Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 19 Apr 2023 08:59:03 -0700 (PDT)
Subject:     Re: [PATCH v8 0/7] Add non-coherent DMA support for AX45MP
In-Reply-To: <ZDzs3eYIKPFcv0HQ@infradead.org>
CC:     Conor Dooley <conor@kernel.org>, prabhakar.csengg@gmail.com,
        Arnd Bergmann <arnd@arndb.de>,
        Conor Dooley <conor.dooley@microchip.com>,
        geert+renesas@glider.be, heiko@sntech.de, guoren@kernel.org,
        ajones@ventanamicro.com, Paul Walmsley <paul.walmsley@sifive.com>,
        aou@eecs.berkeley.edu, samuel@sholland.org,
        linux-riscv@lists.infradead.org, robh+dt@kernel.org,
        krzysztof.kozlowski+dt@linaro.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-renesas-soc@vger.kernel.org,
        biju.das.jz@bp.renesas.com, prabhakar.mahadev-lad.rj@bp.renesas.com
From:   Palmer Dabbelt <palmer@dabbelt.com>
To:     Christoph Hellwig <hch@infradead.org>
Message-ID: <mhng-e296c86c-71b1-46f8-88c6-16eda3590a3d@palmer-ri-x1c9>
Mime-Version: 1.0 (MHng)
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Sun, 16 Apr 2023 23:53:17 PDT (-0700), Christoph Hellwig wrote:
> On Wed, Apr 12, 2023 at 09:32:30PM +0100, Conor Dooley wrote:
>> On Wed, Apr 12, 2023 at 12:08:53PM +0100, Prabhakar wrote:
>>
>> > Note,
>> > - This series requires testing on Cores with zicbom and T-Head SoCs
>>
>> As I said last time, I dunno what actual Zicbom stuff exists, other than
>> perhaps the Ventana lads having something. I did some tyre kicking on my
>> D1 and it was fine, although nothing has actually changed materially for
>> either of them with this series in v8..
>
> And as saying before, there is absolutely no reason to add non-standard
> non-coherent DMA support and let this cancer creep.  If you want Linux
> support implement Zicbom, be that in hardware or the SBI.

IMO we should just take the support in Linux: trying to hide stuff behind the
SBI leads to more more headaches than it's worth, we end up with a bunch of
broken firmware to try and work around.  We've already got a mess here because
of the D1 support, we might as well just live with it.

In practice there's just going to be a ton of mess in arch/riscv, as the ISA
has been missing many core features for many years and hardware vendors are
allowed to do whatever they want.  That's obviously a huge headache, but I
think there's nothing we can really do about it.