Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH] lib/string: Bring optimized memcmp from glibc
To:     Linus Torvalds <torvalds@linux-foundation.org>,
        David Sterba <dsterba@suse.cz>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Nick Desaulniers <ndesaulniers@google.com>,
        linux-fsdevel <linux-fsdevel@vger.kernel.org>,
        Dave Chinner <david@fromorbit.com>
References: <20210721135926.602840-1-nborisov@suse.com>
 <CAHk-=whqJKKc9wUacLEkvTzXYfYOUDt=kHKX6Fa8Kb4kQftbbQ@mail.gmail.com>
 <20210721201029.GQ19710@twin.jikos.cz>
 <CAHk-=whCygw44p30Pmf+Bt8=LVtmij3_XOxweEA3OQNruhMg+A@mail.gmail.com>
From:   Nikolay Borisov <nborisov@suse.com>
Message-ID: <792949a2-d987-f6a0-a153-8c5fe1e3a073@suse.com>
Date:   Thu, 22 Jul 2021 08:54:14 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.11.0
MIME-Version: 1.0
In-Reply-To: <CAHk-=whCygw44p30Pmf+Bt8=LVtmij3_XOxweEA3OQNruhMg+A@mail.gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 21.07.21 г. 23:27, Linus Torvalds wrote:
> On Wed, Jul 21, 2021 at 1:13 PM David Sterba <dsterba@suse.cz> wrote:
>>
>> adding a memcmp_large that compares by native words or u64 could be
>> the best option.
> 
> Yeah, we could just special-case that one place.

This who thread started because I first implemented a special case just
for dedupe and Dave Chinner suggested instead of playing whack-a-mole to
get something decent for the generic memcmp so that we get an
improvement across the whole of the kernel.

> 
> But see the patches I sent out - I think we can get the best of both worlds.
> 
> A small and simple memcmp() that is good enough and not the
> _completely_ stupid thing we have now.
> 
> The second patch I sent out even gets the mutually aligned case right.
> 
> Of course, the glibc code also ended up unrolling things a bit, but
> honestly, the way it did it was too disgusting for words.
> 
> And if it really turns out that the unrolling makes a big difference -
> although I doubt it's meaningful with any modern core - I can add a
> couple of lines to that simple patch I sent out to do that too.
> Without getting the monster that is that glibc code.
> 
> Of course, my patch depends on the fact that "get_unaligned()" is
> cheap on all CPU's that really matter, and that caches aren't
> direct-mapped any more. The glibc code seems to be written for a world
> where registers are cheap, unaligned accesses are prohibitively
> expensive, and unrolling helps because L1 caches are direct-mapped and
> you really want to do chunking to not get silly way conflicts.
> 
> If old-style Sparc or MIPS was our primary target, that would be one
> thing. But it really isn't.
> 
>               Linus
>