Hello.

On 3/2/2016 10:29 AM, Ramesh Shanmugasundaram wrote:

quoted

quoted

Adds CAN FD controller node for r8a7795.

Note: CAN FD controller register base address specified in R-Car Gen3
Hardware User Manual v0.5E is incorrect. The correct address is:

CAN FD - 0xe66c0000

Signed-off-by: Ramesh Shanmugasundaram
[off-list ref]
---
Hi All,

     This patch is based on linux-next (tag:next-20160225) with the

following

quoted

     patches applied on top.

     [PATCH v2] arm64: dts: r8a7795: Add CAN external clock support
     [PATCH] arm64: dts: r8a7795: Add CAN support

     The respective CAN subsystem changes are submitted separately here
(https://www.mail-archive.com/linux-renesas-soc at vger.kernel.org/msg013
88.html)

Thanks,
Ramesh
---
   arch/arm64/boot/dts/renesas/r8a7795.dtsi | 24 ++++++++++++++++++++++++
   1 file changed, 24 insertions(+)

diff --git a/arch/arm64/boot/dts/renesas/r8a7795.dtsi

b/arch/arm64/boot/dts/renesas/r8a7795.dtsi
index a88f8d8..5049ba6 100644

--- a/arch/arm64/boot/dts/renesas/r8a7795.dtsi
+++ b/arch/arm64/boot/dts/renesas/r8a7795.dtsi

@@ -553,6 +553,30 @@
   			status = "disabled";
   		};

+		canfd: canfd at e66c0000 {

     The node name should still be "can at e66c0000", I think.

Thanks for the review.

The SoC has CAN controller too

    Yes, I figured. BTW, is the CAN-FD identical to the older CAN on the 
hardware level? I haven't see you posting the CAN-FD driver yet...

and hence I chose this node name to differentiate. A grep of canfd on sysfs would tell if the controller is enabled. The channels on "net" would still be named "canx".

I'll change it if you still feel "can at e66c0000" is more appropriate.

    It is -- the node names should be generic and ePAPR even has the "can" 
name listed explicitly in the section 2.2.2.

Thanks,
Ramesh

MBR, Sergei