On Fri, Mar 03, 2017 at 10:54:01AM +0100, Geert Uytterhoeven wrote:
> Hi Simon,
>
> On Thu, Mar 2, 2017 at 3:36 PM, Simon Horman <horms@xxxxxxxxxxxx> wrote:
> > On Tue, Feb 21, 2017 at 03:27:25PM +0100, Geert Uytterhoeven wrote:
> >> On Wed, Dec 7, 2016 at 5:44 PM, Ulrich Hecht
> >> <ulrich.hecht+renesas@xxxxxxxxx> wrote:
> >> > Signed-off-by: Ulrich Hecht <ulrich.hecht+renesas@xxxxxxxxx>
> >> > ---
> >> > arch/arm64/boot/dts/renesas/r8a7796.dtsi | 13 +++++++++++++
> >> > 1 file changed, 13 insertions(+)
> >> >
> >> > diff --git a/arch/arm64/boot/dts/renesas/r8a7796.dtsi b/arch/arm64/boot/dts/renesas/r8a7796.dtsi
> >> > index c5f0df5..782063a 100644
> >> > --- a/arch/arm64/boot/dts/renesas/r8a7796.dtsi
> >> > +++ b/arch/arm64/boot/dts/renesas/r8a7796.dtsi
> >> > @@ -430,6 +430,9 @@
> >> > <&cpg CPG_CORE R8A7796_CLK_S3D1>,
> >> > <&scif_clk>;
> >> > clock-names = "fck", "brg_int", "scif_clk";
> >> > + dmas = <&dmac1 0x51>, <&dmac1 0x50>,
> >> > + <&dmac2 0x51>, <&dmac2 0x50>;
> >> > + dma-names = "tx", "rx", "tx", "rx";
> >> > power-domains = <&sysc R8A7796_PD_ALWAYS_ON>;
> >> > status = "disabled";
> >>
> >> Apparently the DMA properties were added to the HSCIF nodes instead
> >> of the SCIF nodes while applying this patch?
> >>
> >> 3 lines of context is not sufficient to distinguish between the various SCIF
> >> and HSCIF nodes :-(
> >
> > Thanks, I have updated renesas-devel accordingly.
> >
> > I notice in renesas-devel there seems to be one "tx", "rx" DMA per SCIF
> > rather than two added in Ulrich's patch. Is there a reason for that?
>
> You mean renesas-drivers?
Yes.
> Oops, looks like I cherry-picked an older version of the patch, and failed to
> notice after reverting your wrong version
> (perhaps the final version was never part of a pull request from Uli?).
>
> BTW, your version is wrong, too: the second pair of DMA channels
> should point to dmac2, not dmac1, e.g.:
>
> scif0: serial@e6e60000 {
> compatible = "renesas,scif-r8a7796",
> "renesas,rcar-gen3-scif", "renesas,scif";
> reg = <0 0xe6e60000 0 64>;
> interrupts = <GIC_SPI 152 IRQ_TYPE_LEVEL_HIGH>;
> clocks = <&cpg CPG_MOD 207>,
> <&cpg CPG_CORE R8A7796_CLK_S3D1>,
> <&scif_clk>;
> clock-names = "fck", "brg_int", "scif_clk";
> + dmas = <&dmac1 0x51>, <&dmac1 0x50>,
> - + <&dmac1 0x51>, <&dmac1 0x50>;
> + + <&dmac2 0x51>, <&dmac2 0x50>;
>
> + dma-names = "tx", "rx", "tx", "rx";
> power-domains = <&sysc R8A7796_PD_ALWAYS_ON>;
> status = "disabled";
> };
Thanks for noticing. I plan to squash the following incremental change into
the relevant commit. Hopefully that will be the end of this saga.
diff --git a/arch/arm64/boot/dts/renesas/r8a7796.dtsi b/arch/arm64/boot/dts/renesas/r8a7796.dtsi
index 84ecd4248fae..951e351ddae1 100644
--- a/arch/arm64/boot/dts/renesas/r8a7796.dtsi
+++ b/arch/arm64/boot/dts/renesas/r8a7796.dtsi
@@ -569,7 +569,7 @@
<&scif_clk>;
clock-names = "fck", "brg_int", "scif_clk";
dmas = <&dmac1 0x51>, <&dmac1 0x50>,
- <&dmac1 0x51>, <&dmac1 0x50>;
+ <&dmac2 0x51>, <&dmac2 0x50>;
dma-names = "tx", "rx", "tx", "rx";
power-domains = <&sysc R8A7796_PD_ALWAYS_ON>;
status = "disabled";
@@ -585,7 +585,7 @@
<&scif_clk>;
clock-names = "fck", "brg_int", "scif_clk";
dmas = <&dmac1 0x53>, <&dmac1 0x52>,
- <&dmac1 0x53>, <&dmac1 0x52>;
+ <&dmac2 0x53>, <&dmac2 0x52>;
dma-names = "tx", "rx", "tx", "rx";
power-domains = <&sysc R8A7796_PD_ALWAYS_ON>;
status = "disabled";
@@ -644,7 +644,7 @@
<&scif_clk>;
clock-names = "fck", "brg_int", "scif_clk";
dmas = <&dmac1 0x5b>, <&dmac1 0x5a>,
- <&dmac1 0x5b>, <&dmac1 0x5a>;
+ <&dmac2 0x5b>, <&dmac2 0x5a>;
dma-names = "tx", "rx", "tx", "rx";
power-domains = <&sysc R8A7796_PD_ALWAYS_ON>;
status = "disabled";
