diff options
Diffstat (limited to 'tools/testing/selftests/pidfd')
| -rw-r--r-- | tools/testing/selftests/pidfd/.gitignore | 4 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/Makefile | 7 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/config | 3 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd.h | 199 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_bind_mount.c | 116 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_exec_helper.c | 12 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_fdinfo_test.c | 28 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_file_handle_test.c | 503 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_getfd_test.c | 37 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_info_test.c | 693 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_open_test.c | 61 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_poll_test.c | 3 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_setns_test.c | 282 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_test.c | 162 | ||||
| -rw-r--r-- | tools/testing/selftests/pidfd/pidfd_wait.c | 345 |
15 files changed, 2151 insertions, 304 deletions
diff --git a/tools/testing/selftests/pidfd/.gitignore b/tools/testing/selftests/pidfd/.gitignore index 973198a3ec3d..0406a065deb4 100644 --- a/tools/testing/selftests/pidfd/.gitignore +++ b/tools/testing/selftests/pidfd/.gitignore @@ -6,3 +6,7 @@ pidfd_wait pidfd_fdinfo_test pidfd_getfd_test pidfd_setns_test +pidfd_file_handle_test +pidfd_bind_mount +pidfd_info_test +pidfd_exec_helper diff --git a/tools/testing/selftests/pidfd/Makefile b/tools/testing/selftests/pidfd/Makefile index f4a2f28f926b..fcbefc0d77f6 100644 --- a/tools/testing/selftests/pidfd/Makefile +++ b/tools/testing/selftests/pidfd/Makefile @@ -1,8 +1,11 @@ # SPDX-License-Identifier: GPL-2.0-only -CFLAGS += -g -I../../../../usr/include/ -pthread +CFLAGS += -g $(KHDR_INCLUDES) -pthread -Wall TEST_GEN_PROGS := pidfd_test pidfd_fdinfo_test pidfd_open_test \ - pidfd_poll_test pidfd_wait pidfd_getfd_test pidfd_setns_test + pidfd_poll_test pidfd_wait pidfd_getfd_test pidfd_setns_test \ + pidfd_file_handle_test pidfd_bind_mount pidfd_info_test + +TEST_GEN_PROGS_EXTENDED := pidfd_exec_helper include ../lib.mk diff --git a/tools/testing/selftests/pidfd/config b/tools/testing/selftests/pidfd/config index bb11de90c0c9..6133524710f7 100644 --- a/tools/testing/selftests/pidfd/config +++ b/tools/testing/selftests/pidfd/config @@ -3,4 +3,7 @@ CONFIG_IPC_NS=y CONFIG_USER_NS=y CONFIG_PID_NS=y CONFIG_NET_NS=y +CONFIG_TIME_NS=y +CONFIG_GENERIC_VDSO_TIME_NS=y CONFIG_CGROUPS=y +CONFIG_CHECKPOINT_RESTORE=y diff --git a/tools/testing/selftests/pidfd/pidfd.h b/tools/testing/selftests/pidfd/pidfd.h index 8d728eda783d..efd74063126e 100644 --- a/tools/testing/selftests/pidfd/pidfd.h +++ b/tools/testing/selftests/pidfd/pidfd.h @@ -12,36 +12,168 @@ #include <stdlib.h> #include <string.h> #include <syscall.h> -#include <sys/mount.h> +#include <sys/ioctl.h> #include <sys/types.h> #include <sys/wait.h> #include "../kselftest.h" +#include "../clone3/clone3_selftests.h" #ifndef P_PIDFD #define P_PIDFD 3 #endif +#ifndef CLONE_NEWTIME +#define CLONE_NEWTIME 0x00000080 +#endif + #ifndef CLONE_PIDFD #define CLONE_PIDFD 0x00001000 #endif #ifndef __NR_pidfd_open -#define __NR_pidfd_open -1 +#define __NR_pidfd_open 434 #endif #ifndef __NR_pidfd_send_signal -#define __NR_pidfd_send_signal -1 +#define __NR_pidfd_send_signal 424 #endif #ifndef __NR_clone3 -#define __NR_clone3 -1 +#define __NR_clone3 435 #endif #ifndef __NR_pidfd_getfd -#define __NR_pidfd_getfd -1 +#define __NR_pidfd_getfd 438 +#endif + +#ifndef PIDFD_NONBLOCK +#define PIDFD_NONBLOCK O_NONBLOCK +#endif + +#ifndef PIDFD_SELF_THREAD +#define PIDFD_SELF_THREAD -10000 /* Current thread. */ +#endif + +#ifndef PIDFD_SELF_THREAD_GROUP +#define PIDFD_SELF_THREAD_GROUP -20000 /* Current thread group leader. */ +#endif + +#ifndef PIDFD_SELF +#define PIDFD_SELF PIDFD_SELF_THREAD +#endif + +#ifndef PIDFD_SELF_PROCESS +#define PIDFD_SELF_PROCESS PIDFD_SELF_THREAD_GROUP +#endif + +#ifndef PIDFS_IOCTL_MAGIC +#define PIDFS_IOCTL_MAGIC 0xFF +#endif + +#ifndef PIDFD_GET_CGROUP_NAMESPACE +#define PIDFD_GET_CGROUP_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 1) +#endif + +#ifndef PIDFD_GET_IPC_NAMESPACE +#define PIDFD_GET_IPC_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 2) +#endif + +#ifndef PIDFD_GET_MNT_NAMESPACE +#define PIDFD_GET_MNT_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 3) +#endif + +#ifndef PIDFD_GET_NET_NAMESPACE +#define PIDFD_GET_NET_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 4) +#endif + +#ifndef PIDFD_GET_PID_NAMESPACE +#define PIDFD_GET_PID_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 5) +#endif + +#ifndef PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE +#define PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 6) +#endif + +#ifndef PIDFD_GET_TIME_NAMESPACE +#define PIDFD_GET_TIME_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 7) +#endif + +#ifndef PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE +#define PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 8) +#endif + +#ifndef PIDFD_GET_USER_NAMESPACE +#define PIDFD_GET_USER_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 9) +#endif + +#ifndef PIDFD_GET_UTS_NAMESPACE +#define PIDFD_GET_UTS_NAMESPACE _IO(PIDFS_IOCTL_MAGIC, 10) +#endif + +#ifndef PIDFD_GET_INFO +#define PIDFD_GET_INFO _IOWR(PIDFS_IOCTL_MAGIC, 11, struct pidfd_info) +#endif + +#ifndef PIDFD_INFO_PID +#define PIDFD_INFO_PID (1UL << 0) /* Always returned, even if not requested */ +#endif + +#ifndef PIDFD_INFO_CREDS +#define PIDFD_INFO_CREDS (1UL << 1) /* Always returned, even if not requested */ +#endif + +#ifndef PIDFD_INFO_CGROUPID +#define PIDFD_INFO_CGROUPID (1UL << 2) /* Always returned if available, even if not requested */ +#endif + +#ifndef PIDFD_INFO_EXIT +#define PIDFD_INFO_EXIT (1UL << 3) /* Always returned if available, even if not requested */ +#endif + +#ifndef PIDFD_INFO_COREDUMP +#define PIDFD_INFO_COREDUMP (1UL << 4) +#endif + +#ifndef PIDFD_COREDUMPED +#define PIDFD_COREDUMPED (1U << 0) /* Did crash and... */ +#endif + +#ifndef PIDFD_COREDUMP_SKIP +#define PIDFD_COREDUMP_SKIP (1U << 1) /* coredumping generation was skipped. */ +#endif + +#ifndef PIDFD_COREDUMP_USER +#define PIDFD_COREDUMP_USER (1U << 2) /* coredump was done as the user. */ +#endif + +#ifndef PIDFD_COREDUMP_ROOT +#define PIDFD_COREDUMP_ROOT (1U << 3) /* coredump was done as root. */ #endif +#ifndef PIDFD_THREAD +#define PIDFD_THREAD O_EXCL +#endif + +struct pidfd_info { + __u64 mask; + __u64 cgroupid; + __u32 pid; + __u32 tgid; + __u32 ppid; + __u32 ruid; + __u32 rgid; + __u32 euid; + __u32 egid; + __u32 suid; + __u32 sgid; + __u32 fsuid; + __u32 fsgid; + __s32 exit_code; + __u32 coredump_mask; + __u32 __spare1; +}; + /* * The kernel reserves 300 pids via RESERVED_PIDS in kernel/pid.c * That means, when it wraps around any pid < 300 will be skipped. @@ -60,7 +192,12 @@ #define PIDFD_SKIP 3 #define PIDFD_XFAIL 4 -int wait_for_pid(pid_t pid) +static inline int sys_waitid(int which, pid_t pid, siginfo_t *info, int options) +{ + return syscall(__NR_waitid, which, pid, info, options, NULL); +} + +static inline int wait_for_pid(pid_t pid) { int status, ret; @@ -70,13 +207,19 @@ again: if (errno == EINTR) goto again; + ksft_print_msg("waitpid returned -1, errno=%d\n", errno); return -1; } - if (!WIFEXITED(status)) + if (!WIFEXITED(status)) { + ksft_print_msg( + "waitpid !WIFEXITED, WIFSIGNALED=%d, WTERMSIG=%d\n", + WIFSIGNALED(status), WTERMSIG(status)); return -1; + } - return WEXITSTATUS(status); + ret = WEXITSTATUS(status); + return ret; } static inline int sys_pidfd_open(pid_t pid, unsigned int flags) @@ -100,4 +243,44 @@ static inline int sys_memfd_create(const char *name, unsigned int flags) return syscall(__NR_memfd_create, name, flags); } +static inline pid_t create_child(int *pidfd, unsigned flags) +{ + struct __clone_args args = { + .flags = CLONE_PIDFD | flags, + .exit_signal = SIGCHLD, + .pidfd = ptr_to_u64(pidfd), + }; + + return sys_clone3(&args, sizeof(struct __clone_args)); +} + +static inline ssize_t read_nointr(int fd, void *buf, size_t count) +{ + ssize_t ret; + + do { + ret = read(fd, buf, count); + } while (ret < 0 && errno == EINTR); + + return ret; +} + +static inline ssize_t write_nointr(int fd, const void *buf, size_t count) +{ + ssize_t ret; + + do { + ret = write(fd, buf, count); + } while (ret < 0 && errno == EINTR); + + return ret; +} + +static inline int sys_execveat(int dirfd, const char *pathname, + char *const argv[], char *const envp[], + int flags) +{ + return syscall(__NR_execveat, dirfd, pathname, argv, envp, flags); +} + #endif /* __PIDFD_H */ diff --git a/tools/testing/selftests/pidfd/pidfd_bind_mount.c b/tools/testing/selftests/pidfd/pidfd_bind_mount.c new file mode 100644 index 000000000000..c094aeb1c620 --- /dev/null +++ b/tools/testing/selftests/pidfd/pidfd_bind_mount.c @@ -0,0 +1,116 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +// Copyright (c) 2024 Christian Brauner <brauner@kernel.org> + +#define _GNU_SOURCE +#include <fcntl.h> +#include <limits.h> +#include <sched.h> +#include <stdio.h> +#include <string.h> +#include <linux/fs.h> +#include <sys/ioctl.h> +#include <sys/stat.h> +#include <sys/mount.h> +#include <unistd.h> + +#include "pidfd.h" +#include "../kselftest_harness.h" +#include "../filesystems/wrappers.h" + +FIXTURE(pidfd_bind_mount) { + char template[PATH_MAX]; + int fd_tmp; + int pidfd; + struct stat st1; + struct stat st2; + __u32 gen1; + __u32 gen2; + bool must_unmount; +}; + +FIXTURE_SETUP(pidfd_bind_mount) +{ + self->fd_tmp = -EBADF; + self->must_unmount = false; + ASSERT_EQ(unshare(CLONE_NEWNS), 0); + ASSERT_LE(snprintf(self->template, PATH_MAX, "%s", P_tmpdir "/pidfd_bind_mount_XXXXXX"), PATH_MAX); + self->fd_tmp = mkstemp(self->template); + ASSERT_GE(self->fd_tmp, 0); + self->pidfd = sys_pidfd_open(getpid(), 0); + ASSERT_GE(self->pidfd, 0); + ASSERT_GE(fstat(self->pidfd, &self->st1), 0); + ASSERT_EQ(ioctl(self->pidfd, FS_IOC_GETVERSION, &self->gen1), 0); +} + +FIXTURE_TEARDOWN(pidfd_bind_mount) +{ + ASSERT_EQ(close(self->fd_tmp), 0); + if (self->must_unmount) + ASSERT_EQ(umount2(self->template, 0), 0); + ASSERT_EQ(unlink(self->template), 0); +} + +/* + * Test that a detached mount can be created for a pidfd and then + * attached to the filesystem hierarchy. + */ +TEST_F(pidfd_bind_mount, bind_mount) +{ + int fd_tree; + + fd_tree = sys_open_tree(self->pidfd, "", OPEN_TREE_CLONE | OPEN_TREE_CLOEXEC | AT_EMPTY_PATH); + ASSERT_GE(fd_tree, 0); + + ASSERT_EQ(move_mount(fd_tree, "", self->fd_tmp, "", MOVE_MOUNT_F_EMPTY_PATH | MOVE_MOUNT_T_EMPTY_PATH), 0); + self->must_unmount = true; + + ASSERT_EQ(close(fd_tree), 0); +} + +/* Test that a pidfd can be reopened through procfs. */ +TEST_F(pidfd_bind_mount, reopen) +{ + int pidfd; + char proc_path[PATH_MAX]; + + sprintf(proc_path, "/proc/self/fd/%d", self->pidfd); + pidfd = open(proc_path, O_RDONLY | O_NOCTTY | O_CLOEXEC); + ASSERT_GE(pidfd, 0); + + ASSERT_GE(fstat(self->pidfd, &self->st2), 0); + ASSERT_EQ(ioctl(self->pidfd, FS_IOC_GETVERSION, &self->gen2), 0); + + ASSERT_TRUE(self->st1.st_dev == self->st2.st_dev && self->st1.st_ino == self->st2.st_ino); + ASSERT_TRUE(self->gen1 == self->gen2); + + ASSERT_EQ(close(pidfd), 0); +} + +/* + * Test that a detached mount can be created for a pidfd and then + * attached to the filesystem hierarchy and reopened. + */ +TEST_F(pidfd_bind_mount, bind_mount_reopen) +{ + int fd_tree, fd_pidfd_mnt; + + fd_tree = sys_open_tree(self->pidfd, "", OPEN_TREE_CLONE | OPEN_TREE_CLOEXEC | AT_EMPTY_PATH); + ASSERT_GE(fd_tree, 0); + + ASSERT_EQ(move_mount(fd_tree, "", self->fd_tmp, "", MOVE_MOUNT_F_EMPTY_PATH | MOVE_MOUNT_T_EMPTY_PATH), 0); + self->must_unmount = true; + + fd_pidfd_mnt = openat(-EBADF, self->template, O_RDONLY | O_NOCTTY | O_CLOEXEC); + ASSERT_GE(fd_pidfd_mnt, 0); + + ASSERT_GE(fstat(fd_tree, &self->st2), 0); + ASSERT_EQ(ioctl(fd_pidfd_mnt, FS_IOC_GETVERSION, &self->gen2), 0); + + ASSERT_TRUE(self->st1.st_dev == self->st2.st_dev && self->st1.st_ino == self->st2.st_ino); + ASSERT_TRUE(self->gen1 == self->gen2); + + ASSERT_EQ(close(fd_tree), 0); + ASSERT_EQ(close(fd_pidfd_mnt), 0); +} + +TEST_HARNESS_MAIN diff --git a/tools/testing/selftests/pidfd/pidfd_exec_helper.c b/tools/testing/selftests/pidfd/pidfd_exec_helper.c new file mode 100644 index 000000000000..5516808c95f2 --- /dev/null +++ b/tools/testing/selftests/pidfd/pidfd_exec_helper.c @@ -0,0 +1,12 @@ +#define _GNU_SOURCE +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> + +int main(int argc, char *argv[]) +{ + if (pause()) + _exit(EXIT_FAILURE); + + _exit(EXIT_SUCCESS); +} diff --git a/tools/testing/selftests/pidfd/pidfd_fdinfo_test.c b/tools/testing/selftests/pidfd/pidfd_fdinfo_test.c index 22558524f71c..f718aac75068 100644 --- a/tools/testing/selftests/pidfd/pidfd_fdinfo_test.c +++ b/tools/testing/selftests/pidfd/pidfd_fdinfo_test.c @@ -12,6 +12,8 @@ #include <string.h> #include <syscall.h> #include <sys/wait.h> +#include <sys/mman.h> +#include <sys/mount.h> #include "pidfd.h" #include "../kselftest.h" @@ -61,7 +63,7 @@ static void error_report(struct error *err, const char *test_name) break; case PIDFD_PASS: - ksft_test_result_pass("%s test: Passed\n"); + ksft_test_result_pass("%s test: Passed\n", test_name); break; default: @@ -80,7 +82,10 @@ static inline int error_check(struct error *err, const char *test_name) return err->code; } +#define CHILD_STACK_SIZE 8192 + struct child { + char *stack; pid_t pid; int fd; }; @@ -89,17 +94,22 @@ static struct child clone_newns(int (*fn)(void *), void *args, struct error *err) { static int flags = CLONE_PIDFD | CLONE_NEWPID | CLONE_NEWNS | SIGCHLD; - size_t stack_size = 1024; - char *stack[1024] = { 0 }; struct child ret; if (!(flags & CLONE_NEWUSER) && geteuid() != 0) flags |= CLONE_NEWUSER; + ret.stack = mmap(NULL, CHILD_STACK_SIZE, PROT_READ | PROT_WRITE, + MAP_PRIVATE | MAP_ANONYMOUS | MAP_STACK, -1, 0); + if (ret.stack == MAP_FAILED) { + error_set(err, -1, "mmap of stack failed (errno %d)", errno); + return ret; + } + #ifdef __ia64__ - ret.pid = __clone2(fn, stack, stack_size, flags, args, &ret.fd); + ret.pid = __clone2(fn, ret.stack, CHILD_STACK_SIZE, flags, args, &ret.fd); #else - ret.pid = clone(fn, stack + stack_size, flags, args, &ret.fd); + ret.pid = clone(fn, ret.stack + CHILD_STACK_SIZE, flags, args, &ret.fd); #endif if (ret.pid < 0) { @@ -129,6 +139,12 @@ static inline int child_join(struct child *child, struct error *err) else if (r > 0) error_set(err, r, "child %d reported: %d", child->pid, r); + if (munmap(child->stack, CHILD_STACK_SIZE)) { + error_set(err, -1, "munmap of child stack failed (errno %d)", errno); + r = -1; + } + + ksft_print_msg("waitpid WEXITSTATUS=%d\n", r); return r; } @@ -292,5 +308,5 @@ int main(int argc, char **argv) test_pidfd_fdinfo_nspid(); test_pidfd_dead_fdinfo(); - return ksft_exit_pass(); + ksft_exit_pass(); } diff --git a/tools/testing/selftests/pidfd/pidfd_file_handle_test.c b/tools/testing/selftests/pidfd/pidfd_file_handle_test.c new file mode 100644 index 000000000000..439b9c6c0457 --- /dev/null +++ b/tools/testing/selftests/pidfd/pidfd_file_handle_test.c @@ -0,0 +1,503 @@ +// SPDX-License-Identifier: GPL-2.0 + +#define _GNU_SOURCE +#include <errno.h> +#include <fcntl.h> +#include <limits.h> +#include <linux/types.h> +#include <poll.h> +#include <sched.h> +#include <signal.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <syscall.h> +#include <sys/prctl.h> +#include <sys/wait.h> +#include <unistd.h> +#include <sys/socket.h> +#include <linux/kcmp.h> +#include <sys/stat.h> + +#include "pidfd.h" +#include "../kselftest_harness.h" + +FIXTURE(file_handle) +{ + pid_t pid; + int pidfd; + + pid_t child_pid1; + int child_pidfd1; + + pid_t child_pid2; + int child_pidfd2; + + pid_t child_pid3; + int child_pidfd3; +}; + +FIXTURE_SETUP(file_handle) +{ + int ret; + int ipc_sockets[2]; + char c; + + self->pid = getpid(); + self->pidfd = sys_pidfd_open(self->pid, 0); + ASSERT_GE(self->pidfd, 0); + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWUSER); + EXPECT_GE(self->child_pid1, 0); + + if (self->child_pid1 == 0) { + close(ipc_sockets[0]); + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + + pause(); + _exit(EXIT_SUCCESS); + } + + close(ipc_sockets[1]); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + close(ipc_sockets[0]); + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWUSER | CLONE_NEWPID); + EXPECT_GE(self->child_pid2, 0); + + if (self->child_pid2 == 0) { + close(ipc_sockets[0]); + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + + pause(); + _exit(EXIT_SUCCESS); + } + + close(ipc_sockets[1]); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + close(ipc_sockets[0]); + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + self->child_pid3 = create_child(&self->child_pidfd3, CLONE_NEWUSER | CLONE_NEWPID); + EXPECT_GE(self->child_pid3, 0); + + if (self->child_pid3 == 0) { + close(ipc_sockets[0]); + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + + pause(); + _exit(EXIT_SUCCESS); + } + + close(ipc_sockets[1]); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + close(ipc_sockets[0]); +} + +FIXTURE_TEARDOWN(file_handle) +{ + EXPECT_EQ(close(self->pidfd), 0); + + EXPECT_EQ(sys_pidfd_send_signal(self->child_pidfd1, SIGKILL, NULL, 0), 0); + if (self->child_pidfd1 >= 0) + EXPECT_EQ(0, close(self->child_pidfd1)); + + EXPECT_EQ(sys_waitid(P_PID, self->child_pid1, NULL, WEXITED), 0); + + EXPECT_EQ(sys_pidfd_send_signal(self->child_pidfd2, SIGKILL, NULL, 0), 0); + if (self->child_pidfd2 >= 0) + EXPECT_EQ(0, close(self->child_pidfd2)); + + EXPECT_EQ(sys_waitid(P_PID, self->child_pid2, NULL, WEXITED), 0); + + if (self->child_pidfd3 >= 0) { + EXPECT_EQ(sys_pidfd_send_signal(self->child_pidfd3, SIGKILL, NULL, 0), 0); + EXPECT_EQ(0, close(self->child_pidfd3)); + EXPECT_EQ(sys_waitid(P_PID, self->child_pid3, NULL, WEXITED), 0); + } +} + +/* + * Test that we can decode a pidfs file handle in the same pid + * namespace. + */ +TEST_F(file_handle, file_handle_same_pidns) +{ + int mnt_id; + struct file_handle *fh; + int pidfd = -EBADF; + struct stat st1, st2; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd1, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + ASSERT_EQ(fstat(self->child_pidfd1, &st1), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, O_CLOEXEC); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, O_NONBLOCK); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + free(fh); +} + +/* + * Test that we can decode a pidfs file handle from a child pid + * namespace. + */ +TEST_F(file_handle, file_handle_child_pidns) +{ + int mnt_id; + struct file_handle *fh; + int pidfd = -EBADF; + struct stat st1, st2; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd2, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + ASSERT_EQ(fstat(self->child_pidfd2, &st1), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, O_CLOEXEC); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, O_NONBLOCK); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + free(fh); +} + +/* + * Test that we fail to decode a pidfs file handle from an ancestor + * child pid namespace. + */ +TEST_F(file_handle, file_handle_foreign_pidns) +{ + int mnt_id; + struct file_handle *fh; + pid_t pid; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->pidfd, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + ASSERT_EQ(setns(self->child_pidfd2, CLONE_NEWUSER | CLONE_NEWPID), 0); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + int pidfd = open_by_handle_at(self->pidfd, fh, 0); + if (pidfd >= 0) { + TH_LOG("Managed to open pidfd outside of the caller's pid namespace hierarchy"); + _exit(1); + } + _exit(0); + } + + ASSERT_EQ(wait_for_pid(pid), 0); + + free(fh); +} + +/* + * Test that we can decode a pidfs file handle of a process that has + * exited but not been reaped. + */ +TEST_F(file_handle, pid_has_exited) +{ + int mnt_id, pidfd, child_pidfd3; + struct file_handle *fh; + struct stat st1, st2; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd3, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + ASSERT_EQ(fstat(self->child_pidfd3, &st1), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + child_pidfd3 = self->child_pidfd3; + self->child_pidfd3 = -EBADF; + EXPECT_EQ(sys_pidfd_send_signal(child_pidfd3, SIGKILL, NULL, 0), 0); + EXPECT_EQ(close(child_pidfd3), 0); + EXPECT_EQ(sys_waitid(P_PID, self->child_pid3, NULL, WEXITED | WNOWAIT), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_GE(pidfd, 0); + + EXPECT_EQ(sys_waitid(P_PID, self->child_pid3, NULL, WEXITED), 0); +} + +/* + * Test that we fail to decode a pidfs file handle of a process that has + * already been reaped. + */ +TEST_F(file_handle, pid_has_been_reaped) +{ + int mnt_id, pidfd, child_pidfd3; + struct file_handle *fh; + struct stat st1, st2; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd3, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + ASSERT_EQ(fstat(self->child_pidfd3, &st1), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); + + child_pidfd3 = self->child_pidfd3; + self->child_pidfd3 = -EBADF; + EXPECT_EQ(sys_pidfd_send_signal(child_pidfd3, SIGKILL, NULL, 0), 0); + EXPECT_EQ(close(child_pidfd3), 0); + EXPECT_EQ(sys_waitid(P_PID, self->child_pid3, NULL, WEXITED), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_LT(pidfd, 0); +} + +/* + * Test valid flags to open a pidfd file handle. Note, that + * PIDFD_NONBLOCK is defined as O_NONBLOCK and O_NONBLOCK is an alias to + * O_NDELAY. Also note that PIDFD_THREAD is an alias for O_EXCL. + */ +TEST_F(file_handle, open_by_handle_at_valid_flags) +{ + int mnt_id; + struct file_handle *fh; + int pidfd = -EBADF; + struct stat st1, st2; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd2, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + ASSERT_EQ(fstat(self->child_pidfd2, &st1), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, + O_RDONLY | + O_WRONLY | + O_RDWR | + O_NONBLOCK | + O_NDELAY | + O_CLOEXEC | + O_EXCL); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); +} + +/* + * Test that invalid flags passed to open a pidfd file handle are + * rejected. + */ +TEST_F(file_handle, open_by_handle_at_invalid_flags) +{ + int mnt_id; + struct file_handle *fh; + int pidfd = -EBADF; + static const struct invalid_pidfs_file_handle_flags { + int oflag; + const char *oflag_name; + } invalid_pidfs_file_handle_flags[] = { + { FASYNC, "FASYNC" }, + { O_CREAT, "O_CREAT" }, + { O_NOCTTY, "O_NOCTTY" }, + { O_CREAT, "O_CREAT" }, + { O_TRUNC, "O_TRUNC" }, + { O_APPEND, "O_APPEND" }, + { O_SYNC, "O_SYNC" }, + { O_DSYNC, "O_DSYNC" }, + { O_DIRECT, "O_DIRECT" }, + { O_DIRECTORY, "O_DIRECTORY" }, + { O_NOFOLLOW, "O_NOFOLLOW" }, + { O_NOATIME, "O_NOATIME" }, + { O_PATH, "O_PATH" }, + { O_TMPFILE, "O_TMPFILE" }, + /* + * O_LARGEFILE is added implicitly by + * open_by_handle_at() so pidfs simply masks it off. + */ + }; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd2, "", fh, &mnt_id, AT_EMPTY_PATH), 0); + + for (int i = 0; i < ARRAY_SIZE(invalid_pidfs_file_handle_flags); i++) { + pidfd = open_by_handle_at(self->pidfd, fh, invalid_pidfs_file_handle_flags[i].oflag); + ASSERT_LT(pidfd, 0) { + TH_LOG("open_by_handle_at() succeeded with invalid flags: %s", invalid_pidfs_file_handle_flags[i].oflag_name); + } + } +} + +/* Test that lookup fails. */ +TEST_F(file_handle, lookup_must_fail) +{ + int mnt_id; + struct file_handle *fh; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_NE(name_to_handle_at(self->child_pidfd2, "lookup-is-not-possible-with-pidfs", fh, &mnt_id, AT_EMPTY_PATH), 0); + ASSERT_EQ(errno, ENOTDIR); + ASSERT_NE(name_to_handle_at(self->child_pidfd2, "lookup-is-not-possible-with-pidfs", fh, &mnt_id, 0), 0); + ASSERT_EQ(errno, ENOTDIR); +} + +#ifndef AT_HANDLE_CONNECTABLE +#define AT_HANDLE_CONNECTABLE 0x002 +#endif + +/* + * Test that AT_HANDLE_CONNECTABLE is rejected. Connectable file handles + * don't make sense for pidfs. Note that currently AT_HANDLE_CONNECTABLE + * is rejected because it is incompatible with AT_EMPTY_PATH which is + * required with pidfds as we don't support lookup. + */ +TEST_F(file_handle, invalid_name_to_handle_at_flags) +{ + int mnt_id; + struct file_handle *fh; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_NE(name_to_handle_at(self->child_pidfd2, "", fh, &mnt_id, AT_EMPTY_PATH | AT_HANDLE_CONNECTABLE), 0); +} + +#ifndef AT_HANDLE_FID +#define AT_HANDLE_FID 0x200 +#endif + +/* + * Test that a request with AT_HANDLE_FID always leads to decodable file + * handle as pidfs always provides export operations. + */ +TEST_F(file_handle, valid_name_to_handle_at_flags) +{ + int mnt_id, pidfd; + struct file_handle *fh; + struct stat st1, st2; + + fh = malloc(sizeof(struct file_handle) + MAX_HANDLE_SZ); + ASSERT_NE(fh, NULL); + memset(fh, 0, sizeof(struct file_handle) + MAX_HANDLE_SZ); + fh->handle_bytes = MAX_HANDLE_SZ; + + ASSERT_EQ(name_to_handle_at(self->child_pidfd2, "", fh, &mnt_id, AT_EMPTY_PATH | AT_HANDLE_FID), 0); + + ASSERT_EQ(fstat(self->child_pidfd2, &st1), 0); + + pidfd = open_by_handle_at(self->pidfd, fh, 0); + ASSERT_GE(pidfd, 0); + + ASSERT_EQ(fstat(pidfd, &st2), 0); + ASSERT_TRUE(st1.st_dev == st2.st_dev && st1.st_ino == st2.st_ino); + + ASSERT_EQ(close(pidfd), 0); +} + +TEST_HARNESS_MAIN diff --git a/tools/testing/selftests/pidfd/pidfd_getfd_test.c b/tools/testing/selftests/pidfd/pidfd_getfd_test.c index 84b65ecccb04..cd51d547b751 100644 --- a/tools/testing/selftests/pidfd/pidfd_getfd_test.c +++ b/tools/testing/selftests/pidfd/pidfd_getfd_test.c @@ -5,6 +5,7 @@ #include <fcntl.h> #include <limits.h> #include <linux/types.h> +#include <poll.h> #include <sched.h> #include <signal.h> #include <stdio.h> @@ -18,7 +19,6 @@ #include <linux/kcmp.h> #include "pidfd.h" -#include "../kselftest.h" #include "../kselftest_harness.h" /* @@ -130,6 +130,7 @@ FIXTURE(child) * When it is closed, the child will exit. */ int sk; + bool ignore_child_result; }; FIXTURE_SETUP(child) @@ -166,10 +167,14 @@ FIXTURE_SETUP(child) FIXTURE_TEARDOWN(child) { + int ret; + EXPECT_EQ(0, close(self->pidfd)); EXPECT_EQ(0, close(self->sk)); - EXPECT_EQ(0, wait_for_pid(self->pid)); + ret = wait_for_pid(self->pid); + if (!self->ignore_child_result) + EXPECT_EQ(0, ret); } TEST_F(child, disable_ptrace) @@ -205,7 +210,10 @@ TEST_F(child, fetch_fd) fd = sys_pidfd_getfd(self->pidfd, self->remote_fd, 0); ASSERT_GE(fd, 0); - EXPECT_EQ(0, sys_kcmp(getpid(), self->pid, KCMP_FILE, fd, self->remote_fd)); + ret = sys_kcmp(getpid(), self->pid, KCMP_FILE, fd, self->remote_fd); + if (ret < 0 && errno == ENOSYS) + SKIP(return, "kcmp() syscall not supported"); + EXPECT_EQ(ret, 0); ret = fcntl(fd, F_GETFD); ASSERT_GE(ret, 0); @@ -233,6 +241,29 @@ TEST(flags_set) EXPECT_EQ(errno, EINVAL); } +TEST_F(child, no_strange_EBADF) +{ + struct pollfd fds; + + self->ignore_child_result = true; + + fds.fd = self->pidfd; + fds.events = POLLIN; + + ASSERT_EQ(kill(self->pid, SIGKILL), 0); + ASSERT_EQ(poll(&fds, 1, 5000), 1); + + /* + * It used to be that pidfd_getfd() could race with the exiting thread + * between exit_files() and release_task(), and get a non-null task + * with a NULL files struct, and you'd get EBADF, which was slightly + * confusing. + */ + errno = 0; + EXPECT_EQ(sys_pidfd_getfd(self->pidfd, self->remote_fd, 0), -1); + EXPECT_EQ(errno, ESRCH); +} + #if __NR_pidfd_getfd == -1 int main(void) { diff --git a/tools/testing/selftests/pidfd/pidfd_info_test.c b/tools/testing/selftests/pidfd/pidfd_info_test.c new file mode 100644 index 000000000000..a0eb6e81eaa2 --- /dev/null +++ b/tools/testing/selftests/pidfd/pidfd_info_test.c @@ -0,0 +1,693 @@ +// SPDX-License-Identifier: GPL-2.0 + +#define _GNU_SOURCE +#include <errno.h> +#include <fcntl.h> +#include <limits.h> +#include <linux/types.h> +#include <poll.h> +#include <pthread.h> +#include <sched.h> +#include <signal.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <syscall.h> +#include <sys/prctl.h> +#include <sys/wait.h> +#include <unistd.h> +#include <sys/socket.h> +#include <linux/kcmp.h> +#include <sys/stat.h> + +#include "pidfd.h" +#include "../kselftest_harness.h" + +FIXTURE(pidfd_info) +{ + pid_t child_pid1; + int child_pidfd1; + + pid_t child_pid2; + int child_pidfd2; + + pid_t child_pid3; + int child_pidfd3; + + pid_t child_pid4; + int child_pidfd4; +}; + +FIXTURE_SETUP(pidfd_info) +{ + int ret; + int ipc_sockets[2]; + char c; + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + self->child_pid1 = create_child(&self->child_pidfd1, 0); + EXPECT_GE(self->child_pid1, 0); + + if (self->child_pid1 == 0) { + close(ipc_sockets[0]); + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + + pause(); + _exit(EXIT_SUCCESS); + } + + EXPECT_EQ(close(ipc_sockets[1]), 0); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + EXPECT_EQ(close(ipc_sockets[0]), 0); + + /* SIGKILL but don't reap. */ + EXPECT_EQ(sys_pidfd_send_signal(self->child_pidfd1, SIGKILL, NULL, 0), 0); + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + self->child_pid2 = create_child(&self->child_pidfd2, 0); + EXPECT_GE(self->child_pid2, 0); + + if (self->child_pid2 == 0) { + close(ipc_sockets[0]); + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + + pause(); + _exit(EXIT_SUCCESS); + } + + EXPECT_EQ(close(ipc_sockets[1]), 0); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + EXPECT_EQ(close(ipc_sockets[0]), 0); + + /* SIGKILL and reap. */ + EXPECT_EQ(sys_pidfd_send_signal(self->child_pidfd2, SIGKILL, NULL, 0), 0); + EXPECT_EQ(sys_waitid(P_PID, self->child_pid2, NULL, WEXITED), 0); + + self->child_pid3 = create_child(&self->child_pidfd3, CLONE_NEWUSER | CLONE_NEWPID); + EXPECT_GE(self->child_pid3, 0); + + if (self->child_pid3 == 0) + _exit(EXIT_SUCCESS); + + self->child_pid4 = create_child(&self->child_pidfd4, CLONE_NEWUSER | CLONE_NEWPID); + EXPECT_GE(self->child_pid4, 0); + + if (self->child_pid4 == 0) + _exit(EXIT_SUCCESS); + + EXPECT_EQ(sys_waitid(P_PID, self->child_pid4, NULL, WEXITED), 0); +} + +FIXTURE_TEARDOWN(pidfd_info) +{ + sys_pidfd_send_signal(self->child_pidfd1, SIGKILL, NULL, 0); + if (self->child_pidfd1 >= 0) + EXPECT_EQ(0, close(self->child_pidfd1)); + + sys_waitid(P_PID, self->child_pid1, NULL, WEXITED); + + sys_pidfd_send_signal(self->child_pidfd2, SIGKILL, NULL, 0); + if (self->child_pidfd2 >= 0) + EXPECT_EQ(0, close(self->child_pidfd2)); + + sys_waitid(P_PID, self->child_pid2, NULL, WEXITED); + sys_waitid(P_PID, self->child_pid3, NULL, WEXITED); + sys_waitid(P_PID, self->child_pid4, NULL, WEXITED); +} + +TEST_F(pidfd_info, sigkill_exit) +{ + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID, + }; + + /* Process has exited but not been reaped so this must work. */ + ASSERT_EQ(ioctl(self->child_pidfd1, PIDFD_GET_INFO, &info), 0); + + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(self->child_pidfd1, PIDFD_GET_INFO, &info), 0); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_CREDS)); + /* Process has exited but not been reaped, so no PIDFD_INFO_EXIT information yet. */ + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_EXIT)); +} + +TEST_F(pidfd_info, sigkill_reaped) +{ + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID, + }; + + /* Process has already been reaped and PIDFD_INFO_EXIT hasn't been set. */ + ASSERT_NE(ioctl(self->child_pidfd2, PIDFD_GET_INFO, &info), 0); + ASSERT_EQ(errno, ESRCH); + + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(self->child_pidfd2, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_TRUE(WIFSIGNALED(info.exit_code)); + ASSERT_EQ(WTERMSIG(info.exit_code), SIGKILL); +} + +TEST_F(pidfd_info, success_exit) +{ + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID, + }; + + /* Process has exited but not been reaped so this must work. */ + ASSERT_EQ(ioctl(self->child_pidfd3, PIDFD_GET_INFO, &info), 0); + + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(self->child_pidfd3, PIDFD_GET_INFO, &info), 0); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_CREDS)); + /* Process has exited but not been reaped, so no PIDFD_INFO_EXIT information yet. */ + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_EXIT)); +} + +TEST_F(pidfd_info, success_reaped) +{ + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID, + }; + + /* Process has already been reaped and PIDFD_INFO_EXIT hasn't been set. */ + ASSERT_NE(ioctl(self->child_pidfd4, PIDFD_GET_INFO, &info), 0); + ASSERT_EQ(errno, ESRCH); + + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(self->child_pidfd4, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_TRUE(WIFEXITED(info.exit_code)); + ASSERT_EQ(WEXITSTATUS(info.exit_code), 0); +} + +TEST_F(pidfd_info, success_reaped_poll) +{ + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT, + }; + struct pollfd fds = {}; + int nevents; + + fds.events = POLLIN; + fds.fd = self->child_pidfd2; + + nevents = poll(&fds, 1, -1); + ASSERT_EQ(nevents, 1); + ASSERT_TRUE(!!(fds.revents & POLLIN)); + ASSERT_TRUE(!!(fds.revents & POLLHUP)); + + ASSERT_EQ(ioctl(self->child_pidfd2, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_TRUE(WIFSIGNALED(info.exit_code)); + ASSERT_EQ(WTERMSIG(info.exit_code), SIGKILL); +} + +static void *pidfd_info_pause_thread(void *arg) +{ + pid_t pid_thread = gettid(); + int ipc_socket = *(int *)arg; + + /* Inform the grand-parent what the tid of this thread is. */ + if (write_nointr(ipc_socket, &pid_thread, sizeof(pid_thread)) != sizeof(pid_thread)) + return NULL; + + close(ipc_socket); + + /* Sleep untill we're killed. */ + pause(); + return NULL; +} + +TEST_F(pidfd_info, thread_group) +{ + pid_t pid_leader, pid_poller, pid_thread; + pthread_t thread; + int nevents, pidfd_leader, pidfd_thread, pidfd_leader_thread, ret; + int ipc_sockets[2]; + struct pollfd fds = {}; + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT, + }, info2; + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + pid_leader = create_child(&pidfd_leader, 0); + EXPECT_GE(pid_leader, 0); + + if (pid_leader == 0) { + close(ipc_sockets[0]); + + /* The thread will outlive the thread-group leader. */ + if (pthread_create(&thread, NULL, pidfd_info_pause_thread, &ipc_sockets[1])) + syscall(__NR_exit, EXIT_FAILURE); + + /* Make the thread-group leader exit prematurely. */ + syscall(__NR_exit, EXIT_SUCCESS); + } + + /* + * Opening a PIDFD_THREAD aka thread-specific pidfd based on a + * thread-group leader must succeed. + */ + pidfd_leader_thread = sys_pidfd_open(pid_leader, PIDFD_THREAD); + ASSERT_GE(pidfd_leader_thread, 0); + + pid_poller = fork(); + ASSERT_GE(pid_poller, 0); + if (pid_poller == 0) { + /* + * We can't poll and wait for the old thread-group + * leader to exit using a thread-specific pidfd. The + * thread-group leader exited prematurely and + * notification is delayed until all subthreads have + * exited. + */ + fds.events = POLLIN; + fds.fd = pidfd_leader_thread; + nevents = poll(&fds, 1, 10000 /* wait 5 seconds */); + if (nevents != 0) + _exit(EXIT_FAILURE); + if (fds.revents & POLLIN) + _exit(EXIT_FAILURE); + if (fds.revents & POLLHUP) + _exit(EXIT_FAILURE); + _exit(EXIT_SUCCESS); + } + + /* Retrieve the tid of the thread. */ + EXPECT_EQ(close(ipc_sockets[1]), 0); + ASSERT_EQ(read_nointr(ipc_sockets[0], &pid_thread, sizeof(pid_thread)), sizeof(pid_thread)); + EXPECT_EQ(close(ipc_sockets[0]), 0); + + /* Opening a thread as a thread-group leader must fail. */ + pidfd_thread = sys_pidfd_open(pid_thread, 0); + ASSERT_LT(pidfd_thread, 0); + ASSERT_EQ(errno, ENOENT); + + /* Opening a thread as a PIDFD_THREAD must succeed. */ + pidfd_thread = sys_pidfd_open(pid_thread, PIDFD_THREAD); + ASSERT_GE(pidfd_thread, 0); + + ASSERT_EQ(wait_for_pid(pid_poller), 0); + + /* + * Note that pidfd_leader is a thread-group pidfd, so polling on it + * would only notify us once all thread in the thread-group have + * exited. So we can't poll before we have taken down the whole + * thread-group. + */ + + /* Get PIDFD_GET_INFO using the thread-group leader pidfd. */ + ASSERT_EQ(ioctl(pidfd_leader, PIDFD_GET_INFO, &info), 0); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_CREDS)); + /* Process has exited but not been reaped, so no PIDFD_INFO_EXIT information yet. */ + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_EQ(info.pid, pid_leader); + + /* + * Now retrieve the same info using the thread specific pidfd + * for the thread-group leader. + */ + info2.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader_thread, PIDFD_GET_INFO, &info2), 0); + ASSERT_TRUE(!!(info2.mask & PIDFD_INFO_CREDS)); + /* Process has exited but not been reaped, so no PIDFD_INFO_EXIT information yet. */ + ASSERT_FALSE(!!(info2.mask & PIDFD_INFO_EXIT)); + ASSERT_EQ(info2.pid, pid_leader); + + /* Now try the thread-specific pidfd. */ + ASSERT_EQ(ioctl(pidfd_thread, PIDFD_GET_INFO, &info), 0); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_CREDS)); + /* The thread hasn't exited, so no PIDFD_INFO_EXIT information yet. */ + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_EQ(info.pid, pid_thread); + + /* + * Take down the whole thread-group. The thread-group leader + * exited successfully but the thread will now be SIGKILLed. + * This must be reflected in the recorded exit information. + */ + EXPECT_EQ(sys_pidfd_send_signal(pidfd_leader, SIGKILL, NULL, 0), 0); + EXPECT_EQ(sys_waitid(P_PIDFD, pidfd_leader, NULL, WEXITED), 0); + + fds.events = POLLIN; + fds.fd = pidfd_leader; + nevents = poll(&fds, 1, -1); + ASSERT_EQ(nevents, 1); + ASSERT_TRUE(!!(fds.revents & POLLIN)); + /* The thread-group leader has been reaped. */ + ASSERT_TRUE(!!(fds.revents & POLLHUP)); + + /* + * Retrieve exit information for the thread-group leader via the + * thread-group leader pidfd. + */ + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + /* Even though the thread-group exited successfully it will still report the group exit code. */ + ASSERT_TRUE(WIFSIGNALED(info.exit_code)); + ASSERT_EQ(WTERMSIG(info.exit_code), SIGKILL); + + /* + * Retrieve exit information for the thread-group leader via the + * thread-specific pidfd. + */ + info2.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader_thread, PIDFD_GET_INFO, &info2), 0); + ASSERT_FALSE(!!(info2.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info2.mask & PIDFD_INFO_EXIT)); + + /* Even though the thread-group exited successfully it will still report the group exit code. */ + ASSERT_TRUE(WIFSIGNALED(info2.exit_code)); + ASSERT_EQ(WTERMSIG(info2.exit_code), SIGKILL); + + /* Retrieve exit information for the thread. */ + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_thread, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + + /* The thread got SIGKILLed. */ + ASSERT_TRUE(WIFSIGNALED(info.exit_code)); + ASSERT_EQ(WTERMSIG(info.exit_code), SIGKILL); + + EXPECT_EQ(close(pidfd_leader), 0); + EXPECT_EQ(close(pidfd_thread), 0); +} + +static void *pidfd_info_thread_exec(void *arg) +{ + pid_t pid_thread = gettid(); + int ipc_socket = *(int *)arg; + + /* Inform the grand-parent what the tid of this thread is. */ + if (write_nointr(ipc_socket, &pid_thread, sizeof(pid_thread)) != sizeof(pid_thread)) + return NULL; + + if (read_nointr(ipc_socket, &pid_thread, sizeof(pid_thread)) != sizeof(pid_thread)) + return NULL; + + close(ipc_socket); + + sys_execveat(AT_FDCWD, "pidfd_exec_helper", NULL, NULL, 0); + return NULL; +} + +TEST_F(pidfd_info, thread_group_exec) +{ + pid_t pid_leader, pid_poller, pid_thread; + pthread_t thread; + int nevents, pidfd_leader, pidfd_leader_thread, pidfd_thread, ret; + int ipc_sockets[2]; + struct pollfd fds = {}; + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT, + }; + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + pid_leader = create_child(&pidfd_leader, 0); + EXPECT_GE(pid_leader, 0); + + if (pid_leader == 0) { + close(ipc_sockets[0]); + + /* The thread will outlive the thread-group leader. */ + if (pthread_create(&thread, NULL, pidfd_info_thread_exec, &ipc_sockets[1])) + syscall(__NR_exit, EXIT_FAILURE); + + /* Make the thread-group leader exit prematurely. */ + syscall(__NR_exit, EXIT_SUCCESS); + } + + /* Open a thread-specific pidfd for the thread-group leader. */ + pidfd_leader_thread = sys_pidfd_open(pid_leader, PIDFD_THREAD); + ASSERT_GE(pidfd_leader_thread, 0); + + pid_poller = fork(); + ASSERT_GE(pid_poller, 0); + if (pid_poller == 0) { + /* + * We can't poll and wait for the old thread-group + * leader to exit using a thread-specific pidfd. The + * thread-group leader exited prematurely and + * notification is delayed until all subthreads have + * exited. + * + * When the thread has execed it will taken over the old + * thread-group leaders struct pid. Calling poll after + * the thread execed will thus block again because a new + * thread-group has started. + */ + fds.events = POLLIN; + fds.fd = pidfd_leader_thread; + nevents = poll(&fds, 1, 10000 /* wait 5 seconds */); + if (nevents != 0) + _exit(EXIT_FAILURE); + if (fds.revents & POLLIN) + _exit(EXIT_FAILURE); + if (fds.revents & POLLHUP) + _exit(EXIT_FAILURE); + _exit(EXIT_SUCCESS); + } + + /* Retrieve the tid of the thread. */ + EXPECT_EQ(close(ipc_sockets[1]), 0); + ASSERT_EQ(read_nointr(ipc_sockets[0], &pid_thread, sizeof(pid_thread)), sizeof(pid_thread)); + + /* Opening a thread as a PIDFD_THREAD must succeed. */ + pidfd_thread = sys_pidfd_open(pid_thread, PIDFD_THREAD); + ASSERT_GE(pidfd_thread, 0); + + /* Now that we've opened a thread-specific pidfd the thread can exec. */ + ASSERT_EQ(write_nointr(ipc_sockets[0], &pid_thread, sizeof(pid_thread)), sizeof(pid_thread)); + EXPECT_EQ(close(ipc_sockets[0]), 0); + + ASSERT_EQ(wait_for_pid(pid_poller), 0); + + /* Wait until the kernel has SIGKILLed the thread. */ + fds.events = POLLHUP; + fds.fd = pidfd_thread; + nevents = poll(&fds, 1, -1); + ASSERT_EQ(nevents, 1); + /* The thread has been reaped. */ + ASSERT_TRUE(!!(fds.revents & POLLHUP)); + + /* Retrieve thread-specific exit info from pidfd. */ + ASSERT_EQ(ioctl(pidfd_thread, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + /* + * While the kernel will have SIGKILLed the whole thread-group + * during exec it will cause the individual threads to exit + * cleanly. + */ + ASSERT_TRUE(WIFEXITED(info.exit_code)); + ASSERT_EQ(WEXITSTATUS(info.exit_code), 0); + + /* + * The thread-group leader is still alive, the thread has taken + * over its struct pid and thus its pid number. + */ + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader, PIDFD_GET_INFO, &info), 0); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_EQ(info.pid, pid_leader); + + /* Take down the thread-group leader. */ + EXPECT_EQ(sys_pidfd_send_signal(pidfd_leader, SIGKILL, NULL, 0), 0); + + /* + * Afte the exec we're dealing with an empty thread-group so now + * we must see an exit notification on the thread-specific pidfd + * for the thread-group leader as there's no subthread that can + * revive the struct pid. + */ + fds.events = POLLIN; + fds.fd = pidfd_leader_thread; + nevents = poll(&fds, 1, -1); + ASSERT_EQ(nevents, 1); + ASSERT_TRUE(!!(fds.revents & POLLIN)); + ASSERT_FALSE(!!(fds.revents & POLLHUP)); + + EXPECT_EQ(sys_waitid(P_PIDFD, pidfd_leader, NULL, WEXITED), 0); + + /* Retrieve exit information for the thread-group leader. */ + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + + EXPECT_EQ(close(pidfd_leader), 0); + EXPECT_EQ(close(pidfd_thread), 0); +} + +static void *pidfd_info_thread_exec_sane(void *arg) +{ + pid_t pid_thread = gettid(); + int ipc_socket = *(int *)arg; + + /* Inform the grand-parent what the tid of this thread is. */ + if (write_nointr(ipc_socket, &pid_thread, sizeof(pid_thread)) != sizeof(pid_thread)) + return NULL; + + if (read_nointr(ipc_socket, &pid_thread, sizeof(pid_thread)) != sizeof(pid_thread)) + return NULL; + + close(ipc_socket); + + sys_execveat(AT_FDCWD, "pidfd_exec_helper", NULL, NULL, 0); + return NULL; +} + +TEST_F(pidfd_info, thread_group_exec_thread) +{ + pid_t pid_leader, pid_poller, pid_thread; + pthread_t thread; + int nevents, pidfd_leader, pidfd_leader_thread, pidfd_thread, ret; + int ipc_sockets[2]; + struct pollfd fds = {}; + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT, + }; + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + pid_leader = create_child(&pidfd_leader, 0); + EXPECT_GE(pid_leader, 0); + + if (pid_leader == 0) { + close(ipc_sockets[0]); + + /* The thread will outlive the thread-group leader. */ + if (pthread_create(&thread, NULL, pidfd_info_thread_exec_sane, &ipc_sockets[1])) + syscall(__NR_exit, EXIT_FAILURE); + + /* + * Pause the thread-group leader. It will be killed once + * the subthread execs. + */ + pause(); + syscall(__NR_exit, EXIT_SUCCESS); + } + + /* Retrieve the tid of the thread. */ + EXPECT_EQ(close(ipc_sockets[1]), 0); + ASSERT_EQ(read_nointr(ipc_sockets[0], &pid_thread, sizeof(pid_thread)), sizeof(pid_thread)); + + /* Opening a thread as a PIDFD_THREAD must succeed. */ + pidfd_thread = sys_pidfd_open(pid_thread, PIDFD_THREAD); + ASSERT_GE(pidfd_thread, 0); + + /* Open a thread-specific pidfd for the thread-group leader. */ + pidfd_leader_thread = sys_pidfd_open(pid_leader, PIDFD_THREAD); + ASSERT_GE(pidfd_leader_thread, 0); + + pid_poller = fork(); + ASSERT_GE(pid_poller, 0); + if (pid_poller == 0) { + /* + * The subthread will now exec. The struct pid of the old + * thread-group leader will be assumed by the subthread which + * becomes the new thread-group leader. So no exit notification + * must be generated. Wait for 5 seconds and call it a success + * if no notification has been received. + */ + fds.events = POLLIN; + fds.fd = pidfd_leader_thread; + nevents = poll(&fds, 1, 10000 /* wait 5 seconds */); + if (nevents != 0) + _exit(EXIT_FAILURE); + if (fds.revents & POLLIN) + _exit(EXIT_FAILURE); + if (fds.revents & POLLHUP) + _exit(EXIT_FAILURE); + _exit(EXIT_SUCCESS); + } + + /* Now that we've opened a thread-specific pidfd the thread can exec. */ + ASSERT_EQ(write_nointr(ipc_sockets[0], &pid_thread, sizeof(pid_thread)), sizeof(pid_thread)); + EXPECT_EQ(close(ipc_sockets[0]), 0); + ASSERT_EQ(wait_for_pid(pid_poller), 0); + + /* Wait until the kernel has SIGKILLed the thread. */ + fds.events = POLLHUP; + fds.fd = pidfd_thread; + nevents = poll(&fds, 1, -1); + ASSERT_EQ(nevents, 1); + /* The thread has been reaped. */ + ASSERT_TRUE(!!(fds.revents & POLLHUP)); + + /* Retrieve thread-specific exit info from pidfd. */ + ASSERT_EQ(ioctl(pidfd_thread, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + /* + * While the kernel will have SIGKILLed the whole thread-group + * during exec it will cause the individual threads to exit + * cleanly. + */ + ASSERT_TRUE(WIFEXITED(info.exit_code)); + ASSERT_EQ(WEXITSTATUS(info.exit_code), 0); + + /* + * The thread-group leader is still alive, the thread has taken + * over its struct pid and thus its pid number. + */ + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader, PIDFD_GET_INFO, &info), 0); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_EXIT)); + ASSERT_EQ(info.pid, pid_leader); + + /* Take down the thread-group leader. */ + EXPECT_EQ(sys_pidfd_send_signal(pidfd_leader, SIGKILL, NULL, 0), 0); + + /* + * Afte the exec we're dealing with an empty thread-group so now + * we must see an exit notification on the thread-specific pidfd + * for the thread-group leader as there's no subthread that can + * revive the struct pid. + */ + fds.events = POLLIN; + fds.fd = pidfd_leader_thread; + nevents = poll(&fds, 1, -1); + ASSERT_EQ(nevents, 1); + ASSERT_TRUE(!!(fds.revents & POLLIN)); + ASSERT_FALSE(!!(fds.revents & POLLHUP)); + + EXPECT_EQ(sys_waitid(P_PIDFD, pidfd_leader, NULL, WEXITED), 0); + + /* Retrieve exit information for the thread-group leader. */ + info.mask = PIDFD_INFO_CGROUPID | PIDFD_INFO_EXIT; + ASSERT_EQ(ioctl(pidfd_leader, PIDFD_GET_INFO, &info), 0); + ASSERT_FALSE(!!(info.mask & PIDFD_INFO_CREDS)); + ASSERT_TRUE(!!(info.mask & PIDFD_INFO_EXIT)); + + EXPECT_EQ(close(pidfd_leader), 0); + EXPECT_EQ(close(pidfd_thread), 0); +} + +TEST_HARNESS_MAIN diff --git a/tools/testing/selftests/pidfd/pidfd_open_test.c b/tools/testing/selftests/pidfd/pidfd_open_test.c index b9fe75fc3e51..cd3de40e4977 100644 --- a/tools/testing/selftests/pidfd/pidfd_open_test.c +++ b/tools/testing/selftests/pidfd/pidfd_open_test.c @@ -6,7 +6,6 @@ #include <inttypes.h> #include <limits.h> #include <linux/types.h> -#include <linux/wait.h> #include <sched.h> #include <signal.h> #include <stdbool.h> @@ -14,6 +13,7 @@ #include <stdlib.h> #include <string.h> #include <syscall.h> +#include <sys/ioctl.h> #include <sys/mount.h> #include <sys/prctl.h> #include <sys/wait.h> @@ -121,10 +121,13 @@ out: int main(int argc, char **argv) { + struct pidfd_info info = { + .mask = PIDFD_INFO_CGROUPID, + }; int pidfd = -1, ret = 1; pid_t pid; - ksft_set_plan(3); + ksft_set_plan(4); pidfd = sys_pidfd_open(-1, 0); if (pidfd >= 0) { @@ -154,11 +157,63 @@ int main(int argc, char **argv) pid = get_pid_from_fdinfo_file(pidfd, "Pid:", sizeof("Pid:") - 1); ksft_print_msg("pidfd %d refers to process with pid %d\n", pidfd, pid); + if (ioctl(pidfd, PIDFD_GET_INFO, &info) < 0) { + ksft_print_msg("%s - failed to get info from pidfd\n", strerror(errno)); + goto on_error; + } + if (info.pid != pid) { + ksft_print_msg("pid from fdinfo file %d does not match pid from ioctl %d\n", + pid, info.pid); + goto on_error; + } + if (info.ppid != getppid()) { + ksft_print_msg("ppid %d does not match ppid from ioctl %d\n", + pid, info.pid); + goto on_error; + } + if (info.ruid != getuid()) { + ksft_print_msg("uid %d does not match uid from ioctl %d\n", + getuid(), info.ruid); + goto on_error; + } + if (info.rgid != getgid()) { + ksft_print_msg("gid %d does not match gid from ioctl %d\n", + getgid(), info.rgid); + goto on_error; + } + if (info.euid != geteuid()) { + ksft_print_msg("euid %d does not match euid from ioctl %d\n", + geteuid(), info.euid); + goto on_error; + } + if (info.egid != getegid()) { + ksft_print_msg("egid %d does not match egid from ioctl %d\n", + getegid(), info.egid); + goto on_error; + } + if (info.suid != geteuid()) { + ksft_print_msg("suid %d does not match suid from ioctl %d\n", + geteuid(), info.suid); + goto on_error; + } + if (info.sgid != getegid()) { + ksft_print_msg("sgid %d does not match sgid from ioctl %d\n", + getegid(), info.sgid); + goto on_error; + } + if ((info.mask & PIDFD_INFO_CGROUPID) && info.cgroupid == 0) { + ksft_print_msg("cgroupid should not be 0 when PIDFD_INFO_CGROUPID is set\n"); + goto on_error; + } + ksft_test_result_pass("get info from pidfd test: passed\n"); + ret = 0; on_error: if (pidfd >= 0) close(pidfd); - return !ret ? ksft_exit_pass() : ksft_exit_fail(); + if (ret) + ksft_exit_fail(); + ksft_exit_pass(); } diff --git a/tools/testing/selftests/pidfd/pidfd_poll_test.c b/tools/testing/selftests/pidfd/pidfd_poll_test.c index 4b115444dfe9..55d74a50358f 100644 --- a/tools/testing/selftests/pidfd/pidfd_poll_test.c +++ b/tools/testing/selftests/pidfd/pidfd_poll_test.c @@ -3,7 +3,6 @@ #define _GNU_SOURCE #include <errno.h> #include <linux/types.h> -#include <linux/wait.h> #include <poll.h> #include <signal.h> #include <stdbool.h> @@ -113,5 +112,5 @@ int main(int argc, char **argv) } ksft_test_result_pass("pidfd poll test: pass\n"); - return ksft_exit_pass(); + ksft_exit_pass(); } diff --git a/tools/testing/selftests/pidfd/pidfd_setns_test.c b/tools/testing/selftests/pidfd/pidfd_setns_test.c index 9418108eae13..e6a079b3d5e2 100644 --- a/tools/testing/selftests/pidfd/pidfd_setns_test.c +++ b/tools/testing/selftests/pidfd/pidfd_setns_test.c @@ -16,11 +16,8 @@ #include <unistd.h> #include <sys/socket.h> #include <sys/stat.h> -#include <linux/kcmp.h> #include "pidfd.h" -#include "../clone3/clone3_selftests.h" -#include "../kselftest.h" #include "../kselftest_harness.h" enum { @@ -32,21 +29,26 @@ enum { PIDFD_NS_NET, PIDFD_NS_CGROUP, PIDFD_NS_PIDCLD, + PIDFD_NS_TIME, + PIDFD_NS_TIMECLD, PIDFD_NS_MAX }; const struct ns_info { const char *name; int flag; + unsigned int pidfd_ioctl; } ns_info[] = { - [PIDFD_NS_USER] = { "user", CLONE_NEWUSER, }, - [PIDFD_NS_MNT] = { "mnt", CLONE_NEWNS, }, - [PIDFD_NS_PID] = { "pid", CLONE_NEWPID, }, - [PIDFD_NS_UTS] = { "uts", CLONE_NEWUTS, }, - [PIDFD_NS_IPC] = { "ipc", CLONE_NEWIPC, }, - [PIDFD_NS_NET] = { "net", CLONE_NEWNET, }, - [PIDFD_NS_CGROUP] = { "cgroup", CLONE_NEWCGROUP, }, - [PIDFD_NS_PIDCLD] = { "pid_for_children", 0, }, + [PIDFD_NS_USER] = { "user", CLONE_NEWUSER, PIDFD_GET_USER_NAMESPACE, }, + [PIDFD_NS_MNT] = { "mnt", CLONE_NEWNS, PIDFD_GET_MNT_NAMESPACE, }, + [PIDFD_NS_PID] = { "pid", CLONE_NEWPID, PIDFD_GET_PID_NAMESPACE, }, + [PIDFD_NS_UTS] = { "uts", CLONE_NEWUTS, PIDFD_GET_UTS_NAMESPACE, }, + [PIDFD_NS_IPC] = { "ipc", CLONE_NEWIPC, PIDFD_GET_IPC_NAMESPACE, }, + [PIDFD_NS_NET] = { "net", CLONE_NEWNET, PIDFD_GET_NET_NAMESPACE, }, + [PIDFD_NS_CGROUP] = { "cgroup", CLONE_NEWCGROUP, PIDFD_GET_CGROUP_NAMESPACE, }, + [PIDFD_NS_TIME] = { "time", CLONE_NEWTIME, PIDFD_GET_TIME_NAMESPACE, }, + [PIDFD_NS_PIDCLD] = { "pid_for_children", 0, PIDFD_GET_PID_FOR_CHILDREN_NAMESPACE, }, + [PIDFD_NS_TIMECLD] = { "time_for_children", 0, PIDFD_GET_TIME_FOR_CHILDREN_NAMESPACE, }, }; FIXTURE(current_nsset) @@ -54,6 +56,7 @@ FIXTURE(current_nsset) pid_t pid; int pidfd; int nsfds[PIDFD_NS_MAX]; + int child_pidfd_derived_nsfds[PIDFD_NS_MAX]; pid_t child_pid_exited; int child_pidfd_exited; @@ -61,36 +64,43 @@ FIXTURE(current_nsset) pid_t child_pid1; int child_pidfd1; int child_nsfds1[PIDFD_NS_MAX]; + int child_pidfd_derived_nsfds1[PIDFD_NS_MAX]; pid_t child_pid2; int child_pidfd2; int child_nsfds2[PIDFD_NS_MAX]; + int child_pidfd_derived_nsfds2[PIDFD_NS_MAX]; }; -static int sys_waitid(int which, pid_t pid, int options) +static bool switch_timens(void) { - return syscall(__NR_waitid, which, pid, NULL, options, NULL); -} + int fd, ret; -pid_t create_child(int *pidfd, unsigned flags) -{ - struct clone_args args = { - .flags = CLONE_PIDFD | flags, - .exit_signal = SIGCHLD, - .pidfd = ptr_to_u64(pidfd), - }; + if (unshare(CLONE_NEWTIME)) + return false; + + fd = open("/proc/self/ns/time_for_children", O_RDONLY | O_CLOEXEC); + if (fd < 0) + return false; - return sys_clone3(&args, sizeof(struct clone_args)); + ret = setns(fd, CLONE_NEWTIME); + close(fd); + return ret == 0; } FIXTURE_SETUP(current_nsset) { int i, proc_fd, ret; + int ipc_sockets[2]; + char c; for (i = 0; i < PIDFD_NS_MAX; i++) { - self->nsfds[i] = -EBADF; - self->child_nsfds1[i] = -EBADF; - self->child_nsfds2[i] = -EBADF; + self->nsfds[i] = -EBADF; + self->child_nsfds1[i] = -EBADF; + self->child_nsfds2[i] = -EBADF; + self->child_pidfd_derived_nsfds[i] = -EBADF; + self->child_pidfd_derived_nsfds1[i] = -EBADF; + self->child_pidfd_derived_nsfds2[i] = -EBADF; } proc_fd = open("/proc/self/ns", O_DIRECTORY | O_CLOEXEC); @@ -99,6 +109,11 @@ FIXTURE_SETUP(current_nsset) } self->pid = getpid(); + self->pidfd = sys_pidfd_open(self->pid, 0); + EXPECT_GT(self->pidfd, 0) { + TH_LOG("%m - Failed to open pidfd for process %d", self->pid); + } + for (i = 0; i < PIDFD_NS_MAX; i++) { const struct ns_info *info = &ns_info[i]; self->nsfds[i] = openat(proc_fd, info->name, O_RDONLY | O_CLOEXEC); @@ -108,53 +123,144 @@ FIXTURE_SETUP(current_nsset) info->name, self->pid); } } - } - self->pidfd = sys_pidfd_open(self->pid, 0); - EXPECT_GT(self->pidfd, 0) { - TH_LOG("%m - Failed to open pidfd for process %d", self->pid); + self->child_pidfd_derived_nsfds[i] = ioctl(self->pidfd, info->pidfd_ioctl, 0); + if (self->child_pidfd_derived_nsfds[i] < 0) { + EXPECT_EQ(errno, EOPNOTSUPP) { + TH_LOG("%m - Failed to derive %s namespace from pidfd of process %d", + info->name, self->pid); + } + } } /* Create task that exits right away. */ - self->child_pid_exited = create_child(&self->child_pidfd_exited, - CLONE_NEWUSER | CLONE_NEWNET); - EXPECT_GT(self->child_pid_exited, 0); - - if (self->child_pid_exited == 0) + self->child_pid_exited = create_child(&self->child_pidfd_exited, 0); + EXPECT_GE(self->child_pid_exited, 0); + + if (self->child_pid_exited == 0) { + if (self->nsfds[PIDFD_NS_USER] >= 0 && unshare(CLONE_NEWUSER) < 0) + _exit(EXIT_FAILURE); + if (self->nsfds[PIDFD_NS_NET] >= 0 && unshare(CLONE_NEWNET) < 0) + _exit(EXIT_FAILURE); _exit(EXIT_SUCCESS); + } - ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, WEXITED | WNOWAIT), 0); + ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, NULL, WEXITED | WNOWAIT), 0); self->pidfd = sys_pidfd_open(self->pid, 0); EXPECT_GE(self->pidfd, 0) { TH_LOG("%m - Failed to open pidfd for process %d", self->pid); } + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + /* Create tasks that will be stopped. */ - self->child_pid1 = create_child(&self->child_pidfd1, - CLONE_NEWUSER | CLONE_NEWNS | - CLONE_NEWCGROUP | CLONE_NEWIPC | - CLONE_NEWUTS | CLONE_NEWPID | - CLONE_NEWNET); + if (self->nsfds[PIDFD_NS_USER] >= 0 && self->nsfds[PIDFD_NS_PID] >= 0) + self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWUSER | CLONE_NEWPID); + else if (self->nsfds[PIDFD_NS_PID] >= 0) + self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWPID); + else if (self->nsfds[PIDFD_NS_USER] >= 0) + self->child_pid1 = create_child(&self->child_pidfd1, CLONE_NEWUSER); + else + self->child_pid1 = create_child(&self->child_pidfd1, 0); EXPECT_GE(self->child_pid1, 0); if (self->child_pid1 == 0) { + close(ipc_sockets[0]); + + if (self->nsfds[PIDFD_NS_MNT] >= 0 && unshare(CLONE_NEWNS) < 0) { + TH_LOG("%m - Failed to unshare mount namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_CGROUP] >= 0 && unshare(CLONE_NEWCGROUP) < 0) { + TH_LOG("%m - Failed to unshare cgroup namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_IPC] >= 0 && unshare(CLONE_NEWIPC) < 0) { + TH_LOG("%m - Failed to unshare ipc namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_UTS] >= 0 && unshare(CLONE_NEWUTS) < 0) { + TH_LOG("%m - Failed to unshare uts namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_NET] >= 0 && unshare(CLONE_NEWNET) < 0) { + TH_LOG("%m - Failed to unshare net namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_TIME] >= 0 && !switch_timens()) { + TH_LOG("%m - Failed to unshare time namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + pause(); _exit(EXIT_SUCCESS); } - self->child_pid2 = create_child(&self->child_pidfd2, - CLONE_NEWUSER | CLONE_NEWNS | - CLONE_NEWCGROUP | CLONE_NEWIPC | - CLONE_NEWUTS | CLONE_NEWPID | - CLONE_NEWNET); + close(ipc_sockets[1]); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + close(ipc_sockets[0]); + + ret = socketpair(AF_LOCAL, SOCK_STREAM | SOCK_CLOEXEC, 0, ipc_sockets); + EXPECT_EQ(ret, 0); + + if (self->nsfds[PIDFD_NS_USER] >= 0 && self->nsfds[PIDFD_NS_PID] >= 0) + self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWUSER | CLONE_NEWPID); + else if (self->nsfds[PIDFD_NS_PID] >= 0) + self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWPID); + else if (self->nsfds[PIDFD_NS_USER] >= 0) + self->child_pid2 = create_child(&self->child_pidfd2, CLONE_NEWUSER); + else + self->child_pid2 = create_child(&self->child_pidfd2, 0); EXPECT_GE(self->child_pid2, 0); if (self->child_pid2 == 0) { + close(ipc_sockets[0]); + + if (self->nsfds[PIDFD_NS_MNT] >= 0 && unshare(CLONE_NEWNS) < 0) { + TH_LOG("%m - Failed to unshare mount namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_CGROUP] >= 0 && unshare(CLONE_NEWCGROUP) < 0) { + TH_LOG("%m - Failed to unshare cgroup namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_IPC] >= 0 && unshare(CLONE_NEWIPC) < 0) { + TH_LOG("%m - Failed to unshare ipc namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_UTS] >= 0 && unshare(CLONE_NEWUTS) < 0) { + TH_LOG("%m - Failed to unshare uts namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_NET] >= 0 && unshare(CLONE_NEWNET) < 0) { + TH_LOG("%m - Failed to unshare net namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + if (self->nsfds[PIDFD_NS_TIME] >= 0 && !switch_timens()) { + TH_LOG("%m - Failed to unshare time namespace for process %d", self->pid); + _exit(EXIT_FAILURE); + } + + if (write_nointr(ipc_sockets[1], "1", 1) < 0) + _exit(EXIT_FAILURE); + + close(ipc_sockets[1]); + pause(); _exit(EXIT_SUCCESS); } + close(ipc_sockets[1]); + ASSERT_EQ(read_nointr(ipc_sockets[0], &c, 1), 1); + close(ipc_sockets[0]); + for (i = 0; i < PIDFD_NS_MAX; i++) { char p[100]; @@ -193,6 +299,22 @@ FIXTURE_SETUP(current_nsset) info->name, self->child_pid1); } } + + self->child_pidfd_derived_nsfds1[i] = ioctl(self->child_pidfd1, info->pidfd_ioctl, 0); + if (self->child_pidfd_derived_nsfds1[i] < 0) { + EXPECT_EQ(errno, EOPNOTSUPP) { + TH_LOG("%m - Failed to derive %s namespace from pidfd of process %d", + info->name, self->child_pid1); + } + } + + self->child_pidfd_derived_nsfds2[i] = ioctl(self->child_pidfd2, info->pidfd_ioctl, 0); + if (self->child_pidfd_derived_nsfds2[i] < 0) { + EXPECT_EQ(errno, EOPNOTSUPP) { + TH_LOG("%m - Failed to derive %s namespace from pidfd of process %d", + info->name, self->child_pid2); + } + } } close(proc_fd); @@ -214,15 +336,21 @@ FIXTURE_TEARDOWN(current_nsset) close(self->child_nsfds1[i]); if (self->child_nsfds2[i] >= 0) close(self->child_nsfds2[i]); + if (self->child_pidfd_derived_nsfds[i] >= 0) + close(self->child_pidfd_derived_nsfds[i]); + if (self->child_pidfd_derived_nsfds1[i] >= 0) + close(self->child_pidfd_derived_nsfds1[i]); + if (self->child_pidfd_derived_nsfds2[i] >= 0) + close(self->child_pidfd_derived_nsfds2[i]); } if (self->child_pidfd1 >= 0) EXPECT_EQ(0, close(self->child_pidfd1)); if (self->child_pidfd2 >= 0) EXPECT_EQ(0, close(self->child_pidfd2)); - ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, WEXITED), 0); - ASSERT_EQ(sys_waitid(P_PID, self->child_pid1, WEXITED), 0); - ASSERT_EQ(sys_waitid(P_PID, self->child_pid2, WEXITED), 0); + ASSERT_EQ(sys_waitid(P_PID, self->child_pid_exited, NULL, WEXITED), 0); + ASSERT_EQ(sys_waitid(P_PID, self->child_pid1, NULL, WEXITED), 0); + ASSERT_EQ(sys_waitid(P_PID, self->child_pid2, NULL, WEXITED), 0); } static int preserve_ns(const int pid, const char *ns) @@ -372,6 +500,42 @@ TEST_F(current_nsset, nsfd_incremental_setns) } } +TEST_F(current_nsset, pidfd_derived_nsfd_incremental_setns) +{ + int i; + pid_t pid; + + pid = getpid(); + for (i = 0; i < PIDFD_NS_MAX; i++) { + const struct ns_info *info = &ns_info[i]; + int nsfd; + + if (self->child_pidfd_derived_nsfds1[i] < 0) + continue; + + if (info->flag) { + ASSERT_EQ(setns(self->child_pidfd_derived_nsfds1[i], info->flag), 0) { + TH_LOG("%m - Failed to setns to %s namespace of %d via nsfd %d", + info->name, self->child_pid1, + self->child_pidfd_derived_nsfds1[i]); + } + } + + /* Verify that we have changed to the correct namespaces. */ + if (info->flag == CLONE_NEWPID) + nsfd = self->child_pidfd_derived_nsfds[i]; + else + nsfd = self->child_pidfd_derived_nsfds1[i]; + ASSERT_EQ(in_same_namespace(nsfd, pid, info->name), 1) { + TH_LOG("setns failed to place us correctly into %s namespace of %d via nsfd %d", + info->name, self->child_pid1, + self->child_pidfd_derived_nsfds1[i]); + } + TH_LOG("Managed to correctly setns to %s namespace of %d via nsfd %d", + info->name, self->child_pid1, self->child_pidfd_derived_nsfds1[i]); + } +} + TEST_F(current_nsset, pidfd_one_shot_setns) { unsigned flags = 0; @@ -468,6 +632,28 @@ TEST_F(current_nsset, no_foul_play) info->name, self->child_pid2, self->child_nsfds2[i]); } + + /* + * Can't setns to a user namespace outside of our hierarchy since we + * don't have caps in there and didn't create it. That means that under + * no circumstances should we be able to setns to any of the other + * ones since they aren't owned by our user namespace. + */ + for (i = 0; i < PIDFD_NS_MAX; i++) { + const struct ns_info *info = &ns_info[i]; + + if (self->child_pidfd_derived_nsfds2[i] < 0 || !info->flag) + continue; + + ASSERT_NE(setns(self->child_pidfd_derived_nsfds2[i], info->flag), 0) { + TH_LOG("Managed to setns to %s namespace of %d via nsfd %d", + info->name, self->child_pid2, + self->child_pidfd_derived_nsfds2[i]); + } + TH_LOG("%m - Correctly failed to setns to %s namespace of %d via nsfd %d", + info->name, self->child_pid2, + self->child_pidfd_derived_nsfds2[i]); + } } TEST(setns_einval) diff --git a/tools/testing/selftests/pidfd/pidfd_test.c b/tools/testing/selftests/pidfd/pidfd_test.c index 7aff2d3b42c0..fcd85cad9f18 100644 --- a/tools/testing/selftests/pidfd/pidfd_test.c +++ b/tools/testing/selftests/pidfd/pidfd_test.c @@ -8,6 +8,7 @@ #include <sched.h> #include <signal.h> #include <stdio.h> +#include <stdbool.h> #include <stdlib.h> #include <string.h> #include <syscall.h> @@ -27,6 +28,8 @@ #define MAX_EVENTS 5 +static bool have_pidfd_send_signal; + static pid_t pidfd_clone(int flags, int *pidfd, int (*fn)(void *)) { size_t stack_size = 1024; @@ -39,12 +42,41 @@ static pid_t pidfd_clone(int flags, int *pidfd, int (*fn)(void *)) #endif } -static int signal_received; +static pthread_t signal_received; static void set_signal_received_on_sigusr1(int sig) { if (sig == SIGUSR1) - signal_received = 1; + signal_received = pthread_self(); +} + +static int send_signal(int pidfd) +{ + int ret = 0; + + if (sys_pidfd_send_signal(pidfd, SIGUSR1, NULL, 0) < 0) { + ret = -EINVAL; + goto exit; + } + + if (signal_received != pthread_self()) { + ret = -EINVAL; + goto exit; + } + +exit: + signal_received = 0; + return ret; +} + +static void *send_signal_worker(void *arg) +{ + int pidfd = (int)(intptr_t)arg; + int ret; + + /* We forward any errors for the caller to handle. */ + ret = send_signal(pidfd); + return (void *)(intptr_t)ret; } /* @@ -53,28 +85,58 @@ static void set_signal_received_on_sigusr1(int sig) */ static int test_pidfd_send_signal_simple_success(void) { - int pidfd, ret; + int pidfd; const char *test_name = "pidfd_send_signal send SIGUSR1"; + pthread_t thread; + void *thread_res; + int err; + + if (!have_pidfd_send_signal) { + ksft_test_result_skip( + "%s test: pidfd_send_signal() syscall not supported\n", + test_name); + return 0; + } + signal(SIGUSR1, set_signal_received_on_sigusr1); + + /* Try sending a signal to ourselves via /proc/self. */ pidfd = open("/proc/self", O_DIRECTORY | O_CLOEXEC); if (pidfd < 0) ksft_exit_fail_msg( "%s test: Failed to open process file descriptor\n", test_name); + err = send_signal(pidfd); + if (err) + ksft_exit_fail_msg( + "%s test: Error %d on sending pidfd signal\n", + test_name, err); + close(pidfd); - signal(SIGUSR1, set_signal_received_on_sigusr1); + /* Now try the same thing only using PIDFD_SELF_THREAD_GROUP. */ + err = send_signal(PIDFD_SELF_THREAD_GROUP); + if (err) + ksft_exit_fail_msg( + "%s test: Error %d on PIDFD_SELF_THREAD_GROUP signal\n", + test_name, err); - ret = sys_pidfd_send_signal(pidfd, SIGUSR1, NULL, 0); - close(pidfd); - if (ret < 0) - ksft_exit_fail_msg("%s test: Failed to send signal\n", + /* + * Now try the same thing in a thread and assert thread ID is equal to + * worker thread ID. + */ + if (pthread_create(&thread, NULL, send_signal_worker, + (void *)(intptr_t)PIDFD_SELF_THREAD)) + ksft_exit_fail_msg("%s test: Failed to create thread\n", test_name); - - if (signal_received != 1) - ksft_exit_fail_msg("%s test: Failed to receive signal\n", + if (pthread_join(thread, &thread_res)) + ksft_exit_fail_msg("%s test: Failed to join thread\n", test_name); + err = (int)(intptr_t)thread_res; + if (err) + ksft_exit_fail_msg( + "%s test: Error %d on PIDFD_SELF_THREAD signal\n", + test_name, err); - signal_received = 0; ksft_test_result_pass("%s test: Sent signal\n", test_name); return 0; } @@ -86,6 +148,13 @@ static int test_pidfd_send_signal_exited_fail(void) pid_t pid; const char *test_name = "pidfd_send_signal signal exited process"; + if (!have_pidfd_send_signal) { + ksft_test_result_skip( + "%s test: pidfd_send_signal() syscall not supported\n", + test_name); + return 0; + } + pid = fork(); if (pid < 0) ksft_exit_fail_msg("%s test: Failed to create new process\n", @@ -98,7 +167,8 @@ static int test_pidfd_send_signal_exited_fail(void) pidfd = open(buf, O_DIRECTORY | O_CLOEXEC); - (void)wait_for_pid(pid); + ret = wait_for_pid(pid); + ksft_print_msg("waitpid WEXITSTATUS=%d\n", ret); if (pidfd < 0) ksft_exit_fail_msg( @@ -137,16 +207,34 @@ static int test_pidfd_send_signal_recycled_pid_fail(void) pid_t pid1; const char *test_name = "pidfd_send_signal signal recycled pid"; + if (!have_pidfd_send_signal) { + ksft_test_result_skip( + "%s test: pidfd_send_signal() syscall not supported\n", + test_name); + return 0; + } + ret = unshare(CLONE_NEWPID); - if (ret < 0) + if (ret < 0) { + if (errno == EPERM) { + ksft_test_result_skip("%s test: Unsharing pid namespace not permitted\n", + test_name); + return 0; + } ksft_exit_fail_msg("%s test: Failed to unshare pid namespace\n", test_name); + } ret = unshare(CLONE_NEWNS); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: Failed to unshare mount namespace\n", - test_name); + if (ret < 0) { + if (errno == EPERM) { + ksft_test_result_skip("%s test: Unsharing mount namespace not permitted\n", + test_name); + return 0; + } + ksft_exit_fail_msg("%s test: Failed to unshare mount namespace\n", + test_name); + } ret = mount(NULL, "/", NULL, MS_REC | MS_PRIVATE, 0); if (ret < 0) @@ -295,7 +383,7 @@ static int test_pidfd_send_signal_recycled_pid_fail(void) ksft_exit_fail_msg("%s test: Failed to recycle pid %d\n", test_name, PID_RECYCLE); case PIDFD_SKIP: - ksft_print_msg("%s test: Skipping test\n", test_name); + ksft_test_result_skip("%s test: Skipping test\n", test_name); ret = 0; break; case PIDFD_XFAIL: @@ -325,15 +413,17 @@ static int test_pidfd_send_signal_syscall_support(void) ret = sys_pidfd_send_signal(pidfd, 0, NULL, 0); if (ret < 0) { - if (errno == ENOSYS) - ksft_exit_skip( + if (errno == ENOSYS) { + ksft_test_result_skip( "%s test: pidfd_send_signal() syscall not supported\n", test_name); - + return 0; + } ksft_exit_fail_msg("%s test: Failed to send signal\n", test_name); } + have_pidfd_send_signal = true; close(pidfd); ksft_test_result_pass( "%s test: pidfd_send_signal() syscall is supported. Tests can be executed\n", @@ -343,13 +433,13 @@ static int test_pidfd_send_signal_syscall_support(void) static void *test_pidfd_poll_exec_thread(void *priv) { - ksft_print_msg("Child Thread: starting. pid %d tid %d ; and sleeping\n", + ksft_print_msg("Child Thread: starting. pid %d tid %ld ; and sleeping\n", getpid(), syscall(SYS_gettid)); ksft_print_msg("Child Thread: doing exec of sleep\n"); execl("/bin/sleep", "sleep", str(CHILD_THREAD_MIN_WAIT), (char *)NULL); - ksft_print_msg("Child Thread: DONE. pid %d tid %d\n", + ksft_print_msg("Child Thread: DONE. pid %d tid %ld\n", getpid(), syscall(SYS_gettid)); return NULL; } @@ -376,7 +466,7 @@ static void poll_pidfd(const char *test_name, int pidfd) c = epoll_wait(epoll_fd, events, MAX_EVENTS, 5000); if (c != 1 || !(events[0].events & EPOLLIN)) - ksft_exit_fail_msg("%s test: Unexpected epoll_wait result (c=%d, events=%x) ", + ksft_exit_fail_msg("%s test: Unexpected epoll_wait result (c=%d, events=%x) " "(errno %d)\n", test_name, c, events[0].events, errno); @@ -389,7 +479,7 @@ static int child_poll_exec_test(void *args) { pthread_t t1; - ksft_print_msg("Child (pidfd): starting. pid %d tid %d\n", getpid(), + ksft_print_msg("Child (pidfd): starting. pid %d tid %ld\n", getpid(), syscall(SYS_gettid)); pthread_create(&t1, NULL, test_pidfd_poll_exec_thread, NULL); /* @@ -398,13 +488,14 @@ static int child_poll_exec_test(void *args) */ while (1) sleep(1); + + return 0; } static void test_pidfd_poll_exec(int use_waitpid) { int pid, pidfd = 0; int status, ret; - pthread_t t1; time_t prog_start = time(NULL); const char *test_name = "pidfd_poll check for premature notification on child thread exec"; @@ -441,10 +532,10 @@ static void test_pidfd_poll_exec(int use_waitpid) static void *test_pidfd_poll_leader_exit_thread(void *priv) { - ksft_print_msg("Child Thread: starting. pid %d tid %d ; and sleeping\n", + ksft_print_msg("Child Thread: starting. pid %d tid %ld ; and sleeping\n", getpid(), syscall(SYS_gettid)); sleep(CHILD_THREAD_MIN_WAIT); - ksft_print_msg("Child Thread: DONE. pid %d tid %d\n", getpid(), syscall(SYS_gettid)); + ksft_print_msg("Child Thread: DONE. pid %d tid %ld\n", getpid(), syscall(SYS_gettid)); return NULL; } @@ -453,23 +544,24 @@ static int child_poll_leader_exit_test(void *args) { pthread_t t1, t2; - ksft_print_msg("Child: starting. pid %d tid %d\n", getpid(), syscall(SYS_gettid)); + ksft_print_msg("Child: starting. pid %d tid %ld\n", getpid(), syscall(SYS_gettid)); pthread_create(&t1, NULL, test_pidfd_poll_leader_exit_thread, NULL); pthread_create(&t2, NULL, test_pidfd_poll_leader_exit_thread, NULL); /* - * glibc exit calls exit_group syscall, so explicity call exit only + * glibc exit calls exit_group syscall, so explicitly call exit only * so that only the group leader exits, leaving the threads alone. */ *child_exit_secs = time(NULL); syscall(SYS_exit, 0); + /* Never reached, but appeases compiler thinking we should return. */ + exit(0); } static void test_pidfd_poll_leader_exit(int use_waitpid) { int pid, pidfd = 0; - int status, ret; - time_t prog_start = time(NULL); + int status, ret = 0; const char *test_name = "pidfd_poll check for premature notification on non-empty" "group leader exit"; @@ -521,7 +613,7 @@ static void test_pidfd_poll_leader_exit(int use_waitpid) int main(int argc, char **argv) { ksft_print_header(); - ksft_set_plan(4); + ksft_set_plan(8); test_pidfd_poll_exec(0); test_pidfd_poll_exec(1); @@ -532,5 +624,5 @@ int main(int argc, char **argv) test_pidfd_send_signal_exited_fail(); test_pidfd_send_signal_recycled_pid_fail(); - return ksft_exit_pass(); + ksft_exit_pass(); } diff --git a/tools/testing/selftests/pidfd/pidfd_wait.c b/tools/testing/selftests/pidfd/pidfd_wait.c index 7079f8eef792..1e2d49751cde 100644 --- a/tools/testing/selftests/pidfd/pidfd_wait.c +++ b/tools/testing/selftests/pidfd/pidfd_wait.c @@ -17,255 +17,206 @@ #include <unistd.h> #include "pidfd.h" -#include "../kselftest.h" +#include "../kselftest_harness.h" #define ptr_to_u64(ptr) ((__u64)((uintptr_t)(ptr))) -static pid_t sys_clone3(struct clone_args *args) -{ - return syscall(__NR_clone3, args, sizeof(struct clone_args)); -} - -static int sys_waitid(int which, pid_t pid, siginfo_t *info, int options, - struct rusage *ru) -{ - return syscall(__NR_waitid, which, pid, info, options, ru); -} +/* Attempt to de-conflict with the selftests tree. */ +#ifndef SKIP +#define SKIP(s, ...) XFAIL(s, ##__VA_ARGS__) +#endif -static int test_pidfd_wait_simple(void) +TEST(wait_simple) { - const char *test_name = "pidfd wait simple"; - int pidfd = -1, status = 0; + int pidfd = -1; pid_t parent_tid = -1; - struct clone_args args = { + struct __clone_args args = { .parent_tid = ptr_to_u64(&parent_tid), .pidfd = ptr_to_u64(&pidfd), .flags = CLONE_PIDFD | CLONE_PARENT_SETTID, .exit_signal = SIGCHLD, }; - int ret; pid_t pid; siginfo_t info = { .si_signo = 0, }; pidfd = open("/proc/self", O_DIRECTORY | O_RDONLY | O_CLOEXEC); - if (pidfd < 0) - ksft_exit_fail_msg("%s test: failed to open /proc/self %s\n", - test_name, strerror(errno)); + ASSERT_GE(pidfd, 0); - pid = sys_waitid(P_PIDFD, pidfd, &info, WEXITED, NULL); - if (pid == 0) - ksft_exit_fail_msg( - "%s test: succeeded to wait on invalid pidfd %s\n", - test_name, strerror(errno)); - close(pidfd); + pid = sys_waitid(P_PIDFD, pidfd, &info, WEXITED); + ASSERT_NE(pid, 0); + EXPECT_EQ(close(pidfd), 0); pidfd = -1; pidfd = open("/dev/null", O_RDONLY | O_CLOEXEC); - if (pidfd == 0) - ksft_exit_fail_msg("%s test: failed to open /dev/null %s\n", - test_name, strerror(errno)); + ASSERT_GE(pidfd, 0); - pid = sys_waitid(P_PIDFD, pidfd, &info, WEXITED, NULL); - if (pid == 0) - ksft_exit_fail_msg( - "%s test: succeeded to wait on invalid pidfd %s\n", - test_name, strerror(errno)); - close(pidfd); + pid = sys_waitid(P_PIDFD, pidfd, &info, WEXITED); + ASSERT_NE(pid, 0); + EXPECT_EQ(close(pidfd), 0); pidfd = -1; - pid = sys_clone3(&args); - if (pid < 0) - ksft_exit_fail_msg("%s test: failed to create new process %s\n", - test_name, strerror(errno)); + pid = sys_clone3(&args, sizeof(args)); + ASSERT_GE(pid, 0); if (pid == 0) exit(EXIT_SUCCESS); - pid = sys_waitid(P_PIDFD, pidfd, &info, WEXITED, NULL); - if (pid < 0) - ksft_exit_fail_msg( - "%s test: failed to wait on process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - if (!WIFEXITED(info.si_status) || WEXITSTATUS(info.si_status)) - ksft_exit_fail_msg( - "%s test: unexpected status received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - close(pidfd); - - if (info.si_signo != SIGCHLD) - ksft_exit_fail_msg( - "%s test: unexpected si_signo value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_signo, parent_tid, pidfd, - strerror(errno)); - - if (info.si_code != CLD_EXITED) - ksft_exit_fail_msg( - "%s test: unexpected si_code value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_code, parent_tid, pidfd, - strerror(errno)); - - if (info.si_pid != parent_tid) - ksft_exit_fail_msg( - "%s test: unexpected si_pid value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_pid, parent_tid, pidfd, - strerror(errno)); - - ksft_test_result_pass("%s test: Passed\n", test_name); - return 0; + pid = sys_waitid(P_PIDFD, pidfd, &info, WEXITED); + ASSERT_GE(pid, 0); + ASSERT_EQ(WIFEXITED(info.si_status), true); + ASSERT_EQ(WEXITSTATUS(info.si_status), 0); + EXPECT_EQ(close(pidfd), 0); + + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_EXITED); + ASSERT_EQ(info.si_pid, parent_tid); } -static int test_pidfd_wait_states(void) +TEST(wait_states) { - const char *test_name = "pidfd wait states"; - int pidfd = -1, status = 0; + int pidfd = -1; pid_t parent_tid = -1; - struct clone_args args = { + struct __clone_args args = { .parent_tid = ptr_to_u64(&parent_tid), .pidfd = ptr_to_u64(&pidfd), .flags = CLONE_PIDFD | CLONE_PARENT_SETTID, .exit_signal = SIGCHLD, }; - int ret; + int pfd[2]; pid_t pid; siginfo_t info = { .si_signo = 0, }; - pid = sys_clone3(&args); - if (pid < 0) - ksft_exit_fail_msg("%s test: failed to create new process %s\n", - test_name, strerror(errno)); + ASSERT_EQ(pipe(pfd), 0); + pid = sys_clone3(&args, sizeof(args)); + ASSERT_GE(pid, 0); if (pid == 0) { + char buf[2]; + + close(pfd[1]); kill(getpid(), SIGSTOP); + ASSERT_EQ(read(pfd[0], buf, 1), 1); + close(pfd[0]); kill(getpid(), SIGSTOP); exit(EXIT_SUCCESS); } - ret = sys_waitid(P_PIDFD, pidfd, &info, WSTOPPED, NULL); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: failed to wait on WSTOPPED process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - if (info.si_signo != SIGCHLD) - ksft_exit_fail_msg( - "%s test: unexpected si_signo value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_signo, parent_tid, pidfd, - strerror(errno)); - - if (info.si_code != CLD_STOPPED) - ksft_exit_fail_msg( - "%s test: unexpected si_code value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_code, parent_tid, pidfd, - strerror(errno)); - - if (info.si_pid != parent_tid) - ksft_exit_fail_msg( - "%s test: unexpected si_pid value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_pid, parent_tid, pidfd, - strerror(errno)); - - ret = sys_pidfd_send_signal(pidfd, SIGCONT, NULL, 0); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: failed to send signal to process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - ret = sys_waitid(P_PIDFD, pidfd, &info, WCONTINUED, NULL); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: failed to wait WCONTINUED on process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - if (info.si_signo != SIGCHLD) - ksft_exit_fail_msg( - "%s test: unexpected si_signo value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_signo, parent_tid, pidfd, - strerror(errno)); - - if (info.si_code != CLD_CONTINUED) - ksft_exit_fail_msg( - "%s test: unexpected si_code value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_code, parent_tid, pidfd, - strerror(errno)); - - if (info.si_pid != parent_tid) - ksft_exit_fail_msg( - "%s test: unexpected si_pid value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_pid, parent_tid, pidfd, - strerror(errno)); - - ret = sys_waitid(P_PIDFD, pidfd, &info, WUNTRACED, NULL); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: failed to wait on WUNTRACED process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - if (info.si_signo != SIGCHLD) - ksft_exit_fail_msg( - "%s test: unexpected si_signo value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_signo, parent_tid, pidfd, - strerror(errno)); - - if (info.si_code != CLD_STOPPED) - ksft_exit_fail_msg( - "%s test: unexpected si_code value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_code, parent_tid, pidfd, - strerror(errno)); - - if (info.si_pid != parent_tid) - ksft_exit_fail_msg( - "%s test: unexpected si_pid value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_pid, parent_tid, pidfd, - strerror(errno)); - - ret = sys_pidfd_send_signal(pidfd, SIGKILL, NULL, 0); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: failed to send SIGKILL to process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - ret = sys_waitid(P_PIDFD, pidfd, &info, WEXITED, NULL); - if (ret < 0) - ksft_exit_fail_msg( - "%s test: failed to wait on WEXITED process with pid %d and pidfd %d: %s\n", - test_name, parent_tid, pidfd, strerror(errno)); - - if (info.si_signo != SIGCHLD) - ksft_exit_fail_msg( - "%s test: unexpected si_signo value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_signo, parent_tid, pidfd, - strerror(errno)); - - if (info.si_code != CLD_KILLED) - ksft_exit_fail_msg( - "%s test: unexpected si_code value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_code, parent_tid, pidfd, - strerror(errno)); - - if (info.si_pid != parent_tid) - ksft_exit_fail_msg( - "%s test: unexpected si_pid value %d received after waiting on process with pid %d and pidfd %d: %s\n", - test_name, info.si_pid, parent_tid, pidfd, - strerror(errno)); - - close(pidfd); - - ksft_test_result_pass("%s test: Passed\n", test_name); - return 0; + close(pfd[0]); + ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WSTOPPED), 0); + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_STOPPED); + ASSERT_EQ(info.si_pid, parent_tid); + + ASSERT_EQ(sys_pidfd_send_signal(pidfd, SIGCONT, NULL, 0), 0); + + ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WCONTINUED), 0); + ASSERT_EQ(write(pfd[1], "C", 1), 1); + close(pfd[1]); + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_CONTINUED); + ASSERT_EQ(info.si_pid, parent_tid); + + ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WUNTRACED), 0); + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_STOPPED); + ASSERT_EQ(info.si_pid, parent_tid); + + ASSERT_EQ(sys_pidfd_send_signal(pidfd, SIGKILL, NULL, 0), 0); + + ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WEXITED), 0); + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_KILLED); + ASSERT_EQ(info.si_pid, parent_tid); + + EXPECT_EQ(close(pidfd), 0); } -int main(int argc, char **argv) +TEST(wait_nonblock) { - ksft_print_header(); - ksft_set_plan(2); + int pidfd; + unsigned int flags = 0; + pid_t parent_tid = -1; + struct __clone_args args = { + .parent_tid = ptr_to_u64(&parent_tid), + .flags = CLONE_PARENT_SETTID, + .exit_signal = SIGCHLD, + }; + int ret; + pid_t pid; + siginfo_t info = { + .si_signo = 0, + }; + + /* + * Callers need to see ECHILD with non-blocking pidfds when no child + * processes exists. + */ + pidfd = sys_pidfd_open(getpid(), PIDFD_NONBLOCK); + EXPECT_GE(pidfd, 0) { + /* pidfd_open() doesn't support PIDFD_NONBLOCK. */ + ASSERT_EQ(errno, EINVAL); + SKIP(return, "Skipping PIDFD_NONBLOCK test"); + } - test_pidfd_wait_simple(); - test_pidfd_wait_states(); + ret = sys_waitid(P_PIDFD, pidfd, &info, WEXITED); + ASSERT_LT(ret, 0); + ASSERT_EQ(errno, ECHILD); + EXPECT_EQ(close(pidfd), 0); - return ksft_exit_pass(); + pid = sys_clone3(&args, sizeof(args)); + ASSERT_GE(pid, 0); + + if (pid == 0) { + kill(getpid(), SIGSTOP); + exit(EXIT_SUCCESS); + } + + pidfd = sys_pidfd_open(pid, PIDFD_NONBLOCK); + EXPECT_GE(pidfd, 0) { + /* pidfd_open() doesn't support PIDFD_NONBLOCK. */ + ASSERT_EQ(errno, EINVAL); + SKIP(return, "Skipping PIDFD_NONBLOCK test"); + } + + flags = fcntl(pidfd, F_GETFL, 0); + ASSERT_GT(flags, 0); + ASSERT_GT((flags & O_NONBLOCK), 0); + + /* + * Callers need to see EAGAIN/EWOULDBLOCK with non-blocking pidfd when + * child processes exist but none have exited. + */ + ret = sys_waitid(P_PIDFD, pidfd, &info, WEXITED); + ASSERT_LT(ret, 0); + ASSERT_EQ(errno, EAGAIN); + + /* + * Callers need to continue seeing 0 with non-blocking pidfd and + * WNOHANG raised explicitly when child processes exist but none have + * exited. + */ + ret = sys_waitid(P_PIDFD, pidfd, &info, WEXITED | WNOHANG); + ASSERT_EQ(ret, 0); + + ASSERT_EQ(fcntl(pidfd, F_SETFL, (flags & ~O_NONBLOCK)), 0); + + ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WSTOPPED), 0); + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_STOPPED); + ASSERT_EQ(info.si_pid, parent_tid); + + ASSERT_EQ(sys_pidfd_send_signal(pidfd, SIGCONT, NULL, 0), 0); + + ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WEXITED), 0); + ASSERT_EQ(info.si_signo, SIGCHLD); + ASSERT_EQ(info.si_code, CLD_EXITED); + ASSERT_EQ(info.si_pid, parent_tid); + + EXPECT_EQ(close(pidfd), 0); } + +TEST_HARNESS_MAIN |