Received-SPF: pass (google.com: domain of linux-kernel+bounces-60530-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Precedence: bulk
MIME-Version: 1.0
References: <20240125-mmc-no-blk-bounce-high-v1-1-d0f92a30e085@linaro.org>
 <8ca6a46e-551a-4400-965f-f4ad60bff072@app.fastmail.com> <Zcaooa0bTKo3OdvV@qmqm.qmqm.pl>
 <CACRpkdb=aDd+CuJhckE9xZjA4MXRS_BD=qd3sc3Qjnm7CNbATA@mail.gmail.com> <87e339bf-6ca9-406a-9f15-d744f90c9c40@app.fastmail.com>
In-Reply-To: <87e339bf-6ca9-406a-9f15-d744f90c9c40@app.fastmail.com>
From: Linus Walleij <linus.walleij@linaro.org>
Date: Sat, 10 Feb 2024 20:38:27 +0100
Message-ID: <CACRpkdYtUFAWweawpJquz2BUT81Ako0cZnKgZeyj8Jj93ru8fA@mail.gmail.com>
Subject: Re: [PATCH] mmc: core Drop BLK_BOUNCE_HIGH
To: Arnd Bergmann <arnd@arndb.de>
Cc: =?UTF-8?B?TWljaGHFgiBNaXJvc8WCYXc=?= <mirq-linux@rere.qmqm.pl>, 
	Christoph Hellwig <hch@lst.de>, Ulf Hansson <ulf.hansson@linaro.org>, 
	"linux-mmc @ vger . kernel . org" <linux-mmc@vger.kernel.org>, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Sat, Feb 10, 2024 at 12:58=E2=80=AFPM Arnd Bergmann <arnd@arndb.de> wrot=
e:
> On Sat, Feb 10, 2024, at 00:41, Linus Walleij wrote:

> > The only difference is where the CPU have to read/write the
> > buffers really, before the change those were all guaranteed to
> > be in lowmem (bounced there by the block core), now they can
> > also be in highmem, but sg_miter will deal with it for sure.
>
> Yes, that was my point: The sg_miter() code is meant to
> handle exactly this case with highmem data, but as far
> as I can tell, that code path has never been tested on
> 32-bit systems with highmem but without BLK_BOUNCE_HIGH.

It's actually possible to enforce testing of highmem scatterlists
to an MMC card (one need to be careful as this is destructive
testing!)
drivers/mmc/core/mmc_test.c

..but the one relevant target I have is a Kirkwood and it only
has 128 MB of memory so highmem won't be exercised.

I'll put this into the cover letter on the other series (fixing a bunch
of drivers to use sg_miter) though.

Yours,
Linus Walleij