posix: Add pidfd_spawn and pidfd_spawnp (BZ 30349)

Returning a pidfd allows a process to keep a race-free handle for a
child process, otherwise, the caller will need to either use pidfd_open
(which still might be subject to TOCTOU) or keep the old racy interface
base on pid_t.

To correct use pifd_spawn, the kernel must support not only returning
the pidfd with clone/clone3 but also waitid (P_PIDFD) (added on Linux
5.4).  If kernel does not support the waitid, pidfd return ENOSYS.
It avoids the need to racy workarounds, such as reading the procfs
fdinfo to get the pid to use along with other wait interfaces.

These interfaces are similar to the posix_spawn and posix_spawnp, with
the only difference being it returns a process file descriptor (int)
instead of a process ID (pid_t).  Their prototypes are:

  int pidfd_spawn (int *restrict pidfd,
                   const char *restrict file,
                   const posix_spawn_file_actions_t *restrict facts,
                   const posix_spawnattr_t *restrict attrp,
                   char *const argv[restrict],
                   char *const envp[restrict])

  int pidfd_spawnp (int *restrict pidfd,
                    const char *restrict path,
                    const posix_spawn_file_actions_t *restrict facts,
                    const posix_spawnattr_t *restrict attrp,
                    char *const argv[restrict_arr],
                    char *const envp[restrict_arr]);

A new symbol is used instead of a posix_spawn extension to avoid
possible issues with language bindings that might track the return
argument lifetime.  Although on Linux pid_t and int are interchangeable,
POSIX only states that pid_t should be a signed integer.

Both symbols reuse the posix_spawn posix_spawn_file_actions_t and
posix_spawnattr_t, to void rehash posix_spawn API or add a new one. It
also means that both interfaces support the same attribute and file
actions, and a new flag or file action on posix_spawn is also added
automatically for pidfd_spawn.

Also, using posix_spawn plumbing allows the reusing of most of the
current testing with some changes:

  - waitid is used instead of waitpid since it is a more generic
    interface.

  - tst-posix_spawn-setsid.c is adapted to take into consideration that
    the caller can check for session id directly.  The test now spawns
itself and writes the session id as a file instead.

  - tst-spawn3.c need to know where pidfd_spawn is used so it keeps an
    extra file description unused.

Checked on x86_64-linux-gnu on Linux 4.15 (no CLONE_PIDFD or waitid
support), Linux 5.4 (full support), and Linux 6.2.
Reviewed-by: Florian Weimer <fweimer@redhat.com>

1 files changed, 125 insertions, 44 deletions

diff --git a/posix/tst-posix_spawn-setsid.c b/posix/tst-posix_spawn-setsid.c
index 124d878ce2..b47eb16cc5 100644
--- a/posix/tst-posix_spawn-setsid.c
+++ b/posix/tst-posix_spawn-setsid.c
@@ -18,78 +18,159 @@
 
 #include <errno.h>
 #include <fcntl.h>
+#include <getopt.h>
+#include <intprops.h>
+#include <paths.h>
 #include <spawn.h>
 #include <stdbool.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <sys/resource.h>
+#include <sys/wait.h>
 #include <unistd.h>
 
 #include <support/check.h>
+#include <support/xunistd.h>
+#include <support/temp_file.h>
+#include <tst-spawn.h>
 
+/* Nonzero if the program gets called via `exec'.  */
+static int restart;
+
+/* Hold the four initial argument used to respawn the process, plus
+   the extra '--direct' and '--restart', and a final NULL.  */
+static char *initial_argv[7];
+static int initial_argv_count;
+
+#define CMDLINE_OPTIONS \
+  { "restart", no_argument, &restart, 1 },
+
+static char *pidfile;
+
+static pid_t
+read_child_sid (void)
+{
+  int pidfd = xopen (pidfile, O_RDONLY, 0);
+
+  char buf[INT_STRLEN_BOUND (pid_t)];
+  ssize_t n = read (pidfd, buf, sizeof (buf));
+  TEST_VERIFY (n < sizeof buf && n >= 0);
+  buf[n] = '\0';
+
+  /* We only expect to read the PID.  */
+  char *endp;
+  long int rpid = strtol (buf, &endp, 10);
+  TEST_VERIFY (endp != buf);
+
+  xclose (pidfd);
+
+  return rpid;
+}
+
+/* Called on process re-execution, write down the session id on PIDFILE.  */
 static void
-do_test_setsid (bool test_setsid)
+handle_restart (const char *pidfile)
 {
-  pid_t sid, child_sid;
-  int res;
+  int pidfd = xopen (pidfile, O_WRONLY, 0);
+
+  char buf[INT_STRLEN_BOUND (pid_t)];
+  int s = snprintf (buf, sizeof buf, "%d", getsid (0));
+  size_t n = write (pidfd, buf, s);
+  TEST_VERIFY (n == s);
+
+  xclose (pidfd);
+}
 
+static void
+do_test_setsid (bool test_setsid)
+{
   /* Current session ID.  */
-  sid = getsid(0);
-  if (sid == (pid_t) -1)
-    FAIL_EXIT1 ("getsid (0): %m");
+  pid_t sid = getsid (0);
+  TEST_VERIFY (sid != (pid_t) -1);
 
   posix_spawnattr_t attrp;
-  /* posix_spawnattr_init should not fail (it basically memset the
-     attribute).  */
-  posix_spawnattr_init (&attrp);
+  TEST_COMPARE (posix_spawnattr_init (&attrp), 0);
   if (test_setsid)
-    {
-      res = posix_spawnattr_setflags (&attrp, POSIX_SPAWN_SETSID);
-      if (res != 0)
-	{
-	  errno = res;
-	  FAIL_EXIT1 ("posix_spawnattr_setflags: %m");
-	}
-    }
-
-  /* Program to run.  */
-  char *args[2] = { (char *) "true", NULL };
-  pid_t child;
-
-  res = posix_spawnp (&child, "true", NULL, &attrp, args, environ);
-  /* posix_spawnattr_destroy is noop.  */
-  posix_spawnattr_destroy (&attrp);
-
-  if (res != 0)
-    {
-      errno = res;
-      FAIL_EXIT1 ("posix_spawnp: %m");
-    }
+    TEST_COMPARE (posix_spawnattr_setflags (&attrp, POSIX_SPAWN_SETSID), 0);
+
+  /* 1 or 4 elements from initial_argv:
+       + path to ld.so          optional
+       + --library-path         optional
+       + the library path       optional
+       + application name
+       + --direct
+       + --restart
+       + pidfile  */
+  int argv_size = initial_argv_count + 2;
+  char *args[argv_size];
+  int argc = 0;
+
+  for (char **arg = initial_argv; *arg != NULL; arg++)
+    args[argc++] = *arg;
+  args[argc++] = pidfile;
+  args[argc] = NULL;
+  TEST_VERIFY (argc < argv_size);
+
+  PID_T_TYPE pid;
+  TEST_COMPARE (POSIX_SPAWN (&pid, args[0], NULL, &attrp, args, environ), 0);
+  TEST_COMPARE (posix_spawnattr_destroy (&attrp), 0);
+
+  siginfo_t sinfo;
+  TEST_COMPARE (WAITID (P_PID, pid, &sinfo, WEXITED), 0);
+  TEST_COMPARE (sinfo.si_code, CLD_EXITED);
+  TEST_COMPARE (sinfo.si_status, 0);
+
+  pid_t child_sid = read_child_sid ();
 
   /* Child should have a different session ID than parent.  */
-  child_sid = getsid (child);
-
-  if (child_sid == (pid_t) -1)
-    FAIL_EXIT1 ("getsid (%i): %m", child);
+  TEST_VERIFY (child_sid != (pid_t) -1);
 
   if (test_setsid)
-    {
-      if (child_sid == sid)
-	FAIL_EXIT1 ("child session ID matched parent one");
-    }
+    TEST_VERIFY (child_sid != sid);
   else
-    {
-      if (child_sid != sid)
-	FAIL_EXIT1 ("child session ID did not match parent one");
-    }
+    TEST_VERIFY (child_sid == sid);
 }
 
 static int
-do_test (void)
+do_test (int argc, char *argv[])
 {
+  /* We must have either:
+
+     - one or four parameters if called initially:
+       + argv[1]: path for ld.so        optional
+       + argv[2]: "--library-path"      optional
+       + argv[3]: the library path      optional
+       + argv[4]: the application name
+
+     - six parameters left if called through re-execution:
+       + argv[5/1]: the application name
+       + argv[6/2]: the pidfile
+
+     * When built with --enable-hardcoded-path-in-tests or issued without
+       using the loader directly.  */
+
+  if (restart)
+    {
+      handle_restart (argv[1]);
+      return 0;
+    }
+
+  TEST_VERIFY_EXIT (argc == 2 || argc == 5);
+
+  int i;
+  for (i = 0; i < argc - 1; i++)
+    initial_argv[i] = argv[i + 1];
+  initial_argv[i++] = (char *) "--direct";
+  initial_argv[i++] = (char *) "--restart";
+  initial_argv_count = i;
+
+  create_temp_file ("tst-posix_spawn-setsid-", &pidfile);
+
   do_test_setsid (false);
   do_test_setsid (true);
 
   return 0;
 }
 
+#define TEST_FUNCTION_ARGV do_test
 #include <support/test-driver.c>