Re: [PATCH 0/3] dmaengine: Stear users towards dma_request_slave_chan()

From: Geert Uytterhoeven <geert@linux-m68k.org>
To: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Cc: Peter Ujfalusi <peter.ujfalusi@ti.com>,
	Andy Shevchenko <andy.shevchenko@gmail.com>,
	Vinod Koul <vkoul@kernel.org>,
	dmaengine <dmaengine@vger.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Dan Williams <dan.j.williams@intel.com>,
	Linux-sh list <linux-sh@vger.kernel.org>,
	Linux-Renesas <linux-renesas-soc@vger.kernel.org>
Subject: Re: [PATCH 0/3] dmaengine: Stear users towards dma_request_slave_chan()
Date: Tue, 4 Feb 2020 09:13:27 +0100	[thread overview]
Message-ID: <CAMuHMdWKHOm8WjMx3Lm-MwZ_VZVaFz_otGe0V3pKp01v81mqZA@mail.gmail.com> (raw)
In-Reply-To: <00734e40-da0b-9d7f-20bc-ce1f9658dcd3@physik.fu-berlin.de>

Hi Adrian,

On Mon, Feb 3, 2020 at 10:26 PM John Paul Adrian Glaubitz
<glaubitz@physik.fu-berlin.de> wrote:
> On 2/3/20 9:34 PM, Geert Uytterhoeven wrote:
> > On Mon, Feb 3, 2020 at 9:21 PM John Paul Adrian Glaubitz
> > <glaubitz@physik.fu-berlin.de> wrote:
> >> On 2/3/20 2:32 PM, Geert Uytterhoeven wrote:
> >>> Both rspi and sh-msiof have users on legacy SH (i.e. without DT):
> >>
> >> FWIW, there is a patch set by Yoshinori Sato to add device tree support
> >> for classical SuperH hardware. It was never merged, unfortunately :(.
> >
> > True.
>
> Would it be possible to keep DMA support if device tree support was
> added for SuperH? I think Rich eventually wanted to merge the patches,
> there were just some minor issues with them.

Adding DT support would definitely make things easier, but would be a lot
of work.
However, using dma_slave_map would be an alternative.

> >>> Anyone who cares for DMA on SuperH?
> >>
> >> What is DMA used for on SuperH? Wouldn't dropping it cut support for
> >> essential hardware features?
> >
> > It may make a few things slower.
> >
> > Does any of your SuperH boards use DMA?
> > Anything interesting in /proc or /sys w.r.t. DMA?
>
> I have:
>
> root@tirpitz:/sys> find . -iname "*dma*"
> ./bus/dma
> ./bus/dma/devices/dma0
> ./bus/dma/devices/dma1
> ./bus/dma/devices/dma2
> ./bus/dma/devices/dma3
> ./bus/dma/devices/dma4
> ./bus/dma/devices/dma5
> ./bus/dma/devices/dma6
> ./bus/dma/devices/dma7
> ./bus/dma/devices/dma8
> ./bus/dma/devices/dma9
> ./bus/dma/devices/dma10
> ./bus/dma/devices/dma11
> ./bus/platform/devices/sh_dmac
> ./bus/platform/devices/sh-dma-engine.0
> ./bus/platform/devices/sh-dma-engine.1

So you do have the two dmaengines...

> On my SH-7785LCR.

... but are they actually used?

git grep -E "(SHDMA|sh_dmae_slave_config)" -- "arch/sh/*7785*"
doesn't come up with any matches, so I don't think any sh7785 platform
is wired to use DMA (yet), only sh7757 and sh772[234].

To double-check:
With current upstream, you can look for "slave" symlinks in sysfs.
With older kernels, you can look at the interrupt counts for the DMACs
in /proc/interrupts.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds