From: Hannes Reinecke <hare@suse.de>
To: Nic Bellinger <nab@daterainc.com>
Cc: target-devel@vger.kernel.org, linux-scsi@vger.kernel.org,
Hannes Reinecke <hare@suse.de>
Subject: [PATCH] iscsi_target: race condition on shutdown
Date: Thu, 5 Dec 2013 14:54:25 +0100 [thread overview]
Message-ID: <1386251665-39173-1-git-send-email-hare@suse.de> (raw)
When shutting down a target there is a race condition between
iscsit_del_np() and __iscsi_target_login_thread().
The latter sets the thread pointer to NULL, and the former
tries to issue kthread_stop() on that pointer without any
synchronization.
This patchs adds proper synchronization pointer between those
calls to ensure that a) the thread is correctly terminate and
b) kthread_stop() isn't called with a NULL pointer.
In the long run iscsi_target_login_thread() should be converted
into a workqueue.
Signed-off-by: Hannes Reinecke <hare@suse.de>
---
drivers/target/iscsi/iscsi_target.c | 12 +++++++++---
drivers/target/iscsi/iscsi_target_login.c | 9 ++++++---
2 files changed, 15 insertions(+), 6 deletions(-)
diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c
index bf76fc4..c7bf3c9 100644
--- a/drivers/target/iscsi/iscsi_target.c
+++ b/drivers/target/iscsi/iscsi_target.c
@@ -457,15 +457,21 @@ int iscsit_del_np(struct iscsi_np *np)
}
np->np_thread_state = ISCSI_NP_THREAD_SHUTDOWN;
spin_unlock_bh(&np->np_thread_lock);
-
- if (np->np_thread) {
+ /* Give __iscsi_target_login_thread() a chance to run */
+ schedule();
+ spin_lock_bh(&np->np_thread_lock);
+ if ((np->np_thread_state == ISCSI_NP_THREAD_SHUTDOWN)
+ && np->np_thread) {
+ np->np_thread_state = ISCSI_NP_THREAD_EXIT;
+ spin_unlock_bh(&np->np_thread_lock);
/*
* We need to send the signal to wakeup Linux/Net
* which may be sleeping in sock_accept()..
*/
send_sig(SIGINT, np->np_thread, 1);
kthread_stop(np->np_thread);
- }
+ } else
+ spin_unlock_bh(&np->np_thread_lock);
np->np_transport->iscsit_free_np(np);
diff --git a/drivers/target/iscsi/iscsi_target_login.c b/drivers/target/iscsi/iscsi_target_login.c
index 4eb93b2..b375d26 100644
--- a/drivers/target/iscsi/iscsi_target_login.c
+++ b/drivers/target/iscsi/iscsi_target_login.c
@@ -1405,7 +1405,8 @@ out:
stop = kthread_should_stop();
if (!stop && signal_pending(current)) {
spin_lock_bh(&np->np_thread_lock);
- stop = (np->np_thread_state == ISCSI_NP_THREAD_SHUTDOWN);
+ stop = (np->np_thread_state == ISCSI_NP_THREAD_SHUTDOWN ||
+ np->np_thread_state == ISCSI_NP_THREAD_EXIT);
spin_unlock_bh(&np->np_thread_lock);
}
/* Wait for another socket.. */
@@ -1414,8 +1415,10 @@ out:
exit:
iscsi_stop_login_thread_timer(np);
spin_lock_bh(&np->np_thread_lock);
- np->np_thread_state = ISCSI_NP_THREAD_EXIT;
- np->np_thread = NULL;
+ if (np->np_thread_state != ISCSI_NP_THREAD_EXIT) {
+ np->np_thread_state = ISCSI_NP_THREAD_EXIT;
+ np->np_thread = NULL;
+ }
spin_unlock_bh(&np->np_thread_lock);
return 0;
--
1.7.12.4
next reply other threads:[~2013-12-05 13:54 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-12-05 13:54 Hannes Reinecke [this message]
2013-12-11 23:44 ` [PATCH] iscsi_target: race condition on shutdown Nicholas A. Bellinger
2013-12-12 7:18 ` Hannes Reinecke
2013-12-12 7:54 ` Nicholas A. Bellinger
2013-12-12 8:05 ` Hannes Reinecke
2013-12-16 19:49 ` Nicholas A. Bellinger
2013-12-18 2:27 ` Nicholas A. Bellinger
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1386251665-39173-1-git-send-email-hare@suse.de \
--to=hare@suse.de \
--cc=linux-scsi@vger.kernel.org \
--cc=nab@daterainc.com \
--cc=target-devel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.