Re: [RFC][PATCH 0/3] perf fixes

2015-01-26 Thread Vince Weaver
On Mon, 26 Jan 2015, Jiri Olsa wrote: > I trigered the same one, should be addressed by the fix Peter > sent in earlier email ahh, yes. gmail (which runs my university's mail servers) is for some reason sending all e-mails from PeterZ to the spam folder. I'll try out that patch. Vince --

Re: [RFC][PATCH 0/3] perf fixes

2015-01-26 Thread Jiri Olsa
On Mon, Jan 26, 2015 at 01:03:47PM -0500, Vince Weaver wrote: > On Mon, 26 Jan 2015, Vince Weaver wrote: > > > I applied this to (just before) 3.19-rc6 and let it run on my > > troublemaking Haswell machine overnight. It is looking good! > > spoke too soon, managed to trigger one of the new war

Re: [RFC][PATCH 0/3] perf fixes

2015-01-26 Thread Vince Weaver
On Mon, 26 Jan 2015, Vince Weaver wrote: > I applied this to (just before) 3.19-rc6 and let it run on my > troublemaking Haswell machine overnight. It is looking good! spoke too soon, managed to trigger one of the new warnings you added. This is WARN_ON_ONCE(event->ctx != ctx); in even

Re: [RFC][PATCH 0/3] perf fixes

2015-01-26 Thread Vince Weaver
On Fri, 23 Jan 2015, Peter Zijlstra wrote: > Here are a few perf patches that (hopefully) fix a number of reported > (and at least one unreported afaik) issues as triggered by Vince's > fuzzer. > > After a few days of staring at that event->ctx mess and writing > increasingly horrible patches, I