Re: [PATCH] sched: Allow sleeping spinlocks on PREEMPT_RT within non_block_start() section

[David Woodhouse]

On Thu, 2026-08-21 at 11:57 +0200, Sebastian Andrzej Siewior wrote:
> Since spinlock_t (and rwlock_t) are not problem, add a sleeping_lock
> argument to __might_resched() signaling if this is scheduling request
> is from a sleeping lock (true) or regular scheduling request (false).


Acked-by: David Woodhouse <dwmw@xxxxxxxxxxxx>
Fixes: 312364f3534c ("kernel.h: Add non_block_start/end()")
Reported-by: syzbot+c3178b6b512446632bac@xxxxxxxxxxxxxxxxxxxxxxxxx
Closes: https://syzkaller.appspot.com/bug?extid=c3178b6b512446632bac

I've prodded syzbot; let's see if we can have a Tested-by: from it too.

I kind of also want synchronize_srcu() to be permitted too, but I guess 
https://lore.kernel.org/all/6eed3fe3461e9690b486ca98fa7563f60d3940ff.camel@xxxxxxxxxxxxx/
would need to be resolved first.

Thanks.

Attachment: smime.p7s
Description: S/MIME cryptographic signature