On Tue, Aug 9, 2011 at 2:59 PM, Lennert Buytenhek buytenh@wantstofly.org wrote:
On Tue, Aug 09, 2011 at 02:49:42PM +0100, Niels de Vos wrote:
= There is a new schedulear/timer patch from Lennert -- mystery minutes-long hangs should be gone for real now.
Do you have any references like a git-commit or discussion on a mailinglist for this? I'd like to match this with some other ARM hardware (non/OLPC) issues I have seen.
There's a description of our problem at:
http://dev.laptop.org/git/olpc-kernel/commit/?h=arm-3.0&id=b8923ed0809f2...
Nice description! As it's mach-mmp dependent, it's unlikely that I can apply this for mx51 based systems. The information in the git-commit is clear and I'll try to investigate if 'sporadic hangs' could be caused due to missing events. (Which sounds sane, as it might happen when the system is under high load.)
Looking at arch/arm/plat-mxc/time.c and its *_set_next_event() functions, it appears to correctly handle the case where the timeout has already expired, so it's not likely that mx51 is suffering from the same problem.
Wow, thats a swift reply! Thanks for checking :) Niels