lkml.org 
[lkml]   [2011]   [Aug]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH v2 6/6] perf tool: Parse general/raw events from sysfs
From
Date
On Mon, 2011-08-08 at 16:57 +0800, Lin Ming wrote:
> On Mon, 2011-08-08 at 16:48 +0800, Peter Zijlstra wrote:
> > On Mon, 2011-08-08 at 09:08 +0800, Lin Ming wrote:
> > >
> > > We only need to register one pmu with two same graphics cards.
> > > Then we can overload pid argument of sys_perf_event_open() to
> > > differentiate each instance of graphic card.
> > >
> > > Just like perf cgroup does.
> >
> > I don't like that much at all.. For one it will make it impossible to
> > put the pmu device in the right sysfs topology.
>
> So do you prefer adding a number to gpu name, as Stephane mentioned?

Not quite having read back that far (reading my email in date-reverse
manner this morning), yeah. That allows links from /sys/bus/pci/... back
to the pmu device.


\
 
 \ /
  Last update: 2011-08-08 11:07    [W:0.118 / U:0.108 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site