linux-kernel - Re: [RFC PATCH] sched: avoid unnecessary atomic_read when sync_core_before

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: <802b944d-d168-d9c8-add3-1fe17f3985f5@linux.alibaba.com>
Date:   Wed, 6 Apr 2022 10:47:22 +0800
From:   Tianchen Ding <dtcccc@...ux.alibaba.com>
To:     Andrew Morton <akpm@...ux-foundation.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Fenghua Yu <fenghua.yu@...el.com>,
        Borislav Petkov <bp@...e.de>,
        Pavel Tatashin <pasha.tatashin@...een.com>,
        NeilBrown <neilb@...e.de>,
        Vasily Averin <vasily.averin@...ux.dev>,
        "Matthew Wilcox (Oracle)" <willy@...radead.org>
Cc:     linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH] sched: avoid unnecessary atomic_read when
 sync_core_before_usermode() is empty

We've run schbench and found wakeup latency on some arm64 machines worse 
than others. Perf shows there's a hotspot on 
atomic_read(&mm->membarrier_state);

We're still working for the real reason behind it (maybe cache or sth 
hardware related), and we do see remove this function can help improve 
performance.

Thanks.

On 2022/4/2 11:08, Tianchen Ding wrote:
> On archs except x86, CONFIG_ARCH_HAS_SYNC_CORE_BEFORE_USERMODE is not
> defined. We found membarrier_mm_sync_core_before_usermode() looks like
> this when compiled by gcc10:
> 
> 	if (current->mm != mm)
> 		return;
> 	atomic_read(&mm->membarrier_state);
> 
> This memory access is unnecessary. Remove it to improve performance.
> 
> Signed-off-by: Tianchen Ding <dtcccc@...ux.alibaba.com>
> ---
>   include/linux/sched/mm.h | 6 ++++++
>   1 file changed, 6 insertions(+)
> 
> diff --git a/include/linux/sched/mm.h b/include/linux/sched/mm.h
> index a80356e9dc69..3ded68d9f913 100644
> --- a/include/linux/sched/mm.h
> +++ b/include/linux/sched/mm.h
> @@ -401,6 +401,7 @@ enum {
>   #include <asm/membarrier.h>
>   #endif
>   
> +#ifdef CONFIG_ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
>   static inline void membarrier_mm_sync_core_before_usermode(struct mm_struct *mm)
>   {
>   	if (current->mm != mm)
> @@ -410,6 +411,11 @@ static inline void membarrier_mm_sync_core_before_usermode(struct mm_struct *mm)
>   		return;
>   	sync_core_before_usermode();
>   }
> +#else
> +static inline void membarrier_mm_sync_core_before_usermode(struct mm_struct *mm)
> +{
> +}
> +#endif
>   
>   extern void membarrier_exec_mmap(struct mm_struct *mm);
>