c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2024-11-17 18:11:00 +03:00
Code Issues Packages Projects Releases Wiki Activity
1b457bcbe5
mirror_zfs/module/splat/splat-condvar.c
Chunwei Chen e843553d03 Don't hold mutex until release cv in cv_wait 
If a thread is holding mutex when doing cv_destroy, it might end up waiting a
thread in cv_wait. The waiter would wake up trying to aquire the same mutex
and cause deadlock.

We solve this by move the mutex_enter to the bottom of cv_wait, so that
the waiter will release the cv first, allowing cv_destroy to succeed and have
a chance to free the mutex.

This would create race condition on the cv_mutex. We use xchg to set and check
it to ensure we won't be harmed by the race. This would result in the cv_mutex
debugging becomes best-effort.

Also, the change reveals a race, which was unlikely before, where we call
mutex_destroy while test threads are still holding the mutex. We use
kthread_stop to make sure the threads are exit before mutex_destroy.

Signed-off-by: Chunwei Chen <tuxoko@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tim Chase <tim@chase2k.com>
Issue zfsonlinux/zfs#4166
Issue zfsonlinux/zfs#4106
2016-01-12 15:18:44 -08:00

512 lines
15 KiB
C
Raw Blame History

/*****************************************************************************\
* Copyright (C) 2007-2010 Lawrence Livermore National Security, LLC.
* Copyright (C) 2007 The Regents of the University of California.
* Produced at Lawrence Livermore National Laboratory (cf, DISCLAIMER).
* Written by Brian Behlendorf <behlendorf1@llnl.gov>.
* UCRL-CODE-235197
*
* This file is part of the SPL, Solaris Porting Layer.
* For details, see <http://zfsonlinux.org/>.
*
* The SPL is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License as published by the
* Free Software Foundation; either version 2 of the License, or (at your
* option) any later version.
*
* The SPL 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 General Public License
* for more details.
*
* You should have received a copy of the GNU General Public License along
* with the SPL. If not, see <http://www.gnu.org/licenses/>.
*****************************************************************************
* Solaris Porting LAyer Tests (SPLAT) Condition Variable Tests.
\*****************************************************************************/
#include <sys/condvar.h>
#include <sys/timer.h>
#include <sys/thread.h>
#include "splat-internal.h"
#define SPLAT_CONDVAR_NAME "condvar"
#define SPLAT_CONDVAR_DESC "Kernel Condition Variable Tests"
#define SPLAT_CONDVAR_TEST1_ID 0x0501
#define SPLAT_CONDVAR_TEST1_NAME "signal1"
#define SPLAT_CONDVAR_TEST1_DESC "Wake a single thread, cv_wait()/cv_signal()"
#define SPLAT_CONDVAR_TEST2_ID 0x0502
#define SPLAT_CONDVAR_TEST2_NAME "broadcast1"
#define SPLAT_CONDVAR_TEST2_DESC "Wake all threads, cv_wait()/cv_broadcast()"
#define SPLAT_CONDVAR_TEST3_ID 0x0503
#define SPLAT_CONDVAR_TEST3_NAME "signal2"
#define SPLAT_CONDVAR_TEST3_DESC "Wake a single thread, cv_wait_timeout()/cv_signal()"
#define SPLAT_CONDVAR_TEST4_ID 0x0504
#define SPLAT_CONDVAR_TEST4_NAME "broadcast2"
#define SPLAT_CONDVAR_TEST4_DESC "Wake all threads, cv_wait_timeout()/cv_broadcast()"
#define SPLAT_CONDVAR_TEST5_ID 0x0505
#define SPLAT_CONDVAR_TEST5_NAME "timeout"
#define SPLAT_CONDVAR_TEST5_DESC "Timeout thread, cv_wait_timeout()"
#define SPLAT_CONDVAR_TEST_MAGIC 0x115599DDUL
#define SPLAT_CONDVAR_TEST_NAME "condvar"
#define SPLAT_CONDVAR_TEST_COUNT 8
typedef struct condvar_priv {
unsigned long cv_magic;
struct file *cv_file;
kcondvar_t cv_condvar;
kmutex_t cv_mtx;
} condvar_priv_t;
typedef struct condvar_thr {
const char *ct_name;
condvar_priv_t *ct_cvp;
struct task_struct *ct_thread;
int ct_rc;
} condvar_thr_t;
int
splat_condvar_test12_thread(void *arg)
{
condvar_thr_t *ct = (condvar_thr_t *)arg;
condvar_priv_t *cv = ct->ct_cvp;
ASSERT(cv->cv_magic == SPLAT_CONDVAR_TEST_MAGIC);
mutex_enter(&cv->cv_mtx);
splat_vprint(cv->cv_file, ct->ct_name,
"%s thread sleeping with %d waiters\n",
ct->ct_thread->comm, atomic_read(&cv->cv_condvar.cv_waiters));
cv_wait(&cv->cv_condvar, &cv->cv_mtx);
splat_vprint(cv->cv_file, ct->ct_name,
"%s thread woken %d waiters remain\n",
ct->ct_thread->comm, atomic_read(&cv->cv_condvar.cv_waiters));
mutex_exit(&cv->cv_mtx);
/* wait for main thread reap us */
while (!kthread_should_stop())
schedule();
return 0;
}
static int
splat_condvar_test1(struct file *file, void *arg)
{
int i, count = 0, rc = 0;
condvar_thr_t ct[SPLAT_CONDVAR_TEST_COUNT];
condvar_priv_t cv;
cv.cv_magic = SPLAT_CONDVAR_TEST_MAGIC;
cv.cv_file = file;
mutex_init(&cv.cv_mtx, SPLAT_CONDVAR_TEST_NAME, MUTEX_DEFAULT, NULL);
cv_init(&cv.cv_condvar, NULL, CV_DEFAULT, NULL);
/* Create some threads, the exact number isn't important just as
* long as we know how many we managed to create and should expect. */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
ct[i].ct_cvp = &cv;
ct[i].ct_name = SPLAT_CONDVAR_TEST1_NAME;
ct[i].ct_rc = 0;
ct[i].ct_thread = spl_kthread_create(splat_condvar_test12_thread,
&ct[i], "%s/%d", SPLAT_CONDVAR_TEST_NAME, i);
if (!IS_ERR(ct[i].ct_thread)) {
wake_up_process(ct[i].ct_thread);
count++;
}
}
/* Wait until all threads are waiting on the condition variable */
while (atomic_read(&cv.cv_condvar.cv_waiters) != count)
schedule();
/* Wake a single thread at a time, wait until it exits */
for (i = 1; i <= count; i++) {
cv_signal(&cv.cv_condvar);
while (atomic_read(&cv.cv_condvar.cv_waiters) > (count - i))
schedule();
/* Correct behavior 1 thread woken */
if (atomic_read(&cv.cv_condvar.cv_waiters) == (count - i))
continue;
splat_vprint(file, SPLAT_CONDVAR_TEST1_NAME, "Attempted to "
"wake %d thread but work %d threads woke\n",
1, count - atomic_read(&cv.cv_condvar.cv_waiters));
rc = -EINVAL;
break;
}
if (!rc)
splat_vprint(file, SPLAT_CONDVAR_TEST1_NAME, "Correctly woke "
"%d sleeping threads %d at a time\n", count, 1);
/* Wait until that last nutex is dropped */
while (mutex_owner(&cv.cv_mtx))
schedule();
/* Wake everything for the failure case */
cv_broadcast(&cv.cv_condvar);
cv_destroy(&cv.cv_condvar);
/* wait for threads to exit */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
if (!IS_ERR(ct[i].ct_thread))
kthread_stop(ct[i].ct_thread);
}
mutex_destroy(&cv.cv_mtx);
return rc;
}
static int
splat_condvar_test2(struct file *file, void *arg)
{
int i, count = 0, rc = 0;
condvar_thr_t ct[SPLAT_CONDVAR_TEST_COUNT];
condvar_priv_t cv;
cv.cv_magic = SPLAT_CONDVAR_TEST_MAGIC;
cv.cv_file = file;
mutex_init(&cv.cv_mtx, SPLAT_CONDVAR_TEST_NAME, MUTEX_DEFAULT, NULL);
cv_init(&cv.cv_condvar, NULL, CV_DEFAULT, NULL);
/* Create some threads, the exact number isn't important just as
* long as we know how many we managed to create and should expect. */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
ct[i].ct_cvp = &cv;
ct[i].ct_name = SPLAT_CONDVAR_TEST2_NAME;
ct[i].ct_rc = 0;
ct[i].ct_thread = spl_kthread_create(splat_condvar_test12_thread,
&ct[i], "%s/%d", SPLAT_CONDVAR_TEST_NAME, i);
if (!IS_ERR(ct[i].ct_thread)) {
wake_up_process(ct[i].ct_thread);
count++;
}
}
/* Wait until all threads are waiting on the condition variable */
while (atomic_read(&cv.cv_condvar.cv_waiters) != count)
schedule();
/* Wake all threads waiting on the condition variable */
cv_broadcast(&cv.cv_condvar);
/* Wait until all threads have exited */
while ((atomic_read(&cv.cv_condvar.cv_waiters) > 0) || mutex_owner(&cv.cv_mtx))
schedule();
splat_vprint(file, SPLAT_CONDVAR_TEST2_NAME, "Correctly woke all "
"%d sleeping threads at once\n", count);
/* Wake everything for the failure case */
cv_destroy(&cv.cv_condvar);
/* wait for threads to exit */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
if (!IS_ERR(ct[i].ct_thread))
kthread_stop(ct[i].ct_thread);
}
mutex_destroy(&cv.cv_mtx);
return rc;
}
int
splat_condvar_test34_thread(void *arg)
{
condvar_thr_t *ct = (condvar_thr_t *)arg;
condvar_priv_t *cv = ct->ct_cvp;
clock_t rc;
ASSERT(cv->cv_magic == SPLAT_CONDVAR_TEST_MAGIC);
mutex_enter(&cv->cv_mtx);
splat_vprint(cv->cv_file, ct->ct_name,
"%s thread sleeping with %d waiters\n",
ct->ct_thread->comm, atomic_read(&cv->cv_condvar.cv_waiters));
/* Sleep no longer than 3 seconds, for this test we should
* actually never sleep that long without being woken up. */
rc = cv_timedwait(&cv->cv_condvar, &cv->cv_mtx, lbolt + HZ * 3);
if (rc == -1) {
ct->ct_rc = -ETIMEDOUT;
splat_vprint(cv->cv_file, ct->ct_name, "%s thread timed out, "
"should have been woken\n", ct->ct_thread->comm);
} else {
splat_vprint(cv->cv_file, ct->ct_name,
"%s thread woken %d waiters remain\n",
ct->ct_thread->comm,
atomic_read(&cv->cv_condvar.cv_waiters));
}
mutex_exit(&cv->cv_mtx);
/* wait for main thread reap us */
while (!kthread_should_stop())
schedule();
return 0;
}
static int
splat_condvar_test3(struct file *file, void *arg)
{
int i, count = 0, rc = 0;
condvar_thr_t ct[SPLAT_CONDVAR_TEST_COUNT];
condvar_priv_t cv;
cv.cv_magic = SPLAT_CONDVAR_TEST_MAGIC;
cv.cv_file = file;
mutex_init(&cv.cv_mtx, SPLAT_CONDVAR_TEST_NAME, MUTEX_DEFAULT, NULL);
cv_init(&cv.cv_condvar, NULL, CV_DEFAULT, NULL);
/* Create some threads, the exact number isn't important just as
* long as we know how many we managed to create and should expect. */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
ct[i].ct_cvp = &cv;
ct[i].ct_name = SPLAT_CONDVAR_TEST3_NAME;
ct[i].ct_rc = 0;
ct[i].ct_thread = spl_kthread_create(splat_condvar_test34_thread,
&ct[i], "%s/%d", SPLAT_CONDVAR_TEST_NAME, i);
if (!IS_ERR(ct[i].ct_thread)) {
wake_up_process(ct[i].ct_thread);
count++;
}
}
/* Wait until all threads are waiting on the condition variable */
while (atomic_read(&cv.cv_condvar.cv_waiters) != count)
schedule();
/* Wake a single thread at a time, wait until it exits */
for (i = 1; i <= count; i++) {
cv_signal(&cv.cv_condvar);
while (atomic_read(&cv.cv_condvar.cv_waiters) > (count - i))
schedule();
/* Correct behavior 1 thread woken */
if (atomic_read(&cv.cv_condvar.cv_waiters) == (count - i))
continue;
splat_vprint(file, SPLAT_CONDVAR_TEST3_NAME, "Attempted to "
"wake %d thread but work %d threads woke\n",
1, count - atomic_read(&cv.cv_condvar.cv_waiters));
rc = -EINVAL;
break;
}
/* Validate no waiting thread timed out early */
for (i = 0; i < count; i++)
if (ct[i].ct_rc)
rc = ct[i].ct_rc;
if (!rc)
splat_vprint(file, SPLAT_CONDVAR_TEST3_NAME, "Correctly woke "
"%d sleeping threads %d at a time\n", count, 1);
/* Wait until that last nutex is dropped */
while (mutex_owner(&cv.cv_mtx))
schedule();
/* Wake everything for the failure case */
cv_broadcast(&cv.cv_condvar);
cv_destroy(&cv.cv_condvar);
/* wait for threads to exit */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
if (!IS_ERR(ct[i].ct_thread))
kthread_stop(ct[i].ct_thread);
}
mutex_destroy(&cv.cv_mtx);
return rc;
}
static int
splat_condvar_test4(struct file *file, void *arg)
{
int i, count = 0, rc = 0;
condvar_thr_t ct[SPLAT_CONDVAR_TEST_COUNT];
condvar_priv_t cv;
cv.cv_magic = SPLAT_CONDVAR_TEST_MAGIC;
cv.cv_file = file;
mutex_init(&cv.cv_mtx, SPLAT_CONDVAR_TEST_NAME, MUTEX_DEFAULT, NULL);
cv_init(&cv.cv_condvar, NULL, CV_DEFAULT, NULL);
/* Create some threads, the exact number isn't important just as
* long as we know how many we managed to create and should expect. */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
ct[i].ct_cvp = &cv;
ct[i].ct_name = SPLAT_CONDVAR_TEST3_NAME;
ct[i].ct_rc = 0;
ct[i].ct_thread = spl_kthread_create(splat_condvar_test34_thread,
&ct[i], "%s/%d", SPLAT_CONDVAR_TEST_NAME, i);
if (!IS_ERR(ct[i].ct_thread)) {
wake_up_process(ct[i].ct_thread);
count++;
}
}
/* Wait until all threads are waiting on the condition variable */
while (atomic_read(&cv.cv_condvar.cv_waiters) != count)
schedule();
/* Wake a single thread at a time, wait until it exits */
for (i = 1; i <= count; i++) {
cv_signal(&cv.cv_condvar);
while (atomic_read(&cv.cv_condvar.cv_waiters) > (count - i))
schedule();
/* Correct behavior 1 thread woken */
if (atomic_read(&cv.cv_condvar.cv_waiters) == (count - i))
continue;
splat_vprint(file, SPLAT_CONDVAR_TEST3_NAME, "Attempted to "
"wake %d thread but work %d threads woke\n",
1, count - atomic_read(&cv.cv_condvar.cv_waiters));
rc = -EINVAL;
break;
}
/* Validate no waiting thread timed out early */
for (i = 0; i < count; i++)
if (ct[i].ct_rc)
rc = ct[i].ct_rc;
if (!rc)
splat_vprint(file, SPLAT_CONDVAR_TEST3_NAME, "Correctly woke "
"%d sleeping threads %d at a time\n", count, 1);
/* Wait until that last nutex is dropped */
while (mutex_owner(&cv.cv_mtx))
schedule();
/* Wake everything for the failure case */
cv_broadcast(&cv.cv_condvar);
cv_destroy(&cv.cv_condvar);
/* wait for threads to exit */
for (i = 0; i < SPLAT_CONDVAR_TEST_COUNT; i++) {
if (!IS_ERR(ct[i].ct_thread))
kthread_stop(ct[i].ct_thread);
}
mutex_destroy(&cv.cv_mtx);
return rc;
}
static int
splat_condvar_test5(struct file *file, void *arg)
{
kcondvar_t condvar;
kmutex_t mtx;
clock_t time_left, time_before, time_after, time_delta;
uint64_t whole_delta;
uint32_t remain_delta;
int rc = 0;
mutex_init(&mtx, SPLAT_CONDVAR_TEST_NAME, MUTEX_DEFAULT, NULL);
cv_init(&condvar, NULL, CV_DEFAULT, NULL);
splat_vprint(file, SPLAT_CONDVAR_TEST5_NAME, "Thread going to sleep for "
"%d second and expecting to be woken by timeout\n", 1);
/* Allow a 1 second timeout, plenty long to validate correctness. */
time_before = lbolt;
mutex_enter(&mtx);
time_left = cv_timedwait(&condvar, &mtx, lbolt + HZ);
mutex_exit(&mtx);
time_after = lbolt;
time_delta = time_after - time_before; /* XXX - Handle jiffie wrap */
whole_delta = time_delta;
remain_delta = do_div(whole_delta, HZ);
if (time_left == -1) {
if (time_delta >= HZ) {
splat_vprint(file, SPLAT_CONDVAR_TEST5_NAME,
"Thread correctly timed out and was asleep "
"for %d.%d seconds (%d second min)\n",
(int)whole_delta, (int)remain_delta, 1);
} else {
splat_vprint(file, SPLAT_CONDVAR_TEST5_NAME,
"Thread correctly timed out but was only "
"asleep for %d.%d seconds (%d second "
"min)\n", (int)whole_delta,
(int)remain_delta, 1);
rc = -ETIMEDOUT;
}
} else {
splat_vprint(file, SPLAT_CONDVAR_TEST5_NAME,
"Thread exited after only %d.%d seconds, it "
"did not hit the %d second timeout\n",
(int)whole_delta, (int)remain_delta, 1);
rc = -ETIMEDOUT;
}
cv_destroy(&condvar);
mutex_destroy(&mtx);
return rc;
}
splat_subsystem_t *
splat_condvar_init(void)
{
splat_subsystem_t *sub;
sub = kmalloc(sizeof(*sub), GFP_KERNEL);
if (sub == NULL)
return NULL;
memset(sub, 0, sizeof(*sub));
strncpy(sub->desc.name, SPLAT_CONDVAR_NAME, SPLAT_NAME_SIZE);
strncpy(sub->desc.desc, SPLAT_CONDVAR_DESC, SPLAT_DESC_SIZE);
INIT_LIST_HEAD(&sub->subsystem_list);
INIT_LIST_HEAD(&sub->test_list);
spin_lock_init(&sub->test_lock);
sub->desc.id = SPLAT_SUBSYSTEM_CONDVAR;
SPLAT_TEST_INIT(sub, SPLAT_CONDVAR_TEST1_NAME, SPLAT_CONDVAR_TEST1_DESC,
SPLAT_CONDVAR_TEST1_ID, splat_condvar_test1);
SPLAT_TEST_INIT(sub, SPLAT_CONDVAR_TEST2_NAME, SPLAT_CONDVAR_TEST2_DESC,
SPLAT_CONDVAR_TEST2_ID, splat_condvar_test2);
SPLAT_TEST_INIT(sub, SPLAT_CONDVAR_TEST3_NAME, SPLAT_CONDVAR_TEST3_DESC,
SPLAT_CONDVAR_TEST3_ID, splat_condvar_test3);
SPLAT_TEST_INIT(sub, SPLAT_CONDVAR_TEST4_NAME, SPLAT_CONDVAR_TEST4_DESC,
SPLAT_CONDVAR_TEST4_ID, splat_condvar_test4);
SPLAT_TEST_INIT(sub, SPLAT_CONDVAR_TEST5_NAME, SPLAT_CONDVAR_TEST5_DESC,
SPLAT_CONDVAR_TEST5_ID, splat_condvar_test5);
return sub;
}
void
splat_condvar_fini(splat_subsystem_t *sub)
{
ASSERT(sub);
SPLAT_TEST_FINI(sub, SPLAT_CONDVAR_TEST5_ID);
SPLAT_TEST_FINI(sub, SPLAT_CONDVAR_TEST4_ID);
SPLAT_TEST_FINI(sub, SPLAT_CONDVAR_TEST3_ID);
SPLAT_TEST_FINI(sub, SPLAT_CONDVAR_TEST2_ID);
SPLAT_TEST_FINI(sub, SPLAT_CONDVAR_TEST1_ID);
kfree(sub);
}
int
splat_condvar_id(void) {
return SPLAT_SUBSYSTEM_CONDVAR;
}
Reference in New Issue View Git Blame Copy Permalink