Messages in this thread | | | Date | Sun, 24 Mar 2013 20:46:38 +0700 | Subject | Re: ipc,sem: sysv semaphore scalability | From | Emmanuel Benisty <> |
| |
On Sun, Mar 24, 2013 at 2:45 AM, Linus Torvalds <torvalds@linux-foundation.org> wrote: > On Fri, Mar 22, 2013 at 8:19 PM, Emmanuel Benisty <benisty.e@gmail.com> wrote: >> >> I could reproduce it but could you please let me know what would be >> the right tools I should use to catch the original oops? >> This is what I got but I doubt it will be helpful: >> http://i.imgur.com/Mewi1hC.jpg > > In this case, I think the best thing to do would be to comment out all > of drm_warn_on_modeset_not_all_locked(), because those warnings make > the original problem (that probably caused the lock problem in the > first place that it is warning about) scroll away. > > That said, you should also take the oneliner fix that Rik posted to > patch 7 (subject line: "[PATCH 7/7 part3] fix for sem_lock") and apply > that, just to make sure that you aren't possibly hitting a bug with > the shared-memory locking introduced by that (unusual) case.
Thanks Linus. I hope I got this right, here's the result (3.9-rc4, 7+1 patches): http://i.imgur.com/BebGZxV.jpg
| |