linux-kernel - Re: [PATCH 1/1] mm/percpu.c: simplify grouping cpu logic in pcpu_build_alloc

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 for Android: free password hash cracker in your pocket

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

Date:   Fri, 23 Sep 2016 15:23:34 -0400
From:   Tejun Heo <tj@...nel.org>
To:     zijun_hu <zijun_hu@...o.com>
Cc:     Andrew Morton <akpm@...ux-foundation.org>, zijun_hu@....com,
        linux-mm@...ck.org,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        cl@...ux.com
Subject: Re: [PATCH 1/1] mm/percpu.c: simplify grouping cpu logic in
 pcpu_build_alloc_info()

On Sat, Sep 24, 2016 at 02:15:09AM +0800, zijun_hu wrote:
> From: zijun_hu <zijun_hu@....com>
> 
> simplify grouping cpu logic in pcpu_build_alloc_info() to improve
> readability and performance, it discards the goto statement too
> 
> for every possible cpu, decide whether it can share group id of any
> lower index CPU, use the group id if so, otherwise a new group id
> is allocated to it
> 
> Signed-off-by: zijun_hu <zijun_hu@....com>

I'm not gonna change that code unless there are clear upsides.  It's a
complicated code path which is run once during boot.  It's not worth
optimizing, the author doesn't explain how the change has been tested
or verified and doesn't respond to people pointing out that these
drive-by patches aren't helpful.

I won't engage with his patches until he changes his approach and
think that it's advisable for others to do so too.

Nacked-by: Tejun Heo <tj@...nel.org>

Thanks.

-- 
tejun