[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <9002989d-3d87-dee1-23a3-6a652243d5b0@amd.com>
Date: Tue, 22 Nov 2022 13:43:10 +0530
From: Ravi Bangoria <ravi.bangoria@....com>
To: peterz@...radead.org
Cc: Ian Rogers <irogers@...gle.com>, acme@...nel.org,
mark.rutland@....com, jolsa@...hat.com, namhyung@...nel.org,
eranian@...gle.com, kan.liang@...ux.intel.com, bp@...en8.de,
x86@...nel.org, linux-perf-users@...r.kernel.org,
linux-kernel@...r.kernel.org, sandipan.das@....com,
ananth.narayan@....com, santosh.shukla@....com,
Ravi Bangoria <ravi.bangoria@....com>
Subject: Re: [PATCH] perf/amd/ibs: Make IBS a core pmu
On 16-Nov-22 11:15 AM, Ravi Bangoria wrote:
> On 15-Nov-22 10:37 PM, Ian Rogers wrote:
>> On Tue, Nov 15, 2022 at 1:39 AM Ravi Bangoria <ravi.bangoria@....com> wrote:
>>>
>>> So far, only one pmu was allowed to be registered as core pmu and thus
>>> IBS pmus were being registered as uncore. However, with the event context
>>> rewrite, that limitation no longer exists and thus IBS pmus can also be
>>> registered as core pmu. This makes IBS much more usable, for ex, user
>>> will be able to do per-process precise monitoring on AMD:
>>>
>>> Before patch:
>>> $ sudo perf record -e cycles:pp ls
>>> Error:
>>> Invalid event (cycles:pp) in per-thread mode, enable system wide with '-a'
>>>
>>> After patch:
>>> $ sudo perf record -e cycles:pp ls
>>> [ perf record: Woken up 1 times to write data ]
>>> [ perf record: Captured and wrote 0.017 MB perf.data (33 samples) ]
>>>
>>> Signed-off-by: Ravi Bangoria <ravi.bangoria@....com>
>>
>> Acked-by: Ian Rogers <irogers@...gle.com>
>>
>> This is awesome!
>> Ian
>
> Thanks Ian.
>
> Peter, can we push this along with rewrite patch?
Gentle Ping, Peter!
Thanks,
Ravi
Powered by blists - more mailing lists