Re: [PATCH v2 12/12] sched: Remove sched_class::balance()

[Kayra Cizmeci <kayracizmeci@gmail.com>]

Hi!

On Wed, Nov 11, 2026 at 05:15:22PM +0100, Peter Zijlstra wrote:
> On Wed, Nov 11, 2026 at 04:47:24PM +0100, Peter Zijlstra wrote:

> > 
> > Gah, you're right. This don't work right. Let me try again.
> 
> Here, something like so then?
> 
> ---
> Subject: sched/deadline: Fix wakeup_preempt() vs dl_server
> From: Peter Zijlstra <peterz@infradead.org>
> Date: Wed Nov 11 14:12:25 CET 2026
> ---
>  kernel/sched/deadline.c |   40 +++++++++++++++++++++++++---------------
>  1 file changed, 25 insertions(+), 15 deletions(-)
> 
> --- a/kernel/sched/deadline.c
> +++ b/kernel/sched/deadline.c
> @@ -715,8 +715,7 @@ static struct rq *dl_task_offline_migrat
>  	return later_rq;
>  }
>  
> -static void
> -enqueue_dl_entity(struct sched_dl_entity *dl_se, int flags);
> +static void enqueue_dl_entity(struct rq *rq, struct sched_dl_entity *dl_se, int flags);
>  static void enqueue_task_dl(struct rq *rq, struct task_struct *p, int flags);
>  static void dequeue_dl_entity(struct sched_dl_entity *dl_se, int flags);
>  static void wakeup_preempt_dl(struct rq *rq, struct task_struct *p, int flags);
> @@ -1184,10 +1183,7 @@ static enum hrtimer_restart dl_server_ti
>  			dl_se->dl_defer_running = 1;
>  		}
>  
> -		enqueue_dl_entity(dl_se, ENQUEUE_REPLENISH);
> -
> -		if (!dl_task(dl_se->rq->curr) || dl_entity_preempt(dl_se, &dl_se->rq->curr->dl))
> -			resched_curr(rq);
> +		enqueue_dl_entity(rq, dl_se, ENQUEUE_REPLENISH);
>  
>  		__push_dl_task(rq, rf);
>  	}
> @@ -1520,7 +1516,7 @@ static void update_curr_dl_se(struct rq
>  					replenish_dl_new_period(dl_se, rq);
>  					start_dl_timer(dl_se);
>  				} else {
> -					enqueue_dl_entity(dl_se, ENQUEUE_REPLENISH);
> +					enqueue_dl_entity(rq, dl_se, ENQUEUE_REPLENISH);
>  				}
>  			} else {
>  				enqueue_task_dl(rq, dl_task_of(dl_se), ENQUEUE_REPLENISH);
> @@ -1812,9 +1808,7 @@ void dl_server_start(struct sched_dl_ent
>  
>  	trace_sched_dl_server_start_tp(dl_se, cpu_of(rq), dl_get_type(dl_se, rq));
>  	dl_se->dl_server_active = 1;
> -	enqueue_dl_entity(dl_se, ENQUEUE_WAKEUP);
> -	if (!dl_task(dl_se->rq->curr) || dl_entity_preempt(dl_se, &rq->curr->dl))
> -		resched_curr(dl_se->rq);
> +	enqueue_dl_entity(rq, dl_se, ENQUEUE_WAKEUP);
>  }
>  
>  void dl_server_stop(struct sched_dl_entity *dl_se)
> @@ -2379,8 +2373,8 @@ static void __dequeue_dl_entity(struct s
>  	dec_dl_tasks(dl_se, dl_rq);
>  }
>  
> -static void
> -enqueue_dl_entity(struct sched_dl_entity *dl_se, int flags)
> +static bool
> +_enqueue_dl_entity(struct sched_dl_entity *dl_se, int flags)
>  {
>  	WARN_ON_ONCE(on_dl_rq(dl_se));
>  
> @@ -2418,7 +2412,7 @@ enqueue_dl_entity(struct sched_dl_entity
>  		if (flags & ENQUEUE_WAKEUP)
>  			task_contending(dl_se, flags);
>  
> -		return;
> +		return false;
>  	}
>  
>  	/*
> @@ -2442,7 +2436,7 @@ enqueue_dl_entity(struct sched_dl_entity
>  	 * deferred task and still got to wait, don't enqueue.
>  	 */
>  	if (dl_se->dl_throttled && start_dl_timer(dl_se))
> -		return;
> +		return false;
>  
>  	/*
>  	 * We're about to enqueue, make sure we're not ->dl_throttled!
> @@ -2457,6 +2451,22 @@ enqueue_dl_entity(struct sched_dl_entity
>  	}
>  
>  	__enqueue_dl_entity(dl_se);
> +	return true;
> +}
> +
> +static void enqueue_dl_entity(struct rq *rq, struct sched_dl_entity *dl_se, int flags)
> +{
> +	/*
> +	 * Deadline can enqueue entities from various timers; and esp.
> +	 * dl_server() entities will not pass through wakeup_preempt().
> +	 */
> +	if (!_enqueue_dl_entity(dl_se, flags))
> +		return;
> +
> +	if (!dl_task(rq->curr) || dl_entity_preempt(dl_se, &rq->curr->dl))
> +		resched_curr(dl_se->rq);
> +
> +	rq->next_class = &dl_sched_class;
>  }
>  
>  static void dequeue_dl_entity(struct sched_dl_entity *dl_se, int flags)
> @@ -2534,7 +2544,7 @@ static void enqueue_task_dl(struct rq *r
>  	if (task_on_rq_migrating(p))
>  		flags |= ENQUEUE_MIGRATING;
>  
> -	enqueue_dl_entity(dl_se, flags);
> +	enqueue_dl_entity(rq, dl_se, flags);
>  
>  	if (dl_server(dl_se))
>  		return;

I tried running the tests, and the second I tried it crashed :/.

Then I added WARN_ON_ONCE(!dl_se->rq); to catch if dl_se->rq were
causing all this. And changed resched_curr(dl_se->rq)
to resched_curr(rq).

Warning triggered, and because I changed it I was able to test it,
it was around 13 us. (Base was around 12 us so I would count it the same)

Changing it to rq seem to work correct. If I'm not missing any gears in my
brain...


Here's the warning that I added since I wasn't able to get the real
crash logs:

[   45.193366] ------------[ cut here ]------------
[   45.193472] WARNING: kernel/sched/deadline.c:2466 at enqueue_dl_entity+0x4a/0x70, CPU#1: stress-ng-cycli/1918
[   45.193480] Modules linked in:
[   45.193573] CPU: 1 UID: 0 PID: 1918 Comm: stress-ng-cycli Not tainted 7.4.0-rc1+ #36 PREEMPT(lazy)
[   45.193576] Hardware name: System manufacturer System Product Name/TUF B450M-PRO GAMING, BIOS 4604 03/22/2024
[   45.193579] RIP: 0010:enqueue_dl_entity+0x4a/0x70
[   45.193583] Code: 00 00 00 74 25 48 8b 43 20 8b 50 6c 85 d2 78 1e 48 89 df e8 48 36 fc ff 48 c7 83 a8 0c 00 00 30 e7 d1 96 5b 5d e9 c1 a4 c6 ff <0f> 0b eb d7 f6 45 53 10 75 dc 48 8b 80 50 02 00 00 48 39 45 48 79
[   45.193586] RSP: 0018:ffffd38d4b597c70 EFLAGS: 00010046
[   45.193590] RAX: 0000000000000001 RBX: ffff8b5a162b6b00 RCX: ffffd38d4b597bd4
[   45.193592] RDX: 0000000000000003 RSI: ffff8b5713e1b0a8 RDI: ffff8b5713e1b090
[   45.193594] RBP: ffff8b57059d0208 R08: ffff8b5713e1b090 R09: ffff8b57059d1048
[   45.193596] R10: ffff8b57059d0000 R11: 0000000000000000 R12: ffff8b57059d0208
[   45.193598] R13: ffff8b5a162b6b00 R14: ffff8b57059d0000 R15: ffffd38d4b597d78
[   45.193600] FS:  00007f3fa07444c0(0000) GS:ffff8b5a7d950000(0000) knlGS:0000000000000000
[   45.193603] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   45.193605] CR2: 00007ffd6f7c1688 CR3: 000000012dccc000 CR4: 0000000000f50ef0
[   45.193608] PKRU: 55555554
[   45.193610] Call Trace:
[   45.193612]  <TASK>
[   45.193615]  enqueue_task_dl+0x5b/0x170
[   45.193622]  enqueue_task+0x88/0x210
[   45.193629]  sched_change_end+0x147/0x1d0
[   45.193636]  __sched_setscheduler+0x23f/0xd10
[   45.193648]  __x64_sys_sched_setattr+0xb4/0x1f0
[   45.193659]  do_syscall_64+0xed/0x1560
[   45.193664]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193668]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193671]  ? find_held_lock+0x2b/0x80
[   45.193677]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193680]  ? __lock_release.isra.0+0x57/0x170
[   45.193684]  ? count_memcg_events+0x1c9/0x290
[   45.193689]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193692]  ? find_held_lock+0x2b/0x80
[   45.193698]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193701]  ? __lock_release.isra.0+0x57/0x170
[   45.193704]  ? handle_mm_fault+0x218/0x380
[   45.193712]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193715]  ? lock_release.part.0+0x1c/0x50
[   45.193718]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193721]  ? do_user_addr_fault+0x4ba/0x7e0
[   45.193728]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193731]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193734]  ? lockdep_hardirqs_on_prepare.part.0+0x9b/0x140
[   45.193738]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193741]  ? trace_hardirqs_off_finish+0xc6/0xe0
[   45.193745]  ? srso_alias_return_thunk+0x5/0xfbef5
[   45.193749]  ? do_syscall_64+0x5e/0x1560
--
[   45.193815] ---[ end trace 0000000000000000 ]---
[   57.602578] ------------[ cut here ]------------
[   57.602584] WARNING: kernel/sched/deadline.c:2466 at enqueue_dl_entity+0x4a/0x70, CPU#0: stress-ng-cycli/1956
[   57.602593] Modules linked in:
[   57.602601] CPU: 0 UID: 0 PID: 1956 Comm: stress-ng-cycli Tainted: G        W           7.4.0-rc1+ #36 PREEMPT(lazy)
[   57.602606] Tainted: [W]=WARN
[   57.602609] Hardware name: System manufacturer System Product Name/TUF B450M-PRO GAMING, BIOS 4604 03/22/2024
[   57.602612] RIP: 0010:enqueue_dl_entity+0x4a/0x70
[   57.602617] Code: 00 00 00 74 25 48 8b 43 20 8b 50 6c 85 d2 78 1e 48 89 df e8 48 36 fc ff 48 c7 83 a8 0c 00 00 30 e7 d1 96 5b 5d e9 c1 a4 c6 ff <0f> 0b eb d7 f6 45 53 10 75 dc 48 8b 80 50 02 00 00 48 39 45 48 79
[   57.602621] RSP: 0018:ffffd38d4b6ffa70 EFLAGS: 00010046
[   57.602626] RAX: 0000000000000001 RBX: ffff8b5a16236b00 RCX: ffffd38d4b6ff9d4
[   57.602629] RDX: 0000000000000003 RSI: ffff8b5713e1b0a8 RDI: ffff8b5713e1b090
[   57.602632] RBP: ffff8b57059c0208 R08: ffff8b5713e1b090 R09: ffff8b57059c1048
[   57.602634] R10: ffff8b57059c0000 R11: 0000000000000000 R12: ffff8b57059c0208
[   57.602637] R13: ffff8b5a16236b00 R14: ffff8b57059c0000 R15: ffffd38d4b6ffb78
[   57.602640] FS:  00007f56864024c0(0000) GS:ffff8b5a7d8d0000(0000) knlGS:0000000000000000
[   57.602644] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   57.602647] CR2: 00007fff971de0e8 CR3: 0000000127ad3000 CR4: 0000000000f50ef0
[   57.602650] PKRU: 55555554
[   57.602653] Call Trace:
[   57.602656]  <TASK>
[   57.602661]  enqueue_task_dl+0x5b/0x170
[   57.602670]  enqueue_task+0x88/0x210
[   57.602680]  sched_change_end+0x147/0x1d0
[   57.602689]  __sched_setscheduler+0x23f/0xd10
[   57.602705]  __x64_sys_sched_setattr+0xb4/0x1f0
[   57.602720]  do_syscall_64+0xed/0x1560


Cheers,

Kayra