glibc/sysdeps/unix/sysv/linux/tst-process_madvise.c
Stefan Liebler 11f09947f3 tst-process_madvise: Check process_madvise-syscall support.
So far this test checks if pidfd_open-syscall is supported,
which was introduced with linux 5.3.

The process_madvise-syscall was introduced with linux 5.10.
Thus you'll get FAILs if you are running a kernel in between.

This patch adds a check if the first process_madvise-syscall
returns ENOSYS and in this case will fail with UNSUPPORTED.
Reviewed-by: Florian Weimer <fweimer@redhat.com>
2022-08-11 12:21:05 +02:00

145 lines
4 KiB
C

/* Basic tests for Linux process_madvise.
Copyright (C) 2022 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<https://www.gnu.org/licenses/>. */
#include <array_length.h>
#include <errno.h>
#include <limits.h>
#include <stdlib.h>
#include <support/check.h>
#include <support/process_state.h>
#include <support/support.h>
#include <support/xsocket.h>
#include <support/xunistd.h>
#include <sys/mman.h>
#include <sys/pidfd.h>
#include <sys/wait.h>
/* The pair of sockets used for coordination. The subprocess uses
sockets[1]. */
static int sockets[2];
static long int page_size;
static void
exit_subprocess (int dummy)
{
exit (EXIT_FAILURE);
}
static void
subprocess (void)
{
/* In case something goes wrong with parent before pidfd_send_signal. */
support_create_timer (5, 0, false, exit_subprocess);
void *p1 = xmmap (NULL, page_size * 2, PROT_READ | PROT_WRITE,
MAP_PRIVATE | MAP_ANONYMOUS, -1);
void *p2 = xmmap (NULL, page_size, PROT_READ | PROT_WRITE,
MAP_PRIVATE | MAP_ANONYMOUS, -1);
xmunmap(p2, page_size);
xsendto (sockets[1], &(struct iovec) { p1, page_size * 2 },
sizeof (struct iovec), 0, NULL, 0);
xsendto (sockets[1], &(struct iovec) { p2, page_size },
sizeof (struct iovec), 0, NULL, 0);
pause ();
_exit (0);
}
static int
do_test (void)
{
page_size = sysconf (_SC_PAGE_SIZE);
{
int r = pidfd_open (-1, 0);
TEST_COMPARE (r, -1);
if (errno == ENOSYS)
FAIL_UNSUPPORTED ("kernel does not support pidfd_open, skipping test");
TEST_COMPARE (errno, EINVAL);
}
TEST_COMPARE (socketpair (AF_UNIX, SOCK_STREAM, 0, sockets), 0);
pid_t pid = xfork ();
if (pid == 0)
{
xclose (sockets[0]);
subprocess ();
}
xclose (sockets[1]);
int pidfd = pidfd_open (pid, 0);
TEST_VERIFY (pidfd != -1);
/* The target process is going to send us two iovec's. The first one points
to a valid mapping, the other points to a previously valid mapping which
has now been unmapped. */
{
struct iovec iv;
xrecvfrom (sockets[0], &iv, sizeof (iv), 0, NULL, 0);
/* We expect this to succeed in the target process because the mapping
is valid. */
ssize_t ret = process_madvise (pidfd, &iv, 1, MADV_COLD, 0);
if (ret == -1 && errno == ENOSYS)
FAIL_UNSUPPORTED ("kernel does not support process_madvise, skipping"
"test");
TEST_COMPARE (ret, 2 * page_size);
}
{
struct iovec iv;
xrecvfrom (sockets[0], &iv, sizeof (iv), 0, NULL, 0);
/* We expect this to fail in the target process because the second iovec
points to an unmapped region. The target process arranges for this to
be the case. */
TEST_COMPARE (process_madvise (pidfd, &iv, 1, MADV_COLD, 0), -1);
TEST_COMPARE (errno, ENOMEM);
}
{
struct iovec iv[IOV_MAX + 1];
TEST_COMPARE (process_madvise (pidfd, iv, array_length (iv), MADV_COLD,
0), -1);
TEST_COMPARE (errno, EINVAL);
}
TEST_COMPARE (pidfd_send_signal (pidfd, SIGKILL, NULL, 0), 0);
{
siginfo_t info;
int r = waitid (P_PIDFD, pidfd, &info, WEXITED);
TEST_COMPARE (r, 0);
TEST_COMPARE (info.si_status, SIGKILL);
TEST_COMPARE (info.si_code, CLD_KILLED);
}
TEST_COMPARE (pidfd_send_signal (pidfd, SIGKILL, NULL, 0), -1);
TEST_COMPARE (errno, ESRCH);
return 0;
}
#include <support/test-driver.c>