Skip to content

Commit

Permalink
ipmi: Avoid touching possible corrupted lists in the panic context
Browse files Browse the repository at this point in the history
When processing queued messages in the panic context, IPMI driver
tries to do it without any locking to avoid deadlocks.  However,
this means we can touch a corrupted list if the kernel panicked
while manipulating the list.  Fortunately, current `add-tail and
del-from-head' style implementation won't touch the corrupted part,
but it is inherently risky.

To get rid of the risk, this patch re-initializes the message lists
on panic if the related spinlock has already been acquired.  As the
result, we may lose queued messages, but it's not so painful.
Dropping messages on the received message list is also less
problematic because no one can respond the received messages.

Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>

Fixed a comment typo.

Signed-off-by: Corey Minyard <cminyard@mvista.com>
  • Loading branch information
Hidehiro Kawai authored and Corey Minyard committed Sep 3, 2015
1 parent 82802f9 commit 06e5e34
Showing 1 changed file with 17 additions and 0 deletions.
17 changes: 17 additions & 0 deletions drivers/char/ipmi/ipmi_msghandler.c
Original file line number Diff line number Diff line change
Expand Up @@ -4511,6 +4511,23 @@ static int panic_event(struct notifier_block *this,
/* Interface is not ready. */
continue;

/*
* If we were interrupted while locking xmit_msgs_lock or
* waiting_rcv_msgs_lock, the corresponding list may be
* corrupted. In this case, drop items on the list for
* the safety.
*/
if (!spin_trylock(&intf->xmit_msgs_lock)) {
INIT_LIST_HEAD(&intf->xmit_msgs);
INIT_LIST_HEAD(&intf->hp_xmit_msgs);
} else
spin_unlock(&intf->xmit_msgs_lock);

if (!spin_trylock(&intf->waiting_rcv_msgs_lock))
INIT_LIST_HEAD(&intf->waiting_rcv_msgs);
else
spin_unlock(&intf->waiting_rcv_msgs_lock);

intf->run_to_completion = 1;
intf->handlers->set_run_to_completion(intf->send_info, 1);
}
Expand Down

0 comments on commit 06e5e34

Please sign in to comment.