[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <616a4f00-dd1f-20b2-a228-3fa9d7391016@linux.intel.com>
Date: Thu, 20 Feb 2020 14:25:03 -0500
From: "Liang, Kan" <kan.liang@...ux.intel.com>
To: Andi Kleen <ak@...ux.intel.com>
Cc: Jiri Olsa <jolsa@...hat.com>, acme@...nel.org, mingo@...hat.com,
peterz@...radead.org, linux-kernel@...r.kernel.org,
mark.rutland@....com, namhyung@...nel.org,
ravi.bangoria@...ux.ibm.com, yao.jin@...ux.intel.com
Subject: Re: [PATCH 0/5] Support metric group constraint
On 2/20/2020 11:43 AM, Andi Kleen wrote:
>> For other metric groups, even they have cycles, the issue should not be
>> triggered.
>> For example, if they have 4 or less events, the cycles can be scheduled to
>> GP counter instead.
>> If they have 6 or more events, the weak group will be reject anyway.
>> Perf tool will open it as non-group (standalone metrics).
>
> Technically it can also happen for 9 events with Hyper Threading off or
> on Icelake (8 generic counters)
>
> I didn't think we had any of those, but please double check.
>
I checked all public metrics groups. Right, we don't have such metrics
group with 8 GP events + 1 cycles.
We only need to add watchdog constraint for Page_Walks_Utilization for now.
Thanks,
Kan
Powered by blists - more mailing lists