diff options
Diffstat (limited to 'tools/power')
71 files changed, 7195 insertions, 2828 deletions
diff --git a/tools/power/acpi/.gitignore b/tools/power/acpi/.gitignore index f698a0e5bfa6..eada0297ef88 100644 --- a/tools/power/acpi/.gitignore +++ b/tools/power/acpi/.gitignore @@ -1,4 +1,6 @@ +# SPDX-License-Identifier: GPL-2.0-only /acpidbg /acpidump /ec +/pfrut /include/ diff --git a/tools/power/acpi/Makefile b/tools/power/acpi/Makefile index ebd3e1a1c28e..5ff1d9c864d0 100644 --- a/tools/power/acpi/Makefile +++ b/tools/power/acpi/Makefile @@ -7,18 +7,20 @@ include ../../scripts/Makefile.include -all: acpidbg acpidump ec -clean: acpidbg_clean acpidump_clean ec_clean -install: acpidbg_install acpidump_install ec_install -uninstall: acpidbg_uninstall acpidump_uninstall ec_uninstall +.NOTPARALLEL: -acpidbg acpidump ec: FORCE +all: acpidbg acpidump ec pfrut +clean: acpidbg_clean acpidump_clean ec_clean pfrut_clean +install: acpidbg_install acpidump_install ec_install pfrut_install +uninstall: acpidbg_uninstall acpidump_uninstall ec_uninstall pfrut_uninstall + +acpidbg acpidump ec pfrut: FORCE $(call descend,tools/$@,all) -acpidbg_clean acpidump_clean ec_clean: +acpidbg_clean acpidump_clean ec_clean pfrut_clean: $(call descend,tools/$(@:_clean=),clean) -acpidbg_install acpidump_install ec_install: +acpidbg_install acpidump_install ec_install pfrut_install: $(call descend,tools/$(@:_install=),install) -acpidbg_uninstall acpidump_uninstall ec_uninstall: +acpidbg_uninstall acpidump_uninstall ec_uninstall pfrut_uninstall: $(call descend,tools/$(@:_uninstall=),uninstall) .PHONY: FORCE diff --git a/tools/power/acpi/Makefile.config b/tools/power/acpi/Makefile.config index 54a2857c2510..cd7106876a5f 100644 --- a/tools/power/acpi/Makefile.config +++ b/tools/power/acpi/Makefile.config @@ -54,7 +54,6 @@ INSTALL_SCRIPT = ${INSTALL_PROGRAM} CROSS = #/usr/i386-linux-uclibc/usr/bin/i386-uclibc- CROSS_COMPILE ?= $(CROSS) LD = $(CC) -HOSTCC = gcc # check if compiler option is supported cc-supports = ${shell if $(CC) ${1} -S -o /dev/null -x c /dev/null > /dev/null 2>&1; then echo "$(1)"; fi;} @@ -70,6 +69,7 @@ KERNEL_INCLUDE := $(OUTPUT)include ACPICA_INCLUDE := $(srctree)/../../../drivers/acpi/acpica CFLAGS += -D_LINUX -I$(KERNEL_INCLUDE) -I$(ACPICA_INCLUDE) CFLAGS += $(WARNINGS) +MKDIR = mkdir ifeq ($(strip $(V)),false) QUIET=@ diff --git a/tools/power/acpi/Makefile.rules b/tools/power/acpi/Makefile.rules index 2a6c170b57cd..b71aada77688 100644 --- a/tools/power/acpi/Makefile.rules +++ b/tools/power/acpi/Makefile.rules @@ -9,7 +9,7 @@ objdir := $(OUTPUT)tools/$(TOOL)/ toolobjs := $(addprefix $(objdir),$(TOOL_OBJS)) $(OUTPUT)$(TOOL): $(toolobjs) FORCE $(ECHO) " LD " $(subst $(OUTPUT),,$@) - $(QUIET) $(LD) $(CFLAGS) $(LDFLAGS) $(toolobjs) -L$(OUTPUT) -o $@ + $(QUIET) $(LD) $(CFLAGS) $(toolobjs) $(LDFLAGS) -L$(OUTPUT) -o $@ $(ECHO) " STRIP " $(subst $(OUTPUT),,$@) $(QUIET) $(STRIPCMD) $@ @@ -21,6 +21,7 @@ $(KERNEL_INCLUDE): $(objdir)%.o: %.c $(KERNEL_INCLUDE) $(ECHO) " CC " $(subst $(OUTPUT),,$@) + $(QUIET) $(MKDIR) -p $(objdir) 2>/dev/null $(QUIET) $(CC) -c $(CFLAGS) -o $@ $< all: $(OUTPUT)$(TOOL) diff --git a/tools/power/acpi/common/cmfsize.c b/tools/power/acpi/common/cmfsize.c index d1d18ff5c911..38f9b9da8170 100644 --- a/tools/power/acpi/common/cmfsize.c +++ b/tools/power/acpi/common/cmfsize.c @@ -1,9 +1,9 @@ // SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0 /****************************************************************************** * - * Module Name: cfsize - Common get file size function + * Module Name: cmfsize - Common get file size function * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/common/getopt.c b/tools/power/acpi/common/getopt.c index c3708f30ab3a..96fd6cec78e2 100644 --- a/tools/power/acpi/common/getopt.c +++ b/tools/power/acpi/common/getopt.c @@ -3,7 +3,7 @@ * * Module Name: getopt * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/man/pfrut.8 b/tools/power/acpi/man/pfrut.8 new file mode 100644 index 000000000000..3db574770e8d --- /dev/null +++ b/tools/power/acpi/man/pfrut.8 @@ -0,0 +1,137 @@ +.TH "PFRUT" "8" "October 2021" "pfrut 1.0" "" +.hy +.SH Name +.PP +pfrut \- Platform Firmware Runtime Update and Telemetry tool +.SH SYNOPSIS +.PP +\f[B]pfrut\f[R] [\f[I]Options\f[R]] +.SH DESCRIPTION +.PP +The PFRUT(Platform Firmware Runtime Update and Telemetry) kernel interface is designed +to +.PD 0 +.P +.PD +interact with the platform firmware interface defined in the +.PD 0 +.P +.PD +Management Mode Firmware Runtime +Update (https://uefi.org/sites/default/files/resources/Intel_MM_OS_Interface_Spec_Rev100.pdf) +.PD 0 +.P +.PD +\f[B]pfrut\f[R] is the tool to interact with the kernel interface. +.PD 0 +.P +.PD +.SH OPTIONS +.TP +.B \f[B]\-h\f[R], \f[B]\-\-help\f[R] +Display helper information. +.TP +.B \f[B]\-l\f[R], \f[B]\-\-load\f[R] +Load the capsule file into the system. +To be more specific, the capsule file will be copied to the +communication buffer. +.TP +.B \f[B]\-s\f[R], \f[B]\-\-stage\f[R] +Stage the capsule image from communication buffer into Management Mode +and perform authentication. +.TP +.B \f[B]\-a\f[R], \f[B]\-\-activate\f[R] +Activate a previous staged capsule image. +.TP +.B \f[B]\-u\f[R], \f[B]\-\-update\f[R] +Perform both stage and activation actions. +.TP +.B \f[B]\-q\f[R], \f[B]\-\-query\f[R] +Query the update capability. +.TP +.B \f[B]\-d\f[R], \f[B]\-\-setrev\f[R] +Set the revision ID of code injection/driver update. +.TP +.B \f[B]\-D\f[R], \f[B]\-\-setrevlog\f[R] +Set the revision ID of telemetry. +.TP +.B \f[B]\-G\f[R], \f[B]\-\-getloginfo\f[R] +Get telemetry log information and print it out. +.TP +.B \f[B]\-T\f[R], \f[B]\-\-type\f[R] +Set the telemetry log data type. +.TP +.B \f[B]\-L\f[R], \f[B]\-\-level\f[R] +Set the telemetry log level. +.TP +.B \f[B]\-R\f[R], \f[B]\-\-read\f[R] +Read all the telemetry data and print it out. +.SH EXAMPLES +.PP +\f[B]pfrut \-G\f[R] +.PP +log_level:4 +.PD 0 +.P +.PD +log_type:0 +.PD 0 +.P +.PD +log_revid:2 +.PD 0 +.P +.PD +max_data_size:65536 +.PD 0 +.P +.PD +chunk1_size:0 +.PD 0 +.P +.PD +chunk2_size:1401 +.PD 0 +.P +.PD +rollover_cnt:0 +.PD 0 +.P +.PD +reset_cnt:4 +.PP +\f[B]pfru \-q\f[R] +.PP +code injection image type:794bf8b2\-6e7b\-454e\-885f\-3fb9bb185402 +.PD 0 +.P +.PD +fw_version:0 +.PD 0 +.P +.PD +code_rt_version:1 +.PD 0 +.P +.PD +driver update image type:0e5f0b14\-f849\-7945\-ad81\-bc7b6d2bb245 +.PD 0 +.P +.PD +drv_rt_version:0 +.PD 0 +.P +.PD +drv_svn:0 +.PD 0 +.P +.PD +platform id:39214663\-b1a8\-4eaa\-9024\-f2bb53ea4723 +.PD 0 +.P +.PD +oem id:a36db54f\-ea2a\-e14e\-b7c4\-b5780e51ba3d +.PP +\f[B]pfrut \-l yours.cap \-u \-T 1 \-L 4\f[R] +.SH AUTHORS +Chen Yu. diff --git a/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c b/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c index 5aaddc79adf7..bd08f36df4a7 100644 --- a/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c +++ b/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c @@ -3,7 +3,7 @@ * * Module Name: oslinuxtbl - Linux OSL for obtaining ACPI tables * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ @@ -35,7 +35,7 @@ static acpi_status osl_add_table_to_list(char *signature, u32 instance); static acpi_status osl_read_table_from_file(char *filename, acpi_size file_offset, - char *signature, struct acpi_table_header **table); + struct acpi_table_header **table); static acpi_status osl_map_table(acpi_size address, @@ -110,7 +110,7 @@ u32 gbl_table_count = 0; * * RETURN: Status; Converted from errno. * - * DESCRIPTION: Get last errno and conver it to acpi_status. + * DESCRIPTION: Get last errno and convert it to acpi_status. * *****************************************************************************/ @@ -1184,8 +1184,6 @@ osl_table_name_from_file(char *filename, char *signature, u32 *instance) * * PARAMETERS: filename - File that contains the desired table * file_offset - Offset of the table in file - * signature - Optional ACPI Signature for desired table. - * A null terminated 4-character string. * table - Where a pointer to the table is returned * * RETURN: Status; Table buffer is returned if AE_OK. @@ -1197,7 +1195,7 @@ osl_table_name_from_file(char *filename, char *signature, u32 *instance) static acpi_status osl_read_table_from_file(char *filename, acpi_size file_offset, - char *signature, struct acpi_table_header **table) + struct acpi_table_header **table) { FILE *table_file; struct acpi_table_header header; @@ -1225,6 +1223,8 @@ osl_read_table_from_file(char *filename, goto exit; } +#ifdef ACPI_OBSOLETE_FUNCTIONS + /* If signature is specified, it must match the table */ if (signature) { @@ -1244,6 +1244,7 @@ osl_read_table_from_file(char *filename, goto exit; } } +#endif table_length = ap_get_table_length(&header); if (table_length == 0) { @@ -1366,7 +1367,7 @@ osl_get_customized_table(char *pathname, /* There is no physical address saved for customized tables, use zero */ *address = 0; - status = osl_read_table_from_file(table_filename, 0, NULL, table); + status = osl_read_table_from_file(table_filename, 0, table); return (status); } diff --git a/tools/power/acpi/os_specific/service_layers/osunixdir.c b/tools/power/acpi/os_specific/service_layers/osunixdir.c index fd05ddee240f..5107892d054b 100644 --- a/tools/power/acpi/os_specific/service_layers/osunixdir.c +++ b/tools/power/acpi/os_specific/service_layers/osunixdir.c @@ -3,7 +3,7 @@ * * Module Name: osunixdir - Unix directory access interfaces * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/os_specific/service_layers/osunixmap.c b/tools/power/acpi/os_specific/service_layers/osunixmap.c index c565546e85bc..6ff4edd8dc3b 100644 --- a/tools/power/acpi/os_specific/service_layers/osunixmap.c +++ b/tools/power/acpi/os_specific/service_layers/osunixmap.c @@ -3,7 +3,7 @@ * * Module Name: osunixmap - Unix OSL for file mappings * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/os_specific/service_layers/osunixxf.c b/tools/power/acpi/os_specific/service_layers/osunixxf.c index 5b2fd968535f..b3651a04d68c 100644 --- a/tools/power/acpi/os_specific/service_layers/osunixxf.c +++ b/tools/power/acpi/os_specific/service_layers/osunixxf.c @@ -3,7 +3,7 @@ * * Module Name: osunixxf - UNIX OSL interfaces * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/tools/acpidump/acpidump.h b/tools/power/acpi/tools/acpidump/acpidump.h index 26a5eae9f87f..153249c87fd7 100644 --- a/tools/power/acpi/tools/acpidump/acpidump.h +++ b/tools/power/acpi/tools/acpidump/acpidump.h @@ -3,7 +3,7 @@ * * Module Name: acpidump.h - Include file for acpi_dump utility * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/tools/acpidump/apdump.c b/tools/power/acpi/tools/acpidump/apdump.c index 76433296055d..d54dde02b87d 100644 --- a/tools/power/acpi/tools/acpidump/apdump.c +++ b/tools/power/acpi/tools/acpidump/apdump.c @@ -3,7 +3,7 @@ * * Module Name: apdump - Dump routines for ACPI tables (acpidump) * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/tools/acpidump/apfiles.c b/tools/power/acpi/tools/acpidump/apfiles.c index a682bae4e6f6..2d9b45a9b526 100644 --- a/tools/power/acpi/tools/acpidump/apfiles.c +++ b/tools/power/acpi/tools/acpidump/apfiles.c @@ -3,7 +3,7 @@ * * Module Name: apfiles - File-related functions for acpidump utility * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/tools/acpidump/apmain.c b/tools/power/acpi/tools/acpidump/apmain.c index 046e6b8d6baa..44b23fc53dd9 100644 --- a/tools/power/acpi/tools/acpidump/apmain.c +++ b/tools/power/acpi/tools/acpidump/apmain.c @@ -3,7 +3,7 @@ * * Module Name: apmain - Main module for the acpidump utility * - * Copyright (C) 2000 - 2020, Intel Corp. + * Copyright (C) 2000 - 2022, Intel Corp. * *****************************************************************************/ diff --git a/tools/power/acpi/tools/pfrut/Makefile b/tools/power/acpi/tools/pfrut/Makefile new file mode 100644 index 000000000000..61c1a96fd433 --- /dev/null +++ b/tools/power/acpi/tools/pfrut/Makefile @@ -0,0 +1,23 @@ +# SPDX-License-Identifier: GPL-2.0+ + +include ../../Makefile.config + +TOOL = pfrut +EXTRA_INSTALL = install-man +EXTRA_UNINSTALL = uninstall-man + +CFLAGS += -Wall -O2 +CFLAGS += -DPFRUT_HEADER='"../../../../../include/uapi/linux/pfrut.h"' +LDFLAGS += -luuid + +TOOL_OBJS = \ + pfrut.o + +include ../../Makefile.rules + +install-man: $(srctree)/man/pfrut.8 + $(ECHO) " INST " pfrut.8 + $(QUIET) $(INSTALL_DATA) -D $< $(DESTDIR)$(mandir)/man8/pfrut.8 +uninstall-man: + $(ECHO) " UNINST " pfrut.8 + $(QUIET) rm -f $(DESTDIR)$(mandir)/man8/pfrut.8 diff --git a/tools/power/acpi/tools/pfrut/pfrut.c b/tools/power/acpi/tools/pfrut/pfrut.c new file mode 100644 index 000000000000..52aa0351533c --- /dev/null +++ b/tools/power/acpi/tools/pfrut/pfrut.c @@ -0,0 +1,424 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Platform Firmware Runtime Update tool to do Management + * Mode code injection/driver update and telemetry retrieval. + * + * This tool uses the interfaces provided by pfr_update and + * pfr_telemetry drivers. These interfaces are exposed via + * /dev/pfr_update and /dev/pfr_telemetry. Write operation + * on the /dev/pfr_update is to load the EFI capsule into + * kernel space. Mmap/read operations on /dev/pfr_telemetry + * could be used to read the telemetry data to user space. + */ +#define _GNU_SOURCE +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <unistd.h> +#include <getopt.h> +#include <sys/ioctl.h> +#include <sys/mman.h> +#include <uuid/uuid.h> +#include PFRUT_HEADER + +char *capsule_name; +int action, query_cap, log_type, log_level, log_read, log_getinfo, + revid, log_revid; +int set_log_level, set_log_type, + set_revid, set_log_revid; + +char *progname; + +#define LOG_ERR 0 +#define LOG_WARN 1 +#define LOG_INFO 2 +#define LOG_VERB 4 +#define LOG_EXEC_IDX 0 +#define LOG_HISTORY_IDX 1 +#define REVID_1 1 +#define REVID_2 2 + +static int valid_log_level(int level) +{ + return level == LOG_ERR || level == LOG_WARN || + level == LOG_INFO || level == LOG_VERB; +} + +static int valid_log_type(int type) +{ + return type == LOG_EXEC_IDX || type == LOG_HISTORY_IDX; +} + +static inline int valid_log_revid(int id) +{ + return id == REVID_1 || id == REVID_2; +} + +static void help(void) +{ + fprintf(stderr, + "usage: %s [OPTIONS]\n" + " code injection:\n" + " -l, --load\n" + " -s, --stage\n" + " -a, --activate\n" + " -u, --update [stage and activate]\n" + " -q, --query\n" + " -d, --revid update\n" + " telemetry:\n" + " -G, --getloginfo\n" + " -T, --type(0:execution, 1:history)\n" + " -L, --level(0, 1, 2, 4)\n" + " -R, --read\n" + " -D, --revid log\n", + progname); +} + +char *option_string = "l:sauqd:GT:L:RD:h"; +static struct option long_options[] = { + {"load", required_argument, 0, 'l'}, + {"stage", no_argument, 0, 's'}, + {"activate", no_argument, 0, 'a'}, + {"update", no_argument, 0, 'u'}, + {"query", no_argument, 0, 'q'}, + {"getloginfo", no_argument, 0, 'G'}, + {"type", required_argument, 0, 'T'}, + {"level", required_argument, 0, 'L'}, + {"read", no_argument, 0, 'R'}, + {"setrev", required_argument, 0, 'd'}, + {"setrevlog", required_argument, 0, 'D'}, + {"help", no_argument, 0, 'h'}, + {} +}; + +static void parse_options(int argc, char **argv) +{ + int option_index = 0; + char *pathname; + int opt; + + pathname = strdup(argv[0]); + progname = basename(pathname); + + while ((opt = getopt_long_only(argc, argv, option_string, + long_options, &option_index)) != -1) { + switch (opt) { + case 'l': + capsule_name = optarg; + break; + case 's': + action = 1; + break; + case 'a': + action = 2; + break; + case 'u': + action = 3; + break; + case 'q': + query_cap = 1; + break; + case 'G': + log_getinfo = 1; + break; + case 'T': + log_type = atoi(optarg); + set_log_type = 1; + break; + case 'L': + log_level = atoi(optarg); + set_log_level = 1; + break; + case 'R': + log_read = 1; + break; + case 'd': + revid = atoi(optarg); + set_revid = 1; + break; + case 'D': + log_revid = atoi(optarg); + set_log_revid = 1; + break; + case 'h': + help(); + exit(0); + default: + break; + } + } +} + +void print_cap(struct pfru_update_cap_info *cap) +{ + char *uuid; + + uuid = malloc(37); + if (!uuid) { + perror("Can not allocate uuid buffer\n"); + exit(1); + } + + uuid_unparse(cap->code_type, uuid); + printf("code injection image type:%s\n", uuid); + printf("fw_version:%d\n", cap->fw_version); + printf("code_rt_version:%d\n", cap->code_rt_version); + + uuid_unparse(cap->drv_type, uuid); + printf("driver update image type:%s\n", uuid); + printf("drv_rt_version:%d\n", cap->drv_rt_version); + printf("drv_svn:%d\n", cap->drv_svn); + + uuid_unparse(cap->platform_id, uuid); + printf("platform id:%s\n", uuid); + uuid_unparse(cap->oem_id, uuid); + printf("oem id:%s\n", uuid); + printf("oem information length:%d\n", cap->oem_info_len); + + free(uuid); +} + +int main(int argc, char *argv[]) +{ + int fd_update, fd_update_log, fd_capsule; + struct pfrt_log_data_info data_info; + struct pfrt_log_info info; + struct pfru_update_cap_info cap; + void *addr_map_capsule; + struct stat st; + char *log_buf; + int ret; + + if (getuid() != 0) { + printf("Please run the tool as root - Exiting.\n"); + return 1; + } + + parse_options(argc, argv); + + fd_update = open("/dev/acpi_pfr_update0", O_RDWR); + if (fd_update < 0) { + printf("PFRU device not supported - Quit...\n"); + return 1; + } + + fd_update_log = open("/dev/acpi_pfr_telemetry0", O_RDWR); + if (fd_update_log < 0) { + printf("PFRT device not supported - Quit...\n"); + return 1; + } + + if (query_cap) { + ret = ioctl(fd_update, PFRU_IOC_QUERY_CAP, &cap); + if (ret) + perror("Query Update Capability info failed."); + else + print_cap(&cap); + + close(fd_update); + close(fd_update_log); + + return ret; + } + + if (log_getinfo) { + ret = ioctl(fd_update_log, PFRT_LOG_IOC_GET_DATA_INFO, &data_info); + if (ret) { + perror("Get telemetry data info failed."); + close(fd_update); + close(fd_update_log); + + return 1; + } + + ret = ioctl(fd_update_log, PFRT_LOG_IOC_GET_INFO, &info); + if (ret) { + perror("Get telemetry info failed."); + close(fd_update); + close(fd_update_log); + + return 1; + } + + printf("log_level:%d\n", info.log_level); + printf("log_type:%d\n", info.log_type); + printf("log_revid:%d\n", info.log_revid); + printf("max_data_size:%d\n", data_info.max_data_size); + printf("chunk1_size:%d\n", data_info.chunk1_size); + printf("chunk2_size:%d\n", data_info.chunk2_size); + printf("rollover_cnt:%d\n", data_info.rollover_cnt); + printf("reset_cnt:%d\n", data_info.reset_cnt); + + return 0; + } + + info.log_level = -1; + info.log_type = -1; + info.log_revid = -1; + + if (set_log_level) { + if (!valid_log_level(log_level)) { + printf("Invalid log level %d\n", + log_level); + } else { + info.log_level = log_level; + } + } + + if (set_log_type) { + if (!valid_log_type(log_type)) { + printf("Invalid log type %d\n", + log_type); + } else { + info.log_type = log_type; + } + } + + if (set_log_revid) { + if (!valid_log_revid(log_revid)) { + printf("Invalid log revid %d, unchanged.\n", + log_revid); + } else { + info.log_revid = log_revid; + } + } + + ret = ioctl(fd_update_log, PFRT_LOG_IOC_SET_INFO, &info); + if (ret) { + perror("Log information set failed.(log_level, log_type, log_revid)"); + close(fd_update); + close(fd_update_log); + + return 1; + } + + if (set_revid) { + ret = ioctl(fd_update, PFRU_IOC_SET_REV, &revid); + if (ret) { + perror("pfru update revid set failed"); + close(fd_update); + close(fd_update_log); + + return 1; + } + + printf("pfru update revid set to %d\n", revid); + } + + if (capsule_name) { + fd_capsule = open(capsule_name, O_RDONLY); + if (fd_capsule < 0) { + perror("Can not open capsule file..."); + close(fd_update); + close(fd_update_log); + + return 1; + } + + if (fstat(fd_capsule, &st) < 0) { + perror("Can not fstat capsule file..."); + close(fd_capsule); + close(fd_update); + close(fd_update_log); + + return 1; + } + + addr_map_capsule = mmap(NULL, st.st_size, PROT_READ, MAP_SHARED, + fd_capsule, 0); + if (addr_map_capsule == MAP_FAILED) { + perror("Failed to mmap capsule file."); + close(fd_capsule); + close(fd_update); + close(fd_update_log); + + return 1; + } + + ret = write(fd_update, (char *)addr_map_capsule, st.st_size); + printf("Load %d bytes of capsule file into the system\n", + ret); + + if (ret == -1) { + perror("Failed to load capsule file"); + close(fd_capsule); + close(fd_update); + close(fd_update_log); + + return 1; + } + + munmap(addr_map_capsule, st.st_size); + close(fd_capsule); + printf("Load done.\n"); + } + + if (action) { + if (action == 1) { + ret = ioctl(fd_update, PFRU_IOC_STAGE, NULL); + } else if (action == 2) { + ret = ioctl(fd_update, PFRU_IOC_ACTIVATE, NULL); + } else if (action == 3) { + ret = ioctl(fd_update, PFRU_IOC_STAGE_ACTIVATE, NULL); + } else { + close(fd_update); + close(fd_update_log); + + return 1; + } + printf("Update finished, return %d\n", ret); + } + + close(fd_update); + + if (log_read) { + void *p_mmap; + int max_data_sz; + + ret = ioctl(fd_update_log, PFRT_LOG_IOC_GET_DATA_INFO, &data_info); + if (ret) { + perror("Get telemetry data info failed."); + close(fd_update_log); + + return 1; + } + + max_data_sz = data_info.max_data_size; + if (!max_data_sz) { + printf("No telemetry data available.\n"); + close(fd_update_log); + + return 1; + } + + log_buf = malloc(max_data_sz + 1); + if (!log_buf) { + perror("log_buf allocate failed."); + close(fd_update_log); + + return 1; + } + + p_mmap = mmap(NULL, max_data_sz, PROT_READ, MAP_SHARED, fd_update_log, 0); + if (p_mmap == MAP_FAILED) { + perror("mmap error."); + close(fd_update_log); + + return 1; + } + + memcpy(log_buf, p_mmap, max_data_sz); + log_buf[max_data_sz] = '\0'; + printf("%s\n", log_buf); + free(log_buf); + + munmap(p_mmap, max_data_sz); + } + + close(fd_update_log); + + return 0; +} diff --git a/tools/power/cpupower/.gitignore b/tools/power/cpupower/.gitignore index 1f9977cc609c..7677329c42a6 100644 --- a/tools/power/cpupower/.gitignore +++ b/tools/power/cpupower/.gitignore @@ -1,3 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0-only .libs libcpupower.so libcpupower.so.* diff --git a/tools/power/cpupower/Makefile b/tools/power/cpupower/Makefile index c8622497ef23..e9b6de314654 100644 --- a/tools/power/cpupower/Makefile +++ b/tools/power/cpupower/Makefile @@ -51,7 +51,7 @@ DESTDIR ?= # Package-related definitions. Distributions can modify the version # and _should_ modify the PACKAGE_BUGREPORT definition -VERSION= $(shell ./utils/version-gen.sh) +VERSION:= $(shell ./utils/version-gen.sh) LIB_MAJ= 0.0.1 LIB_MIN= 0 @@ -143,9 +143,9 @@ 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 -LIB_HEADERS = lib/cpufreq.h lib/cpupower.h lib/cpuidle.h -LIB_SRC = lib/cpufreq.c lib/cpupower.c lib/cpuidle.c -LIB_OBJS = lib/cpufreq.o lib/cpupower.o lib/cpuidle.o +LIB_HEADERS = lib/cpufreq.h lib/cpupower.h lib/cpuidle.h lib/acpi_cppc.h +LIB_SRC = lib/cpufreq.c lib/cpupower.c lib/cpuidle.c lib/acpi_cppc.c +LIB_OBJS = lib/cpufreq.o lib/cpupower.o lib/cpuidle.o lib/acpi_cppc.o LIB_OBJS := $(addprefix $(OUTPUT),$(LIB_OBJS)) override CFLAGS += -pipe @@ -270,14 +270,14 @@ clean: $(MAKE) -C bench O=$(OUTPUT) clean -install-lib: +install-lib: libcpupower $(INSTALL) -d $(DESTDIR)${libdir} $(CP) $(OUTPUT)libcpupower.so* $(DESTDIR)${libdir}/ $(INSTALL) -d $(DESTDIR)${includedir} $(INSTALL_DATA) lib/cpufreq.h $(DESTDIR)${includedir}/cpufreq.h $(INSTALL_DATA) lib/cpuidle.h $(DESTDIR)${includedir}/cpuidle.h -install-tools: +install-tools: $(OUTPUT)cpupower $(INSTALL) -d $(DESTDIR)${bindir} $(INSTALL_PROGRAM) $(OUTPUT)cpupower $(DESTDIR)${bindir} $(INSTALL) -d $(DESTDIR)${bash_completion_dir} @@ -293,14 +293,14 @@ install-man: $(INSTALL_DATA) -D man/cpupower-info.1 $(DESTDIR)${mandir}/man1/cpupower-info.1 $(INSTALL_DATA) -D man/cpupower-monitor.1 $(DESTDIR)${mandir}/man1/cpupower-monitor.1 -install-gmo: +install-gmo: create-gmo $(INSTALL) -d $(DESTDIR)${localedir} for HLANG in $(LANGUAGES); do \ echo '$(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: +install-bench: compile-bench @#DESTDIR must be set from outside to survive @sbindir=$(sbindir) bindir=$(bindir) docdir=$(docdir) confdir=$(confdir) $(MAKE) -C bench O=$(OUTPUT) install diff --git a/tools/power/cpupower/ToDo b/tools/power/cpupower/TODO index b196a139a3e4..b196a139a3e4 100644 --- a/tools/power/cpupower/ToDo +++ b/tools/power/cpupower/TODO diff --git a/tools/power/cpupower/bench/Makefile b/tools/power/cpupower/bench/Makefile index f68b4bc55273..d9d9923af85c 100644 --- a/tools/power/cpupower/bench/Makefile +++ b/tools/power/cpupower/bench/Makefile @@ -27,7 +27,7 @@ $(OUTPUT)cpufreq-bench: $(OBJS) all: $(OUTPUT)cpufreq-bench -install: +install: $(OUTPUT)cpufreq-bench mkdir -p $(DESTDIR)/$(sbindir) mkdir -p $(DESTDIR)/$(bindir) mkdir -p $(DESTDIR)/$(docdir) diff --git a/tools/power/cpupower/debug/i386/dump_psb.c b/tools/power/cpupower/debug/i386/dump_psb.c index 2c768cf70128..6fb81b42ea61 100644 --- a/tools/power/cpupower/debug/i386/dump_psb.c +++ b/tools/power/cpupower/debug/i386/dump_psb.c @@ -1,7 +1,5 @@ -/* - * dump_psb. (c) 2004, Dave Jones, Red Hat Inc. - * Licensed under the GPL v2. - */ +// SPDX-License-Identifier: GPL-2.0-only +// dump_psb. (c) 2004, Dave Jones, Red Hat Inc. #include <fcntl.h> #include <stdio.h> diff --git a/tools/power/cpupower/debug/i386/intel_gsic.c b/tools/power/cpupower/debug/i386/intel_gsic.c index e5e926f46d6b..befd837f07f8 100644 --- a/tools/power/cpupower/debug/i386/intel_gsic.c +++ b/tools/power/cpupower/debug/i386/intel_gsic.c @@ -71,7 +71,7 @@ int main (void) printf("\tsmi_cmd=0x?? smi_port=0x?? smi_sig=1\n"); printf("\nUnfortunately, you have to know what exactly are " "smi_cmd and smi_port, and this\nis system " - "dependant.\n"); + "dependent.\n"); } return 1; } diff --git a/tools/power/cpupower/lib/acpi_cppc.c b/tools/power/cpupower/lib/acpi_cppc.c new file mode 100644 index 000000000000..c401ac331e9f --- /dev/null +++ b/tools/power/cpupower/lib/acpi_cppc.c @@ -0,0 +1,59 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#include <stdio.h> +#include <errno.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <unistd.h> + +#include "cpupower_intern.h" +#include "acpi_cppc.h" + +/* ACPI CPPC sysfs access ***********************************************/ + +static int acpi_cppc_read_file(unsigned int cpu, const char *fname, + char *buf, size_t buflen) +{ + char path[SYSFS_PATH_MAX]; + + snprintf(path, sizeof(path), PATH_TO_CPU "cpu%u/acpi_cppc/%s", + cpu, fname); + return cpupower_read_sysfs(path, buf, buflen); +} + +static const char * const acpi_cppc_value_files[] = { + [HIGHEST_PERF] = "highest_perf", + [LOWEST_PERF] = "lowest_perf", + [NOMINAL_PERF] = "nominal_perf", + [LOWEST_NONLINEAR_PERF] = "lowest_nonlinear_perf", + [LOWEST_FREQ] = "lowest_freq", + [NOMINAL_FREQ] = "nominal_freq", + [REFERENCE_PERF] = "reference_perf", + [WRAPAROUND_TIME] = "wraparound_time" +}; + +unsigned long acpi_cppc_get_data(unsigned int cpu, enum acpi_cppc_value which) +{ + unsigned long long value; + unsigned int len; + char linebuf[MAX_LINE_LEN]; + char *endp; + + if (which >= MAX_CPPC_VALUE_FILES) + return 0; + + len = acpi_cppc_read_file(cpu, acpi_cppc_value_files[which], + linebuf, sizeof(linebuf)); + if (len == 0) + return 0; + + value = strtoull(linebuf, &endp, 0); + + if (endp == linebuf || errno == ERANGE) + return 0; + + return value; +} diff --git a/tools/power/cpupower/lib/acpi_cppc.h b/tools/power/cpupower/lib/acpi_cppc.h new file mode 100644 index 000000000000..85ca83080316 --- /dev/null +++ b/tools/power/cpupower/lib/acpi_cppc.h @@ -0,0 +1,21 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ + +#ifndef __ACPI_CPPC_H__ +#define __ACPI_CPPC_H__ + +enum acpi_cppc_value { + HIGHEST_PERF, + LOWEST_PERF, + NOMINAL_PERF, + LOWEST_NONLINEAR_PERF, + LOWEST_FREQ, + NOMINAL_FREQ, + REFERENCE_PERF, + WRAPAROUND_TIME, + MAX_CPPC_VALUE_FILES +}; + +unsigned long acpi_cppc_get_data(unsigned int cpu, + enum acpi_cppc_value which); + +#endif /* _ACPI_CPPC_H */ diff --git a/tools/power/cpupower/lib/cpufreq.c b/tools/power/cpupower/lib/cpufreq.c index 6e04304560ca..1516d23c17c9 100644 --- a/tools/power/cpupower/lib/cpufreq.c +++ b/tools/power/cpupower/lib/cpufreq.c @@ -83,20 +83,21 @@ static const char *cpufreq_value_files[MAX_CPUFREQ_VALUE_READ_FILES] = { [STATS_NUM_TRANSITIONS] = "stats/total_trans" }; - -static unsigned long sysfs_cpufreq_get_one_value(unsigned int cpu, - enum cpufreq_value which) +unsigned long cpufreq_get_sysfs_value_from_table(unsigned int cpu, + const char **table, + unsigned int index, + unsigned int size) { unsigned long value; unsigned int len; char linebuf[MAX_LINE_LEN]; char *endp; - if (which >= MAX_CPUFREQ_VALUE_READ_FILES) + if (!table || index >= size || !table[index]) return 0; - len = sysfs_cpufreq_read_file(cpu, cpufreq_value_files[which], - linebuf, sizeof(linebuf)); + len = sysfs_cpufreq_read_file(cpu, table[index], linebuf, + sizeof(linebuf)); if (len == 0) return 0; @@ -109,6 +110,14 @@ static unsigned long sysfs_cpufreq_get_one_value(unsigned int cpu, return value; } +static unsigned long sysfs_cpufreq_get_one_value(unsigned int cpu, + enum cpufreq_value which) +{ + return cpufreq_get_sysfs_value_from_table(cpu, cpufreq_value_files, + which, + MAX_CPUFREQ_VALUE_READ_FILES); +} + /* read access to files which contain one string */ enum cpufreq_string { @@ -124,7 +133,7 @@ static const char *cpufreq_string_files[MAX_CPUFREQ_STRING_FILES] = { static char *sysfs_cpufreq_get_one_string(unsigned int cpu, - enum cpufreq_string which) + enum cpufreq_string which) { char linebuf[MAX_LINE_LEN]; char *result; @@ -285,7 +294,7 @@ struct cpufreq_available_governors *cpufreq_get_available_governors(unsigned } else { first = malloc(sizeof(*first)); if (!first) - goto error_out; + return NULL; current = first; } current->first = first; @@ -362,7 +371,7 @@ struct cpufreq_available_frequencies } else { first = malloc(sizeof(*first)); if (!first) - goto error_out; + return NULL; current = first; } current->first = first; @@ -418,7 +427,7 @@ struct cpufreq_available_frequencies } else { first = malloc(sizeof(*first)); if (!first) - goto error_out; + return NULL; current = first; } current->first = first; @@ -493,7 +502,7 @@ static struct cpufreq_affected_cpus *sysfs_get_cpu_list(unsigned int cpu, } else { first = malloc(sizeof(*first)); if (!first) - goto error_out; + return NULL; current = first; } current->first = first; @@ -726,7 +735,7 @@ struct cpufreq_stats *cpufreq_get_stats(unsigned int cpu, } else { first = malloc(sizeof(*first)); if (!first) - goto error_out; + return NULL; current = first; } current->first = first; diff --git a/tools/power/cpupower/lib/cpufreq.h b/tools/power/cpupower/lib/cpufreq.h index 95f4fd9e2656..2f3c84035806 100644 --- a/tools/power/cpupower/lib/cpufreq.h +++ b/tools/power/cpupower/lib/cpufreq.h @@ -203,6 +203,18 @@ int cpufreq_modify_policy_governor(unsigned int cpu, char *governor); int cpufreq_set_frequency(unsigned int cpu, unsigned long target_frequency); +/* + * get the sysfs value from specific table + * + * Read the value with the sysfs file name from specific table. Does + * only work if the cpufreq driver has the specific sysfs interfaces. + */ + +unsigned long cpufreq_get_sysfs_value_from_table(unsigned int cpu, + const char **table, + unsigned int index, + unsigned int size); + #ifdef __cplusplus } #endif diff --git a/tools/power/cpupower/lib/cpupower.c b/tools/power/cpupower/lib/cpupower.c index 3656e697537e..3f7d0c0c5067 100644 --- a/tools/power/cpupower/lib/cpupower.c +++ b/tools/power/cpupower/lib/cpupower.c @@ -16,8 +16,8 @@ unsigned int cpupower_read_sysfs(const char *path, char *buf, size_t buflen) { - int fd; ssize_t numread; + int fd; fd = open(path, O_RDONLY); if (fd == -1) @@ -35,6 +35,27 @@ unsigned int cpupower_read_sysfs(const char *path, char *buf, size_t buflen) return (unsigned int) numread; } +unsigned int cpupower_write_sysfs(const char *path, char *buf, size_t buflen) +{ + ssize_t numwritten; + int fd; + + fd = open(path, O_WRONLY); + if (fd == -1) + return 0; + + numwritten = write(fd, buf, buflen - 1); + if (numwritten < 1) { + perror(path); + close(fd); + return -1; + } + + close(fd); + + return (unsigned int) numwritten; +} + /* * Detect whether a CPU is online * diff --git a/tools/power/cpupower/lib/cpupower_intern.h b/tools/power/cpupower/lib/cpupower_intern.h index 4887c76d23f8..ac1112b956ec 100644 --- a/tools/power/cpupower/lib/cpupower_intern.h +++ b/tools/power/cpupower/lib/cpupower_intern.h @@ -1,6 +1,11 @@ /* SPDX-License-Identifier: GPL-2.0 */ #define PATH_TO_CPU "/sys/devices/system/cpu/" + +#ifndef MAX_LINE_LEN #define MAX_LINE_LEN 4096 +#endif + #define SYSFS_PATH_MAX 255 unsigned int cpupower_read_sysfs(const char *path, char *buf, size_t buflen); +unsigned int cpupower_write_sysfs(const char *path, char *buf, size_t buflen); diff --git a/tools/power/cpupower/man/cpupower-frequency-info.1 b/tools/power/cpupower/man/cpupower-frequency-info.1 index 6aa8d239dff9..dd545b499480 100644 --- a/tools/power/cpupower/man/cpupower-frequency-info.1 +++ b/tools/power/cpupower/man/cpupower-frequency-info.1 @@ -53,6 +53,9 @@ human\-readable output for the \-f, \-w, \-s and \-y parameters. \fB\-n\fR \fB\-\-no-rounding\fR Output frequencies and latencies without rounding off values. .TP +\fB\-c\fR \fB\-\-perf\fR +Get performances and frequencies capabilities of CPPC, by reading it from hardware (only available on the hardware with CPPC). +.TP .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-idle-info.1 b/tools/power/cpupower/man/cpupower-idle-info.1 index 80a1311fa747..20b6345c53ad 100644 --- a/tools/power/cpupower/man/cpupower-idle-info.1 +++ b/tools/power/cpupower/man/cpupower-idle-info.1 @@ -75,7 +75,7 @@ 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 +https://uefi.org/specifications .SH "FILES" .nf \fI/sys/devices/system/cpu/cpu*/cpuidle/state*\fP diff --git a/tools/power/cpupower/man/cpupower-idle-set.1 b/tools/power/cpupower/man/cpupower-idle-set.1 index 21916cff7516..8cef3c71e19e 100644 --- a/tools/power/cpupower/man/cpupower-idle-set.1 +++ b/tools/power/cpupower/man/cpupower-idle-set.1 @@ -4,7 +4,7 @@ cpupower\-idle\-set \- Utility to set cpu idle state specific kernel options .SH "SYNTAX" .LP -cpupower [ \-c cpulist ] idle\-info [\fIoptions\fP] +cpupower [ \-c cpulist ] idle\-set [\fIoptions\fP] .SH "DESCRIPTION" .LP The cpupower idle\-set subcommand allows to set cpu idle, also called cpu diff --git a/tools/power/cpupower/man/cpupower-monitor.1 b/tools/power/cpupower/man/cpupower-monitor.1 index 70a56476f4b0..8ee737eefa5c 100644 --- a/tools/power/cpupower/man/cpupower-monitor.1 +++ b/tools/power/cpupower/man/cpupower-monitor.1 @@ -170,7 +170,7 @@ displayed. .SH REFERENCES "BIOS and Kernel Developer’s Guide (BKDG) for AMD Family 14h Processors" -http://support.amd.com/us/Processor_TechDocs/43170.pdf +https://support.amd.com/us/Processor_TechDocs/43170.pdf "Intel® Turbo Boost Technology in Intel® Coreâ„¢ Microarchitecture (Nehalem) Based Processors" @@ -178,7 +178,7 @@ http://download.intel.com/design/processor/applnots/320354.pdf "Intel® 64 and IA-32 Architectures Software Developer's Manual Volume 3B: System Programming Guide" -http://www.intel.com/products/processor/manuals +https://www.intel.com/products/processor/manuals .SH FILES .ta diff --git a/tools/power/cpupower/utils/cpufreq-info.c b/tools/power/cpupower/utils/cpufreq-info.c index 6efc0f6b1b11..0646f615fe2d 100644 --- a/tools/power/cpupower/utils/cpufreq-info.c +++ b/tools/power/cpupower/utils/cpufreq-info.c @@ -84,43 +84,6 @@ static void proc_cpufreq_output(void) } static int no_rounding; -static void print_speed(unsigned long speed) -{ - unsigned long tmp; - - if (no_rounding) { - if (speed > 1000000) - printf("%u.%06u GHz", ((unsigned int) speed/1000000), - ((unsigned int) speed%1000000)); - else if (speed > 1000) - printf("%u.%03u MHz", ((unsigned int) speed/1000), - (unsigned int) (speed%1000)); - else - printf("%lu kHz", speed); - } else { - if (speed > 1000000) { - tmp = speed%10000; - if (tmp >= 5000) - speed += 10000; - printf("%u.%02u GHz", ((unsigned int) speed/1000000), - ((unsigned int) (speed%1000000)/10000)); - } else if (speed > 100000) { - tmp = speed%1000; - if (tmp >= 500) - speed += 1000; - printf("%u MHz", ((unsigned int) speed/1000)); - } else if (speed > 1000) { - tmp = speed%100; - if (tmp >= 50) - speed += 100; - printf("%u.%01u MHz", ((unsigned int) speed/1000), - ((unsigned int) (speed%1000)/100)); - } - } - - return; -} - static void print_duration(unsigned long duration) { unsigned long tmp; @@ -183,11 +146,13 @@ static int get_boost_mode_x86(unsigned int cpu) printf(_(" Supported: %s\n"), support ? _("yes") : _("no")); printf(_(" Active: %s\n"), active ? _("yes") : _("no")); - if ((cpupower_cpu_info.vendor == X86_VENDOR_AMD && - cpupower_cpu_info.family >= 0x10) || - cpupower_cpu_info.vendor == X86_VENDOR_HYGON) { - ret = decode_pstates(cpu, cpupower_cpu_info.family, b_states, - pstates, &pstate_no); + if (cpupower_cpu_info.vendor == X86_VENDOR_AMD && + cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_PSTATE) { + return 0; + } else if ((cpupower_cpu_info.vendor == X86_VENDOR_AMD && + cpupower_cpu_info.family >= 0x10) || + cpupower_cpu_info.vendor == X86_VENDOR_HYGON) { + ret = decode_pstates(cpu, b_states, pstates, &pstate_no); if (ret) return ret; @@ -255,11 +220,11 @@ static int get_boost_mode(unsigned int cpu) if (freqs) { printf(_(" boost frequency steps: ")); while (freqs->next) { - print_speed(freqs->frequency); + print_speed(freqs->frequency, no_rounding); printf(", "); freqs = freqs->next; } - print_speed(freqs->frequency); + print_speed(freqs->frequency, no_rounding); printf("\n"); cpufreq_put_available_frequencies(freqs); } @@ -278,7 +243,7 @@ static int get_freq_kernel(unsigned int cpu, unsigned int human) return -EINVAL; } if (human) { - print_speed(freq); + print_speed(freq, no_rounding); } else printf("%lu", freq); printf(_(" (asserted by call to kernel)\n")); @@ -297,7 +262,7 @@ static int get_freq_hardware(unsigned int cpu, unsigned int human) return -EINVAL; } if (human) { - print_speed(freq); + print_speed(freq, no_rounding); } else printf("%lu", freq); printf(_(" (asserted by call to hardware)\n")); @@ -317,9 +282,9 @@ static int get_hardware_limits(unsigned int cpu, unsigned int human) if (human) { printf(_(" hardware limits: ")); - print_speed(min); + print_speed(min, no_rounding); printf(" - "); - print_speed(max); + print_speed(max, no_rounding); printf("\n"); } else { printf("%lu %lu\n", min, max); @@ -351,9 +316,9 @@ static int get_policy(unsigned int cpu) return -EINVAL; } printf(_(" current policy: frequency should be within ")); - print_speed(policy->min); + print_speed(policy->min, no_rounding); printf(_(" and ")); - print_speed(policy->max); + print_speed(policy->max, no_rounding); printf(".\n "); printf(_("The governor \"%s\" may decide which speed to use\n" @@ -437,7 +402,7 @@ static int get_freq_stats(unsigned int cpu, unsigned int human) struct cpufreq_stats *stats = cpufreq_get_stats(cpu, &total_time); while (stats) { if (human) { - print_speed(stats->frequency); + print_speed(stats->frequency, no_rounding); printf(":%.2f%%", (100.0 * stats->time_in_state) / total_time); } else @@ -473,6 +438,17 @@ static int get_latency(unsigned int cpu, unsigned int human) return 0; } +/* --performance / -c */ + +static int get_perf_cap(unsigned int cpu) +{ + if (cpupower_cpu_info.vendor == X86_VENDOR_AMD && + cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_PSTATE) + amd_pstate_show_perf_and_freq(cpu, no_rounding); + + return 0; +} + static void debug_output_one(unsigned int cpu) { struct cpufreq_available_frequencies *freqs; @@ -487,11 +463,11 @@ static void debug_output_one(unsigned int cpu) if (freqs) { printf(_(" available frequency steps: ")); while (freqs->next) { - print_speed(freqs->frequency); + print_speed(freqs->frequency, no_rounding); printf(", "); freqs = freqs->next; } - print_speed(freqs->frequency); + print_speed(freqs->frequency, no_rounding); printf("\n"); cpufreq_put_available_frequencies(freqs); } @@ -501,6 +477,7 @@ static void debug_output_one(unsigned int cpu) if (get_freq_hardware(cpu, 1) < 0) get_freq_kernel(cpu, 1); get_boost_mode(cpu); + get_perf_cap(cpu); } static struct option info_opts[] = { @@ -519,6 +496,7 @@ static struct option info_opts[] = { {"proc", no_argument, NULL, 'o'}, {"human", no_argument, NULL, 'm'}, {"no-rounding", no_argument, NULL, 'n'}, + {"performance", no_argument, NULL, 'c'}, { }, }; @@ -532,7 +510,7 @@ int cmd_freq_info(int argc, char **argv) int output_param = 0; do { - ret = getopt_long(argc, argv, "oefwldpgrasmybn", info_opts, + ret = getopt_long(argc, argv, "oefwldpgrasmybnc", info_opts, NULL); switch (ret) { case '?': @@ -555,6 +533,7 @@ int cmd_freq_info(int argc, char **argv) case 'e': case 's': case 'y': + case 'c': if (output_param) { output_param = -1; cont = 0; @@ -661,6 +640,9 @@ int cmd_freq_info(int argc, char **argv) case 'y': ret = get_latency(cpu, human); break; + case 'c': + ret = get_perf_cap(cpu); + break; } if (ret) return ret; diff --git a/tools/power/cpupower/utils/cpufreq-set.c b/tools/power/cpupower/utils/cpufreq-set.c index 6ed82fba5aaa..c5e60a39cfa6 100644 --- a/tools/power/cpupower/utils/cpufreq-set.c +++ b/tools/power/cpupower/utils/cpufreq-set.c @@ -99,13 +99,17 @@ static unsigned long string_to_frequency(const char *str) continue; if (str[cp] == '.') { - while (power > -1 && isdigit(str[cp+1])) - cp++, power--; + while (power > -1 && isdigit(str[cp+1])) { + cp++; + power--; + } } - if (power >= -1) /* not enough => pad */ + if (power >= -1) { /* not enough => pad */ pad = power + 1; - else /* to much => strip */ - pad = 0, cp += power + 1; + } else { /* too much => strip */ + pad = 0; + cp += power + 1; + } /* check bounds */ if (cp <= 0 || cp + pad > NORM_FREQ_LEN - 1) return 0; @@ -311,6 +315,7 @@ int cmd_freq_set(int argc, char **argv) } } + get_cpustate(); /* loop over CPUs */ for (cpu = bitmask_first(cpus_chosen); @@ -328,5 +333,7 @@ int cmd_freq_set(int argc, char **argv) } } + print_offline_cpus(); + return 0; } diff --git a/tools/power/cpupower/utils/cpuidle-set.c b/tools/power/cpupower/utils/cpuidle-set.c index 569f268f4c7f..46158928f9ad 100644 --- a/tools/power/cpupower/utils/cpuidle-set.c +++ b/tools/power/cpupower/utils/cpuidle-set.c @@ -95,6 +95,8 @@ int cmd_idle_set(int argc, char **argv) exit(EXIT_FAILURE); } + get_cpustate(); + /* Default is: set all CPUs */ if (bitmask_isallclear(cpus_chosen)) bitmask_setall(cpus_chosen); @@ -181,5 +183,7 @@ int cmd_idle_set(int argc, char **argv) break; } } + + print_offline_cpus(); return EXIT_SUCCESS; } diff --git a/tools/power/cpupower/utils/cpupower-info.c b/tools/power/cpupower/utils/cpupower-info.c index d3755ea70d4d..06345b543786 100644 --- a/tools/power/cpupower/utils/cpupower-info.c +++ b/tools/power/cpupower/utils/cpupower-info.c @@ -62,7 +62,7 @@ int cmd_info(int argc, char **argv) default: print_wrong_arg_exit(); } - }; + } if (!params.params) params.params = 0x7; @@ -101,7 +101,7 @@ int cmd_info(int argc, char **argv) } if (params.perf_bias) { - ret = msr_intel_get_perf_bias(cpu); + ret = cpupower_intel_get_perf_bias(cpu); if (ret < 0) { fprintf(stderr, _("Could not read perf-bias value[%d]\n"), ret); diff --git a/tools/power/cpupower/utils/cpupower-set.c b/tools/power/cpupower/utils/cpupower-set.c index 3cca6f715dd9..180d5ba877e6 100644 --- a/tools/power/cpupower/utils/cpupower-set.c +++ b/tools/power/cpupower/utils/cpupower-set.c @@ -72,7 +72,7 @@ int cmd_set(int argc, char **argv) default: print_wrong_arg_exit(); } - }; + } if (!params.params) print_wrong_arg_exit(); @@ -95,7 +95,7 @@ int cmd_set(int argc, char **argv) } if (params.perf_bias) { - ret = msr_intel_set_perf_bias(cpu, perf_bias); + ret = cpupower_intel_set_perf_bias(cpu, perf_bias); if (ret) { fprintf(stderr, _("Error setting perf-bias " "value on CPU %d\n"), cpu); diff --git a/tools/power/cpupower/utils/cpupower.c b/tools/power/cpupower/utils/cpupower.c index 8e3d08042825..8ac3304a9957 100644 --- a/tools/power/cpupower/utils/cpupower.c +++ b/tools/power/cpupower/utils/cpupower.c @@ -34,6 +34,8 @@ int run_as_root; int base_cpu; /* Affected cpus chosen by -c/--cpu param */ struct bitmask *cpus_chosen; +struct bitmask *online_cpus; +struct bitmask *offline_cpus; #ifdef DEBUG int be_verbose; @@ -178,6 +180,8 @@ int main(int argc, const char *argv[]) char pathname[32]; cpus_chosen = bitmask_alloc(sysconf(_SC_NPROCESSORS_CONF)); + online_cpus = bitmask_alloc(sysconf(_SC_NPROCESSORS_CONF)); + offline_cpus = bitmask_alloc(sysconf(_SC_NPROCESSORS_CONF)); argc--; argv += 1; @@ -230,6 +234,10 @@ int main(int argc, const char *argv[]) ret = p->main(argc, argv); if (cpus_chosen) bitmask_free(cpus_chosen); + if (online_cpus) + bitmask_free(online_cpus); + if (offline_cpus) + bitmask_free(offline_cpus); return ret; } print_help(); diff --git a/tools/power/cpupower/utils/helpers/amd.c b/tools/power/cpupower/utils/helpers/amd.c index 7c4f83a8c973..c519cc89c97f 100644 --- a/tools/power/cpupower/utils/helpers/amd.c +++ b/tools/power/cpupower/utils/helpers/amd.c @@ -8,12 +8,16 @@ #include <pci/pci.h> #include "helpers/helpers.h" +#include "cpufreq.h" +#include "acpi_cppc.h" +/* ACPI P-States Helper Functions for AMD Processors ***************/ #define MSR_AMD_PSTATE_STATUS 0xc0010063 #define MSR_AMD_PSTATE 0xc0010064 #define MSR_AMD_PSTATE_LIMIT 0xc0010061 -union msr_pstate { +union core_pstate { + /* pre fam 17h: */ struct { unsigned fid:6; unsigned did:3; @@ -26,7 +30,8 @@ union msr_pstate { unsigned idddiv:2; unsigned res3:21; unsigned en:1; - } bits; + } pstate; + /* since fam 17h: */ struct { unsigned fid:8; unsigned did:6; @@ -35,37 +40,37 @@ union msr_pstate { unsigned idddiv:2; unsigned res1:31; unsigned en:1; - } fam17h_bits; + } pstatedef; unsigned long long val; }; -static int get_did(int family, union msr_pstate pstate) +static int get_did(union core_pstate pstate) { int t; - if (family == 0x12) + if (cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_PSTATEDEF) + t = pstate.pstatedef.did; + else if (cpupower_cpu_info.family == 0x12) t = pstate.val & 0xf; - else if (family == 0x17 || family == 0x18) - t = pstate.fam17h_bits.did; else - t = pstate.bits.did; + t = pstate.pstate.did; return t; } -static int get_cof(int family, union msr_pstate pstate) +static int get_cof(union core_pstate pstate) { int t; int fid, did, cof; - did = get_did(family, pstate); - if (family == 0x17 || family == 0x18) { - fid = pstate.fam17h_bits.fid; + did = get_did(pstate); + if (cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_PSTATEDEF) { + fid = pstate.pstatedef.fid; cof = 200 * fid / did; } else { t = 0x10; - fid = pstate.bits.fid; - if (family == 0x11) + fid = pstate.pstate.fid; + if (cpupower_cpu_info.family == 0x11) t = 0x8; cof = (100 * (fid + t)) >> did; } @@ -74,8 +79,7 @@ static int get_cof(int family, union msr_pstate pstate) /* Needs: * cpu -> the cpu that gets evaluated - * cpu_family -> The cpu's family (0x10, 0x12,...) - * boots_states -> how much boost states the machines support + * boost_states -> how much boost states the machines support * * Fills up: * pstates -> a pointer to an array of size MAX_HW_PSTATES @@ -85,31 +89,23 @@ static int get_cof(int family, union msr_pstate pstate) * * returns zero on success, -1 on failure */ -int decode_pstates(unsigned int cpu, unsigned int cpu_family, - int boost_states, unsigned long *pstates, int *no) +int decode_pstates(unsigned int cpu, int boost_states, + unsigned long *pstates, int *no) { - int i, psmax, pscur; - union msr_pstate pstate; + int i, psmax; + union core_pstate pstate; unsigned long long val; - /* Only read out frequencies from HW when CPU might be boostable - to keep the code as short and clean as possible. - Otherwise frequencies are exported via ACPI tables. - */ - if (cpu_family < 0x10 || cpu_family == 0x14) + /* Only read out frequencies from HW if HW Pstate is supported, + * otherwise frequencies are exported via ACPI tables. + */ + if (!(cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_HW_PSTATE)) return -1; if (read_msr(cpu, MSR_AMD_PSTATE_LIMIT, &val)) return -1; psmax = (val >> 4) & 0x7; - - if (read_msr(cpu, MSR_AMD_PSTATE_STATUS, &val)) - return -1; - - pscur = val & 0x7; - - pscur += boost_states; psmax += boost_states; for (i = 0; i <= psmax; i++) { if (i >= MAX_HW_PSTATES) { @@ -119,12 +115,12 @@ int decode_pstates(unsigned int cpu, unsigned int cpu_family, } if (read_msr(cpu, MSR_AMD_PSTATE + i, &pstate.val)) return -1; - if ((cpu_family == 0x17) && (!pstate.fam17h_bits.en)) - continue; - else if (!pstate.bits.en) + + /* The enabled bit (bit 63) is common for all families */ + if (!pstate.pstatedef.en) continue; - pstates[i] = get_cof(cpu_family, pstate); + pstates[i] = get_cof(pstate); } *no = i; return 0; @@ -153,4 +149,78 @@ int amd_pci_get_num_boost_states(int *active, int *states) pci_cleanup(pci_acc); return 0; } + +/* ACPI P-States Helper Functions for AMD Processors ***************/ + +/* AMD P-State Helper Functions ************************************/ +enum amd_pstate_value { + AMD_PSTATE_HIGHEST_PERF, + AMD_PSTATE_MAX_FREQ, + AMD_PSTATE_LOWEST_NONLINEAR_FREQ, + MAX_AMD_PSTATE_VALUE_READ_FILES, +}; + +static const char *amd_pstate_value_files[MAX_AMD_PSTATE_VALUE_READ_FILES] = { + [AMD_PSTATE_HIGHEST_PERF] = "amd_pstate_highest_perf", + [AMD_PSTATE_MAX_FREQ] = "amd_pstate_max_freq", + [AMD_PSTATE_LOWEST_NONLINEAR_FREQ] = "amd_pstate_lowest_nonlinear_freq", +}; + +static unsigned long amd_pstate_get_data(unsigned int cpu, + enum amd_pstate_value value) +{ + return cpufreq_get_sysfs_value_from_table(cpu, + amd_pstate_value_files, + value, + MAX_AMD_PSTATE_VALUE_READ_FILES); +} + +void amd_pstate_boost_init(unsigned int cpu, int *support, int *active) +{ + unsigned long highest_perf, nominal_perf, cpuinfo_min, + cpuinfo_max, amd_pstate_max; + + highest_perf = amd_pstate_get_data(cpu, AMD_PSTATE_HIGHEST_PERF); + nominal_perf = acpi_cppc_get_data(cpu, NOMINAL_PERF); + + *support = highest_perf > nominal_perf ? 1 : 0; + if (!(*support)) + return; + + cpufreq_get_hardware_limits(cpu, &cpuinfo_min, &cpuinfo_max); + amd_pstate_max = amd_pstate_get_data(cpu, AMD_PSTATE_MAX_FREQ); + + *active = cpuinfo_max == amd_pstate_max ? 1 : 0; +} + +void amd_pstate_show_perf_and_freq(unsigned int cpu, int no_rounding) +{ + printf(_(" AMD PSTATE Highest Performance: %lu. Maximum Frequency: "), + amd_pstate_get_data(cpu, AMD_PSTATE_HIGHEST_PERF)); + /* + * If boost isn't active, the cpuinfo_max doesn't indicate real max + * frequency. So we read it back from amd-pstate sysfs entry. + */ + print_speed(amd_pstate_get_data(cpu, AMD_PSTATE_MAX_FREQ), no_rounding); + printf(".\n"); + + printf(_(" AMD PSTATE Nominal Performance: %lu. Nominal Frequency: "), + acpi_cppc_get_data(cpu, NOMINAL_PERF)); + print_speed(acpi_cppc_get_data(cpu, NOMINAL_FREQ) * 1000, + no_rounding); + printf(".\n"); + + printf(_(" AMD PSTATE Lowest Non-linear Performance: %lu. Lowest Non-linear Frequency: "), + acpi_cppc_get_data(cpu, LOWEST_NONLINEAR_PERF)); + print_speed(amd_pstate_get_data(cpu, AMD_PSTATE_LOWEST_NONLINEAR_FREQ), + no_rounding); + printf(".\n"); + + printf(_(" AMD PSTATE Lowest Performance: %lu. Lowest Frequency: "), + acpi_cppc_get_data(cpu, LOWEST_PERF)); + print_speed(acpi_cppc_get_data(cpu, LOWEST_FREQ) * 1000, no_rounding); + printf(".\n"); +} + +/* AMD P-State Helper Functions ************************************/ #endif /* defined(__i386__) || defined(__x86_64__) */ diff --git a/tools/power/cpupower/utils/helpers/bitmask.c b/tools/power/cpupower/utils/helpers/bitmask.c index 6c7932f5bd66..649d87cb8b0f 100644 --- a/tools/power/cpupower/utils/helpers/bitmask.c +++ b/tools/power/cpupower/utils/helpers/bitmask.c @@ -26,11 +26,11 @@ struct bitmask *bitmask_alloc(unsigned int n) struct bitmask *bmp; bmp = malloc(sizeof(*bmp)); - if (bmp == 0) + if (!bmp) return 0; bmp->size = n; bmp->maskp = calloc(longsperbits(n), sizeof(unsigned long)); - if (bmp->maskp == 0) { + if (!bmp->maskp) { free(bmp); return 0; } @@ -40,7 +40,7 @@ struct bitmask *bitmask_alloc(unsigned int n) /* Free `struct bitmask` */ void bitmask_free(struct bitmask *bmp) { - if (bmp == 0) + if (!bmp) return; free(bmp->maskp); bmp->maskp = (unsigned long *)0xdeadcdef; /* double free tripwire */ diff --git a/tools/power/cpupower/utils/helpers/cpuid.c b/tools/power/cpupower/utils/helpers/cpuid.c index 73bfafc60e9b..eae91f11d187 100644 --- a/tools/power/cpupower/utils/helpers/cpuid.c +++ b/tools/power/cpupower/utils/helpers/cpuid.c @@ -128,13 +128,40 @@ out: /* AMD or Hygon Boost state enable/disable register */ if (cpu_info->vendor == X86_VENDOR_AMD || cpu_info->vendor == X86_VENDOR_HYGON) { - if (ext_cpuid_level >= 0x80000007 && - (cpuid_edx(0x80000007) & (1 << 9))) - cpu_info->caps |= CPUPOWER_CAP_AMD_CBP; + if (ext_cpuid_level >= 0x80000007) { + if (cpuid_edx(0x80000007) & (1 << 9)) { + cpu_info->caps |= CPUPOWER_CAP_AMD_CPB; + + if (cpu_info->family >= 0x17) + cpu_info->caps |= CPUPOWER_CAP_AMD_CPB_MSR; + } + + if ((cpuid_edx(0x80000007) & (1 << 7)) && + cpu_info->family != 0x14) { + /* HW pstate was not implemented in family 0x14 */ + cpu_info->caps |= CPUPOWER_CAP_AMD_HW_PSTATE; + + if (cpu_info->family >= 0x17) + cpu_info->caps |= CPUPOWER_CAP_AMD_PSTATEDEF; + } + } if (ext_cpuid_level >= 0x80000008 && cpuid_ebx(0x80000008) & (1 << 4)) cpu_info->caps |= CPUPOWER_CAP_AMD_RDPRU; + + if (cpupower_amd_pstate_enabled()) { + cpu_info->caps |= CPUPOWER_CAP_AMD_PSTATE; + + /* + * If AMD P-State is enabled, the firmware will treat + * AMD P-State function as high priority. + */ + cpu_info->caps &= ~CPUPOWER_CAP_AMD_CPB; + cpu_info->caps &= ~CPUPOWER_CAP_AMD_CPB_MSR; + cpu_info->caps &= ~CPUPOWER_CAP_AMD_HW_PSTATE; + cpu_info->caps &= ~CPUPOWER_CAP_AMD_PSTATEDEF; + } } if (cpu_info->vendor == X86_VENDOR_INTEL) { diff --git a/tools/power/cpupower/utils/helpers/helpers.h b/tools/power/cpupower/utils/helpers/helpers.h index c258eeccd05f..96e4bede078b 100644 --- a/tools/power/cpupower/utils/helpers/helpers.h +++ b/tools/power/cpupower/utils/helpers/helpers.h @@ -11,6 +11,7 @@ #include <libintl.h> #include <locale.h> +#include <stdbool.h> #include "helpers/bitmask.h" #include <cpupower.h> @@ -64,12 +65,16 @@ enum cpupower_cpu_vendor {X86_VENDOR_UNKNOWN = 0, X86_VENDOR_INTEL, #define CPUPOWER_CAP_INV_TSC 0x00000001 #define CPUPOWER_CAP_APERF 0x00000002 -#define CPUPOWER_CAP_AMD_CBP 0x00000004 +#define CPUPOWER_CAP_AMD_CPB 0x00000004 #define CPUPOWER_CAP_PERF_BIAS 0x00000008 #define CPUPOWER_CAP_HAS_TURBO_RATIO 0x00000010 #define CPUPOWER_CAP_IS_SNB 0x00000020 #define CPUPOWER_CAP_INTEL_IDA 0x00000040 #define CPUPOWER_CAP_AMD_RDPRU 0x00000080 +#define CPUPOWER_CAP_AMD_HW_PSTATE 0x00000100 +#define CPUPOWER_CAP_AMD_PSTATEDEF 0x00000200 +#define CPUPOWER_CAP_AMD_CPB_MSR 0x00000400 +#define CPUPOWER_CAP_AMD_PSTATE 0x00000800 #define CPUPOWER_AMD_CPBDIS 0x02000000 @@ -94,6 +99,8 @@ struct cpupower_cpu_info { */ extern int get_cpu_info(struct cpupower_cpu_info *cpu_info); extern struct cpupower_cpu_info cpupower_cpu_info; + + /* cpuid and cpuinfo helpers **************************/ /* X86 ONLY ****************************************/ @@ -105,8 +112,8 @@ extern struct cpupower_cpu_info cpupower_cpu_info; extern int read_msr(int cpu, unsigned int idx, unsigned long long *val); extern int write_msr(int cpu, unsigned int idx, unsigned long long val); -extern int msr_intel_set_perf_bias(unsigned int cpu, unsigned int val); -extern int msr_intel_get_perf_bias(unsigned int cpu); +extern int cpupower_intel_set_perf_bias(unsigned int cpu, unsigned int val); +extern int cpupower_intel_get_perf_bias(unsigned int cpu); extern unsigned long long msr_intel_get_turbo_ratio(unsigned int cpu); /* Read/Write msr ****************************/ @@ -123,13 +130,23 @@ extern struct pci_dev *pci_slot_func_init(struct pci_access **pacc, /* AMD HW pstate decoding **************************/ -extern int decode_pstates(unsigned int cpu, unsigned int cpu_family, - int boost_states, unsigned long *pstates, int *no); +extern int decode_pstates(unsigned int cpu, int boost_states, + unsigned long *pstates, int *no); /* AMD HW pstate decoding **************************/ extern int cpufreq_has_boost_support(unsigned int cpu, int *support, int *active, int * states); + +/* AMD P-State stuff **************************/ +bool cpupower_amd_pstate_enabled(void); +void amd_pstate_boost_init(unsigned int cpu, + int *support, int *active); +void amd_pstate_show_perf_and_freq(unsigned int cpu, + int no_rounding); + +/* AMD P-State stuff **************************/ + /* * CPUID functions returning a single datum */ @@ -141,18 +158,17 @@ unsigned int cpuid_edx(unsigned int op); /* cpuid and cpuinfo helpers **************************/ /* X86 ONLY ********************************************/ #else -static inline int decode_pstates(unsigned int cpu, unsigned int cpu_family, - int boost_states, unsigned long *pstates, - int *no) +static inline int decode_pstates(unsigned int cpu, int boost_states, + unsigned long *pstates, int *no) { return -1; }; static inline int read_msr(int cpu, unsigned int idx, unsigned long long *val) { return -1; }; static inline int write_msr(int cpu, unsigned int idx, unsigned long long val) { return -1; }; -static inline int msr_intel_set_perf_bias(unsigned int cpu, unsigned int val) +static inline int cpupower_intel_set_perf_bias(unsigned int cpu, unsigned int val) { return -1; }; -static inline int msr_intel_get_perf_bias(unsigned int cpu) +static inline int cpupower_intel_get_perf_bias(unsigned int cpu) { return -1; }; static inline unsigned long long msr_intel_get_turbo_ratio(unsigned int cpu) { return 0; }; @@ -163,6 +179,15 @@ static inline int cpufreq_has_boost_support(unsigned int cpu, int *support, int *active, int * states) { return -1; } +static inline bool cpupower_amd_pstate_enabled(void) +{ return false; } +static inline void amd_pstate_boost_init(unsigned int cpu, int *support, + int *active) +{} +static inline void amd_pstate_show_perf_and_freq(unsigned int cpu, + int no_rounding) +{} + /* cpuid and cpuinfo helpers **************************/ static inline unsigned int cpuid_eax(unsigned int op) { return 0; }; @@ -171,4 +196,15 @@ static inline unsigned int cpuid_ecx(unsigned int op) { return 0; }; static inline unsigned int cpuid_edx(unsigned int op) { return 0; }; #endif /* defined(__i386__) || defined(__x86_64__) */ +/* + * CPU State related functions + */ +extern struct bitmask *online_cpus; +extern struct bitmask *offline_cpus; + +void get_cpustate(void); +void print_online_cpus(void); +void print_offline_cpus(void); +void print_speed(unsigned long speed, int no_rounding); + #endif /* __CPUPOWERUTILS_HELPERS__ */ diff --git a/tools/power/cpupower/utils/helpers/misc.c b/tools/power/cpupower/utils/helpers/misc.c index f406adc40bad..9547b29254a7 100644 --- a/tools/power/cpupower/utils/helpers/misc.c +++ b/tools/power/cpupower/utils/helpers/misc.c @@ -1,24 +1,29 @@ // SPDX-License-Identifier: GPL-2.0 -#if defined(__i386__) || defined(__x86_64__) + +#include <stdio.h> +#include <errno.h> +#include <stdlib.h> +#include <string.h> #include "helpers/helpers.h" +#include "helpers/sysfs.h" +#include "cpufreq.h" + +#if defined(__i386__) || defined(__x86_64__) + +#include "cpupower_intern.h" #define MSR_AMD_HWCR 0xc0010015 int cpufreq_has_boost_support(unsigned int cpu, int *support, int *active, int *states) { - struct cpupower_cpu_info cpu_info; int ret; unsigned long long val; *support = *active = *states = 0; - ret = get_cpu_info(&cpu_info); - if (ret) - return ret; - - if (cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_CBP) { + if (cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_CPB) { *support = 1; /* AMD Family 0x17 does not utilize PCI D18F4 like prior @@ -26,7 +31,7 @@ int cpufreq_has_boost_support(unsigned int cpu, int *support, int *active, * has Hardware determined variable increments instead. */ - if (cpu_info.family == 0x17 || cpu_info.family == 0x18) { + if (cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_CPB_MSR) { if (!read_msr(cpu, MSR_AMD_HWCR, &val)) { if (!(val & CPUPOWER_AMD_CPBDIS)) *active = 1; @@ -36,8 +41,166 @@ int cpufreq_has_boost_support(unsigned int cpu, int *support, int *active, if (ret) return ret; } + } else if (cpupower_cpu_info.caps & CPUPOWER_CAP_AMD_PSTATE) { + amd_pstate_boost_init(cpu, support, active); } else if (cpupower_cpu_info.caps & CPUPOWER_CAP_INTEL_IDA) *support = *active = 1; return 0; } + +int cpupower_intel_get_perf_bias(unsigned int cpu) +{ + char linebuf[MAX_LINE_LEN]; + char path[SYSFS_PATH_MAX]; + unsigned long val; + char *endp; + + if (!(cpupower_cpu_info.caps & CPUPOWER_CAP_PERF_BIAS)) + return -1; + + snprintf(path, sizeof(path), PATH_TO_CPU "cpu%u/power/energy_perf_bias", cpu); + + if (cpupower_read_sysfs(path, linebuf, MAX_LINE_LEN) == 0) + return -1; + + val = strtol(linebuf, &endp, 0); + if (endp == linebuf || errno == ERANGE) + return -1; + + return val; +} + +int cpupower_intel_set_perf_bias(unsigned int cpu, unsigned int val) +{ + char path[SYSFS_PATH_MAX]; + char linebuf[3] = {}; + + if (!(cpupower_cpu_info.caps & CPUPOWER_CAP_PERF_BIAS)) + return -1; + + snprintf(path, sizeof(path), PATH_TO_CPU "cpu%u/power/energy_perf_bias", cpu); + snprintf(linebuf, sizeof(linebuf), "%d", val); + + if (cpupower_write_sysfs(path, linebuf, 3) <= 0) + return -1; + + return 0; +} + +bool cpupower_amd_pstate_enabled(void) +{ + char *driver = cpufreq_get_driver(0); + bool ret = false; + + if (!driver) + return ret; + + if (!strcmp(driver, "amd-pstate")) + ret = true; + + cpufreq_put_driver(driver); + + return ret; +} + #endif /* #if defined(__i386__) || defined(__x86_64__) */ + +/* get_cpustate + * + * Gather the information of all online CPUs into bitmask struct + */ +void get_cpustate(void) +{ + unsigned int cpu = 0; + + bitmask_clearall(online_cpus); + bitmask_clearall(offline_cpus); + + for (cpu = bitmask_first(cpus_chosen); + cpu <= bitmask_last(cpus_chosen); cpu++) { + + if (cpupower_is_cpu_online(cpu) == 1) + bitmask_setbit(online_cpus, cpu); + else + bitmask_setbit(offline_cpus, cpu); + + continue; + } +} + +/* print_online_cpus + * + * Print the CPU numbers of all CPUs that are online currently + */ +void print_online_cpus(void) +{ + int str_len = 0; + char *online_cpus_str = NULL; + + str_len = online_cpus->size * 5; + online_cpus_str = (void *)malloc(sizeof(char) * str_len); + + if (!bitmask_isallclear(online_cpus)) { + bitmask_displaylist(online_cpus_str, str_len, online_cpus); + printf(_("Following CPUs are online:\n%s\n"), online_cpus_str); + } +} + +/* print_offline_cpus + * + * Print the CPU numbers of all CPUs that are offline currently + */ +void print_offline_cpus(void) +{ + int str_len = 0; + char *offline_cpus_str = NULL; + + str_len = offline_cpus->size * 5; + offline_cpus_str = (void *)malloc(sizeof(char) * str_len); + + if (!bitmask_isallclear(offline_cpus)) { + bitmask_displaylist(offline_cpus_str, str_len, offline_cpus); + printf(_("Following CPUs are offline:\n%s\n"), offline_cpus_str); + printf(_("cpupower set operation was not performed on them\n")); + } +} + +/* + * print_speed + * + * Print the exact CPU frequency with appropriate unit + */ +void print_speed(unsigned long speed, int no_rounding) +{ + unsigned long tmp; + + if (no_rounding) { + if (speed > 1000000) + printf("%u.%06u GHz", ((unsigned int)speed / 1000000), + ((unsigned int)speed % 1000000)); + else if (speed > 1000) + printf("%u.%03u MHz", ((unsigned int)speed / 1000), + (unsigned int)(speed % 1000)); + else + printf("%lu kHz", speed); + } else { + if (speed > 1000000) { + tmp = speed % 10000; + if (tmp >= 5000) + speed += 10000; + printf("%u.%02u GHz", ((unsigned int)speed / 1000000), + ((unsigned int)(speed % 1000000) / 10000)); + } else if (speed > 100000) { + tmp = speed % 1000; + if (tmp >= 500) + speed += 1000; + printf("%u MHz", ((unsigned int)speed / 1000)); + } else if (speed > 1000) { + tmp = speed % 100; + if (tmp >= 50) + speed += 100; + printf("%u.%01u MHz", ((unsigned int)speed / 1000), + ((unsigned int)(speed % 1000) / 100)); + } + } +} diff --git a/tools/power/cpupower/utils/helpers/msr.c b/tools/power/cpupower/utils/helpers/msr.c index ab9950748838..8b0b6be74bb8 100644 --- a/tools/power/cpupower/utils/helpers/msr.c +++ b/tools/power/cpupower/utils/helpers/msr.c @@ -11,7 +11,6 @@ /* Intel specific MSRs */ #define MSR_IA32_PERF_STATUS 0x198 #define MSR_IA32_MISC_ENABLES 0x1a0 -#define MSR_IA32_ENERGY_PERF_BIAS 0x1b0 #define MSR_NEHALEM_TURBO_RATIO_LIMIT 0x1ad /* @@ -73,33 +72,6 @@ int write_msr(int cpu, unsigned int idx, unsigned long long val) return -1; } -int msr_intel_get_perf_bias(unsigned int cpu) -{ - unsigned long long val; - int ret; - - if (!(cpupower_cpu_info.caps & CPUPOWER_CAP_PERF_BIAS)) - return -1; - - ret = read_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, &val); - if (ret) - return ret; - return val; -} - -int msr_intel_set_perf_bias(unsigned int cpu, unsigned int val) -{ - int ret; - - if (!(cpupower_cpu_info.caps & CPUPOWER_CAP_PERF_BIAS)) - return -1; - - ret = write_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, val); - if (ret) - return ret; - return 0; -} - unsigned long long msr_intel_get_turbo_ratio(unsigned int cpu) { unsigned long long val; 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 20f46348271b..5edd35bd9ee9 100644 --- a/tools/power/cpupower/utils/idle_monitor/amd_fam14h_idle.c +++ b/tools/power/cpupower/utils/idle_monitor/amd_fam14h_idle.c @@ -117,7 +117,7 @@ static int amd_fam14h_get_pci_info(struct cstate *state, break; default: return -1; - }; + } return 0; } diff --git a/tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c b/tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c index a65f7d011513..8b42c2f0a5b0 100644 --- a/tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c +++ b/tools/power/cpupower/utils/idle_monitor/cpuidle_sysfs.c @@ -53,7 +53,7 @@ static int cpuidle_start(void) dprint("CPU %d - State: %d - Val: %llu\n", cpu, state, previous_count[cpu][state]); } - }; + } return 0; } @@ -72,7 +72,7 @@ static int cpuidle_stop(void) dprint("CPU %d - State: %d - Val: %llu\n", cpu, state, previous_count[cpu][state]); } - }; + } return 0; } @@ -172,7 +172,7 @@ static struct cpuidle_monitor *cpuidle_register(void) cpuidle_cstates[num].id = num; cpuidle_cstates[num].get_count_percent = cpuidle_get_count_percent; - }; + } /* Free this at program termination */ previous_count = malloc(sizeof(long long *) * cpu_count); diff --git a/tools/power/cpupower/utils/idle_monitor/hsw_ext_idle.c b/tools/power/cpupower/utils/idle_monitor/hsw_ext_idle.c index 97ad3233a521..55e55b6b42f9 100644 --- a/tools/power/cpupower/utils/idle_monitor/hsw_ext_idle.c +++ b/tools/power/cpupower/utils/idle_monitor/hsw_ext_idle.c @@ -79,7 +79,7 @@ static int hsw_ext_get_count(enum intel_hsw_ext_id id, unsigned long long *val, break; default: return -1; - }; + } if (read_msr(cpu, msr, val)) return -1; return 0; diff --git a/tools/power/cpupower/utils/idle_monitor/nhm_idle.c b/tools/power/cpupower/utils/idle_monitor/nhm_idle.c index 114271165182..16eaf006f61f 100644 --- a/tools/power/cpupower/utils/idle_monitor/nhm_idle.c +++ b/tools/power/cpupower/utils/idle_monitor/nhm_idle.c @@ -91,7 +91,7 @@ static int nhm_get_count(enum intel_nhm_id id, unsigned long long *val, break; default: return -1; - }; + } if (read_msr(cpu, msr, val)) return -1; diff --git a/tools/power/cpupower/utils/idle_monitor/snb_idle.c b/tools/power/cpupower/utils/idle_monitor/snb_idle.c index df8b223cc096..811d63ab17a7 100644 --- a/tools/power/cpupower/utils/idle_monitor/snb_idle.c +++ b/tools/power/cpupower/utils/idle_monitor/snb_idle.c @@ -77,7 +77,7 @@ static int snb_get_count(enum intel_snb_id id, unsigned long long *val, break; default: return -1; - }; + } if (read_msr(cpu, msr, val)) return -1; return 0; diff --git a/tools/power/pm-graph/Makefile b/tools/power/pm-graph/Makefile index 845541544570..b5310832c19c 100644 --- a/tools/power/pm-graph/Makefile +++ b/tools/power/pm-graph/Makefile @@ -41,6 +41,10 @@ uninstall : if [ -d $(DESTDIR)$(PREFIX)/lib/pm-graph/config ] ; then \ rmdir $(DESTDIR)$(PREFIX)/lib/pm-graph/config; \ fi; + rm -f $(DESTDIR)$(PREFIX)/lib/pm-graph/__pycache__/* + if [ -d $(DESTDIR)$(PREFIX)/lib/pm-graph/__pycache__ ] ; then \ + rmdir $(DESTDIR)$(PREFIX)/lib/pm-graph/__pycache__; \ + fi; rm -f $(DESTDIR)$(PREFIX)/lib/pm-graph/* if [ -d $(DESTDIR)$(PREFIX)/lib/pm-graph ] ; then \ rmdir $(DESTDIR)$(PREFIX)/lib/pm-graph; \ diff --git a/tools/power/pm-graph/README b/tools/power/pm-graph/README index 96259f6e5715..3213dbe63b74 100644 --- a/tools/power/pm-graph/README +++ b/tools/power/pm-graph/README @@ -1,26 +1,27 @@ - p m - g r a p h + _ + _ __ _ __ ___ __ _ _ __ __ _ _ __ | |__ + | '_ \| '_ ` _ \ _____ / _` | '__/ _` | '_ \| '_ \ + | |_) | | | | | |_____| (_| | | | (_| | |_) | | | | + | .__/|_| |_| |_| \__, |_| \__,_| .__/|_| |_| + |_| |___/ |_| pm-graph: suspend/resume/boot timing analysis tools - Version: 5.5 + Version: 5.10 Author: Todd Brandt <todd.e.brandt@intel.com> - Home Page: https://01.org/pm-graph + Home Page: https://www.intel.com/content/www/us/en/developer/topic-technology/open/pm-graph/overview.html Report bugs/issues at bugzilla.kernel.org Tools/pm-graph - https://bugzilla.kernel.org/buglist.cgi?component=pm-graph&product=Tools Full documentation available online & in man pages - Getting Started: - https://01.org/pm-graph/documentation/getting-started + https://www.intel.com/content/www/us/en/developer/articles/technical/usage.html - - Config File Format: - https://01.org/pm-graph/documentation/3-config-file-format + - Feature Summary: + https://www.intel.com/content/www/us/en/developer/topic-technology/open/pm-graph/features.html - upstream version in git: - https://github.com/intel/pm-graph/ - - Requirements: - - runs with python2 or python3, choice is made by /usr/bin/python link - - python2 now requires python-configparser be installed + git clone https://github.com/intel/pm-graph/ Table of Contents - Overview @@ -29,6 +30,8 @@ - Basic Usage - Dev Mode Usage - Proc Mode Usage + - Endurance Testing + - Usage Examples - Configuration Files - Usage Examples - Config File Options @@ -54,15 +57,18 @@ | SETUP | ------------------------------------------------------------------ - These packages are required to execute the scripts + Package Requirements + - runs with python2 or python3, choice is made by /usr/bin/python link - python - - python-requests + - python-configparser (for python2 sleepgraph) + - python-requests (for stresstester.py) + - linux-tools-common (for turbostat usage in sleepgraph) Ubuntu: - sudo apt-get install python python-requests + sudo apt-get install python python-configparser python-requests linux-tools-common Fedora: - sudo dnf install python python-requests + sudo dnf install python python-configparser python-requests linux-tools-common The tools can most easily be installed via git clone and make install @@ -91,8 +97,8 @@ (kernel/pre-3.15/enable_trace_events_suspend_resume.patch) (kernel/pre-3.15/enable_trace_events_device_pm_callback.patch) - If you're using a kernel older than 3.15.0, the following - additional kernel parameters are required: + If you're using bootgraph, or sleepgraph with a kernel older than 3.15.0, + the following additional kernel parameters are required: (e.g. in file /etc/default/grub) GRUB_CMDLINE_LINUX_DEFAULT="... initcall_debug log_buf_len=32M ..." @@ -190,6 +196,104 @@ _______________ %> sudo ./sleepgraph.py -config config/suspend-proc.cfg +------------------------------------------------------------------ +| ENDURANCE TESTING | +------------------------------------------------------------------ + + The best way to gauge the health of a system is to run a series of + suspend/resumes over an extended period and analyze the behavior. This can be + accomplished with sleepgraph's -multi argument. You specify two numbers: the + number of tests to run OR the duration in days, hours, or minutes, and the + delay in seconds between them. For instance, -multi 20 5: execute 20 tests with + a 5 second delay between each, or -multi 24h 0: execute tests over a 24 hour + period with no delay between tests. You can include any other options you like + to generate the data you want. It's most useful to collect dev mode timelines + as the kprobes don't alter the performance much and you get more insight. + + On completion, the output folder contains a series of folders for the + individual test data and a set of summary pages in the root. The summary.html + file is a tabular list of the tests with relevant info and links. The + summary-issue.html and summary-devices.html files include data taken from + all tests on kernel issues and device performance. The folder looks like this: + + suspend-xN-{date}-{time}: + summary.html + summary-issues.html + summary-devices.html + suspend-{date}-{time} (1) + suspend-{date}-{time} (2) + ... + + These are the relevant arguments to use for testing: + + -m mode + Mode to initiate for suspend e.g. mem, freeze, standby (default: mem). + + -rtcwake t + Use rtcwake to autoresume after t seconds (default: 15). + + -gzip (optional) + Gzip the trace and dmesg logs to save space. The tool can also read in + gzipped logs for processing. This reduces the multitest folder size. + + -dev (optional) + Add kernel source calls and threads to the timeline (default: disabled). + + -multi n d + Execute n consecutive tests at d seconds intervals. The outputs will be + created in a new subdirectory: suspend-xN-{date}-{time}. When the multitest + run is done, the -summary command is called automatically to create summary + html files for all the data (unless you use -skiphtml). -skiphtml will + speed up the testing by not creating timelines or summary html files. You + can then run the tool again at a later time with -summary and -genhtml to + create the timelines. + + -skiphtml (optional) + Run the test and capture the trace logs, but skip the timeline and summary + html generation. This can greatly speed up overall testing. You can then + copy the data to a faster host machine and run -summary -genhtml to + generate the timelines and summary. + + These are the relevant commands to use after testing is complete: + + -summary indir + Generate or regenerate the summary for a -multi test run. Creates three + files: summary.html, summary-issues.html, and summary-devices.html in the + current folder. summary.html is a table of tests with relevant info sorted + by kernel/host/mode, and links to the test html files. summary-issues.html + is a list of kernel issues found in dmesg from all the tests. + summary-devices.html is a list of devices and times from all the tests. + + -genhtml + Used with -summary to regenerate any missing html timelines from their + dmesg and ftrace logs. This will require a significant amount of time if + there are thousands of tests. + +Usage Examples +_______________ + + A multitest is initiated like this: + + %> sudo ./sleepgraph.py -m mem -rtcwake 10 -dev -gzip -multi 2000 0 + + or you can skip timeline generation in order to speed things up + + %> sudo ./sleepgraph.py -m mem -rtcwake 10 -dev -gzip -multi 2000 0 -skiphtml + + The tool will produce an output folder with all the test subfolders inside. + Each test subfolder contains the dmesg/ftrace logs and/or the html timeline + depending on whether you used the -skiphtml option. The root folder contains + the summary.html files. + + The summary for an existing multitest is generated like this: + + %> cd suspend-x2000-{date}-{time} + %> sleepgraph.py -summary . + + or if you need to generate the html timelines you can use -genhtml + + %> cd suspend-xN-{date}-{time} + %> sleepgraph.py -summary . -genhtml ------------------------------------------------------------------ | CONFIGURATION FILES | diff --git a/tools/power/pm-graph/bootgraph.py b/tools/power/pm-graph/bootgraph.py index d3b99a1e92d6..f96f50e0c336 100755 --- a/tools/power/pm-graph/bootgraph.py +++ b/tools/power/pm-graph/bootgraph.py @@ -1,4 +1,4 @@ -#!/usr/bin/python +#!/usr/bin/env python3 # SPDX-License-Identifier: GPL-2.0-only # # Tool for analyzing boot timing @@ -69,22 +69,24 @@ class SystemValues(aslib.SystemValues): bootloader = 'grub' blexec = [] def __init__(self): - self.hostname = platform.node() + self.kernel, self.hostname = 'unknown', platform.node() self.testtime = datetime.now().strftime('%Y-%m-%d_%H:%M:%S') if os.path.exists('/proc/version'): fp = open('/proc/version', 'r') - val = fp.read().strip() + self.kernel = self.kernelVersion(fp.read().strip()) fp.close() - self.kernel = self.kernelVersion(val) - else: - self.kernel = 'unknown' self.testdir = datetime.now().strftime('boot-%y%m%d-%H%M%S') def kernelVersion(self, msg): - return msg.split()[2] + m = re.match('^[Ll]inux *[Vv]ersion *(?P<v>\S*) .*', msg) + if m: + return m.group('v') + return 'unknown' def checkFtraceKernelVersion(self): - val = tuple(map(int, self.kernel.split('-')[0].split('.'))) - if val >= (4, 10, 0): - return True + m = re.match('^(?P<x>[0-9]*)\.(?P<y>[0-9]*)\.(?P<z>[0-9]*).*', self.kernel) + if m: + val = tuple(map(int, m.groups())) + if val >= (4, 10, 0): + return True return False def kernelParams(self): cmdline = 'initcall_debug log_buf_len=32M' diff --git a/tools/power/pm-graph/sleepgraph.8 b/tools/power/pm-graph/sleepgraph.8 index 43aee64316df..643271b6fc6f 100644 --- a/tools/power/pm-graph/sleepgraph.8 +++ b/tools/power/pm-graph/sleepgraph.8 @@ -74,8 +74,13 @@ after the test is complete. Switch the display to the requested mode for the test using the xset command. This helps maintain the consistency of test data for better comparison. .TP -\fB-skiphtml\fR -Run the test and capture the trace logs, but skip the timeline generation. +\fB-wifi\fR +If a wifi connection is available, check that it reconnects after resume. Include +the reconnect time in the total resume time calculation and treat wifi timeouts +as resume failures. +.TP +\fB-wifitrace\fR +Trace through the wifi reconnect time and include it in the timeline. .SS "advanced" .TP @@ -117,8 +122,24 @@ Include \fIt\fR ms delay before 1st suspend (default: 0 ms). Include \fIt\fR ms delay after last resume (default: 0 ms). .TP \fB-multi \fIn d\fR -Execute \fIn\fR consecutive tests at \fId\fR seconds intervals. The outputs will -be created in a new subdirectory with a summary page: suspend-xN-{date}-{time}. +Used for endurance testing. If \fIn\fR is entirely numeric, it's treated as a count: +Execute \fIn\fR consecutive tests at \fId\fR second intervals. +If \fIn\fR is an integer followed by a "d", "h", or "m", it's treated as a duration: +Execute tests continuously over \fIn\fR days, hours, or minutes at \fId\fR second intervals. +The outputs will be created in a new subdirectory, for count: suspend-{date}-{time}-xN, +for duration: suspend-{date}-{time}-Nm. When the multitest run is done, the \fI-summary\fR +command is called automatically to create summary html files for all the data (unless you +use \fI-skiphtml\fR). \fI-skiphtml\fR will speed up the testing by not creating timelines +or summary html files. You can then run the tool again at a later time with \fI-summary\fR +and \fI-genhtml\fR to create the timelines. +.TP +\fB-maxfail \fIn\fR +Abort a -multi run after \fIn\fR consecutive fails. 0 means never abort (default = 0). +.TP +\fB-skiphtml\fR +Run the test and capture the trace logs, but skip the timeline generation. +You can generate the html timelines later with \fI-dmesg\fR & \fI-ftrace\fR, or +by running \fI-summary\fR and \fI-genhtml\fR. .SS "ftrace debug" .TP @@ -173,11 +194,20 @@ Set trace buffer size to N kilo-bytes (default: all of free memory up to 3GB) .SH COMMANDS .TP \fB-summary \fIindir\fR -Create a summary page of all tests in \fIindir\fR. Creates summary.html -in the current folder. The output page is a table of tests with -suspend and resume values sorted by suspend mode, host, and kernel. -Includes test averages by mode and links to the test html files. -Use -genhtml to include tests with missing html. +Create a set of summary pages for all tests in \fIindir\fR recursively. +Creates summary.html, summary-issues.html, and summary-devices.html in the current folder. +summary.html is a table of tests with relevant info sorted by kernel/host/mode, +and links to the test html files. It identifies the minimum, maximum, and median +suspend and resume times for you with highlights and links in the header. +summary-issues.html is a list of kernel issues found in dmesg from all the tests. +summary-devices.html is a list of devices and times from all the tests. + +Use \fI-genhtml\fR to regenerate any tests with missing html. +.TP +\fB-genhtml\fR +Used with \fI-summary\fR to regenerate any missing html timelines from their +dmesg and ftrace logs. This will require a significant amount of time if there +are thousands of tests. .TP \fB-modes\fR List available suspend modes. @@ -189,10 +219,7 @@ with any options you intend to use to see if they will work. \fB-fpdt\fR Print out the contents of the ACPI Firmware Performance Data Table. .TP -\fB-battery\fR -Print out battery status and current charge. -.TP -\fB-wifi\fR +\fB-wificheck\fR Print out wifi status and connection details. .TP \fB-xon/-xoff/-xstandby/-xsuspend\fR @@ -208,6 +235,9 @@ Print out system info extracted from BIOS. Reads /dev/mem directly instead of go \fB-devinfo\fR Print out the pm settings of all devices which support runtime suspend. .TP +\fB-cmdinfo\fR +Print out all the platform data collected from the system that makes it into the logs. +.TP \fB-flist\fR Print the list of ftrace functions currently being captured. Functions that are not available as symbols in the current kernel are shown in red. @@ -272,14 +302,20 @@ Run two suspends back to back, include a 500ms delay before, after, and in betwe .IP \f(CW$ sudo sleepgraph -m mem -rtcwake 15 -x2 -predelay 500 -x2delay 500 -postdelay 500\fR .PP +Execute a suspend using a custom command. +.IP +\f(CW$ sudo sleepgraph -cmd "echo mem > /sys/power/state" -rtcwake 15\fR +.PP + +.SS "endurance testing using -multi" +.PP Do a batch run of 10 freezes with 30 seconds delay between runs. .IP \f(CW$ sudo sleepgraph -m freeze -rtcwake 15 -multi 10 30\fR .PP -Execute a suspend using a custom command. +Do a batch run of freezes for 24 hours. .IP -\f(CW$ sudo sleepgraph -cmd "echo mem > /sys/power/state" -rtcwake 15\fR -.PP +\f(CW$ sudo sleepgraph -m freeze -rtcwake 15 -multi 24h 0\fR .SS "adding callgraph data" Add device callgraphs. Limit the trace depth and only show callgraphs 10ms or larger. diff --git a/tools/power/pm-graph/sleepgraph.py b/tools/power/pm-graph/sleepgraph.py index f7d1c1f62f86..cfe343306e08 100755 --- a/tools/power/pm-graph/sleepgraph.py +++ b/tools/power/pm-graph/sleepgraph.py @@ -1,4 +1,4 @@ -#!/usr/bin/python +#!/usr/bin/env python3 # SPDX-License-Identifier: GPL-2.0-only # # Tool for analyzing suspend/resume timing @@ -58,7 +58,7 @@ import re import platform import signal import codecs -from datetime import datetime +from datetime import datetime, timedelta import struct import configparser import gzip @@ -66,8 +66,13 @@ from threading import Thread from subprocess import call, Popen, PIPE import base64 +debugtiming = False +mystarttime = time.time() def pprint(msg): - print(msg) + if debugtiming: + print('[%09.3f] %s' % (time.time()-mystarttime, msg)) + else: + print(msg) sys.stdout.flush() def ascii(text): @@ -81,23 +86,28 @@ def ascii(text): # store system values and test parameters class SystemValues: title = 'SleepGraph' - version = '5.5' + version = '5.10' ansi = False rs = 0 display = '' gzip = False sync = False + wifi = False + netfix = False verbose = False testlog = True dmesglog = True ftracelog = False + acpidebug = True tstat = True - mindevlen = 0.0 + wifitrace = False + mindevlen = 0.0001 mincglen = 0.0 cgphase = '' cgtest = -1 cgskip = '' - multitest = {'run': False, 'count': 0, 'delay': 0} + maxfail = 0 + multitest = {'run': False, 'count': 1000000, 'delay': 0} max_graph_depth = 0 callloopmaxgap = 0.0001 callloopmaxlen = 0.005 @@ -105,6 +115,7 @@ class SystemValues: cpucount = 0 memtotal = 204800 memfree = 204800 + osversion = '' srgap = 0 cgexp = False testdir = '' @@ -113,6 +124,9 @@ class SystemValues: fpdtpath = '/sys/firmware/acpi/tables/FPDT' epath = '/sys/kernel/debug/tracing/events/power/' pmdpath = '/sys/power/pm_debug_messages' + s0ixpath = '/sys/module/intel_pmc_core/parameters/warn_on_s0ix_failures' + s0ixres = '/sys/devices/system/cpu/cpuidle/low_power_idle_system_residency_us' + acpipath='/sys/module/acpi/parameters/debug_level' traceevents = [ 'suspend_resume', 'wakeup_source_activate', @@ -148,10 +162,11 @@ class SystemValues: x2delay = 0 skiphtml = False usecallgraph = False - ftopfunc = 'suspend_devices_and_enter' + ftopfunc = 'pm_suspend' ftop = False usetraceevents = False usetracemarkers = True + useftrace = True usekprobes = True usedevsrc = False useprocmon = False @@ -160,11 +175,14 @@ class SystemValues: devdump = False mixedphaseheight = True devprops = dict() + cfgdef = dict() platinfo = [] predelay = 0 postdelay = 0 - pmdebug = '' + tmstart = 'SUSPEND START %Y%m%d-%H:%M:%S.%f' + tmend = 'RESUME COMPLETE %Y%m%d-%H:%M:%S.%f' tracefuncs = { + 'async_synchronize_full': {}, 'sys_sync': {}, 'ksys_sync': {}, '__pm_notifier_call_chain': {}, @@ -183,9 +201,11 @@ class SystemValues: 'acpi_s2idle_sync': {}, 'acpi_s2idle_begin': {}, 'acpi_s2idle_prepare': {}, + 'acpi_s2idle_prepare_late': {}, 'acpi_s2idle_wake': {}, 'acpi_s2idle_wakeup': {}, 'acpi_s2idle_restore': {}, + 'acpi_s2idle_restore_early': {}, 'hibernate_preallocate_memory': {}, 'create_basic_memory_bitmaps': {}, 'swsusp_write': {}, @@ -270,12 +290,30 @@ class SystemValues: 'intel_opregion_init': {}, 'intel_fbdev_set_suspend': {}, } + infocmds = [ + [0, 'sysinfo', 'uname', '-a'], + [0, 'cpuinfo', 'head', '-7', '/proc/cpuinfo'], + [0, 'kparams', 'cat', '/proc/cmdline'], + [0, 'mcelog', 'mcelog'], + [0, 'pcidevices', 'lspci', '-tv'], + [0, 'usbdevices', 'lsusb', '-tv'], + [0, 'acpidevices', 'sh', '-c', 'ls -l /sys/bus/acpi/devices/*/physical_node'], + [0, 's0ix_require', 'cat', '/sys/kernel/debug/pmc_core/substate_requirements'], + [0, 's0ix_debug', 'cat', '/sys/kernel/debug/pmc_core/slp_s0_debug_status'], + [1, 's0ix_residency', 'cat', '/sys/kernel/debug/pmc_core/slp_s0_residency_usec'], + [1, 'interrupts', 'cat', '/proc/interrupts'], + [1, 'wakeups', 'cat', '/sys/kernel/debug/wakeup_sources'], + [2, 'gpecounts', 'sh', '-c', 'grep -v invalid /sys/firmware/acpi/interrupts/*'], + [2, 'suspendstats', 'sh', '-c', 'grep -v invalid /sys/power/suspend_stats/*'], + [2, 'cpuidle', 'sh', '-c', 'grep -v invalid /sys/devices/system/cpu/cpu*/cpuidle/state*/s2idle/*'], + [2, 'battery', 'sh', '-c', 'grep -v invalid /sys/class/power_supply/*/*'], + [2, 'thermal', 'sh', '-c', 'grep . /sys/class/thermal/thermal_zone*/temp'], + ] cgblacklist = [] kprobes = dict() timeformat = '%.3f' cmdline = '%s %s' % \ (os.path.basename(sys.argv[0]), ' '.join(sys.argv[1:])) - kparams = '' sudouser = '' def __init__(self): self.archargs = 'args_'+platform.machine() @@ -295,6 +333,9 @@ class SystemValues: if os.getuid() == 0 and 'SUDO_USER' in os.environ and \ os.environ['SUDO_USER']: self.sudouser = os.environ['SUDO_USER'] + def resetlog(self): + self.logmsg = '' + self.platinfo = [] def vprint(self, msg): self.logmsg += msg+'\n' if self.verbose or msg.startswith('WARNING:'): @@ -304,11 +345,11 @@ class SystemValues: return signame = self.signames[signum] if signum in self.signames else 'UNKNOWN' msg = 'Signal %s caused a tool exit, line %d' % (signame, frame.f_lineno) - sysvals.outputResult({'error':msg}) + self.outputResult({'error':msg}) sys.exit(3) def signalHandlerInit(self): capture = ['BUS', 'SYS', 'XCPU', 'XFSZ', 'PWR', 'HUP', 'INT', 'QUIT', - 'ILL', 'ABRT', 'FPE', 'SEGV', 'TERM', 'TSTP'] + 'ILL', 'ABRT', 'FPE', 'SEGV', 'TERM'] self.signames = dict() for i in capture: s = 'SIG'+i @@ -336,6 +377,19 @@ class SystemValues: self.outputResult({'error':msg}) sys.exit(1) return False + def usable(self, file, ishtml=False): + if not os.path.exists(file) or os.path.getsize(file) < 1: + return False + if ishtml: + try: + fp = open(file, 'r') + res = fp.read(1000) + fp.close() + except: + return False + if '<html>' not in res: + return False + return True def getExec(self, cmd): try: fp = Popen(['which', cmd], stdout=PIPE, stderr=PIPE).stdout @@ -389,18 +443,16 @@ class SystemValues: r = info['bios-release-date'] if 'bios-release-date' in info else '' self.sysstamp = '# sysinfo | man:%s | plat:%s | cpu:%s | bios:%s | biosdate:%s | numcpu:%d | memsz:%d | memfr:%d' % \ (m, p, c, b, r, self.cpucount, self.memtotal, self.memfree) - try: - kcmd = open('/proc/cmdline', 'r').read().strip() - except: - kcmd = '' - if kcmd: - self.sysstamp += '\n# kparams | %s' % kcmd + if self.osversion: + self.sysstamp += ' | os:%s' % self.osversion def printSystemInfo(self, fatal=False): self.rootCheck(True) out = dmidecode(self.mempath, fatal) if len(out) < 1: return fmt = '%-24s: %s' + if self.osversion: + print(fmt % ('os-version', self.osversion)) for name in sorted(out): print(fmt % (name, out[name])) print(fmt % ('cpucount', ('%d' % self.cpucount))) @@ -408,20 +460,25 @@ class SystemValues: print(fmt % ('memfree', ('%d kB' % self.memfree))) def cpuInfo(self): self.cpucount = 0 - fp = open('/proc/cpuinfo', 'r') - for line in fp: - if re.match('^processor[ \t]*:[ \t]*[0-9]*', line): - self.cpucount += 1 - fp.close() - fp = open('/proc/meminfo', 'r') - for line in fp: - m = re.match('^MemTotal:[ \t]*(?P<sz>[0-9]*) *kB', line) - if m: - self.memtotal = int(m.group('sz')) - m = re.match('^MemFree:[ \t]*(?P<sz>[0-9]*) *kB', line) - if m: - self.memfree = int(m.group('sz')) - fp.close() + if os.path.exists('/proc/cpuinfo'): + with open('/proc/cpuinfo', 'r') as fp: + for line in fp: + if re.match('^processor[ \t]*:[ \t]*[0-9]*', line): + self.cpucount += 1 + if os.path.exists('/proc/meminfo'): + with open('/proc/meminfo', 'r') as fp: + for line in fp: + m = re.match('^MemTotal:[ \t]*(?P<sz>[0-9]*) *kB', line) + if m: + self.memtotal = int(m.group('sz')) + m = re.match('^MemFree:[ \t]*(?P<sz>[0-9]*) *kB', line) + if m: + self.memfree = int(m.group('sz')) + if os.path.exists('/etc/os-release'): + with open('/etc/os-release', 'r') as fp: + for line in fp: + if line.startswith('PRETTY_NAME='): + self.osversion = line[12:].strip().replace('"', '') def initTestOutput(self, name): self.prefix = self.hostname v = open('/proc/version', 'r').read().strip() @@ -441,6 +498,7 @@ class SystemValues: self.testdir+'/'+self.prefix+'_'+self.suspendmode+'.html' if not os.path.isdir(self.testdir): os.makedirs(self.testdir) + self.sudoUserchown(self.testdir) def getValueList(self, value): out = [] for i in value.split(','): @@ -473,9 +531,9 @@ class SystemValues: call('echo 0 > %s/wakealarm' % self.rtcpath, shell=True) def initdmesg(self): # get the latest time stamp from the dmesg log - fp = Popen('dmesg', stdout=PIPE).stdout + lines = Popen('dmesg', stdout=PIPE).stdout.readlines() ktime = '0' - for line in fp: + for line in reversed(lines): line = ascii(line).replace('\r\n', '') idx = line.find('[') if idx > 1: @@ -483,10 +541,10 @@ class SystemValues: m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) if(m): ktime = m.group('ktime') - fp.close() + break self.dmesgstart = float(ktime) def getdmesg(self, testdata): - op = self.writeDatafileHeader(sysvals.dmesgfile, testdata) + op = self.writeDatafileHeader(self.dmesgfile, testdata) # store all new dmesg lines since initdmesg was called fp = Popen('dmesg', stdout=PIPE).stdout for line in fp: @@ -679,6 +737,8 @@ class SystemValues: return False return True def fsetVal(self, val, path): + if not self.useftrace: + return False return self.setVal(val, self.tpath+path) def getVal(self, file): res = '' @@ -692,14 +752,14 @@ class SystemValues: pass return res def fgetVal(self, path): + if not self.useftrace: + return '' return self.getVal(self.tpath+path) def cleanupFtrace(self): - if(self.usecallgraph or self.usetraceevents or self.usedevsrc): + if self.useftrace: self.fsetVal('0', 'events/kprobes/enable') self.fsetVal('', 'kprobe_events') self.fsetVal('1024', 'buffer_size_kb') - if self.pmdebug: - self.setVal(self.pmdebug, self.pmdpath) def setupAllKprobes(self): for name in self.tracefuncs: self.defaultKprobe(name, self.tracefuncs[name]) @@ -716,17 +776,15 @@ class SystemValues: if name == f: return True return False - def initFtrace(self): - self.printSystemInfo(False) - pprint('INITIALIZING FTRACE...') + def initFtrace(self, quiet=False): + if not self.useftrace: + return + if not quiet: + sysvals.printSystemInfo(False) + pprint('INITIALIZING FTRACE') # turn trace off self.fsetVal('0', 'tracing_on') self.cleanupFtrace() - # pm debug messages - pv = self.getVal(self.pmdpath) - if pv != '1': - self.setVal('1', self.pmdpath) - self.pmdebug = pv # set the trace clock to global self.fsetVal('global', 'trace_clock') self.fsetVal('nop', 'current_tracer') @@ -746,12 +804,16 @@ class SystemValues: if tgtsize < 65536: tgtsize = int(self.fgetVal('buffer_size_kb')) * cpus break - pprint('Setting trace buffers to %d kB (%d kB per cpu)' % (tgtsize, tgtsize/cpus)) + self.vprint('Setting trace buffers to %d kB (%d kB per cpu)' % (tgtsize, tgtsize/cpus)) # initialize the callgraph trace if(self.usecallgraph): # set trace type self.fsetVal('function_graph', 'current_tracer') self.fsetVal('', 'set_ftrace_filter') + # temporary hack to fix https://bugzilla.kernel.org/show_bug.cgi?id=212761 + fp = open(self.tpath+'set_ftrace_notrace', 'w') + fp.write('native_queued_spin_lock_slowpath\ndev_driver_string') + fp.close() # set trace format options self.fsetVal('print-parent', 'trace_options') self.fsetVal('funcgraph-abstime', 'trace_options') @@ -782,7 +844,8 @@ class SystemValues: if self.usedevsrc: for name in self.dev_tracefuncs: self.defaultKprobe(name, self.dev_tracefuncs[name]) - pprint('INITIALIZING KPROBES...') + if not quiet: + pprint('INITIALIZING KPROBES') self.addKprobes(self.verbose) if(self.usetraceevents): # turn trace events on @@ -827,21 +890,12 @@ class SystemValues: fw = test['fw'] if(fw): fp.write('# fwsuspend %u fwresume %u\n' % (fw[0], fw[1])) - if 'mcelog' in test: - fp.write('# mcelog %s\n' % test['mcelog']) if 'turbo' in test: fp.write('# turbostat %s\n' % test['turbo']) - if 'bat' in test: - (a1, c1), (a2, c2) = test['bat'] - fp.write('# battery %s %d %s %d\n' % (a1, c1, a2, c2)) if 'wifi' in test: - wstr = [] - for wifi in test['wifi']: - tmp = [] - for key in sorted(wifi): - tmp.append('%s:%s' % (key, wifi[key])) - wstr.append('|'.join(tmp)) - fp.write('# wifi %s\n' % (','.join(wstr))) + fp.write('# wifi %s\n' % test['wifi']) + if 'netfix' in test: + fp.write('# netfix %s\n' % test['netfix']) if test['error'] or len(testdata) > 1: fp.write('# enter_sleep_error %s\n' % test['error']) return fp @@ -861,6 +915,8 @@ class SystemValues: fp.write('error%s: %s\n' % (n, testdata['error'])) else: fp.write('result%s: pass\n' % n) + if 'mode' in testdata: + fp.write('mode%s: %s\n' % (n, testdata['mode'])) for v in ['suspend', 'resume', 'boot', 'lastinit']: if v in testdata: fp.write('%s%s: %.3f\n' % (v, n, testdata[v])) @@ -892,6 +948,16 @@ class SystemValues: if isgz: return gzip.open(filename, mode+'t') return open(filename, mode) + def putlog(self, filename, text): + with self.openlog(filename, 'a') as fp: + fp.write(text) + fp.close() + def dlog(self, text): + if not self.dmesgfile: + return + self.putlog(self.dmesgfile, '# %s\n' % text) + def flog(self, text): + self.putlog(self.ftracefile, text) def b64unzip(self, data): try: out = codecs.decode(base64.b64decode(data), 'zlib').decode() @@ -901,23 +967,7 @@ class SystemValues: def b64zip(self, data): out = base64.b64encode(codecs.encode(data.encode(), 'zlib')).decode() return out - def mcelog(self, clear=False): - cmd = self.getExec('mcelog') - if not cmd: - return '' - if clear: - call(cmd+' > /dev/null 2>&1', shell=True) - return '' - try: - fp = Popen([cmd], stdout=PIPE, stderr=PIPE).stdout - out = ascii(fp.read()).strip() - fp.close() - except: - return '' - if not out: - return '' - return self.b64zip(out) - def platforminfo(self): + def platforminfo(self, cmdafter): # add platform info on to a completed ftrace file if not os.path.exists(self.ftracefile): return False @@ -932,10 +982,7 @@ class SystemValues: tp = TestProps() tf = self.openlog(self.ftracefile, 'r') for line in tf: - # determine the trace data type (required for further parsing) - m = re.match(tp.tracertypefmt, line) - if(m): - tp.setTracerType(m.group('t')) + if tp.stampInfo(line, self): continue # parse only valid lines, if this is not one move on m = re.match(tp.ftrace_line_fmt, line) @@ -961,34 +1008,31 @@ class SystemValues: dirname = props[dev].syspath if not dirname or not os.path.exists(dirname): continue - with open(dirname+'/power/async') as fp: - text = fp.read() - props[dev].isasync = False - if 'enabled' in text: + props[dev].isasync = False + if os.path.exists(dirname+'/power/async'): + fp = open(dirname+'/power/async') + if 'enabled' in fp.read(): props[dev].isasync = True + fp.close() fields = os.listdir(dirname) - if 'product' in fields: - with open(dirname+'/product', 'rb') as fp: - props[dev].altname = ascii(fp.read()) - elif 'name' in fields: - with open(dirname+'/name', 'rb') as fp: - props[dev].altname = ascii(fp.read()) - elif 'model' in fields: - with open(dirname+'/model', 'rb') as fp: - props[dev].altname = ascii(fp.read()) - elif 'description' in fields: - with open(dirname+'/description', 'rb') as fp: - props[dev].altname = ascii(fp.read()) - elif 'id' in fields: - with open(dirname+'/id', 'rb') as fp: - props[dev].altname = ascii(fp.read()) - elif 'idVendor' in fields and 'idProduct' in fields: - idv, idp = '', '' - with open(dirname+'/idVendor', 'rb') as fp: - idv = ascii(fp.read()).strip() - with open(dirname+'/idProduct', 'rb') as fp: - idp = ascii(fp.read()).strip() - props[dev].altname = '%s:%s' % (idv, idp) + for file in ['product', 'name', 'model', 'description', 'id', 'idVendor']: + if file not in fields: + continue + try: + with open(os.path.join(dirname, file), 'rb') as fp: + props[dev].altname = ascii(fp.read()) + except: + continue + if file == 'idVendor': + idv, idp = props[dev].altname.strip(), '' + try: + with open(os.path.join(dirname, 'idProduct'), 'rb') as fp: + idp = ascii(fp.read()).strip() + except: + props[dev].altname = '' + break + props[dev].altname = '%s:%s' % (idv, idp) + break if props[dev].altname: out = props[dev].altname.strip().replace('\n', ' ')\ .replace(',', ' ').replace(';', ' ') @@ -1001,31 +1045,107 @@ class SystemValues: footer += '# platform-devinfo: %s\n' % self.b64zip(out) # add a line for each of these commands with their outputs - cmds = [ - ['pcidevices', 'lspci', '-tv'], - ['interrupts', 'cat', '/proc/interrupts'], - ['gpecounts', 'sh', '-c', 'grep -v invalid /sys/firmware/acpi/interrupts/gpe*'], - ] - for cargs in cmds: - name = cargs[0] - cmdline = ' '.join(cargs[1:]) - cmdpath = self.getExec(cargs[1]) - if not cmdpath: + for name, cmdline, info in cmdafter: + footer += '# platform-%s: %s | %s\n' % (name, cmdline, self.b64zip(info)) + self.flog(footer) + return True + def commonPrefix(self, list): + if len(list) < 2: + return '' + prefix = list[0] + for s in list[1:]: + while s[:len(prefix)] != prefix and prefix: + prefix = prefix[:len(prefix)-1] + if not prefix: + break + if '/' in prefix and prefix[-1] != '/': + prefix = prefix[0:prefix.rfind('/')+1] + return prefix + def dictify(self, text, format): + out = dict() + header = True if format == 1 else False + delim = ' ' if format == 1 else ':' + for line in text.split('\n'): + if header: + header, out['@'] = False, line continue - cmd = [cmdpath] + cargs[2:] + line = line.strip() + if delim in line: + data = line.split(delim, 1) + num = re.search(r'[\d]+', data[1]) + if format == 2 and num: + out[data[0].strip()] = num.group() + else: + out[data[0].strip()] = data[1] + return out + def cmdinfo(self, begin, debug=False): + out = [] + if begin: + self.cmd1 = dict() + for cargs in self.infocmds: + delta, name = cargs[0], cargs[1] + cmdline, cmdpath = ' '.join(cargs[2:]), self.getExec(cargs[2]) + if not cmdpath or (begin and not delta): + continue + self.dlog('[%s]' % cmdline) try: - fp = Popen(cmd, stdout=PIPE, stderr=PIPE).stdout + fp = Popen([cmdpath]+cargs[3:], stdout=PIPE, stderr=PIPE).stdout info = ascii(fp.read()).strip() fp.close() except: continue - if not info: - continue - footer += '# platform-%s: %s | %s\n' % (name, cmdline, self.b64zip(info)) - - with self.openlog(self.ftracefile, 'a') as fp: - fp.write(footer) - return True + if not debug and begin: + self.cmd1[name] = self.dictify(info, delta) + elif not debug and delta and name in self.cmd1: + before, after = self.cmd1[name], self.dictify(info, delta) + dinfo = ('\t%s\n' % before['@']) if '@' in before and len(before) > 1 else '' + prefix = self.commonPrefix(list(before.keys())) + for key in sorted(before): + if key in after and before[key] != after[key]: + title = key.replace(prefix, '') + if delta == 2: + dinfo += '\t%s : %s -> %s\n' % \ + (title, before[key].strip(), after[key].strip()) + else: + dinfo += '%10s (start) : %s\n%10s (after) : %s\n' % \ + (title, before[key], title, after[key]) + dinfo = '\tnothing changed' if not dinfo else dinfo.rstrip() + out.append((name, cmdline, dinfo)) + else: + out.append((name, cmdline, '\tnothing' if not info else info)) + return out + def testVal(self, file, fmt='basic', value=''): + if file == 'restoreall': + for f in self.cfgdef: + if os.path.exists(f): + fp = open(f, 'w') + fp.write(self.cfgdef[f]) + fp.close() + self.cfgdef = dict() + elif value and os.path.exists(file): + fp = open(file, 'r+') + if fmt == 'radio': + m = re.match('.*\[(?P<v>.*)\].*', fp.read()) + if m: + self.cfgdef[file] = m.group('v') + elif fmt == 'acpi': + line = fp.read().strip().split('\n')[-1] + m = re.match('.* (?P<v>[0-9A-Fx]*) .*', line) + if m: + self.cfgdef[file] = m.group('v') + else: + self.cfgdef[file] = fp.read().strip() + fp.write(value) + fp.close() + def s0ixSupport(self): + if not os.path.exists(self.s0ixres) or not os.path.exists(self.mempowerfile): + return False + fp = open(sysvals.mempowerfile, 'r') + data = fp.read().strip() + fp.close() + if '[s2idle]' in data: + return True + return False def haveTurbostat(self): if not self.tstat: return False @@ -1035,11 +1155,11 @@ class SystemValues: fp = Popen([cmd, '-v'], stdout=PIPE, stderr=PIPE).stderr out = ascii(fp.read()).strip() fp.close() - if re.match('turbostat version [0-9\.]* .*', out): - sysvals.vprint(out) + if re.match('turbostat version .*', out): + self.vprint(out) return True return False - def turbostat(self): + def turbostat(self, s0ixready): cmd = self.getExec('turbostat') rawout = keyline = valline = '' fullcmd = '%s -q -S echo freeze > %s' % (cmd, self.powerfile) @@ -1056,42 +1176,58 @@ class SystemValues: fp.close() if not keyline or not valline or len(keyline) != len(valline): errmsg = 'unrecognized turbostat output:\n'+rawout.strip() - sysvals.vprint(errmsg) - if not sysvals.verbose: + self.vprint(errmsg) + if not self.verbose: pprint(errmsg) return '' - if sysvals.verbose: + if self.verbose: pprint(rawout.strip()) out = [] for key in keyline: idx = keyline.index(key) val = valline[idx] + if key == 'SYS%LPI' and not s0ixready and re.match('^[0\.]*$', val): + continue out.append('%s=%s' % (key, val)) return '|'.join(out) - def checkWifi(self): - out = dict() - iwcmd, ifcmd = self.getExec('iwconfig'), self.getExec('ifconfig') - if not iwcmd or not ifcmd: - return out - fp = Popen(iwcmd, stdout=PIPE, stderr=PIPE).stdout - for line in fp: - m = re.match('(?P<dev>\S*) .* ESSID:(?P<ess>\S*)', ascii(line)) - if not m: - continue - out['device'] = m.group('dev') - if '"' in m.group('ess'): - out['essid'] = m.group('ess').strip('"') - break + def netfixon(self, net='both'): + cmd = self.getExec('netfix') + if not cmd: + return '' + fp = Popen([cmd, '-s', net, 'on'], stdout=PIPE, stderr=PIPE).stdout + out = ascii(fp.read()).strip() fp.close() - if 'device' in out: - fp = Popen([ifcmd, out['device']], stdout=PIPE, stderr=PIPE).stdout - for line in fp: - m = re.match('.* inet (?P<ip>[0-9\.]*)', ascii(line)) - if m: - out['ip'] = m.group('ip') - break - fp.close() return out + def wifiDetails(self, dev): + try: + info = open('/sys/class/net/%s/device/uevent' % dev, 'r').read().strip() + except: + return dev + vals = [dev] + for prop in info.split('\n'): + if prop.startswith('DRIVER=') or prop.startswith('PCI_ID='): + vals.append(prop.split('=')[-1]) + return ':'.join(vals) + def checkWifi(self, dev=''): + try: + w = open('/proc/net/wireless', 'r').read().strip() + except: + return '' + for line in reversed(w.split('\n')): + m = re.match(' *(?P<dev>.*): (?P<stat>[0-9a-f]*) .*', line) + if not m or (dev and dev != m.group('dev')): + continue + return m.group('dev') + return '' + def pollWifi(self, dev, timeout=10): + start = time.time() + while (time.time() - start) < timeout: + w = self.checkWifi(dev) + if w: + return '%s reconnected %.2f' % \ + (self.wifiDetails(dev), max(0, time.time() - start)) + time.sleep(0.01) + return '%s timeout %d' % (self.wifiDetails(dev), timeout) def errorSummary(self, errinfo, msg): found = False for entry in errinfo: @@ -1113,8 +1249,9 @@ class SystemValues: arr[j] = arr[j]\ .replace('\\', '\\\\').replace(']', '\]').replace('[', '\[')\ .replace('.', '\.').replace('+', '\+').replace('*', '\*')\ - .replace('(', '\(').replace(')', '\)') - mstr = ' '.join(arr) + .replace('(', '\(').replace(')', '\)').replace('}', '\}')\ + .replace('{', '\{') + mstr = ' *'.join(arr) entry = { 'line': msg, 'match': mstr, @@ -1122,15 +1259,118 @@ class SystemValues: 'urls': {self.hostname: [self.htmlfile]} } errinfo.append(entry) + def multistat(self, start, idx, finish): + if 'time' in self.multitest: + id = '%d Duration=%dmin' % (idx+1, self.multitest['time']) + else: + id = '%d/%d' % (idx+1, self.multitest['count']) + t = time.time() + if 'start' not in self.multitest: + self.multitest['start'] = self.multitest['last'] = t + self.multitest['total'] = 0.0 + pprint('TEST (%s) START' % id) + return + dt = t - self.multitest['last'] + if not start: + if idx == 0 and self.multitest['delay'] > 0: + self.multitest['total'] += self.multitest['delay'] + pprint('TEST (%s) COMPLETE -- Duration %.1fs' % (id, dt)) + return + self.multitest['total'] += dt + self.multitest['last'] = t + avg = self.multitest['total'] / idx + if 'time' in self.multitest: + left = finish - datetime.now() + left -= timedelta(microseconds=left.microseconds) + else: + left = timedelta(seconds=((self.multitest['count'] - idx) * int(avg))) + pprint('TEST (%s) START - Avg Duration %.1fs, Time left %s' % \ + (id, avg, str(left))) + def multiinit(self, c, d): + sz, unit = 'count', 'm' + if c.endswith('d') or c.endswith('h') or c.endswith('m'): + sz, unit, c = 'time', c[-1], c[:-1] + self.multitest['run'] = True + self.multitest[sz] = getArgInt('multi: n d (exec count)', c, 1, 1000000, False) + self.multitest['delay'] = getArgInt('multi: n d (delay between tests)', d, 0, 3600, False) + if unit == 'd': + self.multitest[sz] *= 1440 + elif unit == 'h': + self.multitest[sz] *= 60 + def displayControl(self, cmd): + xset, ret = 'timeout 10 xset -d :0.0 {0}', 0 + if self.sudouser: + xset = 'sudo -u %s %s' % (self.sudouser, xset) + if cmd == 'init': + ret = call(xset.format('dpms 0 0 0'), shell=True) + if not ret: + ret = call(xset.format('s off'), shell=True) + elif cmd == 'reset': + ret = call(xset.format('s reset'), shell=True) + elif cmd in ['on', 'off', 'standby', 'suspend']: + b4 = self.displayControl('stat') + ret = call(xset.format('dpms force %s' % cmd), shell=True) + if not ret: + curr = self.displayControl('stat') + self.vprint('Display Switched: %s -> %s' % (b4, curr)) + if curr != cmd: + self.vprint('WARNING: Display failed to change to %s' % cmd) + if ret: + self.vprint('WARNING: Display failed to change to %s with xset' % cmd) + return ret + elif cmd == 'stat': + fp = Popen(xset.format('q').split(' '), stdout=PIPE).stdout + ret = 'unknown' + for line in fp: + m = re.match('[\s]*Monitor is (?P<m>.*)', ascii(line)) + if(m and len(m.group('m')) >= 2): + out = m.group('m').lower() + ret = out[3:] if out[0:2] == 'in' else out + break + fp.close() + return ret + def setRuntimeSuspend(self, before=True): + if before: + # runtime suspend disable or enable + if self.rs > 0: + self.rstgt, self.rsval, self.rsdir = 'on', 'auto', 'enabled' + else: + self.rstgt, self.rsval, self.rsdir = 'auto', 'on', 'disabled' + pprint('CONFIGURING RUNTIME SUSPEND...') + self.rslist = deviceInfo(self.rstgt) + for i in self.rslist: + self.setVal(self.rsval, i) + pprint('runtime suspend %s on all devices (%d changed)' % (self.rsdir, len(self.rslist))) + pprint('waiting 5 seconds...') + time.sleep(5) + else: + # runtime suspend re-enable or re-disable + for i in self.rslist: + self.setVal(self.rstgt, i) + pprint('runtime suspend settings restored on %d devices' % len(self.rslist)) + def start(self, pm): + if self.useftrace: + self.dlog('start ftrace tracing') + self.fsetVal('1', 'tracing_on') + if self.useprocmon: + self.dlog('start the process monitor') + pm.start() + def stop(self, pm): + if self.useftrace: + if self.useprocmon: + self.dlog('stop the process monitor') + pm.stop() + self.dlog('stop ftrace tracing') + self.fsetVal('0', 'tracing_on') sysvals = SystemValues() switchvalues = ['enable', 'disable', 'on', 'off', 'true', 'false', '1', '0'] switchoff = ['disable', 'off', 'false', '0'] suspendmodename = { - 'freeze': 'Freeze (S0)', - 'standby': 'Standby (S1)', - 'mem': 'Suspend (S3)', - 'disk': 'Hibernate (S4)' + 'standby': 'standby (S1)', + 'freeze': 'freeze (S2idle)', + 'mem': 'suspend (S3)', + 'disk': 'hibernate (S4)' } # Class: DevProps @@ -1162,8 +1402,8 @@ class DevProps: if self.xtraclass: return ' '+self.xtraclass if self.isasync: - return ' async_device' - return ' sync_device' + return ' (async)' + return ' (sync)' # Class: DeviceNode # Description: @@ -1210,25 +1450,32 @@ class Data: 'resume_complete': {'order': 9, 'color': '#FFFFCC'}, } errlist = { - 'HWERROR' : '.*\[ *Hardware Error *\].*', - 'FWBUG' : '.*\[ *Firmware Bug *\].*', - 'BUG' : '.*BUG.*', - 'ERROR' : '.*ERROR.*', - 'WARNING' : '.*WARNING.*', - 'IRQ' : '.*genirq: .*', - 'TASKFAIL': '.*Freezing of tasks *.*', - 'ACPI' : '.*ACPI *(?P<b>[A-Za-z]*) *Error[: ].*', - 'DEVFAIL' : '.* failed to (?P<b>[a-z]*) async: .*', - 'DISKFULL': '.*No space left on device.*', - 'USBERR' : '.*usb .*device .*, error [0-9-]*', - 'ATAERR' : ' *ata[0-9\.]*: .*failed.*', - 'MEIERR' : ' *mei.*: .*failed.*', - 'TPMERR' : '(?i) *tpm *tpm[0-9]*: .*error.*', + 'HWERROR' : r'.*\[ *Hardware Error *\].*', + 'FWBUG' : r'.*\[ *Firmware Bug *\].*', + 'BUG' : r'(?i).*\bBUG\b.*', + 'ERROR' : r'(?i).*\bERROR\b.*', + 'WARNING' : r'(?i).*\bWARNING\b.*', + 'FAULT' : r'(?i).*\bFAULT\b.*', + 'FAIL' : r'(?i).*\bFAILED\b.*', + 'INVALID' : r'(?i).*\bINVALID\b.*', + 'CRASH' : r'(?i).*\bCRASHED\b.*', + 'TIMEOUT' : r'(?i).*\bTIMEOUT\b.*', + 'ABORT' : r'(?i).*\bABORT\b.*', + 'IRQ' : r'.*\bgenirq: .*', + 'TASKFAIL': r'.*Freezing of tasks *.*', + 'ACPI' : r'.*\bACPI *(?P<b>[A-Za-z]*) *Error[: ].*', + 'DISKFULL': r'.*\bNo space left on device.*', + 'USBERR' : r'.*usb .*device .*, error [0-9-]*', + 'ATAERR' : r' *ata[0-9\.]*: .*failed.*', + 'MEIERR' : r' *mei.*: .*failed.*', + 'TPMERR' : r'(?i) *tpm *tpm[0-9]*: .*error.*', } def __init__(self, num): idchar = 'abcdefghij' self.start = 0.0 # test start self.end = 0.0 # test end + self.hwstart = 0 # rtc test start + self.hwend = 0 # rtc test end self.tSuspended = 0.0 # low-level suspend start self.tResumed = 0.0 # low-level resume start self.tKernSus = 0.0 # kernel level suspend start @@ -1240,10 +1487,8 @@ class Data: self.stamp = 0 self.outfile = '' self.kerror = False - self.battery = 0 - self.wifi = 0 + self.wifi = dict() self.turbostat = 0 - self.mcelog = 0 self.enterfail = '' self.currphase = '' self.pstl = dict() # process timeline @@ -1273,11 +1518,11 @@ class Data: if self.dmesg[p]['order'] == order: return p return '' - def lastPhase(self): + def lastPhase(self, depth=1): plist = self.sortedPhases() - if len(plist) < 1: + if len(plist) < depth: return '' - return plist[-1] + return plist[-1*depth] def turbostatInfo(self): tp = TestProps() out = {'syslpi':'N/A','pkgpc10':'N/A'} @@ -1297,9 +1542,12 @@ class Data: if len(self.dmesgtext) < 1 and sysvals.dmesgfile: lf = sysvals.openlog(sysvals.dmesgfile, 'r') i = 0 + tp = TestProps() list = [] for line in lf: i += 1 + if tp.stampInfo(line, sysvals): + continue m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) if not m: continue @@ -1308,25 +1556,36 @@ class Data: continue dir = 'suspend' if t < self.tSuspended else 'resume' msg = m.group('msg') + if re.match('capability: warning: .*', msg): + continue for err in self.errlist: if re.match(self.errlist[err], msg): list.append((msg, err, dir, t, i, i)) self.kerror = True break - msglist = [] + tp.msglist = [] for msg, type, dir, t, idx1, idx2 in list: - msglist.append(msg) - sysvals.vprint('kernel %s found in %s at %f' % (type, dir, t)) + tp.msglist.append(msg) self.errorinfo[dir].append((type, t, idx1, idx2)) if self.kerror: sysvals.dmesglog = True if len(self.dmesgtext) < 1 and sysvals.dmesgfile: lf.close() - return msglist - def setStart(self, time): + return tp + def setStart(self, time, msg=''): self.start = time - def setEnd(self, time): + if msg: + try: + self.hwstart = datetime.strptime(msg, sysvals.tmstart) + except: + self.hwstart = 0 + def setEnd(self, time, msg=''): self.end = time + if msg: + try: + self.hwend = datetime.strptime(msg, sysvals.tmend) + except: + self.hwend = 0 def isTraceEventOutsideDeviceCalls(self, pid, time): for phase in self.sortedPhases(): list = self.dmesg[phase]['list'] @@ -1400,19 +1659,20 @@ class Data: ubiquitous = False if kprobename in dtf and 'ub' in dtf[kprobename]: ubiquitous = True - title = cdata+' '+rdata - mstr = '\(.*\) *(?P<args>.*) *\((?P<caller>.*)\+.* arg1=(?P<ret>.*)' - m = re.match(mstr, title) - if m: - c = m.group('caller') - a = m.group('args').strip() - r = m.group('ret') + mc = re.match('\(.*\) *(?P<args>.*)', cdata) + mr = re.match('\((?P<caller>\S*).* arg1=(?P<ret>.*)', rdata) + if mc and mr: + c = mr.group('caller').split('+')[0] + a = mc.group('args').strip() + r = mr.group('ret') if len(r) > 6: r = '' else: r = 'ret=%s ' % r if ubiquitous and c in dtf and 'ub' in dtf[c]: return False + else: + return False color = sysvals.kprobeColor(kprobename) e = DevFunction(displayname, a, c, r, start, end, ubiquitous, proc, pid, color) tgtdev['src'].append(e) @@ -1527,6 +1787,16 @@ class Data: if('src' in d): for e in d['src']: e.time = self.trimTimeVal(e.time, t0, dT, left) + e.end = self.trimTimeVal(e.end, t0, dT, left) + e.length = e.end - e.time + if('cpuexec' in d): + cpuexec = dict() + for e in d['cpuexec']: + c0, cN = e + c0 = self.trimTimeVal(c0, t0, dT, left) + cN = self.trimTimeVal(cN, t0, dT, left) + cpuexec[(c0, cN)] = d['cpuexec'][e] + d['cpuexec'] = cpuexec for dir in ['suspend', 'resume']: list = [] for e in self.errorinfo[dir]: @@ -1541,11 +1811,29 @@ class Data: if 'resume_machine' in phase and 'suspend_machine' in lp: tS, tR = self.dmesg[lp]['end'], self.dmesg[phase]['start'] tL = tR - tS - if tL > 0: - left = True if tR > tZero else False - self.trimTime(tS, tL, left) - self.tLow.append('%.0f'%(tL*1000)) + if tL <= 0: + continue + left = True if tR > tZero else False + self.trimTime(tS, tL, left) + if 'waking' in self.dmesg[lp]: + tCnt = self.dmesg[lp]['waking'][0] + if self.dmesg[lp]['waking'][1] >= 0.001: + tTry = '%.0f' % (round(self.dmesg[lp]['waking'][1] * 1000)) + else: + tTry = '%.3f' % (self.dmesg[lp]['waking'][1] * 1000) + text = '%.0f (%s ms waking %d times)' % (tL * 1000, tTry, tCnt) + else: + text = '%.0f' % (tL * 1000) + self.tLow.append(text) lp = phase + def getMemTime(self): + if not self.hwstart or not self.hwend: + return + stime = (self.tSuspended - self.start) * 1000000 + rtime = (self.end - self.tResumed) * 1000000 + hws = self.hwstart + timedelta(microseconds=stime) + hwr = self.hwend - timedelta(microseconds=rtime) + self.tLow.append('%.0f'%((hwr - hws).total_seconds() * 1000)) def getTimeValues(self): sktime = (self.tSuspended - self.tKernSus) * 1000 rktime = (self.tKernRes - self.tResumed) * 1000 @@ -1672,7 +1960,7 @@ class Data: length = -1.0 if(start >= 0 and end >= 0): length = end - start - if pid == -2: + if pid == -2 or name not in sysvals.tracefuncs.keys(): i = 2 origname = name while(name in list): @@ -1685,6 +1973,15 @@ class Data: if color: list[name]['color'] = color return name + def findDevice(self, phase, name): + list = self.dmesg[phase]['list'] + mydev = '' + for devname in sorted(list): + if name == devname or re.match('^%s\[(?P<num>[0-9]*)\]$' % name, devname): + mydev = devname + if mydev: + return list[mydev] + return False def deviceChildren(self, devname, phase): devlist = [] list = self.dmesg[phase]['list'] @@ -1800,7 +2097,7 @@ class Data: for dev in list: length = (list[dev]['end'] - list[dev]['start']) * 1000 width = widfmt % (((list[dev]['end']-list[dev]['start'])*100)/tTotal) - if width != '0.000000' and length >= mindevlen: + if length >= mindevlen: devlist.append(dev) self.tdevlist[phase] = devlist def addHorizontalDivider(self, devname, devend): @@ -1814,78 +2111,46 @@ class Data: return d def addProcessUsageEvent(self, name, times): # get the start and end times for this process - maxC = 0 - tlast = 0 - start = -1 - end = -1 + cpuexec = dict() + tlast = start = end = -1 for t in sorted(times): - if tlast == 0: + if tlast < 0: tlast = t continue - if name in self.pstl[t]: - if start == -1 or tlast < start: + if name in self.pstl[t] and self.pstl[t][name] > 0: + if start < 0: start = tlast - if end == -1 or t > end: - end = t + end, key = t, (tlast, t) + maxj = (t - tlast) * 1024.0 + cpuexec[key] = min(1.0, float(self.pstl[t][name]) / maxj) tlast = t - if start == -1 or end == -1: - return 0 + if start < 0 or end < 0: + return # add a new action for this process and get the object out = self.newActionGlobal(name, start, end, -3) - if not out: - return 0 - phase, devname = out - dev = self.dmesg[phase]['list'][devname] - # get the cpu exec data - tlast = 0 - clast = 0 - cpuexec = dict() - for t in sorted(times): - if tlast == 0 or t <= start or t > end: - tlast = t - continue - list = self.pstl[t] - c = 0 - if name in list: - c = list[name] - if c > maxC: - maxC = c - if c != clast: - key = (tlast, t) - cpuexec[key] = c - tlast = t - clast = c - dev['cpuexec'] = cpuexec - return maxC + if out: + phase, devname = out + dev = self.dmesg[phase]['list'][devname] + dev['cpuexec'] = cpuexec def createProcessUsageEvents(self): - # get an array of process names - proclist = [] - for t in sorted(self.pstl): - pslist = self.pstl[t] - for ps in sorted(pslist): - if ps not in proclist: - proclist.append(ps) - # get a list of data points for suspend and resume - tsus = [] - tres = [] + # get an array of process names and times + proclist = {'sus': dict(), 'res': dict()} + tdata = {'sus': [], 'res': []} for t in sorted(self.pstl): - if t < self.tSuspended: - tsus.append(t) - else: - tres.append(t) + dir = 'sus' if t < self.tSuspended else 'res' + for ps in sorted(self.pstl[t]): + if ps not in proclist[dir]: + proclist[dir][ps] = 0 + tdata[dir].append(t) # process the events for suspend and resume - if len(proclist) > 0: + if len(proclist['sus']) > 0 or len(proclist['res']) > 0: sysvals.vprint('Process Execution:') - for ps in proclist: - c = self.addProcessUsageEvent(ps, tsus) - if c > 0: - sysvals.vprint('%25s (sus): %d' % (ps, c)) - c = self.addProcessUsageEvent(ps, tres) - if c > 0: - sysvals.vprint('%25s (res): %d' % (ps, c)) - def handleEndMarker(self, time): + for dir in ['sus', 'res']: + for ps in sorted(proclist[dir]): + self.addProcessUsageEvent(ps, tdata[dir]) + def handleEndMarker(self, time, msg=''): dm = self.dmesg - self.setEnd(time) + self.setEnd(time, msg) self.initDevicegroups() # give suspend_prepare an end if needed if 'suspend_prepare' in dm and dm['suspend_prepare']['end'] < 0: @@ -1904,6 +2169,30 @@ class Data: # set resume complete to end at end marker if 'resume_complete' in dm: dm['resume_complete']['end'] = time + def initcall_debug_call(self, line, quick=False): + m = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) .* (?P<f>.*)\: '+\ + 'PM: *calling .* @ (?P<n>.*), parent: (?P<p>.*)', line) + if not m: + m = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) .* (?P<f>.*)\: '+\ + 'calling .* @ (?P<n>.*), parent: (?P<p>.*)', line) + if not m: + m = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) calling '+\ + '(?P<f>.*)\+ @ (?P<n>.*), parent: (?P<p>.*)', line) + if m: + return True if quick else m.group('t', 'f', 'n', 'p') + return False if quick else ('', '', '', '') + def initcall_debug_return(self, line, quick=False): + m = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) .* (?P<f>.*)\: PM: '+\ + '.* returned (?P<r>[0-9]*) after (?P<dt>[0-9]*) usecs', line) + if not m: + m = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) .* (?P<f>.*)\: '+\ + '.* returned (?P<r>[0-9]*) after (?P<dt>[0-9]*) usecs', line) + if not m: + m = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) call '+\ + '(?P<f>.*)\+ returned .* after (?P<dt>.*) usecs', line) + if m: + return True if quick else m.group('t', 'f', 'dt') + return False if quick else ('', '', '') def debugPrint(self): for p in self.sortedPhases(): list = self.dmesg[p]['list'] @@ -2071,7 +2360,7 @@ class FTraceLine: if not self.fevent: return False if sysvals.usetracemarkers: - if(self.name == 'SUSPEND START'): + if(self.name.startswith('SUSPEND START')): return True return False else: @@ -2084,7 +2373,7 @@ class FTraceLine: if not self.fevent: return False if sysvals.usetracemarkers: - if(self.name == 'RESUME COMPLETE'): + if(self.name.startswith('RESUME COMPLETE')): return True return False else: @@ -2444,7 +2733,7 @@ class Timeline: def createHeader(self, sv, stamp): if(not stamp['time']): return - self.html += '<div class="version"><a href="https://01.org/suspendresume">%s v%s</a></div>' \ + self.html += '<div class="version"><a href="https://01.org/pm-graph">%s v%s</a></div>' \ % (sv.title, sv.version) if sv.logmsg and sv.testlog: self.html += '<button id="showtest" class="logbtn btnfmt">log</button>' @@ -2670,36 +2959,36 @@ class TestProps: stampfmt = '# [a-z]*-(?P<m>[0-9]{2})(?P<d>[0-9]{2})(?P<y>[0-9]{2})-'+\ '(?P<H>[0-9]{2})(?P<M>[0-9]{2})(?P<S>[0-9]{2})'+\ ' (?P<host>.*) (?P<mode>.*) (?P<kernel>.*)$' - batteryfmt = '^# battery (?P<a1>\w*) (?P<c1>\d*) (?P<a2>\w*) (?P<c2>\d*)' - wififmt = '^# wifi (?P<w>.*)' + wififmt = '^# wifi *(?P<d>\S*) *(?P<s>\S*) *(?P<t>[0-9\.]+).*' tstatfmt = '^# turbostat (?P<t>\S*)' - mcelogfmt = '^# mcelog (?P<m>\S*)' testerrfmt = '^# enter_sleep_error (?P<e>.*)' sysinfofmt = '^# sysinfo .*' cmdlinefmt = '^# command \| (?P<cmd>.*)' kparamsfmt = '^# kparams \| (?P<kp>.*)' devpropfmt = '# Device Properties: .*' - pinfofmt = '# platform-(?P<val>[a-z,A-Z,0-9]*): (?P<info>.*)' + pinfofmt = '# platform-(?P<val>[a-z,A-Z,0-9,_]*): (?P<info>.*)' tracertypefmt = '# tracer: (?P<t>.*)' firmwarefmt = '# fwsuspend (?P<s>[0-9]*) fwresume (?P<r>[0-9]*)$' procexecfmt = 'ps - (?P<ps>.*)$' + procmultifmt = '@(?P<n>[0-9]*)\|(?P<ps>.*)$' ftrace_line_fmt_fg = \ '^ *(?P<time>[0-9\.]*) *\| *(?P<cpu>[0-9]*)\)'+\ ' *(?P<proc>.*)-(?P<pid>[0-9]*) *\|'+\ '[ +!#\*@$]*(?P<dur>[0-9\.]*) .*\| (?P<msg>.*)' ftrace_line_fmt_nop = \ ' *(?P<proc>.*)-(?P<pid>[0-9]*) *\[(?P<cpu>[0-9]*)\] *'+\ - '(?P<flags>.{4}) *(?P<time>[0-9\.]*): *'+\ + '(?P<flags>\S*) *(?P<time>[0-9\.]*): *'+\ '(?P<msg>.*)' + machinesuspend = 'machine_suspend\[.*' + multiproclist = dict() + multiproctime = 0.0 + multiproccnt = 0 def __init__(self): self.stamp = '' self.sysinfo = '' self.cmdline = '' - self.kparams = '' self.testerror = [] - self.mcelog = [] self.turbostat = [] - self.battery = [] self.wifi = [] self.fwdata = [] self.ftrace_line_fmt = self.ftrace_line_fmt_nop @@ -2714,28 +3003,16 @@ class TestProps: self.ftrace_line_fmt = self.ftrace_line_fmt_nop else: doError('Invalid tracer format: [%s]' % tracer) - def stampInfo(self, line): + def stampInfo(self, line, sv): if re.match(self.stampfmt, line): self.stamp = line return True elif re.match(self.sysinfofmt, line): self.sysinfo = line return True - elif re.match(self.kparamsfmt, line): - self.kparams = line - return True - elif re.match(self.cmdlinefmt, line): - self.cmdline = line - return True - elif re.match(self.mcelogfmt, line): - self.mcelog.append(line) - return True elif re.match(self.tstatfmt, line): self.turbostat.append(line) return True - elif re.match(self.batteryfmt, line): - self.battery.append(line) - return True elif re.match(self.wififmt, line): self.wifi.append(line) return True @@ -2745,10 +3022,26 @@ class TestProps: elif re.match(self.firmwarefmt, line): self.fwdata.append(line) return True + elif(re.match(self.devpropfmt, line)): + self.parseDevprops(line, sv) + return True + elif(re.match(self.pinfofmt, line)): + self.parsePlatformInfo(line, sv) + return True + m = re.match(self.cmdlinefmt, line) + if m: + self.cmdline = m.group('cmd') + return True + m = re.match(self.tracertypefmt, line) + if(m): + self.setTracerType(m.group('t')) + return True return False def parseStamp(self, data, sv): # global test data m = re.match(self.stampfmt, self.stamp) + if not self.stamp or not m: + doError('data does not include the expected stamp') data.stamp = {'time': '', 'host': '', 'mode': ''} dt = datetime(int(m.group('y'))+2000, int(m.group('m')), int(m.group('d')), int(m.group('H')), int(m.group('M')), @@ -2767,9 +3060,13 @@ class TestProps: data.stamp[key] = val sv.hostname = data.stamp['host'] sv.suspendmode = data.stamp['mode'] + if sv.suspendmode == 'freeze': + self.machinesuspend = 'timekeeping_freeze\[.*' + else: + self.machinesuspend = 'machine_suspend\[.*' if sv.suspendmode == 'command' and sv.ftracefile != '': modes = ['on', 'freeze', 'standby', 'mem', 'disk'] - fp = sysvals.openlog(sv.ftracefile, 'r') + fp = sv.openlog(sv.ftracefile, 'r') for line in fp: m = re.match('.* machine_suspend\[(?P<mode>.*)\]', line) if m and m.group('mode') in ['1', '2', '3', '4']: @@ -2777,13 +3074,7 @@ class TestProps: data.stamp['mode'] = sv.suspendmode break fp.close() - m = re.match(self.cmdlinefmt, self.cmdline) - if m: - sv.cmdline = m.group('cmd') - if self.kparams: - m = re.match(self.kparamsfmt, self.kparams) - if m: - sv.kparams = m.group('kp') + sv.cmdline = self.cmdline if not sv.stamp: sv.stamp = data.stamp # firmware data @@ -2793,26 +3084,18 @@ class TestProps: data.fwSuspend, data.fwResume = int(m.group('s')), int(m.group('r')) if(data.fwSuspend > 0 or data.fwResume > 0): data.fwValid = True - # mcelog data - if len(self.mcelog) > data.testnumber: - m = re.match(self.mcelogfmt, self.mcelog[data.testnumber]) - if m: - data.mcelog = sv.b64unzip(m.group('m')) # turbostat data if len(self.turbostat) > data.testnumber: m = re.match(self.tstatfmt, self.turbostat[data.testnumber]) if m: data.turbostat = m.group('t') - # battery data - if len(self.battery) > data.testnumber: - m = re.match(self.batteryfmt, self.battery[data.testnumber]) - if m: - data.battery = m.groups() # wifi data if len(self.wifi) > data.testnumber: m = re.match(self.wififmt, self.wifi[data.testnumber]) if m: - data.wifi = m.group('w') + data.wifi = {'dev': m.group('d'), 'stat': m.group('s'), + 'time': float(m.group('t'))} + data.stamp['wifi'] = m.group('d') # sleep mode enter errors if len(self.testerror) > data.testnumber: m = re.match(self.testerrfmt, self.testerror[data.testnumber]) @@ -2870,6 +3153,7 @@ class TestRun: self.ttemp = dict() class ProcessMonitor: + maxchars = 512 def __init__(self): self.proclist = dict() self.running = False @@ -2895,19 +3179,23 @@ class ProcessMonitor: if ujiff > 0 or kjiff > 0: running[pid] = ujiff + kjiff process.wait() - out = '' + out = [''] for pid in running: jiffies = running[pid] val = self.proclist[pid] - if out: - out += ',' - out += '%s-%s %d' % (val['name'], pid, jiffies) - return 'ps - '+out + if len(out[-1]) > self.maxchars: + out.append('') + elif len(out[-1]) > 0: + out[-1] += ',' + out[-1] += '%s-%s %d' % (val['name'], pid, jiffies) + if len(out) > 1: + for line in out: + sysvals.fsetVal('ps - @%d|%s' % (len(out), line), 'trace_marker') + else: + sysvals.fsetVal('ps - %s' % out[0], 'trace_marker') def processMonitor(self, tid): while self.running: - out = self.procstat() - if out: - sysvals.fsetVal(out, 'trace_marker') + self.procstat() def start(self): self.thread = Thread(target=self.processMonitor, args=(0,)) self.running = True @@ -2923,7 +3211,7 @@ class ProcessMonitor: # markers, and/or kprobes required for primary parsing. def doesTraceLogHaveTraceEvents(): kpcheck = ['_cal: (', '_ret: ('] - techeck = ['suspend_resume', 'device_pm_callback'] + techeck = ['suspend_resume', 'device_pm_callback', 'tracing_mark_write'] tmcheck = ['SUSPEND START', 'RESUME COMPLETE'] sysvals.usekprobes = False fp = sysvals.openlog(sysvals.ftracefile, 'r') @@ -2946,12 +3234,11 @@ def doesTraceLogHaveTraceEvents(): check.remove(i) tmcheck = check fp.close() - sysvals.usetraceevents = True if len(techeck) < 2 else False + sysvals.usetraceevents = True if len(techeck) < 3 else False sysvals.usetracemarkers = True if len(tmcheck) == 0 else False # Function: appendIncompleteTraceLog # Description: -# [deprecated for kernel 3.15 or newer] # Adds callgraph data which lacks trace event data. This is only # for timelines generated from 3.15 or older # Arguments: @@ -2973,20 +3260,7 @@ def appendIncompleteTraceLog(testruns): for line in tf: # remove any latent carriage returns line = line.replace('\r\n', '') - if tp.stampInfo(line): - continue - # determine the trace data type (required for further parsing) - m = re.match(tp.tracertypefmt, line) - if(m): - tp.setTracerType(m.group('t')) - continue - # device properties line - if(re.match(tp.devpropfmt, line)): - tp.parseDevprops(line, sysvals) - continue - # platform info line - if(re.match(tp.pinfofmt, line)): - tp.parsePlatformInfo(line, sysvals) + if tp.stampInfo(line, sysvals): continue # parse only valid lines, if this is not one move on m = re.match(tp.ftrace_line_fmt, line) @@ -3012,13 +3286,13 @@ def appendIncompleteTraceLog(testruns): if(t.startMarker()): data = testrun[testidx].data tp.parseStamp(data, sysvals) - data.setStart(t.time) + data.setStart(t.time, t.name) continue if(not data): continue # find the end of resume if(t.endMarker()): - data.setEnd(t.time) + data.setEnd(t.time, t.name) testidx += 1 if(testidx >= testcnt): break @@ -3066,6 +3340,61 @@ def appendIncompleteTraceLog(testruns): dev['ftrace'] = cg break +# Function: loadTraceLog +# Description: +# load the ftrace file into memory and fix up any ordering issues +# Output: +# TestProps instance and an array of lines in proper order +def loadTraceLog(): + tp, data, lines, trace = TestProps(), dict(), [], [] + tf = sysvals.openlog(sysvals.ftracefile, 'r') + for line in tf: + # remove any latent carriage returns + line = line.replace('\r\n', '') + if tp.stampInfo(line, sysvals): + continue + # ignore all other commented lines + if line[0] == '#': + continue + # ftrace line: parse only valid lines + m = re.match(tp.ftrace_line_fmt, line) + if(not m): + continue + dur = m.group('dur') if tp.cgformat else 'traceevent' + info = (m.group('time'), m.group('proc'), m.group('pid'), + m.group('msg'), dur) + # group the data by timestamp + t = float(info[0]) + if t in data: + data[t].append(info) + else: + data[t] = [info] + # we only care about trace event ordering + if (info[3].startswith('suspend_resume:') or \ + info[3].startswith('tracing_mark_write:')) and t not in trace: + trace.append(t) + tf.close() + for t in sorted(data): + first, last, blk = [], [], data[t] + if len(blk) > 1 and t in trace: + # move certain lines to the start or end of a timestamp block + for i in range(len(blk)): + if 'SUSPEND START' in blk[i][3]: + first.append(i) + elif re.match('.* timekeeping_freeze.*begin', blk[i][3]): + last.append(i) + elif re.match('.* timekeeping_freeze.*end', blk[i][3]): + first.append(i) + elif 'RESUME COMPLETE' in blk[i][3]: + last.append(i) + if len(first) == 1 and len(last) == 0: + blk.insert(0, blk.pop(first[0])) + elif len(last) == 1 and len(first) == 0: + blk.append(blk.pop(last[0])) + for info in blk: + lines.append(info) + return (tp, lines) + # Function: parseTraceLog # Description: # Analyze an ftrace log output file generated from this app during @@ -3081,56 +3410,22 @@ def parseTraceLog(live=False): doError('%s does not exist' % sysvals.ftracefile) if not live: sysvals.setupAllKprobes() - ksuscalls = ['pm_prepare_console'] + ksuscalls = ['ksys_sync', 'pm_prepare_console'] krescalls = ['pm_restore_console'] tracewatch = ['irq_wakeup'] if sysvals.usekprobes: tracewatch += ['sync_filesystems', 'freeze_processes', 'syscore_suspend', 'syscore_resume', 'resume_console', 'thaw_processes', 'CPU_ON', - 'CPU_OFF', 'timekeeping_freeze', 'acpi_suspend'] + 'CPU_OFF', 'acpi_suspend'] # extract the callgraph and traceevent data - tp = TestProps() - testruns = [] - testdata = [] - testrun = 0 - data = 0 - tf = sysvals.openlog(sysvals.ftracefile, 'r') + s2idle_enter = hwsus = False + testruns, testdata = [], [] + testrun, data, limbo = 0, 0, True phase = 'suspend_prepare' - for line in tf: - # remove any latent carriage returns - line = line.replace('\r\n', '') - if tp.stampInfo(line): - continue - # tracer type line: determine the trace data type - m = re.match(tp.tracertypefmt, line) - if(m): - tp.setTracerType(m.group('t')) - continue - # device properties line - if(re.match(tp.devpropfmt, line)): - tp.parseDevprops(line, sysvals) - continue - # platform info line - if(re.match(tp.pinfofmt, line)): - tp.parsePlatformInfo(line, sysvals) - continue - # ignore all other commented lines - if line[0] == '#': - continue - # ftrace line: parse only valid lines - m = re.match(tp.ftrace_line_fmt, line) - if(not m): - continue + tp, tf = loadTraceLog() + for m_time, m_proc, m_pid, m_msg, m_param3 in tf: # gather the basic message data from the line - m_time = m.group('time') - m_proc = m.group('proc') - m_pid = m.group('pid') - m_msg = m.group('msg') - if(tp.cgformat): - m_param3 = m.group('dur') - else: - m_param3 = 'traceevent' if(m_time and m_pid and m_msg): t = FTraceLine(m_time, m_msg, m_param3) pid = int(m_pid) @@ -3141,40 +3436,59 @@ def parseTraceLog(live=False): continue # find the start of suspend if(t.startMarker()): - data = Data(len(testdata)) + data, limbo = Data(len(testdata)), False testdata.append(data) testrun = TestRun(data) testruns.append(testrun) tp.parseStamp(data, sysvals) - data.setStart(t.time) + data.setStart(t.time, t.name) data.first_suspend_prepare = True phase = data.setPhase('suspend_prepare', t.time, True) continue - if(not data): + if(not data or limbo): continue # process cpu exec line if t.type == 'tracing_mark_write': + if t.name == 'CMD COMPLETE' and data.tKernRes == 0: + data.tKernRes = t.time m = re.match(tp.procexecfmt, t.name) if(m): - proclist = dict() - for ps in m.group('ps').split(','): + parts, msg = 1, m.group('ps') + m = re.match(tp.procmultifmt, msg) + if(m): + parts, msg = int(m.group('n')), m.group('ps') + if tp.multiproccnt == 0: + tp.multiproctime = t.time + tp.multiproclist = dict() + proclist = tp.multiproclist + tp.multiproccnt += 1 + else: + proclist = dict() + tp.multiproccnt = 0 + for ps in msg.split(','): val = ps.split() - if not val: + if not val or len(val) != 2: continue name = val[0].replace('--', '-') proclist[name] = int(val[1]) - data.pstl[t.time] = proclist + if parts == 1: + data.pstl[t.time] = proclist + elif parts == tp.multiproccnt: + data.pstl[tp.multiproctime] = proclist + tp.multiproccnt = 0 continue # find the end of resume if(t.endMarker()): - data.handleEndMarker(t.time) + if data.tKernRes == 0: + data.tKernRes = t.time + data.handleEndMarker(t.time, t.name) if(not sysvals.usetracemarkers): # no trace markers? then quit and be sure to finish recording # the event we used to trigger resume end if('thaw_processes' in testrun.ttemp and len(testrun.ttemp['thaw_processes']) > 0): # if an entry exists, assume this is its end testrun.ttemp['thaw_processes'][-1]['end'] = t.time - break + limbo = True continue # trace event processing if(t.fevent): @@ -3197,7 +3511,7 @@ def parseTraceLog(live=False): # -- phase changes -- # start of kernel suspend if(re.match('suspend_enter\[.*', t.name)): - if(isbegin): + if(isbegin and data.tKernSus == 0): data.tKernSus = t.time continue # suspend_prepare start @@ -3222,16 +3536,30 @@ def parseTraceLog(live=False): phase = data.setPhase('suspend_noirq', t.time, isbegin) continue # suspend_machine/resume_machine - elif(re.match('machine_suspend\[.*', t.name)): + elif(re.match(tp.machinesuspend, t.name)): + lp = data.lastPhase() if(isbegin): - lp = data.lastPhase() - if lp == 'resume_machine': - data.dmesg[lp]['end'] = t.time + hwsus = True + if lp.startswith('resume_machine'): + # trim out s2idle loops, track time trying to freeze + llp = data.lastPhase(2) + if llp.startswith('suspend_machine'): + if 'waking' not in data.dmesg[llp]: + data.dmesg[llp]['waking'] = [0, 0.0] + data.dmesg[llp]['waking'][0] += 1 + data.dmesg[llp]['waking'][1] += \ + t.time - data.dmesg[lp]['start'] + data.currphase = '' + del data.dmesg[lp] + continue phase = data.setPhase('suspend_machine', data.dmesg[lp]['end'], True) data.setPhase(phase, t.time, False) if data.tSuspended == 0: data.tSuspended = t.time else: + if lp.startswith('resume_machine'): + data.dmesg[lp]['end'] = t.time + continue phase = data.setPhase('resume_machine', t.time, True) if(sysvals.suspendmode in ['mem', 'disk']): susp = phase.replace('resume', 'suspend') @@ -3262,6 +3590,19 @@ def parseTraceLog(live=False): # global events (outside device calls) are graphed if(name not in testrun.ttemp): testrun.ttemp[name] = [] + # special handling for s2idle_enter + if name == 'machine_suspend': + if hwsus: + s2idle_enter = hwsus = False + elif s2idle_enter and not isbegin: + if(len(testrun.ttemp[name]) > 0): + testrun.ttemp[name][-1]['end'] = t.time + testrun.ttemp[name][-1]['loop'] += 1 + elif not s2idle_enter and isbegin: + s2idle_enter = True + testrun.ttemp[name].append({'begin': t.time, + 'end': t.time, 'pid': pid, 'loop': 0}) + continue if(isbegin): # create a new list entry testrun.ttemp[name].append(\ @@ -3293,9 +3634,8 @@ def parseTraceLog(live=False): if(not m): continue n = m.group('d') - list = data.dmesg[phase]['list'] - if(n in list): - dev = list[n] + dev = data.findDevice(phase, n) + if dev: dev['length'] = t.time - dev['start'] dev['end'] = t.time # kprobe event processing @@ -3320,7 +3660,8 @@ def parseTraceLog(live=False): 'proc': m_proc, }) # start of kernel resume - if(phase == 'suspend_prepare' and kprobename in ksuscalls): + if(data.tKernSus == 0 and phase == 'suspend_prepare' \ + and kprobename in ksuscalls): data.tKernSus = t.time elif(t.freturn): if(key not in tp.ktemp) or len(tp.ktemp[key]) < 1: @@ -3328,6 +3669,9 @@ def parseTraceLog(live=False): e = next((x for x in reversed(tp.ktemp[key]) if x['end'] < 0), 0) if not e: continue + if (t.time - e['begin']) * 1000 < sysvals.mindevlen: + tp.ktemp[key].pop() + continue e['end'] = t.time e['rdata'] = kprobedata # end of kernel resume @@ -3350,12 +3694,11 @@ def parseTraceLog(live=False): testrun.ftemp[key].append(FTraceCallGraph(pid, sysvals)) if(res == -1): testrun.ftemp[key][-1].addLine(t) - tf.close() if len(testdata) < 1: sysvals.vprint('WARNING: ftrace start marker is missing') if data and not data.devicegroups: sysvals.vprint('WARNING: ftrace end marker is missing') - data.handleEndMarker(t.time) + data.handleEndMarker(t.time, t.name) if sysvals.suspendmode == 'command': for test in testruns: @@ -3397,7 +3740,12 @@ def parseTraceLog(live=False): # add actual trace funcs for name in sorted(test.ttemp): for event in test.ttemp[name]: - data.newActionGlobal(name, event['begin'], event['end'], event['pid']) + if event['end'] - event['begin'] <= 0: + continue + title = name + if name == 'machine_suspend' and 'loop' in event: + title = 's2idle_enter_%dx' % event['loop'] + data.newActionGlobal(title, event['begin'], event['end'], event['pid']) # add the kprobe based virtual tracefuncs as actual devices for key in sorted(tp.ktemp): name, pid = key @@ -3466,8 +3814,15 @@ def parseTraceLog(live=False): for p in sorted(phasedef, key=lambda k:phasedef[k]['order']): if p not in data.dmesg: if not terr: - pprint('TEST%s FAILED: %s failed in %s phase' % (tn, sysvals.suspendmode, lp)) - terr = '%s%s failed in %s phase' % (sysvals.suspendmode, tn, lp) + ph = p if 'machine' in p else lp + if p == 'suspend_machine': + sm = sysvals.suspendmode + if sm in suspendmodename: + sm = suspendmodename[sm] + terr = 'test%s did not enter %s power mode' % (tn, sm) + else: + terr = '%s%s failed in %s phase' % (sysvals.suspendmode, tn, ph) + pprint('TEST%s FAILED: %s' % (tn, terr)) error.append(terr) if data.tSuspended == 0: data.tSuspended = data.dmesg[lp]['end'] @@ -3476,6 +3831,10 @@ def parseTraceLog(live=False): data.fwValid = False sysvals.vprint('WARNING: phase "%s" is missing!' % p) lp = p + if not terr and 'dev' in data.wifi and data.wifi['stat'] == 'timeout': + terr = '%s%s failed in wifi_resume <i>(%s %.0fs timeout)</i>' % \ + (sysvals.suspendmode, tn, data.wifi['dev'], data.wifi['time']) + error.append(terr) if not terr and data.enterfail: pprint('test%s FAILED: enter %s failed with %s' % (tn, sysvals.suspendmode, data.enterfail)) terr = 'test%s failed to enter %s mode' % (tn, sysvals.suspendmode) @@ -3503,9 +3862,7 @@ def parseTraceLog(live=False): # Function: loadKernelLog # Description: -# [deprecated for kernel 3.15.0 or newer] # load the dmesg file into memory and fix up any ordering issues -# The dmesg filename is taken from sysvals # Output: # An array of empty Data objects with only their dmesgtext attributes set def loadKernelLog(): @@ -3525,13 +3882,14 @@ def loadKernelLog(): idx = line.find('[') if idx > 1: line = line[idx:] - if tp.stampInfo(line): + if tp.stampInfo(line, sysvals): continue m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) if(not m): continue msg = m.group("msg") - if(re.match('PM: Syncing filesystems.*', msg)): + if re.match('PM: Syncing filesystems.*', msg) or \ + re.match('PM: suspend entry.*', msg): if(data): testruns.append(data) data = Data(len(testruns)) @@ -3542,11 +3900,17 @@ def loadKernelLog(): if(m): sysvals.stamp['kernel'] = m.group('k') m = re.match('PM: Preparing system for (?P<m>.*) sleep', msg) - if(m): + if not m: + m = re.match('PM: Preparing system for sleep \((?P<m>.*)\)', msg) + if m: sysvals.stamp['mode'] = sysvals.suspendmode = m.group('m') data.dmesgtext.append(line) lf.close() + if sysvals.suspendmode == 's2idle': + sysvals.suspendmode = 'freeze' + elif sysvals.suspendmode == 'deep': + sysvals.suspendmode = 'mem' if data: testruns.append(data) if len(testruns) < 1: @@ -3557,12 +3921,9 @@ def loadKernelLog(): for data in testruns: last = '' for line in data.dmesgtext: - mc = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) calling '+\ - '(?P<f>.*)\+ @ .*, parent: .*', line) - mr = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) call '+\ - '(?P<f>.*)\+ returned .* after (?P<dt>.*) usecs', last) - if(mc and mr and (mc.group('t') == mr.group('t')) and - (mc.group('f') == mr.group('f'))): + ct, cf, n, p = data.initcall_debug_call(line) + rt, rf, l = data.initcall_debug_return(last) + if ct and rt and ct == rt and cf == rf: i = data.dmesgtext.index(last) j = data.dmesgtext.index(line) data.dmesgtext[i] = line @@ -3572,7 +3933,6 @@ def loadKernelLog(): # Function: parseKernelLog # Description: -# [deprecated for kernel 3.15.0 or newer] # Analyse a dmesg log output file generated from this app during # the execution phase. Create a set of device structures in memory # for subsequent formatting in the html output file @@ -3591,30 +3951,30 @@ def parseKernelLog(data): # dmesg phase match table dm = { - 'suspend_prepare': ['PM: Syncing filesystems.*'], - 'suspend': ['PM: Entering [a-z]* sleep.*', 'Suspending console.*'], - 'suspend_late': ['PM: suspend of devices complete after.*'], - 'suspend_noirq': ['PM: late suspend of devices complete after.*'], - 'suspend_machine': ['PM: noirq suspend of devices complete after.*'], - 'resume_machine': ['ACPI: Low-level resume complete.*'], - 'resume_noirq': ['ACPI: Waking up from system sleep state.*'], - 'resume_early': ['PM: noirq resume of devices complete after.*'], - 'resume': ['PM: early resume of devices complete after.*'], - 'resume_complete': ['PM: resume of devices complete after.*'], + 'suspend_prepare': ['PM: Syncing filesystems.*', 'PM: suspend entry.*'], + 'suspend': ['PM: Entering [a-z]* sleep.*', 'Suspending console.*', + 'PM: Suspending system .*'], + 'suspend_late': ['PM: suspend of devices complete after.*', + 'PM: freeze of devices complete after.*'], + 'suspend_noirq': ['PM: late suspend of devices complete after.*', + 'PM: late freeze of devices complete after.*'], + 'suspend_machine': ['PM: suspend-to-idle', + 'PM: noirq suspend of devices complete after.*', + 'PM: noirq freeze of devices complete after.*'], + 'resume_machine': ['PM: Timekeeping suspended for.*', + 'ACPI: Low-level resume complete.*', + 'ACPI: resume from mwait', + 'Suspended for [0-9\.]* seconds'], + 'resume_noirq': ['PM: resume from suspend-to-idle', + 'ACPI: Waking up from system sleep state.*'], + 'resume_early': ['PM: noirq resume of devices complete after.*', + 'PM: noirq restore of devices complete after.*'], + 'resume': ['PM: early resume of devices complete after.*', + 'PM: early restore of devices complete after.*'], + 'resume_complete': ['PM: resume of devices complete after.*', + 'PM: restore of devices complete after.*'], 'post_resume': ['.*Restarting tasks \.\.\..*'], } - if(sysvals.suspendmode == 'standby'): - dm['resume_machine'] = ['PM: Restoring platform NVS memory'] - elif(sysvals.suspendmode == 'disk'): - dm['suspend_late'] = ['PM: freeze of devices complete after.*'] - dm['suspend_noirq'] = ['PM: late freeze of devices complete after.*'] - dm['suspend_machine'] = ['PM: noirq freeze of devices complete after.*'] - dm['resume_machine'] = ['PM: Restoring platform NVS memory'] - dm['resume_early'] = ['PM: noirq restore of devices complete after.*'] - dm['resume'] = ['PM: early restore of devices complete after.*'] - dm['resume_complete'] = ['PM: restore of devices complete after.*'] - elif(sysvals.suspendmode == 'freeze'): - dm['resume_machine'] = ['ACPI: resume from mwait'] # action table (expected events that occur and show up in dmesg) at = { @@ -3662,12 +4022,13 @@ def parseKernelLog(data): for s in dm[p]: if(re.match(s, msg)): phasechange, phase = True, p + dm[p] = [s] break # hack for determining resume_machine end for freeze if(not sysvals.usetraceevents and sysvals.suspendmode == 'freeze' \ and phase == 'resume_machine' and \ - re.match('calling (?P<f>.*)\+ @ .*, parent: .*', msg)): + data.initcall_debug_call(line, True)): data.setPhase(phase, ktime, False) phase = 'resume_noirq' data.setPhase(phase, ktime, True) @@ -3740,26 +4101,18 @@ def parseKernelLog(data): # -- device callbacks -- if(phase in data.sortedPhases()): # device init call - if(re.match('calling (?P<f>.*)\+ @ .*, parent: .*', msg)): - sm = re.match('calling (?P<f>.*)\+ @ '+\ - '(?P<n>.*), parent: (?P<p>.*)', msg); - f = sm.group('f') - n = sm.group('n') - p = sm.group('p') - if(f and n and p): - data.newAction(phase, f, int(n), p, ktime, -1, '') - # device init return - elif(re.match('call (?P<f>.*)\+ returned .* after '+\ - '(?P<t>.*) usecs', msg)): - sm = re.match('call (?P<f>.*)\+ returned .* after '+\ - '(?P<t>.*) usecs(?P<a>.*)', msg); - f = sm.group('f') - t = sm.group('t') - list = data.dmesg[phase]['list'] - if(f in list): - dev = list[f] - dev['length'] = int(t) - dev['end'] = ktime + t, f, n, p = data.initcall_debug_call(line) + if t and f and n and p: + data.newAction(phase, f, int(n), p, ktime, -1, '') + else: + # device init return + t, f, l = data.initcall_debug_return(line) + if t and f and l: + list = data.dmesg[phase]['list'] + if(f in list): + dev = list[f] + dev['length'] = int(l) + dev['end'] = ktime # if trace events are not available, these are better than nothing if(not sysvals.usetraceevents): @@ -3801,6 +4154,8 @@ def parseKernelLog(data): # fill in any missing phases phasedef = data.phasedef terr, lp = '', 'suspend_prepare' + if lp not in data.dmesg: + doError('dmesg log format has changed, could not find start of suspend') for p in sorted(phasedef, key=lambda k:phasedef[k]['order']): if p not in data.dmesg: if not terr: @@ -3856,6 +4211,8 @@ def callgraphHTML(sv, hf, num, cg, title, color, devid): fmt = '<n>(%.3f ms @ '+sv.timeformat+')</n>' flen = fmt % (line.length*1000, line.time) if line.isLeaf(): + if line.length * 1000 < sv.mincglen: + continue hf.write(html_func_leaf.format(line.name, flen)) elif line.freturn: hf.write(html_func_end) @@ -3873,18 +4230,20 @@ def addCallgraphs(sv, hf, data): if sv.cgphase and p != sv.cgphase: continue list = data.dmesg[p]['list'] - for devname in data.sortedDevices(p): - if len(sv.cgfilter) > 0 and devname not in sv.cgfilter: + for d in data.sortedDevices(p): + if len(sv.cgfilter) > 0 and d not in sv.cgfilter: continue - dev = list[devname] + dev = list[d] color = 'white' if 'color' in data.dmesg[p]: color = data.dmesg[p]['color'] if 'color' in dev: color = dev['color'] - name = devname - if(devname in sv.devprops): - name = sv.devprops[devname].altName(devname) + name = d if '[' not in d else d.split('[')[0] + if(d in sv.devprops): + name = sv.devprops[d].altName(d) + if 'drv' in dev and dev['drv']: + name += ' {%s}' % dev['drv'] if sv.suspendmode in suspendmodename: name += ' '+p if('ftrace' in dev): @@ -3933,7 +4292,7 @@ def createHTMLSummarySimple(testruns, htmlfile, title): tAvg, tMin, tMax, tMed = [0.0, 0.0], [0.0, 0.0], [0.0, 0.0], [dict(), dict()] iMin, iMed, iMax = [0, 0], [0, 0], [0, 0] num = 0 - useturbo = False + useturbo = usewifi = False lastmode = '' cnt = dict() for data in sorted(testruns, key=lambda v:(v['mode'], v['host'], v['kernel'], v['time'])): @@ -3952,17 +4311,17 @@ def createHTMLSummarySimple(testruns, htmlfile, title): tAvg, tMin, tMax, tMed = [0.0, 0.0], [0.0, 0.0], [0.0, 0.0], [dict(), dict()] iMin, iMed, iMax = [0, 0], [0, 0], [0, 0] num = 0 - pkgpc10 = syslpi = '' + pkgpc10 = syslpi = wifi = '' if 'pkgpc10' in data and 'syslpi' in data: - pkgpc10 = data['pkgpc10'] - syslpi = data['syslpi'] - useturbo = True + pkgpc10, syslpi, useturbo = data['pkgpc10'], data['syslpi'], True + if 'wifi' in data: + wifi, usewifi = data['wifi'], True res = data['result'] tVal = [float(data['suspend']), float(data['resume'])] list[mode]['data'].append([data['host'], data['kernel'], data['time'], tVal[0], tVal[1], data['url'], res, data['issues'], data['sus_worst'], data['sus_worsttime'], - data['res_worst'], data['res_worsttime'], pkgpc10, syslpi]) + data['res_worst'], data['res_worsttime'], pkgpc10, syslpi, wifi]) idx = len(list[mode]['data']) - 1 if res.startswith('fail in'): res = 'fail' @@ -4002,7 +4361,12 @@ def createHTMLSummarySimple(testruns, htmlfile, title): td = '\t<td>{0}</td>\n' tdh = '\t<td{1}>{0}</td>\n' tdlink = '\t<td><a href="{0}">html</a></td>\n' - colspan = '14' if useturbo else '12' + cols = 12 + if useturbo: + cols += 2 + if usewifi: + cols += 1 + colspan = '%d' % cols # table header html += '<table>\n<tr>\n' + th.format('#') +\ @@ -4013,6 +4377,8 @@ def createHTMLSummarySimple(testruns, htmlfile, title): th.format('Worst Resume Device') + th.format('RD Time') if useturbo: html += th.format('PkgPC10') + th.format('SysLPI') + if usewifi: + html += th.format('Wifi') html += th.format('Detail')+'</tr>\n' # export list into html head = '<tr class="head"><td>{0}</td><td>{1}</td>'+\ @@ -4076,6 +4442,8 @@ def createHTMLSummarySimple(testruns, htmlfile, title): if useturbo: html += td.format(d[12]) # pkg_pc10 html += td.format(d[13]) # syslpi + if usewifi: + html += td.format(d[14]) # wifi html += tdlink.format(d[5]) if d[5] else td.format('') # url html += '</tr>\n' num += 1 @@ -4224,6 +4592,8 @@ def createHTML(testruns, testfail): kerror = True if(sysvals.suspendmode in ['freeze', 'standby']): data.trimFreezeTime(testruns[-1].tSuspended) + else: + data.getMemTime() # html function templates html_error = '<div id="{1}" title="kernel error/warning" class="err" style="right:{0}%">{2}→</div>\n' @@ -4242,13 +4612,10 @@ def createHTML(testruns, testfail): '<td class="green">Execution Time: <b>{0} ms</b></td>'\ '<td class="yellow">Command: <b>{1}</b></td>'\ '</tr>\n</table>\n' - html_timegroups = '<table class="time2">\n<tr>'\ - '<td class="green" title="time from kernel enter_state({5}) to firmware mode [kernel time only]">{4}Kernel Suspend: {0} ms</td>'\ - '<td class="purple">{4}Firmware Suspend: {1} ms</td>'\ - '<td class="purple">{4}Firmware Resume: {2} ms</td>'\ - '<td class="yellow" title="time from firmware mode to return from kernel enter_state({5}) [kernel time only]">{4}Kernel Resume: {3} ms</td>'\ - '</tr>\n</table>\n' html_fail = '<table class="testfail"><tr><td>{0}</td></tr></table>\n' + html_kdesc = '<td class="{3}" title="time spent in kernel execution">{0}Kernel {2}: {1} ms</td>' + html_fwdesc = '<td class="{3}" title="time spent in firmware">{0}Firmware {2}: {1} ms</td>' + html_wifdesc = '<td class="yellow" title="time for wifi to reconnect after resume complete ({2})">{0}Wifi Resume: {1}</td>' # html format variables scaleH = 20 @@ -4264,13 +4631,10 @@ def createHTML(testruns, testfail): # Generate the header for this timeline for data in testruns: tTotal = data.end - data.start - sktime, rktime = data.getTimeValues() if(tTotal == 0): doError('No timeline data') - if(len(data.tLow) > 0): - low_time = '+'.join(data.tLow) if sysvals.suspendmode == 'command': - run_time = '%.0f'%((data.end-data.start)*1000) + run_time = '%.0f' % (tTotal * 1000) if sysvals.testcommand: testdesc = sysvals.testcommand else: @@ -4279,43 +4643,55 @@ def createHTML(testruns, testfail): testdesc = ordinal(data.testnumber+1)+' '+testdesc thtml = html_timetotal3.format(run_time, testdesc) devtl.html += thtml - elif data.fwValid: - suspend_time = '%.0f'%(sktime + (data.fwSuspend/1000000.0)) - resume_time = '%.0f'%(rktime + (data.fwResume/1000000.0)) - testdesc1 = 'Total' - testdesc2 = '' - stitle = 'time from kernel enter_state(%s) to low-power mode [kernel & firmware time]' % sysvals.suspendmode - rtitle = 'time from low-power mode to return from kernel enter_state(%s) [firmware & kernel time]' % sysvals.suspendmode - if(len(testruns) > 1): - testdesc1 = testdesc2 = ordinal(data.testnumber+1) - testdesc2 += ' ' - if(len(data.tLow) == 0): - thtml = html_timetotal.format(suspend_time, \ - resume_time, testdesc1, stitle, rtitle) - else: - thtml = html_timetotal2.format(suspend_time, low_time, \ - resume_time, testdesc1, stitle, rtitle) - devtl.html += thtml + continue + # typical full suspend/resume header + stot, rtot = sktime, rktime = data.getTimeValues() + ssrc, rsrc, testdesc, testdesc2 = ['kernel'], ['kernel'], 'Kernel', '' + if data.fwValid: + stot += (data.fwSuspend/1000000.0) + rtot += (data.fwResume/1000000.0) + ssrc.append('firmware') + rsrc.append('firmware') + testdesc = 'Total' + if 'time' in data.wifi and data.wifi['stat'] != 'timeout': + rtot += data.end - data.tKernRes + (data.wifi['time'] * 1000.0) + rsrc.append('wifi') + testdesc = 'Total' + suspend_time, resume_time = '%.3f' % stot, '%.3f' % rtot + stitle = 'time from kernel suspend start to %s mode [%s time]' % \ + (sysvals.suspendmode, ' & '.join(ssrc)) + rtitle = 'time from %s mode to kernel resume complete [%s time]' % \ + (sysvals.suspendmode, ' & '.join(rsrc)) + if(len(testruns) > 1): + testdesc = testdesc2 = ordinal(data.testnumber+1) + testdesc2 += ' ' + if(len(data.tLow) == 0): + thtml = html_timetotal.format(suspend_time, \ + resume_time, testdesc, stitle, rtitle) + else: + low_time = '+'.join(data.tLow) + thtml = html_timetotal2.format(suspend_time, low_time, \ + resume_time, testdesc, stitle, rtitle) + devtl.html += thtml + if not data.fwValid and 'dev' not in data.wifi: + continue + # extra detail when the times come from multiple sources + thtml = '<table class="time2">\n<tr>' + thtml += html_kdesc.format(testdesc2, '%.3f'%sktime, 'Suspend', 'green') + if data.fwValid: sftime = '%.3f'%(data.fwSuspend / 1000000.0) rftime = '%.3f'%(data.fwResume / 1000000.0) - devtl.html += html_timegroups.format('%.3f'%sktime, \ - sftime, rftime, '%.3f'%rktime, testdesc2, sysvals.suspendmode) - else: - suspend_time = '%.3f' % sktime - resume_time = '%.3f' % rktime - testdesc = 'Kernel' - stitle = 'time from kernel enter_state(%s) to firmware mode [kernel time only]' % sysvals.suspendmode - rtitle = 'time from firmware mode to return from kernel enter_state(%s) [kernel time only]' % sysvals.suspendmode - if(len(testruns) > 1): - testdesc = ordinal(data.testnumber+1)+' '+testdesc - if(len(data.tLow) == 0): - thtml = html_timetotal.format(suspend_time, \ - resume_time, testdesc, stitle, rtitle) + thtml += html_fwdesc.format(testdesc2, sftime, 'Suspend', 'green') + thtml += html_fwdesc.format(testdesc2, rftime, 'Resume', 'yellow') + thtml += html_kdesc.format(testdesc2, '%.3f'%rktime, 'Resume', 'yellow') + if 'time' in data.wifi: + if data.wifi['stat'] != 'timeout': + wtime = '%.0f ms'%(data.end - data.tKernRes + (data.wifi['time'] * 1000.0)) else: - thtml = html_timetotal2.format(suspend_time, low_time, \ - resume_time, testdesc, stitle, rtitle) - devtl.html += thtml - + wtime = 'TIMEOUT' + thtml += html_wifdesc.format(testdesc2, wtime, data.wifi['dev']) + thtml += '</tr>\n</table>\n' + devtl.html += thtml if testfail: devtl.html += html_fail.format(testfail) @@ -4414,12 +4790,9 @@ def createHTML(testruns, testfail): # draw the devices for this phase phaselist = data.dmesg[b]['list'] for d in sorted(data.tdevlist[b]): - name = d - drv = '' - dev = phaselist[d] - xtraclass = '' - xtrainfo = '' - xtrastyle = '' + dname = d if ('[' not in d or 'CPU' in d) else d.split('[')[0] + name, dev = dname, phaselist[d] + drv = xtraclass = xtrainfo = xtrastyle = '' if 'htmlclass' in dev: xtraclass = dev['htmlclass'] if 'color' in dev: @@ -4450,24 +4823,23 @@ def createHTML(testruns, testfail): title += b devtl.html += devtl.html_device.format(dev['id'], \ title, left, top, '%.3f'%rowheight, width, \ - d+drv, xtraclass, xtrastyle) + dname+drv, xtraclass, xtrastyle) if('cpuexec' in dev): for t in sorted(dev['cpuexec']): start, end = t - j = float(dev['cpuexec'][t]) / 5 - if j > 1.0: - j = 1.0 height = '%.3f' % (rowheight/3) top = '%.3f' % (rowtop + devtl.scaleH + 2*rowheight/3) left = '%f' % (((start-m0)*100)/mTotal) width = '%f' % ((end-start)*100/mTotal) - color = 'rgba(255, 0, 0, %f)' % j + color = 'rgba(255, 0, 0, %f)' % dev['cpuexec'][t] devtl.html += \ html_cpuexec.format(left, top, height, width, color) if('src' not in dev): continue # draw any trace events for this device for e in dev['src']: + if e.length == 0: + continue height = '%.3f' % devtl.rowH top = '%.3f' % (rowtop + devtl.scaleH + (e.row*devtl.rowH)) left = '%f' % (((e.time-m0)*100)/mTotal) @@ -5054,157 +5426,142 @@ def addScriptCode(hf, testruns): '</script>\n' hf.write(script_code); -def setRuntimeSuspend(before=True): - global sysvals - sv = sysvals - if sv.rs == 0: - return - if before: - # runtime suspend disable or enable - if sv.rs > 0: - sv.rstgt, sv.rsval, sv.rsdir = 'on', 'auto', 'enabled' - else: - sv.rstgt, sv.rsval, sv.rsdir = 'auto', 'on', 'disabled' - pprint('CONFIGURING RUNTIME SUSPEND...') - sv.rslist = deviceInfo(sv.rstgt) - for i in sv.rslist: - sv.setVal(sv.rsval, i) - pprint('runtime suspend %s on all devices (%d changed)' % (sv.rsdir, len(sv.rslist))) - pprint('waiting 5 seconds...') - time.sleep(5) - else: - # runtime suspend re-enable or re-disable - for i in sv.rslist: - sv.setVal(sv.rstgt, i) - pprint('runtime suspend settings restored on %d devices' % len(sv.rslist)) - # Function: executeSuspend # Description: # Execute system suspend through the sysfs interface, then copy the output # dmesg and ftrace files to the test output directory. -def executeSuspend(): - pm = ProcessMonitor() - tp = sysvals.tpath - wifi = sysvals.checkWifi() +def executeSuspend(quiet=False): + sv, tp, pm = sysvals, sysvals.tpath, ProcessMonitor() + if sv.wifi: + wifi = sv.checkWifi() + sv.dlog('wifi check, connected device is "%s"' % wifi) testdata = [] - battery = True if getBattery() else False # run these commands to prepare the system for suspend - if sysvals.display: - pprint('SET DISPLAY TO %s' % sysvals.display.upper()) - displayControl(sysvals.display) + if sv.display: + if not quiet: + pprint('SET DISPLAY TO %s' % sv.display.upper()) + ret = sv.displayControl(sv.display) + sv.dlog('xset display %s, ret = %d' % (sv.display, ret)) time.sleep(1) - if sysvals.sync: - pprint('SYNCING FILESYSTEMS') + if sv.sync: + if not quiet: + pprint('SYNCING FILESYSTEMS') + sv.dlog('syncing filesystems') call('sync', shell=True) - # mark the start point in the kernel ring buffer just as we start - sysvals.initdmesg() - # start ftrace - if(sysvals.usecallgraph or sysvals.usetraceevents): - pprint('START TRACING') - sysvals.fsetVal('1', 'tracing_on') - if sysvals.useprocmon: - pm.start() + sv.dlog('read dmesg') + sv.initdmesg() + sv.dlog('cmdinfo before') + sv.cmdinfo(True) + sv.start(pm) # execute however many s/r runs requested - for count in range(1,sysvals.execcount+1): + for count in range(1,sv.execcount+1): # x2delay in between test runs - if(count > 1 and sysvals.x2delay > 0): - sysvals.fsetVal('WAIT %d' % sysvals.x2delay, 'trace_marker') - time.sleep(sysvals.x2delay/1000.0) - sysvals.fsetVal('WAIT END', 'trace_marker') + if(count > 1 and sv.x2delay > 0): + sv.fsetVal('WAIT %d' % sv.x2delay, 'trace_marker') + time.sleep(sv.x2delay/1000.0) + sv.fsetVal('WAIT END', 'trace_marker') # start message - if sysvals.testcommand != '': + if sv.testcommand != '': pprint('COMMAND START') else: - if(sysvals.rtcwake): + if(sv.rtcwake): pprint('SUSPEND START') else: pprint('SUSPEND START (press a key to resume)') - sysvals.mcelog(True) - bat1 = getBattery() if battery else False # set rtcwake - if(sysvals.rtcwake): - pprint('will issue an rtcwake in %d seconds' % sysvals.rtcwaketime) - sysvals.rtcWakeAlarmOn() + if(sv.rtcwake): + if not quiet: + pprint('will issue an rtcwake in %d seconds' % sv.rtcwaketime) + sv.dlog('enable RTC wake alarm') + sv.rtcWakeAlarmOn() # start of suspend trace marker - if(sysvals.usecallgraph or sysvals.usetraceevents): - sysvals.fsetVal('SUSPEND START', 'trace_marker') + sv.fsetVal(datetime.now().strftime(sv.tmstart), 'trace_marker') # predelay delay - if(count == 1 and sysvals.predelay > 0): - sysvals.fsetVal('WAIT %d' % sysvals.predelay, 'trace_marker') - time.sleep(sysvals.predelay/1000.0) - sysvals.fsetVal('WAIT END', 'trace_marker') + if(count == 1 and sv.predelay > 0): + sv.fsetVal('WAIT %d' % sv.predelay, 'trace_marker') + time.sleep(sv.predelay/1000.0) + sv.fsetVal('WAIT END', 'trace_marker') # initiate suspend or command + sv.dlog('system executing a suspend') tdata = {'error': ''} - if sysvals.testcommand != '': - res = call(sysvals.testcommand+' 2>&1', shell=True); + if sv.testcommand != '': + res = call(sv.testcommand+' 2>&1', shell=True); if res != 0: tdata['error'] = 'cmd returned %d' % res else: - mode = sysvals.suspendmode - if sysvals.memmode and os.path.exists(sysvals.mempowerfile): + s0ixready = sv.s0ixSupport() + mode = sv.suspendmode + if sv.memmode and os.path.exists(sv.mempowerfile): mode = 'mem' - pf = open(sysvals.mempowerfile, 'w') - pf.write(sysvals.memmode) - pf.close() - if sysvals.diskmode and os.path.exists(sysvals.diskpowerfile): + sv.testVal(sv.mempowerfile, 'radio', sv.memmode) + if sv.diskmode and os.path.exists(sv.diskpowerfile): mode = 'disk' - pf = open(sysvals.diskpowerfile, 'w') - pf.write(sysvals.diskmode) - pf.close() - if mode == 'freeze' and sysvals.haveTurbostat(): + sv.testVal(sv.diskpowerfile, 'radio', sv.diskmode) + if sv.acpidebug: + sv.testVal(sv.acpipath, 'acpi', '0xe') + if ((mode == 'freeze') or (sv.memmode == 's2idle')) \ + and sv.haveTurbostat(): # execution will pause here - turbo = sysvals.turbostat() + turbo = sv.turbostat(s0ixready) if turbo: tdata['turbo'] = turbo else: - pf = open(sysvals.powerfile, 'w') + pf = open(sv.powerfile, 'w') pf.write(mode) # execution will pause here try: pf.close() except Exception as e: tdata['error'] = str(e) - if(sysvals.rtcwake): - sysvals.rtcWakeAlarmOff() + sv.fsetVal('CMD COMPLETE', 'trace_marker') + sv.dlog('system returned') + # reset everything + sv.testVal('restoreall') + if(sv.rtcwake): + sv.dlog('disable RTC wake alarm') + sv.rtcWakeAlarmOff() # postdelay delay - if(count == sysvals.execcount and sysvals.postdelay > 0): - sysvals.fsetVal('WAIT %d' % sysvals.postdelay, 'trace_marker') - time.sleep(sysvals.postdelay/1000.0) - sysvals.fsetVal('WAIT END', 'trace_marker') + if(count == sv.execcount and sv.postdelay > 0): + sv.fsetVal('WAIT %d' % sv.postdelay, 'trace_marker') + time.sleep(sv.postdelay/1000.0) + sv.fsetVal('WAIT END', 'trace_marker') # return from suspend pprint('RESUME COMPLETE') - if(sysvals.usecallgraph or sysvals.usetraceevents): - sysvals.fsetVal('RESUME COMPLETE', 'trace_marker') - if(sysvals.suspendmode == 'mem' or sysvals.suspendmode == 'command'): + if(count < sv.execcount): + sv.fsetVal(datetime.now().strftime(sv.tmend), 'trace_marker') + elif(not sv.wifitrace): + sv.fsetVal(datetime.now().strftime(sv.tmend), 'trace_marker') + sv.stop(pm) + if sv.wifi and wifi: + tdata['wifi'] = sv.pollWifi(wifi) + sv.dlog('wifi check, %s' % tdata['wifi']) + if(count == sv.execcount and sv.wifitrace): + sv.fsetVal(datetime.now().strftime(sv.tmend), 'trace_marker') + sv.stop(pm) + if sv.netfix: + tdata['netfix'] = sv.netfixon() + sv.dlog('netfix, %s' % tdata['netfix']) + if(sv.suspendmode == 'mem' or sv.suspendmode == 'command'): + sv.dlog('read the ACPI FPDT') tdata['fw'] = getFPDT(False) - mcelog = sysvals.mcelog() - if mcelog: - tdata['mcelog'] = mcelog - bat2 = getBattery() if battery else False - if battery and bat1 and bat2: - tdata['bat'] = (bat1, bat2) - if 'device' in wifi and 'ip' in wifi: - tdata['wifi'] = (wifi, sysvals.checkWifi()) testdata.append(tdata) - # stop ftrace - if(sysvals.usecallgraph or sysvals.usetraceevents): - if sysvals.useprocmon: - pm.stop() - sysvals.fsetVal('0', 'tracing_on') + sv.dlog('cmdinfo after') + cmdafter = sv.cmdinfo(False) # grab a copy of the dmesg output - pprint('CAPTURING DMESG') - sysvals.getdmesg(testdata) + if not quiet: + pprint('CAPTURING DMESG') + sv.getdmesg(testdata) # grab a copy of the ftrace output - if(sysvals.usecallgraph or sysvals.usetraceevents): - pprint('CAPTURING TRACE') - op = sysvals.writeDatafileHeader(sysvals.ftracefile, testdata) + if sv.useftrace: + if not quiet: + pprint('CAPTURING TRACE') + op = sv.writeDatafileHeader(sv.ftracefile, testdata) fp = open(tp+'trace', 'r') for line in fp: op.write(line) op.close() - sysvals.fsetVal('', 'trace') - sysvals.platforminfo() - return testdata + sv.fsetVal('', 'trace') + sv.platforminfo(cmdafter) def readFile(file): if os.path.islink(file): @@ -5447,58 +5804,6 @@ def dmidecode(mempath, fatal=False): count += 1 return out -def getBattery(): - p, charge, bat = '/sys/class/power_supply', 0, {} - if not os.path.exists(p): - return False - for d in os.listdir(p): - type = sysvals.getVal(os.path.join(p, d, 'type')).strip().lower() - if type != 'battery': - continue - for v in ['status', 'energy_now', 'capacity_now']: - bat[v] = sysvals.getVal(os.path.join(p, d, v)).strip().lower() - break - if 'status' not in bat: - return False - ac = False if 'discharging' in bat['status'] else True - for v in ['energy_now', 'capacity_now']: - if v in bat and bat[v]: - charge = int(bat[v]) - return (ac, charge) - -def displayControl(cmd): - xset, ret = 'timeout 10 xset -d :0.0 {0}', 0 - if sysvals.sudouser: - xset = 'sudo -u %s %s' % (sysvals.sudouser, xset) - if cmd == 'init': - ret = call(xset.format('dpms 0 0 0'), shell=True) - if not ret: - ret = call(xset.format('s off'), shell=True) - elif cmd == 'reset': - ret = call(xset.format('s reset'), shell=True) - elif cmd in ['on', 'off', 'standby', 'suspend']: - b4 = displayControl('stat') - ret = call(xset.format('dpms force %s' % cmd), shell=True) - if not ret: - curr = displayControl('stat') - sysvals.vprint('Display Switched: %s -> %s' % (b4, curr)) - if curr != cmd: - sysvals.vprint('WARNING: Display failed to change to %s' % cmd) - if ret: - sysvals.vprint('WARNING: Display failed to change to %s with xset' % cmd) - return ret - elif cmd == 'stat': - fp = Popen(xset.format('q').split(' '), stdout=PIPE).stdout - ret = 'unknown' - for line in fp: - m = re.match('[\s]*Monitor is (?P<m>.*)', ascii(line)) - if(m and len(m.group('m')) >= 2): - out = m.group('m').lower() - ret = out[3:] if out[0:2] == 'in' else out - break - fp.close() - return ret - # Function: getFPDT # Description: # Read the acpi bios tables and pull out FPDT, the firmware data @@ -5678,13 +5983,19 @@ def statusCheck(probecheck=False): pprint(' please choose one with -m') # check if ftrace is available - res = sysvals.colorText('NO') - ftgood = sysvals.verifyFtrace() - if(ftgood): - res = 'YES' - elif(sysvals.usecallgraph): - status = 'ftrace is not properly supported' - pprint(' is ftrace supported: %s' % res) + if sysvals.useftrace: + res = sysvals.colorText('NO') + sysvals.useftrace = sysvals.verifyFtrace() + efmt = '"{0}" uses ftrace, and it is not properly supported' + if sysvals.useftrace: + res = 'YES' + elif sysvals.usecallgraph: + status = efmt.format('-f') + elif sysvals.usedevsrc: + status = efmt.format('-dev') + elif sysvals.useprocmon: + status = efmt.format('-proc') + pprint(' is ftrace supported: %s' % res) # check if kprobes are available if sysvals.usekprobes: @@ -5697,8 +6008,8 @@ def statusCheck(probecheck=False): pprint(' are kprobes supported: %s' % res) # what data source are we using - res = 'DMESG' - if(ftgood): + res = 'DMESG (very limited, ftrace is preferred)' + if sysvals.useftrace: sysvals.usetraceevents = True for e in sysvals.traceevents: if not os.path.exists(sysvals.epath+e): @@ -5715,6 +6026,17 @@ def statusCheck(probecheck=False): status = 'rtcwake is not properly supported' pprint(' is rtcwake supported: %s' % res) + # check info commands + pprint(' optional commands this tool may use for info:') + no = sysvals.colorText('MISSING') + yes = sysvals.colorText('FOUND', 32) + for c in ['turbostat', 'mcelog', 'lspci', 'lsusb', 'netfix']: + if c == 'turbostat': + res = yes if sysvals.haveTurbostat() else no + else: + res = yes if sysvals.getExec(c) else no + pprint(' %s: %s' % (c, res)) + if not probecheck: return status @@ -5780,8 +6102,9 @@ def getArgFloat(name, args, min, max, main=True): doError(name+': value should be between %f and %f' % (min, max), True) return val -def processData(live=False): - pprint('PROCESSING DATA') +def processData(live=False, quiet=False): + if not quiet: + pprint('PROCESSING: %s' % sysvals.htmlfile) sysvals.vprint('usetraceevents=%s, usetracemarkers=%s, usekprobes=%s' % \ (sysvals.usetraceevents, sysvals.usetracemarkers, sysvals.usekprobes)) error = '' @@ -5796,20 +6119,17 @@ def processData(live=False): parseKernelLog(data) if(sysvals.ftracefile and (sysvals.usecallgraph or sysvals.usetraceevents)): appendIncompleteTraceLog(testruns) - shown = ['bios', 'biosdate', 'cpu', 'host', 'kernel', 'man', 'memfr', - 'memsz', 'mode', 'numcpu', 'plat', 'time'] + if not sysvals.stamp: + pprint('ERROR: data does not include the expected stamp') + return (testruns, {'error': 'timeline generation failed'}) + shown = ['os', 'bios', 'biosdate', 'cpu', 'host', 'kernel', 'man', 'memfr', + 'memsz', 'mode', 'numcpu', 'plat', 'time', 'wifi'] sysvals.vprint('System Info:') for key in sorted(sysvals.stamp): if key in shown: sysvals.vprint(' %-8s : %s' % (key.upper(), sysvals.stamp[key])) - if sysvals.kparams: - sysvals.vprint('Kparams:\n %s' % sysvals.kparams) sysvals.vprint('Command:\n %s' % sysvals.cmdline) for data in testruns: - if data.mcelog: - sysvals.vprint('MCELOG Data:') - for line in data.mcelog.split('\n'): - sysvals.vprint(' %s' % line) if data.turbostat: idx, s = 0, 'Turbostat:\n ' for val in data.turbostat.split('|'): @@ -5819,23 +6139,13 @@ def processData(live=False): s += '\n ' s += val + ' ' sysvals.vprint(s) - if data.battery: - a1, c1, a2, c2 = data.battery - s = 'Battery:\n Before - AC: %s, Charge: %d\n After - AC: %s, Charge: %d' % \ - (a1, int(c1), a2, int(c2)) - sysvals.vprint(s) - if data.wifi: - w = data.wifi.replace('|', ' ').split(',') - s = 'Wifi:\n Before %s\n After %s' % \ - (w[0], w[1]) - sysvals.vprint(s) data.printDetails() - if len(sysvals.platinfo) > 0: - sysvals.vprint('\nPlatform Info:') - for info in sysvals.platinfo: - sysvals.vprint(info[0]+' - '+info[1]) - sysvals.vprint(info[2]) - sysvals.vprint('') + if len(sysvals.platinfo) > 0: + sysvals.vprint('\nPlatform Info:') + for info in sysvals.platinfo: + sysvals.vprint('[%s - %s]' % (info[0], info[1])) + sysvals.vprint(info[2]) + sysvals.vprint('') if sysvals.cgdump: for data in testruns: data.debugPrint() @@ -5845,7 +6155,8 @@ def processData(live=False): return (testruns, {'error': 'timeline generation failed'}) sysvals.vprint('Creating the html timeline (%s)...' % sysvals.htmlfile) createHTML(testruns, error) - pprint('DONE') + if not quiet: + pprint('DONE: %s' % sysvals.htmlfile) data = testruns[0] stamp = data.stamp stamp['suspend'], stamp['resume'] = data.getTimeValues() @@ -5872,31 +6183,41 @@ def rerunTest(htmlfile=''): doError('a directory already exists with this name: %s' % sysvals.htmlfile) elif not os.access(sysvals.htmlfile, os.W_OK): doError('missing permission to write to %s' % sysvals.htmlfile) - testruns, stamp = processData(False) - sysvals.logmsg = '' + testruns, stamp = processData() + sysvals.resetlog() return stamp # Function: runTest # Description: # execute a suspend/resume, gather the logs, and generate the output -def runTest(n=0): +def runTest(n=0, quiet=False): # prepare for the test - sysvals.initFtrace() sysvals.initTestOutput('suspend') + op = sysvals.writeDatafileHeader(sysvals.dmesgfile, []) + op.write('# EXECUTION TRACE START\n') + op.close() + if n <= 1: + if sysvals.rs != 0: + sysvals.dlog('%sabling runtime suspend' % ('en' if sysvals.rs > 0 else 'dis')) + sysvals.setRuntimeSuspend(True) + if sysvals.display: + ret = sysvals.displayControl('init') + sysvals.dlog('xset display init, ret = %d' % ret) + sysvals.testVal(sysvals.pmdpath, 'basic', '1') + sysvals.testVal(sysvals.s0ixpath, 'basic', 'Y') + sysvals.dlog('initialize ftrace') + sysvals.initFtrace(quiet) # execute the test - testdata = executeSuspend() + executeSuspend(quiet) sysvals.cleanupFtrace() if sysvals.skiphtml: + sysvals.outputResult({}, n) sysvals.sudoUserchown(sysvals.testdir) return - if not testdata[0]['error']: - testruns, stamp = processData(True) - for data in testruns: - del data - else: - stamp = testdata[0] - + testruns, stamp = processData(True, quiet) + for data in testruns: + del data sysvals.sudoUserchown(sysvals.testdir) sysvals.outputResult(stamp, n) if 'error' in stamp: @@ -5904,24 +6225,27 @@ def runTest(n=0): return 0 def find_in_html(html, start, end, firstonly=True): - n, out = 0, [] - while n < len(html): - m = re.search(start, html[n:]) - if not m: - break - i = m.end() - m = re.search(end, html[n+i:]) + cnt, out, list = len(html), [], [] + if firstonly: + m = re.search(start, html) + if m: + list.append(m) + else: + list = re.finditer(start, html) + for match in list: + s = match.end() + e = cnt if (len(out) < 1 or s + 10000 > cnt) else s + 10000 + m = re.search(end, html[s:e]) if not m: break - j = m.start() - str = html[n+i:n+i+j] + e = s + m.start() + str = html[s:e] if end == 'ms': num = re.search(r'[-+]?\d*\.\d+|\d+', str) str = num.group() if num else 'NaN' if firstonly: return str out.append(str) - n += i+j if firstonly: return '' return out @@ -5945,7 +6269,7 @@ def data_from_html(file, outpath, issues, fulldetail=False): tstr = dt.strftime('%Y/%m/%d %H:%M:%S') error = find_in_html(html, '<table class="testfail"><tr><td>', '</td>') if error: - m = re.match('[a-z]* failed in (?P<p>[a-z0-9_]*) phase', error) + m = re.match('[a-z0-9]* failed in (?P<p>\S*).*', error) if m: result = 'fail in %s' % m.group('p') else: @@ -5953,7 +6277,7 @@ def data_from_html(file, outpath, issues, fulldetail=False): else: result = 'pass' # extract error info - ilist = [] + tp, ilist = False, [] extra = dict() log = find_in_html(html, '<div id="dmesglog" style="display:none;">', '</div>').strip() @@ -5961,8 +6285,8 @@ def data_from_html(file, outpath, issues, fulldetail=False): d = Data(0) d.end = 999999999 d.dmesgtext = log.split('\n') - msglist = d.extractErrorInfo() - for msg in msglist: + tp = d.extractErrorInfo() + for msg in tp.msglist: sysvals.errorSummary(issues, msg) if stmp[2] == 'freeze': extra = d.turbostatInfo() @@ -5974,9 +6298,23 @@ def data_from_html(file, outpath, issues, fulldetail=False): elist[err[0]] += 1 for i in elist: ilist.append('%sx%d' % (i, elist[i]) if elist[i] > 1 else i) + line = find_in_html(log, '# wifi ', '\n') + if line: + extra['wifi'] = line + line = find_in_html(log, '# netfix ', '\n') + if line: + extra['netfix'] = line low = find_in_html(html, 'freeze time: <b>', ' ms</b>') - if low and '|' in low: - issue = 'FREEZEx%d' % len(low.split('|')) + for lowstr in ['waking', '+']: + if not low: + break + if lowstr not in low: + continue + if lowstr == '+': + issue = 'S2LOOPx%d' % len(low.split('+')) + else: + m = re.match('.*waking *(?P<n>[0-9]*) *times.*', low) + issue = 'S2WAKEx%s' % m.group('n') if m else 'S2WAKExNaN' match = [i for i in issues if i['match'] == issue] if len(match) > 0: match[0]['count'] += 1 @@ -6000,6 +6338,8 @@ def data_from_html(file, outpath, issues, fulldetail=False): if not m: continue name, time, phase = m.group('n'), m.group('t'), m.group('p') + if name == 'async_synchronize_full': + continue if ' async' in name or ' sync' in name: name = ' '.join(name.split(' ')[:-1]) if phase.startswith('suspend'): @@ -6042,19 +6382,26 @@ def data_from_html(file, outpath, issues, fulldetail=False): data[key] = extra[key] if fulldetail: data['funclist'] = find_in_html(html, '<div title="', '" class="traceevent"', False) + if tp: + for arg in ['-multi ', '-info ']: + if arg in tp.cmdline: + data['target'] = tp.cmdline[tp.cmdline.find(arg):].split()[1] + break return data def genHtml(subdir, force=False): for dirname, dirnames, filenames in os.walk(subdir): sysvals.dmesgfile = sysvals.ftracefile = sysvals.htmlfile = '' for filename in filenames: - if(re.match('.*_dmesg.txt', filename)): - sysvals.dmesgfile = os.path.join(dirname, filename) - elif(re.match('.*_ftrace.txt', filename)): - sysvals.ftracefile = os.path.join(dirname, filename) + file = os.path.join(dirname, filename) + if sysvals.usable(file): + if(re.match('.*_dmesg.txt', filename)): + sysvals.dmesgfile = file + elif(re.match('.*_ftrace.txt', filename)): + sysvals.ftracefile = file sysvals.setOutputFile() - if sysvals.ftracefile and sysvals.htmlfile and \ - (force or not os.path.exists(sysvals.htmlfile)): + if (sysvals.dmesgfile or sysvals.ftracefile) and sysvals.htmlfile and \ + (force or not sysvals.usable(sysvals.htmlfile, True)): pprint('FTRACE: %s' % sysvals.ftracefile) if sysvals.dmesgfile: pprint('DMESG : %s' % sysvals.dmesgfile) @@ -6069,8 +6416,7 @@ def runSummary(subdir, local=True, genhtml=False): pprint('Generating a summary of folder:\n %s' % inpath) if genhtml: genHtml(subdir) - issues = [] - testruns = [] + target, issues, testruns = '', [], [] desc = {'host':[],'mode':[],'kernel':[]} for dirname, dirnames, filenames in os.walk(subdir): for filename in filenames: @@ -6079,6 +6425,8 @@ def runSummary(subdir, local=True, genhtml=False): data = data_from_html(os.path.join(dirname, filename), outpath, issues) if(not data): continue + if 'target' in data: + target = data['target'] testruns.append(data) for key in desc: if data[key] not in desc[key]: @@ -6086,6 +6434,8 @@ def runSummary(subdir, local=True, genhtml=False): pprint('Summary files:') if len(desc['host']) == len(desc['mode']) == len(desc['kernel']) == 1: title = '%s %s %s' % (desc['host'][0], desc['kernel'][0], desc['mode'][0]) + if target: + title += ' %s' % target else: title = inpath createHTMLSummarySimple(testruns, os.path.join(outpath, 'summary.html'), title) @@ -6169,9 +6519,9 @@ def configFromFile(file): sysvals.cgtest = getArgInt('cgtest', value, 0, 1, False) elif(option == 'cgphase'): d = Data(0) - if value not in d.sortedPhases(): + if value not in d.phasedef: doError('invalid phase --> (%s: %s), valid phases are %s'\ - % (option, value, d.sortedPhases()), True) + % (option, value, d.phasedef.keys()), True) sysvals.cgphase = value elif(option == 'fadd'): file = sysvals.configFile(value) @@ -6184,9 +6534,7 @@ def configFromFile(file): nums = value.split() if len(nums) != 2: doError('multi requires 2 integers (exec_count and delay)', True) - sysvals.multitest['run'] = True - sysvals.multitest['count'] = getArgInt('multi: n d (exec count)', nums[0], 2, 1000000, False) - sysvals.multitest['delay'] = getArgInt('multi: n d (delay between tests)', nums[1], 0, 3600, False) + sysvals.multiinit(nums[0], nums[1]) elif(option == 'devicefilter'): sysvals.setDeviceFilter(value) elif(option == 'expandcg'): @@ -6342,6 +6690,9 @@ def printHelp(): ' -srgap Add a visible gap in the timeline between sus/res (default: disabled)\n'\ ' -skiphtml Run the test and capture the trace logs, but skip the timeline (default: disabled)\n'\ ' -result fn Export a results table to a text file for parsing.\n'\ + ' -wifi If a wifi connection is available, check that it reconnects after resume.\n'\ + ' -wifitrace Trace kernel execution through wifi reconnect.\n'\ + ' -netfix Use netfix to reset the network in the event it fails to resume.\n'\ ' [testprep]\n'\ ' -sync Sync the filesystems before starting the test\n'\ ' -rs on/off Enable/disable runtime suspend for all devices, restore all after test\n'\ @@ -6356,8 +6707,10 @@ def printHelp(): ' -predelay t Include t ms delay before 1st suspend (default: 0 ms)\n'\ ' -postdelay t Include t ms delay after last resume (default: 0 ms)\n'\ ' -mindev ms Discard all device blocks shorter than ms milliseconds (e.g. 0.001 for us)\n'\ - ' -multi n d Execute <n> consecutive tests at <d> seconds intervals. The outputs will\n'\ - ' be created in a new subdirectory with a summary page.\n'\ + ' -multi n d Execute <n> consecutive tests at <d> seconds intervals. If <n> is followed\n'\ + ' by a "d", "h", or "m" execute for <n> days, hours, or mins instead.\n'\ + ' The outputs will be created in a new subdirectory with a summary page.\n'\ + ' -maxfail n Abort a -multi run after n consecutive fails (default is 0 = never abort)\n'\ ' [debug]\n'\ ' -f Use ftrace to create device callgraphs (default: disabled)\n'\ ' -ftop Use ftrace on the top level call: "%s" (default: disabled)\n'\ @@ -6379,11 +6732,11 @@ def printHelp(): ' -modes List available suspend modes\n'\ ' -status Test to see if the system is enabled to run this tool\n'\ ' -fpdt Print out the contents of the ACPI Firmware Performance Data Table\n'\ - ' -battery Print out battery info (if available)\n'\ - ' -wifi Print out wifi connection info (if wireless-tools and device exists)\n'\ + ' -wificheck Print out wifi connection info\n'\ ' -x<mode> Test xset by toggling the given mode (on/off/standby/suspend)\n'\ ' -sysinfo Print out system info extracted from BIOS\n'\ ' -devinfo Print out the pm settings of all devices which support runtime suspend\n'\ + ' -cmdinfo Print out all the platform info collected before and after suspend/resume\n'\ ' -flist Print the list of functions currently being captured in ftrace\n'\ ' -flistall Print all functions capable of being captured in ftrace\n'\ ' -summary dir Create a summary of tests in this dir [-genhtml builds missing html]\n'\ @@ -6399,8 +6752,8 @@ if __name__ == '__main__': genhtml = False cmd = '' simplecmds = ['-sysinfo', '-modes', '-fpdt', '-flist', '-flistall', - '-devinfo', '-status', '-battery', '-xon', '-xoff', '-xstandby', - '-xsuspend', '-xinit', '-xreset', '-xstat', '-wifi'] + '-devinfo', '-status', '-xon', '-xoff', '-xstandby', '-xsuspend', + '-xinit', '-xreset', '-xstat', '-wificheck', '-cmdinfo'] if '-f' in sys.argv: sysvals.cgskip = sysvals.configFile('cgskip.txt') # loop through the command line arguments @@ -6422,6 +6775,8 @@ if __name__ == '__main__': elif(arg == '-v'): pprint("Version %s" % sysvals.version) sys.exit(0) + elif(arg == '-debugtiming'): + debugtiming = True elif(arg == '-x2'): sysvals.execcount = 2 elif(arg == '-x2delay'): @@ -6462,8 +6817,24 @@ if __name__ == '__main__': sysvals.usedevsrc = True elif(arg == '-sync'): sysvals.sync = True + elif(arg == '-wifi'): + sysvals.wifi = True + elif(arg == '-wifitrace'): + sysvals.wifitrace = True + elif(arg == '-netfix'): + sysvals.netfix = True elif(arg == '-gzip'): sysvals.gzip = True + elif(arg == '-info'): + try: + val = next(args) + except: + doError('-info requires one string argument', True) + elif(arg == '-desc'): + try: + val = next(args) + except: + doError('-desc requires one string argument', True) elif(arg == '-rs'): try: val = next(args) @@ -6555,10 +6926,14 @@ if __name__ == '__main__': sysvals.cgexp = True elif(arg == '-srgap'): sysvals.srgap = 5 + elif(arg == '-maxfail'): + sysvals.maxfail = getArgInt('-maxfail', args, 0, 1000000) elif(arg == '-multi'): - sysvals.multitest['run'] = True - sysvals.multitest['count'] = getArgInt('-multi n d (exec count)', args, 2, 1000000) - sysvals.multitest['delay'] = getArgInt('-multi n d (delay between tests)', args, 0, 3600) + try: + c, d = next(args), next(args) + except: + doError('-multi requires two values', True) + sysvals.multiinit(c, d) elif(arg == '-o'): try: val = next(args) @@ -6655,13 +7030,6 @@ if __name__ == '__main__': elif(cmd == 'fpdt'): if not getFPDT(True): ret = 1 - elif(cmd == 'battery'): - out = getBattery() - if out: - pprint('AC Connect : %s\nBattery Charge: %d' % out) - else: - pprint('no battery found') - ret = 1 elif(cmd == 'sysinfo'): sysvals.printSystemInfo(True) elif(cmd == 'devinfo'): @@ -6676,16 +7044,18 @@ if __name__ == '__main__': runSummary(sysvals.outdir, True, genhtml) elif(cmd in ['xon', 'xoff', 'xstandby', 'xsuspend', 'xinit', 'xreset']): sysvals.verbose = True - ret = displayControl(cmd[1:]) + ret = sysvals.displayControl(cmd[1:]) elif(cmd == 'xstat'): - pprint('Display Status: %s' % displayControl('stat').upper()) - elif(cmd == 'wifi'): - out = sysvals.checkWifi() - if 'device' not in out: - pprint('WIFI interface not found') + pprint('Display Status: %s' % sysvals.displayControl('stat').upper()) + elif(cmd == 'wificheck'): + dev = sysvals.checkWifi() + if dev: + print('%s is connected' % sysvals.wifiDetails(dev)) else: - for key in sorted(out): - pprint('%6s: %s' % (key.upper(), out[key])) + print('No wifi connection found') + elif(cmd == 'cmdinfo'): + for out in sysvals.cmdinfo(False, True): + print('[%s - %s]\n%s\n' % out) sys.exit(ret) # if instructed, re-analyze existing data files @@ -6714,30 +7084,39 @@ if __name__ == '__main__': if mode.startswith('disk-'): sysvals.diskmode = mode.split('-', 1)[-1] sysvals.suspendmode = 'disk' - sysvals.systemInfo(dmidecode(sysvals.mempath)) - setRuntimeSuspend(True) - if sysvals.display: - displayControl('init') - ret = 0 + failcnt, ret = 0, 0 if sysvals.multitest['run']: # run multiple tests in a separate subdirectory if not sysvals.outdir: - s = 'suspend-x%d' % sysvals.multitest['count'] - sysvals.outdir = datetime.now().strftime(s+'-%y%m%d-%H%M%S') + if 'time' in sysvals.multitest: + s = '-%dm' % sysvals.multitest['time'] + else: + s = '-x%d' % sysvals.multitest['count'] + sysvals.outdir = datetime.now().strftime('suspend-%y%m%d-%H%M%S'+s) if not os.path.isdir(sysvals.outdir): os.makedirs(sysvals.outdir) + sysvals.sudoUserchown(sysvals.outdir) + finish = datetime.now() + if 'time' in sysvals.multitest: + finish += timedelta(minutes=sysvals.multitest['time']) for i in range(sysvals.multitest['count']): - if(i != 0): + sysvals.multistat(True, i, finish) + if i != 0 and sysvals.multitest['delay'] > 0: pprint('Waiting %d seconds...' % (sysvals.multitest['delay'])) time.sleep(sysvals.multitest['delay']) - pprint('TEST (%d/%d) START' % (i+1, sysvals.multitest['count'])) fmt = 'suspend-%y%m%d-%H%M%S' sysvals.testdir = os.path.join(sysvals.outdir, datetime.now().strftime(fmt)) - ret = runTest(i+1) - pprint('TEST (%d/%d) COMPLETE' % (i+1, sysvals.multitest['count'])) - sysvals.logmsg = '' + ret = runTest(i+1, not sysvals.verbose) + failcnt = 0 if not ret else failcnt + 1 + if sysvals.maxfail > 0 and failcnt >= sysvals.maxfail: + pprint('Maximum fail count of %d reached, aborting multitest' % (sysvals.maxfail)) + break + sysvals.resetlog() + sysvals.multistat(False, i, finish) + if 'time' in sysvals.multitest and datetime.now() >= finish: + break if not sysvals.skiphtml: runSummary(sysvals.outdir, False, False) sysvals.sudoUserchown(sysvals.outdir) @@ -6746,7 +7125,10 @@ if __name__ == '__main__': sysvals.testdir = sysvals.outdir # run the test in the current directory ret = runTest() + + # reset to default values after testing if sysvals.display: - displayControl('reset') - setRuntimeSuspend(False) + sysvals.displayControl('reset') + if sysvals.rs != 0: + sysvals.setRuntimeSuspend(False) sys.exit(ret) diff --git a/tools/power/x86/amd_pstate_tracer/amd_pstate_trace.py b/tools/power/x86/amd_pstate_tracer/amd_pstate_trace.py new file mode 100755 index 000000000000..2dea4032ac56 --- /dev/null +++ b/tools/power/x86/amd_pstate_tracer/amd_pstate_trace.py @@ -0,0 +1,354 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0-only +# -*- coding: utf-8 -*- +# +""" This utility can be used to debug and tune the performance of the +AMD P-State driver. It imports intel_pstate_tracer to analyze AMD P-State +trace event. + +Prerequisites: + Python version 2.7.x or higher + gnuplot 5.0 or higher + gnuplot-py 1.8 or higher + (Most of the distributions have these required packages. They may be called + gnuplot-py, phython-gnuplot or phython3-gnuplot, gnuplot-nox, ... ) + + Kernel config for Linux trace is enabled + + see print_help(): for Usage and Output details + +""" +from __future__ import print_function +from datetime import datetime +import subprocess +import os +import time +import re +import signal +import sys +import getopt +import Gnuplot +from numpy import * +from decimal import * +sys.path.append('../intel_pstate_tracer') +#import intel_pstate_tracer +import intel_pstate_tracer as ipt + +__license__ = "GPL version 2" + +MAX_CPUS = 256 +# Define the csv file columns +C_COMM = 15 +C_ELAPSED = 14 +C_SAMPLE = 13 +C_DURATION = 12 +C_LOAD = 11 +C_TSC = 10 +C_APERF = 9 +C_MPERF = 8 +C_FREQ = 7 +C_MAX_PERF = 6 +C_DES_PERF = 5 +C_MIN_PERF = 4 +C_USEC = 3 +C_SEC = 2 +C_CPU = 1 + +global sample_num, last_sec_cpu, last_usec_cpu, start_time, test_name, trace_file + +getcontext().prec = 11 + +sample_num =0 +last_sec_cpu = [0] * MAX_CPUS +last_usec_cpu = [0] * MAX_CPUS + +def plot_per_cpu_freq(cpu_index): + """ Plot per cpu frequency """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_frequency.png" % cpu_index + g_plot = ipt.common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + g_plot('set yrange [0:7]') + g_plot('set ytics 0, 1') + g_plot('set ylabel "CPU Frequency (GHz)"') + g_plot('set title "{} : frequency : CPU {:0>3} : {:%F %H:%M}"'.format(test_name, cpu_index, datetime.now())) + g_plot('set ylabel "CPU frequency"') + g_plot('set key off') + ipt.set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y1'.format(C_ELAPSED, C_FREQ)) + +def plot_per_cpu_des_perf(cpu_index): + """ Plot per cpu desired perf """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_des_perf.png" % cpu_index + g_plot = ipt.common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + g_plot('set yrange [0:255]') + g_plot('set ylabel "des perf"') + g_plot('set title "{} : cpu des perf : CPU {:0>3} : {:%F %H:%M}"'.format(test_name, cpu_index, datetime.now())) + g_plot('set key off') + ipt.set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y1'.format(C_ELAPSED, C_DES_PERF)) + +def plot_per_cpu_load(cpu_index): + """ Plot per cpu load """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_load.png" % cpu_index + g_plot = ipt.common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + g_plot('set yrange [0:100]') + g_plot('set ytics 0, 10') + g_plot('set ylabel "CPU load (percent)"') + g_plot('set title "{} : cpu load : CPU {:0>3} : {:%F %H:%M}"'.format(test_name, cpu_index, datetime.now())) + g_plot('set key off') + ipt.set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y1'.format(C_ELAPSED, C_LOAD)) + +def plot_all_cpu_frequency(): + """ Plot all cpu frequencies """ + + output_png = 'all_cpu_frequencies.png' + g_plot = ipt.common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + g_plot('set ylabel "CPU Frequency (GHz)"') + g_plot('set title "{} : cpu frequencies : {:%F %H:%M}"'.format(test_name, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).decode('utf-8').replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_FREQ) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_all_cpu_des_perf(): + """ Plot all cpu desired perf """ + + output_png = 'all_cpu_des_perf.png' + g_plot = ipt.common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + g_plot('set ylabel "des perf"') + g_plot('set title "{} : cpu des perf : {:%F %H:%M}"'.format(test_name, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).decode('utf-8').replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 255 ps 1 title i".format(C_ELAPSED, C_DES_PERF) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_all_cpu_load(): + """ Plot all cpu load """ + + output_png = 'all_cpu_load.png' + g_plot = ipt.common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + g_plot('set yrange [0:100]') + g_plot('set ylabel "CPU load (percent)"') + g_plot('set title "{} : cpu load : {:%F %H:%M}"'.format(test_name, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).decode('utf-8').replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 255 ps 1 title i".format(C_ELAPSED, C_LOAD) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def store_csv(cpu_int, time_pre_dec, time_post_dec, min_perf, des_perf, max_perf, freq_ghz, mperf, aperf, tsc, common_comm, load, duration_ms, sample_num, elapsed_time, cpu_mask): + """ Store master csv file information """ + + global graph_data_present + + if cpu_mask[cpu_int] == 0: + return + + try: + f_handle = open('cpu.csv', 'a') + string_buffer = "CPU_%03u, %05u, %06u, %u, %u, %u, %.4f, %u, %u, %u, %.2f, %.3f, %u, %.3f, %s\n" % (cpu_int, int(time_pre_dec), int(time_post_dec), int(min_perf), int(des_perf), int(max_perf), freq_ghz, int(mperf), int(aperf), int(tsc), load, duration_ms, sample_num, elapsed_time, common_comm) + f_handle.write(string_buffer) + f_handle.close() + except: + print('IO error cpu.csv') + return + + graph_data_present = True; + + +def cleanup_data_files(): + """ clean up existing data files """ + + if os.path.exists('cpu.csv'): + os.remove('cpu.csv') + f_handle = open('cpu.csv', 'a') + f_handle.write('common_cpu, common_secs, common_usecs, min_perf, des_perf, max_perf, freq, mperf, aperf, tsc, load, duration_ms, sample_num, elapsed_time, common_comm') + f_handle.write('\n') + f_handle.close() + +def read_trace_data(file_name, cpu_mask): + """ Read and parse trace data """ + + global current_max_cpu + global sample_num, last_sec_cpu, last_usec_cpu, start_time + + try: + data = open(file_name, 'r').read() + except: + print('Error opening ', file_name) + sys.exit(2) + + for line in data.splitlines(): + search_obj = \ + re.search(r'(^(.*?)\[)((\d+)[^\]])(.*?)(\d+)([.])(\d+)(.*?amd_min_perf=)(\d+)(.*?amd_des_perf=)(\d+)(.*?amd_max_perf=)(\d+)(.*?freq=)(\d+)(.*?mperf=)(\d+)(.*?aperf=)(\d+)(.*?tsc=)(\d+)' + , line) + + if search_obj: + cpu = search_obj.group(3) + cpu_int = int(cpu) + cpu = str(cpu_int) + + time_pre_dec = search_obj.group(6) + time_post_dec = search_obj.group(8) + min_perf = search_obj.group(10) + des_perf = search_obj.group(12) + max_perf = search_obj.group(14) + freq = search_obj.group(16) + mperf = search_obj.group(18) + aperf = search_obj.group(20) + tsc = search_obj.group(22) + + common_comm = search_obj.group(2).replace(' ', '') + + if sample_num == 0 : + start_time = Decimal(time_pre_dec) + Decimal(time_post_dec) / Decimal(1000000) + sample_num += 1 + + if last_sec_cpu[cpu_int] == 0 : + last_sec_cpu[cpu_int] = time_pre_dec + last_usec_cpu[cpu_int] = time_post_dec + else : + duration_us = (int(time_pre_dec) - int(last_sec_cpu[cpu_int])) * 1000000 + (int(time_post_dec) - int(last_usec_cpu[cpu_int])) + duration_ms = Decimal(duration_us) / Decimal(1000) + last_sec_cpu[cpu_int] = time_pre_dec + last_usec_cpu[cpu_int] = time_post_dec + elapsed_time = Decimal(time_pre_dec) + Decimal(time_post_dec) / Decimal(1000000) - start_time + load = Decimal(int(mperf)*100)/ Decimal(tsc) + freq_ghz = Decimal(freq)/Decimal(1000000) + store_csv(cpu_int, time_pre_dec, time_post_dec, min_perf, des_perf, max_perf, freq_ghz, mperf, aperf, tsc, common_comm, load, duration_ms, sample_num, elapsed_time, cpu_mask) + + if cpu_int > current_max_cpu: + current_max_cpu = cpu_int +# Now separate the main overall csv file into per CPU csv files. + ipt.split_csv(current_max_cpu, cpu_mask) + + +def signal_handler(signal, frame): + print(' SIGINT: Forcing cleanup before exit.') + if interval: + ipt.disable_trace(trace_file) + ipt.clear_trace_file() + ipt.free_trace_buffer() + sys.exit(0) + +trace_file = "/sys/kernel/debug/tracing/events/amd_cpu/enable" +signal.signal(signal.SIGINT, signal_handler) + +interval = "" +file_name = "" +cpu_list = "" +test_name = "" +memory = "10240" +graph_data_present = False; + +valid1 = False +valid2 = False + +cpu_mask = zeros((MAX_CPUS,), dtype=int) + + +try: + opts, args = getopt.getopt(sys.argv[1:],"ht:i:c:n:m:",["help","trace_file=","interval=","cpu=","name=","memory="]) +except getopt.GetoptError: + ipt.print_help('amd_pstate') + sys.exit(2) +for opt, arg in opts: + if opt == '-h': + print() + sys.exit() + elif opt in ("-t", "--trace_file"): + valid1 = True + location = os.path.realpath(os.path.join(os.getcwd(), os.path.dirname(__file__))) + file_name = os.path.join(location, arg) + elif opt in ("-i", "--interval"): + valid1 = True + interval = arg + elif opt in ("-c", "--cpu"): + cpu_list = arg + elif opt in ("-n", "--name"): + valid2 = True + test_name = arg + elif opt in ("-m", "--memory"): + memory = arg + +if not (valid1 and valid2): + ipt.print_help('amd_pstate') + sys.exit() + +if cpu_list: + for p in re.split("[,]", cpu_list): + if int(p) < MAX_CPUS : + cpu_mask[int(p)] = 1 +else: + for i in range (0, MAX_CPUS): + cpu_mask[i] = 1 + +if not os.path.exists('results'): + os.mkdir('results') + ipt.fix_ownership('results') + +os.chdir('results') +if os.path.exists(test_name): + print('The test name directory already exists. Please provide a unique test name. Test re-run not supported, yet.') + sys.exit() +os.mkdir(test_name) +ipt.fix_ownership(test_name) +os.chdir(test_name) + +cur_version = sys.version_info +print('python version (should be >= 2.7):') +print(cur_version) + +cleanup_data_files() + +if interval: + file_name = "/sys/kernel/debug/tracing/trace" + ipt.clear_trace_file() + ipt.set_trace_buffer_size(memory) + ipt.enable_trace(trace_file) + time.sleep(int(interval)) + ipt.disable_trace(trace_file) + +current_max_cpu = 0 + +read_trace_data(file_name, cpu_mask) + +if interval: + ipt.clear_trace_file() + ipt.free_trace_buffer() + +if graph_data_present == False: + print('No valid data to plot') + sys.exit(2) + +for cpu_no in range(0, current_max_cpu + 1): + plot_per_cpu_freq(cpu_no) + plot_per_cpu_des_perf(cpu_no) + plot_per_cpu_load(cpu_no) + +plot_all_cpu_des_perf() +plot_all_cpu_frequency() +plot_all_cpu_load() + +for root, dirs, files in os.walk('.'): + for f in files: + ipt.fix_ownership(f) + +os.chdir('../../') diff --git a/tools/power/x86/intel-speed-select/.gitignore b/tools/power/x86/intel-speed-select/.gitignore index f61145925ce9..a814f89fe75f 100644 --- a/tools/power/x86/intel-speed-select/.gitignore +++ b/tools/power/x86/intel-speed-select/.gitignore @@ -1,2 +1,3 @@ +# SPDX-License-Identifier: GPL-2.0-only include/ intel-speed-select diff --git a/tools/power/x86/intel-speed-select/Build b/tools/power/x86/intel-speed-select/Build index b61456d75190..81e36bd578b1 100644 --- a/tools/power/x86/intel-speed-select/Build +++ b/tools/power/x86/intel-speed-select/Build @@ -1 +1 @@ -intel-speed-select-y += isst-config.o isst-core.o isst-display.o +intel-speed-select-y += isst-config.o isst-core.o isst-display.o isst-daemon.o hfi-events.o diff --git a/tools/power/x86/intel-speed-select/Makefile b/tools/power/x86/intel-speed-select/Makefile index 12c6939dca2a..7221f2f55e8b 100644 --- a/tools/power/x86/intel-speed-select/Makefile +++ b/tools/power/x86/intel-speed-select/Makefile @@ -13,8 +13,8 @@ endif # Do not use make's built-in rules # (this improves performance and avoids hard-to-debug behaviour); MAKEFLAGS += -r - -override CFLAGS += -O2 -Wall -g -D_GNU_SOURCE -I$(OUTPUT)include +override CFLAGS += -O2 -Wall -g -D_GNU_SOURCE -I$(OUTPUT)include -I/usr/include/libnl3 +override LDFLAGS += -lnl-genl-3 -lnl-3 ALL_TARGETS := intel-speed-select ALL_PROGRAMS := $(patsubst %,$(OUTPUT)%,$(ALL_TARGETS)) @@ -31,19 +31,23 @@ $(OUTPUT)include/linux/isst_if.h: ../../../../include/uapi/linux/isst_if.h mkdir -p $(OUTPUT)include/linux 2>&1 || true ln -sf $(CURDIR)/../../../../include/uapi/linux/isst_if.h $@ -prepare: $(OUTPUT)include/linux/isst_if.h +$(OUTPUT)include/linux/thermal.h: ../../../../include/uapi/linux/thermal.h + mkdir -p $(OUTPUT)include/linux 2>&1 || true + ln -sf $(CURDIR)/../../../../include/uapi/linux/thermal.h $@ + +prepare: $(OUTPUT)include/linux/isst_if.h $(OUTPUT)include/linux/thermal.h ISST_IN := $(OUTPUT)intel-speed-select-in.o $(ISST_IN): prepare FORCE $(Q)$(MAKE) $(build)=intel-speed-select $(OUTPUT)intel-speed-select: $(ISST_IN) - $(QUIET_LINK)$(CC) $(CFLAGS) $(LDFLAGS) $< -o $@ + $(QUIET_LINK)$(CC) $(CFLAGS) $< $(LDFLAGS) -o $@ clean: rm -f $(ALL_PROGRAMS) rm -rf $(OUTPUT)include/linux/isst_if.h - find $(if $(OUTPUT),$(OUTPUT),.) -name '*.o' -delete -o -name '\.*.d' -delete + find $(or $(OUTPUT),.) -name '*.o' -delete -o -name '\.*.d' -delete install: $(ALL_PROGRAMS) install -d -m 755 $(DESTDIR)$(bindir); \ diff --git a/tools/power/x86/intel-speed-select/hfi-events.c b/tools/power/x86/intel-speed-select/hfi-events.c new file mode 100644 index 000000000000..be96e90cc2a1 --- /dev/null +++ b/tools/power/x86/intel-speed-select/hfi-events.c @@ -0,0 +1,312 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Speed Select -- Read HFI events for OOB + * Copyright (c) 2022 Intel Corporation. + */ + +/* + * This file incorporates work covered by the following copyright and + * permission notice: + + * WPA Supplicant - driver interaction with Linux nl80211/cfg80211 + * Copyright (c) 2003-2008, Jouni Malinen <j@w1.fi> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Alternatively, this software may be distributed under the terms of + * BSD license. + * + * Requires + * libnl-genl-3-dev + * + * For Fedora/CenOS + * dnf install libnl3-devel + * For Ubuntu + * apt install libnl-3-dev libnl-genl-3-dev + */ + +#include <stdio.h> +#include <stdlib.h> +#include <stdarg.h> +#include <string.h> +#include <unistd.h> +#include <fcntl.h> +#include <sys/file.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <errno.h> +#include <getopt.h> +#include <signal.h> +#include <netlink/genl/genl.h> +#include <netlink/genl/family.h> +#include <netlink/genl/ctrl.h> + +#include <linux/thermal.h> +#include "isst.h" + +struct hfi_event_data { + struct nl_sock *nl_handle; + struct nl_cb *nl_cb; +}; + +struct hfi_event_data drv; + +static int ack_handler(struct nl_msg *msg, void *arg) +{ + int *err = arg; + *err = 0; + return NL_STOP; +} + +static int finish_handler(struct nl_msg *msg, void *arg) +{ + int *ret = arg; + *ret = 0; + return NL_SKIP; +} + +static int error_handler(struct sockaddr_nl *nla, struct nlmsgerr *err, + void *arg) +{ + int *ret = arg; + *ret = err->error; + return NL_SKIP; +} + +static int seq_check_handler(struct nl_msg *msg, void *arg) +{ + return NL_OK; +} + +static int send_and_recv_msgs(struct hfi_event_data *drv, + struct nl_msg *msg, + int (*valid_handler)(struct nl_msg *, void *), + void *valid_data) +{ + struct nl_cb *cb; + int err = -ENOMEM; + + cb = nl_cb_clone(drv->nl_cb); + if (!cb) + goto out; + + err = nl_send_auto_complete(drv->nl_handle, msg); + if (err < 0) + goto out; + + err = 1; + + nl_cb_err(cb, NL_CB_CUSTOM, error_handler, &err); + nl_cb_set(cb, NL_CB_FINISH, NL_CB_CUSTOM, finish_handler, &err); + nl_cb_set(cb, NL_CB_ACK, NL_CB_CUSTOM, ack_handler, &err); + + if (valid_handler) + nl_cb_set(cb, NL_CB_VALID, NL_CB_CUSTOM, + valid_handler, valid_data); + + while (err > 0) + nl_recvmsgs(drv->nl_handle, cb); + out: + nl_cb_put(cb); + nlmsg_free(msg); + return err; +} + +struct family_data { + const char *group; + int id; +}; + +static int family_handler(struct nl_msg *msg, void *arg) +{ + struct family_data *res = arg; + struct nlattr *tb[CTRL_ATTR_MAX + 1]; + struct genlmsghdr *gnlh = nlmsg_data(nlmsg_hdr(msg)); + struct nlattr *mcgrp; + int i; + + nla_parse(tb, CTRL_ATTR_MAX, genlmsg_attrdata(gnlh, 0), + genlmsg_attrlen(gnlh, 0), NULL); + if (!tb[CTRL_ATTR_MCAST_GROUPS]) + return NL_SKIP; + + nla_for_each_nested(mcgrp, tb[CTRL_ATTR_MCAST_GROUPS], i) { + struct nlattr *tb2[CTRL_ATTR_MCAST_GRP_MAX + 1]; + nla_parse(tb2, CTRL_ATTR_MCAST_GRP_MAX, nla_data(mcgrp), + nla_len(mcgrp), NULL); + if (!tb2[CTRL_ATTR_MCAST_GRP_NAME] || + !tb2[CTRL_ATTR_MCAST_GRP_ID] || + strncmp(nla_data(tb2[CTRL_ATTR_MCAST_GRP_NAME]), + res->group, + nla_len(tb2[CTRL_ATTR_MCAST_GRP_NAME])) != 0) + continue; + res->id = nla_get_u32(tb2[CTRL_ATTR_MCAST_GRP_ID]); + break; + } + + return 0; +} + +static int nl_get_multicast_id(struct hfi_event_data *drv, + const char *family, const char *group) +{ + struct nl_msg *msg; + int ret = -1; + struct family_data res = { group, -ENOENT }; + + msg = nlmsg_alloc(); + if (!msg) + return -ENOMEM; + genlmsg_put(msg, 0, 0, genl_ctrl_resolve(drv->nl_handle, "nlctrl"), + 0, 0, CTRL_CMD_GETFAMILY, 0); + NLA_PUT_STRING(msg, CTRL_ATTR_FAMILY_NAME, family); + + ret = send_and_recv_msgs(drv, msg, family_handler, &res); + msg = NULL; + if (ret == 0) + ret = res.id; + +nla_put_failure: + nlmsg_free(msg); + return ret; +} + +struct perf_cap { + int cpu; + int perf; + int eff; +}; + +static void process_hfi_event(struct perf_cap *perf_cap) +{ + struct isst_id id; + + set_isst_id(&id, perf_cap->cpu); + process_level_change(&id); +} + +static int handle_event(struct nl_msg *n, void *arg) +{ + struct nlmsghdr *nlh = nlmsg_hdr(n); + struct genlmsghdr *genlhdr = genlmsg_hdr(nlh); + struct nlattr *attrs[THERMAL_GENL_ATTR_MAX + 1]; + int ret; + struct perf_cap perf_cap = {0}; + + ret = genlmsg_parse(nlh, 0, attrs, THERMAL_GENL_ATTR_MAX, NULL); + + debug_printf("Received event %d parse_rer:%d\n", genlhdr->cmd, ret); + if (genlhdr->cmd == THERMAL_GENL_EVENT_CPU_CAPABILITY_CHANGE) { + struct nlattr *cap; + int j, index = 0; + + debug_printf("THERMAL_GENL_EVENT_CPU_CAPABILITY_CHANGE\n"); + nla_for_each_nested(cap, attrs[THERMAL_GENL_ATTR_CPU_CAPABILITY], j) { + switch (index) { + case 0: + perf_cap.cpu = nla_get_u32(cap); + break; + case 1: + perf_cap.perf = nla_get_u32(cap); + break; + case 2: + perf_cap.eff = nla_get_u32(cap); + break; + default: + break; + } + ++index; + if (index == 3) { + index = 0; + process_hfi_event(&perf_cap); + } + } + } + + return 0; +} + +static int _hfi_exit; + +static int check_hf_suport(void) +{ + unsigned int eax = 0, ebx = 0, ecx = 0, edx = 0; + + __cpuid(6, eax, ebx, ecx, edx); + if (eax & BIT(19)) + return 1; + + return 0; +} + +int hfi_main(void) +{ + struct nl_sock *sock; + struct nl_cb *cb; + int err = 0; + int mcast_id; + int no_block = 0; + + if (!check_hf_suport()) { + fprintf(stderr, "CPU Doesn't support HFI\n"); + return -1; + } + + sock = nl_socket_alloc(); + if (!sock) { + fprintf(stderr, "nl_socket_alloc failed\n"); + return -1; + } + + if (genl_connect(sock)) { + fprintf(stderr, "genl_connect(sk_event) failed\n"); + goto free_sock; + } + + drv.nl_handle = sock; + drv.nl_cb = cb = nl_cb_alloc(NL_CB_DEFAULT); + if (drv.nl_cb == NULL) { + printf("Failed to allocate netlink callbacks"); + goto free_sock; + } + + mcast_id = nl_get_multicast_id(&drv, THERMAL_GENL_FAMILY_NAME, + THERMAL_GENL_EVENT_GROUP_NAME); + if (mcast_id < 0) { + fprintf(stderr, "nl_get_multicast_id failed\n"); + goto free_sock; + } + + if (nl_socket_add_membership(sock, mcast_id)) { + fprintf(stderr, "nl_socket_add_membership failed"); + goto free_sock; + } + + nl_cb_set(cb, NL_CB_SEQ_CHECK, NL_CB_CUSTOM, seq_check_handler, 0); + nl_cb_set(cb, NL_CB_VALID, NL_CB_CUSTOM, handle_event, NULL); + + if (no_block) + nl_socket_set_nonblocking(sock); + + debug_printf("hfi is initialized\n"); + + while (!_hfi_exit && !err) { + err = nl_recvmsgs(sock, cb); + debug_printf("nl_recv_message err:%d\n", err); + } + + return 0; + + /* Netlink library doesn't have calls to dealloc cb or disconnect */ +free_sock: + nl_socket_free(sock); + + return -1; +} + +void hfi_exit(void) +{ + _hfi_exit = 1; +} diff --git a/tools/power/x86/intel-speed-select/isst-config.c b/tools/power/x86/intel-speed-select/isst-config.c index 2b2b8167c65b..a160bad291eb 100644 --- a/tools/power/x86/intel-speed-select/isst-config.c +++ b/tools/power/x86/intel-speed-select/isst-config.c @@ -15,7 +15,8 @@ struct process_cmd_struct { int arg; }; -static const char *version_str = "v1.2"; +static const char *version_str = "v1.13"; + static const int supported_api_ver = 1; static struct isst_if_platform_info isst_platform_info; static char *progname; @@ -25,7 +26,7 @@ static FILE *outf; static int cpu_model; static int cpu_stepping; -#define MAX_CPUS_IN_ONE_REQ 64 +#define MAX_CPUS_IN_ONE_REQ 256 static short max_target_cpus; static unsigned short target_cpus[MAX_CPUS_IN_ONE_REQ]; @@ -42,6 +43,10 @@ static int out_format_json; static int cmd_help; static int force_online_offline; static int auto_mode; +static int fact_enable_fail; + +static int mbox_delay; +static int mbox_retries = 3; /* clos related */ static int current_clos = -1; @@ -58,9 +63,22 @@ struct _cpu_map { unsigned short die_id; unsigned short punit_cpu; unsigned short punit_cpu_core; + unsigned short initialized; }; struct _cpu_map *cpu_map; +struct cpu_topology { + short cpu; + short core_id; + short pkg_id; + short die_id; +}; + +FILE *get_output_file(void) +{ + return outf; +} + void debug_printf(const char *format, ...) { va_list args; @@ -82,6 +100,30 @@ int is_clx_n_platform(void) return 0; } +int is_skx_based_platform(void) +{ + if (cpu_model == 0x55) + return 1; + + return 0; +} + +int is_spr_platform(void) +{ + if (cpu_model == 0x8F) + return 1; + + return 0; +} + +int is_icx_platform(void) +{ + if (cpu_model == 0x6A || cpu_model == 0x6C) + return 1; + + return 0; +} + static int update_cpu_model(void) { unsigned int ebx, ecx, edx; @@ -175,31 +217,192 @@ int out_format_is_json(void) return out_format_json; } -int get_physical_package_id(int cpu) +static int get_stored_topology_info(int cpu, int *core_id, int *pkg_id, int *die_id) +{ + const char *pathname = "/var/run/isst_cpu_topology.dat"; + struct cpu_topology cpu_top; + FILE *fp; + int ret; + + fp = fopen(pathname, "rb"); + if (!fp) + return -1; + + ret = fseek(fp, cpu * sizeof(cpu_top), SEEK_SET); + if (ret) + goto err_ret; + + ret = fread(&cpu_top, sizeof(cpu_top), 1, fp); + if (ret != 1) { + ret = -1; + goto err_ret; + } + + *pkg_id = cpu_top.pkg_id; + *core_id = cpu_top.core_id; + *die_id = cpu_top.die_id; + ret = 0; + +err_ret: + fclose(fp); + + return ret; +} + +static void store_cpu_topology(void) +{ + const char *pathname = "/var/run/isst_cpu_topology.dat"; + FILE *fp; + int i; + + fp = fopen(pathname, "rb"); + if (fp) { + /* Mapping already exists */ + fclose(fp); + return; + } + + fp = fopen(pathname, "wb"); + if (!fp) { + fprintf(stderr, "Can't create file:%s\n", pathname); + return; + } + + fprintf(stderr, "Caching topology information\n"); + + for (i = 0; i < topo_max_cpus; ++i) { + struct cpu_topology cpu_top; + + cpu_top.core_id = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/core_id", i); + if (cpu_top.core_id < 0) + cpu_top.core_id = -1; + + cpu_top.pkg_id = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/physical_package_id", i); + if (cpu_top.pkg_id < 0) + cpu_top.pkg_id = -1; + + cpu_top.die_id = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/die_id", i); + if (cpu_top.die_id < 0) + cpu_top.die_id = -1; + + cpu_top.cpu = i; + + if (fwrite(&cpu_top, sizeof(cpu_top), 1, fp) != 1) { + fprintf(stderr, "Can't write to:%s\n", pathname); + break; + } + } + + fclose(fp); +} + +static int get_physical_package_id(int cpu) { - return parse_int_file( - 0, "/sys/devices/system/cpu/cpu%d/topology/physical_package_id", - cpu); + int ret; + + if (cpu < 0) + return -1; + + if (cpu_map && cpu_map[cpu].initialized) + return cpu_map[cpu].pkg_id; + + ret = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/physical_package_id", + cpu); + if (ret < 0) { + int core_id, pkg_id, die_id; + + ret = get_stored_topology_info(cpu, &core_id, &pkg_id, &die_id); + if (!ret) + return pkg_id; + } + + return ret; } -int get_physical_core_id(int cpu) +static int get_physical_core_id(int cpu) { - return parse_int_file( - 0, "/sys/devices/system/cpu/cpu%d/topology/core_id", cpu); + int ret; + + if (cpu < 0) + return -1; + + if (cpu_map && cpu_map[cpu].initialized) + return cpu_map[cpu].core_id; + + ret = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/core_id", + cpu); + if (ret < 0) { + int core_id, pkg_id, die_id; + + ret = get_stored_topology_info(cpu, &core_id, &pkg_id, &die_id); + if (!ret) + return core_id; + } + + return ret; } -int get_physical_die_id(int cpu) +static int get_physical_die_id(int cpu) { int ret; - ret = parse_int_file(0, "/sys/devices/system/cpu/cpu%d/topology/die_id", - cpu); + if (cpu < 0) + return -1; + + if (cpu_map && cpu_map[cpu].initialized) + return cpu_map[cpu].die_id; + + ret = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/die_id", + cpu); + if (ret < 0) { + int core_id, pkg_id, die_id; + + ret = get_stored_topology_info(cpu, &core_id, &pkg_id, &die_id); + if (!ret) { + if (die_id < 0) + die_id = 0; + + return die_id; + } + } + if (ret < 0) ret = 0; return ret; } +void set_isst_id(struct isst_id *id, int cpu) +{ + id->cpu = cpu; + + id->pkg = get_physical_package_id(cpu); + if (id < 0 || id->pkg >= MAX_PACKAGE_COUNT) + id->pkg = -1; + + id->die = get_physical_die_id(cpu); + if (id < 0 || id->die >= MAX_DIE_PER_PACKAGE) + id->die = -1; +} + +int is_cpu_in_power_domain(int cpu, struct isst_id *id) +{ + struct isst_id tid; + + set_isst_id(&tid, cpu); + + if (id->pkg == tid.pkg && id->die == tid.die) + return 1; + + return 0; +} + int get_cpufreq_base_freq(int cpu) { return parse_int_file(0, "/sys/devices/system/cpu/cpu%d/cpufreq/base_frequency", cpu); @@ -210,7 +413,7 @@ int get_topo_max_cpus(void) return topo_max_cpus; } -static void set_cpu_online_offline(int cpu, int state) +void set_cpu_online_offline(int cpu, int state) { char buffer[128]; int fd, ret; @@ -219,8 +422,14 @@ static void set_cpu_online_offline(int cpu, int state) "/sys/devices/system/cpu/cpu%d/online", cpu); fd = open(buffer, O_WRONLY); - if (fd < 0) + if (fd < 0) { + if (!cpu && state) { + fprintf(stderr, "This system is not configured for CPU 0 online/offline\n"); + fprintf(stderr, "Ignoring online request for CPU 0 as this is already online\n"); + return; + } err(-1, "%s open failed", buffer); + } if (state) ret = write(fd, "1\n", 2); @@ -233,15 +442,26 @@ static void set_cpu_online_offline(int cpu, int state) close(fd); } -#define MAX_PACKAGE_COUNT 8 -#define MAX_DIE_PER_PACKAGE 2 -static void for_each_online_package_in_set(void (*callback)(int, void *, void *, - void *, void *), - void *arg1, void *arg2, void *arg3, - void *arg4) +static void force_all_cpus_online(void) +{ + int i; + + fprintf(stderr, "Forcing all CPUs online\n"); + + for (i = 0; i < topo_max_cpus; ++i) + set_cpu_online_offline(i, 1); + + unlink("/var/run/isst_cpu_topology.dat"); +} + +void for_each_online_package_in_set(void (*callback)(struct isst_id *, void *, void *, + void *, void *), + void *arg1, void *arg2, void *arg3, + void *arg4) { int max_packages[MAX_PACKAGE_COUNT * MAX_PACKAGE_COUNT]; int pkg_index = 0, i; + struct isst_id id; memset(max_packages, 0xff, sizeof(max_packages)); for (i = 0; i < topo_max_cpus; ++i) { @@ -259,7 +479,12 @@ static void for_each_online_package_in_set(void (*callback)(int, void *, void *, die_id = get_physical_die_id(i); if (die_id < 0) die_id = 0; - pkg_id = get_physical_package_id(i); + + pkg_id = parse_int_file(0, + "/sys/devices/system/cpu/cpu%d/topology/physical_package_id", i); + if (pkg_id < 0) + continue; + /* Create an unique id for package, die combination to store */ pkg_id = (MAX_PACKAGE_COUNT * pkg_id + die_id); @@ -270,18 +495,20 @@ static void for_each_online_package_in_set(void (*callback)(int, void *, void *, } } + set_isst_id(&id, i); if (!skip && online && callback) { - callback(i, arg1, arg2, arg3, arg4); + callback(&id, arg1, arg2, arg3, arg4); max_packages[pkg_index++] = pkg_id; } } } static void for_each_online_target_cpu_in_set( - void (*callback)(int, void *, void *, void *, void *), void *arg1, + void (*callback)(struct isst_id *, void *, void *, void *, void *), void *arg1, void *arg2, void *arg3, void *arg4) { - int i; + int i, found = 0; + struct isst_id id; for (i = 0; i < topo_max_cpus; ++i) { int online; @@ -295,9 +522,15 @@ static void for_each_online_target_cpu_in_set( online = 1; /* online entry for CPU 0 needs some special configs */ - if (online && callback) - callback(i, arg1, arg2, arg3, arg4); + set_isst_id(&id, i); + if (online && callback) { + callback(&id, arg1, arg2, arg3, arg4); + found = 1; + } } + + if (!found) + fprintf(stderr, "No valid CPU in the list\n"); } #define BITMASK_SIZE 32 @@ -305,14 +538,27 @@ static void set_max_cpu_num(void) { FILE *filep; unsigned long dummy; + int i; topo_max_cpus = 0; - filep = fopen_or_exit( - "/sys/devices/system/cpu/cpu0/topology/thread_siblings", "r"); + for (i = 0; i < 256; ++i) { + char path[256]; + + snprintf(path, sizeof(path), + "/sys/devices/system/cpu/cpu%d/topology/thread_siblings", i); + filep = fopen(path, "r"); + if (filep) + break; + } + + if (!filep) { + fprintf(stderr, "Can't get max cpu number\n"); + exit(0); + } + while (fscanf(filep, "%lx,", &dummy) == 1) topo_max_cpus += BITMASK_SIZE; fclose(filep); - topo_max_cpus--; /* 0 based */ debug_printf("max cpus %d\n", topo_max_cpus); } @@ -338,103 +584,85 @@ void free_cpu_set(cpu_set_t *cpu_set) } static int cpu_cnt[MAX_PACKAGE_COUNT][MAX_DIE_PER_PACKAGE]; -static long long core_mask[MAX_PACKAGE_COUNT][MAX_DIE_PER_PACKAGE]; -static void set_cpu_present_cpu_mask(void) -{ - size_t size; - DIR *dir; - int i; - size = alloc_cpu_set(&present_cpumask); - present_cpumask_size = size; - for (i = 0; i < topo_max_cpus; ++i) { - char buffer[256]; - - snprintf(buffer, sizeof(buffer), - "/sys/devices/system/cpu/cpu%d", i); - dir = opendir(buffer); - if (dir) { - int pkg_id, die_id; - - CPU_SET_S(i, size, present_cpumask); - die_id = get_physical_die_id(i); - if (die_id < 0) - die_id = 0; - - pkg_id = get_physical_package_id(i); - if (pkg_id < MAX_PACKAGE_COUNT && - die_id < MAX_DIE_PER_PACKAGE) { - int core_id = get_physical_core_id(i); - - cpu_cnt[pkg_id][die_id]++; - core_mask[pkg_id][die_id] |= (1ULL << core_id); - } - } - closedir(dir); - } -} - -int get_core_count(int pkg_id, int die_id) +int get_max_punit_core_id(struct isst_id *id) { - int cnt = 0; + int max_id = 0; + int i; - if (pkg_id < MAX_PACKAGE_COUNT && die_id < MAX_DIE_PER_PACKAGE) { - int i; + for (i = 0; i < topo_max_cpus; ++i) + { + if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) + continue; - for (i = 0; i < sizeof(long long) * 8; ++i) { - if (core_mask[pkg_id][die_id] & (1ULL << i)) - cnt++; - } + if (is_cpu_in_power_domain(i, id) && + cpu_map[i].punit_cpu_core > max_id) + max_id = cpu_map[i].punit_cpu_core; } - return cnt; -} - -int get_cpu_count(int pkg_id, int die_id) -{ - if (pkg_id < MAX_PACKAGE_COUNT && die_id < MAX_DIE_PER_PACKAGE) - return cpu_cnt[pkg_id][die_id]; - - return 0; + return max_id; } -static void set_cpu_target_cpu_mask(void) +int get_cpu_count(struct isst_id *id) { - size_t size; - int i; - - size = alloc_cpu_set(&target_cpumask); - target_cpumask_size = size; - for (i = 0; i < max_target_cpus; ++i) { - if (!CPU_ISSET_S(target_cpus[i], present_cpumask_size, - present_cpumask)) - continue; + if (id->pkg < 0 || id->die < 0) + return 0; - CPU_SET_S(target_cpus[i], size, target_cpumask); - } + return cpu_cnt[id->pkg][id->die]; } static void create_cpu_map(void) { const char *pathname = "/dev/isst_interface"; + size_t size; + DIR *dir; int i, fd = 0; struct isst_if_cpu_maps map; - cpu_map = malloc(sizeof(*cpu_map) * topo_max_cpus); + /* Use calloc to make sure the memory is initialized to Zero */ + cpu_map = calloc(topo_max_cpus, sizeof(*cpu_map)); if (!cpu_map) err(3, "cpumap"); fd = open(pathname, O_RDWR); - if (fd < 0) + if (fd < 0 && !is_clx_n_platform()) err(-1, "%s open failed", pathname); + size = alloc_cpu_set(&present_cpumask); + present_cpumask_size = size; + for (i = 0; i < topo_max_cpus; ++i) { - if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) + char buffer[256]; + int pkg_id, die_id, core_id; + + /* check if CPU is online */ + snprintf(buffer, sizeof(buffer), + "/sys/devices/system/cpu/cpu%d", i); + dir = opendir(buffer); + if (!dir) + continue; + closedir(dir); + + CPU_SET_S(i, size, present_cpumask); + + pkg_id = get_physical_package_id(i); + die_id = get_physical_die_id(i); + core_id = get_physical_core_id(i); + + if (pkg_id < 0 || die_id < 0 || core_id < 0) continue; + cpu_map[i].pkg_id = pkg_id; + cpu_map[i].die_id = die_id; + cpu_map[i].core_id = core_id; + cpu_map[i].initialized = 1; + + cpu_cnt[pkg_id][die_id]++; + + if (fd < 0) + continue; map.cmd_count = 1; map.cpu_map[0].logical_cpu = i; - debug_printf(" map logical_cpu:%d\n", map.cpu_map[0].logical_cpu); if (ioctl(fd, ISST_IF_GET_PHY_ID, &map) == -1) { @@ -443,9 +671,6 @@ static void create_cpu_map(void) map.cpu_map[0].logical_cpu); continue; } - cpu_map[i].core_id = get_physical_core_id(i); - cpu_map[i].pkg_id = get_physical_package_id(i); - cpu_map[i].die_id = get_physical_die_id(i); cpu_map[i].punit_cpu = map.cpu_map[0].physical_cpu; cpu_map[i].punit_cpu_core = (map.cpu_map[0].physical_cpu >> 1); // shift to get core id @@ -456,46 +681,37 @@ static void create_cpu_map(void) cpu_map[i].pkg_id, cpu_map[i].punit_cpu, cpu_map[i].punit_cpu_core); } - - if (fd) + if (fd >= 0) close(fd); -} -int find_logical_cpu(int pkg_id, int die_id, int punit_core_id) -{ - int i; + size = alloc_cpu_set(&target_cpumask); + target_cpumask_size = size; + for (i = 0; i < max_target_cpus; ++i) { + if (!CPU_ISSET_S(target_cpus[i], present_cpumask_size, + present_cpumask)) + continue; - for (i = 0; i < topo_max_cpus; ++i) { - if (cpu_map[i].pkg_id == pkg_id && - cpu_map[i].die_id == die_id && - cpu_map[i].punit_cpu_core == punit_core_id) - return i; + CPU_SET_S(target_cpus[i], size, target_cpumask); } - - return -EINVAL; } -void set_cpu_mask_from_punit_coremask(int cpu, unsigned long long core_mask, +void set_cpu_mask_from_punit_coremask(struct isst_id *id, unsigned long long core_mask, size_t core_cpumask_size, cpu_set_t *core_cpumask, int *cpu_cnt) { int i, cnt = 0; - int die_id, pkg_id; *cpu_cnt = 0; - die_id = get_physical_die_id(cpu); - pkg_id = get_physical_package_id(cpu); for (i = 0; i < 64; ++i) { - if (core_mask & BIT(i)) { + if (core_mask & BIT_ULL(i)) { int j; for (j = 0; j < topo_max_cpus; ++j) { if (!CPU_ISSET_S(j, present_cpumask_size, present_cpumask)) continue; - if (cpu_map[j].pkg_id == pkg_id && - cpu_map[j].die_id == die_id && + if (is_cpu_in_power_domain(j, id) && cpu_map[j].punit_cpu_core == i) { CPU_SET_S(j, core_cpumask_size, core_cpumask); @@ -542,7 +758,11 @@ static int isst_send_mmio_command(unsigned int cpu, unsigned int reg, int write, } if (ioctl(fd, cmd, &io_regs) == -1) { - perror("ISST_IF_IO_CMD"); + if (errno == ENOTTY) { + perror("ISST_IF_IO_COMMAND\n"); + fprintf(stderr, "Check presence of kernel modules: isst_if_mmio\n"); + exit(0); + } fprintf(outf, "Error: mmio_cmd cpu:%d reg:%x read_write:%x\n", cpu, reg, write); } else { @@ -564,14 +784,15 @@ int isst_send_mbox_command(unsigned int cpu, unsigned char command, unsigned int req_data, unsigned int *resp) { const char *pathname = "/dev/isst_interface"; - int fd; + int fd, retry; struct isst_if_mbox_cmds mbox_cmds = { 0 }; debug_printf( "mbox_send: cpu:%d command:%x sub_command:%x parameter:%x req_data:%x\n", cpu, command, sub_command, parameter, req_data); - if (isst_platform_info.mmio_supported && command == CONFIG_CLOS) { + if (!is_skx_based_platform() && command == CONFIG_CLOS && + sub_command != CLOS_PM_QOS_CONFIG) { unsigned int value; int write = 0; int clos_id, core_id, ret = 0; @@ -615,25 +836,42 @@ int isst_send_mbox_command(unsigned int cpu, unsigned char command, mbox_cmds.mbox_cmd[0].parameter = parameter; mbox_cmds.mbox_cmd[0].req_data = req_data; + if (mbox_delay) + usleep(mbox_delay * 1000); + fd = open(pathname, O_RDWR); if (fd < 0) err(-1, "%s open failed", pathname); - if (ioctl(fd, ISST_IF_MBOX_COMMAND, &mbox_cmds) == -1) { - perror("ISST_IF_MBOX_COMMAND"); - fprintf(outf, - "Error: mbox_cmd cpu:%d command:%x sub_command:%x parameter:%x req_data:%x\n", - cpu, command, sub_command, parameter, req_data); - return -1; - } else { - *resp = mbox_cmds.mbox_cmd[0].resp_data; - debug_printf( - "mbox_cmd response: cpu:%d command:%x sub_command:%x parameter:%x req_data:%x resp:%x\n", - cpu, command, sub_command, parameter, req_data, *resp); - } + retry = mbox_retries; + + do { + if (ioctl(fd, ISST_IF_MBOX_COMMAND, &mbox_cmds) == -1) { + if (errno == ENOTTY) { + perror("ISST_IF_MBOX_COMMAND\n"); + fprintf(stderr, "Check presence of kernel modules: isst_if_mbox_pci or isst_if_mbox_msr\n"); + exit(0); + } + debug_printf( + "Error: mbox_cmd cpu:%d command:%x sub_command:%x parameter:%x req_data:%x errorno:%d\n", + cpu, command, sub_command, parameter, req_data, errno); + --retry; + } else { + *resp = mbox_cmds.mbox_cmd[0].resp_data; + debug_printf( + "mbox_cmd response: cpu:%d command:%x sub_command:%x parameter:%x req_data:%x resp:%x\n", + cpu, command, sub_command, parameter, req_data, *resp); + break; + } + } while (retry); close(fd); + if (!retry) { + debug_printf("Failed mbox command even after retries\n"); + return -1; + + } return 0; } @@ -656,7 +894,7 @@ int isst_send_msr_command(unsigned int cpu, unsigned int msr, int write, msr_cmds.msr_cmd[0].data = *req_resp; if (ioctl(fd, ISST_IF_MSR_COMMAND, &msr_cmds) == -1) { - perror("ISST_IF_MSR_COMMAD"); + perror("ISST_IF_MSR_COMMAND"); fprintf(outf, "Error: msr_cmd cpu:%d msr:%x read_write:%d\n", cpu, msr, write); } else { @@ -697,12 +935,95 @@ static int isst_fill_platform_info(void) return 0; } +static void isst_print_extended_platform_info(void) +{ + int cp_state, cp_cap, fact_support = 0, pbf_support = 0; + struct isst_pkg_ctdp_level_info ctdp_level; + struct isst_pkg_ctdp pkg_dev; + int ret, i, j; + FILE *filep; + struct isst_id id; + + for (i = 0; i < 256; ++i) { + char path[256]; + + snprintf(path, sizeof(path), + "/sys/devices/system/cpu/cpu%d/topology/thread_siblings", i); + filep = fopen(path, "r"); + if (filep) + break; + } + + if (!filep) + return; + + fclose(filep); + + set_isst_id(&id, i); + ret = isst_get_ctdp_levels(&id, &pkg_dev); + if (ret) + return; + + if (pkg_dev.enabled) { + fprintf(outf, "Intel(R) SST-PP (feature perf-profile) is supported\n"); + } else { + fprintf(outf, "Intel(R) SST-PP (feature perf-profile) is not supported\n"); + fprintf(outf, "Only performance level 0 (base level) is present\n"); + } + + if (pkg_dev.locked) + fprintf(outf, "TDP level change control is locked\n"); + else + fprintf(outf, "TDP level change control is unlocked, max level: %d \n", pkg_dev.levels); + + for (j = 0; j <= pkg_dev.levels; ++j) { + ret = isst_get_ctdp_control(&id, j, &ctdp_level); + if (ret) + continue; + + if (!fact_support && ctdp_level.fact_support) + fact_support = 1; + + if (!pbf_support && ctdp_level.pbf_support) + pbf_support = 1; + } + + if (fact_support) + fprintf(outf, "Intel(R) SST-TF (feature turbo-freq) is supported\n"); + else + fprintf(outf, "Intel(R) SST-TF (feature turbo-freq) is not supported\n"); + + if (pbf_support) + fprintf(outf, "Intel(R) SST-BF (feature base-freq) is supported\n"); + else + fprintf(outf, "Intel(R) SST-BF (feature base-freq) is not supported\n"); + + ret = isst_read_pm_config(&id, &cp_state, &cp_cap); + if (ret) { + fprintf(outf, "Intel(R) SST-CP (feature core-power) status is unknown\n"); + return; + } + if (cp_cap) + fprintf(outf, "Intel(R) SST-CP (feature core-power) is supported\n"); + else + fprintf(outf, "Intel(R) SST-CP (feature core-power) is not supported\n"); +} + static void isst_print_platform_information(void) { struct isst_if_platform_info platform_info; const char *pathname = "/dev/isst_interface"; int fd; + if (is_clx_n_platform()) { + fprintf(stderr, "\nThis option in not supported on this platform\n"); + exit(0); + } + + /* Early initialization to create working cpu_map */ + set_max_cpu_num(); + create_cpu_map(); + fd = open(pathname, O_RDWR); if (fd < 0) err(-1, "%s open failed", pathname); @@ -718,6 +1039,7 @@ static void isst_print_platform_information(void) platform_info.mbox_supported); fprintf(outf, "Platform: mmio supported : %d\n", platform_info.mmio_supported); + isst_print_extended_platform_info(); } close(fd); @@ -725,22 +1047,24 @@ static void isst_print_platform_information(void) exit(0); } -static void exec_on_get_ctdp_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static char *local_str0, *local_str1; +static void exec_on_get_ctdp_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { - int (*fn_ptr)(int cpu, void *arg); + int (*fn_ptr)(struct isst_id *id, void *arg); int ret; fn_ptr = arg1; - ret = fn_ptr(cpu, arg2); + ret = fn_ptr(id, arg2); if (ret) - perror("get_tdp_*"); + isst_display_error_info_message(1, "get_tdp_* failed", 0, 0); else - isst_ctdp_display_core_info(cpu, outf, arg3, - *(unsigned int *)arg4); + isst_ctdp_display_core_info(id, outf, arg3, + *(unsigned int *)arg4, + local_str0, local_str1); } -#define _get_tdp_level(desc, suffix, object, help) \ +#define _get_tdp_level(desc, suffix, object, help, str0, str1) \ static void get_tdp_##object(int arg) \ { \ struct isst_pkg_ctdp ctdp; \ @@ -751,6 +1075,8 @@ static void exec_on_get_ctdp_cpu(int cpu, void *arg1, void *arg2, void *arg3, help); \ exit(0); \ } \ + local_str0 = str0; \ + local_str1 = str1; \ isst_ctdp_display_information_start(outf); \ if (max_target_cpus) \ for_each_online_target_cpu_in_set( \ @@ -764,12 +1090,12 @@ static void exec_on_get_ctdp_cpu(int cpu, void *arg1, void *arg2, void *arg3, isst_ctdp_display_information_end(outf); \ } -_get_tdp_level("get-config-levels", levels, levels, "TDP levels"); -_get_tdp_level("get-config-version", levels, version, "TDP version"); -_get_tdp_level("get-config-enabled", levels, enabled, "TDP enable status"); +_get_tdp_level("get-config-levels", levels, levels, "Max TDP level", NULL, NULL); +_get_tdp_level("get-config-version", levels, version, "TDP version", NULL, NULL); +_get_tdp_level("get-config-enabled", levels, enabled, "perf-profile enable status", "disabled", "enabled"); _get_tdp_level("get-config-current_level", levels, current_level, - "Current TDP Level"); -_get_tdp_level("get-lock-status", levels, locked, "TDP lock status"); + "Current TDP Level", NULL, NULL); +_get_tdp_level("get-lock-status", levels, locked, "TDP lock status", "unlocked", "locked"); struct isst_pkg_ctdp clx_n_pkg_dev; @@ -801,9 +1127,9 @@ static int clx_n_get_base_ratio(void) return (int)(value); } -static int clx_n_config(int cpu) +static int clx_n_config(struct isst_id *id) { - int i, ret, pkg_id, die_id; + int i, ret; unsigned long cpu_bf; struct isst_pkg_ctdp_level_info *ctdp_level; struct isst_pbf_info *pbf_info; @@ -825,15 +1151,11 @@ static int clx_n_config(int cpu) pbf_info->p1_high = 0; pbf_info->p1_low = ~0; - pkg_id = get_physical_package_id(cpu); - die_id = get_physical_die_id(cpu); - for (i = 0; i < topo_max_cpus; i++) { if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) continue; - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; CPU_SET_S(i, ctdp_level->core_cpumask_size, @@ -870,8 +1192,7 @@ static int clx_n_config(int cpu) if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) continue; - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; cpu_bf = parse_int_file(1, @@ -897,39 +1218,47 @@ error_ret: return ret; } -static void dump_clx_n_config_for_cpu(int cpu, void *arg1, void *arg2, +static void dump_clx_n_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int ret; - ret = clx_n_config(cpu); + if (tdp_level != 0xff && tdp_level != 0) { + isst_display_error_info_message(1, "Invalid level", 1, tdp_level); + exit(0); + } + + ret = clx_n_config(id); if (ret) { - perror("isst_get_process_ctdp"); + debug_printf("clx_n_config failed"); } else { struct isst_pkg_ctdp_level_info *ctdp_level; struct isst_pbf_info *pbf_info; ctdp_level = &clx_n_pkg_dev.ctdp_level[0]; pbf_info = &ctdp_level->pbf_info; - isst_ctdp_display_information(cpu, outf, tdp_level, &clx_n_pkg_dev); + clx_n_pkg_dev.processed = 1; + isst_ctdp_display_information(id, outf, tdp_level, &clx_n_pkg_dev); free_cpu_set(ctdp_level->core_cpumask); free_cpu_set(pbf_info->core_cpumask); } } -static void dump_isst_config_for_cpu(int cpu, void *arg1, void *arg2, +static void dump_isst_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { struct isst_pkg_ctdp pkg_dev; int ret; memset(&pkg_dev, 0, sizeof(pkg_dev)); - ret = isst_get_process_ctdp(cpu, tdp_level, &pkg_dev); + ret = isst_get_process_ctdp(id, tdp_level, &pkg_dev); if (ret) { - perror("isst_get_process_ctdp"); + isst_display_error_info_message(1, "Failed to get perf-profile info on cpu", 1, id->cpu); + isst_ctdp_display_information_end(outf); + exit(1); } else { - isst_ctdp_display_information(cpu, outf, tdp_level, &pkg_dev); - isst_get_process_ctdp_complete(cpu, &pkg_dev); + isst_ctdp_display_information(id, outf, tdp_level, &pkg_dev); + isst_get_process_ctdp_complete(id, &pkg_dev); } } @@ -963,34 +1292,44 @@ static void dump_isst_config(int arg) isst_ctdp_display_information_end(outf); } -static void set_tdp_level_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void adjust_scaling_max_from_base_freq(int cpu); + +static void set_tdp_level_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int ret; - ret = isst_set_tdp_level(cpu, tdp_level); - if (ret) - perror("set_tdp_level_for_cpu"); - else { - isst_display_result(cpu, outf, "perf-profile", "set_tdp_level", + ret = isst_set_tdp_level(id, tdp_level); + if (ret) { + isst_display_error_info_message(1, "Set TDP level failed", 0, 0); + isst_ctdp_display_information_end(outf); + exit(1); + } else { + isst_display_result(id, outf, "perf-profile", "set_tdp_level", ret); if (force_online_offline) { struct isst_pkg_ctdp_level_info ctdp_level; - int pkg_id = get_physical_package_id(cpu); - int die_id = get_physical_die_id(cpu); + + /* Wait for updated base frequencies */ + usleep(2000); fprintf(stderr, "Option is set to online/offline\n"); ctdp_level.core_cpumask_size = alloc_cpu_set(&ctdp_level.core_cpumask); - isst_get_coremask_info(cpu, tdp_level, &ctdp_level); + ret = isst_get_coremask_info(id, tdp_level, &ctdp_level); + if (ret) { + isst_display_error_info_message(1, "Can't get coremask, online/offline option is ignored", 0, 0); + return; + } if (ctdp_level.cpu_count) { int i, max_cpus = get_topo_max_cpus(); for (i = 0; i < max_cpus; ++i) { - if (pkg_id != get_physical_package_id(i) || die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; if (CPU_ISSET_S(i, ctdp_level.core_cpumask_size, ctdp_level.core_cpumask)) { fprintf(stderr, "online cpu %d\n", i); set_cpu_online_offline(i, 1); + adjust_scaling_max_from_base_freq(i); } else { fprintf(stderr, "offline cpu %d\n", i); set_cpu_online_offline(i, 0); @@ -1009,11 +1348,13 @@ static void set_tdp_level(int arg) "\t Arguments: -l|--level : Specify tdp level\n"); fprintf(stderr, "\t Optional Arguments: -o | online : online/offline for the tdp level\n"); + fprintf(stderr, + "\t online/offline operation has limitations, refer to Linux hotplug documentation\n"); exit(0); } if (tdp_level == 0xff) { - fprintf(outf, "Invalid command: specify tdp_level\n"); + isst_display_error_info_message(1, "Invalid command: specify tdp_level", 0, 0); exit(1); } isst_ctdp_display_information_start(outf); @@ -1026,37 +1367,39 @@ static void set_tdp_level(int arg) isst_ctdp_display_information_end(outf); } -static void clx_n_dump_pbf_config_for_cpu(int cpu, void *arg1, void *arg2, +static void clx_n_dump_pbf_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int ret; - ret = clx_n_config(cpu); + ret = clx_n_config(id); if (ret) { - perror("isst_get_process_ctdp"); + isst_display_error_info_message(1, "clx_n_config failed", 0, 0); } else { struct isst_pkg_ctdp_level_info *ctdp_level; struct isst_pbf_info *pbf_info; ctdp_level = &clx_n_pkg_dev.ctdp_level[0]; pbf_info = &ctdp_level->pbf_info; - isst_pbf_display_information(cpu, outf, tdp_level, pbf_info); + isst_pbf_display_information(id, outf, tdp_level, pbf_info); free_cpu_set(ctdp_level->core_cpumask); free_cpu_set(pbf_info->core_cpumask); } } -static void dump_pbf_config_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void dump_pbf_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { struct isst_pbf_info pbf_info; int ret; - ret = isst_get_pbf_info(cpu, tdp_level, &pbf_info); + ret = isst_get_pbf_info(id, tdp_level, &pbf_info); if (ret) { - perror("isst_get_pbf_info"); + isst_display_error_info_message(1, "Failed to get base-freq info at this level", 1, tdp_level); + isst_ctdp_display_information_end(outf); + exit(1); } else { - isst_pbf_display_information(cpu, outf, tdp_level, &pbf_info); + isst_pbf_display_information(id, outf, tdp_level, &pbf_info); isst_get_pbf_info_complete(&pbf_info); } } @@ -1074,7 +1417,7 @@ static void dump_pbf_config(int arg) } if (tdp_level == 0xff) { - fprintf(outf, "Invalid command: specify tdp_level\n"); + isst_display_error_info_message(1, "Invalid command: specify tdp_level", 0, 0); exit(1); } @@ -1093,23 +1436,23 @@ static void dump_pbf_config(int arg) isst_ctdp_display_information_end(outf); } -static int set_clos_param(int cpu, int clos, int epp, int wt, int min, int max) +static int set_clos_param(struct isst_id *id, int clos, int epp, int wt, int min, int max) { struct isst_clos_config clos_config; int ret; - ret = isst_pm_get_clos(cpu, clos, &clos_config); + ret = isst_pm_get_clos(id, clos, &clos_config); if (ret) { - perror("isst_pm_get_clos"); + isst_display_error_info_message(1, "isst_pm_get_clos failed", 0, 0); return ret; } clos_config.clos_min = min; clos_config.clos_max = max; clos_config.epp = epp; clos_config.clos_prop_prio = wt; - ret = isst_set_clos(cpu, clos, &clos_config); + ret = isst_set_clos(id, clos, &clos_config); if (ret) { - perror("isst_pm_set_clos"); + isst_display_error_info_message(1, "isst_set_clos failed", 0, 0); return ret; } @@ -1144,16 +1487,41 @@ static int set_cpufreq_scaling_min_max(int cpu, int max, int freq) return 0; } -static int set_clx_pbf_cpufreq_scaling_min_max(int cpu) +static int no_turbo(void) +{ + return parse_int_file(0, "/sys/devices/system/cpu/intel_pstate/no_turbo"); +} + +static void adjust_scaling_max_from_base_freq(int cpu) +{ + int base_freq, scaling_max_freq; + + scaling_max_freq = parse_int_file(0, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_max_freq", cpu); + base_freq = get_cpufreq_base_freq(cpu); + if (scaling_max_freq < base_freq || no_turbo()) + set_cpufreq_scaling_min_max(cpu, 1, base_freq); +} + +static void adjust_scaling_min_from_base_freq(int cpu) +{ + int base_freq, scaling_min_freq; + + scaling_min_freq = parse_int_file(0, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_min_freq", cpu); + base_freq = get_cpufreq_base_freq(cpu); + if (scaling_min_freq < base_freq) + set_cpufreq_scaling_min_max(cpu, 0, base_freq); +} + +static int set_clx_pbf_cpufreq_scaling_min_max(struct isst_id *id) { struct isst_pkg_ctdp_level_info *ctdp_level; struct isst_pbf_info *pbf_info; - int i, pkg_id, die_id, freq, freq_high, freq_low; + int i, freq, freq_high, freq_low; int ret; - ret = clx_n_config(cpu); + ret = clx_n_config(id); if (ret) { - perror("set_clx_pbf_cpufreq_scaling_min_max"); + debug_printf("cpufreq_scaling_min_max failed for CLX"); return ret; } @@ -1162,11 +1530,8 @@ static int set_clx_pbf_cpufreq_scaling_min_max(int cpu) freq_high = pbf_info->p1_high * 100000; freq_low = pbf_info->p1_low * 100000; - pkg_id = get_physical_package_id(cpu); - die_id = get_physical_die_id(cpu); for (i = 0; i < get_topo_max_cpus(); ++i) { - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; if (CPU_ISSET_S(i, pbf_info->core_cpumask_size, @@ -1229,83 +1594,75 @@ static int set_cpufreq_scaling_min_max_from_cpuinfo(int cpu, int cpuinfo_max, in return 0; } -static void set_scaling_min_to_cpuinfo_max(int cpu) +static void set_scaling_min_to_cpuinfo_max(struct isst_id *id) { - int i, pkg_id, die_id; + int i; - pkg_id = get_physical_package_id(cpu); - die_id = get_physical_die_id(cpu); for (i = 0; i < get_topo_max_cpus(); ++i) { - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; + adjust_scaling_max_from_base_freq(i); set_cpufreq_scaling_min_max_from_cpuinfo(i, 1, 0); + adjust_scaling_min_from_base_freq(i); } } -static void set_scaling_min_to_cpuinfo_min(int cpu) +static void set_scaling_min_to_cpuinfo_min(struct isst_id *id) { - int i, pkg_id, die_id; + int i; - pkg_id = get_physical_package_id(cpu); - die_id = get_physical_die_id(cpu); for (i = 0; i < get_topo_max_cpus(); ++i) { - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; + adjust_scaling_max_from_base_freq(i); set_cpufreq_scaling_min_max_from_cpuinfo(i, 0, 0); } } -static void set_scaling_max_to_cpuinfo_max(int cpu) +static void set_scaling_max_to_cpuinfo_max(struct isst_id *id) { - int i, pkg_id, die_id; + int i; - pkg_id = get_physical_package_id(cpu); - die_id = get_physical_die_id(cpu); for (i = 0; i < get_topo_max_cpus(); ++i) { - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; set_cpufreq_scaling_min_max_from_cpuinfo(i, 1, 1); } } -static int set_core_priority_and_min(int cpu, int mask_size, +static int set_core_priority_and_min(struct isst_id *id, int mask_size, cpu_set_t *cpu_mask, int min_high, int min_low) { - int pkg_id, die_id, ret, i; + int ret, i; if (!CPU_COUNT_S(mask_size, cpu_mask)) return -1; - ret = set_clos_param(cpu, 0, 0, 0, min_high, 0xff); + ret = set_clos_param(id, 0, 0, 0, min_high, 0xff); if (ret) return ret; - ret = set_clos_param(cpu, 1, 15, 15, min_low, 0xff); + ret = set_clos_param(id, 1, 15, 15, min_low, 0xff); if (ret) return ret; - ret = set_clos_param(cpu, 2, 15, 15, min_low, 0xff); + ret = set_clos_param(id, 2, 15, 15, min_low, 0xff); if (ret) return ret; - ret = set_clos_param(cpu, 3, 15, 15, min_low, 0xff); + ret = set_clos_param(id, 3, 15, 15, min_low, 0xff); if (ret) return ret; - pkg_id = get_physical_package_id(cpu); - die_id = get_physical_die_id(cpu); for (i = 0; i < get_topo_max_cpus(); ++i) { int clos; + struct isst_id tid; - if (pkg_id != get_physical_package_id(i) || - die_id != get_physical_die_id(i)) + if (!is_cpu_in_power_domain(i, id)) continue; if (CPU_ISSET_S(i, mask_size, cpu_mask)) @@ -1314,9 +1671,10 @@ static int set_core_priority_and_min(int cpu, int mask_size, clos = 3; debug_printf("Associate cpu: %d clos: %d\n", i, clos); - ret = isst_clos_associate(i, clos); + set_isst_id(&tid, i); + ret = isst_clos_associate(&tid, clos); if (ret) { - perror("isst_clos_associate"); + isst_display_error_info_message(1, "isst_clos_associate failed", 0, 0); return ret; } } @@ -1324,95 +1682,111 @@ static int set_core_priority_and_min(int cpu, int mask_size, return 0; } -static int set_pbf_core_power(int cpu) +static int set_pbf_core_power(struct isst_id *id) { struct isst_pbf_info pbf_info; struct isst_pkg_ctdp pkg_dev; int ret; - ret = isst_get_ctdp_levels(cpu, &pkg_dev); + ret = isst_get_ctdp_levels(id, &pkg_dev); if (ret) { - perror("isst_get_ctdp_levels"); + debug_printf("isst_get_ctdp_levels failed"); return ret; } debug_printf("Current_level: %d\n", pkg_dev.current_level); - ret = isst_get_pbf_info(cpu, pkg_dev.current_level, &pbf_info); + ret = isst_get_pbf_info(id, pkg_dev.current_level, &pbf_info); if (ret) { - perror("isst_get_pbf_info"); + debug_printf("isst_get_pbf_info failed"); return ret; } debug_printf("p1_high: %d p1_low: %d\n", pbf_info.p1_high, pbf_info.p1_low); - ret = set_core_priority_and_min(cpu, pbf_info.core_cpumask_size, + ret = set_core_priority_and_min(id, pbf_info.core_cpumask_size, pbf_info.core_cpumask, pbf_info.p1_high, pbf_info.p1_low); if (ret) { - perror("set_core_priority_and_min"); + debug_printf("set_core_priority_and_min failed"); return ret; } - ret = isst_pm_qos_config(cpu, 1, 1); + ret = isst_pm_qos_config(id, 1, 1); if (ret) { - perror("isst_pm_qos_config"); + debug_printf("isst_pm_qos_config failed"); return ret; } return 0; } -static void set_pbf_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void set_pbf_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { + struct isst_pkg_ctdp_level_info ctdp_level; + struct isst_pkg_ctdp pkg_dev; int ret; int status = *(int *)arg4; if (is_clx_n_platform()) { + ret = 0; if (status) { - ret = 0; - if (auto_mode) - set_clx_pbf_cpufreq_scaling_min_max(cpu); + set_clx_pbf_cpufreq_scaling_min_max(id); } else { - ret = -1; - if (auto_mode) { - set_scaling_max_to_cpuinfo_max(cpu); - set_scaling_min_to_cpuinfo_min(cpu); - } + set_scaling_max_to_cpuinfo_max(id); + set_scaling_min_to_cpuinfo_min(id); } goto disp_result; } + ret = isst_get_ctdp_levels(id, &pkg_dev); + if (ret) { + isst_display_error_info_message(1, "Failed to get number of levels", 0, 0); + goto disp_result; + } + + ret = isst_get_ctdp_control(id, pkg_dev.current_level, &ctdp_level); + if (ret) { + isst_display_error_info_message(1, "Failed to get current level", 0, 0); + goto disp_result; + } + + if (!ctdp_level.pbf_support) { + isst_display_error_info_message(1, "base-freq feature is not present at this level", 1, pkg_dev.current_level); + ret = -1; + goto disp_result; + } + if (auto_mode && status) { - ret = set_pbf_core_power(cpu); + ret = set_pbf_core_power(id); if (ret) goto disp_result; } - ret = isst_set_pbf_fact_status(cpu, 1, status); + ret = isst_set_pbf_fact_status(id, 1, status); if (ret) { - perror("isst_set_pbf"); + debug_printf("isst_set_pbf_fact_status failed"); if (auto_mode) - isst_pm_qos_config(cpu, 0, 0); + isst_pm_qos_config(id, 0, 0); } else { if (auto_mode) { if (status) - set_scaling_min_to_cpuinfo_max(cpu); + set_scaling_min_to_cpuinfo_max(id); else - set_scaling_min_to_cpuinfo_min(cpu); + set_scaling_min_to_cpuinfo_min(id); } } if (auto_mode && !status) - isst_pm_qos_config(cpu, 0, 0); + isst_pm_qos_config(id, 0, 1); disp_result: if (status) - isst_display_result(cpu, outf, "base-freq", "enable", + isst_display_result(id, outf, "base-freq", "enable", ret); else - isst_display_result(cpu, outf, "base-freq", "disable", + isst_display_result(id, outf, "base-freq", "disable", ret); } @@ -1424,10 +1798,25 @@ static void set_pbf_enable(int arg) if (enable) { fprintf(stderr, "Enable Intel Speed Select Technology base frequency feature\n"); + if (is_clx_n_platform()) { + fprintf(stderr, + "\tOn this platform this command doesn't enable feature in the hardware.\n"); + fprintf(stderr, + "\tIt updates the cpufreq scaling_min_freq to match cpufreq base_frequency.\n"); + exit(0); + + } fprintf(stderr, "\tOptional Arguments: -a|--auto : Use priority of cores to set core-power associations\n"); } else { + if (is_clx_n_platform()) { + fprintf(stderr, + "\tOn this platform this command doesn't disable feature in the hardware.\n"); + fprintf(stderr, + "\tIt updates the cpufreq scaling_min_freq to match cpuinfo_min_freq\n"); + exit(0); + } fprintf(stderr, "Disable Intel Speed Select Technology base frequency feature\n"); fprintf(stderr, @@ -1446,18 +1835,21 @@ static void set_pbf_enable(int arg) isst_ctdp_display_information_end(outf); } -static void dump_fact_config_for_cpu(int cpu, void *arg1, void *arg2, +static void dump_fact_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { struct isst_fact_info fact_info; int ret; - ret = isst_get_fact_info(cpu, tdp_level, &fact_info); - if (ret) - perror("isst_get_fact_bucket_info"); - else - isst_fact_display_information(cpu, outf, tdp_level, fact_bucket, + ret = isst_get_fact_info(id, tdp_level, fact_bucket, &fact_info); + if (ret) { + isst_display_error_info_message(1, "Failed to get turbo-freq info at this level", 1, tdp_level); + isst_ctdp_display_information_end(outf); + exit(1); + } else { + isst_fact_display_information(id, outf, tdp_level, fact_bucket, fact_avx, &fact_info); + } } static void dump_fact_config(int arg) @@ -1475,7 +1867,7 @@ static void dump_fact_config(int arg) } if (tdp_level == 0xff) { - fprintf(outf, "Invalid command: specify tdp_level\n"); + isst_display_error_info_message(1, "Invalid command: specify tdp_level\n", 0, 0); exit(1); } @@ -1489,23 +1881,49 @@ static void dump_fact_config(int arg) isst_ctdp_display_information_end(outf); } -static void set_fact_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void set_fact_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { + struct isst_pkg_ctdp_level_info ctdp_level; + struct isst_pkg_ctdp pkg_dev; int ret; int status = *(int *)arg4; - if (auto_mode && status) { - ret = isst_pm_qos_config(cpu, 1, 1); + if (status && no_turbo()) { + isst_display_error_info_message(1, "Turbo mode is disabled", 0, 0); + ret = -1; + goto disp_results; + } + + ret = isst_get_ctdp_levels(id, &pkg_dev); + if (ret) { + isst_display_error_info_message(1, "Failed to get number of levels", 0, 0); + goto disp_results; + } + + ret = isst_get_ctdp_control(id, pkg_dev.current_level, &ctdp_level); + if (ret) { + isst_display_error_info_message(1, "Failed to get current level", 0, 0); + goto disp_results; + } + + if (!ctdp_level.fact_support) { + isst_display_error_info_message(1, "turbo-freq feature is not present at this level", 1, pkg_dev.current_level); + ret = -1; + goto disp_results; + } + + if (status) { + ret = isst_pm_qos_config(id, 1, 1); if (ret) goto disp_results; } - ret = isst_set_pbf_fact_status(cpu, 0, status); + ret = isst_set_pbf_fact_status(id, 0, status); if (ret) { - perror("isst_set_fact"); + debug_printf("isst_set_pbf_fact_status failed"); if (auto_mode) - isst_pm_qos_config(cpu, 0, 0); + isst_pm_qos_config(id, 0, 0); goto disp_results; } @@ -1514,29 +1932,32 @@ static void set_fact_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, if (status) { struct isst_pkg_ctdp pkg_dev; - ret = isst_get_ctdp_levels(cpu, &pkg_dev); + ret = isst_get_ctdp_levels(id, &pkg_dev); if (!ret) - ret = isst_set_trl(cpu, fact_trl); + ret = isst_set_trl(id, fact_trl); if (ret && auto_mode) - isst_pm_qos_config(cpu, 0, 0); + isst_pm_qos_config(id, 0, 0); } else { if (auto_mode) - isst_pm_qos_config(cpu, 0, 0); + isst_pm_qos_config(id, 0, 0); } disp_results: if (status) { - isst_display_result(cpu, outf, "turbo-freq", "enable", ret); + isst_display_result(id, outf, "turbo-freq", "enable", ret); + if (ret) + fact_enable_fail = ret; } else { /* Since we modified TRL during Fact enable, restore it */ - isst_set_trl_from_current_tdp(cpu, fact_trl); - isst_display_result(cpu, outf, "turbo-freq", "disable", ret); + isst_set_trl_from_current_tdp(id, fact_trl); + isst_display_result(id, outf, "turbo-freq", "disable", ret); } } static void set_fact_enable(int arg) { int i, ret, enable = arg; + struct isst_id id; if (cmd_help) { if (enable) { @@ -1568,7 +1989,7 @@ static void set_fact_enable(int arg) NULL, &enable); isst_ctdp_display_information_end(outf); - if (enable && auto_mode) { + if (!fact_enable_fail && enable && auto_mode) { /* * When we adjust CLOS param, we have to set for siblings also. * So for the each user specified CPU, also add the sibling @@ -1610,19 +2031,20 @@ static void set_fact_enable(int arg) if (!CPU_ISSET_S(i, present_cpumask_size, present_cpumask)) continue; - ret = set_clos_param(i, 0, 0, 0, 0, 0xff); + set_isst_id(&id, i); + ret = set_clos_param(&id, 0, 0, 0, 0, 0xff); if (ret) goto error_disp; - ret = set_clos_param(i, 1, 15, 15, 0, 0xff); + ret = set_clos_param(&id, 1, 15, 15, 0, 0xff); if (ret) goto error_disp; - ret = set_clos_param(i, 2, 15, 15, 0, 0xff); + ret = set_clos_param(&id, 2, 15, 15, 0, 0xff); if (ret) goto error_disp; - ret = set_clos_param(i, 3, 15, 15, 0, 0xff); + ret = set_clos_param(&id, 3, 15, 15, 0, 0xff); if (ret) goto error_disp; @@ -1632,35 +2054,39 @@ static void set_fact_enable(int arg) clos = 3; debug_printf("Associate cpu: %d clos: %d\n", i, clos); - ret = isst_clos_associate(i, clos); + ret = isst_clos_associate(&id, clos); if (ret) goto error_disp; } - isst_display_result(-1, outf, "turbo-freq --auto", "enable", 0); + set_isst_id(&id, -1); + isst_display_result(&id, outf, "turbo-freq --auto", "enable", 0); } return; error_disp: - isst_display_result(i, outf, "turbo-freq --auto", "enable", ret); + isst_display_result(&id, outf, "turbo-freq --auto", "enable", ret); } -static void enable_clos_qos_config(int cpu, void *arg1, void *arg2, void *arg3, +static void enable_clos_qos_config(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int ret; int status = *(int *)arg4; - ret = isst_pm_qos_config(cpu, status, clos_priority_type); + if (is_skx_based_platform()) + clos_priority_type = 1; + + ret = isst_pm_qos_config(id, status, clos_priority_type); if (ret) - perror("isst_pm_qos_config"); + isst_display_error_info_message(1, "isst_pm_qos_config failed", 0, 0); if (status) - isst_display_result(cpu, outf, "core-power", "enable", + isst_display_result(id, outf, "core-power", "enable", ret); else - isst_display_result(cpu, outf, "core-power", "disable", + isst_display_result(id, outf, "core-power", "disable", ret); } @@ -1672,9 +2098,11 @@ static void set_clos_enable(int arg) if (enable) { fprintf(stderr, "Enable core-power for a package/die\n"); - fprintf(stderr, - "\tClos Enable: Specify priority type with [--priority|-p]\n"); - fprintf(stderr, "\t\t 0: Proportional, 1: Ordered\n"); + if (!is_skx_based_platform()) { + fprintf(stderr, + "\tClos Enable: Specify priority type with [--priority|-p]\n"); + fprintf(stderr, "\t\t 0: Proportional, 1: Ordered\n"); + } } else { fprintf(stderr, "Disable core-power: [No command arguments are required]\n"); @@ -1697,17 +2125,17 @@ static void set_clos_enable(int arg) isst_ctdp_display_information_end(outf); } -static void dump_clos_config_for_cpu(int cpu, void *arg1, void *arg2, +static void dump_clos_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { struct isst_clos_config clos_config; int ret; - ret = isst_pm_get_clos(cpu, current_clos, &clos_config); + ret = isst_pm_get_clos(id, current_clos, &clos_config); if (ret) - perror("isst_pm_get_clos"); + isst_display_error_info_message(1, "isst_pm_get_clos failed", 0, 0); else - isst_clos_display_information(cpu, outf, current_clos, + isst_clos_display_information(id, outf, current_clos, &clos_config); } @@ -1721,7 +2149,8 @@ static void dump_clos_config(int arg) exit(0); } if (current_clos < 0 || current_clos > 3) { - fprintf(stderr, "Invalid clos id\n"); + isst_display_error_info_message(1, "Invalid clos id\n", 0, 0); + isst_ctdp_display_information_end(outf); exit(0); } @@ -1735,16 +2164,21 @@ static void dump_clos_config(int arg) isst_ctdp_display_information_end(outf); } -static void get_clos_info_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void get_clos_info_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int enable, ret, prio_type; - ret = isst_clos_get_clos_information(cpu, &enable, &prio_type); + ret = isst_clos_get_clos_information(id, &enable, &prio_type); if (ret) - perror("isst_clos_get_info"); - else - isst_clos_display_clos_information(cpu, outf, enable, prio_type); + isst_display_error_info_message(1, "isst_clos_get_info failed", 0, 0); + else { + int cp_state, cp_cap; + + isst_read_pm_config(id, &cp_state, &cp_cap); + isst_clos_display_clos_information(id, outf, enable, prio_type, + cp_state, cp_cap); + } } static void dump_clos_info(int arg) @@ -1752,42 +2186,37 @@ static void dump_clos_info(int arg) if (cmd_help) { fprintf(stderr, "Print Intel Speed Select Technology core power information\n"); - fprintf(stderr, "\tSpecify targeted cpu id with [--cpu|-c]\n"); - exit(0); - } - - if (!max_target_cpus) { - fprintf(stderr, - "Invalid target cpu. Specify with [-c|--cpu]\n"); + fprintf(stderr, "\t Optionally specify targeted cpu id with [--cpu|-c]\n"); exit(0); } isst_ctdp_display_information_start(outf); - for_each_online_target_cpu_in_set(get_clos_info_for_cpu, NULL, - NULL, NULL, NULL); + if (max_target_cpus) + for_each_online_target_cpu_in_set(get_clos_info_for_cpu, NULL, + NULL, NULL, NULL); + else + for_each_online_package_in_set(get_clos_info_for_cpu, NULL, + NULL, NULL, NULL); isst_ctdp_display_information_end(outf); } -static void set_clos_config_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void set_clos_config_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { struct isst_clos_config clos_config; int ret; - clos_config.pkg_id = get_physical_package_id(cpu); - clos_config.die_id = get_physical_die_id(cpu); - clos_config.epp = clos_epp; clos_config.clos_prop_prio = clos_prop_prio; clos_config.clos_min = clos_min; clos_config.clos_max = clos_max; clos_config.clos_desired = clos_desired; - ret = isst_set_clos(cpu, current_clos, &clos_config); + ret = isst_set_clos(id, current_clos, &clos_config); if (ret) - perror("isst_set_clos"); + isst_display_error_info_message(1, "isst_set_clos failed", 0, 0); else - isst_display_result(cpu, outf, "core-power", "config", ret); + isst_display_result(id, outf, "core-power", "config", ret); } static void set_clos_config(int arg) @@ -1797,26 +2226,27 @@ static void set_clos_config(int arg) "Set core-power configuration for one of the four clos ids\n"); fprintf(stderr, "\tSpecify targeted clos id with [--clos|-c]\n"); - fprintf(stderr, "\tSpecify clos EPP with [--epp|-e]\n"); - fprintf(stderr, - "\tSpecify clos Proportional Priority [--weight|-w]\n"); + if (!is_skx_based_platform()) { + fprintf(stderr, "\tSpecify clos EPP with [--epp|-e]\n"); + fprintf(stderr, + "\tSpecify clos Proportional Priority [--weight|-w]\n"); + } fprintf(stderr, "\tSpecify clos min in MHz with [--min|-n]\n"); fprintf(stderr, "\tSpecify clos max in MHz with [--max|-m]\n"); - fprintf(stderr, "\tSpecify clos desired in MHz with [--desired|-d]\n"); exit(0); } if (current_clos < 0 || current_clos > 3) { - fprintf(stderr, "Invalid clos id\n"); + isst_display_error_info_message(1, "Invalid clos id\n", 0, 0); exit(0); } - if (clos_epp < 0 || clos_epp > 0x0F) { - fprintf(stderr, "clos epp is not specified, default: 0\n"); + if (!is_skx_based_platform() && (clos_epp < 0 || clos_epp > 0x0F)) { + fprintf(stderr, "clos epp is not specified or invalid, default: 0\n"); clos_epp = 0; } - if (clos_prop_prio < 0 || clos_prop_prio > 0x0F) { + if (!is_skx_based_platform() && (clos_prop_prio < 0 || clos_prop_prio > 0x0F)) { fprintf(stderr, - "clos frequency weight is not specified, default: 0\n"); + "clos frequency weight is not specified or invalid, default: 0\n"); clos_prop_prio = 0; } if (clos_min < 0) { @@ -1824,11 +2254,11 @@ static void set_clos_config(int arg) clos_min = 0; } if (clos_max < 0) { - fprintf(stderr, "clos max is not specified, default: 25500 MHz\n"); + fprintf(stderr, "clos max is not specified, default: Max frequency (ratio 0xff)\n"); clos_max = 0xff; } - if (clos_desired < 0) { - fprintf(stderr, "clos desired is not specified, default: 0\n"); + if (clos_desired) { + fprintf(stderr, "clos desired is not supported on this platform\n"); clos_desired = 0x00; } @@ -1842,16 +2272,16 @@ static void set_clos_config(int arg) isst_ctdp_display_information_end(outf); } -static void set_clos_assoc_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void set_clos_assoc_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int ret; - ret = isst_clos_associate(cpu, current_clos); + ret = isst_clos_associate(id, current_clos); if (ret) - perror("isst_clos_associate"); + debug_printf("isst_clos_associate failed"); else - isst_display_result(cpu, outf, "core-power", "assoc", ret); + isst_display_result(id, outf, "core-power", "assoc", ret); } static void set_clos_assoc(int arg) @@ -1860,32 +2290,35 @@ static void set_clos_assoc(int arg) fprintf(stderr, "Associate a clos id to a CPU\n"); fprintf(stderr, "\tSpecify targeted clos id with [--clos|-c]\n"); + fprintf(stderr, + "\tFor example to associate clos 1 to CPU 0: issue\n"); + fprintf(stderr, + "\tintel-speed-select --cpu 0 core-power assoc --clos 1\n"); exit(0); } if (current_clos < 0 || current_clos > 3) { - fprintf(stderr, "Invalid clos id\n"); + isst_display_error_info_message(1, "Invalid clos id\n", 0, 0); exit(0); } if (max_target_cpus) for_each_online_target_cpu_in_set(set_clos_assoc_for_cpu, NULL, NULL, NULL, NULL); else { - fprintf(stderr, - "Invalid target cpu. Specify with [-c|--cpu]\n"); + isst_display_error_info_message(1, "Invalid target cpu. Specify with [-c|--cpu]", 0, 0); } } -static void get_clos_assoc_for_cpu(int cpu, void *arg1, void *arg2, void *arg3, +static void get_clos_assoc_for_cpu(struct isst_id *id, void *arg1, void *arg2, void *arg3, void *arg4) { int clos, ret; - ret = isst_clos_get_assoc_status(cpu, &clos); + ret = isst_clos_get_assoc_status(id, &clos); if (ret) - perror("isst_clos_get_assoc_status"); + isst_display_error_info_message(1, "isst_clos_get_assoc_status failed", 0, 0); else - isst_clos_display_assoc_information(cpu, outf, clos); + isst_clos_display_assoc_information(id, outf, clos); } static void get_clos_assoc(int arg) @@ -1897,8 +2330,7 @@ static void get_clos_assoc(int arg) } if (!max_target_cpus) { - fprintf(stderr, - "Invalid target cpu. Specify with [-c|--cpu]\n"); + isst_display_error_info_message(1, "Invalid target cpu. Specify with [-c|--cpu]", 0, 0); exit(0); } @@ -1908,6 +2340,105 @@ static void get_clos_assoc(int arg) isst_ctdp_display_information_end(outf); } +static void set_turbo_mode_for_cpu(struct isst_id *id, int status) +{ + int base_freq; + + if (status) { + base_freq = get_cpufreq_base_freq(id->cpu); + set_cpufreq_scaling_min_max(id->cpu, 1, base_freq); + } else { + set_scaling_max_to_cpuinfo_max(id); + } + + if (status) { + isst_display_result(id, outf, "turbo-mode", "enable", 0); + } else { + isst_display_result(id, outf, "turbo-mode", "disable", 0); + } +} + +static void set_turbo_mode(int arg) +{ + int i, enable = arg; + struct isst_id id; + + if (cmd_help) { + if (enable) + fprintf(stderr, "Set turbo mode enable\n"); + else + fprintf(stderr, "Set turbo mode disable\n"); + exit(0); + } + + isst_ctdp_display_information_start(outf); + + for (i = 0; i < topo_max_cpus; ++i) { + int online; + + if (i) + online = parse_int_file( + 1, "/sys/devices/system/cpu/cpu%d/online", i); + else + online = + 1; /* online entry for CPU 0 needs some special configs */ + + if (online) { + set_isst_id(&id, i); + set_turbo_mode_for_cpu(&id, enable); + } + + } + isst_ctdp_display_information_end(outf); +} + +static void get_set_trl(struct isst_id *id, void *arg1, void *arg2, void *arg3, + void *arg4) +{ + unsigned long long trl; + int set = *(int *)arg4; + int ret; + + if (set && !fact_trl) { + isst_display_error_info_message(1, "Invalid TRL. Specify with [-t|--trl]", 0, 0); + exit(0); + } + + if (set) { + ret = isst_set_trl(id, fact_trl); + isst_display_result(id, outf, "turbo-mode", "set-trl", ret); + return; + } + + ret = isst_get_trl(id, &trl); + if (ret) + isst_display_result(id, outf, "turbo-mode", "get-trl", ret); + else + isst_trl_display_information(id, outf, trl); +} + +static void process_trl(int arg) +{ + if (cmd_help) { + if (arg) { + fprintf(stderr, "Set TRL (turbo ratio limits)\n"); + fprintf(stderr, "\t t|--trl: Specify turbo ratio limit for setting TRL\n"); + } else { + fprintf(stderr, "Get TRL (turbo ratio limits)\n"); + } + exit(0); + } + + isst_ctdp_display_information_start(outf); + if (max_target_cpus) + for_each_online_target_cpu_in_set(get_set_trl, NULL, + NULL, NULL, &arg); + else + for_each_online_package_in_set(get_set_trl, NULL, + NULL, NULL, &arg); + isst_ctdp_display_information_end(outf); +} + static struct process_cmd_struct clx_n_cmds[] = { { "perf-profile", "info", dump_isst_config, 0 }, { "base-freq", "info", dump_pbf_config, 0 }, @@ -1938,6 +2469,10 @@ static struct process_cmd_struct isst_cmds[] = { { "core-power", "get-config", dump_clos_config, 0 }, { "core-power", "assoc", set_clos_assoc, 0 }, { "core-power", "get-assoc", get_clos_assoc, 0 }, + { "turbo-mode", "enable", set_turbo_mode, 0 }, + { "turbo-mode", "disable", set_turbo_mode, 1 }, + { "turbo-mode", "get-trl", process_trl, 0 }, + { "turbo-mode", "set-trl", process_trl, 1 }, { NULL, NULL, NULL } }; @@ -2035,7 +2570,7 @@ static void parse_cmd_args(int argc, int start, char **argv) option_index = start; optind = start + 1; - while ((opt = getopt_long(argc, argv, "b:l:t:c:d:e:n:m:p:w:hoa", + while ((opt = getopt_long(argc, argv, "b:l:t:c:d:e:n:m:p:w:r:hoa", long_options, &option_index)) != -1) { switch (opt) { case 'a': @@ -2061,7 +2596,7 @@ static void parse_cmd_args(int argc, int start, char **argv) fact_avx = 0x01; } else if (!strncmp(optarg, "avx2", 4)) { fact_avx = 0x02; - } else if (!strncmp(optarg, "avx512", 4)) { + } else if (!strncmp(optarg, "avx512", 6)) { fact_avx = 0x04; } else { fprintf(outf, "Invalid sse,avx options\n"); @@ -2078,6 +2613,10 @@ static void parse_cmd_args(int argc, int start, char **argv) break; case 'e': clos_epp = atoi(optarg); + if (is_skx_based_platform()) { + isst_display_error_info_message(1, "epp can't be specified on this platform", 0, 0); + exit(0); + } break; case 'n': clos_min = atoi(optarg); @@ -2089,14 +2628,25 @@ static void parse_cmd_args(int argc, int start, char **argv) break; case 'p': clos_priority_type = atoi(optarg); + if (is_skx_based_platform() && !clos_priority_type) { + isst_display_error_info_message(1, "Invalid clos priority type: proportional for this platform", 0, 0); + exit(0); + } break; case 'w': clos_prop_prio = atoi(optarg); + if (is_skx_based_platform()) { + isst_display_error_info_message(1, "weight can't be specified on this platform", 0, 0); + exit(0); + } break; default: - printf("no match\n"); + printf("Unknown option: ignore\n"); } } + + if (argv[optind]) + printf("Garbage at the end of command: ignore\n"); } static void isst_help(void) @@ -2138,6 +2688,16 @@ static void fact_help(void) printf("\tcommand : disable\n"); } +static void turbo_mode_help(void) +{ + printf("turbo-mode:\tEnables users to enable/disable turbo mode by adjusting frequency settings. Also allows to get and set turbo ratio limits (TRL).\n"); + printf("\tcommand : enable\n"); + printf("\tcommand : disable\n"); + printf("\tcommand : get-trl\n"); + printf("\tcommand : set-trl\n"); +} + + static void core_power_help(void) { printf("core-power:\tInterface that allows user to define per core/tile\n\ @@ -2162,6 +2722,7 @@ static struct process_cmd_help_struct isst_help_cmds[] = { { "base-freq", pbf_help }, { "turbo-freq", fact_help }, { "core-power", core_power_help }, + { "turbo-mode", turbo_mode_help }, { NULL, NULL } }; @@ -2193,9 +2754,6 @@ void process_command(int argc, char **argv, } } - if (!is_clx_n_platform()) - create_cpu_map(); - i = 0; while (cmds[i].feature) { if (!strcmp(cmds[i].feature, feature) && @@ -2214,11 +2772,18 @@ void process_command(int argc, char **argv, static void usage(void) { - printf("Intel(R) Speed Select Technology\n"); + if (is_clx_n_platform()) { + fprintf(stderr, "\nThere is limited support of Intel Speed Select features on this platform.\n"); + fprintf(stderr, "Everything is pre-configured using BIOS options, this tool can't enable any feature in the hardware.\n\n"); + } + printf("\nUsage:\n"); printf("intel-speed-select [OPTIONS] FEATURE COMMAND COMMAND_ARGUMENTS\n"); - printf("\nUse this tool to enumerate and control the Intel Speed Select Technology features,\n"); - printf("\nFEATURE : [perf-profile|base-freq|turbo-freq|core-power]\n"); + printf("\nUse this tool to enumerate and control the Intel Speed Select Technology features:\n"); + if (is_clx_n_platform()) + printf("\nFEATURE : [perf-profile|base-freq]\n"); + else + printf("\nFEATURE : [perf-profile|base-freq|turbo-freq|core-power|turbo-mode]\n"); printf("\nFor help on each feature, use -h|--help\n"); printf("\tFor example: intel-speed-select perf-profile -h\n"); @@ -2231,48 +2796,101 @@ static void usage(void) printf("\t\tDefault: Die scoped for all dies in the system with multiple dies/package\n"); printf("\t\t\t Or Package scoped for all Packages when each package contains one die\n"); printf("\t[-d|--debug] : Debug mode\n"); + printf("\t[-f|--format] : output format [json|text]. Default: text\n"); printf("\t[-h|--help] : Print help\n"); printf("\t[-i|--info] : Print platform information\n"); + printf("\t[-a|--all-cpus-online] : Force online every CPU in the system\n"); printf("\t[-o|--out] : Output file\n"); printf("\t\t\tDefault : stderr\n"); - printf("\t[-f|--format] : output format [json|text]. Default: text\n"); + printf("\t[-p|--pause] : Delay between two mail box commands in milliseconds\n"); + printf("\t[-r|--retry] : Retry count for mail box commands on failure, default 3\n"); printf("\t[-v|--version] : Print version\n"); - + printf("\t[-b|--oob : Start a daemon to process HFI events for perf profile change from Out of Band agent.\n"); + printf("\t[-n|--no-daemon : Don't run as daemon. By default --oob will turn on daemon mode\n"); + printf("\t[-w|--delay : Delay for reading config level state change in OOB poll mode.\n"); printf("\nResult format\n"); printf("\tResult display uses a common format for each command:\n"); printf("\tResults are formatted in text/JSON with\n"); printf("\t\tPackage, Die, CPU, and command specific results.\n"); + + printf("\nExamples\n"); + printf("\tTo get platform information:\n"); + printf("\t\tintel-speed-select --info\n"); + printf("\tTo get full perf-profile information dump:\n"); + printf("\t\tintel-speed-select perf-profile info\n"); + printf("\tTo get full base-freq information dump:\n"); + printf("\t\tintel-speed-select base-freq info -l 0\n"); + if (!is_clx_n_platform()) { + printf("\tTo get full turbo-freq information dump:\n"); + printf("\t\tintel-speed-select turbo-freq info -l 0\n"); + } exit(1); } static void print_version(void) { fprintf(outf, "Version %s\n", version_str); - fprintf(outf, "Build date %s time %s\n", __DATE__, __TIME__); exit(0); } static void cmdline(int argc, char **argv) { - int opt; + const char *pathname = "/dev/isst_interface"; + char *ptr; + FILE *fp; + int opt, force_cpus_online = 0; int option_index = 0; int ret; + int oob_mode = 0; + int poll_interval = -1; + int no_daemon = 0; static struct option long_options[] = { + { "all-cpus-online", no_argument, 0, 'a' }, { "cpu", required_argument, 0, 'c' }, { "debug", no_argument, 0, 'd' }, { "format", required_argument, 0, 'f' }, { "help", no_argument, 0, 'h' }, { "info", no_argument, 0, 'i' }, + { "pause", required_argument, 0, 'p' }, { "out", required_argument, 0, 'o' }, + { "retry", required_argument, 0, 'r' }, { "version", no_argument, 0, 'v' }, + { "oob", no_argument, 0, 'b' }, + { "no-daemon", no_argument, 0, 'n' }, + { "poll-interval", required_argument, 0, 'w' }, { 0, 0, 0, 0 } }; + if (geteuid() != 0) { + fprintf(stderr, "Must run as root\n"); + exit(0); + } + + ret = update_cpu_model(); + if (ret) + err(-1, "Invalid CPU model (%d)\n", cpu_model); + printf("Intel(R) Speed Select Technology\n"); + printf("Executing on CPU model:%d[0x%x]\n", cpu_model, cpu_model); + + if (!is_clx_n_platform()) { + fp = fopen(pathname, "rb"); + if (!fp) { + fprintf(stderr, "Intel speed select drivers are not loaded on this system.\n"); + fprintf(stderr, "Verify that kernel config includes CONFIG_INTEL_SPEED_SELECT_INTERFACE.\n"); + fprintf(stderr, "If the config is included then this is not a supported platform.\n"); + exit(0); + } + fclose(fp); + } + progname = argv[0]; - while ((opt = getopt_long_only(argc, argv, "+c:df:hio:v", long_options, + while ((opt = getopt_long_only(argc, argv, "+c:df:hio:vabw:n", long_options, &option_index)) != -1) { switch (opt) { + case 'a': + force_cpus_online = 1; + break; case 'c': parse_cpu_command(optarg); break; @@ -2295,31 +2913,61 @@ static void cmdline(int argc, char **argv) fclose(outf); outf = fopen_or_exit(optarg, "w"); break; + case 'p': + ret = strtol(optarg, &ptr, 10); + if (!ret) + fprintf(stderr, "Invalid pause interval, ignore\n"); + else + mbox_delay = ret; + break; + case 'r': + ret = strtol(optarg, &ptr, 10); + if (!ret) + fprintf(stderr, "Invalid retry count, ignore\n"); + else + mbox_retries = ret; + break; case 'v': print_version(); break; + case 'b': + oob_mode = 1; + break; + case 'n': + no_daemon = 1; + break; + case 'w': + ret = strtol(optarg, &ptr, 10); + if (!ret) { + fprintf(stderr, "Invalid poll interval count\n"); + exit(0); + } + poll_interval = ret; + break; default: usage(); } } - if (geteuid() != 0) { - fprintf(stderr, "Must run as root\n"); + if (optind > (argc - 2) && !oob_mode) { + usage(); exit(0); } + set_max_cpu_num(); + if (force_cpus_online) + force_all_cpus_online(); + store_cpu_topology(); + create_cpu_map(); - if (optind > (argc - 2)) { - fprintf(stderr, "Feature name and|or command not specified\n"); - exit(0); + if (oob_mode) { + if (debug_flag) + fprintf(stderr, "OOB mode is enabled in debug mode\n"); + + ret = isst_daemon(debug_flag, poll_interval, no_daemon); + if (ret) + fprintf(stderr, "OOB mode enable failed\n"); + goto out; } - ret = update_cpu_model(); - if (ret) - err(-1, "Invalid CPU model (%d)\n", cpu_model); - printf("Intel(R) Speed Select Technology\n"); - printf("Executing on CPU model:%d[0x%x]\n", cpu_model, cpu_model); - set_max_cpu_num(); - set_cpu_present_cpu_mask(); - set_cpu_target_cpu_mask(); if (!is_clx_n_platform()) { ret = isst_fill_platform_info(); diff --git a/tools/power/x86/intel-speed-select/isst-core.c b/tools/power/x86/intel-speed-select/isst-core.c index 81a119f688a3..f701b45c832c 100644 --- a/tools/power/x86/intel-speed-select/isst-core.c +++ b/tools/power/x86/intel-speed-select/isst-core.c @@ -6,7 +6,7 @@ #include "isst.h" -int isst_write_pm_config(int cpu, int cp_state) +int isst_write_pm_config(struct isst_id *id, int cp_state) { unsigned int req, resp; int ret; @@ -16,27 +16,27 @@ int isst_write_pm_config(int cpu, int cp_state) else req = 0; - ret = isst_send_mbox_command(cpu, WRITE_PM_CONFIG, PM_FEATURE, 0, req, + ret = isst_send_mbox_command(id->cpu, WRITE_PM_CONFIG, PM_FEATURE, 0, req, &resp); if (ret) return ret; - debug_printf("cpu:%d WRITE_PM_CONFIG resp:%x\n", cpu, resp); + debug_printf("cpu:%d WRITE_PM_CONFIG resp:%x\n", id->cpu, resp); return 0; } -int isst_read_pm_config(int cpu, int *cp_state, int *cp_cap) +int isst_read_pm_config(struct isst_id *id, int *cp_state, int *cp_cap) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, READ_PM_CONFIG, PM_FEATURE, 0, 0, + ret = isst_send_mbox_command(id->cpu, READ_PM_CONFIG, PM_FEATURE, 0, 0, &resp); if (ret) return ret; - debug_printf("cpu:%d READ_PM_CONFIG resp:%x\n", cpu, resp); + debug_printf("cpu:%d READ_PM_CONFIG resp:%x\n", id->cpu, resp); *cp_state = resp & BIT(16); *cp_cap = resp & BIT(0) ? 1 : 0; @@ -44,12 +44,12 @@ int isst_read_pm_config(int cpu, int *cp_state, int *cp_cap) return 0; } -int isst_get_ctdp_levels(int cpu, struct isst_pkg_ctdp *pkg_dev) +int isst_get_ctdp_levels(struct isst_id *id, struct isst_pkg_ctdp *pkg_dev) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_LEVELS_INFO, 0, 0, &resp); if (ret) { pkg_dev->levels = 0; @@ -60,7 +60,7 @@ int isst_get_ctdp_levels(int cpu, struct isst_pkg_ctdp *pkg_dev) return 0; } - debug_printf("cpu:%d CONFIG_TDP_GET_LEVELS_INFO resp:%x\n", cpu, resp); + debug_printf("cpu:%d CONFIG_TDP_GET_LEVELS_INFO resp:%x\n", id->cpu, resp); pkg_dev->version = resp & 0xff; pkg_dev->levels = (resp >> 8) & 0xff; @@ -71,14 +71,14 @@ int isst_get_ctdp_levels(int cpu, struct isst_pkg_ctdp *pkg_dev) return 0; } -int isst_get_ctdp_control(int cpu, int config_index, +int isst_get_ctdp_control(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { int cp_state, cp_cap; unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_TDP_CONTROL, 0, config_index, &resp); if (ret) @@ -89,51 +89,53 @@ int isst_get_ctdp_control(int cpu, int config_index, ctdp_level->fact_enabled = !!(resp & BIT(16)); ctdp_level->pbf_enabled = !!(resp & BIT(17)); - ret = isst_read_pm_config(cpu, &cp_state, &cp_cap); + ret = isst_read_pm_config(id, &cp_state, &cp_cap); if (ret) { - debug_printf("cpu:%d pm_config is not supported \n", cpu); + debug_printf("cpu:%d pm_config is not supported\n", id->cpu); } else { - debug_printf("cpu:%d pm_config SST-CP state:%d cap:%d \n", cpu, cp_state, cp_cap); + debug_printf("cpu:%d pm_config SST-CP state:%d cap:%d\n", id->cpu, cp_state, cp_cap); ctdp_level->sst_cp_support = cp_cap; ctdp_level->sst_cp_enabled = cp_state; } debug_printf( "cpu:%d CONFIG_TDP_GET_TDP_CONTROL resp:%x fact_support:%d pbf_support: %d fact_enabled:%d pbf_enabled:%d\n", - cpu, resp, ctdp_level->fact_support, ctdp_level->pbf_support, + id->cpu, resp, ctdp_level->fact_support, ctdp_level->pbf_support, ctdp_level->fact_enabled, ctdp_level->pbf_enabled); return 0; } -int isst_get_tdp_info(int cpu, int config_index, +int isst_get_tdp_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_TDP_INFO, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_TDP_INFO, 0, config_index, &resp); - if (ret) + if (ret) { + isst_display_error_info_message(1, "Invalid level, Can't get TDP information at level", 1, config_index); return ret; + } ctdp_level->pkg_tdp = resp & GENMASK(14, 0); ctdp_level->tdp_ratio = (resp & GENMASK(23, 16)) >> 16; debug_printf( "cpu:%d ctdp:%d CONFIG_TDP_GET_TDP_INFO resp:%x tdp_ratio:%d pkg_tdp:%d\n", - cpu, config_index, resp, ctdp_level->tdp_ratio, + id->cpu, config_index, resp, ctdp_level->tdp_ratio, ctdp_level->pkg_tdp); return 0; } -int isst_get_pwr_info(int cpu, int config_index, +int isst_get_pwr_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_PWR_INFO, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_PWR_INFO, 0, config_index, &resp); if (ret) return ret; @@ -143,18 +145,18 @@ int isst_get_pwr_info(int cpu, int config_index, debug_printf( "cpu:%d ctdp:%d CONFIG_TDP_GET_PWR_INFO resp:%x pkg_max_power:%d pkg_min_power:%d\n", - cpu, config_index, resp, ctdp_level->pkg_max_power, + id->cpu, config_index, resp, ctdp_level->pkg_max_power, ctdp_level->pkg_min_power); return 0; } -void isst_get_uncore_p0_p1_info(int cpu, int config_index, +void isst_get_uncore_p0_p1_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_UNCORE_P0_P1_INFO, 0, config_index, &resp); if (ret) { @@ -167,16 +169,16 @@ void isst_get_uncore_p0_p1_info(int cpu, int config_index, ctdp_level->uncore_p1 = (resp & GENMASK(15, 8)) >> 8; debug_printf( "cpu:%d ctdp:%d CONFIG_TDP_GET_UNCORE_P0_P1_INFO resp:%x uncore p0:%d uncore p1:%d\n", - cpu, config_index, resp, ctdp_level->uncore_p0, + id->cpu, config_index, resp, ctdp_level->uncore_p0, ctdp_level->uncore_p1); } -void isst_get_p1_info(int cpu, int config_index, +void isst_get_p1_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_P1_INFO, 0, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_P1_INFO, 0, config_index, &resp); if (ret) { ctdp_level->sse_p1 = 0; @@ -190,16 +192,17 @@ void isst_get_p1_info(int cpu, int config_index, ctdp_level->avx512_p1 = (resp & GENMASK(23, 16)) >> 16; debug_printf( "cpu:%d ctdp:%d CONFIG_TDP_GET_P1_INFO resp:%x sse_p1:%d avx2_p1:%d avx512_p1:%d\n", - cpu, config_index, resp, ctdp_level->sse_p1, + id->cpu, config_index, resp, ctdp_level->sse_p1, ctdp_level->avx2_p1, ctdp_level->avx512_p1); } -void isst_get_uncore_mem_freq(int cpu, int config_index, +void isst_get_uncore_mem_freq(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_MEM_FREQ, + + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_MEM_FREQ, 0, config_index, &resp); if (ret) { ctdp_level->mem_freq = 0; @@ -207,18 +210,32 @@ void isst_get_uncore_mem_freq(int cpu, int config_index, } ctdp_level->mem_freq = resp & GENMASK(7, 0); + if (is_spr_platform()) { + ctdp_level->mem_freq *= 200; + } else if (is_icx_platform()) { + if (ctdp_level->mem_freq < 7) { + ctdp_level->mem_freq = (12 - ctdp_level->mem_freq) * 133.33 * 2 * 10; + ctdp_level->mem_freq /= 10; + if (ctdp_level->mem_freq % 10 > 5) + ctdp_level->mem_freq++; + } else { + ctdp_level->mem_freq = 0; + } + } else { + ctdp_level->mem_freq = 0; + } debug_printf( "cpu:%d ctdp:%d CONFIG_TDP_GET_MEM_FREQ resp:%x uncore mem_freq:%d\n", - cpu, config_index, resp, ctdp_level->mem_freq); + id->cpu, config_index, resp, ctdp_level->mem_freq); } -int isst_get_tjmax_info(int cpu, int config_index, +int isst_get_tjmax_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_TJMAX_INFO, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_TJMAX_INFO, 0, config_index, &resp); if (ret) return ret; @@ -227,12 +244,12 @@ int isst_get_tjmax_info(int cpu, int config_index, debug_printf( "cpu:%d ctdp:%d CONFIG_TDP_GET_TJMAX_INFO resp:%x t_proc_hot:%d\n", - cpu, config_index, resp, ctdp_level->t_proc_hot); + id->cpu, config_index, resp, ctdp_level->t_proc_hot); return 0; } -int isst_get_coremask_info(int cpu, int config_index, +int isst_get_coremask_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level) { unsigned int resp; @@ -243,7 +260,7 @@ int isst_get_coremask_info(int cpu, int config_index, unsigned long long mask; int cpu_count = 0; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_CORE_MASK, 0, (i << 8) | config_index, &resp); if (ret) @@ -251,27 +268,27 @@ int isst_get_coremask_info(int cpu, int config_index, debug_printf( "cpu:%d ctdp:%d mask:%d CONFIG_TDP_GET_CORE_MASK resp:%x\n", - cpu, config_index, i, resp); + id->cpu, config_index, i, resp); mask = (unsigned long long)resp << (32 * i); - set_cpu_mask_from_punit_coremask(cpu, mask, + set_cpu_mask_from_punit_coremask(id, mask, ctdp_level->core_cpumask_size, ctdp_level->core_cpumask, &cpu_count); ctdp_level->cpu_count += cpu_count; - debug_printf("cpu:%d ctdp:%d mask:%d cpu count:%d\n", cpu, + debug_printf("cpu:%d ctdp:%d mask:%d cpu count:%d\n", id->cpu, config_index, i, ctdp_level->cpu_count); } return 0; } -int isst_get_get_trl_from_msr(int cpu, int *trl) +int isst_get_get_trl_from_msr(struct isst_id *id, int *trl) { unsigned long long msr_trl; int ret; - ret = isst_send_msr_command(cpu, 0x1AD, 0, &msr_trl); + ret = isst_send_msr_command(id->cpu, 0x1AD, 0, &msr_trl); if (ret) return ret; @@ -287,13 +304,13 @@ int isst_get_get_trl_from_msr(int cpu, int *trl) return 0; } -int isst_get_get_trl(int cpu, int level, int avx_level, int *trl) +int isst_get_get_trl(struct isst_id *id, int level, int avx_level, int *trl) { unsigned int req, resp; int ret; req = level | (avx_level << 16); - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_TURBO_LIMIT_RATIOS, 0, req, &resp); if (ret) @@ -301,7 +318,7 @@ int isst_get_get_trl(int cpu, int level, int avx_level, int *trl) debug_printf( "cpu:%d CONFIG_TDP_GET_TURBO_LIMIT_RATIOS req:%x resp:%x\n", - cpu, req, resp); + id->cpu, req, resp); trl[0] = resp & GENMASK(7, 0); trl[1] = (resp & GENMASK(15, 8)) >> 8; @@ -309,13 +326,13 @@ int isst_get_get_trl(int cpu, int level, int avx_level, int *trl) trl[3] = (resp & GENMASK(31, 24)) >> 24; req = level | BIT(8) | (avx_level << 16); - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_TURBO_LIMIT_RATIOS, 0, req, &resp); if (ret) return ret; - debug_printf("cpu:%d CONFIG_TDP_GET_TURBO_LIMIT req:%x resp:%x\n", cpu, + debug_printf("cpu:%d CONFIG_TDP_GET_TURBO_LIMIT req:%x resp:%x\n", id->cpu, req, resp); trl[4] = resp & GENMASK(7, 0); @@ -326,76 +343,83 @@ int isst_get_get_trl(int cpu, int level, int avx_level, int *trl) return 0; } -int isst_get_trl_bucket_info(int cpu, unsigned long long *buckets_info) +int isst_get_trl_bucket_info(struct isst_id *id, unsigned long long *buckets_info) { int ret; - debug_printf("cpu:%d bucket info via MSR\n", cpu); + debug_printf("cpu:%d bucket info via MSR\n", id->cpu); *buckets_info = 0; - ret = isst_send_msr_command(cpu, 0x1ae, 0, buckets_info); + ret = isst_send_msr_command(id->cpu, 0x1ae, 0, buckets_info); if (ret) return ret; - debug_printf("cpu:%d bucket info via MSR successful 0x%llx\n", cpu, + debug_printf("cpu:%d bucket info via MSR successful 0x%llx\n", id->cpu, *buckets_info); return 0; } -int isst_set_tdp_level_msr(int cpu, int tdp_level) +int isst_set_tdp_level(struct isst_id *id, int tdp_level) { - unsigned long long level = tdp_level; + unsigned int resp; int ret; - debug_printf("cpu: tdp_level via MSR %d\n", cpu, tdp_level); - if (isst_get_config_tdp_lock_status(cpu)) { - debug_printf("cpu: tdp_locked %d\n", cpu); + if (isst_get_config_tdp_lock_status(id)) { + isst_display_error_info_message(1, "TDP is locked", 0, 0); return -1; - } - if (tdp_level > 2) - return -1; /* invalid value */ + } - ret = isst_send_msr_command(cpu, 0x64b, 1, &level); - if (ret) + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_SET_LEVEL, 0, + tdp_level, &resp); + if (ret) { + isst_display_error_info_message(1, "Set TDP level failed for level", 1, tdp_level); return ret; - - debug_printf("cpu: tdp_level via MSR successful %d\n", cpu, tdp_level); + } return 0; } -int isst_set_tdp_level(int cpu, int tdp_level) +int isst_get_pbf_info(struct isst_id *id, int level, struct isst_pbf_info *pbf_info) { - unsigned int resp; - int ret; + struct isst_pkg_ctdp_level_info ctdp_level; + struct isst_pkg_ctdp pkg_dev; + int i, ret, max_punit_core, max_mask_index; + unsigned int req, resp; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_SET_LEVEL, 0, - tdp_level, &resp); - if (ret) - return isst_set_tdp_level_msr(cpu, tdp_level); + ret = isst_get_ctdp_levels(id, &pkg_dev); + if (ret) { + isst_display_error_info_message(1, "Failed to get number of levels", 0, 0); + return ret; + } - return 0; -} + if (level > pkg_dev.levels) { + isst_display_error_info_message(1, "Invalid level", 1, level); + return -1; + } -int isst_get_pbf_info(int cpu, int level, struct isst_pbf_info *pbf_info) -{ - int i, ret, core_cnt, max; - unsigned int req, resp; + ret = isst_get_ctdp_control(id, level, &ctdp_level); + if (ret) + return ret; + + if (!ctdp_level.pbf_support) { + isst_display_error_info_message(1, "base-freq feature is not present at this level", 1, level); + return -1; + } pbf_info->core_cpumask_size = alloc_cpu_set(&pbf_info->core_cpumask); - core_cnt = get_core_count(get_physical_package_id(cpu), get_physical_die_id(cpu)); - max = core_cnt > 32 ? 2 : 1; + max_punit_core = get_max_punit_core_id(id); + max_mask_index = max_punit_core > 32 ? 2 : 1; - for (i = 0; i < max; ++i) { + for (i = 0; i < max_mask_index; ++i) { unsigned long long mask; int count; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_CORE_MASK_INFO, 0, (i << 8) | level, &resp); if (ret) @@ -403,23 +427,23 @@ int isst_get_pbf_info(int cpu, int level, struct isst_pbf_info *pbf_info) debug_printf( "cpu:%d CONFIG_TDP_PBF_GET_CORE_MASK_INFO resp:%x\n", - cpu, resp); + id->cpu, resp); mask = (unsigned long long)resp << (32 * i); - set_cpu_mask_from_punit_coremask(cpu, mask, + set_cpu_mask_from_punit_coremask(id, mask, pbf_info->core_cpumask_size, pbf_info->core_cpumask, &count); } req = level; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_P1HI_P1LO_INFO, 0, req, &resp); if (ret) return ret; - debug_printf("cpu:%d CONFIG_TDP_PBF_GET_P1HI_P1LO_INFO resp:%x\n", cpu, + debug_printf("cpu:%d CONFIG_TDP_PBF_GET_P1HI_P1LO_INFO resp:%x\n", id->cpu, resp); pbf_info->p1_low = resp & 0xff; @@ -427,21 +451,21 @@ int isst_get_pbf_info(int cpu, int level, struct isst_pbf_info *pbf_info) req = level; ret = isst_send_mbox_command( - cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_TDP_INFO, 0, req, &resp); + id->cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_TDP_INFO, 0, req, &resp); if (ret) return ret; - debug_printf("cpu:%d CONFIG_TDP_PBF_GET_TDP_INFO resp:%x\n", cpu, resp); + debug_printf("cpu:%d CONFIG_TDP_PBF_GET_TDP_INFO resp:%x\n", id->cpu, resp); pbf_info->tdp = resp & 0xffff; req = level; ret = isst_send_mbox_command( - cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_TJ_MAX_INFO, 0, req, &resp); + id->cpu, CONFIG_TDP, CONFIG_TDP_PBF_GET_TJ_MAX_INFO, 0, req, &resp); if (ret) return ret; - debug_printf("cpu:%d CONFIG_TDP_PBF_GET_TJ_MAX_INFO resp:%x\n", cpu, + debug_printf("cpu:%d CONFIG_TDP_PBF_GET_TJ_MAX_INFO resp:%x\n", id->cpu, resp); pbf_info->t_control = (resp >> 8) & 0xff; pbf_info->t_prochot = resp & 0xff; @@ -454,7 +478,7 @@ void isst_get_pbf_info_complete(struct isst_pbf_info *pbf_info) free_cpu_set(pbf_info->core_cpumask); } -int isst_set_pbf_fact_status(int cpu, int pbf, int enable) +int isst_set_pbf_fact_status(struct isst_id *id, int pbf, int enable) { struct isst_pkg_ctdp pkg_dev; struct isst_pkg_ctdp_level_info ctdp_level; @@ -462,13 +486,13 @@ int isst_set_pbf_fact_status(int cpu, int pbf, int enable) unsigned int req = 0, resp; int ret; - ret = isst_get_ctdp_levels(cpu, &pkg_dev); + ret = isst_get_ctdp_levels(id, &pkg_dev); if (ret) - debug_printf("cpu:%d No support for dynamic ISST\n", cpu); + debug_printf("cpu:%d No support for dynamic ISST\n", id->cpu); current_level = pkg_dev.current_level; - ret = isst_get_ctdp_control(cpu, current_level, &ctdp_level); + ret = isst_get_ctdp_control(id, current_level, &ctdp_level); if (ret) return ret; @@ -481,6 +505,10 @@ int isst_set_pbf_fact_status(int cpu, int pbf, int enable) else req &= ~BIT(17); } else { + + if (enable && !ctdp_level.sst_cp_enabled) + isst_display_error_info_message(0, "Make sure to execute before: core-power enable", 0, 0); + if (ctdp_level.pbf_enabled) req = BIT(17); @@ -490,18 +518,18 @@ int isst_set_pbf_fact_status(int cpu, int pbf, int enable) req &= ~BIT(16); } - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_SET_TDP_CONTROL, 0, req, &resp); if (ret) return ret; debug_printf("cpu:%d CONFIG_TDP_SET_TDP_CONTROL pbf/fact:%d req:%x\n", - cpu, pbf, req); + id->cpu, pbf, req); return 0; } -int isst_get_fact_bucket_info(int cpu, int level, +int isst_get_fact_bucket_info(struct isst_id *id, int level, struct isst_fact_bucket_info *bucket_info) { unsigned int resp; @@ -511,7 +539,7 @@ int isst_get_fact_bucket_info(int cpu, int level, int j; ret = isst_send_mbox_command( - cpu, CONFIG_TDP, + id->cpu, CONFIG_TDP, CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_NUMCORES, 0, (i << 8) | level, &resp); if (ret) @@ -519,7 +547,7 @@ int isst_get_fact_bucket_info(int cpu, int level, debug_printf( "cpu:%d CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_NUMCORES index:%d level:%d resp:%x\n", - cpu, i, level, resp); + id->cpu, i, level, resp); for (j = 0; j < 4; ++j) { bucket_info[j + (i * 4)].high_priority_cores_count = @@ -532,7 +560,7 @@ int isst_get_fact_bucket_info(int cpu, int level, int j; ret = isst_send_mbox_command( - cpu, CONFIG_TDP, + id->cpu, CONFIG_TDP, CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_RATIOS, 0, (k << 16) | (i << 8) | level, &resp); if (ret) @@ -540,7 +568,7 @@ int isst_get_fact_bucket_info(int cpu, int level, debug_printf( "cpu:%d CONFIG_TDP_GET_FACT_HP_TURBO_LIMIT_RATIOS index:%d level:%d avx:%d resp:%x\n", - cpu, i, level, k, resp); + id->cpu, i, level, k, resp); for (j = 0; j < 4; ++j) { switch (k) { @@ -566,44 +594,94 @@ int isst_get_fact_bucket_info(int cpu, int level, return 0; } -int isst_get_fact_info(int cpu, int level, struct isst_fact_info *fact_info) +int isst_get_fact_info(struct isst_id *id, int level, int fact_bucket, struct isst_fact_info *fact_info) { + struct isst_pkg_ctdp_level_info ctdp_level; + struct isst_pkg_ctdp pkg_dev; unsigned int resp; - int ret; + int j, ret, print; - ret = isst_send_mbox_command(cpu, CONFIG_TDP, + ret = isst_get_ctdp_levels(id, &pkg_dev); + if (ret) { + isst_display_error_info_message(1, "Failed to get number of levels", 0, 0); + return ret; + } + + if (level > pkg_dev.levels) { + isst_display_error_info_message(1, "Invalid level", 1, level); + return -1; + } + + ret = isst_get_ctdp_control(id, level, &ctdp_level); + if (ret) + return ret; + + if (!ctdp_level.fact_support) { + isst_display_error_info_message(1, "turbo-freq feature is not present at this level", 1, level); + return -1; + } + + ret = isst_send_mbox_command(id->cpu, CONFIG_TDP, CONFIG_TDP_GET_FACT_LP_CLIPPING_RATIO, 0, level, &resp); if (ret) return ret; debug_printf("cpu:%d CONFIG_TDP_GET_FACT_LP_CLIPPING_RATIO resp:%x\n", - cpu, resp); + id->cpu, resp); fact_info->lp_clipping_ratio_license_sse = resp & 0xff; fact_info->lp_clipping_ratio_license_avx2 = (resp >> 8) & 0xff; fact_info->lp_clipping_ratio_license_avx512 = (resp >> 16) & 0xff; - ret = isst_get_fact_bucket_info(cpu, level, fact_info->bucket_info); + ret = isst_get_fact_bucket_info(id, level, fact_info->bucket_info); + if (ret) + return ret; - return ret; + print = 0; + for (j = 0; j < ISST_FACT_MAX_BUCKETS; ++j) { + if (fact_bucket != 0xff && fact_bucket != j) + continue; + + if (!fact_info->bucket_info[j].high_priority_cores_count) + break; + + print = 1; + } + if (!print) { + isst_display_error_info_message(1, "Invalid bucket", 0, 0); + return -1; + } + + return 0; } -int isst_set_trl(int cpu, unsigned long long trl) +int isst_get_trl(struct isst_id *id, unsigned long long *trl) +{ + int ret; + + ret = isst_send_msr_command(id->cpu, 0x1AD, 0, trl); + if (ret) + return ret; + + return 0; +} + +int isst_set_trl(struct isst_id *id, unsigned long long trl) { int ret; if (!trl) trl = 0xFFFFFFFFFFFFFFFFULL; - ret = isst_send_msr_command(cpu, 0x1AD, 1, &trl); + ret = isst_send_msr_command(id->cpu, 0x1AD, 1, &trl); if (ret) return ret; return 0; } -int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl) +int isst_set_trl_from_current_tdp(struct isst_id *id, unsigned long long trl) { unsigned long long msr_trl; int ret; @@ -615,11 +693,11 @@ int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl) int trl[8]; int i; - ret = isst_get_ctdp_levels(cpu, &pkg_dev); + ret = isst_get_ctdp_levels(id, &pkg_dev); if (ret) return ret; - ret = isst_get_get_trl(cpu, pkg_dev.current_level, 0, trl); + ret = isst_get_get_trl(id, pkg_dev.current_level, 0, trl); if (ret) return ret; @@ -630,7 +708,7 @@ int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl) msr_trl |= (_trl << (i * 8)); } } - ret = isst_send_msr_command(cpu, 0x1AD, 1, &msr_trl); + ret = isst_send_msr_command(id->cpu, 0x1AD, 1, &msr_trl); if (ret) return ret; @@ -638,12 +716,12 @@ int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl) } /* Return 1 if locked */ -int isst_get_config_tdp_lock_status(int cpu) +int isst_get_config_tdp_lock_status(struct isst_id *id) { unsigned long long tdp_control = 0; int ret; - ret = isst_send_msr_command(cpu, 0x64b, 0, &tdp_control); + ret = isst_send_msr_command(id->cpu, 0x64b, 0, &tdp_control); if (ret) return ret; @@ -652,7 +730,7 @@ int isst_get_config_tdp_lock_status(int cpu) return ret; } -void isst_get_process_ctdp_complete(int cpu, struct isst_pkg_ctdp *pkg_dev) +void isst_get_process_ctdp_complete(struct isst_id *id, struct isst_pkg_ctdp *pkg_dev) { int i; @@ -669,126 +747,138 @@ void isst_get_process_ctdp_complete(int cpu, struct isst_pkg_ctdp *pkg_dev) } } -int isst_get_process_ctdp(int cpu, int tdp_level, struct isst_pkg_ctdp *pkg_dev) +int isst_get_process_ctdp(struct isst_id *id, int tdp_level, struct isst_pkg_ctdp *pkg_dev) { - int i, ret; + int i, ret, valid = 0; if (pkg_dev->processed) return 0; - ret = isst_get_ctdp_levels(cpu, pkg_dev); + ret = isst_get_ctdp_levels(id, pkg_dev); if (ret) return ret; debug_printf("cpu: %d ctdp enable:%d current level: %d levels:%d\n", - cpu, pkg_dev->enabled, pkg_dev->current_level, + id->cpu, pkg_dev->enabled, pkg_dev->current_level, pkg_dev->levels); + if (tdp_level != 0xff && tdp_level > pkg_dev->levels) { + isst_display_error_info_message(1, "Invalid level", 0, 0); + return -1; + } + + if (!pkg_dev->enabled) + isst_display_error_info_message(0, "perf-profile feature is not supported, just base-config level 0 is valid", 0, 0); + for (i = 0; i <= pkg_dev->levels; ++i) { struct isst_pkg_ctdp_level_info *ctdp_level; if (tdp_level != 0xff && i != tdp_level) continue; - debug_printf("cpu:%d Get Information for TDP level:%d\n", cpu, + debug_printf("cpu:%d Get Information for TDP level:%d\n", id->cpu, i); ctdp_level = &pkg_dev->ctdp_level[i]; ctdp_level->level = i; - ctdp_level->control_cpu = cpu; - ctdp_level->pkg_id = get_physical_package_id(cpu); - ctdp_level->die_id = get_physical_die_id(cpu); + ctdp_level->control_cpu = id->cpu; + ctdp_level->pkg_id = id->pkg; + ctdp_level->die_id = id->die; - ret = isst_get_ctdp_control(cpu, i, ctdp_level); + ret = isst_get_ctdp_control(id, i, ctdp_level); if (ret) continue; + valid = 1; pkg_dev->processed = 1; ctdp_level->processed = 1; if (ctdp_level->pbf_support) { - ret = isst_get_pbf_info(cpu, i, &ctdp_level->pbf_info); + ret = isst_get_pbf_info(id, i, &ctdp_level->pbf_info); if (!ret) ctdp_level->pbf_found = 1; } if (ctdp_level->fact_support) { - ret = isst_get_fact_info(cpu, i, + ret = isst_get_fact_info(id, i, 0xff, &ctdp_level->fact_info); if (ret) return ret; } - if (!pkg_dev->enabled) { + if (!pkg_dev->enabled && is_skx_based_platform()) { int freq; - freq = get_cpufreq_base_freq(cpu); + freq = get_cpufreq_base_freq(id->cpu); if (freq > 0) { ctdp_level->sse_p1 = freq / 100000; ctdp_level->tdp_ratio = ctdp_level->sse_p1; } - isst_get_get_trl_from_msr(cpu, ctdp_level->trl_sse_active_cores); - isst_get_trl_bucket_info(cpu, &ctdp_level->buckets_info); + isst_get_get_trl_from_msr(id, ctdp_level->trl_sse_active_cores); + isst_get_trl_bucket_info(id, &ctdp_level->buckets_info); continue; } - ret = isst_get_tdp_info(cpu, i, ctdp_level); + ret = isst_get_tdp_info(id, i, ctdp_level); if (ret) return ret; - ret = isst_get_pwr_info(cpu, i, ctdp_level); + ret = isst_get_pwr_info(id, i, ctdp_level); if (ret) return ret; - ret = isst_get_tjmax_info(cpu, i, ctdp_level); + ret = isst_get_tjmax_info(id, i, ctdp_level); if (ret) return ret; ctdp_level->core_cpumask_size = alloc_cpu_set(&ctdp_level->core_cpumask); - ret = isst_get_coremask_info(cpu, i, ctdp_level); + ret = isst_get_coremask_info(id, i, ctdp_level); if (ret) return ret; - ret = isst_get_trl_bucket_info(cpu, &ctdp_level->buckets_info); + ret = isst_get_trl_bucket_info(id, &ctdp_level->buckets_info); if (ret) return ret; - ret = isst_get_get_trl(cpu, i, 0, + ret = isst_get_get_trl(id, i, 0, ctdp_level->trl_sse_active_cores); if (ret) return ret; - ret = isst_get_get_trl(cpu, i, 1, + ret = isst_get_get_trl(id, i, 1, ctdp_level->trl_avx_active_cores); if (ret) return ret; - ret = isst_get_get_trl(cpu, i, 2, + ret = isst_get_get_trl(id, i, 2, ctdp_level->trl_avx_512_active_cores); if (ret) return ret; - isst_get_uncore_p0_p1_info(cpu, i, ctdp_level); - isst_get_p1_info(cpu, i, ctdp_level); - isst_get_uncore_mem_freq(cpu, i, ctdp_level); + isst_get_uncore_p0_p1_info(id, i, ctdp_level); + isst_get_p1_info(id, i, ctdp_level); + isst_get_uncore_mem_freq(id, i, ctdp_level); } + if (!valid) + isst_display_error_info_message(0, "Invalid level, Can't get TDP control information at specified levels on cpu", 1, id->cpu); + return 0; } -int isst_clos_get_clos_information(int cpu, int *enable, int *type) +int isst_clos_get_clos_information(struct isst_id *id, int *enable, int *type) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, 0, 0, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, 0, 0, &resp); if (ret) return ret; - debug_printf("cpu:%d CLOS_PM_QOS_CONFIG resp:%x\n", cpu, resp); + debug_printf("cpu:%d CLOS_PM_QOS_CONFIG resp:%x\n", id->cpu, resp); if (resp & BIT(1)) *enable = 1; @@ -803,7 +893,7 @@ int isst_clos_get_clos_information(int cpu, int *enable, int *type) return 0; } -int isst_pm_qos_config(int cpu, int enable_clos, int priority_type) +int isst_pm_qos_config(struct isst_id *id, int enable_clos, int priority_type) { unsigned int req, resp; int ret; @@ -812,36 +902,38 @@ int isst_pm_qos_config(int cpu, int enable_clos, int priority_type) struct isst_pkg_ctdp pkg_dev; struct isst_pkg_ctdp_level_info ctdp_level; - ret = isst_get_ctdp_levels(cpu, &pkg_dev); + ret = isst_get_ctdp_levels(id, &pkg_dev); if (ret) { debug_printf("isst_get_ctdp_levels\n"); return ret; } - ret = isst_get_ctdp_control(cpu, pkg_dev.current_level, + ret = isst_get_ctdp_control(id, pkg_dev.current_level, &ctdp_level); if (ret) return ret; if (ctdp_level.fact_enabled) { - debug_printf("Turbo-freq feature must be disabled first\n"); + isst_display_error_info_message(1, "Ignoring request, turbo-freq feature is still enabled", 0, 0); return -EINVAL; } - ret = isst_write_pm_config(cpu, 0); + ret = isst_write_pm_config(id, 0); if (ret) - perror("isst_write_pm_config\n"); + isst_display_error_info_message(0, "WRITE_PM_CONFIG command failed, ignoring error", 0, 0); } else { - ret = isst_write_pm_config(cpu, 1); + ret = isst_write_pm_config(id, 1); if (ret) - perror("isst_write_pm_config\n"); + isst_display_error_info_message(0, "WRITE_PM_CONFIG command failed, ignoring error", 0, 0); } - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, 0, 0, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, 0, 0, &resp); - if (ret) + if (ret) { + isst_display_error_info_message(1, "CLOS_PM_QOS_CONFIG command failed", 0, 0); return ret; + } - debug_printf("cpu:%d CLOS_PM_QOS_CONFIG resp:%x\n", cpu, resp); + debug_printf("cpu:%d CLOS_PM_QOS_CONFIG resp:%x\n", id->cpu, resp); req = resp; @@ -850,35 +942,35 @@ int isst_pm_qos_config(int cpu, int enable_clos, int priority_type) else req = req & ~BIT(1); + if (priority_type > 1) + isst_display_error_info_message(1, "Invalid priority type: Changing type to ordered", 0, 0); + if (priority_type) req = req | BIT(2); else req = req & ~BIT(2); - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PM_QOS_CONFIG, BIT(MBOX_CMD_WRITE_BIT), req, &resp); if (ret) return ret; - debug_printf("cpu:%d CLOS_PM_QOS_CONFIG priority type:%d req:%x\n", cpu, + debug_printf("cpu:%d CLOS_PM_QOS_CONFIG priority type:%d req:%x\n", id->cpu, priority_type, req); return 0; } -int isst_pm_get_clos(int cpu, int clos, struct isst_clos_config *clos_config) +int isst_pm_get_clos(struct isst_id *id, int clos, struct isst_clos_config *clos_config) { unsigned int resp; int ret; - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_CLOS, clos, 0, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PM_CLOS, clos, 0, &resp); if (ret) return ret; - clos_config->pkg_id = get_physical_package_id(cpu); - clos_config->die_id = get_physical_die_id(cpu); - clos_config->epp = resp & 0x0f; clos_config->clos_prop_prio = (resp >> 4) & 0x0f; clos_config->clos_min = (resp >> 8) & 0xff; @@ -888,7 +980,7 @@ int isst_pm_get_clos(int cpu, int clos, struct isst_clos_config *clos_config) return 0; } -int isst_set_clos(int cpu, int clos, struct isst_clos_config *clos_config) +int isst_set_clos(struct isst_id *id, int clos, struct isst_clos_config *clos_config) { unsigned int req, resp; unsigned int param; @@ -902,53 +994,53 @@ int isst_set_clos(int cpu, int clos, struct isst_clos_config *clos_config) param = BIT(MBOX_CMD_WRITE_BIT) | clos; - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PM_CLOS, param, req, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PM_CLOS, param, req, &resp); if (ret) return ret; - debug_printf("cpu:%d CLOS_PM_CLOS param:%x req:%x\n", cpu, param, req); + debug_printf("cpu:%d CLOS_PM_CLOS param:%x req:%x\n", id->cpu, param, req); return 0; } -int isst_clos_get_assoc_status(int cpu, int *clos_id) +int isst_clos_get_assoc_status(struct isst_id *id, int *clos_id) { unsigned int resp; unsigned int param; int core_id, ret; - core_id = find_phy_core_num(cpu); + core_id = find_phy_core_num(id->cpu); param = core_id; - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PQR_ASSOC, param, 0, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PQR_ASSOC, param, 0, &resp); if (ret) return ret; - debug_printf("cpu:%d CLOS_PQR_ASSOC param:%x resp:%x\n", cpu, param, + debug_printf("cpu:%d CLOS_PQR_ASSOC param:%x resp:%x\n", id->cpu, param, resp); *clos_id = (resp >> 16) & 0x03; return 0; } -int isst_clos_associate(int cpu, int clos_id) +int isst_clos_associate(struct isst_id *id, int clos_id) { unsigned int req, resp; unsigned int param; int core_id, ret; req = (clos_id & 0x03) << 16; - core_id = find_phy_core_num(cpu); + core_id = find_phy_core_num(id->cpu); param = BIT(MBOX_CMD_WRITE_BIT) | core_id; - ret = isst_send_mbox_command(cpu, CONFIG_CLOS, CLOS_PQR_ASSOC, param, + ret = isst_send_mbox_command(id->cpu, CONFIG_CLOS, CLOS_PQR_ASSOC, param, req, &resp); if (ret) return ret; - debug_printf("cpu:%d CLOS_PQR_ASSOC param:%x req:%x\n", cpu, param, + debug_printf("cpu:%d CLOS_PQR_ASSOC param:%x req:%x\n", id->cpu, param, req); return 0; diff --git a/tools/power/x86/intel-speed-select/isst-daemon.c b/tools/power/x86/intel-speed-select/isst-daemon.c new file mode 100644 index 000000000000..0699137c0901 --- /dev/null +++ b/tools/power/x86/intel-speed-select/isst-daemon.c @@ -0,0 +1,242 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Intel Speed Select -- Allow speed select to daemonize + * Copyright (c) 2022 Intel Corporation. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <stdarg.h> +#include <string.h> +#include <unistd.h> +#include <fcntl.h> +#include <sys/file.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <errno.h> +#include <getopt.h> +#include <signal.h> +#include <time.h> + +#include "isst.h" + +static int per_package_levels_info[MAX_PACKAGE_COUNT][MAX_DIE_PER_PACKAGE]; +static time_t per_package_levels_tm[MAX_PACKAGE_COUNT][MAX_DIE_PER_PACKAGE]; + +static void init_levels(void) +{ + int i, j; + + for (i = 0; i < MAX_PACKAGE_COUNT; ++i) + for (j = 0; j < MAX_DIE_PER_PACKAGE; ++j) + per_package_levels_info[i][j] = -1; +} + +void process_level_change(struct isst_id *id) +{ + struct isst_pkg_ctdp_level_info ctdp_level; + struct isst_pkg_ctdp pkg_dev; + time_t tm; + int ret; + + if (id->pkg < 0 || id->die < 0) { + debug_printf("Invalid package/die info for cpu:%d\n", id->cpu); + return; + } + + tm = time(NULL); + if (tm - per_package_levels_tm[id->pkg][id->die] < 2) + return; + + per_package_levels_tm[id->pkg][id->die] = tm; + + ret = isst_get_ctdp_levels(id, &pkg_dev); + if (ret) { + debug_printf("Can't get tdp levels for cpu:%d\n", id->cpu); + return; + } + + debug_printf("Get Config level %d pkg:%d die:%d current_level:%d\n", id->cpu, + id->pkg, id->die, pkg_dev.current_level); + + if (pkg_dev.locked) { + debug_printf("config TDP s locked \n"); + return; + } + + if (per_package_levels_info[id->pkg][id->die] == pkg_dev.current_level) + return; + + debug_printf("**Config level change for cpu:%d pkg:%d die:%d from %d to %d\n", + id->cpu, id->pkg, id->die, per_package_levels_info[id->pkg][id->die], + pkg_dev.current_level); + + per_package_levels_info[id->pkg][id->die] = pkg_dev.current_level; + + ctdp_level.core_cpumask_size = + alloc_cpu_set(&ctdp_level.core_cpumask); + ret = isst_get_coremask_info(id, pkg_dev.current_level, &ctdp_level); + if (ret) { + free_cpu_set(ctdp_level.core_cpumask); + debug_printf("Can't get core_mask:%d\n", id->cpu); + return; + } + + if (ctdp_level.cpu_count) { + int i, max_cpus = get_topo_max_cpus(); + for (i = 0; i < max_cpus; ++i) { + if (!is_cpu_in_power_domain(i, id)) + continue; + if (CPU_ISSET_S(i, ctdp_level.core_cpumask_size, ctdp_level.core_cpumask)) { + fprintf(stderr, "online cpu %d\n", i); + set_cpu_online_offline(i, 1); + } else { + fprintf(stderr, "offline cpu %d\n", i); + set_cpu_online_offline(i, 0); + } + } + } + + free_cpu_set(ctdp_level.core_cpumask); +} + +static void _poll_for_config_change(struct isst_id *id, void *arg1, void *arg2, + void *arg3, void *arg4) +{ + process_level_change(id); +} + +static void poll_for_config_change(void) +{ + for_each_online_package_in_set(_poll_for_config_change, NULL, NULL, + NULL, NULL); +} + +static int done = 0; +static int pid_file_handle; + +static void signal_handler(int sig) +{ + switch (sig) { + case SIGINT: + case SIGTERM: + done = 1; + hfi_exit(); + exit(0); + break; + default: + break; + } +} + +static void daemonize(char *rundir, char *pidfile) +{ + int pid, sid, i; + char str[10]; + struct sigaction sig_actions; + sigset_t sig_set; + int ret; + + if (getppid() == 1) + return; + + sigemptyset(&sig_set); + sigaddset(&sig_set, SIGCHLD); + sigaddset(&sig_set, SIGTSTP); + sigaddset(&sig_set, SIGTTOU); + sigaddset(&sig_set, SIGTTIN); + sigprocmask(SIG_BLOCK, &sig_set, NULL); + + sig_actions.sa_handler = signal_handler; + sigemptyset(&sig_actions.sa_mask); + sig_actions.sa_flags = 0; + + sigaction(SIGHUP, &sig_actions, NULL); + sigaction(SIGTERM, &sig_actions, NULL); + sigaction(SIGINT, &sig_actions, NULL); + + pid = fork(); + if (pid < 0) { + /* Could not fork */ + exit(EXIT_FAILURE); + } + if (pid > 0) + exit(EXIT_SUCCESS); + + umask(027); + + sid = setsid(); + if (sid < 0) + exit(EXIT_FAILURE); + + /* close all descriptors */ + for (i = getdtablesize(); i >= 0; --i) + close(i); + + i = open("/dev/null", O_RDWR); + ret = dup(i); + if (ret == -1) + exit(EXIT_FAILURE); + + ret = dup(i); + if (ret == -1) + exit(EXIT_FAILURE); + + ret = chdir(rundir); + if (ret == -1) + exit(EXIT_FAILURE); + + pid_file_handle = open(pidfile, O_RDWR | O_CREAT, 0600); + if (pid_file_handle == -1) { + /* Couldn't open lock file */ + exit(1); + } + /* Try to lock file */ +#ifdef LOCKF_SUPPORT + if (lockf(pid_file_handle, F_TLOCK, 0) == -1) { +#else + if (flock(pid_file_handle, LOCK_EX|LOCK_NB) < 0) { +#endif + /* Couldn't get lock on lock file */ + fprintf(stderr, "Couldn't get lock file %d\n", getpid()); + exit(1); + } + snprintf(str, sizeof(str), "%d\n", getpid()); + ret = write(pid_file_handle, str, strlen(str)); + if (ret == -1) + exit(EXIT_FAILURE); + + close(i); +} + +int isst_daemon(int debug_mode, int poll_interval, int no_daemon) +{ + int ret; + + if (!no_daemon && poll_interval < 0 && !debug_mode) { + fprintf(stderr, "OOB mode is enabled and will run as daemon\n"); + daemonize((char *) "/tmp/", + (char *)"/tmp/hfi-events.pid"); + } else { + signal(SIGINT, signal_handler); + } + + init_levels(); + + if (poll_interval < 0) { + ret = hfi_main(); + if (ret) { + fprintf(stderr, "HFI initialization failed\n"); + } + fprintf(stderr, "Must specify poll-interval\n"); + return ret; + } + + debug_printf("Starting loop\n"); + while (!done) { + sleep(poll_interval); + poll_for_config_change(); + } + + return 0; +} diff --git a/tools/power/x86/intel-speed-select/isst-display.c b/tools/power/x86/intel-speed-select/isst-display.c index 4fb0c1d49d64..b19f57d30f55 100644 --- a/tools/power/x86/intel-speed-select/isst-display.c +++ b/tools/power/x86/intel-speed-select/isst-display.c @@ -25,10 +25,14 @@ static void printcpulist(int str_len, char *str, int mask_size, index = snprintf(&str[curr_index], str_len - curr_index, ","); curr_index += index; + if (curr_index >= str_len) + break; } index = snprintf(&str[curr_index], str_len - curr_index, "%d", i); curr_index += index; + if (curr_index >= str_len) + break; first = 0; } } @@ -64,10 +68,14 @@ static void printcpumask(int str_len, char *str, int mask_size, index = snprintf(&str[curr_index], str_len - curr_index, "%08x", mask[i]); curr_index += index; + if (curr_index >= str_len) + break; if (i) { strncat(&str[curr_index], ",", str_len - curr_index); curr_index++; } + if (curr_index >= str_len) + break; } free(mask); @@ -158,27 +166,34 @@ static void format_and_print(FILE *outf, int level, char *header, char *value) last_level = level; } -static void print_package_info(int cpu, FILE *outf) +static int print_package_info(struct isst_id *id, FILE *outf) { char header[256]; - snprintf(header, sizeof(header), "package-%d", - get_physical_package_id(cpu)); + if (out_format_is_json()) { + snprintf(header, sizeof(header), "package-%d:die-%d:cpu-%d", + id->pkg, id->die, id->cpu); + format_and_print(outf, 1, header, NULL); + return 1; + } + snprintf(header, sizeof(header), "package-%d", id->pkg); format_and_print(outf, 1, header, NULL); - snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); + snprintf(header, sizeof(header), "die-%d", id->die); format_and_print(outf, 2, header, NULL); - snprintf(header, sizeof(header), "cpu-%d", cpu); + snprintf(header, sizeof(header), "cpu-%d", id->cpu); format_and_print(outf, 3, header, NULL); + + return 3; } -static void _isst_pbf_display_information(int cpu, FILE *outf, int level, +static void _isst_pbf_display_information(struct isst_id *id, FILE *outf, int level, struct isst_pbf_info *pbf_info, int disp_level) { char header[256]; - char value[256]; + char value[512]; - snprintf(header, sizeof(header), "speed-select-base-freq"); + snprintf(header, sizeof(header), "speed-select-base-freq-properties"); format_and_print(outf, disp_level, header, NULL); snprintf(header, sizeof(header), "high-priority-base-frequency(MHz)"); @@ -214,7 +229,7 @@ static void _isst_pbf_display_information(int cpu, FILE *outf, int level, format_and_print(outf, disp_level + 1, header, value); } -static void _isst_fact_display_information(int cpu, FILE *outf, int level, +static void _isst_fact_display_information(struct isst_id *id, FILE *outf, int level, int fact_bucket, int fact_avx, struct isst_fact_info *fact_info, int base_level) @@ -222,9 +237,23 @@ static void _isst_fact_display_information(int cpu, FILE *outf, int level, struct isst_fact_bucket_info *bucket_info = fact_info->bucket_info; char header[256]; char value[256]; - int j; + int print = 0, j; + + for (j = 0; j < ISST_FACT_MAX_BUCKETS; ++j) { + if (fact_bucket != 0xff && fact_bucket != j) + continue; + + if (!bucket_info[j].high_priority_cores_count) + break; + + print = 1; + } + if (!print) { + fprintf(stderr, "Invalid bucket\n"); + return; + } - snprintf(header, sizeof(header), "speed-select-turbo-freq"); + snprintf(header, sizeof(header), "speed-select-turbo-freq-properties"); format_and_print(outf, base_level, header, NULL); for (j = 0; j < ISST_FACT_MAX_BUCKETS; ++j) { if (fact_bucket != 0xff && fact_bucket != j) @@ -288,35 +317,47 @@ static void _isst_fact_display_information(int cpu, FILE *outf, int level, format_and_print(outf, base_level + 2, header, value); } -void isst_ctdp_display_core_info(int cpu, FILE *outf, char *prefix, - unsigned int val) +void isst_ctdp_display_core_info(struct isst_id *id, FILE *outf, char *prefix, + unsigned int val, char *str0, char *str1) { char header[256]; char value[256]; + int level = 1; - snprintf(header, sizeof(header), "package-%d", - get_physical_package_id(cpu)); - format_and_print(outf, 1, header, NULL); - snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); - format_and_print(outf, 2, header, NULL); - snprintf(header, sizeof(header), "cpu-%d", cpu); - format_and_print(outf, 3, header, NULL); + if (out_format_is_json()) { + snprintf(header, sizeof(header), "package-%d:die-%d:cpu-%d", + id->pkg, id->die, id->cpu); + format_and_print(outf, level++, header, NULL); + } else { + snprintf(header, sizeof(header), "package-%d", id->pkg); + format_and_print(outf, level++, header, NULL); + snprintf(header, sizeof(header), "die-%d", id->die); + format_and_print(outf, level++, header, NULL); + snprintf(header, sizeof(header), "cpu-%d", id->cpu); + format_and_print(outf, level++, header, NULL); + } - snprintf(value, sizeof(value), "%u", val); - format_and_print(outf, 4, prefix, value); + if (str0 && !val) + snprintf(value, sizeof(value), "%s", str0); + else if (str1 && val) + snprintf(value, sizeof(value), "%s", str1); + else + snprintf(value, sizeof(value), "%u", val); + format_and_print(outf, level, prefix, value); format_and_print(outf, 1, NULL, NULL); } -void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, +void isst_ctdp_display_information(struct isst_id *id, FILE *outf, int tdp_level, struct isst_pkg_ctdp *pkg_dev) { char header[256]; - char value[256]; - int i, base_level = 1; + char value[512]; + static int level; + int i; if (pkg_dev->processed) - print_package_info(cpu, outf); + level = print_package_info(id, outf); for (i = 0; i <= pkg_dev->levels; ++i) { struct isst_pkg_ctdp_level_info *ctdp_level; @@ -328,72 +369,79 @@ void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, snprintf(header, sizeof(header), "perf-profile-level-%d", ctdp_level->level); - format_and_print(outf, base_level + 3, header, NULL); + format_and_print(outf, level + 1, header, NULL); snprintf(header, sizeof(header), "cpu-count"); - j = get_cpu_count(get_physical_die_id(cpu), - get_physical_die_id(cpu)); + j = get_cpu_count(id); snprintf(value, sizeof(value), "%d", j); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); + + j = CPU_COUNT_S(ctdp_level->core_cpumask_size, + ctdp_level->core_cpumask); + if (j) { + snprintf(header, sizeof(header), "enable-cpu-count"); + snprintf(value, sizeof(value), "%d", j); + format_and_print(outf, level + 2, header, value); + } if (ctdp_level->core_cpumask_size) { snprintf(header, sizeof(header), "enable-cpu-mask"); printcpumask(sizeof(value), value, ctdp_level->core_cpumask_size, ctdp_level->core_cpumask); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "enable-cpu-list"); printcpulist(sizeof(value), value, ctdp_level->core_cpumask_size, ctdp_level->core_cpumask); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } snprintf(header, sizeof(header), "thermal-design-power-ratio"); snprintf(value, sizeof(value), "%d", ctdp_level->tdp_ratio); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "base-frequency(MHz)"); if (!ctdp_level->sse_p1) ctdp_level->sse_p1 = ctdp_level->tdp_ratio; snprintf(value, sizeof(value), "%d", ctdp_level->sse_p1 * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); if (ctdp_level->avx2_p1) { snprintf(header, sizeof(header), "base-frequency-avx2(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->avx2_p1 * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } if (ctdp_level->avx512_p1) { snprintf(header, sizeof(header), "base-frequency-avx512(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->avx512_p1 * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } if (ctdp_level->uncore_p1) { snprintf(header, sizeof(header), "uncore-frequency-min(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->uncore_p1 * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } if (ctdp_level->uncore_p0) { snprintf(header, sizeof(header), "uncore-frequency-max(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->uncore_p0 * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } if (ctdp_level->mem_freq) { snprintf(header, sizeof(header), "mem-frequency(MHz)"); snprintf(value, sizeof(value), "%d", - ctdp_level->mem_freq * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 4, header, value); + ctdp_level->mem_freq); + format_and_print(outf, level + 2, header, value); } snprintf(header, sizeof(header), @@ -405,7 +453,7 @@ void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, snprintf(value, sizeof(value), "disabled"); } else snprintf(value, sizeof(value), "unsupported"); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "speed-select-base-freq"); @@ -416,7 +464,7 @@ void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, snprintf(value, sizeof(value), "disabled"); } else snprintf(value, sizeof(value), "unsupported"); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "speed-select-core-power"); @@ -427,241 +475,311 @@ void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, snprintf(value, sizeof(value), "disabled"); } else snprintf(value, sizeof(value), "unsupported"); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); if (is_clx_n_platform()) { if (ctdp_level->pbf_support) - _isst_pbf_display_information(cpu, outf, + _isst_pbf_display_information(id, outf, tdp_level, &ctdp_level->pbf_info, - base_level + 4); + level + 2); continue; } if (ctdp_level->pkg_tdp) { snprintf(header, sizeof(header), "thermal-design-power(W)"); snprintf(value, sizeof(value), "%d", ctdp_level->pkg_tdp); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } if (ctdp_level->t_proc_hot) { snprintf(header, sizeof(header), "tjunction-max(C)"); snprintf(value, sizeof(value), "%d", ctdp_level->t_proc_hot); - format_and_print(outf, base_level + 4, header, value); + format_and_print(outf, level + 2, header, value); } snprintf(header, sizeof(header), "turbo-ratio-limits-sse"); - format_and_print(outf, base_level + 4, header, NULL); + format_and_print(outf, level + 2, header, NULL); for (j = 0; j < 8; ++j) { snprintf(header, sizeof(header), "bucket-%d", j); - format_and_print(outf, base_level + 5, header, NULL); + format_and_print(outf, level + 3, header, NULL); snprintf(header, sizeof(header), "core-count"); snprintf(value, sizeof(value), "%llu", (ctdp_level->buckets_info >> (j * 8)) & 0xff); - format_and_print(outf, base_level + 6, header, value); + format_and_print(outf, level + 4, header, value); snprintf(header, sizeof(header), "max-turbo-frequency(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->trl_sse_active_cores[j] * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 6, header, value); + format_and_print(outf, level + 4, header, value); } if (ctdp_level->trl_avx_active_cores[0]) { snprintf(header, sizeof(header), "turbo-ratio-limits-avx2"); - format_and_print(outf, base_level + 4, header, NULL); + format_and_print(outf, level + 2, header, NULL); for (j = 0; j < 8; ++j) { snprintf(header, sizeof(header), "bucket-%d", j); - format_and_print(outf, base_level + 5, header, NULL); + format_and_print(outf, level + 3, header, NULL); snprintf(header, sizeof(header), "core-count"); snprintf(value, sizeof(value), "%llu", (ctdp_level->buckets_info >> (j * 8)) & 0xff); - format_and_print(outf, base_level + 6, header, value); + format_and_print(outf, level + 4, header, value); snprintf(header, sizeof(header), "max-turbo-frequency(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->trl_avx_active_cores[j] * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 6, header, value); + format_and_print(outf, level + 4, header, value); } } if (ctdp_level->trl_avx_512_active_cores[0]) { snprintf(header, sizeof(header), "turbo-ratio-limits-avx512"); - format_and_print(outf, base_level + 4, header, NULL); + format_and_print(outf, level + 2, header, NULL); for (j = 0; j < 8; ++j) { snprintf(header, sizeof(header), "bucket-%d", j); - format_and_print(outf, base_level + 5, header, NULL); + format_and_print(outf, level + 3, header, NULL); snprintf(header, sizeof(header), "core-count"); snprintf(value, sizeof(value), "%llu", (ctdp_level->buckets_info >> (j * 8)) & 0xff); - format_and_print(outf, base_level + 6, header, value); + format_and_print(outf, level + 4, header, value); snprintf(header, sizeof(header), "max-turbo-frequency(MHz)"); snprintf(value, sizeof(value), "%d", ctdp_level->trl_avx_512_active_cores[j] * DISP_FREQ_MULTIPLIER); - format_and_print(outf, base_level + 6, header, value); + format_and_print(outf, level + 4, header, value); } } if (ctdp_level->pbf_support) - _isst_pbf_display_information(cpu, outf, i, + _isst_pbf_display_information(id, outf, i, &ctdp_level->pbf_info, - base_level + 4); + level + 2); if (ctdp_level->fact_support) - _isst_fact_display_information(cpu, outf, i, 0xff, 0xff, + _isst_fact_display_information(id, outf, i, 0xff, 0xff, &ctdp_level->fact_info, - base_level + 4); + level + 2); } format_and_print(outf, 1, NULL, NULL); } +static int start; void isst_ctdp_display_information_start(FILE *outf) { last_level = 0; format_and_print(outf, 0, "start", NULL); + start = 1; } void isst_ctdp_display_information_end(FILE *outf) { format_and_print(outf, 0, NULL, NULL); + start = 0; } -void isst_pbf_display_information(int cpu, FILE *outf, int level, +void isst_pbf_display_information(struct isst_id *id, FILE *outf, int level, struct isst_pbf_info *pbf_info) { - print_package_info(cpu, outf); - _isst_pbf_display_information(cpu, outf, level, pbf_info, 4); + int _level; + + _level = print_package_info(id, outf); + _isst_pbf_display_information(id, outf, level, pbf_info, _level + 1); format_and_print(outf, 1, NULL, NULL); } -void isst_fact_display_information(int cpu, FILE *outf, int level, +void isst_fact_display_information(struct isst_id *id, FILE *outf, int level, int fact_bucket, int fact_avx, struct isst_fact_info *fact_info) { - print_package_info(cpu, outf); - _isst_fact_display_information(cpu, outf, level, fact_bucket, fact_avx, - fact_info, 4); + int _level; + + _level = print_package_info(id, outf); + _isst_fact_display_information(id, outf, level, fact_bucket, fact_avx, + fact_info, _level + 1); format_and_print(outf, 1, NULL, NULL); } -void isst_clos_display_information(int cpu, FILE *outf, int clos, +void isst_clos_display_information(struct isst_id *id, FILE *outf, int clos, struct isst_clos_config *clos_config) { char header[256]; char value[256]; + int level; - snprintf(header, sizeof(header), "package-%d", - get_physical_package_id(cpu)); - format_and_print(outf, 1, header, NULL); - snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); - format_and_print(outf, 2, header, NULL); - snprintf(header, sizeof(header), "cpu-%d", cpu); - format_and_print(outf, 3, header, NULL); + level = print_package_info(id, outf); snprintf(header, sizeof(header), "core-power"); - format_and_print(outf, 4, header, NULL); + format_and_print(outf, level + 1, header, NULL); snprintf(header, sizeof(header), "clos"); snprintf(value, sizeof(value), "%d", clos); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "epp"); snprintf(value, sizeof(value), "%d", clos_config->epp); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "clos-proportional-priority"); snprintf(value, sizeof(value), "%d", clos_config->clos_prop_prio); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "clos-min"); snprintf(value, sizeof(value), "%d MHz", clos_config->clos_min * DISP_FREQ_MULTIPLIER); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "clos-max"); - snprintf(value, sizeof(value), "%d MHz", clos_config->clos_max * DISP_FREQ_MULTIPLIER); - format_and_print(outf, 5, header, value); + if (clos_config->clos_max == 0xff) + snprintf(value, sizeof(value), "Max Turbo frequency"); + else + snprintf(value, sizeof(value), "%d MHz", clos_config->clos_max * DISP_FREQ_MULTIPLIER); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "clos-desired"); snprintf(value, sizeof(value), "%d MHz", clos_config->clos_desired * DISP_FREQ_MULTIPLIER); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); - format_and_print(outf, 1, NULL, NULL); + format_and_print(outf, level, NULL, NULL); } -void isst_clos_display_clos_information(int cpu, FILE *outf, - int clos_enable, int type) +void isst_clos_display_clos_information(struct isst_id *id, FILE *outf, + int clos_enable, int type, + int state, int cap) { char header[256]; char value[256]; + int level; - snprintf(header, sizeof(header), "package-%d", - get_physical_package_id(cpu)); - format_and_print(outf, 1, header, NULL); - snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); - format_and_print(outf, 2, header, NULL); - snprintf(header, sizeof(header), "cpu-%d", cpu); - format_and_print(outf, 3, header, NULL); + level = print_package_info(id, outf); snprintf(header, sizeof(header), "core-power"); - format_and_print(outf, 4, header, NULL); + format_and_print(outf, level + 1, header, NULL); + + snprintf(header, sizeof(header), "support-status"); + if (cap) + snprintf(value, sizeof(value), "supported"); + else + snprintf(value, sizeof(value), "unsupported"); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "enable-status"); - snprintf(value, sizeof(value), "%d", clos_enable); - format_and_print(outf, 5, header, value); + if (state) + snprintf(value, sizeof(value), "enabled"); + else + snprintf(value, sizeof(value), "disabled"); + format_and_print(outf, level + 2, header, value); + + snprintf(header, sizeof(header), "clos-enable-status"); + if (clos_enable) + snprintf(value, sizeof(value), "enabled"); + else + snprintf(value, sizeof(value), "disabled"); + format_and_print(outf, level + 2, header, value); snprintf(header, sizeof(header), "priority-type"); - snprintf(value, sizeof(value), "%d", type); - format_and_print(outf, 5, header, value); + if (type) + snprintf(value, sizeof(value), "ordered"); + else + snprintf(value, sizeof(value), "proportional"); + format_and_print(outf, level + 2, header, value); - format_and_print(outf, 1, NULL, NULL); + format_and_print(outf, level, NULL, NULL); } -void isst_clos_display_assoc_information(int cpu, FILE *outf, int clos) +void isst_clos_display_assoc_information(struct isst_id *id, FILE *outf, int clos) { char header[256]; char value[256]; + int level; - snprintf(header, sizeof(header), "package-%d", - get_physical_package_id(cpu)); - format_and_print(outf, 1, header, NULL); - snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); - format_and_print(outf, 2, header, NULL); - snprintf(header, sizeof(header), "cpu-%d", cpu); - format_and_print(outf, 3, header, NULL); + level = print_package_info(id, outf); snprintf(header, sizeof(header), "get-assoc"); - format_and_print(outf, 4, header, NULL); + format_and_print(outf, level + 1, header, NULL); snprintf(header, sizeof(header), "clos"); snprintf(value, sizeof(value), "%d", clos); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); - format_and_print(outf, 1, NULL, NULL); + format_and_print(outf, level, NULL, NULL); } -void isst_display_result(int cpu, FILE *outf, char *feature, char *cmd, +void isst_display_result(struct isst_id *id, FILE *outf, char *feature, char *cmd, int result) { char header[256]; char value[256]; + int level = 3; + + if (id->cpu >= 0) + level = print_package_info(id, outf); - if (cpu >= 0) { - snprintf(header, sizeof(header), "package-%d", - get_physical_package_id(cpu)); - format_and_print(outf, 1, header, NULL); - snprintf(header, sizeof(header), "die-%d", get_physical_die_id(cpu)); - format_and_print(outf, 2, header, NULL); - snprintf(header, sizeof(header), "cpu-%d", cpu); - format_and_print(outf, 3, header, NULL); - } snprintf(header, sizeof(header), "%s", feature); - format_and_print(outf, 4, header, NULL); + format_and_print(outf, level + 1, header, NULL); snprintf(header, sizeof(header), "%s", cmd); if (!result) snprintf(value, sizeof(value), "success"); else snprintf(value, sizeof(value), "failed(error %d)", result); - format_and_print(outf, 5, header, value); + format_and_print(outf, level + 2, header, value); + + format_and_print(outf, level, NULL, NULL); +} + +void isst_display_error_info_message(int error, char *msg, int arg_valid, int arg) +{ + FILE *outf = get_output_file(); + static int error_index; + char header[256]; + char value[256]; + + if (!out_format_is_json()) { + if (arg_valid) + snprintf(value, sizeof(value), "%s %d", msg, arg); + else + snprintf(value, sizeof(value), "%s", msg); + if (error) + fprintf(outf, "Error: %s\n", value); + else + fprintf(outf, "Information: %s\n", value); + return; + } + + if (!start) + format_and_print(outf, 0, "start", NULL); + + if (error) + snprintf(header, sizeof(header), "Error%d", error_index++); + else + snprintf(header, sizeof(header), "Information:%d", error_index++); + format_and_print(outf, 1, header, NULL); + + snprintf(header, sizeof(header), "message"); + if (arg_valid) + snprintf(value, sizeof(value), "%s %d", msg, arg); + else + snprintf(value, sizeof(value), "%s", msg); + + format_and_print(outf, 2, header, value); format_and_print(outf, 1, NULL, NULL); + if (!start) + format_and_print(outf, 0, NULL, NULL); +} + +void isst_trl_display_information(struct isst_id *id, FILE *outf, unsigned long long trl) +{ + char header[256]; + char value[256]; + int level; + + level = print_package_info(id, outf); + + snprintf(header, sizeof(header), "get-trl"); + format_and_print(outf, level + 1, header, NULL); + + snprintf(header, sizeof(header), "trl"); + snprintf(value, sizeof(value), "0x%llx", trl); + format_and_print(outf, level + 2, header, value); + + format_and_print(outf, level, NULL, NULL); } diff --git a/tools/power/x86/intel-speed-select/isst.h b/tools/power/x86/intel-speed-select/isst.h index ad5aa6341d0f..409fcc9c8033 100644 --- a/tools/power/x86/intel-speed-select/isst.h +++ b/tools/power/x86/intel-speed-select/isst.h @@ -29,6 +29,7 @@ #include <sys/ioctl.h> #define BIT(x) (1 << (x)) +#define BIT_ULL(nr) (1ULL << (nr)) #define GENMASK(h, l) (((~0UL) << (l)) & (~0UL >> (sizeof(long) * 8 - 1 - (h)))) #define GENMASK_ULL(h, l) \ (((~0ULL) << (l)) & (~0ULL >> (sizeof(long long) * 8 - 1 - (h)))) @@ -75,9 +76,17 @@ #define DISP_FREQ_MULTIPLIER 100 +#define MAX_PACKAGE_COUNT 8 +#define MAX_DIE_PER_PACKAGE 2 + +/* Unified structure to specific a CPU or a Power Domain */ +struct isst_id { + int cpu; + int pkg; + int die; +}; + struct isst_clos_config { - int pkg_id; - int die_id; unsigned char epp; unsigned char clos_prop_prio; unsigned char clos_min; @@ -167,21 +176,20 @@ struct isst_pkg_ctdp { struct isst_pkg_ctdp_level_info ctdp_level[ISST_MAX_TDP_LEVELS]; }; +extern int is_cpu_in_power_domain(int cpu, struct isst_id *id); extern int get_topo_max_cpus(void); -extern int get_cpu_count(int pkg_id, int die_id); -extern int get_core_count(int pkg_id, int die_id); +extern int get_cpu_count(struct isst_id *id); +extern int get_max_punit_core_id(struct isst_id *id); /* Common interfaces */ +FILE *get_output_file(void); extern void debug_printf(const char *format, ...); extern int out_format_is_json(void); -extern int get_physical_package_id(int cpu); -extern int get_physical_die_id(int cpu); +extern void set_isst_id(struct isst_id *id, int cpu); extern size_t alloc_cpu_set(cpu_set_t **cpu_set); extern void free_cpu_set(cpu_set_t *cpu_set); -extern int find_logical_cpu(int pkg_id, int die_id, int phy_cpu); -extern int find_phy_cpu_num(int logical_cpu); extern int find_phy_core_num(int logical_cpu); -extern void set_cpu_mask_from_punit_coremask(int cpu, +extern void set_cpu_mask_from_punit_coremask(struct isst_id *id, unsigned long long core_mask, size_t core_cpumask_size, cpu_set_t *core_cpumask, @@ -195,57 +203,74 @@ extern int isst_send_mbox_command(unsigned int cpu, unsigned char command, extern int isst_send_msr_command(unsigned int cpu, unsigned int command, int write, unsigned long long *req_resp); -extern int isst_get_ctdp_levels(int cpu, struct isst_pkg_ctdp *pkg_dev); -extern int isst_get_coremask_info(int cpu, int config_index, +extern int isst_get_ctdp_levels(struct isst_id *id, struct isst_pkg_ctdp *pkg_dev); +extern int isst_get_ctdp_control(struct isst_id *id, int config_index, + struct isst_pkg_ctdp_level_info *ctdp_level); +extern int isst_get_coremask_info(struct isst_id *id, int config_index, struct isst_pkg_ctdp_level_info *ctdp_level); -extern int isst_get_process_ctdp(int cpu, int tdp_level, +extern int isst_get_process_ctdp(struct isst_id *id, int tdp_level, struct isst_pkg_ctdp *pkg_dev); -extern void isst_get_process_ctdp_complete(int cpu, +extern void isst_get_process_ctdp_complete(struct isst_id *id, struct isst_pkg_ctdp *pkg_dev); -extern void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level, +extern void isst_ctdp_display_information(struct isst_id *id, FILE *outf, int tdp_level, struct isst_pkg_ctdp *pkg_dev); -extern void isst_ctdp_display_core_info(int cpu, FILE *outf, char *prefix, - unsigned int val); +extern void isst_ctdp_display_core_info(struct isst_id *id, FILE *outf, char *prefix, + unsigned int val, char *str0, char *str1); extern void isst_ctdp_display_information_start(FILE *outf); extern void isst_ctdp_display_information_end(FILE *outf); -extern void isst_pbf_display_information(int cpu, FILE *outf, int level, +extern void isst_pbf_display_information(struct isst_id *id, FILE *outf, int level, struct isst_pbf_info *info); -extern int isst_set_tdp_level(int cpu, int tdp_level); -extern int isst_set_tdp_level_msr(int cpu, int tdp_level); -extern int isst_set_pbf_fact_status(int cpu, int pbf, int enable); -extern int isst_get_pbf_info(int cpu, int level, +extern int isst_set_tdp_level(struct isst_id *id, int tdp_level); +extern int isst_set_pbf_fact_status(struct isst_id *id, int pbf, int enable); +extern int isst_get_pbf_info(struct isst_id *id, int level, struct isst_pbf_info *pbf_info); extern void isst_get_pbf_info_complete(struct isst_pbf_info *pbf_info); -extern int isst_get_fact_info(int cpu, int level, +extern int isst_get_fact_info(struct isst_id *id, int level, int fact_bucket, struct isst_fact_info *fact_info); -extern int isst_get_fact_bucket_info(int cpu, int level, +extern int isst_get_fact_bucket_info(struct isst_id *id, int level, struct isst_fact_bucket_info *bucket_info); -extern void isst_fact_display_information(int cpu, FILE *outf, int level, +extern void isst_fact_display_information(struct isst_id *id, FILE *outf, int level, int fact_bucket, int fact_avx, struct isst_fact_info *fact_info); -extern int isst_set_trl(int cpu, unsigned long long trl); -extern int isst_set_trl_from_current_tdp(int cpu, unsigned long long trl); -extern int isst_get_config_tdp_lock_status(int cpu); +extern int isst_set_trl(struct isst_id *id, unsigned long long trl); +extern int isst_get_trl(struct isst_id *id, unsigned long long *trl); +extern int isst_set_trl_from_current_tdp(struct isst_id *id, unsigned long long trl); +extern int isst_get_config_tdp_lock_status(struct isst_id *id); -extern int isst_pm_qos_config(int cpu, int enable_clos, int priority_type); -extern int isst_pm_get_clos(int cpu, int clos, +extern int isst_pm_qos_config(struct isst_id *id, int enable_clos, int priority_type); +extern int isst_pm_get_clos(struct isst_id *id, int clos, struct isst_clos_config *clos_config); -extern int isst_set_clos(int cpu, int clos, +extern int isst_set_clos(struct isst_id *id, int clos, struct isst_clos_config *clos_config); -extern int isst_clos_associate(int cpu, int clos); -extern int isst_clos_get_assoc_status(int cpu, int *clos_id); -extern void isst_clos_display_information(int cpu, FILE *outf, int clos, +extern int isst_clos_associate(struct isst_id *id, int clos); +extern int isst_clos_get_assoc_status(struct isst_id *id, int *clos_id); +extern void isst_clos_display_information(struct isst_id *id, FILE *outf, int clos, struct isst_clos_config *clos_config); -extern void isst_clos_display_assoc_information(int cpu, FILE *outf, int clos); -extern int isst_read_reg(unsigned short reg, unsigned int *val); -extern int isst_write_reg(int reg, unsigned int val); +extern void isst_clos_display_assoc_information(struct isst_id *id, FILE *outf, int clos); -extern void isst_display_result(int cpu, FILE *outf, char *feature, char *cmd, +extern void isst_display_result(struct isst_id *id, FILE *outf, char *feature, char *cmd, int result); -extern int isst_clos_get_clos_information(int cpu, int *enable, int *type); -extern void isst_clos_display_clos_information(int cpu, FILE *outf, - int clos_enable, int type); +extern int isst_clos_get_clos_information(struct isst_id *id, int *enable, int *type); +extern void isst_clos_display_clos_information(struct isst_id *id, FILE *outf, + int clos_enable, int type, + int state, int cap); extern int is_clx_n_platform(void); extern int get_cpufreq_base_freq(int cpu); +extern int isst_read_pm_config(struct isst_id *id, int *cp_state, int *cp_cap); +extern void isst_display_error_info_message(int error, char *msg, int arg_valid, int arg); +extern int is_skx_based_platform(void); +extern int is_spr_platform(void); +extern int is_icx_platform(void); +extern void isst_trl_display_information(struct isst_id *id, FILE *outf, unsigned long long trl); + +extern void set_cpu_online_offline(int cpu, int state); +extern void for_each_online_package_in_set(void (*callback)(struct isst_id *, void *, void *, + void *, void *), + void *arg1, void *arg2, void *arg3, + void *arg4); +extern int isst_daemon(int debug_mode, int poll_interval, int no_daemon); +extern void process_level_change(struct isst_id *id); +extern int hfi_main(void); +extern void hfi_exit(void); #endif diff --git a/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py b/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py index 256199c7a182..b46e9eb8f5aa 100755 --- a/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py +++ b/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py @@ -1,4 +1,4 @@ -#!/usr/bin/python +#!/usr/bin/env python # SPDX-License-Identifier: GPL-2.0-only # -*- coding: utf-8 -*- # @@ -63,7 +63,7 @@ C_USEC = 3 C_SEC = 2 C_CPU = 1 -global sample_num, last_sec_cpu, last_usec_cpu, start_time, testname +global sample_num, last_sec_cpu, last_usec_cpu, start_time, testname, trace_file # 11 digits covers uptime to 115 days getcontext().prec = 11 @@ -72,17 +72,17 @@ sample_num =0 last_sec_cpu = [0] * MAX_CPUS last_usec_cpu = [0] * MAX_CPUS -def print_help(): - print('intel_pstate_tracer.py:') +def print_help(driver_name): + print('%s_tracer.py:'%driver_name) print(' Usage:') print(' If the trace file is available, then to simply parse and plot, use (sudo not required):') - print(' ./intel_pstate_tracer.py [-c cpus] -t <trace_file> -n <test_name>') + print(' ./%s_tracer.py [-c cpus] -t <trace_file> -n <test_name>'%driver_name) print(' Or') - print(' ./intel_pstate_tracer.py [--cpu cpus] ---trace_file <trace_file> --name <test_name>') + print(' ./%s_tracer.py [--cpu cpus] ---trace_file <trace_file> --name <test_name>'%driver_name) print(' To generate trace file, parse and plot, use (sudo required):') - print(' sudo ./intel_pstate_tracer.py [-c cpus] -i <interval> -n <test_name> -m <kbytes>') + print(' sudo ./%s_tracer.py [-c cpus] -i <interval> -n <test_name> -m <kbytes>'%driver_name) print(' Or') - print(' sudo ./intel_pstate_tracer.py [--cpu cpus] --interval <interval> --name <test_name> --memory <kbytes>') + print(' sudo ./%s_tracer.py [--cpu cpus] --interval <interval> --name <test_name> --memory <kbytes>'%driver_name) print(' Optional argument:') print(' cpus: comma separated list of CPUs') print(' kbytes: Kilo bytes of memory per CPU to allocate to the trace buffer. Default: 10240') @@ -235,7 +235,6 @@ def plot_duration_cpu(): output_png = 'all_cpu_durations.png' g_plot = common_all_gnuplot_settings(output_png) # autoscale this one, no set y range - g_plot('set ytics 0, 500') g_plot('set ylabel "Timer Duration (MilliSeconds)"') g_plot('set title "{} : cpu durations : {:%F %H:%M}"'.format(testname, datetime.now())) @@ -324,7 +323,7 @@ def set_4_plot_linestyles(g_plot): g_plot('set style line 3 linetype 1 linecolor rgb "purple" pointtype -1') g_plot('set style line 4 linetype 1 linecolor rgb "blue" pointtype -1') -def store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _to, mperf, aperf, tsc, freq_ghz, io_boost, common_comm, load, duration_ms, sample_num, elapsed_time, tsc_ghz): +def store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _to, mperf, aperf, tsc, freq_ghz, io_boost, common_comm, load, duration_ms, sample_num, elapsed_time, tsc_ghz, cpu_mask): """ Store master csv file information """ global graph_data_present @@ -343,11 +342,9 @@ def store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _t graph_data_present = True; -def split_csv(): +def split_csv(current_max_cpu, cpu_mask): """ seperate the all csv file into per CPU csv files. """ - global current_max_cpu - if os.path.exists('cpu.csv'): for index in range(0, current_max_cpu + 1): if cpu_mask[int(index)] != 0: @@ -382,27 +379,25 @@ def clear_trace_file(): print('IO error clearing trace file ') sys.exit(2) -def enable_trace(): +def enable_trace(trace_file): """ Enable trace """ try: - open('/sys/kernel/debug/tracing/events/power/pstate_sample/enable' - , 'w').write("1") + open(trace_file,'w').write("1") except: print('IO error enabling trace ') sys.exit(2) -def disable_trace(): +def disable_trace(trace_file): """ Disable trace """ try: - open('/sys/kernel/debug/tracing/events/power/pstate_sample/enable' - , 'w').write("0") + open(trace_file, 'w').write("0") except: print('IO error disabling trace ') sys.exit(2) -def set_trace_buffer_size(): +def set_trace_buffer_size(memory): """ Set trace buffer size """ try: @@ -422,7 +417,7 @@ def free_trace_buffer(): print('IO error freeing trace buffer ') sys.exit(2) -def read_trace_data(filename): +def read_trace_data(filename, cpu_mask): """ Read and parse trace data """ global current_max_cpu @@ -482,135 +477,137 @@ def read_trace_data(filename): tsc_ghz = Decimal(0) if duration_ms != Decimal(0) : tsc_ghz = Decimal(tsc)/duration_ms/Decimal(1000000) - store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _to, mperf, aperf, tsc, freq_ghz, io_boost, common_comm, load, duration_ms, sample_num, elapsed_time, tsc_ghz) + store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _to, mperf, aperf, tsc, freq_ghz, io_boost, common_comm, load, duration_ms, sample_num, elapsed_time, tsc_ghz, cpu_mask) if cpu_int > current_max_cpu: current_max_cpu = cpu_int # End of for each trace line loop # Now seperate the main overall csv file into per CPU csv files. - split_csv() + split_csv(current_max_cpu, cpu_mask) def signal_handler(signal, frame): print(' SIGINT: Forcing cleanup before exit.') if interval: - disable_trace() + disable_trace(trace_file) clear_trace_file() # Free the memory free_trace_buffer() sys.exit(0) -signal.signal(signal.SIGINT, signal_handler) +if __name__ == "__main__": + trace_file = "/sys/kernel/debug/tracing/events/power/pstate_sample/enable" + signal.signal(signal.SIGINT, signal_handler) -interval = "" -filename = "" -cpu_list = "" -testname = "" -memory = "10240" -graph_data_present = False; + interval = "" + filename = "" + cpu_list = "" + testname = "" + memory = "10240" + graph_data_present = False; -valid1 = False -valid2 = False + valid1 = False + valid2 = False -cpu_mask = zeros((MAX_CPUS,), dtype=int) + cpu_mask = zeros((MAX_CPUS,), dtype=int) -try: - opts, args = getopt.getopt(sys.argv[1:],"ht:i:c:n:m:",["help","trace_file=","interval=","cpu=","name=","memory="]) -except getopt.GetoptError: - print_help() - sys.exit(2) -for opt, arg in opts: - if opt == '-h': - print() + try: + opts, args = getopt.getopt(sys.argv[1:],"ht:i:c:n:m:",["help","trace_file=","interval=","cpu=","name=","memory="]) + except getopt.GetoptError: + print_help('intel_pstate') + sys.exit(2) + for opt, arg in opts: + if opt == '-h': + print_help('intel_pstate') + sys.exit() + elif opt in ("-t", "--trace_file"): + valid1 = True + location = os.path.realpath(os.path.join(os.getcwd(), os.path.dirname(__file__))) + filename = os.path.join(location, arg) + elif opt in ("-i", "--interval"): + valid1 = True + interval = arg + elif opt in ("-c", "--cpu"): + cpu_list = arg + elif opt in ("-n", "--name"): + valid2 = True + testname = arg + elif opt in ("-m", "--memory"): + memory = arg + + if not (valid1 and valid2): + print_help('intel_pstate') sys.exit() - elif opt in ("-t", "--trace_file"): - valid1 = True - location = os.path.realpath(os.path.join(os.getcwd(), os.path.dirname(__file__))) - filename = os.path.join(location, arg) - elif opt in ("-i", "--interval"): - valid1 = True - interval = arg - elif opt in ("-c", "--cpu"): - cpu_list = arg - elif opt in ("-n", "--name"): - valid2 = True - testname = arg - elif opt in ("-m", "--memory"): - memory = arg - -if not (valid1 and valid2): - print_help() - sys.exit() - -if cpu_list: - for p in re.split("[,]", cpu_list): - if int(p) < MAX_CPUS : - cpu_mask[int(p)] = 1 -else: - for i in range (0, MAX_CPUS): - cpu_mask[i] = 1 - -if not os.path.exists('results'): - os.mkdir('results') + + if cpu_list: + for p in re.split("[,]", cpu_list): + if int(p) < MAX_CPUS : + cpu_mask[int(p)] = 1 + else: + for i in range (0, MAX_CPUS): + cpu_mask[i] = 1 + + if not os.path.exists('results'): + os.mkdir('results') + # The regular user needs to own the directory, not root. + fix_ownership('results') + + os.chdir('results') + if os.path.exists(testname): + print('The test name directory already exists. Please provide a unique test name. Test re-run not supported, yet.') + sys.exit() + os.mkdir(testname) # The regular user needs to own the directory, not root. - fix_ownership('results') - -os.chdir('results') -if os.path.exists(testname): - print('The test name directory already exists. Please provide a unique test name. Test re-run not supported, yet.') - sys.exit() -os.mkdir(testname) -# The regular user needs to own the directory, not root. -fix_ownership(testname) -os.chdir(testname) - -# Temporary (or perhaps not) -cur_version = sys.version_info -print('python version (should be >= 2.7):') -print(cur_version) - -# Left as "cleanup" for potential future re-run ability. -cleanup_data_files() - -if interval: - filename = "/sys/kernel/debug/tracing/trace" - clear_trace_file() - set_trace_buffer_size() - enable_trace() - print('Sleeping for ', interval, 'seconds') - time.sleep(int(interval)) - disable_trace() - -current_max_cpu = 0 - -read_trace_data(filename) - -if interval: - clear_trace_file() - # Free the memory - free_trace_buffer() - -if graph_data_present == False: - print('No valid data to plot') - sys.exit(2) - -for cpu_no in range(0, current_max_cpu + 1): - plot_perf_busy_with_sample(cpu_no) - plot_perf_busy(cpu_no) - plot_durations(cpu_no) - plot_loads(cpu_no) - -plot_pstate_cpu_with_sample() -plot_pstate_cpu() -plot_load_cpu() -plot_frequency_cpu() -plot_duration_cpu() -plot_scaled_cpu() -plot_boost_cpu() -plot_ghz_cpu() - -# It is preferrable, but not necessary, that the regular user owns the files, not root. -for root, dirs, files in os.walk('.'): - for f in files: - fix_ownership(f) - -os.chdir('../../') + fix_ownership(testname) + os.chdir(testname) + + # Temporary (or perhaps not) + cur_version = sys.version_info + print('python version (should be >= 2.7):') + print(cur_version) + + # Left as "cleanup" for potential future re-run ability. + cleanup_data_files() + + if interval: + filename = "/sys/kernel/debug/tracing/trace" + clear_trace_file() + set_trace_buffer_size(memory) + enable_trace(trace_file) + print('Sleeping for ', interval, 'seconds') + time.sleep(int(interval)) + disable_trace(trace_file) + + current_max_cpu = 0 + + read_trace_data(filename, cpu_mask) + + if interval: + clear_trace_file() + # Free the memory + free_trace_buffer() + + if graph_data_present == False: + print('No valid data to plot') + sys.exit(2) + + for cpu_no in range(0, current_max_cpu + 1): + plot_perf_busy_with_sample(cpu_no) + plot_perf_busy(cpu_no) + plot_durations(cpu_no) + plot_loads(cpu_no) + + plot_pstate_cpu_with_sample() + plot_pstate_cpu() + plot_load_cpu() + plot_frequency_cpu() + plot_duration_cpu() + plot_scaled_cpu() + plot_boost_cpu() + plot_ghz_cpu() + + # It is preferrable, but not necessary, that the regular user owns the files, not root. + for root, dirs, files in os.walk('.'): + for f in files: + fix_ownership(f) + + os.chdir('../../') diff --git a/tools/power/x86/turbostat/.gitignore b/tools/power/x86/turbostat/.gitignore index 7521370d3568..e13109b43cd1 100644 --- a/tools/power/x86/turbostat/.gitignore +++ b/tools/power/x86/turbostat/.gitignore @@ -1 +1,2 @@ +# SPDX-License-Identifier: GPL-2.0-only turbostat diff --git a/tools/power/x86/turbostat/Makefile b/tools/power/x86/turbostat/Makefile index 2b6551269e43..92e139b9c792 100644 --- a/tools/power/x86/turbostat/Makefile +++ b/tools/power/x86/turbostat/Makefile @@ -9,14 +9,15 @@ ifeq ("$(origin O)", "command line") endif turbostat : turbostat.c -override CFLAGS += -O2 -Wall -I../../../include +override CFLAGS += -O2 -Wall -Wextra -I../../../include override CFLAGS += -DMSRHEADER='"../../../../arch/x86/include/asm/msr-index.h"' override CFLAGS += -DINTEL_FAMILY_HEADER='"../../../../arch/x86/include/asm/intel-family.h"' +override CFLAGS += -D_FILE_OFFSET_BITS=64 override CFLAGS += -D_FORTIFY_SOURCE=2 %: %.c @mkdir -p $(BUILD_OUTPUT) - $(CC) $(CFLAGS) $< -o $(BUILD_OUTPUT)/$@ $(LDFLAGS) -lcap + $(CC) $(CFLAGS) $< -o $(BUILD_OUTPUT)/$@ $(LDFLAGS) -lcap -lrt .PHONY : clean clean : diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index a6db83a88e85..c7b26a3603af 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -54,12 +54,14 @@ name as necessary to disambiguate it from others is necessary. Note that option .PP \fB--cpu cpu-set\fP limit output to system summary plus the specified cpu-set. If cpu-set is the string "core", then the system summary plus the first CPU in each core are printed -- eg. subsequent HT siblings are not printed. Or if cpu-set is the string "package", then the system summary plus the first CPU in each package is printed. Otherwise, the system summary plus the specified set of CPUs are printed. The cpu-set is ordered from low to high, comma delimited with ".." and "-" permitted to denote a range. eg. 1,2,8,14..17,21-44 .PP -\fB--hide column\fP do not show the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. Use "--hide sysfs" to hide the sysfs statistics columns as a group. +\fB--hide column\fP do not show the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. .PP \fB--enable column\fP show the specified built-in columns, which are otherwise disabled, by default. Currently the only built-in counters disabled by default are "usec", "Time_Of_Day_Seconds", "APIC" and "X2APIC". The column name "all" can be used to enable all disabled-by-default built-in counters. .PP -\fB--show column\fP show only the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. Use "--show sysfs" to show the sysfs statistics columns as a group. +\fB--show column\fP show only the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. +.PP +\fB--show CATEGORY --hide CATEGORY\fP Show and hide also accept a single CATEGORY of columns: "all", "topology", "idle", "frequency", "power", "sysfs", "other". .PP \fB--Dump\fP displays the raw counter values. .PP @@ -90,40 +92,66 @@ displays the statistics gathered since it was forked. .SH ROW DESCRIPTIONS The system configuration dump (if --quiet is not used) is followed by statistics. The first row of the statistics labels the content of each column (below). The second row of statistics is the system summary line. The system summary line has a '-' in the columns for the Package, Core, and CPU. The contents of the system summary line depends on the type of column. Columns that count items (eg. IRQ) show the sum across all CPUs in the system. Columns that show a percentage show the average across all CPUs in the system. Columns that dump raw MSR values simply show 0 in the summary. After the system summary row, each row describes a specific Package/Core/CPU. Note that if the --cpu parameter is used to limit which specific CPUs are displayed, turbostat will still collect statistics for all CPUs in the system and will still show the system summary for all CPUs in the system. .SH COLUMN DESCRIPTIONS -.nf +.PP \fBusec\fP For each CPU, the number of microseconds elapsed during counter collection, including thread migration -- if any. This counter is disabled by default, and is enabled with "--enable usec", or --debug. On the summary row, usec refers to the total elapsed time to collect the counters on all cpus. +.PP \fBTime_Of_Day_Seconds\fP For each CPU, the gettimeofday(2) value (seconds.subsec since Epoch) when the counters ending the measurement interval were collected. This column is disabled by default, and can be enabled with "--enable Time_Of_Day_Seconds" or "--debug". On the summary row, Time_Of_Day_Seconds refers to the timestamp following collection of counters on the last CPU. +.PP \fBCore\fP processor core number. Note that multiple CPUs per core indicate support for Intel(R) Hyper-Threading Technology (HT). +.PP \fBCPU\fP Linux CPU (logical processor) number. Yes, it is okay that on many systems the CPUs are not listed in numerical order -- for efficiency reasons, turbostat runs in topology order, so HT siblings appear together. +.PP \fBPackage\fP processor package number -- not present on systems with a single processor package. +.PP \fBAvg_MHz\fP number of cycles executed divided by time elapsed. Note that this includes idle-time when 0 instructions are executed. +.PP \fBBusy%\fP percent of the measurement interval that the CPU executes instructions, aka. % of time in "C0" state. +.PP \fBBzy_MHz\fP average clock rate while the CPU was not idle (ie. in "c0" state). +.PP \fBTSC_MHz\fP average MHz that the TSC ran during the entire interval. +.PP \fBIRQ\fP The number of interrupts serviced by that CPU during the measurement interval. The system total line is the sum of interrupts serviced across all CPUs. turbostat parses /proc/interrupts to generate this summary. +.PP \fBSMI\fP The number of System Management Interrupts serviced CPU during the measurement interval. While this counter is actually per-CPU, SMI are triggered on all processors, so the number should be the same for all CPUs. +.PP \fBC1, C2, C3...\fP The number times Linux requested the C1, C2, C3 idle state during the measurement interval. The system summary line shows the sum for all CPUs. These are C-state names as exported in /sys/devices/system/cpu/cpu*/cpuidle/state*/name. While their names are generic, their attributes are processor specific. They the system description section of output shows what MWAIT sub-states they are mapped to on each system. +.PP \fBC1%, C2%, C3%\fP The residency percentage that Linux requested C1, C2, C3.... The system summary is the average of all CPUs in the system. Note that these are software, reflecting what was requested. The hardware counters reflect what was actually achieved. +.PP \fBCPU%c1, CPU%c3, CPU%c6, CPU%c7\fP show the percentage residency in hardware core idle states. These numbers are from hardware residency counters. +.PP \fBCoreTmp\fP Degrees Celsius reported by the per-core Digital Thermal Sensor. +.PP \fBPkgTmp\fP Degrees Celsius reported by the per-package Package Thermal Monitor. +.PP \fBGFX%rc6\fP The percentage of time the GPU is in the "render C6" state, rc6, during the measurement interval. From /sys/class/drm/card0/power/rc6_residency_ms. +.PP \fBGFXMHz\fP Instantaneous snapshot of what sysfs presents at the end of the measurement interval. From /sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz. +.PP \fBPkg%pc2, Pkg%pc3, Pkg%pc6, Pkg%pc7\fP percentage residency in hardware package idle states. These numbers are from hardware residency counters. +.PP \fBPkgWatt\fP Watts consumed by the whole package. +.PP \fBCorWatt\fP Watts consumed by the core part of the package. +.PP \fBGFXWatt\fP Watts consumed by the Graphics part of the package -- available only on client processors. +.PP \fBRAMWatt\fP Watts consumed by the DRAM DIMMS -- available only on server processors. +.PP \fBPKG_%\fP percent of the interval that RAPL throttling was active on the Package. Note that the system summary is the sum of the package throttling time, and thus may be higher than 100% on a multi-package system. Note that the meaning of this field is model specific. For example, some hardware increments this counter when RAPL responds to thermal limits, but does not increment this counter when RAPL responds to power limits. Comparing PkgWatt and PkgTmp to system limits is necessary. +.PP \fBRAM_%\fP percent of the interval that RAPL throttling was active on DRAM. -.fi +.PP +\fBUncMHz\fP uncore MHz, instantaneous sample. .SH TOO MUCH INFORMATION EXAMPLE By default, turbostat dumps all possible information -- a system configuration header, followed by columns for all counters. This is ideal for remote debugging, use the "--out" option to save everything to a text file, and get that file to the expert helping you debug. .PP When you are not interested in all that information, and there are several ways to see only what you want. First the "--quiet" option will skip the configuration information, and turbostat will show only the counter columns. Second, you can reduce the columns with the "--hide" and "--show" options. If you use the "--show" option, then turbostat will show only the columns you list. If you use the "--hide" option, turbostat will show all columns, except the ones you list. .PP -To find out what columns are available for --show and --hide, the "--list" option is available. For convenience, the special strings "sysfs" can be used to refer to all of the sysfs C-state counters at once: +To find out what columns are available for --show and --hide, the "--list" option is available. Usually, the CATEGORY names above are used to refer to groups of counters. Also, for convenience, the special string "sysfs" can be used to refer to all of the sysfs C-state counters at once: +.PP .nf sudo ./turbostat --show sysfs --quiet sleep 10 10.003837 sec @@ -156,32 +184,29 @@ Without a command to fork, turbostat displays statistics ever 5 seconds. Periodic output goes to stdout, by default, unless --out is used to specify an output file. The 5-second interval can be changed with the "-i sec" option. .nf -sudo ./turbostat --quiet --hide sysfs,IRQ,SMI,CoreTmp,PkgTmp,GFX%rc6,GFXMHz,PkgWatt,CorWatt,GFXWatt - Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c1 CPU%c3 CPU%c6 CPU%c7 - - - 488 12.52 3900 3498 12.50 0.00 0.00 74.98 - 0 0 5 0.13 3900 3498 99.87 0.00 0.00 0.00 - 0 4 3897 99.99 3900 3498 0.01 - 1 1 0 0.00 3856 3498 0.01 0.00 0.00 99.98 - 1 5 0 0.00 3861 3498 0.01 - 2 2 1 0.02 3889 3498 0.03 0.00 0.00 99.95 - 2 6 0 0.00 3863 3498 0.05 - 3 3 0 0.01 3869 3498 0.02 0.00 0.00 99.97 - 3 7 0 0.00 3878 3498 0.03 - Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c1 CPU%c3 CPU%c6 CPU%c7 - - - 491 12.59 3900 3498 12.42 0.00 0.00 74.99 - 0 0 27 0.69 3900 3498 99.31 0.00 0.00 0.00 - 0 4 3898 99.99 3900 3498 0.01 - 1 1 0 0.00 3883 3498 0.01 0.00 0.00 99.99 - 1 5 0 0.00 3898 3498 0.01 - 2 2 0 0.01 3889 3498 0.02 0.00 0.00 99.98 - 2 6 0 0.00 3889 3498 0.02 - 3 3 0 0.00 3856 3498 0.01 0.00 0.00 99.99 - 3 7 0 0.00 3897 3498 0.01 +sudo turbostat --quiet --show CPU,frequency + Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c7 UncMhz + - - 524 12.48 4198 3096 74.53 3800 + 0 0 4 0.09 4081 3096 98.88 3800 + 0 4 1 0.02 4063 3096 + 1 1 2 0.06 4063 3096 99.60 + 1 5 2 0.05 4070 3096 + 2 2 4178 99.52 4199 3096 0.00 + 2 6 3 0.08 4159 3096 + 3 3 1 0.04 4046 3096 99.66 + 3 7 0 0.01 3989 3096 + Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c7 UncMhz + - - 525 12.52 4198 3096 74.54 3800 + 0 0 4 0.10 4051 3096 99.49 3800 + 0 4 2 0.04 3993 3096 + 1 1 3 0.07 4054 3096 99.56 + 1 5 4 0.10 4018 3096 + 2 2 4178 99.51 4199 3096 0.00 + 2 6 4 0.09 4143 3096 + 3 3 2 0.06 4026 3096 99.10 + 3 7 7 0.17 4074 3096 .fi -This example also shows the use of the --hide option to skip columns that are not wanted. -Note that cpu4 in this example is 99.99% busy, while the other CPUs are all under 1% busy. -Notice that cpu4's HT sibling is cpu0, which is under 1% busy, but can get into CPU%c1 only, -because its cpu4's activity on shared hardware keeps it from entering a deeper C-state. +This example also shows the use of the --show option to show only the desired columns. .SH SYSTEM CONFIGURATION INFORMATION EXAMPLE @@ -189,61 +214,86 @@ By default, turbostat always dumps system configuration information before taking measurements. In the example above, "--quiet" is used to suppress that output. Here is an example of the configuration information: .nf -turbostat version 2017.02.15 - Len Brown <lenb@kernel.org> -CPUID(0): GenuineIntel 13 CPUID levels; family:model:stepping 0x6:3c:3 (6:60:3) -CPUID(1): SSE3 MONITOR - EIST TM2 TSC MSR ACPI-TM TM -CPUID(6): APERF, TURBO, DTS, PTM, No-HWP, No-HWPnotify, No-HWPwindow, No-HWPepp, No-HWPpkg, EPB -cpu4: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EIST No-MWAIT PREFETCH TURBO) -CPUID(7): No-SGX -cpu4: MSR_MISC_PWR_MGMT: 0x00400000 (ENable-EIST_Coordination DISable-EPB DISable-OOB) -RAPL: 3121 sec. Joule Counter Range, at 84 Watts -cpu4: MSR_PLATFORM_INFO: 0x80838f3012300 +turbostat version 2022.04.16 - Len Brown <lenb@kernel.org> +Kernel command line: BOOT_IMAGE=/boot/vmlinuz-5.18.0-rc6-00001-ge6891250e3b5 ... +CPUID(0): GenuineIntel 0x16 CPUID levels +CPUID(1): family:model:stepping 0x6:9e:9 (6:158:9) microcode 0xea +CPUID(0x80000000): max_extended_levels: 0x80000008 +CPUID(1): SSE3 MONITOR - EIST TM2 TSC MSR ACPI-TM HT TM +CPUID(6): APERF, TURBO, DTS, PTM, HWP, HWPnotify, HWPwindow, HWPepp, No-HWPpkg, EPB +cpu7: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EIST MWAIT PREFETCH TURBO) +CPUID(7): SGX +cpu7: MSR_IA32_FEATURE_CONTROL: 0x00000005 (Locked ) +CPUID(0x15): eax_crystal: 2 ebx_tsc: 258 ecx_crystal_hz: 0 +TSC: 3096 MHz (24000000 Hz * 258 / 2 / 1000000) +CPUID(0x16): base_mhz: 3100 max_mhz: 4200 bus_mhz: 100 +cpu7: MSR_MISC_PWR_MGMT: 0x00401cc0 (ENable-EIST_Coordination DISable-EPB DISable-OOB) +RAPL: 5825 sec. Joule Counter Range, at 45 Watts +cpu7: MSR_PLATFORM_INFO: 0x80839f1011f00 8 * 100.0 = 800.0 MHz max efficiency frequency -35 * 100.0 = 3500.0 MHz base frequency -cpu4: MSR_IA32_POWER_CTL: 0x0004005d (C1E auto-promotion: DISabled) -cpu4: MSR_TURBO_RATIO_LIMIT: 0x25262727 -37 * 100.0 = 3700.0 MHz max turbo 4 active cores -38 * 100.0 = 3800.0 MHz max turbo 3 active cores -39 * 100.0 = 3900.0 MHz max turbo 2 active cores -39 * 100.0 = 3900.0 MHz max turbo 1 active cores -cpu4: MSR_CONFIG_TDP_NOMINAL: 0x00000023 (base_ratio=35) -cpu4: MSR_CONFIG_TDP_LEVEL_1: 0x00000000 () -cpu4: MSR_CONFIG_TDP_LEVEL_2: 0x00000000 () -cpu4: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( lock=1) -cpu4: MSR_TURBO_ACTIVATION_RATIO: 0x00000000 (MAX_NON_TURBO_RATIO=0 lock=0) -cpu4: MSR_PKG_CST_CONFIG_CONTROL: 0x1e000400 (UNdemote-C3, UNdemote-C1, demote-C3, demote-C1, UNlocked: pkg-cstate-limit=0: pc0) -cpu4: POLL: CPUIDLE CORE POLL IDLE -cpu4: C1: MWAIT 0x00 -cpu4: C1E: MWAIT 0x01 -cpu4: C3: MWAIT 0x10 -cpu4: C6: MWAIT 0x20 -cpu4: C7s: MWAIT 0x32 -cpu4: MSR_MISC_FEATURE_CONTROL: 0x00000000 (L2-Prefetch L2-Prefetch-pair L1-Prefetch L1-IP-Prefetch) -cpu0: MSR_IA32_ENERGY_PERF_BIAS: 0x00000006 (balanced) -cpu0: MSR_CORE_PERF_LIMIT_REASONS, 0x31200000 (Active: ) (Logged: Transitions, MultiCoreTurbo, Amps, Auto-HWP, ) -cpu0: MSR_GFX_PERF_LIMIT_REASONS, 0x00000000 (Active: ) (Logged: ) -cpu0: MSR_RING_PERF_LIMIT_REASONS, 0x0d000000 (Active: ) (Logged: Amps, PkgPwrL1, PkgPwrL2, ) +31 * 100.0 = 3100.0 MHz base frequency +cpu7: MSR_IA32_POWER_CTL: 0x002c005d (C1E auto-promotion: DISabled) +cpu7: MSR_TURBO_RATIO_LIMIT: 0x2728292a +39 * 100.0 = 3900.0 MHz max turbo 4 active cores +40 * 100.0 = 4000.0 MHz max turbo 3 active cores +41 * 100.0 = 4100.0 MHz max turbo 2 active cores +42 * 100.0 = 4200.0 MHz max turbo 1 active cores +cpu7: MSR_CONFIG_TDP_NOMINAL: 0x0000001f (base_ratio=31) +cpu7: MSR_CONFIG_TDP_LEVEL_1: 0x00000000 () +cpu7: MSR_CONFIG_TDP_LEVEL_2: 0x00000000 () +cpu7: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( lock=1) +cpu7: MSR_TURBO_ACTIVATION_RATIO: 0x00000000 (MAX_NON_TURBO_RATIO=0 lock=0) +cpu7: MSR_PKG_CST_CONFIG_CONTROL: 0x1e008008 (UNdemote-C3, UNdemote-C1, demote-C3, demote-C1, locked, pkg-cstate-limit=8 (unlimited)) +Uncore Frequency pkg0 die0: 800 - 3900 MHz (800 - 3900 MHz) +/dev/cpu_dma_latency: 2000000000 usec (default) +current_driver: intel_idle +current_governor: menu +current_governor_ro: menu +cpu7: POLL: CPUIDLE CORE POLL IDLE +cpu7: C1: MWAIT 0x00 +cpu7: C1E: MWAIT 0x01 +cpu7: C3: MWAIT 0x10 +cpu7: C6: MWAIT 0x20 +cpu7: C7s: MWAIT 0x33 +cpu7: C8: MWAIT 0x40 +cpu7: C9: MWAIT 0x50 +cpu7: C10: MWAIT 0x60 +cpu7: cpufreq driver: intel_pstate +cpu7: cpufreq governor: performance +cpufreq intel_pstate no_turbo: 0 +cpu7: MSR_MISC_FEATURE_CONTROL: 0x00000000 (L2-Prefetch L2-Prefetch-pair L1-Prefetch L1-IP-Prefetch) +cpu0: MSR_PM_ENABLE: 0x00000001 (HWP) +cpu0: MSR_HWP_CAPABILITIES: 0x01101f53 (high 83 guar 31 eff 16 low 1) +cpu0: MSR_HWP_REQUEST: 0x00005353 (min 83 max 83 des 0 epp 0x0 window 0x0 pkg 0x0) +cpu0: MSR_HWP_INTERRUPT: 0x00000001 (EN_Guaranteed_Perf_Change, Dis_Excursion_Min) +cpu0: MSR_HWP_STATUS: 0x00000004 (No-Guaranteed_Perf_Change, No-Excursion_Min) +cpu0: EPB: 6 (balanced) cpu0: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.125000 Watts, 0.000061 Joules, 0.000977 sec.) -cpu0: MSR_PKG_POWER_INFO: 0x000002a0 (84 W TDP, RAPL 0 - 0 W, 0.000000 sec.) -cpu0: MSR_PKG_POWER_LIMIT: 0x428348001a82a0 (UNlocked) -cpu0: PKG Limit #1: ENabled (84.000000 Watts, 8.000000 sec, clamp DISabled) -cpu0: PKG Limit #2: ENabled (105.000000 Watts, 0.002441* sec, clamp DISabled) +cpu0: MSR_PKG_POWER_INFO: 0x00000168 (45 W TDP, RAPL 0 - 0 W, 0.000000 sec.) +cpu0: MSR_PKG_POWER_LIMIT: 0x42820800218208 (UNlocked) +cpu0: PKG Limit #1: ENabled (65.000 Watts, 64.000000 sec, clamp ENabled) +cpu0: PKG Limit #2: ENabled (65.000 Watts, 0.002441* sec, clamp DISabled) +cpu0: MSR_VR_CURRENT_CONFIG: 0x00000000 +cpu0: PKG Limit #4: 0.000000 Watts (UNlocked) +cpu0: MSR_DRAM_POWER_LIMIT: 0x5400de00000000 (UNlocked) +cpu0: DRAM Limit: DISabled (0.000 Watts, 0.000977 sec, clamp DISabled) cpu0: MSR_PP0_POLICY: 0 cpu0: MSR_PP0_POWER_LIMIT: 0x00000000 (UNlocked) -cpu0: Cores Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled) +cpu0: Cores Limit: DISabled (0.000 Watts, 0.000977 sec, clamp DISabled) cpu0: MSR_PP1_POLICY: 0 cpu0: MSR_PP1_POWER_LIMIT: 0x00000000 (UNlocked) -cpu0: GFX Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled) -cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00641400 (100 C) -cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x884c0800 (24 C) -cpu0: MSR_IA32_THERM_STATUS: 0x884c0000 (24 C +/- 1) -cpu1: MSR_IA32_THERM_STATUS: 0x88510000 (19 C +/- 1) -cpu2: MSR_IA32_THERM_STATUS: 0x884e0000 (22 C +/- 1) -cpu3: MSR_IA32_THERM_STATUS: 0x88510000 (19 C +/- 1) -cpu4: MSR_PKGC3_IRTL: 0x00008842 (valid, 67584 ns) -cpu4: MSR_PKGC6_IRTL: 0x00008873 (valid, 117760 ns) -cpu4: MSR_PKGC7_IRTL: 0x00008891 (valid, 148480 ns) +cpu0: GFX Limit: DISabled (0.000 Watts, 0.000977 sec, clamp DISabled) +cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00640000 (100 C) (100 default - 0 offset) +cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x88200800 (68 C) +cpu0: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x00000003 (100 C, 100 C) +cpu7: MSR_PKGC3_IRTL: 0x0000884e (valid, 79872 ns) +cpu7: MSR_PKGC6_IRTL: 0x00008876 (valid, 120832 ns) +cpu7: MSR_PKGC7_IRTL: 0x00008894 (valid, 151552 ns) +cpu7: MSR_PKGC8_IRTL: 0x000088fa (valid, 256000 ns) +cpu7: MSR_PKGC9_IRTL: 0x0000894c (valid, 339968 ns) +cpu7: MSR_PKGC10_IRTL: 0x00008bf2 (valid, 1034240 ns) .fi +.PP The \fBmax efficiency\fP frequency, a.k.a. Low Frequency Mode, is the frequency available at the minimum package voltage. The \fBTSC frequency\fP is the base frequency of the processor -- this should match the brand string @@ -290,7 +340,7 @@ starts a new interval. must be run as root. Alternatively, non-root users can be enabled to run turbostat this way: -# setcap cap_sys_rawio=ep ./turbostat +# setcap cap_sys_admin,cap_sys_rawio,cap_sys_nice=+ep ./turbostat # chmod +r /dev/cpu/*/msr @@ -335,7 +385,7 @@ that they count at TSC rate, which is true on all processors tested to date. .SH REFERENCES Volume 3B: System Programming Guide" -http://www.intel.com/products/processor/manuals/ +https://www.intel.com/products/processor/manuals/ .SH FILES .ta diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 33b370865d16..aba460410dbd 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -3,7 +3,7 @@ * turbostat -- show CPU frequency and C-state residency * on modern Intel and AMD processors. * - * Copyright (c) 2013 Intel Corporation. + * Copyright (c) 2022 Intel Corporation. * Len Brown <len.brown@intel.com> */ @@ -33,13 +33,189 @@ #include <sys/capability.h> #include <errno.h> #include <math.h> +#include <linux/perf_event.h> +#include <asm/unistd.h> +#include <stdbool.h> + +#define UNUSED(x) (void)(x) + +/* + * This list matches the column headers, except + * 1. built-in only, the sysfs counters are not here -- we learn of those at run-time + * 2. Core and CPU are moved to the end, we can't have strings that contain them + * matching on them for --show and --hide. + */ + +/* + * buffer size used by sscanf() for added column names + * Usually truncated to 7 characters, but also handles 18 columns for raw 64-bit counters + */ +#define NAME_BYTES 20 +#define PATH_BYTES 128 + +enum counter_scope { SCOPE_CPU, SCOPE_CORE, SCOPE_PACKAGE }; +enum counter_type { COUNTER_ITEMS, COUNTER_CYCLES, COUNTER_SECONDS, COUNTER_USEC }; +enum counter_format { FORMAT_RAW, FORMAT_DELTA, FORMAT_PERCENT }; + +struct msr_counter { + unsigned int msr_num; + char name[NAME_BYTES]; + char path[PATH_BYTES]; + unsigned int width; + enum counter_type type; + enum counter_format format; + struct msr_counter *next; + unsigned int flags; +#define FLAGS_HIDE (1 << 0) +#define FLAGS_SHOW (1 << 1) +#define SYSFS_PERCPU (1 << 1) +}; + +struct msr_counter bic[] = { + { 0x0, "usec", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Time_Of_Day_Seconds", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Package", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Node", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Avg_MHz", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Busy%", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Bzy_MHz", "", 0, 0, 0, NULL, 0 }, + { 0x0, "TSC_MHz", "", 0, 0, 0, NULL, 0 }, + { 0x0, "IRQ", "", 0, 0, 0, NULL, 0 }, + { 0x0, "SMI", "", 32, 0, FORMAT_DELTA, NULL, 0 }, + { 0x0, "sysfs", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPU%c1", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPU%c3", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPU%c6", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPU%c7", "", 0, 0, 0, NULL, 0 }, + { 0x0, "ThreadC", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CoreTmp", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CoreCnt", "", 0, 0, 0, NULL, 0 }, + { 0x0, "PkgTmp", "", 0, 0, 0, NULL, 0 }, + { 0x0, "GFX%rc6", "", 0, 0, 0, NULL, 0 }, + { 0x0, "GFXMHz", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg%pc2", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg%pc3", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg%pc6", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg%pc7", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg%pc8", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg%pc9", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pk%pc10", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPU%LPI", "", 0, 0, 0, NULL, 0 }, + { 0x0, "SYS%LPI", "", 0, 0, 0, NULL, 0 }, + { 0x0, "PkgWatt", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CorWatt", "", 0, 0, 0, NULL, 0 }, + { 0x0, "GFXWatt", "", 0, 0, 0, NULL, 0 }, + { 0x0, "PkgCnt", "", 0, 0, 0, NULL, 0 }, + { 0x0, "RAMWatt", "", 0, 0, 0, NULL, 0 }, + { 0x0, "PKG_%", "", 0, 0, 0, NULL, 0 }, + { 0x0, "RAM_%", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Pkg_J", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Cor_J", "", 0, 0, 0, NULL, 0 }, + { 0x0, "GFX_J", "", 0, 0, 0, NULL, 0 }, + { 0x0, "RAM_J", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Mod%c6", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Totl%C0", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Any%C0", "", 0, 0, 0, NULL, 0 }, + { 0x0, "GFX%C0", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPUGFX%", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Core", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CPU", "", 0, 0, 0, NULL, 0 }, + { 0x0, "APIC", "", 0, 0, 0, NULL, 0 }, + { 0x0, "X2APIC", "", 0, 0, 0, NULL, 0 }, + { 0x0, "Die", "", 0, 0, 0, NULL, 0 }, + { 0x0, "GFXAMHz", "", 0, 0, 0, NULL, 0 }, + { 0x0, "IPC", "", 0, 0, 0, NULL, 0 }, + { 0x0, "CoreThr", "", 0, 0, 0, NULL, 0 }, + { 0x0, "UncMHz", "", 0, 0, 0, NULL, 0 }, +}; + +#define MAX_BIC (sizeof(bic) / sizeof(struct msr_counter)) +#define BIC_USEC (1ULL << 0) +#define BIC_TOD (1ULL << 1) +#define BIC_Package (1ULL << 2) +#define BIC_Node (1ULL << 3) +#define BIC_Avg_MHz (1ULL << 4) +#define BIC_Busy (1ULL << 5) +#define BIC_Bzy_MHz (1ULL << 6) +#define BIC_TSC_MHz (1ULL << 7) +#define BIC_IRQ (1ULL << 8) +#define BIC_SMI (1ULL << 9) +#define BIC_sysfs (1ULL << 10) +#define BIC_CPU_c1 (1ULL << 11) +#define BIC_CPU_c3 (1ULL << 12) +#define BIC_CPU_c6 (1ULL << 13) +#define BIC_CPU_c7 (1ULL << 14) +#define BIC_ThreadC (1ULL << 15) +#define BIC_CoreTmp (1ULL << 16) +#define BIC_CoreCnt (1ULL << 17) +#define BIC_PkgTmp (1ULL << 18) +#define BIC_GFX_rc6 (1ULL << 19) +#define BIC_GFXMHz (1ULL << 20) +#define BIC_Pkgpc2 (1ULL << 21) +#define BIC_Pkgpc3 (1ULL << 22) +#define BIC_Pkgpc6 (1ULL << 23) +#define BIC_Pkgpc7 (1ULL << 24) +#define BIC_Pkgpc8 (1ULL << 25) +#define BIC_Pkgpc9 (1ULL << 26) +#define BIC_Pkgpc10 (1ULL << 27) +#define BIC_CPU_LPI (1ULL << 28) +#define BIC_SYS_LPI (1ULL << 29) +#define BIC_PkgWatt (1ULL << 30) +#define BIC_CorWatt (1ULL << 31) +#define BIC_GFXWatt (1ULL << 32) +#define BIC_PkgCnt (1ULL << 33) +#define BIC_RAMWatt (1ULL << 34) +#define BIC_PKG__ (1ULL << 35) +#define BIC_RAM__ (1ULL << 36) +#define BIC_Pkg_J (1ULL << 37) +#define BIC_Cor_J (1ULL << 38) +#define BIC_GFX_J (1ULL << 39) +#define BIC_RAM_J (1ULL << 40) +#define BIC_Mod_c6 (1ULL << 41) +#define BIC_Totl_c0 (1ULL << 42) +#define BIC_Any_c0 (1ULL << 43) +#define BIC_GFX_c0 (1ULL << 44) +#define BIC_CPUGFX (1ULL << 45) +#define BIC_Core (1ULL << 46) +#define BIC_CPU (1ULL << 47) +#define BIC_APIC (1ULL << 48) +#define BIC_X2APIC (1ULL << 49) +#define BIC_Die (1ULL << 50) +#define BIC_GFXACTMHz (1ULL << 51) +#define BIC_IPC (1ULL << 52) +#define BIC_CORE_THROT_CNT (1ULL << 53) +#define BIC_UNCORE_MHZ (1ULL << 54) + +#define BIC_TOPOLOGY (BIC_Package | BIC_Node | BIC_CoreCnt | BIC_PkgCnt | BIC_Core | BIC_CPU | BIC_Die ) +#define BIC_THERMAL_PWR ( BIC_CoreTmp | BIC_PkgTmp | BIC_PkgWatt | BIC_CorWatt | BIC_GFXWatt | BIC_RAMWatt | BIC_PKG__ | BIC_RAM__) +#define BIC_FREQUENCY ( BIC_Avg_MHz | BIC_Busy | BIC_Bzy_MHz | BIC_TSC_MHz | BIC_GFXMHz | BIC_GFXACTMHz | BIC_UNCORE_MHZ) +#define BIC_IDLE ( BIC_sysfs | BIC_CPU_c1 | BIC_CPU_c3 | BIC_CPU_c6 | BIC_CPU_c7 | BIC_GFX_rc6 | BIC_Pkgpc2 | BIC_Pkgpc3 | BIC_Pkgpc6 | BIC_Pkgpc7 | BIC_Pkgpc8 | BIC_Pkgpc9 | BIC_Pkgpc10 | BIC_CPU_LPI | BIC_SYS_LPI | BIC_Mod_c6 | BIC_Totl_c0 | BIC_Any_c0 | BIC_GFX_c0 | BIC_CPUGFX) +#define BIC_OTHER ( BIC_IRQ | BIC_SMI | BIC_ThreadC | BIC_CoreTmp | BIC_IPC) + +#define BIC_DISABLED_BY_DEFAULT (BIC_USEC | BIC_TOD | BIC_APIC | BIC_X2APIC) + +unsigned long long bic_enabled = (0xFFFFFFFFFFFFFFFFULL & ~BIC_DISABLED_BY_DEFAULT); +unsigned long long bic_present = BIC_USEC | BIC_TOD | BIC_sysfs | BIC_APIC | BIC_X2APIC; + +#define DO_BIC(COUNTER_NAME) (bic_enabled & bic_present & COUNTER_NAME) +#define DO_BIC_READ(COUNTER_NAME) (bic_present & COUNTER_NAME) +#define ENABLE_BIC(COUNTER_NAME) (bic_enabled |= COUNTER_NAME) +#define BIC_PRESENT(COUNTER_BIT) (bic_present |= COUNTER_BIT) +#define BIC_NOT_PRESENT(COUNTER_BIT) (bic_present &= ~COUNTER_BIT) +#define BIC_IS_ENABLED(COUNTER_BIT) (bic_enabled & COUNTER_BIT) char *proc_stat = "/proc/stat"; FILE *outf; int *fd_percpu; -struct timeval interval_tv = {5, 0}; -struct timespec interval_ts = {5, 0}; +int *fd_instr_count_percpu; +struct timeval interval_tv = { 5, 0 }; +struct timespec interval_ts = { 5, 0 }; + +/* Save original CPU model */ +unsigned int model_orig; + unsigned int num_iterations; +unsigned int header_iterations; unsigned int debug; unsigned int quiet; unsigned int shown; @@ -54,6 +230,8 @@ unsigned int do_slm_cstates; unsigned int use_c1_residency_msr; unsigned int has_aperf; unsigned int has_epb; +unsigned int has_turbo; +unsigned int is_hybrid; unsigned int do_irtl_snb; unsigned int do_irtl_hsw; unsigned int units = 1000000; /* MHz etc */ @@ -75,29 +253,33 @@ char *output_buffer, *outp; unsigned int do_rapl; unsigned int do_dts; unsigned int do_ptm; -unsigned long long gfx_cur_rc6_ms; +unsigned int do_ipc; +unsigned long long gfx_cur_rc6_ms; unsigned long long cpuidle_cur_cpu_lpi_us; unsigned long long cpuidle_cur_sys_lpi_us; unsigned int gfx_cur_mhz; -unsigned int tcc_activation_temp; -unsigned int tcc_activation_temp_override; +unsigned int gfx_act_mhz; +unsigned int tj_max; +unsigned int tj_max_override; +int tcc_offset_bits; double rapl_power_units, rapl_time_units; double rapl_dram_energy_units, rapl_energy_units; double rapl_joule_counter_range; unsigned int do_core_perf_limit_reasons; unsigned int has_automatic_cstate_conversion; +unsigned int dis_cstate_prewake; unsigned int do_gfx_perf_limit_reasons; unsigned int do_ring_perf_limit_reasons; unsigned int crystal_hz; unsigned long long tsc_hz; int base_cpu; double discover_bclk(unsigned int family, unsigned int model); -unsigned int has_hwp; /* IA32_PM_ENABLE, IA32_HWP_CAPABILITIES */ +unsigned int has_hwp; /* IA32_PM_ENABLE, IA32_HWP_CAPABILITIES */ /* IA32_HWP_REQUEST, IA32_HWP_STATUS */ -unsigned int has_hwp_notify; /* IA32_HWP_INTERRUPT */ +unsigned int has_hwp_notify; /* IA32_HWP_INTERRUPT */ unsigned int has_hwp_activity_window; /* IA32_HWP_REQUEST[bits 41:32] */ -unsigned int has_hwp_epp; /* IA32_HWP_REQUEST[bits 31:24] */ -unsigned int has_hwp_pkg; /* IA32_HWP_REQUEST_PKG */ +unsigned int has_hwp_epp; /* IA32_HWP_REQUEST[bits 31:24] */ +unsigned int has_hwp_pkg; /* IA32_HWP_REQUEST_PKG */ unsigned int has_misc_feature_control; unsigned int first_counter_read = 1; int ignore_stdin; @@ -147,13 +329,6 @@ int ignore_stdin; #define MAX(a, b) ((a) > (b) ? (a) : (b)) -/* - * buffer size used by sscanf() for added column names - * Usually truncated to 7 characters, but also handles 18 columns for raw 64-bit counters - */ -#define NAME_BYTES 20 -#define PATH_BYTES 128 - int backwards_count; char *progname; @@ -172,12 +347,14 @@ struct thread_data { unsigned long long aperf; unsigned long long mperf; unsigned long long c1; - unsigned long long irq_count; + unsigned long long instr_count; + unsigned long long irq_count; unsigned int smi_count; unsigned int cpu_id; unsigned int apic_id; unsigned int x2apic_id; unsigned int flags; + bool is_atom; #define CPU_IS_FIRST_THREAD_IN_CORE 0x2 #define CPU_IS_FIRST_CORE_IN_PACKAGE 0x4 unsigned long long counter[MAX_ADDED_THREAD_COUNTERS]; @@ -191,6 +368,7 @@ struct core_data { unsigned int core_temp_c; unsigned int core_energy; /* MSR_CORE_ENERGY_STAT */ unsigned int core_id; + unsigned long long core_throt_cnt; unsigned long long counter[MAX_ADDED_COUNTERS]; } *core_even, *core_odd; @@ -210,14 +388,16 @@ struct pkg_data { unsigned long long pkg_both_core_gfxe_c0; long long gfx_rc6_ms; unsigned int gfx_mhz; + unsigned int gfx_act_mhz; unsigned int package_id; - unsigned int energy_pkg; /* MSR_PKG_ENERGY_STATUS */ - unsigned int energy_dram; /* MSR_DRAM_ENERGY_STATUS */ - unsigned int energy_cores; /* MSR_PP0_ENERGY_STATUS */ - unsigned int energy_gfx; /* MSR_PP1_ENERGY_STATUS */ - unsigned int rapl_pkg_perf_status; /* MSR_PKG_PERF_STATUS */ - unsigned int rapl_dram_perf_status; /* MSR_DRAM_PERF_STATUS */ + unsigned long long energy_pkg; /* MSR_PKG_ENERGY_STATUS */ + unsigned long long energy_dram; /* MSR_DRAM_ENERGY_STATUS */ + unsigned long long energy_cores; /* MSR_PP0_ENERGY_STATUS */ + unsigned long long energy_gfx; /* MSR_PP1_ENERGY_STATUS */ + unsigned long long rapl_pkg_perf_status; /* MSR_PKG_PERF_STATUS */ + unsigned long long rapl_dram_perf_status; /* MSR_DRAM_PERF_STATUS */ unsigned int pkg_temp_c; + unsigned int uncore_mhz; unsigned long long counter[MAX_ADDED_COUNTERS]; } *package_even, *package_odd; @@ -238,27 +418,120 @@ struct pkg_data { ((node_no) * topo.cores_per_node) + \ (core_no)) - #define GET_PKG(pkg_base, pkg_no) (pkg_base + pkg_no) -enum counter_scope {SCOPE_CPU, SCOPE_CORE, SCOPE_PACKAGE}; -enum counter_type {COUNTER_ITEMS, COUNTER_CYCLES, COUNTER_SECONDS, COUNTER_USEC}; -enum counter_format {FORMAT_RAW, FORMAT_DELTA, FORMAT_PERCENT}; +/* + * The accumulated sum of MSR is defined as a monotonic + * increasing MSR, it will be accumulated periodically, + * despite its register's bit width. + */ +enum { + IDX_PKG_ENERGY, + IDX_DRAM_ENERGY, + IDX_PP0_ENERGY, + IDX_PP1_ENERGY, + IDX_PKG_PERF, + IDX_DRAM_PERF, + IDX_COUNT, +}; -struct msr_counter { - unsigned int msr_num; - char name[NAME_BYTES]; - char path[PATH_BYTES]; - unsigned int width; - enum counter_type type; - enum counter_format format; - struct msr_counter *next; - unsigned int flags; -#define FLAGS_HIDE (1 << 0) -#define FLAGS_SHOW (1 << 1) -#define SYSFS_PERCPU (1 << 1) +int get_msr_sum(int cpu, off_t offset, unsigned long long *msr); + +struct msr_sum_array { + /* get_msr_sum() = sum + (get_msr() - last) */ + struct { + /*The accumulated MSR value is updated by the timer */ + unsigned long long sum; + /*The MSR footprint recorded in last timer */ + unsigned long long last; + } entries[IDX_COUNT]; }; +/* The percpu MSR sum array.*/ +struct msr_sum_array *per_cpu_msr_sum; + +off_t idx_to_offset(int idx) +{ + off_t offset; + + switch (idx) { + case IDX_PKG_ENERGY: + if (do_rapl & RAPL_AMD_F17H) + offset = MSR_PKG_ENERGY_STAT; + else + offset = MSR_PKG_ENERGY_STATUS; + break; + case IDX_DRAM_ENERGY: + offset = MSR_DRAM_ENERGY_STATUS; + break; + case IDX_PP0_ENERGY: + offset = MSR_PP0_ENERGY_STATUS; + break; + case IDX_PP1_ENERGY: + offset = MSR_PP1_ENERGY_STATUS; + break; + case IDX_PKG_PERF: + offset = MSR_PKG_PERF_STATUS; + break; + case IDX_DRAM_PERF: + offset = MSR_DRAM_PERF_STATUS; + break; + default: + offset = -1; + } + return offset; +} + +int offset_to_idx(off_t offset) +{ + int idx; + + switch (offset) { + case MSR_PKG_ENERGY_STATUS: + case MSR_PKG_ENERGY_STAT: + idx = IDX_PKG_ENERGY; + break; + case MSR_DRAM_ENERGY_STATUS: + idx = IDX_DRAM_ENERGY; + break; + case MSR_PP0_ENERGY_STATUS: + idx = IDX_PP0_ENERGY; + break; + case MSR_PP1_ENERGY_STATUS: + idx = IDX_PP1_ENERGY; + break; + case MSR_PKG_PERF_STATUS: + idx = IDX_PKG_PERF; + break; + case MSR_DRAM_PERF_STATUS: + idx = IDX_DRAM_PERF; + break; + default: + idx = -1; + } + return idx; +} + +int idx_valid(int idx) +{ + switch (idx) { + case IDX_PKG_ENERGY: + return do_rapl & (RAPL_PKG | RAPL_AMD_F17H); + case IDX_DRAM_ENERGY: + return do_rapl & RAPL_DRAM; + case IDX_PP0_ENERGY: + return do_rapl & RAPL_CORES_ENERGY_STATUS; + case IDX_PP1_ENERGY: + return do_rapl & RAPL_GFX; + case IDX_PKG_PERF: + return do_rapl & RAPL_PKG_PERF_STATUS; + case IDX_DRAM_PERF: + return do_rapl & RAPL_DRAM_PERF_STATUS; + default: + return 0; + } +} + struct sys_counters { unsigned int added_thread_counters; unsigned int added_core_counters; @@ -282,7 +555,7 @@ struct cpu_topology { int logical_node_id; /* 0-based count within the package */ int physical_core_id; int thread_id; - cpu_set_t *put_ids; /* Processing Unit/Thread IDs */ + cpu_set_t *put_ids; /* Processing Unit/Thread IDs */ } *cpus; struct topo_params { @@ -299,7 +572,7 @@ struct topo_params { struct timeval tv_even, tv_odd, tv_delta; -int *irq_column_2_cpu; /* /proc/interrupts column numbers */ +int *irq_column_2_cpu; /* /proc/interrupts column numbers */ int *irqs_per_cpu; /* indexed by cpu_num */ void setup_all_buffers(void); @@ -312,34 +585,31 @@ int cpu_is_not_present(int cpu) { return !CPU_ISSET_S(cpu, cpu_present_setsize, cpu_present_set); } + /* * run func(thread, core, package) in topology order * skip non-present cpus */ -int for_all_cpus(int (func)(struct thread_data *, struct core_data *, struct pkg_data *), - struct thread_data *thread_base, struct core_data *core_base, struct pkg_data *pkg_base) +int for_all_cpus(int (func) (struct thread_data *, struct core_data *, struct pkg_data *), + struct thread_data *thread_base, struct core_data *core_base, struct pkg_data *pkg_base) { int retval, pkg_no, core_no, thread_no, node_no; for (pkg_no = 0; pkg_no < topo.num_packages; ++pkg_no) { for (node_no = 0; node_no < topo.nodes_per_pkg; node_no++) { for (core_no = 0; core_no < topo.cores_per_node; ++core_no) { - for (thread_no = 0; thread_no < - topo.threads_per_core; ++thread_no) { + for (thread_no = 0; thread_no < topo.threads_per_core; ++thread_no) { struct thread_data *t; struct core_data *c; struct pkg_data *p; - t = GET_THREAD(thread_base, thread_no, - core_no, node_no, - pkg_no); + t = GET_THREAD(thread_base, thread_no, core_no, node_no, pkg_no); if (cpu_is_not_present(t->cpu_id)) continue; - c = GET_CORE(core_base, core_no, - node_no, pkg_no); + c = GET_CORE(core_base, core_no, node_no, pkg_no); p = GET_PKG(pkg_base, pkg_no); retval = func(t, c, p); @@ -361,6 +631,7 @@ int cpu_migrate(int cpu) else return 0; } + int get_msr_fd(int cpu) { char pathname[32]; @@ -381,145 +652,57 @@ int get_msr_fd(int cpu) return fd; } -int get_msr(int cpu, off_t offset, unsigned long long *msr) +static long perf_event_open(struct perf_event_attr *hw_event, pid_t pid, int cpu, int group_fd, unsigned long flags) { - ssize_t retval; + return syscall(__NR_perf_event_open, hw_event, pid, cpu, group_fd, flags); +} - retval = pread(get_msr_fd(cpu), msr, sizeof(*msr), offset); +static int perf_instr_count_open(int cpu_num) +{ + struct perf_event_attr pea; + int fd; - if (retval != sizeof *msr) - err(-1, "cpu%d: msr offset 0x%llx read failed", cpu, (unsigned long long)offset); + memset(&pea, 0, sizeof(struct perf_event_attr)); + pea.type = PERF_TYPE_HARDWARE; + pea.size = sizeof(struct perf_event_attr); + pea.config = PERF_COUNT_HW_INSTRUCTIONS; - return 0; + /* counter for cpu_num, including user + kernel and all processes */ + fd = perf_event_open(&pea, -1, cpu_num, -1, 0); + if (fd == -1) { + warn("cpu%d: perf instruction counter", cpu_num); + BIC_NOT_PRESENT(BIC_IPC); + } + + return fd; } -/* - * This list matches the column headers, except - * 1. built-in only, the sysfs counters are not here -- we learn of those at run-time - * 2. Core and CPU are moved to the end, we can't have strings that contain them - * matching on them for --show and --hide. - */ -struct msr_counter bic[] = { - { 0x0, "usec" }, - { 0x0, "Time_Of_Day_Seconds" }, - { 0x0, "Package" }, - { 0x0, "Node" }, - { 0x0, "Avg_MHz" }, - { 0x0, "Busy%" }, - { 0x0, "Bzy_MHz" }, - { 0x0, "TSC_MHz" }, - { 0x0, "IRQ" }, - { 0x0, "SMI", "", 32, 0, FORMAT_DELTA, NULL}, - { 0x0, "sysfs" }, - { 0x0, "CPU%c1" }, - { 0x0, "CPU%c3" }, - { 0x0, "CPU%c6" }, - { 0x0, "CPU%c7" }, - { 0x0, "ThreadC" }, - { 0x0, "CoreTmp" }, - { 0x0, "CoreCnt" }, - { 0x0, "PkgTmp" }, - { 0x0, "GFX%rc6" }, - { 0x0, "GFXMHz" }, - { 0x0, "Pkg%pc2" }, - { 0x0, "Pkg%pc3" }, - { 0x0, "Pkg%pc6" }, - { 0x0, "Pkg%pc7" }, - { 0x0, "Pkg%pc8" }, - { 0x0, "Pkg%pc9" }, - { 0x0, "Pk%pc10" }, - { 0x0, "CPU%LPI" }, - { 0x0, "SYS%LPI" }, - { 0x0, "PkgWatt" }, - { 0x0, "CorWatt" }, - { 0x0, "GFXWatt" }, - { 0x0, "PkgCnt" }, - { 0x0, "RAMWatt" }, - { 0x0, "PKG_%" }, - { 0x0, "RAM_%" }, - { 0x0, "Pkg_J" }, - { 0x0, "Cor_J" }, - { 0x0, "GFX_J" }, - { 0x0, "RAM_J" }, - { 0x0, "Mod%c6" }, - { 0x0, "Totl%C0" }, - { 0x0, "Any%C0" }, - { 0x0, "GFX%C0" }, - { 0x0, "CPUGFX%" }, - { 0x0, "Core" }, - { 0x0, "CPU" }, - { 0x0, "APIC" }, - { 0x0, "X2APIC" }, - { 0x0, "Die" }, -}; +int get_instr_count_fd(int cpu) +{ + if (fd_instr_count_percpu[cpu]) + return fd_instr_count_percpu[cpu]; -#define MAX_BIC (sizeof(bic) / sizeof(struct msr_counter)) -#define BIC_USEC (1ULL << 0) -#define BIC_TOD (1ULL << 1) -#define BIC_Package (1ULL << 2) -#define BIC_Node (1ULL << 3) -#define BIC_Avg_MHz (1ULL << 4) -#define BIC_Busy (1ULL << 5) -#define BIC_Bzy_MHz (1ULL << 6) -#define BIC_TSC_MHz (1ULL << 7) -#define BIC_IRQ (1ULL << 8) -#define BIC_SMI (1ULL << 9) -#define BIC_sysfs (1ULL << 10) -#define BIC_CPU_c1 (1ULL << 11) -#define BIC_CPU_c3 (1ULL << 12) -#define BIC_CPU_c6 (1ULL << 13) -#define BIC_CPU_c7 (1ULL << 14) -#define BIC_ThreadC (1ULL << 15) -#define BIC_CoreTmp (1ULL << 16) -#define BIC_CoreCnt (1ULL << 17) -#define BIC_PkgTmp (1ULL << 18) -#define BIC_GFX_rc6 (1ULL << 19) -#define BIC_GFXMHz (1ULL << 20) -#define BIC_Pkgpc2 (1ULL << 21) -#define BIC_Pkgpc3 (1ULL << 22) -#define BIC_Pkgpc6 (1ULL << 23) -#define BIC_Pkgpc7 (1ULL << 24) -#define BIC_Pkgpc8 (1ULL << 25) -#define BIC_Pkgpc9 (1ULL << 26) -#define BIC_Pkgpc10 (1ULL << 27) -#define BIC_CPU_LPI (1ULL << 28) -#define BIC_SYS_LPI (1ULL << 29) -#define BIC_PkgWatt (1ULL << 30) -#define BIC_CorWatt (1ULL << 31) -#define BIC_GFXWatt (1ULL << 32) -#define BIC_PkgCnt (1ULL << 33) -#define BIC_RAMWatt (1ULL << 34) -#define BIC_PKG__ (1ULL << 35) -#define BIC_RAM__ (1ULL << 36) -#define BIC_Pkg_J (1ULL << 37) -#define BIC_Cor_J (1ULL << 38) -#define BIC_GFX_J (1ULL << 39) -#define BIC_RAM_J (1ULL << 40) -#define BIC_Mod_c6 (1ULL << 41) -#define BIC_Totl_c0 (1ULL << 42) -#define BIC_Any_c0 (1ULL << 43) -#define BIC_GFX_c0 (1ULL << 44) -#define BIC_CPUGFX (1ULL << 45) -#define BIC_Core (1ULL << 46) -#define BIC_CPU (1ULL << 47) -#define BIC_APIC (1ULL << 48) -#define BIC_X2APIC (1ULL << 49) -#define BIC_Die (1ULL << 50) + fd_instr_count_percpu[cpu] = perf_instr_count_open(cpu); -#define BIC_DISABLED_BY_DEFAULT (BIC_USEC | BIC_TOD | BIC_APIC | BIC_X2APIC) + return fd_instr_count_percpu[cpu]; +} -unsigned long long bic_enabled = (0xFFFFFFFFFFFFFFFFULL & ~BIC_DISABLED_BY_DEFAULT); -unsigned long long bic_present = BIC_USEC | BIC_TOD | BIC_sysfs | BIC_APIC | BIC_X2APIC; +int get_msr(int cpu, off_t offset, unsigned long long *msr) +{ + ssize_t retval; -#define DO_BIC(COUNTER_NAME) (bic_enabled & bic_present & COUNTER_NAME) -#define DO_BIC_READ(COUNTER_NAME) (bic_present & COUNTER_NAME) -#define ENABLE_BIC(COUNTER_NAME) (bic_enabled |= COUNTER_NAME) -#define BIC_PRESENT(COUNTER_BIT) (bic_present |= COUNTER_BIT) -#define BIC_NOT_PRESENT(COUNTER_BIT) (bic_present &= ~COUNTER_BIT) + retval = pread(get_msr_fd(cpu), msr, sizeof(*msr), offset); + if (retval != sizeof *msr) + err(-1, "cpu%d: msr offset 0x%llx read failed", cpu, (unsigned long long)offset); + + return 0; +} #define MAX_DEFERRED 16 +char *deferred_add_names[MAX_DEFERRED]; char *deferred_skip_names[MAX_DEFERRED]; +int deferred_add_index; int deferred_skip_index; /* @@ -531,42 +714,42 @@ enum show_hide_mode { SHOW_LIST, HIDE_LIST } global_show_hide_mode = HIDE_LIST; void help(void) { fprintf(outf, - "Usage: turbostat [OPTIONS][(--interval seconds) | COMMAND ...]\n" - "\n" - "Turbostat forks the specified COMMAND and prints statistics\n" - "when COMMAND completes.\n" - "If no COMMAND is specified, turbostat wakes every 5-seconds\n" - "to print statistics, until interrupted.\n" - " -a, --add add a counter\n" - " eg. --add msr0x10,u64,cpu,delta,MY_TSC\n" - " -c, --cpu cpu-set limit output to summary plus cpu-set:\n" - " {core | package | j,k,l..m,n-p }\n" - " -d, --debug displays usec, Time_Of_Day_Seconds and more debugging\n" - " -D, --Dump displays the raw counter values\n" - " -e, --enable [all | column]\n" - " shows all or the specified disabled column\n" - " -H, --hide [column|column,column,...]\n" - " hide the specified column(s)\n" - " -i, --interval sec.subsec\n" - " Override default 5-second measurement interval\n" - " -J, --Joules displays energy in Joules instead of Watts\n" - " -l, --list list column headers only\n" - " -n, --num_iterations num\n" - " number of the measurement iterations\n" - " -o, --out file\n" - " create or truncate \"file\" for all output\n" - " -q, --quiet skip decoding system configuration header\n" - " -s, --show [column|column,column,...]\n" - " show only the specified column(s)\n" - " -S, --Summary\n" - " limits output to 1-line system summary per interval\n" - " -T, --TCC temperature\n" - " sets the Thermal Control Circuit temperature in\n" - " degrees Celsius\n" - " -h, --help print this help message\n" - " -v, --version print version information\n" - "\n" - "For more help, run \"man turbostat\"\n"); + "Usage: turbostat [OPTIONS][(--interval seconds) | COMMAND ...]\n" + "\n" + "Turbostat forks the specified COMMAND and prints statistics\n" + "when COMMAND completes.\n" + "If no COMMAND is specified, turbostat wakes every 5-seconds\n" + "to print statistics, until interrupted.\n" + " -a, --add add a counter\n" + " eg. --add msr0x10,u64,cpu,delta,MY_TSC\n" + " -c, --cpu cpu-set limit output to summary plus cpu-set:\n" + " {core | package | j,k,l..m,n-p }\n" + " -d, --debug displays usec, Time_Of_Day_Seconds and more debugging\n" + " -D, --Dump displays the raw counter values\n" + " -e, --enable [all | column]\n" + " shows all or the specified disabled column\n" + " -H, --hide [column|column,column,...]\n" + " hide the specified column(s)\n" + " -i, --interval sec.subsec\n" + " Override default 5-second measurement interval\n" + " -J, --Joules displays energy in Joules instead of Watts\n" + " -l, --list list column headers only\n" + " -n, --num_iterations num\n" + " number of the measurement iterations\n" + " -N, --header_iterations num\n" + " print header every num iterations\n" + " -o, --out file\n" + " create or truncate \"file\" for all output\n" + " -q, --quiet skip decoding system configuration header\n" + " -s, --show [column|column,column,...]\n" + " show only the specified column(s)\n" + " -S, --Summary\n" + " limits output to 1-line system summary per interval\n" + " -T, --TCC temperature\n" + " sets the Thermal Control Circuit temperature in\n" + " degrees Celsius\n" + " -h, --help print this help message\n" + " -v, --version print version information\n" "\n" "For more help, run \"man turbostat\"\n"); } /* @@ -576,7 +759,7 @@ void help(void) */ unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) { - int i; + unsigned int i; unsigned long long retval = 0; while (name_list) { @@ -587,28 +770,51 @@ unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) if (comma) *comma = '\0'; - if (!strcmp(name_list, "all")) - return ~0; - for (i = 0; i < MAX_BIC; ++i) { if (!strcmp(name_list, bic[i].name)) { retval |= (1ULL << i); break; } + if (!strcmp(name_list, "all")) { + retval |= ~0; + break; + } else if (!strcmp(name_list, "topology")) { + retval |= BIC_TOPOLOGY; + break; + } else if (!strcmp(name_list, "power")) { + retval |= BIC_THERMAL_PWR; + break; + } else if (!strcmp(name_list, "idle")) { + retval |= BIC_IDLE; + break; + } else if (!strcmp(name_list, "frequency")) { + retval |= BIC_FREQUENCY; + break; + } else if (!strcmp(name_list, "other")) { + retval |= BIC_OTHER; + break; + } + } if (i == MAX_BIC) { if (mode == SHOW_LIST) { - fprintf(stderr, "Invalid counter name: %s\n", name_list); - exit(-1); - } - deferred_skip_names[deferred_skip_index++] = name_list; - if (debug) - fprintf(stderr, "deferred \"%s\"\n", name_list); - if (deferred_skip_index >= MAX_DEFERRED) { - fprintf(stderr, "More than max %d un-recognized --skip options '%s'\n", - MAX_DEFERRED, name_list); - help(); - exit(1); + deferred_add_names[deferred_add_index++] = name_list; + if (deferred_add_index >= MAX_DEFERRED) { + fprintf(stderr, "More than max %d un-recognized --add options '%s'\n", + MAX_DEFERRED, name_list); + help(); + exit(1); + } + } else { + deferred_skip_names[deferred_skip_index++] = name_list; + if (debug) + fprintf(stderr, "deferred \"%s\"\n", name_list); + if (deferred_skip_index >= MAX_DEFERRED) { + fprintf(stderr, "More than max %d un-recognized --skip options '%s'\n", + MAX_DEFERRED, name_list); + help(); + exit(1); + } } } @@ -620,7 +826,6 @@ unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) return retval; } - void print_header(char *delim) { struct msr_counter *mp; @@ -653,6 +858,9 @@ void print_header(char *delim) if (DO_BIC(BIC_TSC_MHz)) outp += sprintf(outp, "%sTSC_MHz", (printed++ ? delim : "")); + if (DO_BIC(BIC_IPC)) + outp += sprintf(outp, "%sIPC", (printed++ ? delim : "")); + if (DO_BIC(BIC_IRQ)) { if (sums_need_wide_columns) outp += sprintf(outp, "%s IRQ", (printed++ ? delim : "")); @@ -693,6 +901,9 @@ void print_header(char *delim) if (DO_BIC(BIC_CoreTmp)) outp += sprintf(outp, "%sCoreTmp", (printed++ ? delim : "")); + if (DO_BIC(BIC_CORE_THROT_CNT)) + outp += sprintf(outp, "%sCoreThr", (printed++ ? delim : "")); + if (do_rapl && !rapl_joules) { if (DO_BIC(BIC_CorWatt) && (do_rapl & RAPL_PER_CORE_ENERGY)) outp += sprintf(outp, "%sCorWatt", (printed++ ? delim : "")); @@ -724,6 +935,9 @@ void print_header(char *delim) if (DO_BIC(BIC_GFXMHz)) outp += sprintf(outp, "%sGFXMHz", (printed++ ? delim : "")); + if (DO_BIC(BIC_GFXACTMHz)) + outp += sprintf(outp, "%sGFXAMHz", (printed++ ? delim : "")); + if (DO_BIC(BIC_Totl_c0)) outp += sprintf(outp, "%sTotl%%C0", (printed++ ? delim : "")); if (DO_BIC(BIC_Any_c0)) @@ -779,6 +993,9 @@ void print_header(char *delim) if (DO_BIC(BIC_RAM__)) outp += sprintf(outp, "%sRAM_%%", (printed++ ? delim : "")); } + if (DO_BIC(BIC_UNCORE_MHZ)) + outp += sprintf(outp, "%sUncMHz", (printed++ ? delim : "")); + for (mp = sys.pp; mp; mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 64) @@ -796,8 +1013,7 @@ void print_header(char *delim) outp += sprintf(outp, "\n"); } -int dump_counters(struct thread_data *t, struct core_data *c, - struct pkg_data *p) +int dump_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { int i; struct msr_counter *mp; @@ -805,21 +1021,22 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "t %p, c %p, p %p\n", t, c, p); if (t) { - outp += sprintf(outp, "CPU: %d flags 0x%x\n", - t->cpu_id, t->flags); + outp += sprintf(outp, "CPU: %d flags 0x%x\n", t->cpu_id, t->flags); outp += sprintf(outp, "TSC: %016llX\n", t->tsc); outp += sprintf(outp, "aperf: %016llX\n", t->aperf); outp += sprintf(outp, "mperf: %016llX\n", t->mperf); outp += sprintf(outp, "c1: %016llX\n", t->c1); + if (DO_BIC(BIC_IPC)) + outp += sprintf(outp, "IPC: %lld\n", t->instr_count); + if (DO_BIC(BIC_IRQ)) outp += sprintf(outp, "IRQ: %lld\n", t->irq_count); if (DO_BIC(BIC_SMI)) outp += sprintf(outp, "SMI: %d\n", t->smi_count); for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { - outp += sprintf(outp, "tADDED [%d] msr0x%x: %08llX\n", - i, mp->msr_num, t->counter[i]); + outp += sprintf(outp, "tADDED [%d] msr0x%x: %08llX\n", i, mp->msr_num, t->counter[i]); } } @@ -829,11 +1046,11 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "c6: %016llX\n", c->c6); outp += sprintf(outp, "c7: %016llX\n", c->c7); outp += sprintf(outp, "DTS: %dC\n", c->core_temp_c); + outp += sprintf(outp, "cpu_throt_count: %016llX\n", c->core_throt_cnt); outp += sprintf(outp, "Joules: %0X\n", c->core_energy); for (i = 0, mp = sys.cp; mp; i++, mp = mp->next) { - outp += sprintf(outp, "cADDED [%d] msr0x%x: %08llX\n", - i, mp->msr_num, c->counter[i]); + outp += sprintf(outp, "cADDED [%d] msr0x%x: %08llX\n", i, mp->msr_num, c->counter[i]); } outp += sprintf(outp, "mc6_us: %016llX\n", c->mc6_us); } @@ -858,19 +1075,16 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "pc10: %016llX\n", p->pc10); outp += sprintf(outp, "cpu_lpi: %016llX\n", p->cpu_lpi); outp += sprintf(outp, "sys_lpi: %016llX\n", p->sys_lpi); - outp += sprintf(outp, "Joules PKG: %0X\n", p->energy_pkg); - outp += sprintf(outp, "Joules COR: %0X\n", p->energy_cores); - outp += sprintf(outp, "Joules GFX: %0X\n", p->energy_gfx); - outp += sprintf(outp, "Joules RAM: %0X\n", p->energy_dram); - outp += sprintf(outp, "Throttle PKG: %0X\n", - p->rapl_pkg_perf_status); - outp += sprintf(outp, "Throttle RAM: %0X\n", - p->rapl_dram_perf_status); + outp += sprintf(outp, "Joules PKG: %0llX\n", p->energy_pkg); + outp += sprintf(outp, "Joules COR: %0llX\n", p->energy_cores); + outp += sprintf(outp, "Joules GFX: %0llX\n", p->energy_gfx); + outp += sprintf(outp, "Joules RAM: %0llX\n", p->energy_dram); + outp += sprintf(outp, "Throttle PKG: %0llX\n", p->rapl_pkg_perf_status); + outp += sprintf(outp, "Throttle RAM: %0llX\n", p->rapl_dram_perf_status); outp += sprintf(outp, "PTM: %dC\n", p->pkg_temp_c); for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { - outp += sprintf(outp, "pADDED [%d] msr0x%x: %08llX\n", - i, mp->msr_num, p->counter[i]); + outp += sprintf(outp, "pADDED [%d] msr0x%x: %08llX\n", i, mp->msr_num, p->counter[i]); } } @@ -882,8 +1096,7 @@ int dump_counters(struct thread_data *t, struct core_data *c, /* * column formatting convention & formats */ -int format_counters(struct thread_data *t, struct core_data *c, - struct pkg_data *p) +int format_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { double interval_float, tsc; char *fmt8; @@ -892,17 +1105,16 @@ int format_counters(struct thread_data *t, struct core_data *c, char *delim = "\t"; int printed = 0; - /* if showing only 1st thread in core and this isn't one, bail out */ + /* if showing only 1st thread in core and this isn't one, bail out */ if (show_core_only && !(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) return 0; - /* if showing only 1st thread in pkg and this isn't one, bail out */ + /* if showing only 1st thread in pkg and this isn't one, bail out */ if (show_pkg_only && !(t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE)) return 0; /*if not summary line and --cpu is used */ - if ((t != &average.threads) && - (cpu_subset && !CPU_ISSET_S(t->cpu_id, cpu_subset_size, cpu_subset))) + if ((t != &average.threads) && (cpu_subset && !CPU_ISSET_S(t->cpu_id, cpu_subset_size, cpu_subset))) return 0; if (DO_BIC(BIC_USEC)) { @@ -917,7 +1129,7 @@ int format_counters(struct thread_data *t, struct core_data *c, if (DO_BIC(BIC_TOD)) outp += sprintf(outp, "%10ld.%06ld\t", t->tv_end.tv_sec, t->tv_end.tv_usec); - interval_float = t->tv_delta.tv_sec + t->tv_delta.tv_usec/1000000.0; + interval_float = t->tv_delta.tv_sec + t->tv_delta.tv_usec / 1000000.0; tsc = t->tsc * tsc_tweak; @@ -953,11 +1165,9 @@ int format_counters(struct thread_data *t, struct core_data *c, if (DO_BIC(BIC_Node)) { if (t) outp += sprintf(outp, "%s%d", - (printed++ ? delim : ""), - cpus[t->cpu_id].physical_node_id); + (printed++ ? delim : ""), cpus[t->cpu_id].physical_node_id); else - outp += sprintf(outp, "%s-", - (printed++ ? delim : "")); + outp += sprintf(outp, "%s-", (printed++ ? delim : "")); } if (DO_BIC(BIC_Core)) { if (c) @@ -974,22 +1184,25 @@ int format_counters(struct thread_data *t, struct core_data *c, } if (DO_BIC(BIC_Avg_MHz)) - outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), - 1.0 / units * t->aperf / interval_float); + outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), 1.0 / units * t->aperf / interval_float); if (DO_BIC(BIC_Busy)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->mperf/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->mperf / tsc); if (DO_BIC(BIC_Bzy_MHz)) { if (has_base_hz) - outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), base_hz / units * t->aperf / t->mperf); + outp += + sprintf(outp, "%s%.0f", (printed++ ? delim : ""), base_hz / units * t->aperf / t->mperf); else outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), - tsc / units * t->aperf / t->mperf / interval_float); + tsc / units * t->aperf / t->mperf / interval_float); } if (DO_BIC(BIC_TSC_MHz)) - outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), 1.0 * t->tsc/units/interval_float); + outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), 1.0 * t->tsc / units / interval_float); + + if (DO_BIC(BIC_IPC)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 1.0 * t->instr_count / t->aperf); /* IRQ */ if (DO_BIC(BIC_IRQ)) { @@ -1007,7 +1220,8 @@ int format_counters(struct thread_data *t, struct core_data *c, for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 32) - outp += sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int) t->counter[i]); + outp += + sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int)t->counter[i]); else outp += sprintf(outp, "%s0x%016llx", (printed++ ? delim : ""), t->counter[i]); } else if (mp->format == FORMAT_DELTA) { @@ -1017,27 +1231,28 @@ int format_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), t->counter[i]); } else if (mp->format == FORMAT_PERCENT) { if (mp->type == COUNTER_USEC) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), t->counter[i]/interval_float/10000); + outp += + sprintf(outp, "%s%.2f", (printed++ ? delim : ""), + t->counter[i] / interval_float / 10000); else - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->counter[i]/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->counter[i] / tsc); } } /* C1 */ if (DO_BIC(BIC_CPU_c1)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->c1/tsc); - + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->c1 / tsc); /* print per-core data only for 1st thread in core */ if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) goto done; if (DO_BIC(BIC_CPU_c3)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c3/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c3 / tsc); if (DO_BIC(BIC_CPU_c6)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c6/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c6 / tsc); if (DO_BIC(BIC_CPU_c7)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c7/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c7 / tsc); /* Mod%c6 */ if (DO_BIC(BIC_Mod_c6)) @@ -1046,10 +1261,15 @@ int format_counters(struct thread_data *t, struct core_data *c, if (DO_BIC(BIC_CoreTmp)) outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), c->core_temp_c); + /* Core throttle count */ + if (DO_BIC(BIC_CORE_THROT_CNT)) + outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), c->core_throt_cnt); + for (i = 0, mp = sys.cp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 32) - outp += sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int) c->counter[i]); + outp += + sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int)c->counter[i]); else outp += sprintf(outp, "%s0x%016llx", (printed++ ? delim : ""), c->counter[i]); } else if (mp->format == FORMAT_DELTA) { @@ -1058,21 +1278,15 @@ int format_counters(struct thread_data *t, struct core_data *c, else outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), c->counter[i]); } else if (mp->format == FORMAT_PERCENT) { - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->counter[i]/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->counter[i] / tsc); } } - /* - * If measurement interval exceeds minimum RAPL Joule Counter range, - * indicate that results are suspect by printing "**" in fraction place. - */ - if (interval_float < rapl_joule_counter_range) - fmt8 = "%s%.2f"; - else - fmt8 = "%6.0f**"; + fmt8 = "%s%.2f"; if (DO_BIC(BIC_CorWatt) && (do_rapl & RAPL_PER_CORE_ENERGY)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), c->core_energy * rapl_energy_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), c->core_energy * rapl_energy_units / interval_float); if (DO_BIC(BIC_Cor_J) && (do_rapl & RAPL_PER_CORE_ENERGY)) outp += sprintf(outp, fmt8, (printed++ ? delim : ""), c->core_energy * rapl_energy_units); @@ -1090,7 +1304,7 @@ int format_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "%s**.**", (printed++ ? delim : "")); } else { outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), - p->gfx_rc6_ms / 10.0 / interval_float); + p->gfx_rc6_ms / 10.0 / interval_float); } } @@ -1098,44 +1312,56 @@ int format_counters(struct thread_data *t, struct core_data *c, if (DO_BIC(BIC_GFXMHz)) outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), p->gfx_mhz); + /* GFXACTMHz */ + if (DO_BIC(BIC_GFXACTMHz)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), p->gfx_act_mhz); + /* Totl%C0, Any%C0 GFX%C0 CPUGFX% */ if (DO_BIC(BIC_Totl_c0)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_wtd_core_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_wtd_core_c0 / tsc); if (DO_BIC(BIC_Any_c0)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_any_core_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_any_core_c0 / tsc); if (DO_BIC(BIC_GFX_c0)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_any_gfxe_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_any_gfxe_c0 / tsc); if (DO_BIC(BIC_CPUGFX)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_both_core_gfxe_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_both_core_gfxe_c0 / tsc); if (DO_BIC(BIC_Pkgpc2)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc2/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc2 / tsc); if (DO_BIC(BIC_Pkgpc3)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc3/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc3 / tsc); if (DO_BIC(BIC_Pkgpc6)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc6/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc6 / tsc); if (DO_BIC(BIC_Pkgpc7)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc7/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc7 / tsc); if (DO_BIC(BIC_Pkgpc8)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc8/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc8 / tsc); if (DO_BIC(BIC_Pkgpc9)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc9/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc9 / tsc); if (DO_BIC(BIC_Pkgpc10)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc10/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc10 / tsc); if (DO_BIC(BIC_CPU_LPI)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->cpu_lpi / 1000000.0 / interval_float); + outp += + sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->cpu_lpi / 1000000.0 / interval_float); if (DO_BIC(BIC_SYS_LPI)) - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->sys_lpi / 1000000.0 / interval_float); + outp += + sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->sys_lpi / 1000000.0 / interval_float); if (DO_BIC(BIC_PkgWatt)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_pkg * rapl_energy_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_pkg * rapl_energy_units / interval_float); + if (DO_BIC(BIC_CorWatt) && !(do_rapl & RAPL_PER_CORE_ENERGY)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_cores * rapl_energy_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_cores * rapl_energy_units / interval_float); if (DO_BIC(BIC_GFXWatt)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_gfx * rapl_energy_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_gfx * rapl_energy_units / interval_float); if (DO_BIC(BIC_RAMWatt)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_dram * rapl_dram_energy_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), + p->energy_dram * rapl_dram_energy_units / interval_float); if (DO_BIC(BIC_Pkg_J)) outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_pkg * rapl_energy_units); if (DO_BIC(BIC_Cor_J) && !(do_rapl & RAPL_PER_CORE_ENERGY)) @@ -1145,14 +1371,22 @@ int format_counters(struct thread_data *t, struct core_data *c, if (DO_BIC(BIC_RAM_J)) outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_dram * rapl_dram_energy_units); if (DO_BIC(BIC_PKG__)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), 100.0 * p->rapl_pkg_perf_status * rapl_time_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), + 100.0 * p->rapl_pkg_perf_status * rapl_time_units / interval_float); if (DO_BIC(BIC_RAM__)) - outp += sprintf(outp, fmt8, (printed++ ? delim : ""), 100.0 * p->rapl_dram_perf_status * rapl_time_units / interval_float); + outp += + sprintf(outp, fmt8, (printed++ ? delim : ""), + 100.0 * p->rapl_dram_perf_status * rapl_time_units / interval_float); + /* UncMHz */ + if (DO_BIC(BIC_UNCORE_MHZ)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), p->uncore_mhz); for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 32) - outp += sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int) p->counter[i]); + outp += + sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int)p->counter[i]); else outp += sprintf(outp, "%s0x%016llx", (printed++ ? delim : ""), p->counter[i]); } else if (mp->format == FORMAT_DELTA) { @@ -1161,7 +1395,7 @@ int format_counters(struct thread_data *t, struct core_data *c, else outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), p->counter[i]); } else if (mp->format == FORMAT_PERCENT) { - outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->counter[i]/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->counter[i] / tsc); } } @@ -1186,22 +1420,24 @@ void flush_output_stdout(void) outp = output_buffer; } + void flush_output_stderr(void) { fputs(output_buffer, outf); fflush(outf); outp = output_buffer; } + void format_all_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { - static int printed; + static int count; - if (!printed || !summary_only) + if ((!count || (header_iterations && !(count % header_iterations))) || !summary_only) print_header("\t"); format_counters(&average.threads, &average.cores, &average.packages); - printed = 1; + count++; if (summary_only) return; @@ -1210,19 +1446,13 @@ void format_all_counters(struct thread_data *t, struct core_data *c, struct pkg_ } #define DELTA_WRAP32(new, old) \ - if (new > old) { \ - old = new - old; \ - } else { \ - old = 0x100000000 + new - old; \ - } + old = ((((unsigned long long)new << 32) - ((unsigned long long)old << 32)) >> 32); -int -delta_package(struct pkg_data *new, struct pkg_data *old) +int delta_package(struct pkg_data *new, struct pkg_data *old) { int i; struct msr_counter *mp; - if (DO_BIC(BIC_Totl_c0)) old->pkg_wtd_core_c0 = new->pkg_wtd_core_c0 - old->pkg_wtd_core_c0; if (DO_BIC(BIC_Any_c0)) @@ -1247,19 +1477,21 @@ delta_package(struct pkg_data *new, struct pkg_data *old) old->pkg_temp_c = new->pkg_temp_c; /* flag an error when rc6 counter resets/wraps */ - if (old->gfx_rc6_ms > new->gfx_rc6_ms) + if (old->gfx_rc6_ms > new->gfx_rc6_ms) old->gfx_rc6_ms = -1; else old->gfx_rc6_ms = new->gfx_rc6_ms - old->gfx_rc6_ms; + old->uncore_mhz = new->uncore_mhz; old->gfx_mhz = new->gfx_mhz; + old->gfx_act_mhz = new->gfx_act_mhz; - DELTA_WRAP32(new->energy_pkg, old->energy_pkg); - DELTA_WRAP32(new->energy_cores, old->energy_cores); - DELTA_WRAP32(new->energy_gfx, old->energy_gfx); - DELTA_WRAP32(new->energy_dram, old->energy_dram); - DELTA_WRAP32(new->rapl_pkg_perf_status, old->rapl_pkg_perf_status); - DELTA_WRAP32(new->rapl_dram_perf_status, old->rapl_dram_perf_status); + old->energy_pkg = new->energy_pkg - old->energy_pkg; + old->energy_cores = new->energy_cores - old->energy_cores; + old->energy_gfx = new->energy_gfx - old->energy_gfx; + old->energy_dram = new->energy_dram - old->energy_dram; + old->rapl_pkg_perf_status = new->rapl_pkg_perf_status - old->rapl_pkg_perf_status; + old->rapl_dram_perf_status = new->rapl_dram_perf_status - old->rapl_dram_perf_status; for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) @@ -1271,8 +1503,7 @@ delta_package(struct pkg_data *new, struct pkg_data *old) return 0; } -void -delta_core(struct core_data *new, struct core_data *old) +void delta_core(struct core_data *new, struct core_data *old) { int i; struct msr_counter *mp; @@ -1281,6 +1512,7 @@ delta_core(struct core_data *new, struct core_data *old) old->c6 = new->c6 - old->c6; old->c7 = new->c7 - old->c7; old->core_temp_c = new->core_temp_c; + old->core_throt_cnt = new->core_throt_cnt; old->mc6_us = new->mc6_us - old->mc6_us; DELTA_WRAP32(new->core_energy, old->core_energy); @@ -1304,9 +1536,7 @@ int soft_c1_residency_display(int bic) /* * old = new - old */ -int -delta_thread(struct thread_data *new, struct thread_data *old, - struct core_data *core_delta) +int delta_thread(struct thread_data *new, struct thread_data *old, struct core_data *core_delta) { int i; struct msr_counter *mp; @@ -1337,8 +1567,7 @@ delta_thread(struct thread_data *new, struct thread_data *old, old->c1 = new->c1 - old->c1; - if (DO_BIC(BIC_Avg_MHz) || DO_BIC(BIC_Busy) || DO_BIC(BIC_Bzy_MHz) || - soft_c1_residency_display(BIC_Avg_MHz)) { + if (DO_BIC(BIC_Avg_MHz) || DO_BIC(BIC_Busy) || DO_BIC(BIC_Bzy_MHz) || soft_c1_residency_display(BIC_Avg_MHz)) { if ((new->aperf > old->aperf) && (new->mperf > old->mperf)) { old->aperf = new->aperf - old->aperf; old->mperf = new->mperf - old->mperf; @@ -1347,7 +1576,6 @@ delta_thread(struct thread_data *new, struct thread_data *old, } } - if (use_c1_residency_msr) { /* * Some models have a dedicated C1 residency MSR, @@ -1364,7 +1592,7 @@ delta_thread(struct thread_data *new, struct thread_data *old, else { /* normal case, derive c1 */ old->c1 = (old->tsc * tsc_tweak) - old->mperf - core_delta->c3 - - core_delta->c6 - core_delta->c7; + - core_delta->c6 - core_delta->c7; } } @@ -1374,6 +1602,9 @@ delta_thread(struct thread_data *new, struct thread_data *old, old->mperf = 1; /* divide by 0 protection */ } + if (DO_BIC(BIC_IPC)) + old->instr_count = new->instr_count - old->instr_count; + if (DO_BIC(BIC_IRQ)) old->irq_count = new->irq_count - old->irq_count; @@ -1390,8 +1621,7 @@ delta_thread(struct thread_data *new, struct thread_data *old, } int delta_cpu(struct thread_data *t, struct core_data *c, - struct pkg_data *p, struct thread_data *t2, - struct core_data *c2, struct pkg_data *p2) + struct pkg_data *p, struct thread_data *t2, struct core_data *c2, struct pkg_data *p2) { int retval = 0; @@ -1414,7 +1644,7 @@ int delta_cpu(struct thread_data *t, struct core_data *c, void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { int i; - struct msr_counter *mp; + struct msr_counter *mp; t->tv_begin.tv_sec = 0; t->tv_begin.tv_usec = 0; @@ -1428,6 +1658,8 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data t->mperf = 0; t->c1 = 0; + t->instr_count = 0; + t->irq_count = 0; t->smi_count = 0; @@ -1440,6 +1672,7 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data c->mc6_us = 0; c->core_temp_c = 0; c->core_energy = 0; + c->core_throt_cnt = 0; p->pkg_wtd_core_c0 = 0; p->pkg_any_core_c0 = 0; @@ -1468,7 +1701,9 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data p->pkg_temp_c = 0; p->gfx_rc6_ms = 0; + p->uncore_mhz = 0; p->gfx_mhz = 0; + p->gfx_act_mhz = 0; for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) t->counter[i] = 0; @@ -1478,8 +1713,8 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) p->counter[i] = 0; } -int sum_counters(struct thread_data *t, struct core_data *c, - struct pkg_data *p) + +int sum_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { int i; struct msr_counter *mp; @@ -1502,6 +1737,8 @@ int sum_counters(struct thread_data *t, struct core_data *c, average.threads.mperf += t->mperf; average.threads.c1 += t->c1; + average.threads.instr_count += t->instr_count; + average.threads.irq_count += t->irq_count; average.threads.smi_count += t->smi_count; @@ -1521,6 +1758,7 @@ int sum_counters(struct thread_data *t, struct core_data *c, average.cores.mc6_us += c->mc6_us; average.cores.core_temp_c = MAX(average.cores.core_temp_c, c->core_temp_c); + average.cores.core_throt_cnt = MAX(average.cores.core_throt_cnt, c->core_throt_cnt); average.cores.core_energy += c->core_energy; @@ -1563,7 +1801,9 @@ int sum_counters(struct thread_data *t, struct core_data *c, average.packages.energy_gfx += p->energy_gfx; average.packages.gfx_rc6_ms = p->gfx_rc6_ms; + average.packages.uncore_mhz = p->uncore_mhz; average.packages.gfx_mhz = p->gfx_mhz; + average.packages.gfx_act_mhz = p->gfx_act_mhz; average.packages.pkg_temp_c = MAX(average.packages.pkg_temp_c, p->pkg_temp_c); @@ -1577,12 +1817,12 @@ int sum_counters(struct thread_data *t, struct core_data *c, } return 0; } + /* * sum the counters for all cpus in the system * compute the weighted average */ -void compute_average(struct thread_data *t, struct core_data *c, - struct pkg_data *p) +void compute_average(struct thread_data *t, struct core_data *c, struct pkg_data *p) { int i; struct msr_counter *mp; @@ -1597,6 +1837,7 @@ void compute_average(struct thread_data *t, struct core_data *c, average.threads.tsc /= topo.num_cpus; average.threads.aperf /= topo.num_cpus; average.threads.mperf /= topo.num_cpus; + average.threads.instr_count /= topo.num_cpus; average.threads.c1 /= topo.num_cpus; if (average.threads.irq_count > 9999999) @@ -1662,7 +1903,7 @@ static unsigned long long rdtsc(void) { unsigned int low, high; - asm volatile("rdtsc" : "=a" (low), "=d" (high)); + asm volatile ("rdtsc":"=a" (low), "=d"(high)); return low | ((unsigned long long)high) << 32; } @@ -1678,6 +1919,7 @@ FILE *fopen_or_die(const char *path, const char *mode) err(1, "%s: open failed", path); return filep; } + /* * snapshot_sysfs_counter() * @@ -1709,8 +1951,7 @@ int get_mp(int cpu, struct msr_counter *mp, unsigned long long *counterp) char path[128 + PATH_BYTES]; if (mp->flags & SYSFS_PERCPU) { - sprintf(path, "/sys/devices/system/cpu/cpu%d/%s", - cpu, mp->path); + sprintf(path, "/sys/devices/system/cpu/cpu%d/%s", cpu, mp->path); *counterp = snapshot_sysfs_counter(path); } else { @@ -1721,6 +1962,43 @@ int get_mp(int cpu, struct msr_counter *mp, unsigned long long *counterp) return 0; } +unsigned long long get_uncore_mhz(int package, int die) +{ + char path[128]; + + sprintf(path, "/sys/devices/system/cpu/intel_uncore_frequency/package_0%d_die_0%d/current_freq_khz", package, + die); + + return (snapshot_sysfs_counter(path) / 1000); +} + +int get_epb(int cpu) +{ + char path[128 + PATH_BYTES]; + unsigned long long msr; + int ret, epb = -1; + FILE *fp; + + sprintf(path, "/sys/devices/system/cpu/cpu%d/power/energy_perf_bias", cpu); + + fp = fopen(path, "r"); + if (!fp) + goto msr_fallback; + + ret = fscanf(fp, "%d", &epb); + if (ret != 1) + err(1, "%s(%s)", __func__, path); + + fclose(fp); + + return epb; + +msr_fallback: + get_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, &msr); + + return msr & 0xf; +} + void get_apic_id(struct thread_data *t) { unsigned int eax, ebx, ecx, edx; @@ -1743,7 +2021,7 @@ void get_apic_id(struct thread_data *t) eax = ebx = ecx = edx = 0; __cpuid(0x80000001, eax, ebx, ecx, edx); - topology_extensions = ecx & (1 << 22); + topology_extensions = ecx & (1 << 22); if (topology_extensions == 0) return; @@ -1766,8 +2044,27 @@ void get_apic_id(struct thread_data *t) t->x2apic_id = edx; if (debug && (t->apic_id != (t->x2apic_id & 0xff))) - fprintf(outf, "cpu%d: BIOS BUG: apic 0x%x x2apic 0x%x\n", - t->cpu_id, t->apic_id, t->x2apic_id); + fprintf(outf, "cpu%d: BIOS BUG: apic 0x%x x2apic 0x%x\n", t->cpu_id, t->apic_id, t->x2apic_id); +} + +int get_core_throt_cnt(int cpu, unsigned long long *cnt) +{ + char path[128 + PATH_BYTES]; + unsigned long long tmp; + FILE *fp; + int ret; + + sprintf(path, "/sys/devices/system/cpu/cpu%d/thermal_throttle/core_throttle_count", cpu); + fp = fopen(path, "r"); + if (!fp) + return -1; + ret = fscanf(fp, "%lld", &tmp); + fclose(fp); + if (ret != 1) + return -1; + *cnt = tmp; + + return 0; } /* @@ -1784,7 +2081,7 @@ int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) int i; if (cpu_migrate(cpu)) { - fprintf(outf, "Could not migrate to CPU %d\n", cpu); + fprintf(outf, "get_counters: Could not migrate to CPU %d\n", cpu); return -1; } @@ -1795,8 +2092,7 @@ int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) retry: t->tsc = rdtsc(); /* we are running on local CPU of interest */ - if (DO_BIC(BIC_Avg_MHz) || DO_BIC(BIC_Busy) || DO_BIC(BIC_Bzy_MHz) || - soft_c1_residency_display(BIC_Avg_MHz)) { + if (DO_BIC(BIC_Avg_MHz) || DO_BIC(BIC_Busy) || DO_BIC(BIC_Bzy_MHz) || soft_c1_residency_display(BIC_Avg_MHz)) { unsigned long long tsc_before, tsc_between, tsc_after, aperf_time, mperf_time; /* @@ -1843,8 +2139,7 @@ retry: if (aperf_mperf_retry_count < 5) goto retry; else - warnx("cpu%d jitter %lld %lld", - cpu, aperf_time, mperf_time); + warnx("cpu%d jitter %lld %lld", cpu, aperf_time, mperf_time); } aperf_mperf_retry_count = 0; @@ -1852,6 +2147,10 @@ retry: t->mperf = t->mperf * aperf_mperf_multiplier; } + if (DO_BIC(BIC_IPC)) + if (read(get_instr_count_fd(cpu), &t->instr_count, sizeof(long long)) != sizeof(long long)) + return -4; + if (DO_BIC(BIC_IRQ)) t->irq_count = irqs_per_cpu[cpu]; if (DO_BIC(BIC_SMI)) { @@ -1886,9 +2185,19 @@ retry: return -7; } - if (DO_BIC(BIC_CPU_c7) || soft_c1_residency_display(BIC_CPU_c7)) + if (DO_BIC(BIC_CPU_c7) || soft_c1_residency_display(BIC_CPU_c7)) { if (get_msr(cpu, MSR_CORE_C7_RESIDENCY, &c->c7)) return -8; + else if (t->is_atom) { + /* + * For Atom CPUs that has core cstate deeper than c6, + * MSR_CORE_C6_RESIDENCY returns residency of cc6 and deeper. + * Minus CC7 (and deeper cstates) residency to get + * accturate cc6 residency. + */ + c->c6 -= c->c7; + } + } if (DO_BIC(BIC_Mod_c6)) if (get_msr(cpu, MSR_MODULE_C6_RES_MS, &c->mc6_us)) @@ -1897,9 +2206,12 @@ retry: if (DO_BIC(BIC_CoreTmp)) { if (get_msr(cpu, MSR_IA32_THERM_STATUS, &msr)) return -9; - c->core_temp_c = tcc_activation_temp - ((msr >> 16) & 0x7F); + c->core_temp_c = tj_max - ((msr >> 16) & 0x7F); } + if (DO_BIC(BIC_CORE_THROT_CNT)) + get_core_throt_cnt(cpu, &c->core_throt_cnt); + if (do_rapl & RAPL_AMD_F17H) { if (get_msr(cpu, MSR_CORE_ENERGY_STAT, &msr)) return -14; @@ -1966,52 +2278,59 @@ retry: p->sys_lpi = cpuidle_cur_sys_lpi_us; if (do_rapl & RAPL_PKG) { - if (get_msr(cpu, MSR_PKG_ENERGY_STATUS, &msr)) + if (get_msr_sum(cpu, MSR_PKG_ENERGY_STATUS, &msr)) return -13; - p->energy_pkg = msr & 0xFFFFFFFF; + p->energy_pkg = msr; } if (do_rapl & RAPL_CORES_ENERGY_STATUS) { - if (get_msr(cpu, MSR_PP0_ENERGY_STATUS, &msr)) + if (get_msr_sum(cpu, MSR_PP0_ENERGY_STATUS, &msr)) return -14; - p->energy_cores = msr & 0xFFFFFFFF; + p->energy_cores = msr; } if (do_rapl & RAPL_DRAM) { - if (get_msr(cpu, MSR_DRAM_ENERGY_STATUS, &msr)) + if (get_msr_sum(cpu, MSR_DRAM_ENERGY_STATUS, &msr)) return -15; - p->energy_dram = msr & 0xFFFFFFFF; + p->energy_dram = msr; } if (do_rapl & RAPL_GFX) { - if (get_msr(cpu, MSR_PP1_ENERGY_STATUS, &msr)) + if (get_msr_sum(cpu, MSR_PP1_ENERGY_STATUS, &msr)) return -16; - p->energy_gfx = msr & 0xFFFFFFFF; + p->energy_gfx = msr; } if (do_rapl & RAPL_PKG_PERF_STATUS) { - if (get_msr(cpu, MSR_PKG_PERF_STATUS, &msr)) + if (get_msr_sum(cpu, MSR_PKG_PERF_STATUS, &msr)) return -16; - p->rapl_pkg_perf_status = msr & 0xFFFFFFFF; + p->rapl_pkg_perf_status = msr; } if (do_rapl & RAPL_DRAM_PERF_STATUS) { - if (get_msr(cpu, MSR_DRAM_PERF_STATUS, &msr)) + if (get_msr_sum(cpu, MSR_DRAM_PERF_STATUS, &msr)) return -16; - p->rapl_dram_perf_status = msr & 0xFFFFFFFF; + p->rapl_dram_perf_status = msr; } if (do_rapl & RAPL_AMD_F17H) { - if (get_msr(cpu, MSR_PKG_ENERGY_STAT, &msr)) + if (get_msr_sum(cpu, MSR_PKG_ENERGY_STAT, &msr)) return -13; - p->energy_pkg = msr & 0xFFFFFFFF; + p->energy_pkg = msr; } if (DO_BIC(BIC_PkgTmp)) { if (get_msr(cpu, MSR_IA32_PACKAGE_THERM_STATUS, &msr)) return -17; - p->pkg_temp_c = tcc_activation_temp - ((msr >> 16) & 0x7F); + p->pkg_temp_c = tj_max - ((msr >> 16) & 0x7F); } if (DO_BIC(BIC_GFX_rc6)) p->gfx_rc6_ms = gfx_cur_rc6_ms; + /* n.b. assume die0 uncore frequency applies to whole package */ + if (DO_BIC(BIC_UNCORE_MHZ)) + p->uncore_mhz = get_uncore_mhz(p->package_id, 0); + if (DO_BIC(BIC_GFXMHz)) p->gfx_mhz = gfx_cur_mhz; + if (DO_BIC(BIC_GFXACTMHz)) + p->gfx_act_mhz = gfx_act_mhz; + for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { if (get_mp(cpu, mp, &p->counter[i])) return -10; @@ -2028,45 +2347,81 @@ done: * (>= PCL__7) and to index pkg_cstate_limit_strings[]. */ -#define PCLUKN 0 /* Unknown */ -#define PCLRSV 1 /* Reserved */ -#define PCL__0 2 /* PC0 */ -#define PCL__1 3 /* PC1 */ -#define PCL__2 4 /* PC2 */ -#define PCL__3 5 /* PC3 */ -#define PCL__4 6 /* PC4 */ -#define PCL__6 7 /* PC6 */ -#define PCL_6N 8 /* PC6 No Retention */ -#define PCL_6R 9 /* PC6 Retention */ -#define PCL__7 10 /* PC7 */ -#define PCL_7S 11 /* PC7 Shrink */ -#define PCL__8 12 /* PC8 */ -#define PCL__9 13 /* PC9 */ -#define PCL_10 14 /* PC10 */ -#define PCLUNL 15 /* Unlimited */ +#define PCLUKN 0 /* Unknown */ +#define PCLRSV 1 /* Reserved */ +#define PCL__0 2 /* PC0 */ +#define PCL__1 3 /* PC1 */ +#define PCL__2 4 /* PC2 */ +#define PCL__3 5 /* PC3 */ +#define PCL__4 6 /* PC4 */ +#define PCL__6 7 /* PC6 */ +#define PCL_6N 8 /* PC6 No Retention */ +#define PCL_6R 9 /* PC6 Retention */ +#define PCL__7 10 /* PC7 */ +#define PCL_7S 11 /* PC7 Shrink */ +#define PCL__8 12 /* PC8 */ +#define PCL__9 13 /* PC9 */ +#define PCL_10 14 /* PC10 */ +#define PCLUNL 15 /* Unlimited */ int pkg_cstate_limit = PCLUKN; char *pkg_cstate_limit_strings[] = { "reserved", "unknown", "pc0", "pc1", "pc2", - "pc3", "pc4", "pc6", "pc6n", "pc6r", "pc7", "pc7s", "pc8", "pc9", "pc10", "unlimited"}; + "pc3", "pc4", "pc6", "pc6n", "pc6r", "pc7", "pc7s", "pc8", "pc9", "pc10", "unlimited" +}; -int nhm_pkg_cstate_limits[16] = {PCL__0, PCL__1, PCL__3, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int snb_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL_6N, PCL_6R, PCL__7, PCL_7S, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int hsw_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL__3, PCL__6, PCL__7, PCL_7S, PCL__8, PCL__9, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int slv_pkg_cstate_limits[16] = {PCL__0, PCL__1, PCLRSV, PCLRSV, PCL__4, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCL__6, PCL__7}; -int amt_pkg_cstate_limits[16] = {PCLUNL, PCL__1, PCL__2, PCLRSV, PCLRSV, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int phi_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL_6N, PCL_6R, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int glm_pkg_cstate_limits[16] = {PCLUNL, PCL__1, PCL__3, PCL__6, PCL__7, PCL_7S, PCL__8, PCL__9, PCL_10, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int skx_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL_6N, PCL_6R, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; +int nhm_pkg_cstate_limits[16] = + { PCL__0, PCL__1, PCL__3, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; + +int snb_pkg_cstate_limits[16] = + { PCL__0, PCL__2, PCL_6N, PCL_6R, PCL__7, PCL_7S, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; +int hsw_pkg_cstate_limits[16] = + { PCL__0, PCL__2, PCL__3, PCL__6, PCL__7, PCL_7S, PCL__8, PCL__9, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; -static void -calculate_tsc_tweak() +int slv_pkg_cstate_limits[16] = + { PCL__0, PCL__1, PCLRSV, PCLRSV, PCL__4, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCL__6, PCL__7 +}; + +int amt_pkg_cstate_limits[16] = + { PCLUNL, PCL__1, PCL__2, PCLRSV, PCLRSV, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; + +int phi_pkg_cstate_limits[16] = + { PCL__0, PCL__2, PCL_6N, PCL_6R, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; + +int glm_pkg_cstate_limits[16] = + { PCLUNL, PCL__1, PCL__3, PCL__6, PCL__7, PCL_7S, PCL__8, PCL__9, PCL_10, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; + +int skx_pkg_cstate_limits[16] = + { PCL__0, PCL__2, PCL_6N, PCL_6R, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; + +int icx_pkg_cstate_limits[16] = + { PCL__0, PCL__2, PCL__6, PCL__6, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, + PCLRSV, PCLRSV +}; + +static void calculate_tsc_tweak() { tsc_tweak = base_hz / tsc_hz; } -static void -dump_nhm_platform_info(void) +void prewake_cstate_probe(unsigned int family, unsigned int model); + +static void dump_nhm_platform_info(void) { unsigned long long msr; unsigned int ratio; @@ -2076,22 +2431,23 @@ dump_nhm_platform_info(void) fprintf(outf, "cpu%d: MSR_PLATFORM_INFO: 0x%08llx\n", base_cpu, msr); ratio = (msr >> 40) & 0xFF; - fprintf(outf, "%d * %.1f = %.1f MHz max efficiency frequency\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max efficiency frequency\n", ratio, bclk, ratio * bclk); ratio = (msr >> 8) & 0xFF; - fprintf(outf, "%d * %.1f = %.1f MHz base frequency\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz base frequency\n", ratio, bclk, ratio * bclk); get_msr(base_cpu, MSR_IA32_POWER_CTL, &msr); fprintf(outf, "cpu%d: MSR_IA32_POWER_CTL: 0x%08llx (C1E auto-promotion: %sabled)\n", base_cpu, msr, msr & 0x2 ? "EN" : "DIS"); + /* C-state Pre-wake Disable (CSTATE_PREWAKE_DISABLE) */ + if (dis_cstate_prewake) + fprintf(outf, "C-state Pre-wake: %sabled\n", msr & 0x40000000 ? "DIS" : "EN"); + return; } -static void -dump_hsw_turbo_ratio_limits(void) +static void dump_hsw_turbo_ratio_limits(void) { unsigned long long msr; unsigned int ratio; @@ -2102,18 +2458,15 @@ dump_hsw_turbo_ratio_limits(void) ratio = (msr >> 8) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 18 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 18 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 0) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 17 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 17 active cores\n", ratio, bclk, ratio * bclk); return; } -static void -dump_ivt_turbo_ratio_limits(void) +static void dump_ivt_turbo_ratio_limits(void) { unsigned long long msr; unsigned int ratio; @@ -2124,68 +2477,68 @@ dump_ivt_turbo_ratio_limits(void) ratio = (msr >> 56) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 16 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 16 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 48) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 15 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 15 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 40) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 14 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 14 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 32) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 13 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 13 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 24) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 12 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 12 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 16) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 11 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 11 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 8) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 10 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 10 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 0) & 0xFF; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 9 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 9 active cores\n", ratio, bclk, ratio * bclk); return; } + int has_turbo_ratio_group_limits(int family, int model) { if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_ATOM_GOLDMONT: case INTEL_FAM6_SKYLAKE_X: + case INTEL_FAM6_ICELAKE_X: + case INTEL_FAM6_SAPPHIRERAPIDS_X: case INTEL_FAM6_ATOM_GOLDMONT_D: + case INTEL_FAM6_ATOM_TREMONT_D: return 1; + default: + return 0; } - return 0; } -static void -dump_turbo_ratio_limits(int family, int model) +static void dump_turbo_ratio_limits(int trl_msr_offset, int family, int model) { unsigned long long msr, core_counts; - unsigned int ratio, group_size; + int shift; - get_msr(base_cpu, MSR_TURBO_RATIO_LIMIT, &msr); - fprintf(outf, "cpu%d: MSR_TURBO_RATIO_LIMIT: 0x%08llx\n", base_cpu, msr); + get_msr(base_cpu, trl_msr_offset, &msr); + fprintf(outf, "cpu%d: MSR_%sTURBO_RATIO_LIMIT: 0x%08llx\n", + base_cpu, trl_msr_offset == MSR_SECONDARY_TURBO_RATIO_LIMIT ? "SECONDARY" : "", msr); if (has_turbo_ratio_group_limits(family, model)) { get_msr(base_cpu, MSR_TURBO_RATIO_LIMIT1, &core_counts); @@ -2194,58 +2547,20 @@ dump_turbo_ratio_limits(int family, int model) core_counts = 0x0807060504030201; } - ratio = (msr >> 56) & 0xFF; - group_size = (core_counts >> 56) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); - - ratio = (msr >> 48) & 0xFF; - group_size = (core_counts >> 48) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); - - ratio = (msr >> 40) & 0xFF; - group_size = (core_counts >> 40) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); - - ratio = (msr >> 32) & 0xFF; - group_size = (core_counts >> 32) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); - - ratio = (msr >> 24) & 0xFF; - group_size = (core_counts >> 24) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); - - ratio = (msr >> 16) & 0xFF; - group_size = (core_counts >> 16) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); + for (shift = 56; shift >= 0; shift -= 8) { + unsigned int ratio, group_size; - ratio = (msr >> 8) & 0xFF; - group_size = (core_counts >> 8) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); + ratio = (msr >> shift) & 0xFF; + group_size = (core_counts >> shift) & 0xFF; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); + } - ratio = (msr >> 0) & 0xFF; - group_size = (core_counts >> 0) & 0xFF; - if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", - ratio, bclk, ratio * bclk, group_size); return; } -static void -dump_atom_turbo_ratio_limits(void) +static void dump_atom_turbo_ratio_limits(void) { unsigned long long msr; unsigned int ratio; @@ -2255,45 +2570,37 @@ dump_atom_turbo_ratio_limits(void) ratio = (msr >> 0) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz minimum operating frequency\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz minimum operating frequency\n", ratio, bclk, ratio * bclk); ratio = (msr >> 8) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz low frequency mode (LFM)\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz low frequency mode (LFM)\n", ratio, bclk, ratio * bclk); ratio = (msr >> 16) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz base frequency\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz base frequency\n", ratio, bclk, ratio * bclk); get_msr(base_cpu, MSR_ATOM_CORE_TURBO_RATIOS, &msr); fprintf(outf, "cpu%d: MSR_ATOM_CORE_TURBO_RATIOS: 0x%08llx\n", base_cpu, msr & 0xFFFFFFFF); ratio = (msr >> 24) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 4 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 4 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 16) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 3 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 3 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 8) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 2 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 2 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 0) & 0x3F; if (ratio) - fprintf(outf, "%d * %.1f = %.1f MHz max turbo 1 active core\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 1 active core\n", ratio, bclk, ratio * bclk); } -static void -dump_knl_turbo_ratio_limits(void) +static void dump_knl_turbo_ratio_limits(void) { const unsigned int buckets_no = 7; @@ -2305,10 +2612,9 @@ dump_knl_turbo_ratio_limits(void) get_msr(base_cpu, MSR_TURBO_RATIO_LIMIT, &msr); - fprintf(outf, "cpu%d: MSR_TURBO_RATIO_LIMIT: 0x%08llx\n", - base_cpu, msr); + fprintf(outf, "cpu%d: MSR_TURBO_RATIO_LIMIT: 0x%08llx\n", base_cpu, msr); - /** + /* * Turbo encoding in KNL is as follows: * [0] -- Reserved * [7:1] -- Base value of number of active cores of bucket 1. @@ -2351,8 +2657,7 @@ dump_knl_turbo_ratio_limits(void) ratio[i], bclk, ratio[i] * bclk, cores[i]); } -static void -dump_nhm_cst_cfg(void) +static void dump_nhm_cst_cfg(void) { unsigned long long msr; @@ -2365,14 +2670,11 @@ dump_nhm_cst_cfg(void) (msr & SNB_C1_AUTO_UNDEMOTE) ? "UNdemote-C1, " : "", (msr & NHM_C3_AUTO_DEMOTE) ? "demote-C3, " : "", (msr & NHM_C1_AUTO_DEMOTE) ? "demote-C1, " : "", - (msr & (1 << 15)) ? "" : "UN", - (unsigned int)msr & 0xF, - pkg_cstate_limit_strings[pkg_cstate_limit]); + (msr & (1 << 15)) ? "" : "UN", (unsigned int)msr & 0xF, pkg_cstate_limit_strings[pkg_cstate_limit]); #define AUTOMATIC_CSTATE_CONVERSION (1UL << 16) if (has_automatic_cstate_conversion) { - fprintf(outf, ", automatic c-state conversion=%s", - (msr & AUTOMATIC_CSTATE_CONVERSION) ? "on" : "off"); + fprintf(outf, ", automatic c-state conversion=%s", (msr & AUTOMATIC_CSTATE_CONVERSION) ? "on" : "off"); } fprintf(outf, ")\n"); @@ -2380,8 +2682,7 @@ dump_nhm_cst_cfg(void) return; } -static void -dump_config_tdp(void) +static void dump_config_tdp(void) { unsigned long long msr; @@ -2423,7 +2724,7 @@ dump_config_tdp(void) fprintf(outf, ")\n"); } -unsigned int irtl_time_units[] = {1, 32, 1024, 32768, 1048576, 33554432, 0, 0 }; +unsigned int irtl_time_units[] = { 1, 32, 1024, 32768, 1048576, 33554432, 0, 0 }; void print_irtl(void) { @@ -2463,6 +2764,7 @@ void print_irtl(void) (msr & 0x3FF) * irtl_time_units[(msr >> 10) & 0x3]); } + void free_fd_percpu(void) { int i; @@ -2519,7 +2821,6 @@ void free_all_buffers(void) free(cpus); } - /* * Parse a file containing a single int. * Return 0 if file can not be opened @@ -2594,8 +2895,7 @@ void set_node_data(void) * the logical_node_id */ for (cpux = cpu; cpux <= topo.max_cpu_num; cpux++) { - if ((cpus[cpux].physical_package_id == pkg) && - (cpus[cpux].physical_node_id == node)) { + if ((cpus[cpux].physical_package_id == pkg) && (cpus[cpux].physical_node_id == node)) { cpus[cpux].logical_node_id = lnode; cpu_count++; } @@ -2617,8 +2917,7 @@ int get_physical_node_id(struct cpu_topology *thiscpu) int cpu = thiscpu->logical_cpu_id; for (i = 0; i <= topo.max_cpu_num; i++) { - sprintf(path, "/sys/devices/system/cpu/cpu%d/node%i/cpulist", - cpu, i); + sprintf(path, "/sys/devices/system/cpu/cpu%d/node%i/cpulist", cpu, i); filep = fopen(path, "r"); if (!filep) continue; @@ -2648,9 +2947,13 @@ int get_thread_siblings(struct cpu_topology *thiscpu) size = CPU_ALLOC_SIZE((topo.max_cpu_num + 1)); CPU_ZERO_S(size, thiscpu->put_ids); - sprintf(path, - "/sys/devices/system/cpu/cpu%d/topology/thread_siblings", cpu); - filep = fopen_or_die(path, "r"); + sprintf(path, "/sys/devices/system/cpu/cpu%d/topology/thread_siblings", cpu); + filep = fopen(path, "r"); + + if (!filep) { + warnx("%s: open failed", path); + return -1; + } do { offset -= BITMASK_SIZE; if (fscanf(filep, "%lx%c", &map, &character) != 2) @@ -2661,14 +2964,12 @@ int get_thread_siblings(struct cpu_topology *thiscpu) sib_core = get_core_id(so); if (sib_core == thiscpu->physical_core_id) { CPU_SET_S(so, size, thiscpu->put_ids); - if ((so != cpu) && - (cpus[so].thread_id < 0)) - cpus[so].thread_id = - thread_id++; + if ((so != cpu) && (cpus[so].thread_id < 0)) + cpus[so].thread_id = thread_id++; } } } - } while (!strncmp(&character, ",", 1)); + } while (character == ','); fclose(filep); return CPU_COUNT_S(size, thiscpu->put_ids); @@ -2679,41 +2980,31 @@ int get_thread_siblings(struct cpu_topology *thiscpu) * skip non-present cpus */ -int for_all_cpus_2(int (func)(struct thread_data *, struct core_data *, - struct pkg_data *, struct thread_data *, struct core_data *, - struct pkg_data *), struct thread_data *thread_base, - struct core_data *core_base, struct pkg_data *pkg_base, - struct thread_data *thread_base2, struct core_data *core_base2, - struct pkg_data *pkg_base2) +int for_all_cpus_2(int (func) (struct thread_data *, struct core_data *, + struct pkg_data *, struct thread_data *, struct core_data *, + struct pkg_data *), struct thread_data *thread_base, + struct core_data *core_base, struct pkg_data *pkg_base, + struct thread_data *thread_base2, struct core_data *core_base2, struct pkg_data *pkg_base2) { int retval, pkg_no, node_no, core_no, thread_no; for (pkg_no = 0; pkg_no < topo.num_packages; ++pkg_no) { for (node_no = 0; node_no < topo.nodes_per_pkg; ++node_no) { - for (core_no = 0; core_no < topo.cores_per_node; - ++core_no) { - for (thread_no = 0; thread_no < - topo.threads_per_core; ++thread_no) { + for (core_no = 0; core_no < topo.cores_per_node; ++core_no) { + for (thread_no = 0; thread_no < topo.threads_per_core; ++thread_no) { struct thread_data *t, *t2; struct core_data *c, *c2; struct pkg_data *p, *p2; - t = GET_THREAD(thread_base, thread_no, - core_no, node_no, - pkg_no); + t = GET_THREAD(thread_base, thread_no, core_no, node_no, pkg_no); if (cpu_is_not_present(t->cpu_id)) continue; - t2 = GET_THREAD(thread_base2, thread_no, - core_no, node_no, - pkg_no); + t2 = GET_THREAD(thread_base2, thread_no, core_no, node_no, pkg_no); - c = GET_CORE(core_base, core_no, - node_no, pkg_no); - c2 = GET_CORE(core_base2, core_no, - node_no, - pkg_no); + c = GET_CORE(core_base, core_no, node_no, pkg_no); + c2 = GET_CORE(core_base2, core_no, node_no, pkg_no); p = GET_PKG(pkg_base, pkg_no); p2 = GET_PKG(pkg_base2, pkg_no); @@ -2732,7 +3023,7 @@ int for_all_cpus_2(int (func)(struct thread_data *, struct core_data *, * run func(cpu) on every cpu in /proc/stat * return max_cpu number */ -int for_all_proc_cpus(int (func)(int)) +int for_all_proc_cpus(int (func) (int)) { FILE *fp; int cpu_num; @@ -2752,7 +3043,7 @@ int for_all_proc_cpus(int (func)(int)) retval = func(cpu_num); if (retval) { fclose(fp); - return(retval); + return (retval); } } fclose(fp); @@ -2763,22 +3054,27 @@ void re_initialize(void) { free_all_buffers(); setup_all_buffers(); - printf("turbostat: re-initialized with num_cpus %d\n", topo.num_cpus); + fprintf(outf, "turbostat: re-initialized with num_cpus %d\n", topo.num_cpus); } void set_max_cpu_num(void) { FILE *filep; + int base_cpu; unsigned long dummy; + char pathname[64]; + base_cpu = sched_getcpu(); + if (base_cpu < 0) + err(1, "cannot find calling cpu ID"); + sprintf(pathname, "/sys/devices/system/cpu/cpu%d/topology/thread_siblings", base_cpu); + + filep = fopen_or_die(pathname, "r"); topo.max_cpu_num = 0; - filep = fopen_or_die( - "/sys/devices/system/cpu/cpu0/topology/thread_siblings", - "r"); while (fscanf(filep, "%lx,", &dummy) == 1) topo.max_cpu_num += BITMASK_SIZE; fclose(filep); - topo.max_cpu_num--; /* 0 based */ + topo.max_cpu_num--; /* 0 based */ } /* @@ -2787,9 +3083,12 @@ void set_max_cpu_num(void) */ int count_cpus(int cpu) { + UNUSED(cpu); + topo.num_cpus++; return 0; } + int mark_cpu_present(int cpu) { CPU_SET_S(cpu, cpu_present_setsize, cpu_present_set); @@ -2859,12 +3158,12 @@ int snapshot_proc_interrupts(void) } - while (getc(fp) != '\n') - ; /* flush interrupt description */ + while (getc(fp) != '\n') ; /* flush interrupt description */ } return 0; } + /* * snapshot_gfx_rc6_ms() * @@ -2888,6 +3187,7 @@ int snapshot_gfx_rc6_ms(void) return 0; } + /* * snapshot_gfx_mhz() * @@ -2916,6 +3216,33 @@ int snapshot_gfx_mhz(void) } /* + * snapshot_gfx_cur_mhz() + * + * record snapshot of + * /sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz + * + * return 1 if config change requires a restart, else return 0 + */ +int snapshot_gfx_act_mhz(void) +{ + static FILE *fp; + int retval; + + if (fp == NULL) + fp = fopen_or_die("/sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz", "r"); + else { + rewind(fp); + fflush(fp); + } + + retval = fscanf(fp, "%d", &gfx_act_mhz); + if (retval != 1) + err(1, "GFX ACT MHz"); + + return 0; +} + +/* * snapshot_cpu_lpi() * * record snapshot of @@ -2940,6 +3267,7 @@ int snapshot_cpu_lpi_us(void) return 0; } + /* * snapshot_sys_lpi() * @@ -2963,6 +3291,7 @@ int snapshot_sys_lpi_us(void) return 0; } + /* * snapshot /proc and /sys files * @@ -2980,6 +3309,9 @@ int snapshot_proc_sysfs_files(void) if (DO_BIC(BIC_GFXMHz)) snapshot_gfx_mhz(); + if (DO_BIC(BIC_GFXACTMHz)) + snapshot_gfx_act_mhz(); + if (DO_BIC(BIC_CPU_LPI)) snapshot_cpu_lpi_us(); @@ -2991,7 +3323,7 @@ int snapshot_proc_sysfs_files(void) int exit_requested; -static void signal_handler (int signal) +static void signal_handler(int signal) { switch (signal) { case SIGINT: @@ -3057,15 +3389,158 @@ void do_sleep(void) } } +int get_msr_sum(int cpu, off_t offset, unsigned long long *msr) +{ + int ret, idx; + unsigned long long msr_cur, msr_last; + + if (!per_cpu_msr_sum) + return 1; + + idx = offset_to_idx(offset); + if (idx < 0) + return idx; + /* get_msr_sum() = sum + (get_msr() - last) */ + ret = get_msr(cpu, offset, &msr_cur); + if (ret) + return ret; + msr_last = per_cpu_msr_sum[cpu].entries[idx].last; + DELTA_WRAP32(msr_cur, msr_last); + *msr = msr_last + per_cpu_msr_sum[cpu].entries[idx].sum; + + return 0; +} + +timer_t timerid; + +/* Timer callback, update the sum of MSRs periodically. */ +static int update_msr_sum(struct thread_data *t, struct core_data *c, struct pkg_data *p) +{ + int i, ret; + int cpu = t->cpu_id; + + UNUSED(c); + UNUSED(p); + + for (i = IDX_PKG_ENERGY; i < IDX_COUNT; i++) { + unsigned long long msr_cur, msr_last; + off_t offset; + + if (!idx_valid(i)) + continue; + offset = idx_to_offset(i); + if (offset < 0) + continue; + ret = get_msr(cpu, offset, &msr_cur); + if (ret) { + fprintf(outf, "Can not update msr(0x%llx)\n", (unsigned long long)offset); + continue; + } + + msr_last = per_cpu_msr_sum[cpu].entries[i].last; + per_cpu_msr_sum[cpu].entries[i].last = msr_cur & 0xffffffff; + + DELTA_WRAP32(msr_cur, msr_last); + per_cpu_msr_sum[cpu].entries[i].sum += msr_last; + } + return 0; +} + +static void msr_record_handler(union sigval v) +{ + UNUSED(v); + + for_all_cpus(update_msr_sum, EVEN_COUNTERS); +} + +void msr_sum_record(void) +{ + struct itimerspec its; + struct sigevent sev; + + per_cpu_msr_sum = calloc(topo.max_cpu_num + 1, sizeof(struct msr_sum_array)); + if (!per_cpu_msr_sum) { + fprintf(outf, "Can not allocate memory for long time MSR.\n"); + return; + } + /* + * Signal handler might be restricted, so use thread notifier instead. + */ + memset(&sev, 0, sizeof(struct sigevent)); + sev.sigev_notify = SIGEV_THREAD; + sev.sigev_notify_function = msr_record_handler; + + sev.sigev_value.sival_ptr = &timerid; + if (timer_create(CLOCK_REALTIME, &sev, &timerid) == -1) { + fprintf(outf, "Can not create timer.\n"); + goto release_msr; + } + + its.it_value.tv_sec = 0; + its.it_value.tv_nsec = 1; + /* + * A wraparound time has been calculated early. + * Some sources state that the peak power for a + * microprocessor is usually 1.5 times the TDP rating, + * use 2 * TDP for safety. + */ + its.it_interval.tv_sec = rapl_joule_counter_range / 2; + its.it_interval.tv_nsec = 0; + + if (timer_settime(timerid, 0, &its, NULL) == -1) { + fprintf(outf, "Can not set timer.\n"); + goto release_timer; + } + return; + +release_timer: + timer_delete(timerid); +release_msr: + free(per_cpu_msr_sum); +} + +/* + * set_my_sched_priority(pri) + * return previous + * + * if non-root, do this: + * # /sbin/setcap cap_sys_rawio,cap_sys_nice=+ep /usr/bin/turbostat + */ +int set_my_sched_priority(int priority) +{ + int retval; + int original_priority; + + errno = 0; + original_priority = getpriority(PRIO_PROCESS, 0); + if (errno && (original_priority == -1)) + err(errno, "getpriority"); + + retval = setpriority(PRIO_PROCESS, 0, priority); + if (retval) + err(retval, "setpriority(%d)", priority); + + errno = 0; + retval = getpriority(PRIO_PROCESS, 0); + if (retval != priority) + err(retval, "getpriority(%d) != setpriority(%d)", retval, priority); + + return original_priority; +} void turbostat_loop() { int retval; int restarted = 0; - int done_iters = 0; + unsigned int done_iters = 0; setup_signal_handler(); + /* + * elevate own priority for interval mode + */ + set_my_sched_priority(-20); + restart: restarted++; @@ -3075,7 +3550,7 @@ restart: if (retval < -1) { exit(retval); } else if (retval == -1) { - if (restarted > 1) { + if (restarted > 10) { exit(retval); } re_initialize(); @@ -3146,7 +3621,7 @@ void check_dev_msr() sprintf(pathname, "/dev/cpu/%d/msr", base_cpu); if (stat(pathname, &sb)) - if (system("/sbin/modprobe msr > /dev/null 2>&1")) + if (system("/sbin/modprobe msr > /dev/null 2>&1")) err(-5, "no /dev/cpu/0/msr, Try \"# modprobe msr\" "); } @@ -3168,8 +3643,7 @@ int check_for_cap_sys_rawio(void) err(-6, "cap_get\n"); if (cap_flag_value != CAP_SET) { - warnx("capget(CAP_SYS_RAWIO) failed," - " try \"# setcap cap_sys_rawio=ep %s\"", progname); + warnx("capget(CAP_SYS_RAWIO) failed," " try \"# setcap cap_sys_rawio=ep %s\"", progname); return 1; } @@ -3178,6 +3652,7 @@ int check_for_cap_sys_rawio(void) return 0; } + void check_permissions(void) { int do_exit = 0; @@ -3260,11 +3735,17 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) has_misc_feature_control = 1; break; case INTEL_FAM6_SKYLAKE_X: /* SKX */ + case INTEL_FAM6_SAPPHIRERAPIDS_X: /* SPR */ pkg_cstate_limits = skx_pkg_cstate_limits; has_misc_feature_control = 1; break; + case INTEL_FAM6_ICELAKE_X: /* ICX */ + pkg_cstate_limits = icx_pkg_cstate_limits; + has_misc_feature_control = 1; + break; case INTEL_FAM6_ATOM_SILVERMONT: /* BYT */ no_MSR_MISC_PWR_MGMT = 1; + /* FALLTHRU */ case INTEL_FAM6_ATOM_SILVERMONT_D: /* AVN */ pkg_cstate_limits = slv_pkg_cstate_limits; break; @@ -3279,6 +3760,7 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) case INTEL_FAM6_ATOM_GOLDMONT_PLUS: case INTEL_FAM6_ATOM_GOLDMONT_D: /* DNV */ case INTEL_FAM6_ATOM_TREMONT: /* EHL */ + case INTEL_FAM6_ATOM_TREMONT_D: /* JVL */ pkg_cstate_limits = glm_pkg_cstate_limits; break; default: @@ -3294,6 +3776,7 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) has_base_hz = 1; return 1; } + /* * SLV client has support for unique MSRs: * @@ -3306,6 +3789,9 @@ int has_slv_msrs(unsigned int family, unsigned int model) if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_ATOM_SILVERMONT: case INTEL_FAM6_ATOM_SILVERMONT_MID: @@ -3314,47 +3800,95 @@ int has_slv_msrs(unsigned int family, unsigned int model) } return 0; } + int is_dnv(unsigned int family, unsigned int model) { if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_ATOM_GOLDMONT_D: return 1; } return 0; } + int is_bdx(unsigned int family, unsigned int model) { if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_BROADWELL_X: return 1; } return 0; } + int is_skx(unsigned int family, unsigned int model) { if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_SKYLAKE_X: return 1; } return 0; } + +int is_icx(unsigned int family, unsigned int model) +{ + + if (!genuine_intel) + return 0; + + if (family != 6) + return 0; + + switch (model) { + case INTEL_FAM6_ICELAKE_X: + return 1; + } + return 0; +} + +int is_spr(unsigned int family, unsigned int model) +{ + + if (!genuine_intel) + return 0; + + if (family != 6) + return 0; + + switch (model) { + case INTEL_FAM6_SAPPHIRERAPIDS_X: + return 1; + } + return 0; +} + int is_ehl(unsigned int family, unsigned int model) { if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_ATOM_TREMONT: return 1; @@ -3362,13 +3896,31 @@ int is_ehl(unsigned int family, unsigned int model) return 0; } +int is_jvl(unsigned int family, unsigned int model) +{ + if (!genuine_intel) + return 0; + + if (family != 6) + return 0; + + switch (model) { + case INTEL_FAM6_ATOM_TREMONT_D: + return 1; + } + return 0; +} + int has_turbo_ratio_limit(unsigned int family, unsigned int model) { if (has_slv_msrs(family, model)) return 0; + if (family != 6) + return 0; + switch (model) { - /* Nehalem compatible, but do not include turbo-ratio limit support */ + /* Nehalem compatible, but do not include turbo-ratio limit support */ case INTEL_FAM6_NEHALEM_EX: /* Nehalem-EX Xeon - Beckton */ case INTEL_FAM6_XEON_PHI_KNL: /* PHI - Knights Landing (different MSR definition) */ return 0; @@ -3376,6 +3928,7 @@ int has_turbo_ratio_limit(unsigned int family, unsigned int model) return 1; } } + int has_atom_turbo_ratio_limit(unsigned int family, unsigned int model) { if (has_slv_msrs(family, model)) @@ -3383,6 +3936,7 @@ int has_atom_turbo_ratio_limit(unsigned int family, unsigned int model) return 0; } + int has_ivt_turbo_ratio_limit(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -3399,6 +3953,7 @@ int has_ivt_turbo_ratio_limit(unsigned int family, unsigned int model) return 0; } } + int has_hsw_turbo_ratio_limit(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -3430,6 +3985,7 @@ int has_knl_turbo_ratio_limit(unsigned int family, unsigned int model) return 0; } } + int has_glm_turbo_ratio_limit(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -3441,11 +3997,14 @@ int has_glm_turbo_ratio_limit(unsigned int family, unsigned int model) switch (model) { case INTEL_FAM6_ATOM_GOLDMONT: case INTEL_FAM6_SKYLAKE_X: + case INTEL_FAM6_ICELAKE_X: + case INTEL_FAM6_SAPPHIRERAPIDS_X: return 1; default: return 0; } } + int has_config_tdp(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -3466,7 +4025,8 @@ int has_config_tdp(unsigned int family, unsigned int model) case INTEL_FAM6_SKYLAKE_L: /* SKL */ case INTEL_FAM6_CANNONLAKE_L: /* CNL */ case INTEL_FAM6_SKYLAKE_X: /* SKX */ - + case INTEL_FAM6_ICELAKE_X: /* ICX */ + case INTEL_FAM6_SAPPHIRERAPIDS_X: /* SPR */ case INTEL_FAM6_XEON_PHI_KNL: /* Knights Landing */ return 1; default: @@ -3474,13 +4034,57 @@ int has_config_tdp(unsigned int family, unsigned int model) } } -static void -dump_cstate_pstate_config_info(unsigned int family, unsigned int model) +/* + * tcc_offset_bits: + * 0: Tcc Offset not supported (Default) + * 6: Bit 29:24 of MSR_PLATFORM_INFO + * 4: Bit 27:24 of MSR_PLATFORM_INFO + */ +void check_tcc_offset(int model) { - if (!do_nhm_platform_info) + unsigned long long msr; + + if (!genuine_intel) return; - dump_nhm_platform_info(); + switch (model) { + case INTEL_FAM6_SKYLAKE_L: + case INTEL_FAM6_SKYLAKE: + case INTEL_FAM6_KABYLAKE_L: + case INTEL_FAM6_KABYLAKE: + case INTEL_FAM6_ICELAKE_L: + case INTEL_FAM6_ICELAKE: + case INTEL_FAM6_TIGERLAKE_L: + case INTEL_FAM6_TIGERLAKE: + case INTEL_FAM6_COMETLAKE: + if (!get_msr(base_cpu, MSR_PLATFORM_INFO, &msr)) { + msr = (msr >> 30) & 1; + if (msr) + tcc_offset_bits = 6; + } + return; + default: + return; + } +} + +static void remove_underbar(char *s) +{ + char *to = s; + + while (*s) { + if (*s != '_') + *to++ = *s; + s++; + } + + *to = 0; +} + +static void dump_turbo_ratio_info(unsigned int family, unsigned int model) +{ + if (!has_turbo) + return; if (has_hsw_turbo_ratio_limit(family, model)) dump_hsw_turbo_ratio_limits(); @@ -3488,8 +4092,12 @@ dump_cstate_pstate_config_info(unsigned int family, unsigned int model) if (has_ivt_turbo_ratio_limit(family, model)) dump_ivt_turbo_ratio_limits(); - if (has_turbo_ratio_limit(family, model)) - dump_turbo_ratio_limits(family, model); + if (has_turbo_ratio_limit(family, model)) { + dump_turbo_ratio_limits(MSR_TURBO_RATIO_LIMIT, family, model); + + if (is_hybrid) + dump_turbo_ratio_limits(MSR_SECONDARY_TURBO_RATIO_LIMIT, family, model); + } if (has_atom_turbo_ratio_limit(family, model)) dump_atom_turbo_ratio_limits(); @@ -3499,10 +4107,36 @@ dump_cstate_pstate_config_info(unsigned int family, unsigned int model) if (has_config_tdp(family, model)) dump_config_tdp(); +} +static void dump_cstate_pstate_config_info(unsigned int family, unsigned int model) +{ + if (!do_nhm_platform_info) + return; + + dump_nhm_platform_info(); + dump_turbo_ratio_info(family, model); dump_nhm_cst_cfg(); } +static int read_sysfs_int(char *path) +{ + FILE *input; + int retval = -1; + + input = fopen(path, "r"); + if (input == NULL) { + if (debug) + fprintf(outf, "NSFOD %s\n", path); + return (-1); + } + if (fscanf(input, "%d", &retval) != 1) + err(1, "%s: failed to read int from file", path); + fclose(input); + + return (retval); +} + static void dump_sysfs_file(char *path) { FILE *input; @@ -3520,8 +4154,50 @@ static void dump_sysfs_file(char *path) fprintf(outf, "%s: %s", strrchr(path, '/') + 1, cpuidle_buf); } -static void -dump_sysfs_cstate_config(void) + +static void intel_uncore_frequency_probe(void) +{ + int i, j; + char path[128]; + + if (!genuine_intel) + return; + + if (access("/sys/devices/system/cpu/intel_uncore_frequency/package_00_die_00", R_OK)) + return; + + if (!access("/sys/devices/system/cpu/intel_uncore_frequency/package_00_die_00/current_freq_khz", R_OK)) + BIC_PRESENT(BIC_UNCORE_MHZ); + + if (quiet) + return; + + for (i = 0; i < topo.num_packages; ++i) { + for (j = 0; j < topo.num_die; ++j) { + int k, l; + + sprintf(path, "/sys/devices/system/cpu/intel_uncore_frequency/package_0%d_die_0%d/min_freq_khz", + i, j); + k = read_sysfs_int(path); + sprintf(path, "/sys/devices/system/cpu/intel_uncore_frequency/package_0%d_die_0%d/max_freq_khz", + i, j); + l = read_sysfs_int(path); + fprintf(outf, "Uncore Frequency pkg%d die%d: %d - %d MHz ", i, j, k / 1000, l / 1000); + + sprintf(path, + "/sys/devices/system/cpu/intel_uncore_frequency/package_0%d_die_0%d/initial_min_freq_khz", + i, j); + k = read_sysfs_int(path); + sprintf(path, + "/sys/devices/system/cpu/intel_uncore_frequency/package_0%d_die_0%d/initial_max_freq_khz", + i, j); + l = read_sysfs_int(path); + fprintf(outf, "(%d - %d MHz)\n", k / 1000, l / 1000); + } + } +} + +static void dump_sysfs_cstate_config(void) { char path[64]; char name_buf[16]; @@ -3530,9 +4206,6 @@ dump_sysfs_cstate_config(void) int state; char *sp; - if (!DO_BIC(BIC_sysfs)) - return; - if (access("/sys/devices/system/cpu/cpuidle", R_OK)) { fprintf(outf, "cpuidle not loaded\n"); return; @@ -3544,23 +4217,23 @@ dump_sysfs_cstate_config(void) for (state = 0; state < 10; ++state) { - sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", - base_cpu, state); + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", base_cpu, state); input = fopen(path, "r"); if (input == NULL) continue; if (!fgets(name_buf, sizeof(name_buf), input)) err(1, "%s: failed to read file", path); - /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ + /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ sp = strchr(name_buf, '-'); if (!sp) sp = strchrnul(name_buf, '\n'); *sp = '\0'; fclose(input); - sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/desc", - base_cpu, state); + remove_underbar(name_buf); + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/desc", base_cpu, state); input = fopen(path, "r"); if (input == NULL) continue; @@ -3571,8 +4244,8 @@ dump_sysfs_cstate_config(void) fclose(input); } } -static void -dump_sysfs_pstate_config(void) + +static void dump_sysfs_pstate_config(void) { char path[64]; char driver_buf[64]; @@ -3580,8 +4253,7 @@ dump_sysfs_pstate_config(void) FILE *input; int turbo; - sprintf(path, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_driver", - base_cpu); + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_driver", base_cpu); input = fopen(path, "r"); if (input == NULL) { fprintf(outf, "NSFOD %s\n", path); @@ -3591,8 +4263,7 @@ dump_sysfs_pstate_config(void) err(1, "%s: failed to read file", path); fclose(input); - sprintf(path, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_governor", - base_cpu); + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_governor", base_cpu); input = fopen(path, "r"); if (input == NULL) { fprintf(outf, "NSFOD %s\n", path); @@ -3624,16 +4295,17 @@ dump_sysfs_pstate_config(void) } } - /* * print_epb() * Decode the ENERGY_PERF_BIAS MSR */ int print_epb(struct thread_data *t, struct core_data *c, struct pkg_data *p) { - unsigned long long msr; char *epb_string; - int cpu; + int cpu, epb; + + UNUSED(c); + UNUSED(p); if (!has_epb) return 0; @@ -3645,14 +4317,15 @@ int print_epb(struct thread_data *t, struct core_data *c, struct pkg_data *p) return 0; if (cpu_migrate(cpu)) { - fprintf(outf, "Could not migrate to CPU %d\n", cpu); + fprintf(outf, "print_epb: Could not migrate to CPU %d\n", cpu); return -1; } - if (get_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, &msr)) + epb = get_epb(cpu); + if (epb < 0) return 0; - switch (msr & 0xF) { + switch (epb) { case ENERGY_PERF_BIAS_PERFORMANCE: epb_string = "performance"; break; @@ -3666,10 +4339,11 @@ int print_epb(struct thread_data *t, struct core_data *c, struct pkg_data *p) epb_string = "custom"; break; } - fprintf(outf, "cpu%d: MSR_IA32_ENERGY_PERF_BIAS: 0x%08llx (%s)\n", cpu, msr, epb_string); + fprintf(outf, "cpu%d: EPB: %d (%s)\n", cpu, epb, epb_string); return 0; } + /* * print_hwp() * Decode the MSR_HWP_CAPABILITIES @@ -3679,6 +4353,9 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) unsigned long long msr; int cpu; + UNUSED(c); + UNUSED(p); + if (!has_hwp) return 0; @@ -3689,15 +4366,14 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) return 0; if (cpu_migrate(cpu)) { - fprintf(outf, "Could not migrate to CPU %d\n", cpu); + fprintf(outf, "print_hwp: Could not migrate to CPU %d\n", cpu); return -1; } if (get_msr(cpu, MSR_PM_ENABLE, &msr)) return 0; - fprintf(outf, "cpu%d: MSR_PM_ENABLE: 0x%08llx (%sHWP)\n", - cpu, msr, (msr & (1 << 0)) ? "" : "No-"); + fprintf(outf, "cpu%d: MSR_PM_ENABLE: 0x%08llx (%sHWP)\n", cpu, msr, (msr & (1 << 0)) ? "" : "No-"); /* MSR_PM_ENABLE[1] == 1 if HWP is enabled and MSRs visible */ if ((msr & (1 << 0)) == 0) @@ -3707,25 +4383,23 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) return 0; fprintf(outf, "cpu%d: MSR_HWP_CAPABILITIES: 0x%08llx " - "(high %d guar %d eff %d low %d)\n", - cpu, msr, - (unsigned int)HWP_HIGHEST_PERF(msr), - (unsigned int)HWP_GUARANTEED_PERF(msr), - (unsigned int)HWP_MOSTEFFICIENT_PERF(msr), - (unsigned int)HWP_LOWEST_PERF(msr)); + "(high %d guar %d eff %d low %d)\n", + cpu, msr, + (unsigned int)HWP_HIGHEST_PERF(msr), + (unsigned int)HWP_GUARANTEED_PERF(msr), + (unsigned int)HWP_MOSTEFFICIENT_PERF(msr), (unsigned int)HWP_LOWEST_PERF(msr)); if (get_msr(cpu, MSR_HWP_REQUEST, &msr)) return 0; fprintf(outf, "cpu%d: MSR_HWP_REQUEST: 0x%08llx " - "(min %d max %d des %d epp 0x%x window 0x%x pkg 0x%x)\n", - cpu, msr, - (unsigned int)(((msr) >> 0) & 0xff), - (unsigned int)(((msr) >> 8) & 0xff), - (unsigned int)(((msr) >> 16) & 0xff), - (unsigned int)(((msr) >> 24) & 0xff), - (unsigned int)(((msr) >> 32) & 0xff3), - (unsigned int)(((msr) >> 42) & 0x1)); + "(min %d max %d des %d epp 0x%x window 0x%x pkg 0x%x)\n", + cpu, msr, + (unsigned int)(((msr) >> 0) & 0xff), + (unsigned int)(((msr) >> 8) & 0xff), + (unsigned int)(((msr) >> 16) & 0xff), + (unsigned int)(((msr) >> 24) & 0xff), + (unsigned int)(((msr) >> 32) & 0xff3), (unsigned int)(((msr) >> 42) & 0x1)); if (has_hwp_pkg) { if (get_msr(cpu, MSR_HWP_REQUEST_PKG, &msr)) @@ -3737,8 +4411,7 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) (unsigned int)(((msr) >> 0) & 0xff), (unsigned int)(((msr) >> 8) & 0xff), (unsigned int)(((msr) >> 16) & 0xff), - (unsigned int)(((msr) >> 24) & 0xff), - (unsigned int)(((msr) >> 32) & 0xff3)); + (unsigned int)(((msr) >> 24) & 0xff), (unsigned int)(((msr) >> 32) & 0xff3)); } if (has_hwp_notify) { if (get_msr(cpu, MSR_HWP_INTERRUPT, &msr)) @@ -3746,18 +4419,14 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) fprintf(outf, "cpu%d: MSR_HWP_INTERRUPT: 0x%08llx " "(%s_Guaranteed_Perf_Change, %s_Excursion_Min)\n", - cpu, msr, - ((msr) & 0x1) ? "EN" : "Dis", - ((msr) & 0x2) ? "EN" : "Dis"); + cpu, msr, ((msr) & 0x1) ? "EN" : "Dis", ((msr) & 0x2) ? "EN" : "Dis"); } if (get_msr(cpu, MSR_HWP_STATUS, &msr)) return 0; fprintf(outf, "cpu%d: MSR_HWP_STATUS: 0x%08llx " - "(%sGuaranteed_Perf_Change, %sExcursion_Min)\n", - cpu, msr, - ((msr) & 0x1) ? "" : "No-", - ((msr) & 0x2) ? "" : "No-"); + "(%sGuaranteed_Perf_Change, %sExcursion_Min)\n", + cpu, msr, ((msr) & 0x1) ? "" : "No-", ((msr) & 0x2) ? "" : "No-"); return 0; } @@ -3770,6 +4439,9 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data unsigned long long msr; int cpu; + UNUSED(c); + UNUSED(p); + cpu = t->cpu_id; /* per-package */ @@ -3777,7 +4449,7 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data return 0; if (cpu_migrate(cpu)) { - fprintf(outf, "Could not migrate to CPU %d\n", cpu); + fprintf(outf, "print_perf_limit: Could not migrate to CPU %d\n", cpu); return -1; } @@ -3797,8 +4469,7 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data (msr & 1 << 5) ? "Auto-HWP, " : "", (msr & 1 << 4) ? "Graphics, " : "", (msr & 1 << 2) ? "bit2, " : "", - (msr & 1 << 1) ? "ThermStatus, " : "", - (msr & 1 << 0) ? "PROCHOT, " : ""); + (msr & 1 << 1) ? "ThermStatus, " : "", (msr & 1 << 0) ? "PROCHOT, " : ""); fprintf(outf, " (Logged: %s%s%s%s%s%s%s%s%s%s%s%s%s%s)\n", (msr & 1 << 31) ? "bit31, " : "", (msr & 1 << 30) ? "bit30, " : "", @@ -3812,8 +4483,7 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data (msr & 1 << 21) ? "Auto-HWP, " : "", (msr & 1 << 20) ? "Graphics, " : "", (msr & 1 << 18) ? "bit18, " : "", - (msr & 1 << 17) ? "ThermStatus, " : "", - (msr & 1 << 16) ? "PROCHOT, " : ""); + (msr & 1 << 17) ? "ThermStatus, " : "", (msr & 1 << 16) ? "PROCHOT, " : ""); } if (do_gfx_perf_limit_reasons) { @@ -3826,8 +4496,7 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data (msr & 1 << 6) ? "VR-Therm, " : "", (msr & 1 << 8) ? "Amps, " : "", (msr & 1 << 9) ? "GFXPwr, " : "", - (msr & 1 << 10) ? "PkgPwrL1, " : "", - (msr & 1 << 11) ? "PkgPwrL2, " : ""); + (msr & 1 << 10) ? "PkgPwrL1, " : "", (msr & 1 << 11) ? "PkgPwrL2, " : ""); fprintf(outf, " (Logged: %s%s%s%s%s%s%s%s)\n", (msr & 1 << 16) ? "PROCHOT, " : "", (msr & 1 << 17) ? "ThermStatus, " : "", @@ -3835,8 +4504,7 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data (msr & 1 << 22) ? "VR-Therm, " : "", (msr & 1 << 24) ? "Amps, " : "", (msr & 1 << 25) ? "GFXPwr, " : "", - (msr & 1 << 26) ? "PkgPwrL1, " : "", - (msr & 1 << 27) ? "PkgPwrL2, " : ""); + (msr & 1 << 26) ? "PkgPwrL1, " : "", (msr & 1 << 27) ? "PkgPwrL2, " : ""); } if (do_ring_perf_limit_reasons) { get_msr(cpu, MSR_RING_PERF_LIMIT_REASONS, &msr); @@ -3846,21 +4514,19 @@ int print_perf_limit(struct thread_data *t, struct core_data *c, struct pkg_data (msr & 1 << 1) ? "ThermStatus, " : "", (msr & 1 << 6) ? "VR-Therm, " : "", (msr & 1 << 8) ? "Amps, " : "", - (msr & 1 << 10) ? "PkgPwrL1, " : "", - (msr & 1 << 11) ? "PkgPwrL2, " : ""); + (msr & 1 << 10) ? "PkgPwrL1, " : "", (msr & 1 << 11) ? "PkgPwrL2, " : ""); fprintf(outf, " (Logged: %s%s%s%s%s%s)\n", (msr & 1 << 16) ? "PROCHOT, " : "", (msr & 1 << 17) ? "ThermStatus, " : "", (msr & 1 << 22) ? "VR-Therm, " : "", (msr & 1 << 24) ? "Amps, " : "", - (msr & 1 << 26) ? "PkgPwrL1, " : "", - (msr & 1 << 27) ? "PkgPwrL2, " : ""); + (msr & 1 << 26) ? "PkgPwrL1, " : "", (msr & 1 << 27) ? "PkgPwrL2, " : ""); } return 0; } #define RAPL_POWER_GRANULARITY 0x7FFF /* 15 bit power granularity */ -#define RAPL_TIME_GRANULARITY 0x3F /* 6 bit time granularity */ +#define RAPL_TIME_GRANULARITY 0x3F /* 6 bit time granularity */ double get_tdp_intel(unsigned int model) { @@ -3881,28 +4547,26 @@ double get_tdp_intel(unsigned int model) double get_tdp_amd(unsigned int family) { - switch (family) { - case 0x17: - case 0x18: - default: - /* This is the max stock TDP of HEDT/Server Fam17h chips */ - return 250.0; - } + UNUSED(family); + + /* This is the max stock TDP of HEDT/Server Fam17h+ chips */ + return 280.0; } /* * rapl_dram_energy_units_probe() * Energy units are either hard-coded, or come from RAPL Energy Unit MSR. */ -static double -rapl_dram_energy_units_probe(int model, double rapl_energy_units) +static double rapl_dram_energy_units_probe(int model, double rapl_energy_units) { /* only called for genuine_intel, family 6 */ switch (model) { case INTEL_FAM6_HASWELL_X: /* HSX */ case INTEL_FAM6_BROADWELL_X: /* BDX */ + case INTEL_FAM6_SKYLAKE_X: /* SKX */ case INTEL_FAM6_XEON_PHI_KNL: /* KNL */ + case INTEL_FAM6_ICELAKE_X: /* ICX */ return (rapl_dram_energy_units = 15.3 / 1000000); default: return (rapl_energy_units); @@ -3946,7 +4610,9 @@ void rapl_probe_intel(unsigned int family, unsigned int model) BIC_PRESENT(BIC_PkgWatt); break; case INTEL_FAM6_ATOM_TREMONT: /* EHL */ - do_rapl = RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_GFX | RAPL_PKG_POWER_INFO; + do_rapl = + RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS + | RAPL_GFX | RAPL_PKG_POWER_INFO; if (rapl_joules) { BIC_PRESENT(BIC_Pkg_J); BIC_PRESENT(BIC_Cor_J); @@ -3959,9 +4625,19 @@ void rapl_probe_intel(unsigned int family, unsigned int model) BIC_PRESENT(BIC_GFXWatt); } break; + case INTEL_FAM6_ATOM_TREMONT_D: /* JVL */ + do_rapl = RAPL_PKG | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO; + BIC_PRESENT(BIC_PKG__); + if (rapl_joules) + BIC_PRESENT(BIC_Pkg_J); + else + BIC_PRESENT(BIC_PkgWatt); + break; case INTEL_FAM6_SKYLAKE_L: /* SKL */ case INTEL_FAM6_CANNONLAKE_L: /* CNL */ - do_rapl = RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_GFX | RAPL_PKG_POWER_INFO; + do_rapl = + RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS + | RAPL_GFX | RAPL_PKG_POWER_INFO; BIC_PRESENT(BIC_PKG__); BIC_PRESENT(BIC_RAM__); if (rapl_joules) { @@ -3979,8 +4655,12 @@ void rapl_probe_intel(unsigned int family, unsigned int model) case INTEL_FAM6_HASWELL_X: /* HSX */ case INTEL_FAM6_BROADWELL_X: /* BDX */ case INTEL_FAM6_SKYLAKE_X: /* SKX */ + case INTEL_FAM6_ICELAKE_X: /* ICX */ + case INTEL_FAM6_SAPPHIRERAPIDS_X: /* SPR */ case INTEL_FAM6_XEON_PHI_KNL: /* KNL */ - do_rapl = RAPL_PKG | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO; + do_rapl = + RAPL_PKG | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | + RAPL_PKG_POWER_INFO; BIC_PRESENT(BIC_PKG__); BIC_PRESENT(BIC_RAM__); if (rapl_joules) { @@ -3993,7 +4673,9 @@ void rapl_probe_intel(unsigned int family, unsigned int model) break; case INTEL_FAM6_SANDYBRIDGE_X: case INTEL_FAM6_IVYBRIDGE_X: - do_rapl = RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_PKG_PERF_STATUS | RAPL_DRAM_PERF_STATUS | RAPL_PKG_POWER_INFO; + do_rapl = + RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_PKG_PERF_STATUS | + RAPL_DRAM_PERF_STATUS | RAPL_PKG_POWER_INFO; BIC_PRESENT(BIC_PKG__); BIC_PRESENT(BIC_RAM__); if (rapl_joules) { @@ -4018,7 +4700,9 @@ void rapl_probe_intel(unsigned int family, unsigned int model) } break; case INTEL_FAM6_ATOM_GOLDMONT_D: /* DNV */ - do_rapl = RAPL_PKG | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO | RAPL_CORES_ENERGY_STATUS; + do_rapl = + RAPL_PKG | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | + RAPL_PKG_POWER_INFO | RAPL_CORES_ENERGY_STATUS; BIC_PRESENT(BIC_PKG__); BIC_PRESENT(BIC_RAM__); if (rapl_joules) { @@ -4067,29 +4751,24 @@ void rapl_probe_amd(unsigned int family, unsigned int model) unsigned int has_rapl = 0; double tdp; + UNUSED(model); + if (max_extended_level >= 0x80000007) { __cpuid(0x80000007, eax, ebx, ecx, edx); - /* RAPL (Fam 17h) */ + /* RAPL (Fam 17h+) */ has_rapl = edx & (1 << 14); } - if (!has_rapl) + if (!has_rapl || family < 0x17) return; - switch (family) { - case 0x17: /* Zen, Zen+ */ - case 0x18: /* Hygon Dhyana */ - do_rapl = RAPL_AMD_F17H | RAPL_PER_CORE_ENERGY; - if (rapl_joules) { - BIC_PRESENT(BIC_Pkg_J); - BIC_PRESENT(BIC_Cor_J); - } else { - BIC_PRESENT(BIC_PkgWatt); - BIC_PRESENT(BIC_CorWatt); - } - break; - default: - return; + do_rapl = RAPL_AMD_F17H | RAPL_PER_CORE_ENERGY; + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_Cor_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_CorWatt); } if (get_msr(base_cpu, MSR_RAPL_PWR_UNIT, &msr)) @@ -4132,6 +4811,7 @@ void perf_limit_reasons_probe(unsigned int family, unsigned int model) case INTEL_FAM6_HASWELL_L: /* HSW */ case INTEL_FAM6_HASWELL_G: /* HSW */ do_gfx_perf_limit_reasons = 1; + /* FALLTHRU */ case INTEL_FAM6_HASWELL_X: /* HSX */ do_core_perf_limit_reasons = 1; do_ring_perf_limit_reasons = 1; @@ -4142,8 +4822,20 @@ void perf_limit_reasons_probe(unsigned int family, unsigned int model) void automatic_cstate_conversion_probe(unsigned int family, unsigned int model) { - if (is_skx(family, model) || is_bdx(family, model)) + if (family != 6) + return; + + switch (model) { + case INTEL_FAM6_BROADWELL_X: + case INTEL_FAM6_SKYLAKE_X: has_automatic_cstate_conversion = 1; + } +} + +void prewake_cstate_probe(unsigned int family, unsigned int model) +{ + if (is_icx(family, model) || is_spr(family, model)) + dis_cstate_prewake = 1; } int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p) @@ -4152,6 +4844,9 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p unsigned int dts, dts2; int cpu; + UNUSED(c); + UNUSED(p); + if (!(do_dts || do_ptm)) return 0; @@ -4162,7 +4857,7 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p return 0; if (cpu_migrate(cpu)) { - fprintf(outf, "Could not migrate to CPU %d\n", cpu); + fprintf(outf, "print_thermal: Could not migrate to CPU %d\n", cpu); return -1; } @@ -4171,8 +4866,7 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p return 0; dts = (msr >> 16) & 0x7F; - fprintf(outf, "cpu%d: MSR_IA32_PACKAGE_THERM_STATUS: 0x%08llx (%d C)\n", - cpu, msr, tcc_activation_temp - dts); + fprintf(outf, "cpu%d: MSR_IA32_PACKAGE_THERM_STATUS: 0x%08llx (%d C)\n", cpu, msr, tj_max - dts); if (get_msr(cpu, MSR_IA32_PACKAGE_THERM_INTERRUPT, &msr)) return 0; @@ -4180,10 +4874,9 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p dts = (msr >> 16) & 0x7F; dts2 = (msr >> 8) & 0x7F; fprintf(outf, "cpu%d: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x%08llx (%d C, %d C)\n", - cpu, msr, tcc_activation_temp - dts, tcc_activation_temp - dts2); + cpu, msr, tj_max - dts, tj_max - dts2); } - if (do_dts && debug) { unsigned int resolution; @@ -4193,7 +4886,7 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p dts = (msr >> 16) & 0x7F; resolution = (msr >> 27) & 0xF; fprintf(outf, "cpu%d: MSR_IA32_THERM_STATUS: 0x%08llx (%d C +/- %d)\n", - cpu, msr, tcc_activation_temp - dts, resolution); + cpu, msr, tj_max - dts, resolution); if (get_msr(cpu, MSR_IA32_THERM_INTERRUPT, &msr)) return 0; @@ -4201,7 +4894,7 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p dts = (msr >> 16) & 0x7F; dts2 = (msr >> 8) & 0x7F; fprintf(outf, "cpu%d: MSR_IA32_THERM_INTERRUPT: 0x%08llx (%d C, %d C)\n", - cpu, msr, tcc_activation_temp - dts, tcc_activation_temp - dts2); + cpu, msr, tj_max - dts, tj_max - dts2); } return 0; @@ -4209,11 +4902,11 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p void print_power_limit_msr(int cpu, unsigned long long msr, char *label) { - fprintf(outf, "cpu%d: %s: %sabled (%f Watts, %f sec, clamp %sabled)\n", + fprintf(outf, "cpu%d: %s: %sabled (%0.3f Watts, %f sec, clamp %sabled)\n", cpu, label, ((msr >> 15) & 1) ? "EN" : "DIS", ((msr >> 0) & 0x7FFF) * rapl_power_units, - (1.0 + (((msr >> 22) & 0x3)/4.0)) * (1 << ((msr >> 17) & 0x1F)) * rapl_time_units, + (1.0 + (((msr >> 22) & 0x3) / 4.0)) * (1 << ((msr >> 17) & 0x1F)) * rapl_time_units, (((msr >> 16) & 1) ? "EN" : "DIS")); return; @@ -4225,6 +4918,9 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) const char *msr_name; int cpu; + UNUSED(c); + UNUSED(p); + if (!do_rapl) return 0; @@ -4234,7 +4930,7 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) cpu = t->cpu_id; if (cpu_migrate(cpu)) { - fprintf(outf, "Could not migrate to CPU %d\n", cpu); + fprintf(outf, "print_rapl: Could not migrate to CPU %d\n", cpu); return -1; } @@ -4254,12 +4950,11 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) if (do_rapl & RAPL_PKG_POWER_INFO) { if (get_msr(cpu, MSR_PKG_POWER_INFO, &msr)) - return -5; - + return -5; fprintf(outf, "cpu%d: MSR_PKG_POWER_INFO: 0x%08llx (%.0f W TDP, RAPL %.0f - %.0f W, %f sec.)\n", cpu, msr, - ((msr >> 0) & RAPL_POWER_GRANULARITY) * rapl_power_units, + ((msr >> 0) & RAPL_POWER_GRANULARITY) * rapl_power_units, ((msr >> 16) & RAPL_POWER_GRANULARITY) * rapl_power_units, ((msr >> 32) & RAPL_POWER_GRANULARITY) * rapl_power_units, ((msr >> 48) & RAPL_TIME_GRANULARITY) * rapl_time_units); @@ -4274,21 +4969,28 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) cpu, msr, (msr >> 63) & 1 ? "" : "UN"); print_power_limit_msr(cpu, msr, "PKG Limit #1"); - fprintf(outf, "cpu%d: PKG Limit #2: %sabled (%f Watts, %f* sec, clamp %sabled)\n", + fprintf(outf, "cpu%d: PKG Limit #2: %sabled (%0.3f Watts, %f* sec, clamp %sabled)\n", cpu, ((msr >> 47) & 1) ? "EN" : "DIS", ((msr >> 32) & 0x7FFF) * rapl_power_units, - (1.0 + (((msr >> 54) & 0x3)/4.0)) * (1 << ((msr >> 49) & 0x1F)) * rapl_time_units, + (1.0 + (((msr >> 54) & 0x3) / 4.0)) * (1 << ((msr >> 49) & 0x1F)) * rapl_time_units, ((msr >> 48) & 1) ? "EN" : "DIS"); + + if (get_msr(cpu, MSR_VR_CURRENT_CONFIG, &msr)) + return -9; + + fprintf(outf, "cpu%d: MSR_VR_CURRENT_CONFIG: 0x%08llx\n", cpu, msr); + fprintf(outf, "cpu%d: PKG Limit #4: %f Watts (%slocked)\n", + cpu, ((msr >> 0) & 0x1FFF) * rapl_power_units, (msr >> 31) & 1 ? "" : "UN"); } if (do_rapl & RAPL_DRAM_POWER_INFO) { if (get_msr(cpu, MSR_DRAM_POWER_INFO, &msr)) - return -6; + return -6; fprintf(outf, "cpu%d: MSR_DRAM_POWER_INFO,: 0x%08llx (%.0f W TDP, RAPL %.0f - %.0f W, %f sec.)\n", cpu, msr, - ((msr >> 0) & RAPL_POWER_GRANULARITY) * rapl_power_units, + ((msr >> 0) & RAPL_POWER_GRANULARITY) * rapl_power_units, ((msr >> 16) & RAPL_POWER_GRANULARITY) * rapl_power_units, ((msr >> 32) & RAPL_POWER_GRANULARITY) * rapl_power_units, ((msr >> 48) & RAPL_TIME_GRANULARITY) * rapl_time_units); @@ -4297,7 +4999,7 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) if (get_msr(cpu, MSR_DRAM_POWER_LIMIT, &msr)) return -9; fprintf(outf, "cpu%d: MSR_DRAM_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 31) & 1 ? "" : "UN"); + cpu, msr, (msr >> 31) & 1 ? "" : "UN"); print_power_limit_msr(cpu, msr, "DRAM Limit"); } @@ -4311,7 +5013,7 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) if (get_msr(cpu, MSR_PP0_POWER_LIMIT, &msr)) return -9; fprintf(outf, "cpu%d: MSR_PP0_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 31) & 1 ? "" : "UN"); + cpu, msr, (msr >> 31) & 1 ? "" : "UN"); print_power_limit_msr(cpu, msr, "Cores Limit"); } if (do_rapl & RAPL_GFX) { @@ -4323,7 +5025,7 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) if (get_msr(cpu, MSR_PP1_POWER_LIMIT, &msr)) return -9; fprintf(outf, "cpu%d: MSR_PP1_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 31) & 1 ? "" : "UN"); + cpu, msr, (msr >> 31) & 1 ? "" : "UN"); print_power_limit_msr(cpu, msr, "GFX Limit"); } return 0; @@ -4342,25 +5044,31 @@ int has_snb_msrs(unsigned int family, unsigned int model) if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_SANDYBRIDGE: case INTEL_FAM6_SANDYBRIDGE_X: - case INTEL_FAM6_IVYBRIDGE: /* IVB */ - case INTEL_FAM6_IVYBRIDGE_X: /* IVB Xeon */ - case INTEL_FAM6_HASWELL: /* HSW */ - case INTEL_FAM6_HASWELL_X: /* HSW */ - case INTEL_FAM6_HASWELL_L: /* HSW */ - case INTEL_FAM6_HASWELL_G: /* HSW */ - case INTEL_FAM6_BROADWELL: /* BDW */ - case INTEL_FAM6_BROADWELL_G: /* BDW */ - case INTEL_FAM6_BROADWELL_X: /* BDX */ - case INTEL_FAM6_SKYLAKE_L: /* SKL */ - case INTEL_FAM6_CANNONLAKE_L: /* CNL */ - case INTEL_FAM6_SKYLAKE_X: /* SKX */ - case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ + case INTEL_FAM6_IVYBRIDGE: /* IVB */ + case INTEL_FAM6_IVYBRIDGE_X: /* IVB Xeon */ + case INTEL_FAM6_HASWELL: /* HSW */ + case INTEL_FAM6_HASWELL_X: /* HSW */ + case INTEL_FAM6_HASWELL_L: /* HSW */ + case INTEL_FAM6_HASWELL_G: /* HSW */ + case INTEL_FAM6_BROADWELL: /* BDW */ + case INTEL_FAM6_BROADWELL_G: /* BDW */ + case INTEL_FAM6_BROADWELL_X: /* BDX */ + case INTEL_FAM6_SKYLAKE_L: /* SKL */ + case INTEL_FAM6_CANNONLAKE_L: /* CNL */ + case INTEL_FAM6_SKYLAKE_X: /* SKX */ + case INTEL_FAM6_ICELAKE_X: /* ICX */ + case INTEL_FAM6_SAPPHIRERAPIDS_X: /* SPR */ + case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ case INTEL_FAM6_ATOM_GOLDMONT_PLUS: case INTEL_FAM6_ATOM_GOLDMONT_D: /* DNV */ - case INTEL_FAM6_ATOM_TREMONT: /* EHL */ + case INTEL_FAM6_ATOM_TREMONT: /* EHL */ + case INTEL_FAM6_ATOM_TREMONT_D: /* JVL */ return 1; } return 0; @@ -4383,6 +5091,9 @@ int has_c8910_msrs(unsigned int family, unsigned int model) if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_HASWELL_L: /* HSW */ case INTEL_FAM6_BROADWELL: /* BDW */ @@ -4409,6 +5120,9 @@ int has_skl_msrs(unsigned int family, unsigned int model) if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_SKYLAKE_L: /* SKL */ case INTEL_FAM6_CANNONLAKE_L: /* CNL */ @@ -4421,6 +5135,10 @@ int is_slm(unsigned int family, unsigned int model) { if (!genuine_intel) return 0; + + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_ATOM_SILVERMONT: /* BYT */ case INTEL_FAM6_ATOM_SILVERMONT_D: /* AVN */ @@ -4433,6 +5151,10 @@ int is_knl(unsigned int family, unsigned int model) { if (!genuine_intel) return 0; + + if (family != 6) + return 0; + switch (model) { case INTEL_FAM6_XEON_PHI_KNL: /* KNL */ return 1; @@ -4445,8 +5167,11 @@ int is_cnl(unsigned int family, unsigned int model) if (!genuine_intel) return 0; + if (family != 6) + return 0; + switch (model) { - case INTEL_FAM6_CANNONLAKE_L: /* CNL */ + case INTEL_FAM6_CANNONLAKE_L: /* CNL */ return 1; } @@ -4461,7 +5186,7 @@ unsigned int get_aperf_mperf_multiplier(unsigned int family, unsigned int model) } #define SLM_BCLK_FREQS 5 -double slm_freq_table[SLM_BCLK_FREQS] = { 83.3, 100.0, 133.3, 116.7, 80.0}; +double slm_freq_table[SLM_BCLK_FREQS] = { 83.3, 100.0, 133.3, 116.7, 80.0 }; double slm_bclk(void) { @@ -4495,6 +5220,31 @@ double discover_bclk(unsigned int family, unsigned int model) return 133.33; } +int get_cpu_type(struct thread_data *t, struct core_data *c, struct pkg_data *p) +{ + unsigned int eax, ebx, ecx, edx; + + UNUSED(c); + UNUSED(p); + + if (!genuine_intel) + return 0; + + if (cpu_migrate(t->cpu_id)) { + fprintf(outf, "Could not migrate to CPU %d\n", t->cpu_id); + return -1; + } + + if (max_level < 0x1a) + return 0; + + __cpuid(0x1a, eax, ebx, ecx, edx); + eax = (eax >> 24) & 0xFF; + if (eax == 0x20) + t->is_atom = true; + return 0; +} + /* * MSR_IA32_TEMPERATURE_TARGET indicates the temperature where * the Thermal Control Circuit (TCC) activates. @@ -4510,10 +5260,13 @@ double discover_bclk(unsigned int family, unsigned int model) int set_temperature_target(struct thread_data *t, struct core_data *c, struct pkg_data *p) { unsigned long long msr; - unsigned int target_c_local; + unsigned int tcc_default, tcc_offset; int cpu; - /* tcc_activation_temp is used only for dts or ptm */ + UNUSED(c); + UNUSED(p); + + /* tj_max is used only for dts or ptm */ if (!(do_dts || do_ptm)) return 0; @@ -4527,10 +5280,9 @@ int set_temperature_target(struct thread_data *t, struct core_data *c, struct pk return -1; } - if (tcc_activation_temp_override != 0) { - tcc_activation_temp = tcc_activation_temp_override; - fprintf(outf, "cpu%d: Using cmdline TCC Target (%d C)\n", - cpu, tcc_activation_temp); + if (tj_max_override != 0) { + tj_max = tj_max_override; + fprintf(outf, "cpu%d: Using cmdline TCC Target (%d C)\n", cpu, tj_max); return 0; } @@ -4541,23 +5293,36 @@ int set_temperature_target(struct thread_data *t, struct core_data *c, struct pk if (get_msr(base_cpu, MSR_IA32_TEMPERATURE_TARGET, &msr)) goto guess; - target_c_local = (msr >> 16) & 0xFF; + tcc_default = (msr >> 16) & 0xFF; - if (!quiet) - fprintf(outf, "cpu%d: MSR_IA32_TEMPERATURE_TARGET: 0x%08llx (%d C)\n", - cpu, msr, target_c_local); + if (!quiet) { + switch (tcc_offset_bits) { + case 4: + tcc_offset = (msr >> 24) & 0xF; + fprintf(outf, "cpu%d: MSR_IA32_TEMPERATURE_TARGET: 0x%08llx (%d C) (%d default - %d offset)\n", + cpu, msr, tcc_default - tcc_offset, tcc_default, tcc_offset); + break; + case 6: + tcc_offset = (msr >> 24) & 0x3F; + fprintf(outf, "cpu%d: MSR_IA32_TEMPERATURE_TARGET: 0x%08llx (%d C) (%d default - %d offset)\n", + cpu, msr, tcc_default - tcc_offset, tcc_default, tcc_offset); + break; + default: + fprintf(outf, "cpu%d: MSR_IA32_TEMPERATURE_TARGET: 0x%08llx (%d C)\n", cpu, msr, tcc_default); + break; + } + } - if (!target_c_local) + if (!tcc_default) goto guess; - tcc_activation_temp = target_c_local; + tj_max = tcc_default; return 0; guess: - tcc_activation_temp = TJMAX_DEFAULT; - fprintf(outf, "cpu%d: Guessing tjMax %d C, Please use -T to specify\n", - cpu, tcc_activation_temp); + tj_max = TJMAX_DEFAULT; + fprintf(outf, "cpu%d: Guessing tjMax %d C, Please use -T to specify\n", cpu, tj_max); return 0; } @@ -4568,9 +5333,7 @@ void decode_feature_control_msr(void) if (!get_msr(base_cpu, MSR_IA32_FEAT_CTL, &msr)) fprintf(outf, "cpu%d: MSR_IA32_FEATURE_CONTROL: 0x%08llx (%sLocked %s)\n", - base_cpu, msr, - msr & FEAT_CTL_LOCKED ? "" : "UN-", - msr & (1 << 18) ? "SGX" : ""); + base_cpu, msr, msr & FEAT_CTL_LOCKED ? "" : "UN-", msr & (1 << 18) ? "SGX" : ""); } void decode_misc_enable_msr(void) @@ -4598,13 +5361,12 @@ void decode_misc_feature_control(void) return; if (!get_msr(base_cpu, MSR_MISC_FEATURE_CONTROL, &msr)) - fprintf(outf, "cpu%d: MSR_MISC_FEATURE_CONTROL: 0x%08llx (%sL2-Prefetch %sL2-Prefetch-pair %sL1-Prefetch %sL1-IP-Prefetch)\n", - base_cpu, msr, - msr & (0 << 0) ? "No-" : "", - msr & (1 << 0) ? "No-" : "", - msr & (2 << 0) ? "No-" : "", - msr & (3 << 0) ? "No-" : ""); + fprintf(outf, + "cpu%d: MSR_MISC_FEATURE_CONTROL: 0x%08llx (%sL2-Prefetch %sL2-Prefetch-pair %sL1-Prefetch %sL1-IP-Prefetch)\n", + base_cpu, msr, msr & (0 << 0) ? "No-" : "", msr & (1 << 0) ? "No-" : "", + msr & (2 << 0) ? "No-" : "", msr & (3 << 0) ? "No-" : ""); } + /* * Decode MSR_MISC_PWR_MGMT * @@ -4625,10 +5387,9 @@ void decode_misc_pwr_mgmt_msr(void) if (!get_msr(base_cpu, MSR_MISC_PWR_MGMT, &msr)) fprintf(outf, "cpu%d: MSR_MISC_PWR_MGMT: 0x%08llx (%sable-EIST_Coordination %sable-EPB %sable-OOB)\n", base_cpu, msr, - msr & (1 << 0) ? "DIS" : "EN", - msr & (1 << 1) ? "EN" : "DIS", - msr & (1 << 8) ? "EN" : "DIS"); + msr & (1 << 0) ? "DIS" : "EN", msr & (1 << 1) ? "EN" : "DIS", msr & (1 << 8) ? "EN" : "DIS"); } + /* * Decode MSR_CC6_DEMOTION_POLICY_CONFIG, MSR_MC6_DEMOTION_POLICY_CONFIG * @@ -4654,10 +5415,10 @@ void decode_c6_demotion_policy_msr(void) unsigned int intel_model_duplicates(unsigned int model) { - switch(model) { + switch (model) { case INTEL_FAM6_NEHALEM_EP: /* Core i7, Xeon 5500 series - Bloomfield, Gainstown NHM-EP */ case INTEL_FAM6_NEHALEM: /* Core i7 and i5 Processor - Clarksfield, Lynnfield, Jasper Forest */ - case 0x1F: /* Core i7 and i5 Processor - Nehalem */ + case 0x1F: /* Core i7 and i5 Processor - Nehalem */ case INTEL_FAM6_WESTMERE: /* Westmere Client - Clarkdale, Arrandale */ case INTEL_FAM6_WESTMERE_EP: /* Westmere EP - Gulftown */ return INTEL_FAM6_NEHALEM; @@ -4685,24 +5446,75 @@ unsigned int intel_model_duplicates(unsigned int model) case INTEL_FAM6_ICELAKE_NNPI: case INTEL_FAM6_TIGERLAKE_L: case INTEL_FAM6_TIGERLAKE: + case INTEL_FAM6_ROCKETLAKE: + case INTEL_FAM6_LAKEFIELD: + case INTEL_FAM6_ALDERLAKE: + case INTEL_FAM6_ALDERLAKE_L: + case INTEL_FAM6_ALDERLAKE_N: + case INTEL_FAM6_RAPTORLAKE: + case INTEL_FAM6_RAPTORLAKE_P: + case INTEL_FAM6_RAPTORLAKE_S: + case INTEL_FAM6_METEORLAKE: + case INTEL_FAM6_METEORLAKE_L: return INTEL_FAM6_CANNONLAKE_L; - case INTEL_FAM6_ATOM_TREMONT_D: - return INTEL_FAM6_ATOM_GOLDMONT_D; - case INTEL_FAM6_ATOM_TREMONT_L: return INTEL_FAM6_ATOM_TREMONT; - case INTEL_FAM6_ICELAKE_X: - return INTEL_FAM6_SKYLAKE_X; + case INTEL_FAM6_ICELAKE_D: + return INTEL_FAM6_ICELAKE_X; } return model; } + +void print_dev_latency(void) +{ + char *path = "/dev/cpu_dma_latency"; + int fd; + int value; + int retval; + + fd = open(path, O_RDONLY); + if (fd < 0) { + warn("fopen %s\n", path); + return; + } + + retval = read(fd, (void *)&value, sizeof(int)); + if (retval != sizeof(int)) { + warn("read %s\n", path); + close(fd); + return; + } + fprintf(outf, "/dev/cpu_dma_latency: %d usec (%s)\n", value, value == 2000000000 ? "default" : "constrained"); + + close(fd); +} + +/* + * Linux-perf manages the HW instructions-retired counter + * by enabling when requested, and hiding rollover + */ +void linux_perf_init(void) +{ + if (!BIC_IS_ENABLED(BIC_IPC)) + return; + + if (access("/proc/sys/kernel/perf_event_paranoid", F_OK)) + return; + + fd_instr_count_percpu = calloc(topo.max_cpu_num + 1, sizeof(int)); + if (fd_instr_count_percpu == NULL) + err(-1, "calloc fd_instr_count_percpu"); + + BIC_PRESENT(BIC_IPC); +} + void process_cpuid() { unsigned int eax, ebx, ecx, edx; unsigned int fms, family, model, stepping, ecx_flags, edx_flags; - unsigned int has_turbo; + unsigned long long ucode_patch = 0; eax = ebx = ecx = edx = 0; @@ -4716,8 +5528,8 @@ void process_cpuid() hygon_genuine = 1; if (!quiet) - fprintf(outf, "CPUID(0): %.4s%.4s%.4s ", - (char *)&ebx, (char *)&edx, (char *)&ecx); + fprintf(outf, "CPUID(0): %.4s%.4s%.4s 0x%x CPUID levels\n", + (char *)&ebx, (char *)&edx, (char *)&ecx, max_level); __cpuid(1, fms, ebx, ecx, edx); family = (fms >> 8) & 0xf; @@ -4730,6 +5542,9 @@ void process_cpuid() ecx_flags = ecx; edx_flags = edx; + if (get_msr(sched_getcpu(), MSR_IA32_UCODE_REV, &ucode_patch)) + warnx("get_msr(UCODE)\n"); + /* * check max extended function levels of CPUID. * This is needed to check for invariant TSC. @@ -4739,8 +5554,10 @@ void process_cpuid() __cpuid(0x80000000, max_extended_level, ebx, ecx, edx); if (!quiet) { - fprintf(outf, "0x%x CPUID levels; 0x%x xlevels; family:model:stepping 0x%x:%x:%x (%d:%d:%d)\n", - max_level, max_extended_level, family, model, stepping, family, model, stepping); + fprintf(outf, "CPUID(1): family:model:stepping 0x%x:%x:%x (%d:%d:%d) microcode 0x%x\n", + family, model, stepping, family, model, stepping, + (unsigned int)((ucode_patch >> 32) & 0xFFFFFFFF)); + fprintf(outf, "CPUID(0x80000000): max_extended_levels: 0x%x\n", max_extended_level); fprintf(outf, "CPUID(1): %s %s %s %s %s %s %s %s %s %s\n", ecx_flags & (1 << 0) ? "SSE3" : "-", ecx_flags & (1 << 3) ? "MONITOR" : "-", @@ -4750,11 +5567,12 @@ void process_cpuid() edx_flags & (1 << 4) ? "TSC" : "-", edx_flags & (1 << 5) ? "MSR" : "-", edx_flags & (1 << 22) ? "ACPI-TM" : "-", - edx_flags & (1 << 28) ? "HT" : "-", - edx_flags & (1 << 29) ? "TM" : "-"); + edx_flags & (1 << 28) ? "HT" : "-", edx_flags & (1 << 29) ? "TM" : "-"); } - if (genuine_intel) + if (genuine_intel) { + model_orig = model; model = intel_model_duplicates(model); + } if (!(edx_flags & (1 << 5))) errx(1, "CPUID: no MSR"); @@ -4805,14 +5623,11 @@ void process_cpuid() has_hwp ? "" : "No-", has_hwp_notify ? "" : "No-", has_hwp_activity_window ? "" : "No-", - has_hwp_epp ? "" : "No-", - has_hwp_pkg ? "" : "No-", - has_epb ? "" : "No-"); + has_hwp_epp ? "" : "No-", has_hwp_pkg ? "" : "No-", has_epb ? "" : "No-"); if (!quiet) decode_misc_enable_msr(); - if (max_level >= 0x7 && !quiet) { int has_sgx; @@ -4821,7 +5636,10 @@ void process_cpuid() __cpuid_count(0x7, 0, eax, ebx, ecx, edx); has_sgx = ebx & (1 << 2); - fprintf(outf, "CPUID(7): %sSGX\n", has_sgx ? "" : "No-"); + + is_hybrid = edx & (1 << 15); + + fprintf(outf, "CPUID(7): %sSGX %sHybrid\n", has_sgx ? "" : "No-", is_hybrid ? "" : "No-"); if (has_sgx) decode_feature_control_msr(); @@ -4844,7 +5662,7 @@ void process_cpuid() eax_crystal, ebx_tsc, crystal_hz); if (crystal_hz == 0) - switch(model) { + switch (model) { case INTEL_FAM6_SKYLAKE_L: /* SKL */ crystal_hz = 24000000; /* 24.0 MHz */ break; @@ -4857,13 +5675,13 @@ void process_cpuid() break; default: crystal_hz = 0; - } + } if (crystal_hz) { - tsc_hz = (unsigned long long) crystal_hz * ebx_tsc / eax_crystal; + tsc_hz = (unsigned long long)crystal_hz *ebx_tsc / eax_crystal; if (!quiet) fprintf(outf, "TSC: %lld MHz (%d Hz * %d / %d / 1000000)\n", - tsc_hz / 1000000, crystal_hz, ebx_tsc, eax_crystal); + tsc_hz / 1000000, crystal_hz, ebx_tsc, eax_crystal); } } } @@ -4916,6 +5734,14 @@ void process_cpuid() BIC_PRESENT(BIC_Mod_c6); use_c1_residency_msr = 1; } + if (is_jvl(family, model)) { + BIC_NOT_PRESENT(BIC_CPU_c3); + BIC_NOT_PRESENT(BIC_CPU_c7); + BIC_NOT_PRESENT(BIC_Pkgpc2); + BIC_NOT_PRESENT(BIC_Pkgpc3); + BIC_NOT_PRESENT(BIC_Pkgpc6); + BIC_NOT_PRESENT(BIC_Pkgpc7); + } if (is_dnv(family, model)) { BIC_PRESENT(BIC_CPU_c1); BIC_NOT_PRESENT(BIC_CPU_c3); @@ -4924,7 +5750,7 @@ void process_cpuid() BIC_NOT_PRESENT(BIC_Pkgpc7); use_c1_residency_msr = 1; } - if (is_skx(family, model)) { + if (is_skx(family, model) || is_icx(family, model) || is_spr(family, model)) { BIC_NOT_PRESENT(BIC_CPU_c3); BIC_NOT_PRESENT(BIC_Pkgpc3); BIC_NOT_PRESENT(BIC_CPU_c7); @@ -4935,9 +5761,12 @@ void process_cpuid() BIC_NOT_PRESENT(BIC_Pkgpc7); } if (has_c8910_msrs(family, model)) { - BIC_PRESENT(BIC_Pkgpc8); - BIC_PRESENT(BIC_Pkgpc9); - BIC_PRESENT(BIC_Pkgpc10); + if (pkg_cstate_limit >= PCL__8) + BIC_PRESENT(BIC_Pkgpc8); + if (pkg_cstate_limit >= PCL__9) + BIC_PRESENT(BIC_Pkgpc9); + if (pkg_cstate_limit >= PCL_10) + BIC_PRESENT(BIC_Pkgpc10); } do_irtl_hsw = has_c8910_msrs(family, model); if (has_skl_msrs(family, model)) { @@ -4947,10 +5776,9 @@ void process_cpuid() BIC_PRESENT(BIC_CPUGFX); } do_slm_cstates = is_slm(family, model); - do_knl_cstates = is_knl(family, model); + do_knl_cstates = is_knl(family, model); - if (do_slm_cstates || do_knl_cstates || is_cnl(family, model) || - is_ehl(family, model)) + if (do_slm_cstates || do_knl_cstates || is_cnl(family, model) || is_ehl(family, model)) BIC_NOT_PRESENT(BIC_CPU_c3); if (!quiet) @@ -4963,15 +5791,20 @@ void process_cpuid() perf_limit_reasons_probe(family, model); automatic_cstate_conversion_probe(family, model); + check_tcc_offset(model_orig); + if (!quiet) dump_cstate_pstate_config_info(family, model); + intel_uncore_frequency_probe(); if (!quiet) + print_dev_latency(); + if (!quiet) dump_sysfs_cstate_config(); if (!quiet) dump_sysfs_pstate_config(); - if (has_skl_msrs(family, model)) + if (has_skl_msrs(family, model) || is_ehl(family, model)) calculate_tsc_tweak(); if (!access("/sys/class/drm/card0/power/rc6_residency_ms", R_OK)) @@ -4980,11 +5813,19 @@ void process_cpuid() if (!access("/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz", R_OK)) BIC_PRESENT(BIC_GFXMHz); + if (!access("/sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz", R_OK)) + BIC_PRESENT(BIC_GFXACTMHz); + if (!access("/sys/devices/system/cpu/cpuidle/low_power_idle_cpu_residency_us", R_OK)) BIC_PRESENT(BIC_CPU_LPI); else BIC_NOT_PRESENT(BIC_CPU_LPI); + if (!access("/sys/devices/system/cpu/cpu0/thermal_throttle/core_throttle_count", R_OK)) + BIC_PRESENT(BIC_CORE_THROT_CNT); + else + BIC_NOT_PRESENT(BIC_CORE_THROT_CNT); + if (!access(sys_lpi_file_sysfs, R_OK)) { sys_lpi_file = sys_lpi_file_sysfs; BIC_PRESENT(BIC_SYS_LPI); @@ -5014,11 +5855,6 @@ int dir_filter(const struct dirent *dirp) return 0; } -int open_dev_cpu_msr(int dummy1) -{ - return 0; -} - void topology_probe() { int i; @@ -5037,7 +5873,7 @@ void topology_probe() if (debug > 1) fprintf(outf, "num_cpus %d max_cpu_num %d\n", topo.num_cpus, topo.max_cpu_num); - cpus = calloc(1, (topo.max_cpu_num + 1) * sizeof(struct cpu_topology)); + cpus = calloc(1, (topo.max_cpu_num + 1) * sizeof(struct cpu_topology)); if (cpus == NULL) err(1, "calloc cpus"); @@ -5116,22 +5952,19 @@ void topology_probe() topo.cores_per_node = max_core_id + 1; if (debug > 1) - fprintf(outf, "max_core_id %d, sizing for %d cores per package\n", - max_core_id, topo.cores_per_node); + fprintf(outf, "max_core_id %d, sizing for %d cores per package\n", max_core_id, topo.cores_per_node); if (!summary_only && topo.cores_per_node > 1) BIC_PRESENT(BIC_Core); topo.num_die = max_die_id + 1; if (debug > 1) - fprintf(outf, "max_die_id %d, sizing for %d die\n", - max_die_id, topo.num_die); + fprintf(outf, "max_die_id %d, sizing for %d die\n", max_die_id, topo.num_die); if (!summary_only && topo.num_die > 1) BIC_PRESENT(BIC_Die); topo.num_packages = max_package_id + 1; if (debug > 1) - fprintf(outf, "max_package_id %d, sizing for %d packages\n", - max_package_id, topo.num_packages); + fprintf(outf, "max_package_id %d, sizing for %d packages\n", max_package_id, topo.num_packages); if (!summary_only && topo.num_packages > 1) BIC_PRESENT(BIC_Package); @@ -5154,21 +5987,15 @@ void topology_probe() fprintf(outf, "cpu %d pkg %d die %d node %d lnode %d core %d thread %d\n", i, cpus[i].physical_package_id, cpus[i].die_id, - cpus[i].physical_node_id, - cpus[i].logical_node_id, - cpus[i].physical_core_id, - cpus[i].thread_id); + cpus[i].physical_node_id, cpus[i].logical_node_id, cpus[i].physical_core_id, cpus[i].thread_id); } } -void -allocate_counters(struct thread_data **t, struct core_data **c, - struct pkg_data **p) +void allocate_counters(struct thread_data **t, struct core_data **c, struct pkg_data **p) { int i; - int num_cores = topo.cores_per_node * topo.nodes_per_pkg * - topo.num_packages; + int num_cores = topo.cores_per_node * topo.nodes_per_pkg * topo.num_packages; int num_threads = topo.threads_per_core * num_cores; *t = calloc(num_threads, sizeof(struct thread_data)); @@ -5196,13 +6023,13 @@ allocate_counters(struct thread_data **t, struct core_data **c, error: err(1, "calloc counters"); } + /* * init_counter() * * set FIRST_THREAD_IN_CORE and FIRST_CORE_IN_PACKAGE */ -void init_counter(struct thread_data *thread_base, struct core_data *core_base, - struct pkg_data *pkg_base, int cpu_id) +void init_counter(struct thread_data *thread_base, struct core_data *core_base, struct pkg_data *pkg_base, int cpu_id) { int pkg_id = cpus[cpu_id].physical_package_id; int node_id = cpus[cpu_id].logical_node_id; @@ -5212,7 +6039,6 @@ void init_counter(struct thread_data *thread_base, struct core_data *core_base, struct core_data *c; struct pkg_data *p; - /* Workaround for systems where physical_node_id==-1 * and logical_node_id==(-1 - topo.num_cpus) */ @@ -5234,7 +6060,6 @@ void init_counter(struct thread_data *thread_base, struct core_data *core_base, p->package_id = pkg_id; } - int initialize_counters(int cpu_id) { init_counter(EVEN_COUNTERS, cpu_id); @@ -5249,12 +6074,14 @@ void allocate_output_buffer() if (outp == NULL) err(-1, "calloc output buffer"); } + void allocate_fd_percpu(void) { fd_percpu = calloc(topo.max_cpu_num + 1, sizeof(int)); if (fd_percpu == NULL) err(-1, "calloc fd_percpu"); } + void allocate_irq_buffers(void) { irq_column_2_cpu = calloc(topo.num_cpus, sizeof(int)); @@ -5265,6 +6092,7 @@ void allocate_irq_buffers(void) if (irqs_per_cpu == NULL) err(-1, "calloc %d", topo.max_cpu_num + 1); } + void setup_all_buffers(void) { topology_probe(); @@ -5293,7 +6121,7 @@ void turbostat_init() check_dev_msr(); check_permissions(); process_cpuid(); - + linux_perf_init(); if (!quiet) for_all_cpus(print_hwp, ODD_COUNTERS); @@ -5309,11 +6137,17 @@ void turbostat_init() for_all_cpus(set_temperature_target, ODD_COUNTERS); + for_all_cpus(get_cpu_type, ODD_COUNTERS); + for_all_cpus(get_cpu_type, EVEN_COUNTERS); + if (!quiet) for_all_cpus(print_thermal, ODD_COUNTERS); if (!quiet && do_irtl_snb) print_irtl(); + + if (DO_BIC(BIC_IPC)) + (void)get_instr_count_fd(base_cpu); } int fork_it(char **argv) @@ -5364,7 +6198,7 @@ int fork_it(char **argv) format_all_counters(EVEN_COUNTERS); } - fprintf(outf, "%.6f sec\n", tv_delta.tv_sec + tv_delta.tv_usec/1000000.0); + fprintf(outf, "%.6f sec\n", tv_delta.tv_sec + tv_delta.tv_usec / 1000000.0); flush_output_stderr(); @@ -5389,14 +6223,37 @@ int get_and_dump_counters(void) return status; } -void print_version() { - fprintf(outf, "turbostat version 20.03.20" - " - Len Brown <lenb@kernel.org>\n"); +void print_version() +{ + fprintf(outf, "turbostat version 2022.10.04 - Len Brown <lenb@kernel.org>\n"); +} + +#define COMMAND_LINE_SIZE 2048 + +void print_bootcmd(void) +{ + char bootcmd[COMMAND_LINE_SIZE]; + FILE *fp; + int ret; + + memset(bootcmd, 0, COMMAND_LINE_SIZE); + fp = fopen("/proc/cmdline", "r"); + if (!fp) + return; + + ret = fread(bootcmd, sizeof(char), COMMAND_LINE_SIZE - 1, fp); + if (ret) { + bootcmd[ret] = '\0'; + /* the last character is already '\n' */ + fprintf(outf, "Kernel command line: %s", bootcmd); + } + + fclose(fp); } int add_counter(unsigned int msr_num, char *path, char *name, - unsigned int width, enum counter_scope scope, - enum counter_type type, enum counter_format format, int flags) + unsigned int width, enum counter_scope scope, + enum counter_type type, enum counter_format format, int flags) { struct msr_counter *msrp; @@ -5422,8 +6279,7 @@ int add_counter(unsigned int msr_num, char *path, char *name, sys.tp = msrp; sys.added_thread_counters++; if (sys.added_thread_counters > MAX_ADDED_THREAD_COUNTERS) { - fprintf(stderr, "exceeded max %d added thread counters\n", - MAX_ADDED_COUNTERS); + fprintf(stderr, "exceeded max %d added thread counters\n", MAX_ADDED_COUNTERS); exit(-1); } break; @@ -5433,8 +6289,7 @@ int add_counter(unsigned int msr_num, char *path, char *name, sys.cp = msrp; sys.added_core_counters++; if (sys.added_core_counters > MAX_ADDED_COUNTERS) { - fprintf(stderr, "exceeded max %d added core counters\n", - MAX_ADDED_COUNTERS); + fprintf(stderr, "exceeded max %d added core counters\n", MAX_ADDED_COUNTERS); exit(-1); } break; @@ -5444,8 +6299,7 @@ int add_counter(unsigned int msr_num, char *path, char *name, sys.pp = msrp; sys.added_package_counters++; if (sys.added_package_counters > MAX_ADDED_COUNTERS) { - fprintf(stderr, "exceeded max %d added package counters\n", - MAX_ADDED_COUNTERS); + fprintf(stderr, "exceeded max %d added package counters\n", MAX_ADDED_COUNTERS); exit(-1); } break; @@ -5559,6 +6413,16 @@ next: } } +int is_deferred_add(char *name) +{ + int i; + + for (i = 0; i < deferred_add_index; ++i) + if (!strcmp(name, deferred_add_names[i])) + return 1; + return 0; +} + int is_deferred_skip(char *name) { int i; @@ -5577,65 +6441,67 @@ void probe_sysfs(void) int state; char *sp; - if (!DO_BIC(BIC_sysfs)) - return; - for (state = 10; state >= 0; --state) { - sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", - base_cpu, state); + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", base_cpu, state); input = fopen(path, "r"); if (input == NULL) continue; if (!fgets(name_buf, sizeof(name_buf), input)) err(1, "%s: failed to read file", path); - /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ + /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ sp = strchr(name_buf, '-'); if (!sp) sp = strchrnul(name_buf, '\n'); *sp = '%'; *(sp + 1) = '\0'; + remove_underbar(name_buf); + fclose(input); sprintf(path, "cpuidle/state%d/time", state); + if (!DO_BIC(BIC_sysfs) && !is_deferred_add(name_buf)) + continue; + if (is_deferred_skip(name_buf)) continue; - add_counter(0, path, name_buf, 64, SCOPE_CPU, COUNTER_USEC, - FORMAT_PERCENT, SYSFS_PERCPU); + add_counter(0, path, name_buf, 64, SCOPE_CPU, COUNTER_USEC, FORMAT_PERCENT, SYSFS_PERCPU); } for (state = 10; state >= 0; --state) { - sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", - base_cpu, state); + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", base_cpu, state); input = fopen(path, "r"); if (input == NULL) continue; if (!fgets(name_buf, sizeof(name_buf), input)) err(1, "%s: failed to read file", path); - /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ + /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ sp = strchr(name_buf, '-'); if (!sp) sp = strchrnul(name_buf, '\n'); *sp = '\0'; fclose(input); + remove_underbar(name_buf); + sprintf(path, "cpuidle/state%d/usage", state); + if (!DO_BIC(BIC_sysfs) && !is_deferred_add(name_buf)) + continue; + if (is_deferred_skip(name_buf)) continue; - add_counter(0, path, name_buf, 64, SCOPE_CPU, COUNTER_ITEMS, - FORMAT_DELTA, SYSFS_PERCPU); + add_counter(0, path, name_buf, 64, SCOPE_CPU, COUNTER_ITEMS, FORMAT_DELTA, SYSFS_PERCPU); } } - /* * parse cpuset with following syntax * 1,2,4..6,8-10 and set bits in cpu_subset @@ -5722,36 +6588,36 @@ error: exit(-1); } - void cmdline(int argc, char **argv) { int opt; int option_index = 0; static struct option long_options[] = { - {"add", required_argument, 0, 'a'}, - {"cpu", required_argument, 0, 'c'}, - {"Dump", no_argument, 0, 'D'}, - {"debug", no_argument, 0, 'd'}, /* internal, not documented */ - {"enable", required_argument, 0, 'e'}, - {"interval", required_argument, 0, 'i'}, - {"num_iterations", required_argument, 0, 'n'}, - {"help", no_argument, 0, 'h'}, - {"hide", required_argument, 0, 'H'}, // meh, -h taken by --help - {"Joules", no_argument, 0, 'J'}, - {"list", no_argument, 0, 'l'}, - {"out", required_argument, 0, 'o'}, - {"quiet", no_argument, 0, 'q'}, - {"show", required_argument, 0, 's'}, - {"Summary", no_argument, 0, 'S'}, - {"TCC", required_argument, 0, 'T'}, - {"version", no_argument, 0, 'v' }, - {0, 0, 0, 0 } + { "add", required_argument, 0, 'a' }, + { "cpu", required_argument, 0, 'c' }, + { "Dump", no_argument, 0, 'D' }, + { "debug", no_argument, 0, 'd' }, /* internal, not documented */ + { "enable", required_argument, 0, 'e' }, + { "interval", required_argument, 0, 'i' }, + { "IPC", no_argument, 0, 'I' }, + { "num_iterations", required_argument, 0, 'n' }, + { "header_iterations", required_argument, 0, 'N' }, + { "help", no_argument, 0, 'h' }, + { "hide", required_argument, 0, 'H' }, // meh, -h taken by --help + { "Joules", no_argument, 0, 'J' }, + { "list", no_argument, 0, 'l' }, + { "out", required_argument, 0, 'o' }, + { "quiet", no_argument, 0, 'q' }, + { "show", required_argument, 0, 's' }, + { "Summary", no_argument, 0, 'S' }, + { "TCC", required_argument, 0, 'T' }, + { "version", no_argument, 0, 'v' }, + { 0, 0, 0, 0 } }; progname = argv[0]; - while ((opt = getopt_long_only(argc, argv, "+C:c:Dde:hi:Jn:o:qST:v", - long_options, &option_index)) != -1) { + while ((opt = getopt_long_only(argc, argv, "+C:c:Dde:hi:Jn:o:qST:v", long_options, &option_index)) != -1) { switch (opt) { case 'a': parse_add_command(optarg); @@ -5786,8 +6652,7 @@ void cmdline(int argc, char **argv) double interval = strtod(optarg, NULL); if (interval < 0.001) { - fprintf(outf, "interval %f seconds is too small\n", - interval); + fprintf(outf, "interval %f seconds is too small\n", interval); exit(2); } @@ -5814,8 +6679,15 @@ void cmdline(int argc, char **argv) num_iterations = strtod(optarg, NULL); if (num_iterations <= 0) { - fprintf(outf, "iterations %d should be positive number\n", - num_iterations); + fprintf(outf, "iterations %d should be positive number\n", num_iterations); + exit(2); + } + break; + case 'N': + header_iterations = strtod(optarg, NULL); + + if (header_iterations <= 0) { + fprintf(outf, "iterations %d should be positive number\n", header_iterations); exit(2); } break; @@ -5835,7 +6707,7 @@ void cmdline(int argc, char **argv) summary_only++; break; case 'T': - tcc_activation_temp_override = atoi(optarg); + tj_max_override = atoi(optarg); break; case 'v': print_version(); @@ -5850,13 +6722,17 @@ int main(int argc, char **argv) outf = stderr; cmdline(argc, argv); - if (!quiet) + if (!quiet) { print_version(); + print_bootcmd(); + } probe_sysfs(); turbostat_init(); + msr_sum_record(); + /* dump counters and exit */ if (dump_only) return get_and_dump_counters(); diff --git a/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c b/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c index 3fe1eed900d4..5fd9e594079c 100644 --- a/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c +++ b/tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c @@ -91,6 +91,9 @@ unsigned int has_hwp_request_pkg; /* IA32_HWP_REQUEST_PKG */ unsigned int bdx_highest_ratio; +#define PATH_TO_CPU "/sys/devices/system/cpu/" +#define SYSFS_PATH_MAX 255 + /* * maintain compatibility with original implementation, but don't document it: */ @@ -622,6 +625,57 @@ void cmdline(int argc, char **argv) } } +/* + * Open a file, and exit on failure + */ +FILE *fopen_or_die(const char *path, const char *mode) +{ + FILE *filep = fopen(path, "r"); + + if (!filep) + err(1, "%s: open failed", path); + return filep; +} + +void err_on_hypervisor(void) +{ + FILE *cpuinfo; + char *flags, *hypervisor; + char *buffer; + + /* On VMs /proc/cpuinfo contains a "flags" entry for hypervisor */ + cpuinfo = fopen_or_die("/proc/cpuinfo", "ro"); + + buffer = malloc(4096); + if (!buffer) { + fclose(cpuinfo); + err(-ENOMEM, "buffer malloc fail"); + } + + if (!fread(buffer, 1024, 1, cpuinfo)) { + fclose(cpuinfo); + free(buffer); + err(1, "Reading /proc/cpuinfo failed"); + } + + flags = strstr(buffer, "flags"); + rewind(cpuinfo); + fseek(cpuinfo, flags - buffer, SEEK_SET); + if (!fgets(buffer, 4096, cpuinfo)) { + fclose(cpuinfo); + free(buffer); + err(1, "Reading /proc/cpuinfo failed"); + } + fclose(cpuinfo); + + hypervisor = strstr(buffer, "hypervisor"); + + free(buffer); + + if (hypervisor) + err(-1, + "not supported on this virtual machine"); +} int get_msr(int cpu, int offset, unsigned long long *msr) { @@ -635,8 +689,10 @@ int get_msr(int cpu, int offset, unsigned long long *msr) err(-1, "%s open failed, try chown or chmod +r /dev/cpu/*/msr, or run as root", pathname); retval = pread(fd, msr, sizeof(*msr), offset); - if (retval != sizeof(*msr)) + if (retval != sizeof(*msr)) { + err_on_hypervisor(); err(-1, "%s offset 0x%llx read failed", pathname, (unsigned long long)offset); + } if (debug > 1) fprintf(stderr, "get_msr(cpu%d, 0x%X, 0x%llX)\n", cpu, offset, *msr); @@ -668,6 +724,48 @@ int put_msr(int cpu, int offset, unsigned long long new_msr) return 0; } +static unsigned int read_sysfs(const char *path, char *buf, size_t buflen) +{ + ssize_t numread; + int fd; + + fd = open(path, O_RDONLY); + if (fd == -1) + return 0; + + numread = read(fd, buf, buflen - 1); + if (numread < 1) { + close(fd); + return 0; + } + + buf[numread] = '\0'; + close(fd); + + return (unsigned int) numread; +} + +static unsigned int write_sysfs(const char *path, char *buf, size_t buflen) +{ + ssize_t numwritten; + int fd; + + fd = open(path, O_WRONLY); + if (fd == -1) + return 0; + + numwritten = write(fd, buf, buflen - 1); + if (numwritten < 1) { + perror("write failed\n"); + close(fd); + return -1; + } + + close(fd); + + return (unsigned int) numwritten; +} + void print_hwp_cap(int cpu, struct msr_hwp_cap *cap, char *str) { if (cpu != -1) @@ -745,17 +843,61 @@ void write_hwp_request(int cpu, struct msr_hwp_request *hwp_req, unsigned int ms put_msr(cpu, msr_offset, msr); } +static int get_epb(int cpu) +{ + char path[SYSFS_PATH_MAX]; + char linebuf[3]; + char *endp; + long val; + + if (!has_epb) + return -1; + + snprintf(path, sizeof(path), PATH_TO_CPU "cpu%u/power/energy_perf_bias", cpu); + + if (!read_sysfs(path, linebuf, 3)) + return -1; + + val = strtol(linebuf, &endp, 0); + if (endp == linebuf || errno == ERANGE) + return -1; + + return (int)val; +} + +static int set_epb(int cpu, int val) +{ + char path[SYSFS_PATH_MAX]; + char linebuf[3]; + char *endp; + int ret; + + if (!has_epb) + return -1; + + snprintf(path, sizeof(path), PATH_TO_CPU "cpu%u/power/energy_perf_bias", cpu); + snprintf(linebuf, sizeof(linebuf), "%d", val); + + ret = write_sysfs(path, linebuf, 3); + if (ret <= 0) + return -1; + + val = strtol(linebuf, &endp, 0); + if (endp == linebuf || errno == ERANGE) + return -1; + + return (int)val; +} + int print_cpu_msrs(int cpu) { - unsigned long long msr; struct msr_hwp_request req; struct msr_hwp_cap cap; + int epb; - if (has_epb) { - get_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, &msr); - - printf("cpu%d: EPB %u\n", cpu, (unsigned int) msr); - } + epb = get_epb(cpu); + if (epb >= 0) + printf("cpu%d: EPB %u\n", cpu, (unsigned int) epb); if (!has_hwp) return 0; @@ -1038,15 +1180,15 @@ int enable_hwp_on_cpu(int cpu) int update_cpu_msrs(int cpu) { unsigned long long msr; - + int epb; if (update_epb) { - get_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, &msr); - put_msr(cpu, MSR_IA32_ENERGY_PERF_BIAS, new_epb); + epb = get_epb(cpu); + set_epb(cpu, new_epb); if (verbose) printf("cpu%d: ENERGY_PERF_BIAS old: %d new: %d\n", - cpu, (unsigned int) msr, (unsigned int) new_epb); + cpu, epb, (unsigned int) new_epb); } if (update_turbo) { @@ -1086,18 +1228,6 @@ int update_cpu_msrs(int cpu) return 0; } -/* - * Open a file, and exit on failure - */ -FILE *fopen_or_die(const char *path, const char *mode) -{ - FILE *filep = fopen(path, "r"); - - if (!filep) - err(1, "%s: open failed", path); - return filep; -} - unsigned int get_pkg_num(int cpu) { FILE *fp; |