mbox series

[0/3] Remove shdma DT support

Message ID cover.1623405675.git.geert+renesas@glider.be (mailing list archive)
Headers show
Series Remove shdma DT support | expand

Message

Geert Uytterhoeven June 11, 2021, 10:18 a.m. UTC
Hi all,

Documentation/devicetree/bindings/dma/renesas,shdma.txt is one of the
few^W57% of the DT bindings that haven't been converted to json-schema
yet.  These bindings were originally intended to cover all SH/R-Mobile
SoCs, but the DMA multiplexer node and one DMA controller instance were
only ever added to one .dtsi file, for R-Mobile APE6.  Still, DMA
support for R-Mobile APE6 was never completed to the point that it would
actually work, cfr. commit a19788612f51b787 ("dmaengine: sh: Remove
R-Mobile APE6 support").  Later, the mux idea was dropped when
implementing support for DMA on (very similar) R-Car Gen2, cfr.
renesas,rcar-dmac.yaml.

Hence this series removes the Renesas SHDMA Device Tree bindings, the
SHDMA DMA multiplexer driver, and the corresponding description in the
R-Mobile APE6 DTS.

I plan to queue [PATCH 3/3] in renesas-devel for v5.15.

Thanks for your comments!

Geert Uytterhoeven (3):
  dt-bindings: dmaengine: Remove SHDMA Device Tree bindings
  dmaengine: sh: Remove unused shdma-of driver
  ARM: dts: r8a73a4: Remove non-functional DMA support

 .../devicetree/bindings/dma/renesas,shdma.txt | 84 -------------------
 arch/arm/boot/dts/r8a73a4.dtsi                | 44 ----------
 drivers/dma/sh/Makefile                       |  2 +-
 drivers/dma/sh/shdma-of.c                     | 76 -----------------
 4 files changed, 1 insertion(+), 205 deletions(-)
 delete mode 100644 Documentation/devicetree/bindings/dma/renesas,shdma.txt
 delete mode 100644 drivers/dma/sh/shdma-of.c

Comments

Vinod Koul June 16, 2021, 10:38 a.m. UTC | #1
On 11-06-21, 12:18, Geert Uytterhoeven wrote:
> 	Hi all,
> 
> Documentation/devicetree/bindings/dma/renesas,shdma.txt is one of the
> few^W57% of the DT bindings that haven't been converted to json-schema
> yet.  These bindings were originally intended to cover all SH/R-Mobile
> SoCs, but the DMA multiplexer node and one DMA controller instance were
> only ever added to one .dtsi file, for R-Mobile APE6.  Still, DMA
> support for R-Mobile APE6 was never completed to the point that it would
> actually work, cfr. commit a19788612f51b787 ("dmaengine: sh: Remove
> R-Mobile APE6 support").  Later, the mux idea was dropped when
> implementing support for DMA on (very similar) R-Car Gen2, cfr.
> renesas,rcar-dmac.yaml.
> 
> Hence this series removes the Renesas SHDMA Device Tree bindings, the
> SHDMA DMA multiplexer driver, and the corresponding description in the
> R-Mobile APE6 DTS.

Applied 1 & 2, thanks
John Paul Adrian Glaubitz June 16, 2021, 10:42 a.m. UTC | #2
Hi Geert!

On 6/11/21 12:18 PM, Geert Uytterhoeven wrote:
> Hence this series removes the Renesas SHDMA Device Tree bindings, the
> SHDMA DMA multiplexer driver, and the corresponding description in the
> R-Mobile APE6 DTS.
Do these changes make life harder in case we want to convert SH to device
tree as already prepared by Yoshinori Sato? [1]

Adrian

> [1] https://lore.kernel.org/patchwork/cover/693910/
Geert Uytterhoeven June 16, 2021, 11:18 a.m. UTC | #3
Hi Adrian,

On Wed, Jun 16, 2021 at 12:42 PM John Paul Adrian Glaubitz
<glaubitz@physik.fu-berlin.de> wrote:
> On 6/11/21 12:18 PM, Geert Uytterhoeven wrote:
> > Hence this series removes the Renesas SHDMA Device Tree bindings, the
> > SHDMA DMA multiplexer driver, and the corresponding description in the
> > R-Mobile APE6 DTS.
> Do these changes make life harder in case we want to convert SH to device
> tree as already prepared by Yoshinori Sato? [1]

Probably not. The only modern DT-aware DMAC drivers for Renesas
hardware are drivers/dma/sh/{rcar,usb}-dmac.c.  Soon there will be
drivers/dma/sh/rz-dmac.c, for RZ/G2L (and RZ/A1 and RZ/A2 later).
Given the R-Car DMAC is very similar to the SH/R-Mobile DMAC, the
latter may be made to work with rcar-dmac.c, if anyone is willing
to spend cycles on that.  Likewise, the RZ/A DMAC probably has its
roots in older SH SoCs, so that may be helpful for you, too.

Chris (CCed) may know better...

> > [1] https://lore.kernel.org/patchwork/cover/693910/

None of these handle DMA?

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