Re: [PATCH v2 3/7] perf, x86: Introduce x86 special perf event context

[Yan, Zheng]

On 07/05/2013 08:45 PM, Peter Zijlstra wrote:
> On Fri, Jul 05, 2013 at 11:19:31AM +0800, Yan, Zheng wrote:
>> On 07/04/2013 08:41 PM, Peter Zijlstra wrote:
>>> It should be *optional*, also wtf is that parent_ctx thing for?
>>
>> parent_ctx is for the fork() case, used for checking if the callstack feature
>> is enabled for the parent task. If yes, clone  parent task's LBR stack.
>> For the simple program below:
> 
> So there's a problem with all this; contexts aren't strict per task, we
> play games with them in perf_event_context_sched_out(). We'd have to
> disable that context switch optimization for LBR stack to work and
> that's massively expensive.
> 

Sorry, I didn't realize the optimization. I will describe my change in more details
in the further.

Yan, Zheng

> If you actually did that, you again fail for not having mentioned this
> in your changelog.
> 
> I'm starting to not want to read patches from you; going through them is
> just too much effort, I might as well write the stuff myself :/
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/