Messages in this thread | | | Date | Tue, 12 Jun 2012 17:15:32 +0400 | Subject | Re: [RFC] [PATCH 0/5] Teach perf tool to profile sleep times (V4) | From | Andrew Wagin <> |
| |
Hello,
> So I really don't like this stuff either.. :-) > > Mostly because it adds very similar stuff to what's already there and > then doesn't try and consolidate things. > > And partly because you're measuring the 'wrong' thing, the delay between > sched-out and sched-in includes a number of things, only one part of > that is the delay due to blocking. > > And as a result you need that hideous prev_state crap. >
Yes, you are right.
> > > Would something like the below work?
Yes, I like it. Are you going to commit this? Thank for this work.
| |