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;
MIME-Version: 1.0
References: <20210617152754.17960-1-mcroce@linux.microsoft.com>
 <20210617152754.17960-2-mcroce@linux.microsoft.com> <YNChl0tkofSGzvIX@infradead.org>
 <17bb90eef20145cd9cca1b8e72a514ad@AcuMS.aculab.com>
In-Reply-To: <17bb90eef20145cd9cca1b8e72a514ad@AcuMS.aculab.com>
From:   Matteo Croce <mcroce@linux.microsoft.com>
Date:   Wed, 23 Jun 2021 00:53:35 +0200
Message-ID: <CAFnufp06+s43z5SxL48LB73=GdzYFiq578MCHZ_NMUNZkwdo9g@mail.gmail.com>
Subject: Re: [PATCH v3 1/3] riscv: optimized memcpy
To:     David Laight <David.Laight@aculab.com>
Cc:     Christoph Hellwig <hch@infradead.org>,
        "linux-riscv@lists.infradead.org" <linux-riscv@lists.infradead.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "linux-arch@vger.kernel.org" <linux-arch@vger.kernel.org>,
        Paul Walmsley <paul.walmsley@sifive.com>,
        Palmer Dabbelt <palmer@dabbelt.com>,
        Albert Ou <aou@eecs.berkeley.edu>,
        Atish Patra <atish.patra@wdc.com>,
        Emil Renner Berthing <kernel@esmil.dk>,
        Akira Tsukamoto <akira.tsukamoto@gmail.com>,
        Drew Fustini <drew@beagleboard.org>,
        Bin Meng <bmeng.cn@gmail.com>, Guo Ren <guoren@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Jun 22, 2021 at 10:19 AM David Laight <David.Laight@aculab.com> wrote:
>
> From: Christoph Hellwig
> > Sent: 21 June 2021 15:27
> ...
> > > +           for (next = s.ulong[0]; count >= bytes_long + mask; count -= bytes_long) {
> >
> > Please avoid the pointlessly overlong line.  And (just as a matter of
> > personal preference) I find for loop that don't actually use a single
> > iterator rather confusing.  Wjy not simply:
> >
> >               next = s.ulong[0];
> >               while (count >= bytes_long + mask) {
> >                       ...
> >                       count -= bytes_long;
> >               }
>
> My fist attack on long 'for' statements is just to move the
> initialisation to the previous line.
> Then make sure there is nothing in the comparison that needs
> to be calculated every iteration.
> I suspect you can subtract 'mask' from 'count'.
> Giving:
>                 count -= mask;
>                 next = s.ulong[0];
>                 for (;; count > bytes_long; count -= bytes_long) {
>

This way we'll lose the remainder, as count is used at the end to copy
the leftover.
Anyway, both bytes_long and mask are constant, I doubt they get summed
at every cycle.

> Next is to shorten the variable names!
>
>         David
>

-- 
per aspera ad upstream