index 000000000000..ac56639dd4a9 --- /dev/null +++ b/tools/testing/selftests/vm/map_hugetlb.c @@ -0,0 +1,79 @@ +/* + * Example of using hugepage memory in a user application using the mmap + * system call with MAP_HUGETLB flag. Before running this program make + * sure the administrator has allocated enough default sized huge pages + * to cover the 256 MB allocation. + * + * For ia64 architecture, Linux kernel reserves Region number 4 for hugepages. + * That means the addresses starting with 0x800000... will need to be + * specified. Specifying a fixed address is not required on ppc64, i386 + * or x86_64. + */ +#include <stdlib.h> +#include <stdio.h> +#include <unistd.h> +#include <sys/mman.h> +#include <fcntl.h> + +#define LENGTH (256UL*1024*1024) +#define PROTECTION (PROT_READ | PROT_WRITE) + +#ifndef MAP_HUGETLB +#define MAP_HUGETLB 0x40000 /* arch specific */ +#endif + +/* Only ia64 requires this */ +#ifdef __ia64__ +#define ADDR (void *)(0x8000000000000000UL) +#define FLAGS (MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB | MAP_FIXED) +#else +#define ADDR (void *)(0x0UL) +#define FLAGS (MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB) +#endif + +static void check_bytes(char *addr) +{ + printf("First hex is %x\n", *((unsigned int *)addr)); +} + +static void write_bytes(char *addr) +{ + unsigned long i; + + for (i = 0; i < LENGTH; i++) + *(addr + i) = (char)i; +} + +static int read_bytes(char *addr) +{ + unsigned long i; + + check_bytes(addr); + for (i = 0; i < LENGTH; i++) + if (*(addr + i) != (char)i) { + printf("Mismatch at %lu\n", i); + return 1; + } + return 0; +} + +int main(void) +{ + void *addr; + int ret; + + addr = mmap(ADDR, LENGTH, PROTECTION, FLAGS, 0, 0); + if (addr == MAP_FAILED) { + perror("mmap"); + exit(1); + } + + printf("Returned address is %p\n", addr); + check_bytes(addr); + write_bytes(addr); + ret = read_bytes(addr); + + munmap(addr, LENGTH); + + return ret; +} diff --git a/tools/testing/selftests/vm/run_vmtests b/tools/testing/selftests/vm/run_vmtests new file mode 100644 index 000000000000..8b40bd5e5cc2 --- /dev/null +++ b/tools/testing/selftests/vm/run_vmtests @@ -0,0 +1,77 @@ +#!/bin/bash +#please run as root + +#we need 256M, below is the size in kB +needmem=262144 +mnt=./huge + +#get pagesize and freepages from /proc/meminfo +while read name size unit; do + if [ "$name" = "HugePages_Free:" ]; then + freepgs=$size + fi + if [ "$name" = "Hugepagesize:" ]; then + pgsize=$size + fi +done < /proc/meminfo + +#set proper nr_hugepages +if [ -n "$freepgs" ] && [ -n "$pgsize" ]; then + nr_hugepgs=`cat /proc/sys/vm/nr_hugepages` + needpgs=`expr $needmem / $pgsize` + if [ $freepgs -lt $needpgs ]; then + lackpgs=$(( $needpgs - $freepgs )) + echo $(( $lackpgs + $nr_hugepgs )) > /proc/sys/vm/nr_hugepages + if [ $? -ne 0 ]; then + echo "Please run this test as root" + exit 1 + fi + fi +else + echo "no hugetlbfs support in kernel?" + exit 1 +fi + +mkdir $mnt +mount -t hugetlbfs none $mnt + +echo "--------------------" +echo "runing hugepage-mmap" +echo "--------------------" +./hugepage-mmap +if [ $? -ne 0 ]; then + echo "[FAIL]" +else + echo "[PASS]" +fi + +shmmax=`cat /proc/sys/kernel/shmmax` +shmall=`cat /proc/sys/kernel/shmall` +echo 268435456 > /proc/sys/kernel/shmmax +echo 4194304 > /proc/sys/kernel/shmall +echo "--------------------" +echo "runing hugepage-shm" +echo "--------------------" +./hugepage-shm +if [ $? -ne 0 ]; then + echo "[FAIL]" +else + echo "[PASS]" +fi +echo $shmmax > /proc/sys/kernel/shmmax +echo $shmall > /proc/sys/kernel/shmall + +echo "--------------------" +echo "runing map_hugetlb" +echo "--------------------" +./map_hugetlb +if [ $? -ne 0 ]; then + echo "[FAIL]" +else + echo "[PASS]" +fi + +#cleanup +umount $mnt +rm -rf $mnt +echo $nr_hugepgs > /proc/sys/vm/nr_hugepages diff --git a/tools/usb/ffs-test.c b/tools/usb/ffs-test.c index 4b107b5e623f..8674b9ec14f6 100644 --- a/tools/usb/ffs-test.c +++ b/tools/usb/ffs-test.c @@ -297,7 +297,7 @@ static void *start_thread_helper(void *arg) ret = t->in(t, t->buf, t->buf_size); if (ret > 0) { - ret = t->out(t, t->buf, t->buf_size); + ret = t->out(t, t->buf, ret); name = out_name; op = "write"; } else { diff --git a/tools/usb/testusb.c b/tools/usb/testusb.c index 6e0f56701e44..82d7c590c026 100644 --- a/tools/usb/testusb.c +++ b/tools/usb/testusb.c @@ -358,6 +358,7 @@ static const char *usbfs_dir_find(void) { static char usbfs_path_0[] = "/dev/usb/devices"; static char usbfs_path_1[] = "/proc/bus/usb/devices"; + static char udev_usb_path[] = "/dev/bus/usb"; static char *const usbfs_paths[] = { usbfs_path_0, usbfs_path_1 @@ -376,6 +377,10 @@ static const char *usbfs_dir_find(void) } } while (++it != end); + /* real device-nodes managed by udev */ + if (access(udev_usb_path, F_OK) == 0) + return udev_usb_path; + return NULL; } diff --git a/tools/virtio/linux/virtio.h b/tools/virtio/linux/virtio.h index 7579f19e61e0..81847dd08bd0 100644 --- a/tools/virtio/linux/virtio.h +++ b/tools/virtio/linux/virtio.h @@ -203,6 +203,7 @@ void *virtqueue_get_buf(struct virtqueue *vq, unsigned int *len); void virtqueue_disable_cb(struct virtqueue *vq); bool virtqueue_enable_cb(struct virtqueue *vq); +bool virtqueue_enable_cb_delayed(struct virtqueue *vq); void *virtqueue_detach_unused_buf(struct virtqueue *vq); struct virtqueue *vring_new_virtqueue(unsigned int num, diff --git a/tools/virtio/virtio_test.c b/tools/virtio/virtio_test.c index 6bf95f995364..e626fa553c5a 100644 --- a/tools/virtio/virtio_test.c +++ b/tools/virtio/virtio_test.c @@ -144,7 +144,8 @@ static void wait_for_interrupt(struct vdev_info *dev) } } -static void run_test(struct vdev_info *dev, struct vq_info *vq, int bufs) +static void run_test(struct vdev_info *dev, struct vq_info *vq, + bool delayed, int bufs) { struct scatterlist sl; long started = 0, completed = 0; @@ -183,8 +184,12 @@ static void run_test(struct vdev_info *dev, struct vq_info *vq, int bufs) assert(started <= bufs); if (completed == bufs) break; - if (virtqueue_enable_cb(vq->vq)) { - wait_for_interrupt(dev); + if (delayed) { + if (virtqueue_enable_cb_delayed(vq->vq)) + wait_for_interrupt(dev); + } else { + if (virtqueue_enable_cb(vq->vq)) + wait_for_interrupt(dev); } } test = 0; @@ -216,6 +221,14 @@ const struct option longopts[] = { .val = 'i', }, { + .name = "delayed-interrupt", + .val = 'D', + }, + { + .name = "no-delayed-interrupt", + .val = 'd', + }, + { } }; @@ -224,6 +237,7 @@ static void help() fprintf(stderr, "Usage: virtio_test [--help]" " [--no-indirect]" " [--no-event-idx]" + " [--delayed-interrupt]" "\n"); } @@ -233,6 +247,7 @@ int main(int argc, char **argv) unsigned long long features = (1ULL << VIRTIO_RING_F_INDIRECT_DESC) | (1ULL << VIRTIO_RING_F_EVENT_IDX); int o; + bool delayed = false; for (;;) { o = getopt_long(argc, argv, optstring, longopts, NULL); @@ -251,6 +266,9 @@ int main(int argc, char **argv) case 'i': features &= ~(1ULL << VIRTIO_RING_F_INDIRECT_DESC); break; + case 'D': + delayed = true; + break; default: assert(0); break; @@ -260,6 +278,6 @@ int main(int argc, char **argv) done: vdev_info_init(&dev, features); vq_info_add(&dev, 256); - run_test(&dev, &dev.vqs[0], 0x100000); + run_test(&dev, &dev.vqs[0], delayed, 0x100000); return 0; } diff --git a/tools/vm/Makefile b/tools/vm/Makefile new file mode 100644 index 000000000000..8e30e5c40f8a --- /dev/null +++ b/tools/vm/Makefile @@ -0,0 +1,11 @@ +# Makefile for vm tools + +CC = $(CROSS_COMPILE)gcc +CFLAGS = -Wall -Wextra + +all: page-types slabinfo +%: %.c + $(CC) $(CFLAGS) -o $@ $^ + +clean: + $(RM) page-types slabinfo diff --git a/tools/vm/page-types.c b/tools/vm/page-types.c new file mode 100644 index 000000000000..f576971f6556 --- /dev/null +++ b/tools/vm/page-types.c @@ -0,0 +1,1076 @@ +/* + * page-types: Tool for querying page flags + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; version 2. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should find a copy of v2 of the GNU General Public License somewhere on + * your Linux system; if not, write to the Free Software Foundation, Inc., 59 + * Temple Place, Suite 330, Boston, MA 02111-1307 USA. + * + * Copyright (C) 2009 Intel corporation + * + * Authors: Wu Fengguang <fengguang.wu@intel.com> + */ + +#define _LARGEFILE64_SOURCE +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> +#include <stdint.h> +#include <stdarg.h> +#include <string.h> +#include <getopt.h> +#include <limits.h> +#include <assert.h> +#include <sys/types.h> +#include <sys/errno.h> +#include <sys/fcntl.h> +#include <sys/mount.h> +#include <sys/statfs.h> +#include "../../include/linux/magic.h" +#include "../../include/linux/kernel-page-flags.h" + + +#ifndef MAX_PATH +# define MAX_PATH 256 +#endif + +#ifndef STR +# define _STR(x) #x +# define STR(x) _STR(x) +#endif + +/* + * pagemap kernel ABI bits + */ + +#define PM_ENTRY_BYTES sizeof(uint64_t) +#define PM_STATUS_BITS 3 +#define PM_STATUS_OFFSET (64 - PM_STATUS_BITS) +#define PM_STATUS_MASK (((1LL << PM_STATUS_BITS) - 1) << PM_STATUS_OFFSET) +#define PM_STATUS(nr) (((nr) << PM_STATUS_OFFSET) & PM_STATUS_MASK) +#define PM_PSHIFT_BITS 6 +#define PM_PSHIFT_OFFSET (PM_STATUS_OFFSET - PM_PSHIFT_BITS) +#define PM_PSHIFT_MASK (((1LL << PM_PSHIFT_BITS) - 1) << PM_PSHIFT_OFFSET) +#define PM_PSHIFT(x) (((u64) (x) << PM_PSHIFT_OFFSET) & PM_PSHIFT_MASK) +#define PM_PFRAME_MASK ((1LL << PM_PSHIFT_OFFSET) - 1) +#define PM_PFRAME(x) ((x) & PM_PFRAME_MASK) + +#define PM_PRESENT PM_STATUS(4LL) +#define PM_SWAP PM_STATUS(2LL) + + +/* + * kernel page flags + */ + +#define KPF_BYTES 8 +#define PROC_KPAGEFLAGS "/proc/kpageflags" + +/* [32-] kernel hacking assistances */ +#define KPF_RESERVED 32 +#define KPF_MLOCKED 33 +#define KPF_MAPPEDTODISK 34 +#define KPF_PRIVATE 35 +#define KPF_PRIVATE_2 36 +#define KPF_OWNER_PRIVATE 37 +#define KPF_ARCH 38 +#define KPF_UNCACHED 39 + +/* [48-] take some arbitrary free slots for expanding overloaded flags + * not part of kernel API + */ +#define KPF_READAHEAD 48 +#define KPF_SLOB_FREE 49 +#define KPF_SLUB_FROZEN 50 +#define KPF_SLUB_DEBUG 51 + +#define KPF_ALL_BITS ((uint64_t)~0ULL) +#define KPF_HACKERS_BITS (0xffffULL << 32) +#define KPF_OVERLOADED_BITS (0xffffULL << 48) +#define BIT(name) (1ULL << KPF_##name) +#define BITS_COMPOUND (BIT(COMPOUND_HEAD) | BIT(COMPOUND_TAIL)) + +static const char * const page_flag_names[] = { + [KPF_LOCKED] = "L:locked", + [KPF_ERROR] = "E:error", + [KPF_REFERENCED] = "R:referenced", + [KPF_UPTODATE] = "U:uptodate", + [KPF_DIRTY] = "D:dirty", + [KPF_LRU] = "l:lru", + [KPF_ACTIVE] = "A:active", + [KPF_SLAB] = "S:slab", + [KPF_WRITEBACK] = "W:writeback", + [KPF_RECLAIM] = "I:reclaim", + [KPF_BUDDY] = "B:buddy", + + [KPF_MMAP] = "M:mmap", + [KPF_ANON] = "a:anonymous", + [KPF_SWAPCACHE] = "s:swapcache", + [KPF_SWAPBACKED] = "b:swapbacked", + [KPF_COMPOUND_HEAD] = "H:compound_head", + [KPF_COMPOUND_TAIL] = "T:compound_tail", + [KPF_HUGE] = "G:huge", + [KPF_UNEVICTABLE] = "u:unevictable", + [KPF_HWPOISON] = "X:hwpoison", + [KPF_NOPAGE] = "n:nopage", + [KPF_KSM] = "x:ksm", + [KPF_THP] = "t:thp", + + [KPF_RESERVED] = "r:reserved", + [KPF_MLOCKED] = "m:mlocked", + [KPF_MAPPEDTODISK] = "d:mappedtodisk", + [KPF_PRIVATE] = "P:private", + [KPF_PRIVATE_2] = "p:private_2", + [KPF_OWNER_PRIVATE] = "O:owner_private", + [KPF_ARCH] = "h:arch", + [KPF_UNCACHED] = "c:uncached", + + [KPF_READAHEAD] = "I:readahead", + [KPF_SLOB_FREE] = "P:slob_free", + [KPF_SLUB_FROZEN] = "A:slub_frozen", + [KPF_SLUB_DEBUG] = "E:slub_debug", +}; + + +static const char * const debugfs_known_mountpoints[] = { + "/sys/kernel/debug", + "/debug", + 0, +}; + +/* + * data structures + */ + +static int opt_raw; /* for kernel developers */ +static int opt_list; /* list pages (in ranges) */ +static int opt_no_summary; /* don't show summary */ +static pid_t opt_pid; /* process to walk */ + +#define MAX_ADDR_RANGES 1024 +static int nr_addr_ranges; +static unsigned long opt_offset[MAX_ADDR_RANGES]; +static unsigned long opt_size[MAX_ADDR_RANGES]; + +#define MAX_VMAS 10240 +static int nr_vmas; +static unsigned long pg_start[MAX_VMAS]; +static unsigned long pg_end[MAX_VMAS]; + +#define MAX_BIT_FILTERS 64 +static int nr_bit_filters; +static uint64_t opt_mask[MAX_BIT_FILTERS]; +static uint64_t opt_bits[MAX_BIT_FILTERS]; + +static int page_size; + +static int pagemap_fd; +static int kpageflags_fd; + +static int opt_hwpoison; +static int opt_unpoison; + +static char hwpoison_debug_fs[MAX_PATH+1]; +static int hwpoison_inject_fd; +static int hwpoison_forget_fd; + +#define HASH_SHIFT 13 +#define HASH_SIZE (1 << HASH_SHIFT) +#define HASH_MASK (HASH_SIZE - 1) +#define HASH_KEY(flags) (flags & HASH_MASK) + +static unsigned long total_pages; +static unsigned long nr_pages[HASH_SIZE]; +static uint64_t page_flags[HASH_SIZE]; + + +/* + * helper functions + */ + +#define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0])) + +#define min_t(type, x, y) ({ \ + type __min1 = (x); \ + type __min2 = (y); \ + __min1 < __min2 ? __min1 : __min2; }) + +#define max_t(type, x, y) ({ \ + type __max1 = (x); \ + type __max2 = (y); \ + __max1 > __max2 ? __max1 : __max2; }) + +static unsigned long pages2mb(unsigned long pages) +{ + return (pages * page_size) >> 20; +} + +static void fatal(const char *x, ...) +{ + va_list ap; + + va_start(ap, x); + vfprintf(stderr, x, ap); + va_end(ap); + exit(EXIT_FAILURE); +} + +static int checked_open(const char *pathname, int flags) +{ + int fd = open(pathname, flags); + + if (fd < 0) { + perror(pathname); + exit(EXIT_FAILURE); + } + + return fd; +} + +/* + * pagemap/kpageflags routines + */ + +static unsigned long do_u64_read(int fd, char *name, + uint64_t *buf, + unsigned long index, + unsigned long count) +{ + long bytes; + + if (index > ULONG_MAX / 8) + fatal("index overflow: %lu\n", index); + + if (lseek(fd, index * 8, SEEK_SET) < 0) { + perror(name); + exit(EXIT_FAILURE); + } + + bytes = read(fd, buf, count * 8); + if (bytes < 0) { + perror(name); + exit(EXIT_FAILURE); + } + if (bytes % 8) + fatal("partial read: %lu bytes\n", bytes); + + return bytes / 8; +} + +static unsigned long kpageflags_read(uint64_t *buf, + unsigned long index, + unsigned long pages) +{ + return do_u64_read(kpageflags_fd, PROC_KPAGEFLAGS, buf, index, pages); +} + +static unsigned long pagemap_read(uint64_t *buf, + unsigned long index, + unsigned long pages) +{ + return do_u64_read(pagemap_fd, "/proc/pid/pagemap", buf, index, pages); +} + +static unsigned long pagemap_pfn(uint64_t val) +{ + unsigned long pfn; + + if (val & PM_PRESENT) + pfn = PM_PFRAME(val); + else + pfn = 0; + + return pfn; +} + + +/* + * page flag names + */ + +static char *page_flag_name(uint64_t flags) +{ + static char buf[65]; + int present; + size_t i, j; + + for (i = 0, j = 0; i < ARRAY_SIZE(page_flag_names); i++) { + present = (flags >> i) & 1; + if (!page_flag_names[i]) { + if (present) + fatal("unknown flag bit %d\n", i); + continue; + } + buf[j++] = present ? page_flag_names[i][0] : '_'; + } + + return buf; +} + +static char *page_flag_longname(uint64_t flags) +{ + static char buf[1024]; + size_t i, n; + + for (i = 0, n = 0; i < ARRAY_SIZE(page_flag_names); i++) { + if (!page_flag_names[i]) + continue; + if ((flags >> i) & 1) + n += snprintf(buf + n, sizeof(buf) - n, "%s,", + page_flag_names[i] + 2); + } + if (n) + n--; + buf[n] = '\0'; + + return buf; +} + + +/* + * page list and summary + */ + +static void show_page_range(unsigned long voffset, + unsigned long offset, uint64_t flags) +{ + static uint64_t flags0; + static unsigned long voff; + static unsigned long index; + static unsigned long count; + + if (flags == flags0 && offset == index + count && + (!opt_pid || voffset == voff + count)) { + count++; + return; + } + + if (count) { + if (opt_pid) + printf("%lx\t", voff); + printf("%lx\t%lx\t%s\n", + index, count, page_flag_name(flags0)); + } + + flags0 = flags; + index = offset; + voff = voffset; + count = 1; +} + +static void show_page(unsigned long voffset, + unsigned long offset, uint64_t flags) +{ + if (opt_pid) + printf("%lx\t", voffset); + printf("%lx\t%s\n", offset, page_flag_name(flags)); +} + +static void show_summary(void) +{ + size_t i; + + printf(" flags\tpage-count MB" + " symbolic-flags\t\t\tlong-symbolic-flags\n"); + + for (i = 0; i < ARRAY_SIZE(nr_pages); i++) { + if (nr_pages[i]) + printf("0x%016llx\t%10lu %8lu %s\t%s\n", + (unsigned long long)page_flags[i], + nr_pages[i], + pages2mb(nr_pages[i]), + page_flag_name(page_flags[i]), + page_flag_longname(page_flags[i])); + } + + printf(" total\t%10lu %8lu\n", + total_pages, pages2mb(total_pages)); +} + + +/* + * page flag filters + */ + +static int bit_mask_ok(uint64_t flags) +{ + int i; + + for (i = 0; i < nr_bit_filters; i++) { + if (opt_bits[i] == KPF_ALL_BITS) { + if ((flags & opt_mask[i]) == 0) + return 0; + } else { + if ((flags & opt_mask[i]) != opt_bits[i]) + return 0; + } + } + + return 1; +} + +static uint64_t expand_overloaded_flags(uint64_t flags) +{ + /* SLOB/SLUB overload several page flags */ + if (flags & BIT(SLAB)) { + if (flags & BIT(PRIVATE)) + flags ^= BIT(PRIVATE) | BIT(SLOB_FREE); + if (flags & BIT(ACTIVE)) + flags ^= BIT(ACTIVE) | BIT(SLUB_FROZEN); + if (flags & BIT(ERROR)) + flags ^= BIT(ERROR) | BIT(SLUB_DEBUG); + } + + /* PG_reclaim is overloaded as PG_readahead in the read path */ + if ((flags & (BIT(RECLAIM) | BIT(WRITEBACK))) == BIT(RECLAIM)) + flags ^= BIT(RECLAIM) | BIT(READAHEAD); + + return flags; +} + +static uint64_t well_known_flags(uint64_t flags) +{ + /* hide flags intended only for kernel hacker */ + flags &= ~KPF_HACKERS_BITS; + + /* hide non-hugeTLB compound pages */ + if ((flags & BITS_COMPOUND) && !(flags & BIT(HUGE))) + flags &= ~BITS_COMPOUND; + + return flags; +} + +static uint64_t kpageflags_flags(uint64_t flags) +{ + flags = expand_overloaded_flags(flags); + + if (!opt_raw) + flags = well_known_flags(flags); + + return flags; +} + +/* verify that a mountpoint is actually a debugfs instance */ +static int debugfs_valid_mountpoint(const char *debugfs) +{ + struct statfs st_fs; + + if (statfs(debugfs, &st_fs) < 0) + return -ENOENT; + else if (st_fs.f_type != (long) DEBUGFS_MAGIC) + return -ENOENT; + + return 0; +} + +/* find the path to the mounted debugfs */ +static const char *debugfs_find_mountpoint(void) +{ + const char *const *ptr; + char type[100]; + FILE *fp; + + ptr = debugfs_known_mountpoints; + while (*ptr) { + if (debugfs_valid_mountpoint(*ptr) == 0) { + strcpy(hwpoison_debug_fs, *ptr); + return hwpoison_debug_fs; + } + ptr++; + } + + /* give up and parse /proc/mounts */ + fp = fopen("/proc/mounts", "r"); + if (fp == NULL) + perror("Can't open /proc/mounts for read"); + + while (fscanf(fp, "%*s %" + STR(MAX_PATH) + "s %99s %*s %*d %*d\n", + hwpoison_debug_fs, type) == 2) { + if (strcmp(type, "debugfs") == 0) + break; + } + fclose(fp); + + if (strcmp(type, "debugfs") != 0) + return NULL; + + return hwpoison_debug_fs; +} + +/* mount the debugfs somewhere if it's not mounted */ + +static void debugfs_mount(void) +{ + const char *const *ptr; + + /* see if it's already mounted */ + if (debugfs_find_mountpoint()) + return; + + ptr = debugfs_known_mountpoints; + while (*ptr) { + if (mount(NULL, *ptr, "debugfs", 0, NULL) == 0) { + /* save the mountpoint */ + strcpy(hwpoison_debug_fs, *ptr); + break; + } + ptr++; + } + + if (*ptr == NULL) { + perror("mount debugfs"); + exit(EXIT_FAILURE); + } +} + +/* + * page actions + */ + +static void prepare_hwpoison_fd(void) +{ + char buf[MAX_PATH + 1]; + + debugfs_mount(); + + if (opt_hwpoison && !hwpoison_inject_fd) { + snprintf(buf, MAX_PATH, "%s/hwpoison/corrupt-pfn", + hwpoison_debug_fs); + hwpoison_inject_fd = checked_open(buf, O_WRONLY); + } + + if (opt_unpoison && !hwpoison_forget_fd) { + snprintf(buf, MAX_PATH, "%s/hwpoison/unpoison-pfn", + hwpoison_debug_fs); + hwpoison_forget_fd = checked_open(buf, O_WRONLY); + } +} + +static int hwpoison_page(unsigned long offset) +{ + char buf[100]; + int len; + + len = sprintf(buf, "0x%lx\n", offset); + len = write(hwpoison_inject_fd, buf, len); + if (len < 0) { + perror("hwpoison inject"); + return len; + } + return 0; +} + +static int unpoison_page(unsigned long offset) +{ + char buf[100]; + int len; + + len = sprintf(buf, "0x%lx\n", offset); + len = write(hwpoison_forget_fd, buf, len); + if (len < 0) { + perror("hwpoison forget"); + return len; + } + return 0; +} + +/* + * page frame walker + */ + +static size_t hash_slot(uint64_t flags) +{ + size_t k = HASH_KEY(flags); + size_t i; + + /* Explicitly reserve slot 0 for flags 0: the following logic + * cannot distinguish an unoccupied slot from slot (flags==0). + */ + if (flags == 0) + return 0; + + /* search through the remaining (HASH_SIZE-1) slots */ + for (i = 1; i < ARRAY_SIZE(page_flags); i++, k++) { + if (!k || k >= ARRAY_SIZE(page_flags)) + k = 1; + if (page_flags[k] == 0) { + page_flags[k] = flags; + return k; + } + if (page_flags[k] == flags) + return k; + } + + fatal("hash table full: bump up HASH_SHIFT?\n"); + exit(EXIT_FAILURE); +} + +static void add_page(unsigned long voffset, + unsigned long offset, uint64_t flags) +{ + flags = kpageflags_flags(flags); + + if (!bit_mask_ok(flags)) + return; + + if (opt_hwpoison) + hwpoison_page(offset); + if (opt_unpoison) + unpoison_page(offset); + + if (opt_list == 1) + show_page_range(voffset, offset, flags); + else if (opt_list == 2) + show_page(voffset, offset, flags); + + nr_pages[hash_slot(flags)]++; + total_pages++; +} + +#define KPAGEFLAGS_BATCH (64 << 10) /* 64k pages */ +static void walk_pfn(unsigned long voffset, + unsigned long index, + unsigned long count) +{ + uint64_t buf[KPAGEFLAGS_BATCH]; + unsigned long batch; + unsigned long pages; + unsigned long i; + + while (count) { + batch = min_t(unsigned long, count, KPAGEFLAGS_BATCH); + pages = kpageflags_read(buf, index, batch); + if (pages == 0) + break; + + for (i = 0; i < pages; i++) + add_page(voffset + i, index + i, buf[i]); + + index += pages; + count -= pages; + } +} + +#define PAGEMAP_BATCH (64 << 10) +static void walk_vma(unsigned long index, unsigned long count) +{ + uint64_t buf[PAGEMAP_BATCH]; + unsigned long batch; + unsigned long pages; + unsigned long pfn; + unsigned long i; + + while (count) { + batch = min_t(unsigned long, count, PAGEMAP_BATCH); + pages = pagemap_read(buf, index, batch); + if (pages == 0) + break; + + for (i = 0; i < pages; i++) { + pfn = pagemap_pfn(buf[i]); + if (pfn) + walk_pfn(index + i, pfn, 1); + } + + index += pages; + count -= pages; + } +} + +static void walk_task(unsigned long index, unsigned long count) +{ + const unsigned long end = index + count; + unsigned long start; + int i = 0; + + while (index < end) { + + while (pg_end[i] <= index) + if (++i >= nr_vmas) + return; + if (pg_start[i] >= end) + return; + + start = max_t(unsigned long, pg_start[i], index); + index = min_t(unsigned long, pg_end[i], end); + + assert(start < index); + walk_vma(start, index - start); + } +} + +static void add_addr_range(unsigned long offset, unsigned long size) +{ + if (nr_addr_ranges >= MAX_ADDR_RANGES) + fatal("too many addr ranges\n"); + + opt_offset[nr_addr_ranges] = offset; + opt_size[nr_addr_ranges] = min_t(unsigned long, size, ULONG_MAX-offset); + nr_addr_ranges++; +} + +static void walk_addr_ranges(void) +{ + int i; + + kpageflags_fd = checked_open(PROC_KPAGEFLAGS, O_RDONLY); + + if (!nr_addr_ranges) + add_addr_range(0, ULONG_MAX); + + for (i = 0; i < nr_addr_ranges; i++) + if (!opt_pid) + walk_pfn(0, opt_offset[i], opt_size[i]); + else + walk_task(opt_offset[i], opt_size[i]); + + close(kpageflags_fd); +} + + +/* + * user interface + */ + +static const char *page_flag_type(uint64_t flag) +{ + if (flag & KPF_HACKERS_BITS) + return "(r)"; + if (flag & KPF_OVERLOADED_BITS) + return "(o)"; + return " "; +} + +static void usage(void) +{ + size_t i, j; + + printf( +"page-types [options]\n" +" -r|--raw Raw mode, for kernel developers\n" +" -d|--describe flags Describe flags\n" +" -a|--addr addr-spec Walk a range of pages\n" +" -b|--bits bits-spec Walk pages with specified bits\n" +" -p|--pid pid Walk process address space\n" +#if 0 /* planned features */ +" -f|--file filename Walk file address space\n" +#endif +" -l|--list Show page details in ranges\n" +" -L|--list-each Show page details one by one\n" +" -N|--no-summary Don't show summary info\n" +" -X|--hwpoison hwpoison pages\n" +" -x|--unpoison unpoison pages\n" +" -h|--help Show this usage message\n" +"flags:\n" +" 0x10 bitfield format, e.g.\n" +" anon bit-name, e.g.\n" +" 0x10,anon comma-separated list, e.g.\n" +"addr-spec:\n" +" N one page at offset N (unit: pages)\n" +" N+M pages range from N to N+M-1\n" +" N,M pages range from N to M-1\n" +" N, pages range from N to end\n" +" ,M pages range from 0 to M-1\n" +"bits-spec:\n" +" bit1,bit2 (flags & (bit1|bit2)) != 0\n" +" bit1,bit2=bit1 (flags & (bit1|bit2)) == bit1\n" +" bit1,~bit2 (flags & (bit1|bit2)) == bit1\n" +" =bit1,bit2 flags == (bit1|bit2)\n" +"bit-names:\n" + ); + + for (i = 0, j = 0; i < ARRAY_SIZE(page_flag_names); i++) { + if (!page_flag_names[i]) + continue; + printf("%16s%s", page_flag_names[i] + 2, + page_flag_type(1ULL << i)); + if (++j > 3) { + j = 0; + putchar('\n'); + } + } + printf("\n " + "(r) raw mode bits (o) overloaded bits\n"); +} + +static unsigned long long parse_number(const char *str) +{ + unsigned long long n; + + n = strtoll(str, NULL, 0); + + if (n == 0 && str[0] != '0') + fatal("invalid name or number: %s\n", str); + + return n; +} + +static void parse_pid(const char *str) +{ + FILE *file; + char buf[5000]; + + opt_pid = parse_number(str); + + sprintf(buf, "/proc/%d/pagemap", opt_pid); + pagemap_fd = checked_open(buf, O_RDONLY); + + sprintf(buf, "/proc/%d/maps", opt_pid); + file = fopen(buf, "r"); + if (!file) { + perror(buf); + exit(EXIT_FAILURE); + } + + while (fgets(buf, sizeof(buf), file) != NULL) { + unsigned long vm_start; + unsigned long vm_end; + unsigned long long pgoff; + int major, minor; + char r, w, x, s; + unsigned long ino; + int n; + + n = sscanf(buf, "%lx-%lx %c%c%c%c %llx %x:%x %lu", + &vm_start, + &vm_end, + &r, &w, &x, &s, + &pgoff, + &major, &minor, + &ino); + if (n < 10) { + fprintf(stderr, "unexpected line: %s\n", buf); + continue; + } + pg_start[nr_vmas] = vm_start / page_size; + pg_end[nr_vmas] = vm_end / page_size; + if (++nr_vmas >= MAX_VMAS) { + fprintf(stderr, "too many VMAs\n"); + break; + } + } + fclose(file); +} + +static void parse_file(const char *name) +{ +} + +static void parse_addr_range(const char *optarg) +{ + unsigned long offset; + unsigned long size; + char *p; + + p = strchr(optarg, ','); + if (!p) + p = strchr(optarg, '+'); + + if (p == optarg) { + offset = 0; + size = parse_number(p + 1); + } else if (p) { + offset = parse_number(optarg); + if (p[1] == '\0') + size = ULONG_MAX; + else { + size = parse_number(p + 1); + if (*p == ',') { + if (size < offset) + fatal("invalid range: %lu,%lu\n", + offset, size); + size -= offset; + } + } + } else { + offset = parse_number(optarg); + size = 1; + } + + add_addr_range(offset, size); +} + +static void add_bits_filter(uint64_t mask, uint64_t bits) +{ + if (nr_bit_filters >= MAX_BIT_FILTERS) + fatal("too much bit filters\n"); + + opt_mask[nr_bit_filters] = mask; + opt_bits[nr_bit_filters] = bits; + nr_bit_filters++; +} + +static uint64_t parse_flag_name(const char *str, int len) +{ + size_t i; + + if (!*str || !len) + return 0; + + if (len <= 8 && !strncmp(str, "compound", len)) + return BITS_COMPOUND; + + for (i = 0; i < ARRAY_SIZE(page_flag_names); i++) { + if (!page_flag_names[i]) + continue; + if (!strncmp(str, page_flag_names[i] + 2, len)) + return 1ULL << i; + } + + return parse_number(str); +} + +static uint64_t parse_flag_names(const char *str, int all) +{ + const char *p = str; + uint64_t flags = 0; + + while (1) { + if (*p == ',' || *p == '=' || *p == '\0') { + if ((*str != '~') || (*str == '~' && all && *++str)) + flags |= parse_flag_name(str, p - str); + if (*p != ',') + break; + str = p + 1; + } + p++; + } + + return flags; +} + +static void parse_bits_mask(const char *optarg) +{ + uint64_t mask; + uint64_t bits; + const char *p; + + p = strchr(optarg, '='); + if (p == optarg) { + mask = KPF_ALL_BITS; + bits = parse_flag_names(p + 1, 0); + } else if (p) { + mask = parse_flag_names(optarg, 0); + bits = parse_flag_names(p + 1, 0); + } else if (strchr(optarg, '~')) { + mask = parse_flag_names(optarg, 1); + bits = parse_flag_names(optarg, 0); + } else { + mask = parse_flag_names(optarg, 0); + bits = KPF_ALL_BITS; + } + + add_bits_filter(mask, bits); +} + +static void describe_flags(const char *optarg) +{ + uint64_t flags = parse_flag_names(optarg, 0); + + printf("0x%016llx\t%s\t%s\n", + (unsigned long long)flags, + page_flag_name(flags), + page_flag_longname(flags)); +} + +static const struct option opts[] = { + { "raw" , 0, NULL, 'r' }, + { "pid" , 1, NULL, 'p' }, + { "file" , 1, NULL, 'f' }, + { "addr" , 1, NULL, 'a' }, + { "bits" , 1, NULL, 'b' }, + { "describe" , 1, NULL, 'd' }, + { "list" , 0, NULL, 'l' }, + { "list-each" , 0, NULL, 'L' }, + { "no-summary", 0, NULL, 'N' }, + { "hwpoison" , 0, NULL, 'X' }, + { "unpoison" , 0, NULL, 'x' }, + { "help" , 0, NULL, 'h' }, + { NULL , 0, NULL, 0 } +}; + +int main(int argc, char *argv[]) +{ + int c; + + page_size = getpagesize(); + + while ((c = getopt_long(argc, argv, + "rp:f:a:b:d:lLNXxh", opts, NULL)) != -1) { + switch (c) { + case 'r': + opt_raw = 1; + break; + case 'p': + parse_pid(optarg); + break; + case 'f': + parse_file(optarg); + break; + case 'a': + parse_addr_range(optarg); + break; + case 'b': + parse_bits_mask(optarg); + break; + case 'd': + describe_flags(optarg); + exit(0); + case 'l': + opt_list = 1; + break; + case 'L': + opt_list = 2; + break; + case 'N': + opt_no_summary = 1; + break; + case 'X': + opt_hwpoison = 1; + prepare_hwpoison_fd(); + break; + case 'x': + opt_unpoison = 1; + prepare_hwpoison_fd(); + break; + case 'h': + usage(); + exit(0); + default: + usage(); + exit(1); + } + } + + if (opt_list && opt_pid) + printf("voffset\t"); + if (opt_list == 1) + printf("offset\tlen\tflags\n"); + if (opt_list == 2) + printf("offset\tflags\n"); + + walk_addr_ranges(); + + if (opt_list == 1) + show_page_range(0, 0, 0); /* drain the buffer */ + + if (opt_no_summary) + return 0; + + if (opt_list) + printf("\n\n"); + + show_summary(); + + return 0; +} diff --git a/tools/slub/slabinfo.c b/tools/vm/slabinfo.c index 164cbcf61106..164cbcf61106 100644 --- a/tools/slub/slabinfo.c +++ b/tools/vm/slabinfo.c |