On Thu, Jun 14, 2012 at 10:13 PM, Luck, Tony <tony.luck@xxxxxxxxx> wrote: >> Maybe you can try my test-case at https://lkml.org/lkml/2012/3/31/90 ? > > There seem to be a few cut & paste errors in that version (threads 3, 4, 5 > think they are "Thread 2"). Also no check for malloc failing (which since > it is done in an infinite loop seems inevitable - unless your machines have > infinite memory :-) Yes. This test-case was written in a hurry to run on ARM and the test-case was just meant to blindly allocate and fault pages till OOM happens. So sorry about that ! Evidently on ARM we OOM much before the mallocs start failing. That maybe due to some overcommit_memory setting which I had on that board. > > Fixed up version attached. Seems to run without any lockups on a kernel > with your patch applied until either the malloc's start to fail, or the OOM > killer zaps it. Thank you for that. Really appreciate your help ! > > I put your patch into my "next" tree - it's in linux-next tag "next-20120614" > and I'll push to Linus in the next merge window. And thanks again for that. :-) > > Thanks > > -Tony > > -- To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
|