linux-kernel - Re: [PATCH v2 3/6] sched/fair: Remove spurious shorter slice preemption

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <202507052215.5HznxNGI-lkp@intel.com>
Date: Sat, 5 Jul 2025 22:15:31 +0800
From: kernel test robot <lkp@...el.com>
To: Vincent Guittot <vincent.guittot@...aro.org>, mingo@...hat.com,
	peterz@...radead.org, juri.lelli@...hat.com,
	dietmar.eggemann@....com, rostedt@...dmis.org, bsegall@...gle.com,
	mgorman@...e.de, vschneid@...hat.com, dhaval@...nis.ca,
	linux-kernel@...r.kernel.org
Cc: oe-kbuild-all@...ts.linux.dev,
	Vincent Guittot <vincent.guittot@...aro.org>
Subject: Re: [PATCH v2 3/6] sched/fair: Remove spurious shorter slice
 preemption

Hi Vincent,

kernel test robot noticed the following build warnings:

[auto build test WARNING on tip/sched/core]
[also build test WARNING on peterz-queue/sched/core linus/master v6.16-rc4 next-20250704]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Vincent-Guittot/sched-fair-Use-protect_slice-instead-of-direct-comparison/20250704-223850
base:   tip/sched/core
patch link:    https://lore.kernel.org/r/20250704143612.998419-4-vincent.guittot%40linaro.org
patch subject: [PATCH v2 3/6] sched/fair: Remove spurious shorter slice preemption
config: s390-randconfig-r072-20250705 (https://download.01.org/0day-ci/archive/20250705/202507052215.5HznxNGI-lkp@intel.com/config)
compiler: s390-linux-gcc (GCC) 11.5.0

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@...el.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202507052215.5HznxNGI-lkp@intel.com/

smatch warnings:
kernel/sched/fair.c:8721 check_preempt_wakeup_fair() warn: inconsistent indenting

vim +8721 kernel/sched/fair.c

  8643	
  8644	/*
  8645	 * Preempt the current task with a newly woken task if needed:
  8646	 */
  8647	static void check_preempt_wakeup_fair(struct rq *rq, struct task_struct *p, int wake_flags)
  8648	{
  8649		struct task_struct *donor = rq->donor;
  8650		struct sched_entity *se = &donor->se, *pse = &p->se;
  8651		struct cfs_rq *cfs_rq = task_cfs_rq(donor);
  8652		int cse_is_idle, pse_is_idle;
  8653		bool do_preempt_short = false;
  8654	
  8655		if (unlikely(se == pse))
  8656			return;
  8657	
  8658		/*
  8659		 * This is possible from callers such as attach_tasks(), in which we
  8660		 * unconditionally wakeup_preempt() after an enqueue (which may have
  8661		 * lead to a throttle).  This both saves work and prevents false
  8662		 * next-buddy nomination below.
  8663		 */
  8664		if (unlikely(throttled_hierarchy(cfs_rq_of(pse))))
  8665			return;
  8666	
  8667		if (sched_feat(NEXT_BUDDY) && !(wake_flags & WF_FORK) && !pse->sched_delayed) {
  8668			set_next_buddy(pse);
  8669		}
  8670	
  8671		/*
  8672		 * We can come here with TIF_NEED_RESCHED already set from new task
  8673		 * wake up path.
  8674		 *
  8675		 * Note: this also catches the edge-case of curr being in a throttled
  8676		 * group (e.g. via set_curr_task), since update_curr() (in the
  8677		 * enqueue of curr) will have resulted in resched being set.  This
  8678		 * prevents us from potentially nominating it as a false LAST_BUDDY
  8679		 * below.
  8680		 */
  8681		if (test_tsk_need_resched(rq->curr))
  8682			return;
  8683	
  8684		if (!sched_feat(WAKEUP_PREEMPTION))
  8685			return;
  8686	
  8687		find_matching_se(&se, &pse);
  8688		WARN_ON_ONCE(!pse);
  8689	
  8690		cse_is_idle = se_is_idle(se);
  8691		pse_is_idle = se_is_idle(pse);
  8692	
  8693		/*
  8694		 * Preempt an idle entity in favor of a non-idle entity (and don't preempt
  8695		 * in the inverse case).
  8696		 */
  8697		if (cse_is_idle && !pse_is_idle) {
  8698			/*
  8699			 * When non-idle entity preempt an idle entity,
  8700			 * don't give idle entity slice protection.
  8701			 */
  8702			do_preempt_short = true;
  8703			goto preempt;
  8704		}
  8705	
  8706		if (cse_is_idle != pse_is_idle)
  8707			return;
  8708	
  8709		/*
  8710		 * BATCH and IDLE tasks do not preempt others.
  8711		 */
  8712		if (unlikely(!normal_policy(p->policy)))
  8713			return;
  8714	
  8715		cfs_rq = cfs_rq_of(se);
  8716		update_curr(cfs_rq);
  8717		/*
  8718		 * If @p has a shorter slice than current and @p is eligible, override
  8719		 * current's slice protection in order to allow preemption.
  8720		 */
> 8721		 do_preempt_short = sched_feat(PREEMPT_SHORT) && (pse->slice < se->slice);
  8722	
  8723		/*
  8724		 * If @p has become the most eligible task, force preemption.
  8725		 */
  8726		if (__pick_eevdf(cfs_rq, !do_preempt_short) == pse)
  8727			goto preempt;
  8728	
  8729		return;
  8730	
  8731	preempt:
  8732		if (do_preempt_short)
  8733			cancel_protect_slice(se);
  8734	
  8735		resched_curr_lazy(rq);
  8736	}
  8737	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki