Archive-Team
/
zfs
2
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 103 Wiki Activity

Additional pthread related fixes for ztest 

There are 3 fixes in thie commit.  First, update ztest_run() to store
the thread id and not the address of the kthread_t.  This will be freed
on thread exit and is not safe to use.  This is pretty close to how
things were done in the original ztest code before I got there.

Second, for extra paranoia update thread_exit() to return a special
TS_MAGIC value via pthread_exit().  This value is then verified in
pthread_join() to ensure the thread exited cleanly.  This can be
done cleanly because the kthread doesn't provide a return code
mechanism we need to worry about.

Third, replace the ztest deadman thread with a signal handler.  We
cannot use the previous approach because the correct behavior for
pthreads is to wait for all threads to exit before terminating the
process.  Since the deadman thread won't call exit by design we
end up hanging in kernel_exit().  To avoid this we just setup a
SIGALRM signal handle and register a deadman alarm.  IMHO this
is simpler and cleaner anyway.
This commit is contained in:
Brian Behlendorf 2010-06-28 23:45:08 -07:00
parent a2e73b7516
commit 6fe1e37a5e
3 changed files with 23 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
cmd/ztest/ztest.c
View File

@ -4904,21 +4904,12 @@ ztest_resume_thread(void *arg)
return (NULL);
}
static void *
ztest_deadman_thread(void *arg)
#define GRACE 300
static void
ztest_deadman_alarm(int sig)
{
ztest_shared_t *zs = arg;
int grace = 300;
hrtime_t delta;
delta = (zs->zs_thread_stop - zs->zs_thread_start) / NANOSEC + grace;
(void) poll(NULL, 0, (int)(1000 * delta));
fatal(0, "failed to complete within %d seconds of deadline", grace);
thread_exit();
return (NULL);
fatal(0, "failed to complete within %d seconds of deadline", GRACE);
}
static void
@ -5106,7 +5097,7 @@ ztest_dataset_close(ztest_shared_t *zs, int d)
static void
ztest_run(ztest_shared_t *zs)
{
kthread_t **tid;
kt_did_t *tid;
spa_t *spa;
kthread_t *resume_thread;
uint64_t object;
@ -5158,10 +5149,10 @@ ztest_run(ztest_shared_t *zs)
spa, TS_RUN, NULL, 0, 0)), !=, NULL);
/*
* Create a deadman thread to abort() if we hang.
* Set a deadman alarm to abort() if we hang.
*/
VERIFY3P(thread_create(NULL, 0, ztest_deadman_thread, zs,
TS_RUN, NULL, 0, 0), !=, NULL);
signal(SIGALRM, ztest_deadman_alarm);
alarm((zs->zs_thread_stop - zs->zs_thread_start) / NANOSEC + GRACE);
/*
* Verify that we can safely inquire about about any object,
@ -5187,7 +5178,7 @@ ztest_run(ztest_shared_t *zs)
}
zs->zs_enospc_count = 0;
tid = umem_zalloc(zopt_threads * sizeof (kthread_t *), UMEM_NOFAIL);
tid = umem_zalloc(zopt_threads * sizeof (kt_did_t), UMEM_NOFAIL);
if (zopt_verbose >= 4)
(void) printf("starting main threads...\n");
@ -5196,11 +5187,14 @@ ztest_run(ztest_shared_t *zs)
* Kick off all the tests that run in parallel.
*/
for (int t = 0; t < zopt_threads; t++) {
kthread_t *thread;
if (t < zopt_datasets && ztest_dataset_open(zs, t) != 0)
return;
VERIFY3P(tid[t] = thread_create(NULL, 0, ztest_thread,
VERIFY3P(thread = thread_create(NULL, 0, ztest_thread,
(void *)(uintptr_t)t, TS_RUN, NULL, 0, 0), !=, NULL);
tid[t] = thread->t_tid;
}
/*
@ -5208,7 +5202,7 @@ ztest_run(ztest_shared_t *zs)
* so we don't close datasets while threads are still using them.
*/
for (int t = zopt_threads - 1; t >= 0; t--) {
thread_join(tid[t]->t_tid);
thread_join(tid[t]);
if (t < zopt_datasets)
ztest_dataset_close(zs, t);
}
@ -5218,7 +5212,7 @@ ztest_run(ztest_shared_t *zs)
zs->zs_alloc = metaslab_class_get_alloc(spa_normal_class(spa));
zs->zs_space = metaslab_class_get_space(spa_normal_class(spa));
umem_free(tid, zopt_threads * sizeof (kthread_t *));
umem_free(tid, zopt_threads * sizeof (kt_did_t));
/* Kill the resume thread */
ztest_exiting = B_TRUE;

4
lib/libzpool/include/sys/zfs_context.h
View File

@ -197,7 +197,8 @@ _NOTE(CONSTCOND) } while (0)
/*
* Threads
*/
#define TS_RUN 0x00000002
#define TS_MAGIC 0x72f158ab4261e538ull
#define TS_RUN 0x00000002
#ifdef _linux_
#define STACK_SIZE 8192 /* Linux x86 and amd64 */
#else
@ -221,7 +222,6 @@ typedef struct kthread {
void * t_arg;
} kthread_t;
/* XXX tsd_create()/tsd_destroy() missing */
#define tsd_get(key) pthread_getspecific(key)
#define tsd_set(key, val) pthread_setspecific(key, val)
#define curthread zk_thread_current()

7
lib/libzpool/kernel.c
View File

@ -192,13 +192,16 @@ zk_thread_exit(void)
pthread_mutex_unlock(&kthread_lock);
pthread_cond_broadcast(&kthread_cond);
pthread_exit(NULL);
pthread_exit((void *)TS_MAGIC);
}
void
zk_thread_join(kt_did_t tid)
{
pthread_join((pthread_t)tid, NULL);
void *ret;
pthread_join((pthread_t)tid, &ret);
VERIFY3P(ret, ==, (void *)TS_MAGIC);
}
/*