From:   "Zhang, Qiang1" <qiang1.zhang@intel.com>
To:     Peter Zijlstra <peterz@infradead.org>
CC:     "mingo@redhat.com" <mingo@redhat.com>,
        "vincent.guittot@linaro.org" <vincent.guittot@linaro.org>,
        "juri.lelli@redhat.com" <juri.lelli@redhat.com>,
        "dietmar.eggemann@arm.com" <dietmar.eggemann@arm.com>,
        "rostedt@goodmis.org" <rostedt@goodmis.org>,
        "vschneid@redhat.com" <vschneid@redhat.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: RE: [PATCH] sched/fair: Make hrtimer callback execute in the hard
 interrupt context
Thread-Topic: [PATCH] sched/fair: Make hrtimer callback execute in the hard
 interrupt context
Thread-Index: AQHZT/n3qfMUi8dWmUq1jZFZUkYwxq7tl1uAgAA3xCA=
Date:   Mon, 6 Mar 2023 14:27:41 +0000
Message-ID: <PH0PR11MB5880550ED109DC115FC0EE17DAB69@PH0PR11MB5880.namprd11.prod.outlook.com>
References: <20230306070938.3099273-1-qiang1.zhang@intel.com>
 <20230306110609.GA1267364@hirez.programming.kicks-ass.net>
In-Reply-To: <20230306110609.GA1267364@hirez.programming.kicks-ass.net>
Accept-Language: en-US
Content-Language: en-US
x-ms-exchange-antispam-messagedata-chunkcount: 1
x-ms-exchange-antispam-messagedata-0: =?us-ascii?Q?K4ssMxBqim75eVDMLcMknit7fB6Ho31OPliLOS5ydfekSq/tgK2Dca+J6sq0?=
 =?us-ascii?Q?Z7GlPSiP+mSPIm+LzshYUXBF7ipZwpLjYO9Sr4IBSUY1h8mMO6lkHyFxD/PF?=
 =?us-ascii?Q?V00SqvJnOsO9DrLletK1QosHajmdg2Nvq78mcnAzxhXMOQjOZfowehksdGec?=
 =?us-ascii?Q?sl11j3f8v99aTxcIJ7ua+MtdHU17pWyWWCP8gZSErG4pZdxH7MQXG1TgroEK?=
 =?us-ascii?Q?CynYV61Xzilpl5UqGmZqRFKC7u1UNCopBypdB/NBGrTkSdvFtvnPgnA7/BvG?=
 =?us-ascii?Q?xjOtr7o+Vhc6/2lOVtbfZ4Hjiyk2A3r+bR8cJ78WC3LdjAA1Po8hvjqHZORR?=
 =?us-ascii?Q?6nRQu29DsKMuYo1SZMKNVtyTo0bwKO7Md9p5vigDqzWXKqxBichZLN4yPv3H?=
 =?us-ascii?Q?bqxQG41Sm2SQOsJZqGG7otpUqbmOTpO4/qiW0ieuvegB0qun6TGJdqq6MsOb?=
 =?us-ascii?Q?z1muCDWu/kQtpLexi2tUuNkCDk6wLpi5ctS1fZSVL/zs7h2pGfwsPk+aJvab?=
 =?us-ascii?Q?nW2wVI5kReDE7BH/3fMwM9OHLD+ebWphe0IDK76J2LjU1Y9LOEc7+hZtFSZD?=
 =?us-ascii?Q?cXxXa6sduTBgO+Bvnsdv98XSjGzdViW4Qo6NOzZei5JuaF7vEPE4OhkEBfLl?=
 =?us-ascii?Q?m1c5yXRHP7rGRkQHr/OXKj2HKgUMM1DK+RR2I0cI4Ma9r8CkiQ7kr9SAK1Li?=
 =?us-ascii?Q?5LfLt8/5w7rkQ18CqSQW82oC1VeDwfzK1DKiUJnARKW+x0L/TMDXMfErm4RB?=
 =?us-ascii?Q?Lfk9jExGeIXwVlpcxxllJxCViG8+RHpr3p8OZMGukS3KTsM9Vr7o1xqaX2Oq?=
 =?us-ascii?Q?p1/qZdY3p/ugLURPJ8Q14VHBaNEh3+hTnDMhMWEgFH0FIIdsZHD7ptHjsnXG?=
 =?us-ascii?Q?uPGQhrD0AvkWMB9jWcvQZA/9bY13fz15xKdhJwav5x+vqq25yu5x+k/xnGz4?=
 =?us-ascii?Q?AzSEKnmwUYovsPOUvNJcUp1bxtz80p+K8Qvz0olfx0qEPs9/rUXBfoZuCdin?=
 =?us-ascii?Q?Bg0v5T6a+RTtr9zQbVpvjc3s4V0fabHqBjNc8wBBHV1bfLZm8JlZq/t6YQs4?=
 =?us-ascii?Q?M8zlctUm5Kcoyidnr3IL9aU0cbXILFAFSR1JeXdZ17DGB3PgyW7k6Mhj1GRC?=
 =?us-ascii?Q?szU84n9AG2gwsFqeH2+CWKtKNSzObRQbhIjKGZ2C4R+zzBzrowu1tsjeW229?=
 =?us-ascii?Q?3DAi6jnQpgPnsH0dSp6KJOzXAyy1xS488n4TTrD6kdhshnsjikZh2jVkGntc?=
 =?us-ascii?Q?EMs6Rq95Ls12Xg5QC91zzhoYshUNbSyiDSUeSwo4BA/CwHuztqnAwK2ltVjj?=
 =?us-ascii?Q?uPvH2bnoOfuWTYDvZVYoYhxkHBafAZF+wAekcW+FGo0a1aG52/9fC0VAPdNy?=
 =?us-ascii?Q?+7br8iwnsXA6+317u09cPiAV80VbKF6fSmR6VdkmVPROXjq8vawCn2RMS2Pl?=
 =?us-ascii?Q?xJqrNu/5DYJ9tETDhEzhPzA5AneVZr8/hZIdCrGFMMSZjiH404cn2aUdWtCD?=
 =?us-ascii?Q?0iyIteO5AGGNdc5D0OdQ1p9FsxIqzYQCqGWvN8NZ3tcza9yR+79B1BBNDZ5T?=
 =?us-ascii?Q?SgovdLGcejFBYuhN1TfGrtotrRAWZ+YJc/GCzzcN?=
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: PH0PR11MB5880.namprd11.prod.outlook.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 85ce521b-deb3-40c3-1021-08db1e4ef182
X-MS-Exchange-CrossTenant-originalarrivaltime: 06 Mar 2023 14:27:41.0547
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 46c98d88-e344-4ed4-8496-4ed7712e255d
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: NAmdv2qXVxq9uPxIY4Pb9q9ViRtomKuS30+4B+UNNiD5dM9IgO7i7jLcZBCjZzKsTdtVcm4VjJ9Rckkih6BHyg==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: MW4PR11MB6839
X-OriginatorOrg: intel.com
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org


> The scheduler related hrtimers callback will be executed in softirqd=20
> context for PREEMPT_RT enabled kernel, this commit therefore mark=20
> hrtimers as harded even on PREEMPT_RT enabled kernels, there is no=20
> functional change.
>
>There very much is a functional change, also your reasoning is non-existan=
t.
>
>Also:

Thanks for  reply,  I will reorganize the reason and resend

Thanks
Zqiang

>
>  https://lkml.kernel.org/r/Y1F5sDVGen7ZVW+U@hirez.programming.kicks-ass.n=
et
>
> Signed-off-by: Zqiang <qiang1.zhang@intel.com>
> ---
>  kernel/sched/fair.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
>=20
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index=20
> c36aa54ae071..98c48d144089 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -5610,7 +5610,7 @@ static void start_cfs_slack_bandwidth(struct=20
> cfs_bandwidth *cfs_b)
> =20
>  	hrtimer_start(&cfs_b->slack_timer,
>  			ns_to_ktime(cfs_bandwidth_slack_period),
> -			HRTIMER_MODE_REL);
> +			HRTIMER_MODE_REL_HARD);
>  }
> =20
>  /* we know any runtime found here is valid as update_curr() precedes=20
> return */ @@ -5813,9 +5813,9 @@ void init_cfs_bandwidth(struct cfs_bandwi=
dth *cfs_b)
>  	cfs_b->burst =3D 0;
> =20
>  	INIT_LIST_HEAD(&cfs_b->throttled_cfs_rq);
> -	hrtimer_init(&cfs_b->period_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PI=
NNED);
> +	hrtimer_init(&cfs_b->period_timer, CLOCK_MONOTONIC,=20
> +HRTIMER_MODE_ABS_PINNED_HARD);
>  	cfs_b->period_timer.function =3D sched_cfs_period_timer;
> -	hrtimer_init(&cfs_b->slack_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
> +	hrtimer_init(&cfs_b->slack_timer, CLOCK_MONOTONIC,=20
> +HRTIMER_MODE_REL_HARD);
>  	cfs_b->slack_timer.function =3D sched_cfs_slack_timer;
>  	cfs_b->slack_started =3D false;
>  }
> @@ -5835,7 +5835,7 @@ void start_cfs_bandwidth(struct cfs_bandwidth=20
> *cfs_b)
> =20
>  	cfs_b->period_active =3D 1;
>  	hrtimer_forward_now(&cfs_b->period_timer, cfs_b->period);
> -	hrtimer_start_expires(&cfs_b->period_timer, HRTIMER_MODE_ABS_PINNED);
> +	hrtimer_start_expires(&cfs_b->period_timer,=20
> +HRTIMER_MODE_ABS_PINNED_HARD);
>  }
> =20
>  static void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
> --
> 2.25.1
>=20