On Tue, Oct 2, 2018 at 1:42 AM Petr Mladek pmladek@suse.com wrote:
Well, I still wonder why it helped and why you do not see it with 4.4. I have a feeling that the console owner switch helped only by chance.
So do I. I don't think Steven had the deadlock in mind when working on that patch, but with that patch and that patch alone, the deadlock disappeared.
In fact, you might be affected by a race in printk_safe_flush_on_panic() that was fixed by the commit:
554755be08fba31c7 printk: drop in_nmi check from printk_safe_flush_on_panic()
The above one commit might be enough. Well, there was one more
Thanks for the pointer. Let me test this out.
NMI-related race that was fixed by:
ba552399954dde1b printk: Split the code for storing a message into the log buffer a338f84dc196f44b printk: Create helper function to queue deferred console handling 03fc7f9c99c1e7ae printk/nmi: Prevent deadlock when accessing the main log buffer in NMI
Best Regards, Petr
On Tue, Oct 2, 2018 at 1:17 AM Sergey Senozhatsky sergey.senozhatsky.work@gmail.com wrote:
Sorry, I'm missing something here. Steven's patch deals with lockups and I understand why you want to backport the patch set; but console output deadlock on panic() is another thing.
Understood. But it did fix the deadlock for me and without it I hit the deadlock pretty consistently. Let me test the patch Petr pointed out and will update here.