* [perf] exporting perf_pmu_register?
@ 2012-03-15 21:41 Vince Weaver
2012-06-06 10:41 ` Peter Zijlstra
0 siblings, 1 reply; 3+ messages in thread
From: Vince Weaver @ 2012-03-15 21:41 UTC (permalink / raw)
To: linux-kernel; +Cc: mingo, a.p.zijlstra, paulus, acme
Hello
I'm attempting to write a perf pmu driver as a kernel module, but
am finding this difficult to do it properly because
perf_pmu_register
is not exported.
Would a patch similar to
https://lkml.org/lkml/2010/8/18/64
be accepted that exported this?
I'm trying to write a RAPL/energy pmu module on a Sandybridge-EP machine
but the sysadmin won't install a custom kernel so I'm stuck trying to do
it entirely via loadable module, and it turns out this doesn't seem to be
possible at all :(
Vince
vweaver1@eecs.utk.edu
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [perf] exporting perf_pmu_register?
2012-03-15 21:41 [perf] exporting perf_pmu_register? Vince Weaver
@ 2012-06-06 10:41 ` Peter Zijlstra
2012-06-06 14:30 ` Vince Weaver
0 siblings, 1 reply; 3+ messages in thread
From: Peter Zijlstra @ 2012-06-06 10:41 UTC (permalink / raw)
To: Vince Weaver; +Cc: linux-kernel, mingo, paulus, acme
On Thu, 2012-03-15 at 17:41 -0400, Vince Weaver wrote:
> Would a patch similar to
> https://lkml.org/lkml/2010/8/18/64
> be accepted that exported this?
>
If it comes with a patch that uses the exports, sure.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [perf] exporting perf_pmu_register?
2012-06-06 10:41 ` Peter Zijlstra
@ 2012-06-06 14:30 ` Vince Weaver
0 siblings, 0 replies; 3+ messages in thread
From: Vince Weaver @ 2012-06-06 14:30 UTC (permalink / raw)
To: Peter Zijlstra; +Cc: linux-kernel, mingo, paulus, acme
On Wed, 6 Jun 2012, Peter Zijlstra wrote:
> On Thu, 2012-03-15 at 17:41 -0400, Vince Weaver wrote:
> > Would a patch similar to
> > https://lkml.org/lkml/2010/8/18/64
> > be accepted that exported this?
> >
> If it comes with a patch that uses the exports, sure.
>
while I understand the logic behind that from a kernel-developer
perspective, it does make things harder in the real world.
In the end PAPI has implemented RAPL support via reading /dev/cpu/*/msr
because it was easier to convince our sysadmin to give us read-only
permission to /dev/cpu/*/msr than to take our in-production Sandybridge-EP
system and patch it up to export perf_pmu_register so we could work on a
proper driver.
Vince
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2012-06-06 14:31 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-03-15 21:41 [perf] exporting perf_pmu_register? Vince Weaver
2012-06-06 10:41 ` Peter Zijlstra
2012-06-06 14:30 ` Vince Weaver
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox