On Thu, Sep 10, 2015 at 01:54:18AM -0700, Stephane Eranian wrote:
> On Fri, Aug 21, 2015 at 1:31 PM, Sasha Levin wrote:
> >
> > On 05/21/2015 07:17 AM, Peter Zijlstra wrote:
> > > --- a/arch/x86/kernel/cpu/perf_event_intel.c
> > > +++ b/arch/x86/kernel/cpu/perf_event_intel.c
> > > @@ -2106,7 +2106
On Fri, Aug 21, 2015 at 1:31 PM, Sasha Levin wrote:
>
> On 05/21/2015 07:17 AM, Peter Zijlstra wrote:
> > --- a/arch/x86/kernel/cpu/perf_event_intel.c
> > +++ b/arch/x86/kernel/cpu/perf_event_intel.c
> > @@ -2106,7 +2106,7 @@ static struct event_constraint *
> > intel_get_event_constraints(struct
Ping?
On 08/21/2015 04:31 PM, Sasha Levin wrote:
> On 05/21/2015 07:17 AM, Peter Zijlstra wrote:
>> --- a/arch/x86/kernel/cpu/perf_event_intel.c
>> +++ b/arch/x86/kernel/cpu/perf_event_intel.c
>> @@ -2106,7 +2106,7 @@ static struct event_constraint *
>> intel_get_event_constraints(struct cpu_hw_e
On 05/21/2015 07:17 AM, Peter Zijlstra wrote:
> --- a/arch/x86/kernel/cpu/perf_event_intel.c
> +++ b/arch/x86/kernel/cpu/perf_event_intel.c
> @@ -2106,7 +2106,7 @@ static struct event_constraint *
> intel_get_event_constraints(struct cpu_hw_events *cpuc, int idx,
> struct
* Stephane Eranian wrote:
> On Thu, May 21, 2015 at 11:49 PM, Ingo Molnar wrote:
> >
> >
> > * Stephane Eranian wrote:
> >
> > > On Thu, May 21, 2015 at 7:03 AM, Peter Zijlstra
> > > wrote:
> > > > On Thu, 2015-05-21 at 06:36 -0700, Stephane Eranian wrote:
> > > >> On Thu, May 21, 2015 at 6:
On Thu, May 21, 2015 at 11:49 PM, Ingo Molnar wrote:
>
>
> * Stephane Eranian wrote:
>
> > On Thu, May 21, 2015 at 7:03 AM, Peter Zijlstra
> > wrote:
> > > On Thu, 2015-05-21 at 06:36 -0700, Stephane Eranian wrote:
> > >> On Thu, May 21, 2015 at 6:29 AM, Peter Zijlstra
> > >> wrote:
> > >> >
* Stephane Eranian wrote:
> On Thu, May 21, 2015 at 7:03 AM, Peter Zijlstra wrote:
> > On Thu, 2015-05-21 at 06:36 -0700, Stephane Eranian wrote:
> >> On Thu, May 21, 2015 at 6:29 AM, Peter Zijlstra
> >> wrote:
> >> > On Thu, 2015-05-21 at 06:27 -0700, Stephane Eranian wrote:
> >> >> Or are y
On Thu, May 21, 2015 at 7:53 AM, Peter Zijlstra wrote:
> On Thu, May 21, 2015 at 01:17:11PM +0200, Peter Zijlstra wrote:
>> --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c
>> +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c
>> @@ -706,9 +706,9 @@ void intel_pmu_pebs_disable(struct perf_
>>
>>
On Thu, May 21, 2015 at 7:03 AM, Peter Zijlstra wrote:
> On Thu, 2015-05-21 at 06:36 -0700, Stephane Eranian wrote:
>> On Thu, May 21, 2015 at 6:29 AM, Peter Zijlstra wrote:
>> > On Thu, 2015-05-21 at 06:27 -0700, Stephane Eranian wrote:
>> >> Or are you talking about a preemption while executing
On Thu, May 21, 2015 at 01:17:11PM +0200, Peter Zijlstra wrote:
> --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c
> +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c
> @@ -706,9 +706,9 @@ void intel_pmu_pebs_disable(struct perf_
>
> cpuc->pebs_enabled &= ~(1ULL << hwc->idx);
>
> - if (e
On Thu, 2015-05-21 at 06:36 -0700, Stephane Eranian wrote:
> On Thu, May 21, 2015 at 6:29 AM, Peter Zijlstra wrote:
> > On Thu, 2015-05-21 at 06:27 -0700, Stephane Eranian wrote:
> >> Or are you talking about a preemption while executing
> >> x86_schedule_events()?
> >
> > That.
> >
> > And we ca
On Thu, May 21, 2015 at 6:29 AM, Peter Zijlstra wrote:
> On Thu, 2015-05-21 at 06:27 -0700, Stephane Eranian wrote:
>> Or are you talking about a preemption while executing x86_schedule_events()?
>
> That.
>
> And we can of course cure that by an earlier patch I send; but I find it
> a much simple
On Thu, 2015-05-21 at 06:27 -0700, Stephane Eranian wrote:
> Or are you talking about a preemption while executing x86_schedule_events()?
That.
And we can of course cure that by an earlier patch I send; but I find it
a much simpler rule to just never allow modifying global state for
validation.
-
On Thu, May 21, 2015 at 6:20 AM, Peter Zijlstra wrote:
> On Thu, May 21, 2015 at 06:18:15AM -0700, Stephane Eranian wrote:
>> Yes, it does modify the cpuc->event_list[]->hwc, because it is used as a
>> cache for *EACH* invocation of the function. It is irrelevant outside the
>> function.
>
> Yes,
On Thu, May 21, 2015 at 06:18:15AM -0700, Stephane Eranian wrote:
> Yes, it does modify the cpuc->event_list[]->hwc, because it is used as a
> cache for *EACH* invocation of the function. It is irrelevant outside the
> function.
Yes, but the problem is that they _nest_.
So you get to use the one
On Thu, May 21, 2015 at 6:09 AM, Peter Zijlstra wrote:
> On Thu, May 21, 2015 at 06:07:20AM -0700, Stephane Eranian wrote:
>> On Thu, May 21, 2015 at 5:56 AM, Peter Zijlstra wrote:
>> > On Thu, May 21, 2015 at 05:35:02AM -0700, Stephane Eranian wrote:
>> >> > Commit e979121b1b15 ("perf/x86/intel:
On Thu, May 21, 2015 at 06:07:20AM -0700, Stephane Eranian wrote:
> On Thu, May 21, 2015 at 5:56 AM, Peter Zijlstra wrote:
> > On Thu, May 21, 2015 at 05:35:02AM -0700, Stephane Eranian wrote:
> >> > Commit e979121b1b15 ("perf/x86/intel: Implement cross-HT corruption
> >> > bug workaround") made t
On Thu, May 21, 2015 at 5:56 AM, Peter Zijlstra wrote:
> On Thu, May 21, 2015 at 05:35:02AM -0700, Stephane Eranian wrote:
>> > Commit e979121b1b15 ("perf/x86/intel: Implement cross-HT corruption
>> > bug workaround") made the situation much worse by actually setting the
>> > event->hw.constraint
On Thu, May 21, 2015 at 05:35:02AM -0700, Stephane Eranian wrote:
> Peter,
>
> On Thu, May 21, 2015 at 4:17 AM, Peter Zijlstra wrote:
> > Commit 43b4578071c0 ("perf/x86: Reduce stack usage of
> > x86_schedule_events()") violated the rule that 'fake' scheduling; as
> > used for event/group validat
Peter,
On Thu, May 21, 2015 at 4:17 AM, Peter Zijlstra wrote:
> Commit 43b4578071c0 ("perf/x86: Reduce stack usage of
> x86_schedule_events()") violated the rule that 'fake' scheduling; as
> used for event/group validation; should not change the event state.
>
> This went mostly un-noticed becaus
Commit 43b4578071c0 ("perf/x86: Reduce stack usage of
x86_schedule_events()") violated the rule that 'fake' scheduling; as
used for event/group validation; should not change the event state.
This went mostly un-noticed because repeated calls of
x86_pmu::get_event_constraints() would give the same
21 matches
Mail list logo