Sched_ext development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Tejun Heo" <tj@kernel.org>
Cc: sched-ext@lists.linux.dev
Subject: Re: [PATCH 2/2] sched_ext: Maintain an online cid mask in the scheduler arena
Date: Tue, 15 Sep 2026 08:58:18 +0000	[thread overview]
Message-ID: <20260915085819.5DB991F000FF@smtp.kernel.org> (raw)
In-Reply-To: <20260915082725.3881071-3-tj@kernel.org>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] sched_ext: bit corruption in online_cmask on 32-bit big-endian architectures
--

commit caa3d27a5f1591e49ffdf405f84e61f1f37a10f5
Author: Tejun Heo <tj@kernel.org>

sched_ext: Maintain an online cid mask in the scheduler arena

This commit adds a kernel-maintained online cid mask in the scheduler's
arena for child schedulers. It populates this mask after the cid mapping
is finalized and tracks online/offline states based on hotplug notifications.

> diff --git a/kernel/sched/ext/ext.c b/kernel/sched/ext/ext.c
> index 10af28a9f2c0c..7fbef7f11d3ae 100644
> --- a/kernel/sched/ext/ext.c
> +++ b/kernel/sched/ext/ext.c
[ ... ]
> @@ -3663,8 +3663,20 @@ static void handle_hotplug(struct rq *rq, bool online)
>  		s16 *tbl = rcu_dereference_check(scx_cpu_to_cid_tbl,
>  						 lockdep_is_cpus_held());
>  
> -		if (tbl)
> +		if (tbl) {
> +			struct scx_sched *pos;
> +
>  			cpu_or_cid = tbl[cpu];
> +
> +			guard(raw_spinlock_irqsave)(&scx_sched_lock);
> +			list_for_each_entry(pos, &scx_sched_all, all) {
> +				struct scx_cmask *mask = pos->online_cmask;
> +
> +				if (mask)
> +					__assign_bit(cpu_or_cid, (unsigned long *)mask->bits,
> +						     online);

[Severity: Medium]
Does casting mask->bits to an unsigned long pointer corrupt the bitmask
layout on 32-bit big-endian architectures?

Since mask->bits is declared as a u64 array in struct scx_cmask, BPF
programs expect a standard 64-bit integer bit layout. However,
__assign_bit() operates on native words. On 32-bit big-endian systems,
the most significant 32 bits are at offset 0, meaning __assign_bit(0)
modifies bit 32 of the underlying u64 instead of bit 0.

This would cause BPF programs to observe bit cpu_or_cid + 32 toggled
instead of cpu_or_cid, which could lead to dispatching tasks to offline
CPUs or starving tasks on online CPUs.

> +			}
> +		}
>  	}
>  
>  	if (online && SCX_HAS_OP(sch, cpu_online))

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260915082725.3881071-1-tj@kernel.org?part=2

  reply	other threads:[~2026-09-15  8:58 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-15  8:27 [PATCHSET v2 sched_ext/for-7.3-fixes] sched_ext: Idle claim recovery and online cid mask Tejun Heo
2026-09-15  8:27 ` [PATCH 1/2] sched_ext: scx_qmap: Restore unused idle claims from ops.dispatch() Tejun Heo
2026-09-15  8:27 ` [PATCH 2/2] sched_ext: Maintain an online cid mask in the scheduler arena Tejun Heo
2026-09-15  8:58   ` sashiko-bot [this message]
2026-09-15 16:54     ` Tejun Heo
2026-09-15 12:27 ` [PATCHSET v2 sched_ext/for-7.3-fixes] sched_ext: Idle claim recovery and online cid mask Andrea Righi
2026-09-15 17:01 ` Tejun Heo
  -- strict thread matches above, loose matches on Subject: below --
2026-09-14 23:42 [PATCHSET sched_ext/for-7.3-fixes] sched_ext: Idle repick notifications " Tejun Heo
2026-09-14 23:42 ` [PATCH 2/2] sched_ext: Maintain an online cid mask in the scheduler arena Tejun Heo
2026-09-15  0:03   ` sashiko-bot
2026-09-15  5:44   ` Andrea Righi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260915085819.5DB991F000FF@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=sashiko-reviews@lists.linux.dev \
    --cc=sched-ext@lists.linux.dev \
    --cc=tj@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox