/* SPDX-License-Identifier: LGPL-2.1-or-later */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #if HAVE_VALGRIND_VALGRIND_H #include #endif #include "alloc-util.h" #include "architecture.h" #include "env-util.h" #include "errno-util.h" #include "escape.h" #include "fd-util.h" #include "fileio.h" #include "fs-util.h" #include "locale-util.h" #include "log.h" #include "macro.h" #include "memory-util.h" #include "missing_sched.h" #include "missing_syscall.h" #include "namespace-util.h" #include "path-util.h" #include "process-util.h" #include "raw-clone.h" #include "rlimit-util.h" #include "signal-util.h" #include "stat-util.h" #include "stdio-util.h" #include "string-table.h" #include "string-util.h" #include "terminal-util.h" #include "user-util.h" #include "utf8.h" /* The kernel limits userspace processes to TASK_COMM_LEN (16 bytes), but allows higher values for its own * workers, e.g. "kworker/u9:3-kcryptd/253:0". Let's pick a fixed smallish limit that will work for the kernel. */ #define COMM_MAX_LEN 128 static int get_process_state(pid_t pid) { _cleanup_free_ char *line = NULL; const char *p; char state; int r; assert(pid >= 0); /* Shortcut: if we are enquired about our own state, we are obviously running */ if (pid == 0 || pid == getpid_cached()) return (unsigned char) 'R'; p = procfs_file_alloca(pid, "stat"); r = read_one_line_file(p, &line); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; p = strrchr(line, ')'); if (!p) return -EIO; p++; if (sscanf(p, " %c", &state) != 1) return -EIO; return (unsigned char) state; } int get_process_comm(pid_t pid, char **ret) { _cleanup_free_ char *escaped = NULL, *comm = NULL; int r; assert(ret); assert(pid >= 0); if (pid == 0 || pid == getpid_cached()) { comm = new0(char, TASK_COMM_LEN + 1); /* Must fit in 16 byte according to prctl(2) */ if (!comm) return -ENOMEM; if (prctl(PR_GET_NAME, comm) < 0) return -errno; } else { const char *p; p = procfs_file_alloca(pid, "comm"); /* Note that process names of kernel threads can be much longer than TASK_COMM_LEN */ r = read_one_line_file(p, &comm); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; } escaped = new(char, COMM_MAX_LEN); if (!escaped) return -ENOMEM; /* Escape unprintable characters, just in case, but don't grow the string beyond the underlying size */ cellescape(escaped, COMM_MAX_LEN, comm); *ret = TAKE_PTR(escaped); return 0; } static int get_process_cmdline_nulstr( pid_t pid, size_t max_size, ProcessCmdlineFlags flags, char **ret, size_t *ret_size) { const char *p; char *t; size_t k; int r; /* Retrieves a process' command line as a "sized nulstr", i.e. possibly without the last NUL, but * with a specified size. * * If PROCESS_CMDLINE_COMM_FALLBACK is specified in flags and the process has no command line set * (the case for kernel threads), or has a command line that resolves to the empty string, will * return the "comm" name of the process instead. This will use at most _SC_ARG_MAX bytes of input * data. * * Returns an error, 0 if output was read but is truncated, 1 otherwise. */ p = procfs_file_alloca(pid, "cmdline"); r = read_virtual_file(p, max_size, &t, &k); /* Let's assume that each input byte results in >= 1 * columns of output. We ignore zero-width codepoints. */ if (r == -ENOENT) return -ESRCH; if (r < 0) return r; if (k == 0) { t = mfree(t); if (!(flags & PROCESS_CMDLINE_COMM_FALLBACK)) return -ENOENT; /* Kernel threads have no argv[] */ _cleanup_free_ char *comm = NULL; r = get_process_comm(pid, &comm); if (r < 0) return r; t = strjoin("[", comm, "]"); if (!t) return -ENOMEM; k = strlen(t); r = k <= max_size; if (r == 0) /* truncation */ t[max_size] = '\0'; } *ret = t; *ret_size = k; return r; } int get_process_cmdline(pid_t pid, size_t max_columns, ProcessCmdlineFlags flags, char **ret) { _cleanup_free_ char *t = NULL; size_t k; char *ans; assert(pid >= 0); assert(ret); /* Retrieve and format a commandline. See above for discussion of retrieval options. * * There are two main formatting modes: * * - when PROCESS_CMDLINE_QUOTE is specified, output is quoted in C/Python style. If no shell special * characters are present, this output can be copy-pasted into the terminal to execute. UTF-8 * output is assumed. * * - otherwise, a compact non-roundtrippable form is returned. Non-UTF8 bytes are replaced by �. The * returned string is of the specified console width at most, abbreviated with an ellipsis. * * Returns -ESRCH if the process doesn't exist, and -ENOENT if the process has no command line (and * PROCESS_CMDLINE_COMM_FALLBACK is not specified). Returns 0 and sets *line otherwise. */ int full = get_process_cmdline_nulstr(pid, max_columns, flags, &t, &k); if (full < 0) return full; if (flags & (PROCESS_CMDLINE_QUOTE | PROCESS_CMDLINE_QUOTE_POSIX)) { ShellEscapeFlags shflags = SHELL_ESCAPE_EMPTY | FLAGS_SET(flags, PROCESS_CMDLINE_QUOTE_POSIX) * SHELL_ESCAPE_POSIX; assert(!(flags & PROCESS_CMDLINE_USE_LOCALE)); _cleanup_strv_free_ char **args = NULL; args = strv_parse_nulstr(t, k); if (!args) return -ENOMEM; /* Drop trailing empty strings. See issue #21186. */ STRV_FOREACH_BACKWARDS(p, args) { if (!isempty(*p)) break; *p = mfree(*p); } ans = quote_command_line(args, shflags); if (!ans) return -ENOMEM; } else { /* Arguments are separated by NULs. Let's replace those with spaces. */ for (size_t i = 0; i < k - 1; i++) if (t[i] == '\0') t[i] = ' '; delete_trailing_chars(t, WHITESPACE); bool eight_bit = (flags & PROCESS_CMDLINE_USE_LOCALE) && !is_locale_utf8(); ans = escape_non_printable_full(t, max_columns, eight_bit * XESCAPE_8_BIT | !full * XESCAPE_FORCE_ELLIPSIS); if (!ans) return -ENOMEM; ans = str_realloc(ans); } *ret = ans; return 0; } static int update_argv(const char name[], size_t l) { static int can_do = -1; if (can_do == 0) return 0; can_do = false; /* We'll set it to true only if the whole process works */ /* Let's not bother with this if we don't have euid == 0. Strictly speaking we should check for the * CAP_SYS_RESOURCE capability which is independent of the euid. In our own code the capability generally is * present only for euid == 0, hence let's use this as quick bypass check, to avoid calling mmap() if * PR_SET_MM_ARG_{START,END} fails with EPERM later on anyway. After all geteuid() is dead cheap to call, but * mmap() is not. */ if (geteuid() != 0) return log_debug_errno(SYNTHETIC_ERRNO(EPERM), "Skipping PR_SET_MM, as we don't have privileges."); static size_t mm_size = 0; static char *mm = NULL; int r; if (mm_size < l+1) { size_t nn_size; char *nn; nn_size = PAGE_ALIGN(l+1); nn = mmap(NULL, nn_size, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0); if (nn == MAP_FAILED) return log_debug_errno(errno, "mmap() failed: %m"); strncpy(nn, name, nn_size); /* Now, let's tell the kernel about this new memory */ if (prctl(PR_SET_MM, PR_SET_MM_ARG_START, (unsigned long) nn, 0, 0) < 0) { if (ERRNO_IS_PRIVILEGE(errno)) return log_debug_errno(errno, "PR_SET_MM_ARG_START failed: %m"); /* HACK: prctl() API is kind of dumb on this point. The existing end address may already be * below the desired start address, in which case the kernel may have kicked this back due * to a range-check failure (see linux/kernel/sys.c:validate_prctl_map() to see this in * action). The proper solution would be to have a prctl() API that could set both start+end * simultaneously, or at least let us query the existing address to anticipate this condition * and respond accordingly. For now, we can only guess at the cause of this failure and try * a workaround--which will briefly expand the arg space to something potentially huge before * resizing it to what we want. */ log_debug_errno(errno, "PR_SET_MM_ARG_START failed, attempting PR_SET_MM_ARG_END hack: %m"); if (prctl(PR_SET_MM, PR_SET_MM_ARG_END, (unsigned long) nn + l + 1, 0, 0) < 0) { r = log_debug_errno(errno, "PR_SET_MM_ARG_END hack failed, proceeding without: %m"); (void) munmap(nn, nn_size); return r; } if (prctl(PR_SET_MM, PR_SET_MM_ARG_START, (unsigned long) nn, 0, 0) < 0) return log_debug_errno(errno, "PR_SET_MM_ARG_START still failed, proceeding without: %m"); } else { /* And update the end pointer to the new end, too. If this fails, we don't really know what * to do, it's pretty unlikely that we can rollback, hence we'll just accept the failure, * and continue. */ if (prctl(PR_SET_MM, PR_SET_MM_ARG_END, (unsigned long) nn + l + 1, 0, 0) < 0) log_debug_errno(errno, "PR_SET_MM_ARG_END failed, proceeding without: %m"); } if (mm) (void) munmap(mm, mm_size); mm = nn; mm_size = nn_size; } else { strncpy(mm, name, mm_size); /* Update the end pointer, continuing regardless of any failure. */ if (prctl(PR_SET_MM, PR_SET_MM_ARG_END, (unsigned long) mm + l + 1, 0, 0) < 0) log_debug_errno(errno, "PR_SET_MM_ARG_END failed, proceeding without: %m"); } can_do = true; return 0; } int rename_process(const char name[]) { bool truncated = false; /* This is a like a poor man's setproctitle(). It changes the comm field, argv[0], and also the glibc's * internally used name of the process. For the first one a limit of 16 chars applies; to the second one in * many cases one of 10 (i.e. length of "/sbin/init") — however if we have CAP_SYS_RESOURCES it is unbounded; * to the third one 7 (i.e. the length of "systemd". If you pass a longer string it will likely be * truncated. * * Returns 0 if a name was set but truncated, > 0 if it was set but not truncated. */ if (isempty(name)) return -EINVAL; /* let's not confuse users unnecessarily with an empty name */ if (!is_main_thread()) return -EPERM; /* Let's not allow setting the process name from other threads than the main one, as we * cache things without locking, and we make assumptions that PR_SET_NAME sets the * process name that isn't correct on any other threads */ size_t l = strlen(name); /* First step, change the comm field. The main thread's comm is identical to the process comm. This means we * can use PR_SET_NAME, which sets the thread name for the calling thread. */ if (prctl(PR_SET_NAME, name) < 0) log_debug_errno(errno, "PR_SET_NAME failed: %m"); if (l >= TASK_COMM_LEN) /* Linux userspace process names can be 15 chars at max */ truncated = true; /* Second step, change glibc's ID of the process name. */ if (program_invocation_name) { size_t k; k = strlen(program_invocation_name); strncpy(program_invocation_name, name, k); if (l > k) truncated = true; } /* Third step, completely replace the argv[] array the kernel maintains for us. This requires privileges, but * has the advantage that the argv[] array is exactly what we want it to be, and not filled up with zeros at * the end. This is the best option for changing /proc/self/cmdline. */ (void) update_argv(name, l); /* Fourth step: in all cases we'll also update the original argv[], so that our own code gets it right too if * it still looks here */ if (saved_argc > 0) { if (saved_argv[0]) { size_t k; k = strlen(saved_argv[0]); strncpy(saved_argv[0], name, k); if (l > k) truncated = true; } for (int i = 1; i < saved_argc; i++) { if (!saved_argv[i]) break; memzero(saved_argv[i], strlen(saved_argv[i])); } } return !truncated; } int is_kernel_thread(pid_t pid) { _cleanup_free_ char *line = NULL; unsigned long long flags; size_t l, i; const char *p; char *q; int r; if (IN_SET(pid, 0, 1) || pid == getpid_cached()) /* pid 1, and we ourselves certainly aren't a kernel thread */ return 0; if (!pid_is_valid(pid)) return -EINVAL; p = procfs_file_alloca(pid, "stat"); r = read_one_line_file(p, &line); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; /* Skip past the comm field */ q = strrchr(line, ')'); if (!q) return -EINVAL; q++; /* Skip 6 fields to reach the flags field */ for (i = 0; i < 6; i++) { l = strspn(q, WHITESPACE); if (l < 1) return -EINVAL; q += l; l = strcspn(q, WHITESPACE); if (l < 1) return -EINVAL; q += l; } /* Skip preceding whitespace */ l = strspn(q, WHITESPACE); if (l < 1) return -EINVAL; q += l; /* Truncate the rest */ l = strcspn(q, WHITESPACE); if (l < 1) return -EINVAL; q[l] = 0; r = safe_atollu(q, &flags); if (r < 0) return r; return !!(flags & PF_KTHREAD); } int get_process_capeff(pid_t pid, char **ret) { const char *p; int r; assert(pid >= 0); assert(ret); p = procfs_file_alloca(pid, "status"); r = get_proc_field(p, "CapEff", WHITESPACE, ret); if (r == -ENOENT) return -ESRCH; return r; } static int get_process_link_contents(pid_t pid, const char *proc_file, char **ret) { const char *p; int r; assert(proc_file); p = procfs_file_alloca(pid, proc_file); r = readlink_malloc(p, ret); return r == -ENOENT ? -ESRCH : r; } int get_process_exe(pid_t pid, char **ret) { char *d; int r; assert(pid >= 0); r = get_process_link_contents(pid, "exe", ret); if (r < 0) return r; if (ret) { d = endswith(*ret, " (deleted)"); if (d) *d = '\0'; } return 0; } static int get_process_id(pid_t pid, const char *field, uid_t *ret) { _cleanup_fclose_ FILE *f = NULL; const char *p; int r; assert(field); assert(ret); if (pid < 0) return -EINVAL; p = procfs_file_alloca(pid, "status"); r = fopen_unlocked(p, "re", &f); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; for (;;) { _cleanup_free_ char *line = NULL; char *l; r = read_line(f, LONG_LINE_MAX, &line); if (r < 0) return r; if (r == 0) break; l = strstrip(line); if (startswith(l, field)) { l += strlen(field); l += strspn(l, WHITESPACE); l[strcspn(l, WHITESPACE)] = 0; return parse_uid(l, ret); } } return -EIO; } int get_process_uid(pid_t pid, uid_t *ret) { if (pid == 0 || pid == getpid_cached()) { *ret = getuid(); return 0; } return get_process_id(pid, "Uid:", ret); } int get_process_gid(pid_t pid, gid_t *ret) { if (pid == 0 || pid == getpid_cached()) { *ret = getgid(); return 0; } assert_cc(sizeof(uid_t) == sizeof(gid_t)); return get_process_id(pid, "Gid:", ret); } int get_process_cwd(pid_t pid, char **ret) { assert(pid >= 0); if (pid == 0 || pid == getpid_cached()) return safe_getcwd(ret); return get_process_link_contents(pid, "cwd", ret); } int get_process_root(pid_t pid, char **ret) { assert(pid >= 0); return get_process_link_contents(pid, "root", ret); } #define ENVIRONMENT_BLOCK_MAX (5U*1024U*1024U) int get_process_environ(pid_t pid, char **ret) { _cleanup_fclose_ FILE *f = NULL; _cleanup_free_ char *outcome = NULL; size_t sz = 0; const char *p; int r; assert(pid >= 0); assert(ret); p = procfs_file_alloca(pid, "environ"); r = fopen_unlocked(p, "re", &f); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; for (;;) { char c; if (sz >= ENVIRONMENT_BLOCK_MAX) return -ENOBUFS; if (!GREEDY_REALLOC(outcome, sz + 5)) return -ENOMEM; r = safe_fgetc(f, &c); if (r < 0) return r; if (r == 0) break; if (c == '\0') outcome[sz++] = '\n'; else sz += cescape_char(c, outcome + sz); } outcome[sz] = '\0'; *ret = TAKE_PTR(outcome); return 0; } int get_process_ppid(pid_t pid, pid_t *ret) { _cleanup_free_ char *line = NULL; unsigned long ppid; const char *p; int r; assert(pid >= 0); if (pid == 0 || pid == getpid_cached()) { if (ret) *ret = getppid(); return 0; } if (pid == 1) /* PID 1 has no parent, shortcut this case */ return -EADDRNOTAVAIL; p = procfs_file_alloca(pid, "stat"); r = read_one_line_file(p, &line); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; /* Let's skip the pid and comm fields. The latter is enclosed in () but does not escape any () in its * value, so let's skip over it manually */ p = strrchr(line, ')'); if (!p) return -EIO; p++; if (sscanf(p, " " "%*c " /* state */ "%lu ", /* ppid */ &ppid) != 1) return -EIO; /* If ppid is zero the process has no parent. Which might be the case for PID 1 but also for * processes originating in other namespaces that are inserted into a pidns. Return a recognizable * error in this case. */ if (ppid == 0) return -EADDRNOTAVAIL; if ((pid_t) ppid < 0 || (unsigned long) (pid_t) ppid != ppid) return -ERANGE; if (ret) *ret = (pid_t) ppid; return 0; } int get_process_umask(pid_t pid, mode_t *ret) { _cleanup_free_ char *m = NULL; const char *p; int r; assert(pid >= 0); assert(ret); p = procfs_file_alloca(pid, "status"); r = get_proc_field(p, "Umask", WHITESPACE, &m); if (r == -ENOENT) return -ESRCH; return parse_mode(m, ret); } int wait_for_terminate(pid_t pid, siginfo_t *status) { siginfo_t dummy; assert(pid >= 1); if (!status) status = &dummy; for (;;) { zero(*status); if (waitid(P_PID, pid, status, WEXITED) < 0) { if (errno == EINTR) continue; return negative_errno(); } return 0; } } /* * Return values: * < 0 : wait_for_terminate() failed to get the state of the * process, the process was terminated by a signal, or * failed for an unknown reason. * >=0 : The process terminated normally, and its exit code is * returned. * * That is, success is indicated by a return value of zero, and an * error is indicated by a non-zero value. * * A warning is emitted if the process terminates abnormally, * and also if it returns non-zero unless check_exit_code is true. */ int wait_for_terminate_and_check(const char *name, pid_t pid, WaitFlags flags) { _cleanup_free_ char *buffer = NULL; siginfo_t status; int r, prio; assert(pid > 1); if (!name) { r = get_process_comm(pid, &buffer); if (r < 0) log_debug_errno(r, "Failed to acquire process name of " PID_FMT ", ignoring: %m", pid); else name = buffer; } prio = flags & WAIT_LOG_ABNORMAL ? LOG_ERR : LOG_DEBUG; r = wait_for_terminate(pid, &status); if (r < 0) return log_full_errno(prio, r, "Failed to wait for %s: %m", strna(name)); if (status.si_code == CLD_EXITED) { if (status.si_status != EXIT_SUCCESS) log_full(flags & WAIT_LOG_NON_ZERO_EXIT_STATUS ? LOG_ERR : LOG_DEBUG, "%s failed with exit status %i.", strna(name), status.si_status); else log_debug("%s succeeded.", name); return status.si_status; } else if (IN_SET(status.si_code, CLD_KILLED, CLD_DUMPED)) { log_full(prio, "%s terminated by signal %s.", strna(name), signal_to_string(status.si_status)); return -EPROTO; } log_full(prio, "%s failed due to unknown reason.", strna(name)); return -EPROTO; } /* * Return values: * * < 0 : wait_for_terminate_with_timeout() failed to get the state of the process, the process timed out, the process * was terminated by a signal, or failed for an unknown reason. * * >=0 : The process terminated normally with no failures. * * Success is indicated by a return value of zero, a timeout is indicated by ETIMEDOUT, and all other child failure * states are indicated by error is indicated by a non-zero value. * * This call assumes SIGCHLD has been blocked already, in particular before the child to wait for has been forked off * to remain entirely race-free. */ int wait_for_terminate_with_timeout(pid_t pid, usec_t timeout) { sigset_t mask; int r; usec_t until; assert_se(sigemptyset(&mask) == 0); assert_se(sigaddset(&mask, SIGCHLD) == 0); /* Drop into a sigtimewait-based timeout. Waiting for the * pid to exit. */ until = usec_add(now(CLOCK_MONOTONIC), timeout); for (;;) { usec_t n; siginfo_t status = {}; n = now(CLOCK_MONOTONIC); if (n >= until) break; r = RET_NERRNO(sigtimedwait(&mask, NULL, TIMESPEC_STORE(until - n))); /* Assuming we woke due to the child exiting. */ if (waitid(P_PID, pid, &status, WEXITED|WNOHANG) == 0) { if (status.si_pid == pid) { /* This is the correct child. */ if (status.si_code == CLD_EXITED) return (status.si_status == 0) ? 0 : -EPROTO; else return -EPROTO; } } /* Not the child, check for errors and proceed appropriately */ if (r < 0) { switch (r) { case -EAGAIN: /* Timed out, child is likely hung. */ return -ETIMEDOUT; case -EINTR: /* Received a different signal and should retry */ continue; default: /* Return any unexpected errors */ return r; } } } return -EPROTO; } void sigkill_wait(pid_t pid) { assert(pid > 1); (void) kill(pid, SIGKILL); (void) wait_for_terminate(pid, NULL); } void sigkill_waitp(pid_t *pid) { PROTECT_ERRNO; if (!pid) return; if (*pid <= 1) return; sigkill_wait(*pid); } void sigterm_wait(pid_t pid) { assert(pid > 1); (void) kill_and_sigcont(pid, SIGTERM); (void) wait_for_terminate(pid, NULL); } int kill_and_sigcont(pid_t pid, int sig) { int r; r = RET_NERRNO(kill(pid, sig)); /* If this worked, also send SIGCONT, unless we already just sent a SIGCONT, or SIGKILL was sent which isn't * affected by a process being suspended anyway. */ if (r >= 0 && !IN_SET(sig, SIGCONT, SIGKILL)) (void) kill(pid, SIGCONT); return r; } int getenv_for_pid(pid_t pid, const char *field, char **ret) { _cleanup_fclose_ FILE *f = NULL; char *value = NULL; const char *path; size_t l, sum = 0; int r; assert(pid >= 0); assert(field); assert(ret); if (pid == 0 || pid == getpid_cached()) { const char *e; e = getenv(field); if (!e) { *ret = NULL; return 0; } value = strdup(e); if (!value) return -ENOMEM; *ret = value; return 1; } if (!pid_is_valid(pid)) return -EINVAL; path = procfs_file_alloca(pid, "environ"); r = fopen_unlocked(path, "re", &f); if (r == -ENOENT) return -ESRCH; if (r < 0) return r; l = strlen(field); for (;;) { _cleanup_free_ char *line = NULL; if (sum > ENVIRONMENT_BLOCK_MAX) /* Give up searching eventually */ return -ENOBUFS; r = read_nul_string(f, LONG_LINE_MAX, &line); if (r < 0) return r; if (r == 0) /* EOF */ break; sum += r; if (strneq(line, field, l) && line[l] == '=') { value = strdup(line + l + 1); if (!value) return -ENOMEM; *ret = value; return 1; } } *ret = NULL; return 0; } int pid_is_my_child(pid_t pid) { pid_t ppid; int r; if (pid <= 1) return false; r = get_process_ppid(pid, &ppid); if (r < 0) return r; return ppid == getpid_cached(); } bool pid_is_unwaited(pid_t pid) { /* Checks whether a PID is still valid at all, including a zombie */ if (pid < 0) return false; if (pid <= 1) /* If we or PID 1 would be dead and have been waited for, this code would not be running */ return true; if (pid == getpid_cached()) return true; if (kill(pid, 0) >= 0) return true; return errno != ESRCH; } bool pid_is_alive(pid_t pid) { int r; /* Checks whether a PID is still valid and not a zombie */ if (pid < 0) return false; if (pid <= 1) /* If we or PID 1 would be a zombie, this code would not be running */ return true; if (pid == getpid_cached()) return true; r = get_process_state(pid); if (IN_SET(r, -ESRCH, 'Z')) return false; return true; } int pid_from_same_root_fs(pid_t pid) { const char *root; if (pid < 0) return false; if (pid == 0 || pid == getpid_cached()) return true; root = procfs_file_alloca(pid, "root"); return files_same(root, "/proc/1/root", 0); } bool is_main_thread(void) { static thread_local int cached = 0; if (_unlikely_(cached == 0)) cached = getpid_cached() == gettid() ? 1 : -1; return cached > 0; } bool oom_score_adjust_is_valid(int oa) { return oa >= OOM_SCORE_ADJ_MIN && oa <= OOM_SCORE_ADJ_MAX; } unsigned long personality_from_string(const char *p) { Architecture architecture; if (!p) return PERSONALITY_INVALID; /* Parse a personality specifier. We use our own identifiers that indicate specific ABIs, rather than just * hints regarding the register size, since we want to keep things open for multiple locally supported ABIs for * the same register size. */ architecture = architecture_from_string(p); if (architecture < 0) return PERSONALITY_INVALID; if (architecture == native_architecture()) return PER_LINUX; #ifdef ARCHITECTURE_SECONDARY if (architecture == ARCHITECTURE_SECONDARY) return PER_LINUX32; #endif return PERSONALITY_INVALID; } const char* personality_to_string(unsigned long p) { Architecture architecture = _ARCHITECTURE_INVALID; if (p == PER_LINUX) architecture = native_architecture(); #ifdef ARCHITECTURE_SECONDARY else if (p == PER_LINUX32) architecture = ARCHITECTURE_SECONDARY; #endif if (architecture < 0) return NULL; return architecture_to_string(architecture); } int safe_personality(unsigned long p) { int ret; /* So here's the deal, personality() is weirdly defined by glibc. In some cases it returns a failure via errno, * and in others as negative return value containing an errno-like value. Let's work around this: this is a * wrapper that uses errno if it is set, and uses the return value otherwise. And then it sets both errno and * the return value indicating the same issue, so that we are definitely on the safe side. * * See https://github.com/systemd/systemd/issues/6737 */ errno = 0; ret = personality(p); if (ret < 0) { if (errno != 0) return -errno; errno = -ret; } return ret; } int opinionated_personality(unsigned long *ret) { int current; /* Returns the current personality, or PERSONALITY_INVALID if we can't determine it. This function is a bit * opinionated though, and ignores all the finer-grained bits and exotic personalities, only distinguishing the * two most relevant personalities: PER_LINUX and PER_LINUX32. */ current = safe_personality(PERSONALITY_INVALID); if (current < 0) return current; if (((unsigned long) current & 0xffff) == PER_LINUX32) *ret = PER_LINUX32; else *ret = PER_LINUX; return 0; } void valgrind_summary_hack(void) { #if HAVE_VALGRIND_VALGRIND_H if (getpid_cached() == 1 && RUNNING_ON_VALGRIND) { pid_t pid; pid = raw_clone(SIGCHLD); if (pid < 0) log_emergency_errno(errno, "Failed to fork off valgrind helper: %m"); else if (pid == 0) exit(EXIT_SUCCESS); else { log_info("Spawned valgrind helper as PID "PID_FMT".", pid); (void) wait_for_terminate(pid, NULL); } } #endif } int pid_compare_func(const pid_t *a, const pid_t *b) { /* Suitable for usage in qsort() */ return CMP(*a, *b); } /* The cached PID, possible values: * * == UNSET [0] → cache not initialized yet * == BUSY [-1] → some thread is initializing it at the moment * any other → the cached PID */ #define CACHED_PID_UNSET ((pid_t) 0) #define CACHED_PID_BUSY ((pid_t) -1) static pid_t cached_pid = CACHED_PID_UNSET; void reset_cached_pid(void) { /* Invoked in the child after a fork(), i.e. at the first moment the PID changed */ cached_pid = CACHED_PID_UNSET; } pid_t getpid_cached(void) { static bool installed = false; pid_t current_value; /* getpid_cached() is much like getpid(), but caches the value in local memory, to avoid having to invoke a * system call each time. This restores glibc behaviour from before 2.24, when getpid() was unconditionally * cached. Starting with 2.24 getpid() started to become prohibitively expensive when used for detecting when * objects were used across fork()s. With this caching the old behaviour is somewhat restored. * * https://bugzilla.redhat.com/show_bug.cgi?id=1443976 * https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=c579f48edba88380635ab98cb612030e3ed8691e */ current_value = __sync_val_compare_and_swap(&cached_pid, CACHED_PID_UNSET, CACHED_PID_BUSY); switch (current_value) { case CACHED_PID_UNSET: { /* Not initialized yet, then do so now */ pid_t new_pid; new_pid = raw_getpid(); if (!installed) { /* __register_atfork() either returns 0 or -ENOMEM, in its glibc implementation. Since it's * only half-documented (glibc doesn't document it but LSB does — though only superficially) * we'll check for errors only in the most generic fashion possible. */ if (pthread_atfork(NULL, NULL, reset_cached_pid) != 0) { /* OOM? Let's try again later */ cached_pid = CACHED_PID_UNSET; return new_pid; } installed = true; } cached_pid = new_pid; return new_pid; } case CACHED_PID_BUSY: /* Somebody else is currently initializing */ return raw_getpid(); default: /* Properly initialized */ return current_value; } } int must_be_root(void) { if (geteuid() == 0) return 0; return log_error_errno(SYNTHETIC_ERRNO(EPERM), "Need to be root."); } static void restore_sigsetp(sigset_t **ssp) { if (*ssp) (void) sigprocmask(SIG_SETMASK, *ssp, NULL); } int safe_fork_full( const char *name, const int except_fds[], size_t n_except_fds, ForkFlags flags, pid_t *ret_pid) { pid_t original_pid, pid; sigset_t saved_ss, ss; _unused_ _cleanup_(restore_sigsetp) sigset_t *saved_ssp = NULL; bool block_signals = false, block_all = false; int prio, r; /* A wrapper around fork(), that does a couple of important initializations in addition to mere forking. Always * returns the child's PID in *ret_pid. Returns == 0 in the child, and > 0 in the parent. */ prio = flags & FORK_LOG ? LOG_ERR : LOG_DEBUG; original_pid = getpid_cached(); if (flags & FORK_FLUSH_STDIO) { fflush(stdout); fflush(stderr); /* This one shouldn't be necessary, stderr should be unbuffered anyway, but let's better be safe than sorry */ } if (flags & (FORK_RESET_SIGNALS|FORK_DEATHSIG)) { /* We temporarily block all signals, so that the new child has them blocked initially. This way, we can * be sure that SIGTERMs are not lost we might send to the child. */ assert_se(sigfillset(&ss) >= 0); block_signals = block_all = true; } else if (flags & FORK_WAIT) { /* Let's block SIGCHLD at least, so that we can safely watch for the child process */ assert_se(sigemptyset(&ss) >= 0); assert_se(sigaddset(&ss, SIGCHLD) >= 0); block_signals = true; } if (block_signals) { if (sigprocmask(SIG_SETMASK, &ss, &saved_ss) < 0) return log_full_errno(prio, errno, "Failed to set signal mask: %m"); saved_ssp = &saved_ss; } if ((flags & (FORK_NEW_MOUNTNS|FORK_NEW_USERNS)) != 0) pid = raw_clone(SIGCHLD| (FLAGS_SET(flags, FORK_NEW_MOUNTNS) ? CLONE_NEWNS : 0) | (FLAGS_SET(flags, FORK_NEW_USERNS) ? CLONE_NEWUSER : 0)); else pid = fork(); if (pid < 0) return log_full_errno(prio, errno, "Failed to fork: %m"); if (pid > 0) { /* We are in the parent process */ log_debug("Successfully forked off '%s' as PID " PID_FMT ".", strna(name), pid); if (flags & FORK_WAIT) { if (block_all) { /* undo everything except SIGCHLD */ ss = saved_ss; assert_se(sigaddset(&ss, SIGCHLD) >= 0); (void) sigprocmask(SIG_SETMASK, &ss, NULL); } r = wait_for_terminate_and_check(name, pid, (flags & FORK_LOG ? WAIT_LOG : 0)); if (r < 0) return r; if (r != EXIT_SUCCESS) /* exit status > 0 should be treated as failure, too */ return -EPROTO; } if (ret_pid) *ret_pid = pid; return 1; } /* We are in the child process */ /* Restore signal mask manually */ saved_ssp = NULL; if (flags & FORK_REOPEN_LOG) { /* Close the logs if requested, before we log anything. And make sure we reopen it if needed. */ log_close(); log_set_open_when_needed(true); } if (name) { r = rename_process(name); if (r < 0) log_full_errno(flags & FORK_LOG ? LOG_WARNING : LOG_DEBUG, r, "Failed to rename process, ignoring: %m"); } if (flags & (FORK_DEATHSIG|FORK_DEATHSIG_SIGINT)) if (prctl(PR_SET_PDEATHSIG, (flags & FORK_DEATHSIG_SIGINT) ? SIGINT : SIGTERM) < 0) { log_full_errno(prio, errno, "Failed to set death signal: %m"); _exit(EXIT_FAILURE); } if (flags & FORK_RESET_SIGNALS) { r = reset_all_signal_handlers(); if (r < 0) { log_full_errno(prio, r, "Failed to reset signal handlers: %m"); _exit(EXIT_FAILURE); } /* This implicitly undoes the signal mask stuff we did before the fork()ing above */ r = reset_signal_mask(); if (r < 0) { log_full_errno(prio, r, "Failed to reset signal mask: %m"); _exit(EXIT_FAILURE); } } else if (block_signals) { /* undo what we did above */ if (sigprocmask(SIG_SETMASK, &saved_ss, NULL) < 0) { log_full_errno(prio, errno, "Failed to restore signal mask: %m"); _exit(EXIT_FAILURE); } } if (flags & FORK_DEATHSIG) { pid_t ppid; /* Let's see if the parent PID is still the one we started from? If not, then the parent * already died by the time we set PR_SET_PDEATHSIG, hence let's emulate the effect */ ppid = getppid(); if (ppid == 0) /* Parent is in a different PID namespace. */; else if (ppid != original_pid) { log_debug("Parent died early, raising SIGTERM."); (void) raise(SIGTERM); _exit(EXIT_FAILURE); } } if (FLAGS_SET(flags, FORK_NEW_MOUNTNS | FORK_MOUNTNS_SLAVE)) { /* Optionally, make sure we never propagate mounts to the host. */ if (mount(NULL, "/", NULL, MS_SLAVE | MS_REC, NULL) < 0) { log_full_errno(prio, errno, "Failed to remount root directory as MS_SLAVE: %m"); _exit(EXIT_FAILURE); } } if (flags & FORK_CLOSE_ALL_FDS) { /* Close the logs here in case it got reopened above, as close_all_fds() would close them for us */ log_close(); r = close_all_fds(except_fds, n_except_fds); if (r < 0) { log_full_errno(prio, r, "Failed to close all file descriptors: %m"); _exit(EXIT_FAILURE); } } /* When we were asked to reopen the logs, do so again now */ if (flags & FORK_REOPEN_LOG) { log_open(); log_set_open_when_needed(false); } if (flags & FORK_NULL_STDIO) { r = make_null_stdio(); if (r < 0) { log_full_errno(prio, r, "Failed to connect stdin/stdout to /dev/null: %m"); _exit(EXIT_FAILURE); } } else if (flags & FORK_STDOUT_TO_STDERR) { if (dup2(STDERR_FILENO, STDOUT_FILENO) < 0) { log_full_errno(prio, errno, "Failed to connect stdout to stderr: %m"); _exit(EXIT_FAILURE); } } if (flags & FORK_RLIMIT_NOFILE_SAFE) { r = rlimit_nofile_safe(); if (r < 0) { log_full_errno(prio, r, "Failed to lower RLIMIT_NOFILE's soft limit to 1K: %m"); _exit(EXIT_FAILURE); } } if (ret_pid) *ret_pid = getpid_cached(); return 0; } int namespace_fork( const char *outer_name, const char *inner_name, const int except_fds[], size_t n_except_fds, ForkFlags flags, int pidns_fd, int mntns_fd, int netns_fd, int userns_fd, int root_fd, pid_t *ret_pid) { int r; /* This is much like safe_fork(), but forks twice, and joins the specified namespaces in the middle * process. This ensures that we are fully a member of the destination namespace, with pidns an all, so that * /proc/self/fd works correctly. */ r = safe_fork_full(outer_name, except_fds, n_except_fds, (flags|FORK_DEATHSIG) & ~(FORK_REOPEN_LOG|FORK_NEW_MOUNTNS|FORK_MOUNTNS_SLAVE), ret_pid); if (r < 0) return r; if (r == 0) { pid_t pid; /* Child */ r = namespace_enter(pidns_fd, mntns_fd, netns_fd, userns_fd, root_fd); if (r < 0) { log_full_errno(FLAGS_SET(flags, FORK_LOG) ? LOG_ERR : LOG_DEBUG, r, "Failed to join namespace: %m"); _exit(EXIT_FAILURE); } /* We mask a few flags here that either make no sense for the grandchild, or that we don't have to do again */ r = safe_fork_full(inner_name, except_fds, n_except_fds, flags & ~(FORK_WAIT|FORK_RESET_SIGNALS|FORK_CLOSE_ALL_FDS|FORK_NULL_STDIO), &pid); if (r < 0) _exit(EXIT_FAILURE); if (r == 0) { /* Child */ if (ret_pid) *ret_pid = pid; return 0; } r = wait_for_terminate_and_check(inner_name, pid, FLAGS_SET(flags, FORK_LOG) ? WAIT_LOG : 0); if (r < 0) _exit(EXIT_FAILURE); _exit(r); } return 1; } int set_oom_score_adjust(int value) { char t[DECIMAL_STR_MAX(int)]; xsprintf(t, "%i", value); return write_string_file("/proc/self/oom_score_adj", t, WRITE_STRING_FILE_VERIFY_ON_FAILURE|WRITE_STRING_FILE_DISABLE_BUFFER); } int get_oom_score_adjust(int *ret) { _cleanup_free_ char *t = NULL; int r, a; r = read_virtual_file("/proc/self/oom_score_adj", SIZE_MAX, &t, NULL); if (r < 0) return r; delete_trailing_chars(t, WHITESPACE); assert_se(safe_atoi(t, &a) >= 0); assert_se(oom_score_adjust_is_valid(a)); if (ret) *ret = a; return 0; } int pidfd_get_pid(int fd, pid_t *ret) { char path[STRLEN("/proc/self/fdinfo/") + DECIMAL_STR_MAX(int)]; _cleanup_free_ char *fdinfo = NULL; char *p; int r; if (fd < 0) return -EBADF; xsprintf(path, "/proc/self/fdinfo/%i", fd); r = read_full_virtual_file(path, &fdinfo, NULL); if (r == -ENOENT) /* if fdinfo doesn't exist we assume the process does not exist */ return -ESRCH; if (r < 0) return r; p = startswith(fdinfo, "Pid:"); if (!p) { p = strstr(fdinfo, "\nPid:"); if (!p) return -ENOTTY; /* not a pidfd? */ p += 5; } p += strspn(p, WHITESPACE); p[strcspn(p, WHITESPACE)] = 0; return parse_pid(p, ret); } static int rlimit_to_nice(rlim_t limit) { if (limit <= 1) return PRIO_MAX-1; /* i.e. 19 */ if (limit >= -PRIO_MIN + PRIO_MAX) return PRIO_MIN; /* i.e. -20 */ return PRIO_MAX - (int) limit; } int setpriority_closest(int priority) { int current, limit, saved_errno; struct rlimit highest; /* Try to set requested nice level */ if (setpriority(PRIO_PROCESS, 0, priority) >= 0) return 1; /* Permission failed */ saved_errno = -errno; if (!ERRNO_IS_PRIVILEGE(saved_errno)) return saved_errno; errno = 0; current = getpriority(PRIO_PROCESS, 0); if (errno != 0) return -errno; if (priority == current) return 1; /* Hmm, we'd expect that raising the nice level from our status quo would always work. If it doesn't, * then the whole setpriority() system call is blocked to us, hence let's propagate the error * right-away */ if (priority > current) return saved_errno; if (getrlimit(RLIMIT_NICE, &highest) < 0) return -errno; limit = rlimit_to_nice(highest.rlim_cur); /* We are already less nice than limit allows us */ if (current < limit) { log_debug("Cannot raise nice level, permissions and the resource limit do not allow it."); return 0; } /* Push to the allowed limit */ if (setpriority(PRIO_PROCESS, 0, limit) < 0) return -errno; log_debug("Cannot set requested nice level (%i), used next best (%i).", priority, limit); return 0; } bool invoked_as(char *argv[], const char *token) { if (!argv || isempty(argv[0])) return false; if (isempty(token)) return false; return strstr(last_path_component(argv[0]), token); } _noreturn_ void freeze(void) { log_close(); /* Make sure nobody waits for us (i.e. on one of our sockets) anymore. Note that we use * close_all_fds_without_malloc() instead of plain close_all_fds() here, since we want this function * to be compatible with being called from signal handlers. */ (void) close_all_fds_without_malloc(NULL, 0); /* Let's not freeze right away, but keep reaping zombies. */ for (;;) { siginfo_t si = {}; if (waitid(P_ALL, 0, &si, WEXITED) < 0 && errno != EINTR) break; } /* waitid() failed with an unexpected error, things are really borked. Freeze now! */ for (;;) pause(); } bool argv_looks_like_help(int argc, char **argv) { char **l; /* Scans the command line for indications the user asks for help. This is supposed to be called by * tools that do not implement getopt() style command line parsing because they are not primarily * user-facing. Detects four ways of asking for help: * * 1. Passing zero arguments * 2. Passing "help" as first argument * 3. Passing --help as any argument * 4. Passing -h as any argument */ if (argc <= 1) return true; if (streq_ptr(argv[1], "help")) return true; l = strv_skip(argv, 1); return strv_contains(l, "--help") || strv_contains(l, "-h"); } static const char *const sigchld_code_table[] = { [CLD_EXITED] = "exited", [CLD_KILLED] = "killed", [CLD_DUMPED] = "dumped", [CLD_TRAPPED] = "trapped", [CLD_STOPPED] = "stopped", [CLD_CONTINUED] = "continued", }; DEFINE_STRING_TABLE_LOOKUP(sigchld_code, int); static const char* const sched_policy_table[] = { [SCHED_OTHER] = "other", [SCHED_BATCH] = "batch", [SCHED_IDLE] = "idle", [SCHED_FIFO] = "fifo", [SCHED_RR] = "rr", }; DEFINE_STRING_TABLE_LOOKUP_WITH_FALLBACK(sched_policy, int, INT_MAX);