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

Fix subtle race in threads test case 

The call to wake_up() must be moved under the spin lock because
once we drop the lock 'tp' may no longer be valid because the
creating thread has exited.  This basic thread implementation
was correct, this was simply a flaw in the test case.
This commit is contained in:
Brian Behlendorf 2010-06-28 12:34:20 -07:00
parent 5be4767ae1
commit 616df2dd8b
1 changed files with 3 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
module/splat/splat-thread.c
View File

@ -67,9 +67,9 @@ splat_thread_work1(void *priv)
spin_lock(&tp->tp_lock);
ASSERT(tp->tp_magic == SPLAT_THREAD_TEST_MAGIC);
tp->tp_rc = 1;
wake_up(&tp->tp_waitq);
spin_unlock(&tp->tp_lock);
wake_up(&tp->tp_waitq);
thread_exit();
}
@ -108,18 +108,17 @@ splat_thread_work2(void *priv)
spin_lock(&tp->tp_lock);
ASSERT(tp->tp_magic == SPLAT_THREAD_TEST_MAGIC);
tp->tp_rc = 1;
wake_up(&tp->tp_waitq);
spin_unlock(&tp->tp_lock);
wake_up(&tp->tp_waitq);
thread_exit();
/* The following code is unreachable when thread_exit() is
* working properly, which is exactly what we're testing */
spin_lock(&tp->tp_lock);
tp->tp_rc = 2;
spin_unlock(&tp->tp_lock);
wake_up(&tp->tp_waitq);
spin_unlock(&tp->tp_lock);
}
static int