Re: [perfmon2] timeout-based multiplexing changes
Status: Beta
Brought to you by:
seranian
From: stephane e. <er...@go...> - 2008-01-31 20:42:40
|
Corey, On Jan 31, 2008 8:45 PM, Corey Ashford <cja...@us...> wrote: > Stephane Eranian wrote: > > PS: The 2.6.24 kernel patch is a bit delayed waiting for one last > > patch from IBM and because I'd like to > > get this timeout- problem solved. > > I assume this means the patch that fixes the "lazy interrupt disabling". I apologize > that between some other projects and my development machine having some problems that it's > taken longer than I wanted to create and test the patch we discussed. > Not a problem. I have lots of other things to do at the same time. > I'm also putting together a few patches to the code that correct some of the bugs > we found in POWER5 in the POWER4 and POWER6 code as well. > Ok, then I'll wait for those as well. > Have you figured out which solution to use for the hrtimer and run queue locks? > I have pushed something into the GIT tree already. Please try it out on Power and make sure you have CONFIG_HIGH_RES_TIMERS enabled. |