Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Song Liu <songliubraving@fb.com>
To:     Namhyung Kim <namhyung@kernel.org>
CC:     Peter Zijlstra <peterz@infradead.org>,
        Ingo Molnar <mingo@kernel.org>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Jiri Olsa <jolsa@redhat.com>,
        Mark Rutland <mark.rutland@arm.com>,
        Alexander Shishkin <alexander.shishkin@linux.intel.com>,
        LKML <linux-kernel@vger.kernel.org>,
        Stephane Eranian <eranian@google.com>,
        Andi Kleen <ak@linux.intel.com>,
        "Ian Rogers" <irogers@google.com>, Tejun Heo <tj@kernel.org>
Subject: Re: [PATCH 1/2] perf/core: Share an event with multiple cgroups
Thread-Topic: [PATCH 1/2] perf/core: Share an event with multiple cgroups
Thread-Index: AQHXIACsCyQjF284/Uq+advPGrds7aqZrFQAgAEyVQCAAT52gA==
Date:   Tue, 30 Mar 2021 06:33:16 +0000
Message-ID: <8B2062CE-9BD5-4F17-B831-38C9274A1509@fb.com>
References: <20210323162156.1340260-1-namhyung@kernel.org>
 <20210323162156.1340260-2-namhyung@kernel.org>
 <180E871A-7C5E-4DB2-80F4-643237A7FA69@fb.com>
 <CAM9d7cg6-D2G7ii0sEJoEynvvmXUz=QsqjXG5K7j947DL1PymQ@mail.gmail.com>
In-Reply-To: <CAM9d7cg6-D2G7ii0sEJoEynvvmXUz=QsqjXG5K7j947DL1PymQ@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
Content-Type: text/plain; charset="us-ascii"
Content-ID: <A7090FB691E6E143A432879816F2187B@namprd15.prod.outlook.com>
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-Network-Message-Id: 32abc2c4-0f90-469c-a775-08d8f345b3e6
X-MS-Exchange-CrossTenant-originalarrivaltime: 30 Mar 2021 06:33:16.8766
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 8ae927fe-1255-47a7-a2af-5f3a069daaa2
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: 35uVJKEbECZMfYSrjEz5JVbNrHgt/FPJ1vRoa6J8sWCHcW7ztk/z5FbYWvsTP0Ptg+6vxncqwE/NAcHQQj0MIw==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BYAPR15MB2583
X-OriginatorOrg: fb.com
X-Proofpoint-GUID: JHgFNZkL_qFfU6dOr0mojEp5XwQNNcz2
X-Proofpoint-ORIG-GUID: JHgFNZkL_qFfU6dOr0mojEp5XwQNNcz2
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369,18.0.761
 definitions=2021-03-30_01:2021-03-26,2021-03-30 signatures=0
X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 suspectscore=0
 malwarescore=0 clxscore=1015 mlxlogscore=999 impostorscore=0 bulkscore=0
 lowpriorityscore=0 spamscore=0 priorityscore=1501 mlxscore=0 phishscore=0
 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1
 engine=8.12.0-2103250000 definitions=main-2103300045
X-FB-Internal: deliver
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org


> On Mar 29, 2021, at 4:33 AM, Namhyung Kim <namhyung@kernel.org> wrote:
>=20
> On Mon, Mar 29, 2021 at 2:17 AM Song Liu <songliubraving@fb.com> wrote:
>>> On Mar 23, 2021, at 9:21 AM, Namhyung Kim <namhyung@kernel.org> wrote:
>>>=20
>>> As we can run many jobs (in container) on a big machine, we want to
>>> measure each job's performance during the run.  To do that, the
>>> perf_event can be associated to a cgroup to measure it only.
>>>=20

[...]

>>> +     return 0;
>>> +}
>>=20
>> Could you please explain why we need this logic in can_attach?
>=20
> IIUC the ss->attach() is called after a task's cgroup membership
> is changed.  But we want to collect the performance numbers for
> the old cgroup just before the change.  As the logic merely checks
> the current task's cgroup, it should be done in the can_attach()
> which is called before the cgroup change.

Thanks for the explanations.=20

Overall, I really like the core idea, especially that the overhead on=20
context switch is bounded (by the depth of cgroup tree).=20

Is it possible to make PERF_EVENT_IOC_ATTACH_CGROUP more flexible?=20
Specifically, if we can have
=20
  PERF_EVENT_IOC_ADD_CGROUP	add a cgroup to the list=20
  PERF_EVENT_IOC_EL_CGROUP	delete a cgroup from the list

we can probably share these events among multiple processes, and=20
these processes don't need to know others' cgroup list. I think=20
this will be useful for users to build customized monitoring in=20
its own container.=20

Does this make sense?

Thanks,
Song