Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190614063341.1672-1-fancer.lancer@gmail.com>
 <20190620174002.tgayzon7dc5d57fh@pburton-laptop> <alpine.LFD.2.21.1906201851580.21654@eddie.linux-mips.org>
 <CAK8P3a28Dp3UygNyomDPDxDmCmey37VS7TJkmDogaKUGZMF2mw@mail.gmail.com> <alpine.LFD.2.21.1906211048360.21654@eddie.linux-mips.org>
In-Reply-To: <alpine.LFD.2.21.1906211048360.21654@eddie.linux-mips.org>
From:   Arnd Bergmann <arnd@arndb.de>
Date:   Fri, 21 Jun 2019 13:09:24 +0200
Message-ID: <CAK8P3a3HWn7RXjcT0KA_qOc+C1SgWd2qXSdCTTAmRKHdc4qNbQ@mail.gmail.com>
Subject: Re: [PATCH] mips: Remove q-accessors from non-64bit platforms
To:     "Maciej W. Rozycki" <macro@linux-mips.org>
Cc:     Paul Burton <paul.burton@mips.com>,
        Serge Semin <fancer.lancer@gmail.com>,
        Ralf Baechle <ralf@linux-mips.org>,
        James Hogan <jhogan@kernel.org>,
        Serge Semin <Sergey.Semin@t-platforms.ru>,
        "Vadim V . Vlasov" <vadim.vlasov@t-platforms.ru>,
        "linux-mips@vger.kernel.org" <linux-mips@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Jun 21, 2019 at 12:09 PM Maciej W. Rozycki <macro@linux-mips.org> wrote:
>
> On Fri, 21 Jun 2019, Arnd Bergmann wrote:
>
> > >  The use of 64-bit operations to access option's packet memory, which is
> > > true SRAM, i.e. no side effects, is to improve throughput only and there's
> > > no need for atomicity here nor also any kind of barriers, except at the
> > > conclusion.  Splitting 64-bit accesses into 32-bit halves in software
> > > would not be a functional error here.
> >
> > The other property of packet memory and similar things is that you
> > basically want memcpy()-behavior with no byteswaps. This is one
> > of the few cases in which __raw_readq() is actually the right accessor
> > in (mostly) portable code.
>
>  Correct, but we're missing an `__raw_readq_relaxed', etc. interface and
> having additional barriers applied on every access would hit performance
> very badly;

How so? __raw_readq() by definition has the least barriers of
all, you can't make it more relaxed than it already is.

> in fact even the barriers `*_relaxed' accessors imply would
> best be removed in this use (which is why defza.c uses `readw_o' vs
> `readw_u', etc. internally), but after all the struggles over the years
> for weakly ordered internal APIs x86 people are so averse to I'm not sure
> if I want to start another one.  We can get away with `readq_relaxed' in
> this use though as all the systems this device can be used with are
> little-endian as is TURBOchannel, so no byte-swapping will ever actually
> occur.

I still don't see any downside of using __raw_readq() here, while the
upsides are:

- makes the driver portable to big-endian kernels (even though we don't
  care)
- avoids all barriers
- fixes the build regression.

      Arnd