freebsd-skq/tools/regression/posixsem/posixsem.c

1442 lines
27 KiB
C
Raw Normal View History

Rework the lifetime management of the kernel implementation of POSIX semaphores. Specifically, semaphores are now represented as new file descriptor type that is set to close on exec. This removes the need for all of the manual process reference counting (and fork, exec, and exit event handlers) as the normal file descriptor operations handle all of that for us nicely. It is also suggested as one possible implementation in the spec and at least one other OS (OS X) uses this approach. Some bugs that were fixed as a result include: - References to a named semaphore whose name is removed still work after the sem_unlink() operation. Prior to this patch, if a semaphore's name was removed, valid handles from sem_open() would get EINVAL errors from sem_getvalue(), sem_post(), etc. This fixes that. - Unnamed semaphores created with sem_init() were not cleaned up when a process exited or exec'd. They were only cleaned up if the process did an explicit sem_destroy(). This could result in a leak of semaphore objects that could never be cleaned up. - On the other hand, if another process guessed the id (kernel pointer to 'struct ksem' of an unnamed semaphore (created via sem_init)) and had write access to the semaphore based on UID/GID checks, then that other process could manipulate the semaphore via sem_destroy(), sem_post(), sem_wait(), etc. - As part of the permission check (UID/GID), the umask of the proces creating the semaphore was not honored. Thus if your umask denied group read/write access but the explicit mode in the sem_init() call allowed it, the semaphore would be readable/writable by other users in the same group, for example. This includes access via the previous bug. - If the module refused to unload because there were active semaphores, then it might have deregistered one or more of the semaphore system calls before it noticed that there was a problem. I'm not sure if this actually happened as the order that modules are discovered by the kernel linker depends on how the actual .ko file is linked. One can make the order deterministic by using a single module with a mod_event handler that explicitly registers syscalls (and deregisters during unload after any checks). This also fixes a race where even if the sem_module unloaded first it would have destroyed locks that the syscalls might be trying to access if they are still executing when they are unloaded. XXX: By the way, deregistering system calls doesn't do any blocking to drain any threads from the calls. - Some minor fixes to errno values on error. For example, sem_init() isn't documented to return ENFILE or EMFILE if we run out of semaphores the way that sem_open() can. Instead, it should return ENOSPC in that case. Other changes: - Kernel semaphores now use a hash table to manage the namespace of named semaphores nearly in a similar fashion to the POSIX shared memory object file descriptors. Kernel semaphores can now also have names longer than 14 chars (up to MAXPATHLEN) and can include subdirectories in their pathname. - The UID/GID permission checks for access to a named semaphore are now done via vaccess() rather than a home-rolled set of checks. - Now that kernel semaphores have an associated file object, the various MAC checks for POSIX semaphores accept both a file credential and an active credential. There is also a new posixsem_check_stat() since it is possible to fstat() a semaphore file descriptor. - A small set of regression tests (using the ksem API directly) is present in src/tools/regression/posixsem. Reported by: kris (1) Tested by: kris Reviewed by: rwatson (lightly) MFC after: 1 month
2008-06-27 05:39:04 +00:00
/*-
* Copyright (c) 2008 Yahoo!, Inc.
* All rights reserved.
* Written by: John Baldwin <jhb@FreeBSD.org>
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/queue.h>
#include <sys/_semaphore.h>
#include <sys/sysctl.h>
#include <sys/time.h>
#include <sys/user.h>
#include <sys/wait.h>
#include <errno.h>
#include <fcntl.h>
#include <kvm.h>
#include <limits.h>
#include <semaphore.h>
#include <signal.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>
#include "test.h"
/* Cut and pasted from kernel header, bah! */
/* Operations on timespecs */
#define timespecclear(tvp) ((tvp)->tv_sec = (tvp)->tv_nsec = 0)
#define timespecisset(tvp) ((tvp)->tv_sec || (tvp)->tv_nsec)
#define timespeccmp(tvp, uvp, cmp) \
(((tvp)->tv_sec == (uvp)->tv_sec) ? \
((tvp)->tv_nsec cmp (uvp)->tv_nsec) : \
((tvp)->tv_sec cmp (uvp)->tv_sec))
#define timespecadd(vvp, uvp) \
do { \
(vvp)->tv_sec += (uvp)->tv_sec; \
(vvp)->tv_nsec += (uvp)->tv_nsec; \
if ((vvp)->tv_nsec >= 1000000000) { \
(vvp)->tv_sec++; \
(vvp)->tv_nsec -= 1000000000; \
} \
} while (0)
#define timespecsub(vvp, uvp) \
do { \
(vvp)->tv_sec -= (uvp)->tv_sec; \
(vvp)->tv_nsec -= (uvp)->tv_nsec; \
if ((vvp)->tv_nsec < 0) { \
(vvp)->tv_sec--; \
(vvp)->tv_nsec += 1000000000; \
} \
} while (0)
#define TEST_PATH "/tmp/posixsem_regression_test"
#define ELAPSED(elapsed, limit) (abs((elapsed) - (limit)) < 100)
/* Macros for passing child status to parent over a pipe. */
#define CSTAT(class, error) ((class) << 16 | (error))
#define CSTAT_CLASS(stat) ((stat) >> 16)
#define CSTAT_ERROR(stat) ((stat) & 0xffff)
/*
* Helper routine for tests that use a child process. This routine
* creates a pipe and forks a child process. The child process runs
* the 'func' routine which returns a status integer. The status
* integer gets written over the pipe to the parent and returned in
* '*stat'. If there is an error in pipe(), fork(), or wait() this
* returns -1 and fails the test.
*/
static int
child_worker(int (*func)(void *arg), void *arg, int *stat)
{
pid_t pid;
int pfd[2], cstat;
if (pipe(pfd) < 0) {
fail_errno("pipe");
return (-1);
}
pid = fork();
switch (pid) {
case -1:
/* Error. */
fail_errno("fork");
close(pfd[0]);
close(pfd[1]);
return (-1);
case 0:
/* Child. */
cstat = func(arg);
write(pfd[1], &cstat, sizeof(cstat));
exit(0);
}
if (read(pfd[0], stat, sizeof(*stat)) < 0) {
fail_errno("read(pipe)");
close(pfd[0]);
close(pfd[1]);
return (-1);
}
if (waitpid(pid, NULL, 0) < 0) {
fail_errno("wait");
close(pfd[0]);
close(pfd[1]);
return (-1);
}
close(pfd[0]);
close(pfd[1]);
return (0);
}
/*
* Attempt a ksem_open() that should fail with an expected error of
* 'error'.
*/
static void
ksem_open_should_fail(const char *path, int flags, mode_t mode, unsigned int
value, int error)
{
semid_t id;
if (ksem_open(&id, path, flags, mode, value) >= 0) {
fail_err("ksem_open() didn't fail");
ksem_close(id);
return;
}
if (errno != error) {
fail_errno("ksem_open");
return;
}
pass();
}
/*
* Attempt a ksem_unlink() that should fail with an expected error of
* 'error'.
*/
static void
ksem_unlink_should_fail(const char *path, int error)
{
if (ksem_unlink(path) >= 0) {
fail_err("ksem_unlink() didn't fail");
return;
}
if (errno != error) {
fail_errno("ksem_unlink");
return;
}
pass();
}
/*
* Attempt a ksem_close() that should fail with an expected error of
* 'error'.
*/
static void
ksem_close_should_fail(semid_t id, int error)
{
if (ksem_close(id) >= 0) {
fail_err("ksem_close() didn't fail");
return;
}
if (errno != error) {
fail_errno("ksem_close");
return;
}
pass();
}
/*
* Attempt a ksem_init() that should fail with an expected error of
* 'error'.
*/
static void
ksem_init_should_fail(unsigned int value, int error)
{
semid_t id;
if (ksem_init(&id, value) >= 0) {
fail_err("ksem_init() didn't fail");
ksem_destroy(id);
return;
}
if (errno != error) {
fail_errno("ksem_init");
return;
}
pass();
}
/*
* Attempt a ksem_destroy() that should fail with an expected error of
* 'error'.
*/
static void
ksem_destroy_should_fail(semid_t id, int error)
{
if (ksem_destroy(id) >= 0) {
fail_err("ksem_destroy() didn't fail");
return;
}
if (errno != error) {
fail_errno("ksem_destroy");
return;
}
pass();
}
/*
* Attempt a ksem_post() that should fail with an expected error of
* 'error'.
*/
static void
ksem_post_should_fail(semid_t id, int error)
{
if (ksem_post(id) >= 0) {
fail_err("ksem_post() didn't fail");
return;
}
if (errno != error) {
fail_errno("ksem_post");
return;
}
pass();
}
static void
open_after_unlink(void)
{
semid_t id;
if (ksem_open(&id, TEST_PATH, O_CREAT, 0777, 1) < 0) {
fail_errno("ksem_open(1)");
return;
}
ksem_close(id);
if (ksem_unlink(TEST_PATH) < 0) {
fail_errno("ksem_unlink");
return;
}
ksem_open_should_fail(TEST_PATH, O_RDONLY, 0777, 1, ENOENT);
}
TEST(open_after_unlink, "open after unlink");
static void
open_invalid_path(void)
{
ksem_open_should_fail("blah", 0, 0777, 1, EINVAL);
}
TEST(open_invalid_path, "open invalid path");
static void
open_extra_flags(void)
{
ksem_open_should_fail(TEST_PATH, O_RDONLY | O_DIRECT, 0777, 1, EINVAL);
}
TEST(open_extra_flags, "open with extra flags");
static void
open_bad_value(void)
{
(void)ksem_unlink(TEST_PATH);
ksem_open_should_fail(TEST_PATH, O_CREAT, 0777, UINT_MAX, EINVAL);
}
TEST(open_bad_value, "open with invalid initial value");
static void
open_bad_path_pointer(void)
{
ksem_open_should_fail((char *)1024, O_RDONLY, 0777, 1, EFAULT);
}
TEST(open_bad_path_pointer, "open bad path pointer");
static void
open_path_too_long(void)
{
char *page;
page = malloc(MAXPATHLEN + 1);
memset(page, 'a', MAXPATHLEN);
page[MAXPATHLEN] = '\0';
ksem_open_should_fail(page, O_RDONLY, 0777, 1, ENAMETOOLONG);
free(page);
}
TEST(open_path_too_long, "open pathname too long");
static void
open_nonexisting_semaphore(void)
{
ksem_open_should_fail("/notreallythere", 0, 0777, 1, ENOENT);
}
TEST(open_nonexisting_semaphore, "open nonexistent semaphore");
static void
exclusive_create_existing_semaphore(void)
{
semid_t id;
if (ksem_open(&id, TEST_PATH, O_CREAT, 0777, 1) < 0) {
fail_errno("ksem_open(O_CREAT)");
return;
}
ksem_close(id);
ksem_open_should_fail(TEST_PATH, O_CREAT | O_EXCL, 0777, 1, EEXIST);
ksem_unlink(TEST_PATH);
}
TEST(exclusive_create_existing_semaphore, "O_EXCL of existing semaphore");
static void
init_bad_value(void)
{
ksem_init_should_fail(UINT_MAX, EINVAL);
}
TEST(init_bad_value, "init with invalid initial value");
static void
unlink_bad_path_pointer(void)
{
ksem_unlink_should_fail((char *)1024, EFAULT);
}
TEST(unlink_bad_path_pointer, "unlink bad path pointer");
static void
unlink_path_too_long(void)
{
char *page;
page = malloc(MAXPATHLEN + 1);
memset(page, 'a', MAXPATHLEN);
page[MAXPATHLEN] = '\0';
ksem_unlink_should_fail(page, ENAMETOOLONG);
free(page);
}
TEST(unlink_path_too_long, "unlink pathname too long");
static void
destroy_named_semaphore(void)
{
semid_t id;
if (ksem_open(&id, TEST_PATH, O_CREAT, 0777, 1) < 0) {
fail_errno("ksem_open(O_CREAT)");
return;
}
ksem_destroy_should_fail(id, EINVAL);
ksem_close(id);
ksem_unlink(TEST_PATH);
}
TEST(destroy_named_semaphore, "destroy named semaphore");
static void
close_unnamed_semaphore(void)
{
semid_t id;
if (ksem_init(&id, 1) < 0) {
fail_errno("ksem_init");
return;
}
ksem_close_should_fail(id, EINVAL);
ksem_destroy(id);
}
TEST(close_unnamed_semaphore, "close unnamed semaphore");
static void
destroy_invalid_fd(void)
{
ksem_destroy_should_fail(STDERR_FILENO, EINVAL);
}
TEST(destroy_invalid_fd, "destroy non-semaphore file descriptor");
static void
close_invalid_fd(void)
{
ksem_close_should_fail(STDERR_FILENO, EINVAL);
}
TEST(close_invalid_fd, "close non-semaphore file descriptor");
static void
create_unnamed_semaphore(void)
{
semid_t id;
if (ksem_init(&id, 1) < 0) {
fail_errno("ksem_init");
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(create_unnamed_semaphore, "create unnamed semaphore");
static void
open_named_semaphore(void)
{
semid_t id;
if (ksem_open(&id, TEST_PATH, O_CREAT, 0777, 1) < 0) {
fail_errno("ksem_open(O_CREAT)");
return;
}
if (ksem_close(id) < 0) {
fail_errno("ksem_close");
return;
}
if (ksem_unlink(TEST_PATH) < 0) {
fail_errno("ksem_unlink");
return;
}
pass();
}
TEST(open_named_semaphore, "create named semaphore");
static void
getvalue_invalid_semaphore(void)
{
int val;
if (ksem_getvalue(STDERR_FILENO, &val) >= 0) {
fail_err("ksem_getvalue() didn't fail");
return;
}
if (errno != EINVAL) {
fail_errno("ksem_getvalue");
return;
}
pass();
}
TEST(getvalue_invalid_semaphore, "get value of invalid semaphore");
static void
post_invalid_semaphore(void)
{
ksem_post_should_fail(STDERR_FILENO, EINVAL);
}
TEST(post_invalid_semaphore, "post of invalid semaphore");
static void
wait_invalid_semaphore(void)
{
if (ksem_wait(STDERR_FILENO) >= 0) {
fail_err("ksem_wait() didn't fail");
return;
}
if (errno != EINVAL) {
fail_errno("ksem_wait");
return;
}
pass();
}
TEST(wait_invalid_semaphore, "wait for invalid semaphore");
static void
trywait_invalid_semaphore(void)
{
if (ksem_trywait(STDERR_FILENO) >= 0) {
fail_err("ksem_trywait() didn't fail");
return;
}
if (errno != EINVAL) {
fail_errno("ksem_trywait");
return;
}
pass();
}
TEST(trywait_invalid_semaphore, "try wait for invalid semaphore");
static void
timedwait_invalid_semaphore(void)
{
if (ksem_timedwait(STDERR_FILENO, NULL) >= 0) {
fail_err("ksem_timedwait() didn't fail");
return;
}
if (errno != EINVAL) {
fail_errno("ksem_timedwait");
return;
}
pass();
}
TEST(timedwait_invalid_semaphore, "timed wait for invalid semaphore");
static int
checkvalue(semid_t id, int expected)
{
int val;
if (ksem_getvalue(id, &val) < 0) {
fail_errno("ksem_getvalue");
return (-1);
}
if (val != expected) {
fail_err("sem value should be %d instead of %d", expected, val);
return (-1);
}
return (0);
}
static void
post_test(void)
{
semid_t id;
if (ksem_init(&id, 1) < 0) {
fail_errno("ksem_init");
return;
}
if (checkvalue(id, 1) < 0) {
ksem_destroy(id);
return;
}
if (ksem_post(id) < 0) {
fail_errno("ksem_post");
ksem_destroy(id);
return;
}
if (checkvalue(id, 2) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(post_test, "simple post");
static void
use_after_unlink_test(void)
{
semid_t id;
/*
* Create named semaphore with value of 1 and then unlink it
* while still retaining the initial reference.
*/
if (ksem_open(&id, TEST_PATH, O_CREAT | O_EXCL, 0777, 1) < 0) {
fail_errno("ksem_open(O_CREAT | O_EXCL)");
return;
}
if (ksem_unlink(TEST_PATH) < 0) {
fail_errno("ksem_unlink");
ksem_close(id);
return;
}
if (checkvalue(id, 1) < 0) {
ksem_close(id);
return;
}
/* Post the semaphore to set its value to 2. */
if (ksem_post(id) < 0) {
fail_errno("ksem_post");
ksem_close(id);
return;
}
if (checkvalue(id, 2) < 0) {
ksem_close(id);
return;
}
/* Wait on the semaphore which should set its value to 1. */
if (ksem_wait(id) < 0) {
fail_errno("ksem_wait");
ksem_close(id);
return;
}
if (checkvalue(id, 1) < 0) {
ksem_close(id);
return;
}
if (ksem_close(id) < 0) {
fail_errno("ksem_close");
return;
}
pass();
}
TEST(use_after_unlink_test, "use named semaphore after unlink");
static void
unlocked_trywait(void)
{
semid_t id;
if (ksem_init(&id, 1) < 0) {
fail_errno("ksem_init");
return;
}
/* This should succeed and decrement the value to 0. */
if (ksem_trywait(id) < 0) {
fail_errno("ksem_trywait()");
ksem_destroy(id);
return;
}
if (checkvalue(id, 0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(unlocked_trywait, "unlocked trywait");
static void
locked_trywait(void)
{
semid_t id;
if (ksem_init(&id, 0) < 0) {
fail_errno("ksem_init");
return;
}
/* This should fail with EAGAIN and leave the value at 0. */
if (ksem_trywait(id) >= 0) {
fail_err("ksem_trywait() didn't fail");
ksem_destroy(id);
return;
}
if (errno != EAGAIN) {
fail_errno("wrong error from ksem_trywait()");
ksem_destroy(id);
return;
}
if (checkvalue(id, 0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(locked_trywait, "locked trywait");
/*
* Use a timer to post a specific semaphore after a timeout. A timer
* is scheduled via schedule_post(). check_alarm() must be called
* afterwards to clean up and check for errors.
*/
static semid_t alarm_id = -1;
static int alarm_errno;
static int alarm_handler_installed;
static void
alarm_handler(int signo)
{
if (ksem_post(alarm_id) < 0)
alarm_errno = errno;
}
static int
check_alarm(int just_clear)
{
struct itimerval it;
bzero(&it, sizeof(it));
if (just_clear) {
setitimer(ITIMER_REAL, &it, NULL);
alarm_errno = 0;
alarm_id = -1;
return (0);
}
if (setitimer(ITIMER_REAL, &it, NULL) < 0) {
fail_errno("setitimer");
return (-1);
}
if (alarm_errno != 0 && !just_clear) {
errno = alarm_errno;
fail_errno("ksem_post() (via timeout)");
alarm_errno = 0;
return (-1);
}
alarm_id = -1;
return (0);
}
static int
schedule_post(semid_t id, u_int msec)
{
struct itimerval it;
if (!alarm_handler_installed) {
if (signal(SIGALRM, alarm_handler) == SIG_ERR) {
fail_errno("signal(SIGALRM)");
return (-1);
}
alarm_handler_installed = 1;
}
if (alarm_id != -1) {
fail_err("ksem_post() already scheduled");
return (-1);
}
alarm_id = id;
bzero(&it, sizeof(it));
it.it_value.tv_sec = msec / 1000;
it.it_value.tv_usec = (msec % 1000) * 1000;
if (setitimer(ITIMER_REAL, &it, NULL) < 0) {
fail_errno("setitimer");
return (-1);
}
return (0);
}
static int
timedwait(semid_t id, u_int msec, u_int *delta, int error)
{
struct timespec start, end;
if (clock_gettime(CLOCK_REALTIME, &start) < 0) {
fail_errno("clock_gettime(CLOCK_REALTIME)");
return (-1);
}
end.tv_sec = msec / 1000;
end.tv_nsec = msec % 1000 * 1000000;
timespecadd(&end, &start);
if (ksem_timedwait(id, &end) < 0) {
if (errno != error) {
fail_errno("ksem_timedwait");
return (-1);
}
} else if (error != 0) {
fail_err("ksem_timedwait() didn't fail");
return (-1);
}
if (clock_gettime(CLOCK_REALTIME, &end) < 0) {
fail_errno("clock_gettime(CLOCK_REALTIME)");
return (-1);
}
timespecsub(&end, &start);
*delta = end.tv_nsec / 1000000;
*delta += end.tv_sec * 1000;
return (0);
}
static void
unlocked_timedwait(void)
{
semid_t id;
u_int elapsed;
if (ksem_init(&id, 1) < 0) {
fail_errno("ksem_init");
return;
}
/* This should succeed right away and set the value to 0. */
if (timedwait(id, 5000, &elapsed, 0) < 0) {
ksem_destroy(id);
return;
}
if (!ELAPSED(elapsed, 0)) {
fail_err("ksem_timedwait() of unlocked sem took %ums", elapsed);
ksem_destroy(id);
return;
}
if (checkvalue(id, 0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(unlocked_timedwait, "unlocked timedwait");
static void
expired_timedwait(void)
{
semid_t id;
u_int elapsed;
if (ksem_init(&id, 0) < 0) {
fail_errno("ksem_init");
return;
}
/* This should fail with a timeout and leave the value at 0. */
if (timedwait(id, 2500, &elapsed, ETIMEDOUT) < 0) {
ksem_destroy(id);
return;
}
if (!ELAPSED(elapsed, 2500)) {
fail_err(
"ksem_timedwait() of locked sem took %ums instead of 2500ms",
elapsed);
ksem_destroy(id);
return;
}
if (checkvalue(id, 0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(expired_timedwait, "locked timedwait timeout");
static void
locked_timedwait(void)
{
semid_t id;
u_int elapsed;
if (ksem_init(&id, 0) < 0) {
fail_errno("ksem_init");
return;
}
/*
* Schedule a post to trigger after 1000 ms. The subsequent
* timedwait should succeed after 1000 ms as a result w/o
* timing out.
*/
if (schedule_post(id, 1000) < 0) {
ksem_destroy(id);
return;
}
if (timedwait(id, 2000, &elapsed, 0) < 0) {
check_alarm(1);
ksem_destroy(id);
return;
}
if (!ELAPSED(elapsed, 1000)) {
fail_err(
"ksem_timedwait() with delayed post took %ums instead of 1000ms",
elapsed);
check_alarm(1);
ksem_destroy(id);
return;
}
if (check_alarm(0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(locked_timedwait, "locked timedwait");
static int
testwait(semid_t id, u_int *delta)
{
struct timespec start, end;
if (clock_gettime(CLOCK_REALTIME, &start) < 0) {
fail_errno("clock_gettime(CLOCK_REALTIME)");
return (-1);
}
if (ksem_wait(id) < 0) {
fail_errno("ksem_wait");
return (-1);
}
if (clock_gettime(CLOCK_REALTIME, &end) < 0) {
fail_errno("clock_gettime(CLOCK_REALTIME)");
return (-1);
}
timespecsub(&end, &start);
*delta = end.tv_nsec / 1000000;
*delta += end.tv_sec * 1000;
return (0);
}
static void
unlocked_wait(void)
{
semid_t id;
u_int elapsed;
if (ksem_init(&id, 1) < 0) {
fail_errno("ksem_init");
return;
}
/* This should succeed right away and set the value to 0. */
if (testwait(id, &elapsed) < 0) {
ksem_destroy(id);
return;
}
if (!ELAPSED(elapsed, 0)) {
fail_err("ksem_wait() of unlocked sem took %ums", elapsed);
ksem_destroy(id);
return;
}
if (checkvalue(id, 0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(unlocked_wait, "unlocked wait");
static void
locked_wait(void)
{
semid_t id;
u_int elapsed;
if (ksem_init(&id, 0) < 0) {
fail_errno("ksem_init");
return;
}
/*
* Schedule a post to trigger after 1000 ms. The subsequent
* wait should succeed after 1000 ms as a result.
*/
if (schedule_post(id, 1000) < 0) {
ksem_destroy(id);
return;
}
if (testwait(id, &elapsed) < 0) {
check_alarm(1);
ksem_destroy(id);
return;
}
if (!ELAPSED(elapsed, 1000)) {
fail_err(
"ksem_wait() with delayed post took %ums instead of 1000ms",
elapsed);
check_alarm(1);
ksem_destroy(id);
return;
}
if (check_alarm(0) < 0) {
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(locked_wait, "locked wait");
/*
* Fork off a child process. The child will open the semaphore via
* the same name. The child will then block on the semaphore waiting
* for the parent to post it.
*/
static int
wait_twoproc_child(void *arg)
{
semid_t id;
if (ksem_open(&id, TEST_PATH, 0, 0, 0) < 0)
return (CSTAT(1, errno));
if (ksem_wait(id) < 0)
return (CSTAT(2, errno));
if (ksem_close(id) < 0)
return (CSTAT(3, errno));
return (CSTAT(0, 0));
}
static void
wait_twoproc_test(void)
{
semid_t id;
int stat;
if (ksem_open(&id, TEST_PATH, O_CREAT, 0777, 0)) {
fail_errno("ksem_open");
return;
}
if (schedule_post(id, 500) < 0) {
ksem_close(id);
ksem_unlink(TEST_PATH);
return;
}
if (child_worker(wait_twoproc_child, NULL, &stat) < 0) {
check_alarm(1);
ksem_close(id);
ksem_unlink(TEST_PATH);
return;
}
errno = CSTAT_ERROR(stat);
switch (CSTAT_CLASS(stat)) {
case 0:
pass();
break;
case 1:
fail_errno("child ksem_open()");
break;
case 2:
fail_errno("child ksem_wait()");
break;
case 3:
fail_errno("child ksem_close()");
break;
default:
fail_err("bad child state %#x", stat);
break;
}
check_alarm(1);
ksem_close(id);
ksem_unlink(TEST_PATH);
}
TEST(wait_twoproc_test, "two proc wait");
static void
maxvalue_test(void)
{
semid_t id;
int val;
if (ksem_init(&id, SEM_VALUE_MAX) < 0) {
fail_errno("ksem_init");
return;
}
if (ksem_getvalue(id, &val) < 0) {
fail_errno("ksem_getvalue");
ksem_destroy(id);
return;
}
if (val != SEM_VALUE_MAX) {
fail_err("value %d != SEM_VALUE_MAX");
ksem_destroy(id);
return;
}
if (val < 0) {
fail_err("value < 0");
ksem_destroy(id);
return;
}
if (ksem_destroy(id) < 0) {
fail_errno("ksem_destroy");
return;
}
pass();
}
TEST(maxvalue_test, "get value of SEM_VALUE_MAX semaphore");
static void
maxvalue_post_test(void)
{
semid_t id;
if (ksem_init(&id, SEM_VALUE_MAX) < 0) {
fail_errno("ksem_init");
return;
}
ksem_post_should_fail(id, EOVERFLOW);
ksem_destroy(id);
}
TEST(maxvalue_post_test, "post SEM_VALUE_MAX semaphore");
static void
busy_destroy_test(void)
{
char errbuf[_POSIX2_LINE_MAX];
struct kinfo_proc *kp;
semid_t id;
pid_t pid;
kvm_t *kd;
int count;
kd = kvm_openfiles(NULL, "/dev/null", NULL, O_RDONLY, errbuf);
if (kd == NULL) {
fail_err("kvm_openfiles: %s", errbuf);
return;
}
if (ksem_init(&id, 0) < 0) {
fail_errno("ksem_init");
kvm_close(kd);
return;
}
pid = fork();
switch (pid) {
case -1:
/* Error. */
fail_errno("fork");
ksem_destroy(id);
kvm_close(kd);
return;
case 0:
/* Child. */
ksem_wait(id);
exit(0);
}
/*
* Wait for the child process to block on the semaphore. This
* is a bit gross.
*/
for (;;) {
kp = kvm_getprocs(kd, KERN_PROC_PID, pid, &count);
if (kp == NULL) {
fail_err("kvm_getprocs: %s", kvm_geterr(kd));
kvm_close(kd);
ksem_destroy(id);
return;
}
if (kp->ki_stat == SSLEEP &&
(strcmp(kp->ki_wmesg, "sem") == 0 ||
strcmp(kp->ki_wmesg, "ksem") == 0))
break;
usleep(1000);
}
kvm_close(kd);
ksem_destroy_should_fail(id, EBUSY);
/* Cleanup. */
ksem_post(id);
waitpid(pid, NULL, 0);
ksem_destroy(id);
}
TEST(busy_destroy_test, "destroy unnamed semaphore with waiter");
static int
exhaust_unnamed_child(void *arg)
{
semid_t id;
int i, max;
max = (intptr_t)arg;
for (i = 0; i < max + 1; i++) {
if (ksem_init(&id, 1) < 0) {
if (errno == ENOSPC)
return (CSTAT(0, 0));
return (CSTAT(1, errno));
}
}
return (CSTAT(2, 0));
}
static void
exhaust_unnamed_sems(void)
{
size_t len;
int nsems_max, stat;
len = sizeof(nsems_max);
if (sysctlbyname("p1003_1b.sem_nsems_max", &nsems_max, &len, NULL, 0) <
0) {
fail_errno("sysctl(p1003_1b.sem_nsems_max)");
return;
}
if (child_worker(exhaust_unnamed_child, (void *)(uintptr_t)nsems_max,
&stat))
Rework the lifetime management of the kernel implementation of POSIX semaphores. Specifically, semaphores are now represented as new file descriptor type that is set to close on exec. This removes the need for all of the manual process reference counting (and fork, exec, and exit event handlers) as the normal file descriptor operations handle all of that for us nicely. It is also suggested as one possible implementation in the spec and at least one other OS (OS X) uses this approach. Some bugs that were fixed as a result include: - References to a named semaphore whose name is removed still work after the sem_unlink() operation. Prior to this patch, if a semaphore's name was removed, valid handles from sem_open() would get EINVAL errors from sem_getvalue(), sem_post(), etc. This fixes that. - Unnamed semaphores created with sem_init() were not cleaned up when a process exited or exec'd. They were only cleaned up if the process did an explicit sem_destroy(). This could result in a leak of semaphore objects that could never be cleaned up. - On the other hand, if another process guessed the id (kernel pointer to 'struct ksem' of an unnamed semaphore (created via sem_init)) and had write access to the semaphore based on UID/GID checks, then that other process could manipulate the semaphore via sem_destroy(), sem_post(), sem_wait(), etc. - As part of the permission check (UID/GID), the umask of the proces creating the semaphore was not honored. Thus if your umask denied group read/write access but the explicit mode in the sem_init() call allowed it, the semaphore would be readable/writable by other users in the same group, for example. This includes access via the previous bug. - If the module refused to unload because there were active semaphores, then it might have deregistered one or more of the semaphore system calls before it noticed that there was a problem. I'm not sure if this actually happened as the order that modules are discovered by the kernel linker depends on how the actual .ko file is linked. One can make the order deterministic by using a single module with a mod_event handler that explicitly registers syscalls (and deregisters during unload after any checks). This also fixes a race where even if the sem_module unloaded first it would have destroyed locks that the syscalls might be trying to access if they are still executing when they are unloaded. XXX: By the way, deregistering system calls doesn't do any blocking to drain any threads from the calls. - Some minor fixes to errno values on error. For example, sem_init() isn't documented to return ENFILE or EMFILE if we run out of semaphores the way that sem_open() can. Instead, it should return ENOSPC in that case. Other changes: - Kernel semaphores now use a hash table to manage the namespace of named semaphores nearly in a similar fashion to the POSIX shared memory object file descriptors. Kernel semaphores can now also have names longer than 14 chars (up to MAXPATHLEN) and can include subdirectories in their pathname. - The UID/GID permission checks for access to a named semaphore are now done via vaccess() rather than a home-rolled set of checks. - Now that kernel semaphores have an associated file object, the various MAC checks for POSIX semaphores accept both a file credential and an active credential. There is also a new posixsem_check_stat() since it is possible to fstat() a semaphore file descriptor. - A small set of regression tests (using the ksem API directly) is present in src/tools/regression/posixsem. Reported by: kris (1) Tested by: kris Reviewed by: rwatson (lightly) MFC after: 1 month
2008-06-27 05:39:04 +00:00
return;
errno = CSTAT_ERROR(stat);
switch (CSTAT_CLASS(stat)) {
case 0:
pass();
break;
case 1:
fail_errno("ksem_init");
break;
case 2:
fail_err("Limit of %d semaphores not enforced", nsems_max);
break;
default:
fail_err("bad child state %#x", stat);
break;
}
}
TEST(exhaust_unnamed_sems, "exhaust unnamed semaphores (1)");
static int
exhaust_named_child(void *arg)
{
char buffer[64];
semid_t id;
int i, max;
max = (intptr_t)arg;
for (i = 0; i < max + 1; i++) {
snprintf(buffer, sizeof(buffer), "%s%d", TEST_PATH, i);
if (ksem_open(&id, buffer, O_CREAT, 0777, 1) < 0) {
if (errno == ENOSPC || errno == EMFILE ||
errno == ENFILE)
return (CSTAT(0, 0));
return (CSTAT(1, errno));
}
}
return (CSTAT(2, errno));
}
static void
exhaust_named_sems(void)
{
char buffer[64];
size_t len;
int i, nsems_max, stat;
len = sizeof(nsems_max);
if (sysctlbyname("p1003_1b.sem_nsems_max", &nsems_max, &len, NULL, 0) <
0) {
fail_errno("sysctl(p1003_1b.sem_nsems_max)");
return;
}
if (child_worker(exhaust_named_child, (void *)(uintptr_t)nsems_max,
&stat) < 0)
Rework the lifetime management of the kernel implementation of POSIX semaphores. Specifically, semaphores are now represented as new file descriptor type that is set to close on exec. This removes the need for all of the manual process reference counting (and fork, exec, and exit event handlers) as the normal file descriptor operations handle all of that for us nicely. It is also suggested as one possible implementation in the spec and at least one other OS (OS X) uses this approach. Some bugs that were fixed as a result include: - References to a named semaphore whose name is removed still work after the sem_unlink() operation. Prior to this patch, if a semaphore's name was removed, valid handles from sem_open() would get EINVAL errors from sem_getvalue(), sem_post(), etc. This fixes that. - Unnamed semaphores created with sem_init() were not cleaned up when a process exited or exec'd. They were only cleaned up if the process did an explicit sem_destroy(). This could result in a leak of semaphore objects that could never be cleaned up. - On the other hand, if another process guessed the id (kernel pointer to 'struct ksem' of an unnamed semaphore (created via sem_init)) and had write access to the semaphore based on UID/GID checks, then that other process could manipulate the semaphore via sem_destroy(), sem_post(), sem_wait(), etc. - As part of the permission check (UID/GID), the umask of the proces creating the semaphore was not honored. Thus if your umask denied group read/write access but the explicit mode in the sem_init() call allowed it, the semaphore would be readable/writable by other users in the same group, for example. This includes access via the previous bug. - If the module refused to unload because there were active semaphores, then it might have deregistered one or more of the semaphore system calls before it noticed that there was a problem. I'm not sure if this actually happened as the order that modules are discovered by the kernel linker depends on how the actual .ko file is linked. One can make the order deterministic by using a single module with a mod_event handler that explicitly registers syscalls (and deregisters during unload after any checks). This also fixes a race where even if the sem_module unloaded first it would have destroyed locks that the syscalls might be trying to access if they are still executing when they are unloaded. XXX: By the way, deregistering system calls doesn't do any blocking to drain any threads from the calls. - Some minor fixes to errno values on error. For example, sem_init() isn't documented to return ENFILE or EMFILE if we run out of semaphores the way that sem_open() can. Instead, it should return ENOSPC in that case. Other changes: - Kernel semaphores now use a hash table to manage the namespace of named semaphores nearly in a similar fashion to the POSIX shared memory object file descriptors. Kernel semaphores can now also have names longer than 14 chars (up to MAXPATHLEN) and can include subdirectories in their pathname. - The UID/GID permission checks for access to a named semaphore are now done via vaccess() rather than a home-rolled set of checks. - Now that kernel semaphores have an associated file object, the various MAC checks for POSIX semaphores accept both a file credential and an active credential. There is also a new posixsem_check_stat() since it is possible to fstat() a semaphore file descriptor. - A small set of regression tests (using the ksem API directly) is present in src/tools/regression/posixsem. Reported by: kris (1) Tested by: kris Reviewed by: rwatson (lightly) MFC after: 1 month
2008-06-27 05:39:04 +00:00
return;
errno = CSTAT_ERROR(stat);
switch (CSTAT_CLASS(stat)) {
case 0:
pass();
break;
case 1:
fail_errno("ksem_open");
break;
case 2:
fail_err("Limit of %d semaphores not enforced", nsems_max);
break;
default:
fail_err("bad child state %#x", stat);
break;
}
/* Cleanup any semaphores created by the child. */
for (i = 0; i < nsems_max + 1; i++) {
snprintf(buffer, sizeof(buffer), "%s%d", TEST_PATH, i);
ksem_unlink(buffer);
}
}
TEST(exhaust_named_sems, "exhaust named semaphores (1)");
static int
fdlimit_set(void *arg)
{
struct rlimit rlim;
int max;
max = (intptr_t)arg;
if (getrlimit(RLIMIT_NOFILE, &rlim) < 0)
return (CSTAT(3, errno));
rlim.rlim_cur = max;
if (setrlimit(RLIMIT_NOFILE, &rlim) < 0)
return (CSTAT(4, errno));
return (0);
}
static int
fdlimit_unnamed_child(void *arg)
{
int stat;
stat = fdlimit_set(arg);
if (stat == 0)
stat = exhaust_unnamed_child(arg);
return (stat);
}
static void
fdlimit_unnamed_sems(void)
{
int nsems_max, stat;
nsems_max = 10;
if (child_worker(fdlimit_unnamed_child, (void *)(uintptr_t)nsems_max,
&stat))
Rework the lifetime management of the kernel implementation of POSIX semaphores. Specifically, semaphores are now represented as new file descriptor type that is set to close on exec. This removes the need for all of the manual process reference counting (and fork, exec, and exit event handlers) as the normal file descriptor operations handle all of that for us nicely. It is also suggested as one possible implementation in the spec and at least one other OS (OS X) uses this approach. Some bugs that were fixed as a result include: - References to a named semaphore whose name is removed still work after the sem_unlink() operation. Prior to this patch, if a semaphore's name was removed, valid handles from sem_open() would get EINVAL errors from sem_getvalue(), sem_post(), etc. This fixes that. - Unnamed semaphores created with sem_init() were not cleaned up when a process exited or exec'd. They were only cleaned up if the process did an explicit sem_destroy(). This could result in a leak of semaphore objects that could never be cleaned up. - On the other hand, if another process guessed the id (kernel pointer to 'struct ksem' of an unnamed semaphore (created via sem_init)) and had write access to the semaphore based on UID/GID checks, then that other process could manipulate the semaphore via sem_destroy(), sem_post(), sem_wait(), etc. - As part of the permission check (UID/GID), the umask of the proces creating the semaphore was not honored. Thus if your umask denied group read/write access but the explicit mode in the sem_init() call allowed it, the semaphore would be readable/writable by other users in the same group, for example. This includes access via the previous bug. - If the module refused to unload because there were active semaphores, then it might have deregistered one or more of the semaphore system calls before it noticed that there was a problem. I'm not sure if this actually happened as the order that modules are discovered by the kernel linker depends on how the actual .ko file is linked. One can make the order deterministic by using a single module with a mod_event handler that explicitly registers syscalls (and deregisters during unload after any checks). This also fixes a race where even if the sem_module unloaded first it would have destroyed locks that the syscalls might be trying to access if they are still executing when they are unloaded. XXX: By the way, deregistering system calls doesn't do any blocking to drain any threads from the calls. - Some minor fixes to errno values on error. For example, sem_init() isn't documented to return ENFILE or EMFILE if we run out of semaphores the way that sem_open() can. Instead, it should return ENOSPC in that case. Other changes: - Kernel semaphores now use a hash table to manage the namespace of named semaphores nearly in a similar fashion to the POSIX shared memory object file descriptors. Kernel semaphores can now also have names longer than 14 chars (up to MAXPATHLEN) and can include subdirectories in their pathname. - The UID/GID permission checks for access to a named semaphore are now done via vaccess() rather than a home-rolled set of checks. - Now that kernel semaphores have an associated file object, the various MAC checks for POSIX semaphores accept both a file credential and an active credential. There is also a new posixsem_check_stat() since it is possible to fstat() a semaphore file descriptor. - A small set of regression tests (using the ksem API directly) is present in src/tools/regression/posixsem. Reported by: kris (1) Tested by: kris Reviewed by: rwatson (lightly) MFC after: 1 month
2008-06-27 05:39:04 +00:00
return;
errno = CSTAT_ERROR(stat);
switch (CSTAT_CLASS(stat)) {
case 0:
pass();
break;
case 1:
fail_errno("ksem_init");
break;
case 2:
fail_err("Limit of %d semaphores not enforced", nsems_max);
break;
case 3:
fail_errno("getrlimit");
break;
case 4:
fail_errno("getrlimit");
break;
default:
fail_err("bad child state %#x", stat);
break;
}
}
TEST(fdlimit_unnamed_sems, "exhaust unnamed semaphores (2)");
static int
fdlimit_named_child(void *arg)
{
int stat;
stat = fdlimit_set(arg);
if (stat == 0)
stat = exhaust_named_child(arg);
return (stat);
}
static void
fdlimit_named_sems(void)
{
char buffer[64];
int i, nsems_max, stat;
nsems_max = 10;
if (child_worker(fdlimit_named_child, (void *)(uintptr_t)nsems_max,
&stat) < 0)
Rework the lifetime management of the kernel implementation of POSIX semaphores. Specifically, semaphores are now represented as new file descriptor type that is set to close on exec. This removes the need for all of the manual process reference counting (and fork, exec, and exit event handlers) as the normal file descriptor operations handle all of that for us nicely. It is also suggested as one possible implementation in the spec and at least one other OS (OS X) uses this approach. Some bugs that were fixed as a result include: - References to a named semaphore whose name is removed still work after the sem_unlink() operation. Prior to this patch, if a semaphore's name was removed, valid handles from sem_open() would get EINVAL errors from sem_getvalue(), sem_post(), etc. This fixes that. - Unnamed semaphores created with sem_init() were not cleaned up when a process exited or exec'd. They were only cleaned up if the process did an explicit sem_destroy(). This could result in a leak of semaphore objects that could never be cleaned up. - On the other hand, if another process guessed the id (kernel pointer to 'struct ksem' of an unnamed semaphore (created via sem_init)) and had write access to the semaphore based on UID/GID checks, then that other process could manipulate the semaphore via sem_destroy(), sem_post(), sem_wait(), etc. - As part of the permission check (UID/GID), the umask of the proces creating the semaphore was not honored. Thus if your umask denied group read/write access but the explicit mode in the sem_init() call allowed it, the semaphore would be readable/writable by other users in the same group, for example. This includes access via the previous bug. - If the module refused to unload because there were active semaphores, then it might have deregistered one or more of the semaphore system calls before it noticed that there was a problem. I'm not sure if this actually happened as the order that modules are discovered by the kernel linker depends on how the actual .ko file is linked. One can make the order deterministic by using a single module with a mod_event handler that explicitly registers syscalls (and deregisters during unload after any checks). This also fixes a race where even if the sem_module unloaded first it would have destroyed locks that the syscalls might be trying to access if they are still executing when they are unloaded. XXX: By the way, deregistering system calls doesn't do any blocking to drain any threads from the calls. - Some minor fixes to errno values on error. For example, sem_init() isn't documented to return ENFILE or EMFILE if we run out of semaphores the way that sem_open() can. Instead, it should return ENOSPC in that case. Other changes: - Kernel semaphores now use a hash table to manage the namespace of named semaphores nearly in a similar fashion to the POSIX shared memory object file descriptors. Kernel semaphores can now also have names longer than 14 chars (up to MAXPATHLEN) and can include subdirectories in their pathname. - The UID/GID permission checks for access to a named semaphore are now done via vaccess() rather than a home-rolled set of checks. - Now that kernel semaphores have an associated file object, the various MAC checks for POSIX semaphores accept both a file credential and an active credential. There is also a new posixsem_check_stat() since it is possible to fstat() a semaphore file descriptor. - A small set of regression tests (using the ksem API directly) is present in src/tools/regression/posixsem. Reported by: kris (1) Tested by: kris Reviewed by: rwatson (lightly) MFC after: 1 month
2008-06-27 05:39:04 +00:00
return;
errno = CSTAT_ERROR(stat);
switch (CSTAT_CLASS(stat)) {
case 0:
pass();
break;
case 1:
fail_errno("ksem_open");
break;
case 2:
fail_err("Limit of %d semaphores not enforced", nsems_max);
break;
case 3:
fail_errno("getrlimit");
break;
case 4:
fail_errno("getrlimit");
break;
default:
fail_err("bad child state %#x", stat);
break;
}
/* Cleanup any semaphores created by the child. */
for (i = 0; i < nsems_max + 1; i++) {
snprintf(buffer, sizeof(buffer), "%s%d", TEST_PATH, i);
ksem_unlink(buffer);
}
}
TEST(fdlimit_named_sems, "exhaust named semaphores (2)");
int
main(int argc, char *argv[])
{
signal(SIGSYS, SIG_IGN);
run_tests();
return (0);
}