linux-kernel - Re: [PATCH] mm/slub: fix comments about fastpath limitation on PREEMPT

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <eff58088-e291-6960-cbfd-b6cc428f5f06@suse.cz>
Date:   Thu, 25 Aug 2022 15:11:32 +0200
From:   Vlastimil Babka <vbabka@...e.cz>
To:     Hyeonggon Yoo <42.hyeyoo@...il.com>,
        Christoph Lameter <cl@...ux.com>,
        Pekka Enberg <penberg@...nel.org>,
        David Rientjes <rientjes@...gle.com>,
        Joonsoo Kim <iamjoonsoo.kim@....com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Roman Gushchin <roman.gushchin@...ux.dev>
Cc:     Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
        Thomas Gleixner <tglx@...utronix.de>,
        Mike Galbraith <efault@....de>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mm/slub: fix comments about fastpath limitation on
 PREEMPT_RT

On 8/25/22 03:57, Hyeonggon Yoo wrote:
> With PREEMPT_RT disabling interrupt is unnecessary as there is
> no user of slab in hardirq context on PREEMPT_RT.
> 
> The limitation of lockless fastpath on PREEMPT_RT comes from the fact
> that local_lock does not disable preemption on PREEMPT_RT.
> 
> Fix comments accordingly.
> 
> Signed-off-by: Hyeonggon Yoo <42.hyeyoo@...il.com>

Just FTR, as "slub: Make PREEMPT_RT support less convoluted" patch dealt
with these comments already, there's now nothing left to apply from below.

> ---
>  mm/slub.c | 23 ++++++++++++++---------
>  1 file changed, 14 insertions(+), 9 deletions(-)
> 
> diff --git a/mm/slub.c b/mm/slub.c
> index 30c2ee9e8a29..aa42ac6013b8 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -100,7 +100,7 @@
>   *   except the stat counters. This is a percpu structure manipulated only by
>   *   the local cpu, so the lock protects against being preempted or interrupted
>   *   by an irq. Fast path operations rely on lockless operations instead.
> - *   On PREEMPT_RT, the local lock does not actually disable irqs (and thus
> + *   On PREEMPT_RT, the local lock does not actually disable preemption (and thus
>   *   prevent the lockless operations), so fastpath operations also need to take
>   *   the lock and are no longer lockless.
>   *
> @@ -3185,10 +3185,12 @@ static __always_inline void *slab_alloc_node(struct kmem_cache *s, struct list_l
>  	slab = c->slab;
>  	/*
>  	 * We cannot use the lockless fastpath on PREEMPT_RT because if a
> -	 * slowpath has taken the local_lock_irqsave(), it is not protected
> -	 * against a fast path operation in an irq handler. So we need to take
> -	 * the slow path which uses local_lock. It is still relatively fast if
> -	 * there is a suitable cpu freelist.
> +	 * slowpath has taken the local_lock which does not disable preemption
> +	 * on PREEMPT_RT, it is not protected against a fast path operation in
> +	 * another thread that does not take the local_lock.
> +	 *
> +	 * So we need to take the slow path which uses local_lock. It is still
> +	 * relatively fast if there is a suitable cpu freelist.
>  	 */
>  	if (IS_ENABLED(CONFIG_PREEMPT_RT) ||
>  	    unlikely(!object || !slab || !node_match(slab, node))) {
> @@ -3457,10 +3459,13 @@ static __always_inline void do_slab_free(struct kmem_cache *s,
>  #else /* CONFIG_PREEMPT_RT */
>  		/*
>  		 * We cannot use the lockless fastpath on PREEMPT_RT because if
> -		 * a slowpath has taken the local_lock_irqsave(), it is not
> -		 * protected against a fast path operation in an irq handler. So
> -		 * we need to take the local_lock. We shouldn't simply defer to
> -		 * __slab_free() as that wouldn't use the cpu freelist at all.
> +		 * a slowpath has taken the local_lock which does not disable
> +		 * preemption on PREEMPT_RT, it is not protected against a
> +		 * fast path operation in another thread that does not take
> +		 * the local_lock.
> +		 *
> +		 * So we need to take the local_lock. We shouldn't simply defer
> +		 * to __slab_free() as that wouldn't use the cpu freelist at all.
>  		 */
>  		void **freelist;
>