linux-kernel - Re: [RFC PATCH] fs: micro-optimization remove branches by adjusting flag values

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: <20200915003858.GK3421308@ZenIV.linux.org.uk>
Date:   Tue, 15 Sep 2020 01:38:58 +0100
From:   Al Viro <viro@...iv.linux.org.uk>
To:     Matthew Wilcox <willy@...radead.org>
Cc:     mateusznosek0@...il.com, linux-fsdevel@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH] fs: micro-optimization remove branches by adjusting
 flag values

On Mon, Sep 14, 2020 at 07:06:29PM +0100, Matthew Wilcox wrote:
> $ objdump -d test.o
> 0000000000000000 <a>:
>    0:   89 f8                   mov    %edi,%eax
>    2:   83 e0 03                and    $0x3,%eax
>    5:   c3                      retq   
> 
> Please stop submitting uglifying patches without checking they actually
> improve anything.  GCC is smarter than you think it is.

His main point isn't that - it's reshuffling LOOKUP_... bits to make that
kind of optimisation possible.  However, doing that sets us up for PITA
down the road (e.g. reshuffling LOOKUP_... bits becomes forbidden, etc.)
and I'd rather not go there unless we have a real-world evidence that it
does buy us anything.