[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Ben Dooks]

The r8a7790.dtsi file has three sdhi nodes which all have the wrong resource
size for their register block. This causes the sh_modbile_sdhi driver to
fail to communicate with card at-all.

Change each sdhi node size from 0x100 to 0x200 to correct this.

Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
Cc: Magnus Damm <magnus.damm@gmail.com>
Cc: Simon Horman <horms@verge.net.au>
Cc: Linux SH <linux-sh@vger.kernel.org>
Cc: Linux ARM <linux-arm-kernel@lists.infradead.org>
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Tested-by: William Towle <william.towle@codethink.co.uk>
---
 arch/arm/boot/dts/r8a7790.dtsi | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/arm/boot/dts/r8a7790.dtsi b/arch/arm/boot/dts/r8a7790.dtsi
index 68b7b87..e218f25 100644
--- a/arch/arm/boot/dts/r8a7790.dtsi
+++ b/arch/arm/boot/dts/r8a7790.dtsi
@@ -254,7 +254,7 @@
 
 	sdhi0: sd at ee100000 {
 		compatible = "renesas,sdhi-r8a7790";
-		reg = <0 0xee100000 0 0x100>;
+		reg = <0 0xee100000 0 0x200>;
 		interrupt-parent = <&gic>;
 		interrupts = <0 165 IRQ_TYPE_LEVEL_HIGH>;
 		cap-sd-highspeed;
@@ -263,7 +263,7 @@
 
 	sdhi1: sd at ee120000 {
 		compatible = "renesas,sdhi-r8a7790";
-		reg = <0 0xee120000 0 0x100>;
+		reg = <0 0xee120000 0 0x200>;
 		interrupt-parent = <&gic>;
 		interrupts = <0 166 IRQ_TYPE_LEVEL_HIGH>;
 		cap-sd-highspeed;
@@ -272,7 +272,7 @@
 
 	sdhi2: sd at ee140000 {
 		compatible = "renesas,sdhi-r8a7790";
-		reg = <0 0xee140000 0 0x100>;
+		reg = <0 0xee140000 0 0x200>;
 		interrupt-parent = <&gic>;
 		interrupts = <0 167 IRQ_TYPE_LEVEL_HIGH>;
 		cap-sd-highspeed;
@@ -281,7 +281,7 @@
 
 	sdhi3: sd at ee160000 {
 		compatible = "renesas,sdhi-r8a7790";
-		reg = <0 0xee160000 0 0x100>;
+		reg = <0 0xee160000 0 0x200>;
 		interrupt-parent = <&gic>;
 		interrupts = <0 168 IRQ_TYPE_LEVEL_HIGH>;
 		cap-sd-highspeed;
-- 
1.8.5.1

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Kuninori Morimoto]

Hi Ben

Thank you for your patch

> The r8a7790.dtsi file has three sdhi nodes which all have the wrong resource
> size for their register block. This causes the sh_modbile_sdhi driver to
> fail to communicate with card at-all.
> 
> Change each sdhi node size from 0x100 to 0x200 to correct this.
> 
> Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
> Cc: Magnus Damm <magnus.damm@gmail.com>
> Cc: Simon Horman <horms@verge.net.au>
> Cc: Linux SH <linux-sh@vger.kernel.org>
> Cc: Linux ARM <linux-arm-kernel@lists.infradead.org>
> Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
> Tested-by: William Towle <william.towle@codethink.co.uk>
> ---

Unfortunately, sdhi resource size 0x100 is corrent on Renesas SDHI.
The wrong is SDHI/TMIO driver side, not SoC side.
Now, I'm working/sending sh_modbile_sdhi driver fixup patches for R-Car H2,
but it need more time (= there is merge timing issue)

Best regards
---
Kuninori Morimoto

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Ben Dooks]

On 13/12/13 02:37, Kuninori Morimoto wrote:
>
> Hi Ben
>
> Thank you for your patch
>
>> The r8a7790.dtsi file has three sdhi nodes which all have the wrong resource
>> size for their register block. This causes the sh_modbile_sdhi driver to
>> fail to communicate with card at-all.
>>
>> Change each sdhi node size from 0x100 to 0x200 to correct this.
>>
>> Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
>> Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
>> Cc: Magnus Damm <magnus.damm@gmail.com>
>> Cc: Simon Horman <horms@verge.net.au>
>> Cc: Linux SH <linux-sh@vger.kernel.org>
>> Cc: Linux ARM <linux-arm-kernel@lists.infradead.org>
>> Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
>> Tested-by: William Towle <william.towle@codethink.co.uk>
>> ---
>
> Unfortunately, sdhi resource size 0x100 is corrent on Renesas SDHI.
> The wrong is SDHI/TMIO driver side, not SoC side.
> Now, I'm working/sending sh_modbile_sdhi driver fixup patches for R-Car H2,
> but it need more time (= there is merge timing issue)

Ok, thanks. We will keep this in our tree until the SDHI driver
is sorted out.


-- 
Ben Dooks				http://www.codethink.co.uk/
Senior Engineer				Codethink - Providing Genius

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Magnus Damm]

Hey Morimoto-san,

On Fri, Dec 13, 2013 at 11:37 AM, Kuninori Morimoto
<kuninori.morimoto.gx@gmail.com> wrote:
>
> Hi Ben
>
> Thank you for your patch
>
>> The r8a7790.dtsi file has three sdhi nodes which all have the wrong resource
>> size for their register block. This causes the sh_modbile_sdhi driver to
>> fail to communicate with card at-all.
>>
>> Change each sdhi node size from 0x100 to 0x200 to correct this.
>>
>> Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
>> Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
>> Cc: Magnus Damm <magnus.damm@gmail.com>
>> Cc: Simon Horman <horms@verge.net.au>
>> Cc: Linux SH <linux-sh@vger.kernel.org>
>> Cc: Linux ARM <linux-arm-kernel@lists.infradead.org>
>> Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
>> Tested-by: William Towle <william.towle@codethink.co.uk>
>> ---
>
> Unfortunately, sdhi resource size 0x100 is corrent on Renesas SDHI.
> The wrong is SDHI/TMIO driver side, not SoC side.
> Now, I'm working/sending sh_modbile_sdhi driver fixup patches for R-Car H2,
> but it need more time (= there is merge timing issue)

Thanks for supporting Ben regarding this SDHI issue.

Would it be possible for you to share a list of patches needed to get
SDHI working? If some parts are missing then please post them so the
patches are available on public lists.

I would like to make it possible for Ben to test your patch stack if
he happens to have time.

My plan is to try out r8a7791 SDHI early next week.

Cheers,

/ magnus

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Ben Dooks]

On 13/12/13 10:56, Magnus Damm wrote:
> Hey Morimoto-san,
>
> On Fri, Dec 13, 2013 at 11:37 AM, Kuninori Morimoto
> <kuninori.morimoto.gx@gmail.com> wrote:
>>
>> Hi Ben
>>
>> Thank you for your patch
>>
>>> The r8a7790.dtsi file has three sdhi nodes which all have the wrong resource
>>> size for their register block. This causes the sh_modbile_sdhi driver to
>>> fail to communicate with card at-all.
>>>
>>> Change each sdhi node size from 0x100 to 0x200 to correct this.
>>>
>>> Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
>>> Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
>>> Cc: Magnus Damm <magnus.damm@gmail.com>
>>> Cc: Simon Horman <horms@verge.net.au>
>>> Cc: Linux SH <linux-sh@vger.kernel.org>
>>> Cc: Linux ARM <linux-arm-kernel@lists.infradead.org>
>>> Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
>>> Tested-by: William Towle <william.towle@codethink.co.uk>
>>> ---
>>
>> Unfortunately, sdhi resource size 0x100 is corrent on Renesas SDHI.
>> The wrong is SDHI/TMIO driver side, not SoC side.
>> Now, I'm working/sending sh_modbile_sdhi driver fixup patches for R-Car H2,
>> but it need more time (= there is merge timing issue)
>
> Thanks for supporting Ben regarding this SDHI issue.
>
> Would it be possible for you to share a list of patches needed to get
> SDHI working? If some parts are missing then please post them so the
> patches are available on public lists.
>
> I would like to make it possible for Ben to test your patch stack if
> he happens to have time.

Thanks, I think I have all the previous ones that where pushed to the
linux-sh list. If there's a git tree I could pull from that would make
our job easier.

We have got SDHI0 and SDHI2 working with device tree, but we're
currently seeing less than a megabyte a second when direcly dd-ing
blocks from the card. Is this a noted issue? We got ~11MiB without
DMA on the 3.4-ltsi series.

As a note, I will be away from the 20th December untill Janurary 6th
and it is likely the rest of the team here will be away from 21st.

-- 
Ben Dooks				http://www.codethink.co.uk/
Senior Engineer				Codethink - Providing Genius

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Simon Horman]

On Fri, Dec 13, 2013 at 11:04:15AM +0000, Ben Dooks wrote:
> On 13/12/13 10:56, Magnus Damm wrote:
> >Hey Morimoto-san,
> >
> >On Fri, Dec 13, 2013 at 11:37 AM, Kuninori Morimoto
> ><kuninori.morimoto.gx@gmail.com> wrote:
> >>
> >>Hi Ben
> >>
> >>Thank you for your patch
> >>
> >>>The r8a7790.dtsi file has three sdhi nodes which all have the wrong resource
> >>>size for their register block. This causes the sh_modbile_sdhi driver to
> >>>fail to communicate with card at-all.
> >>>
> >>>Change each sdhi node size from 0x100 to 0x200 to correct this.
> >>>
> >>>Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> >>>Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
> >>>Cc: Magnus Damm <magnus.damm@gmail.com>
> >>>Cc: Simon Horman <horms@verge.net.au>
> >>>Cc: Linux SH <linux-sh@vger.kernel.org>
> >>>Cc: Linux ARM <linux-arm-kernel@lists.infradead.org>
> >>>Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
> >>>Tested-by: William Towle <william.towle@codethink.co.uk>
> >>>---
> >>
> >>Unfortunately, sdhi resource size 0x100 is corrent on Renesas SDHI.
> >>The wrong is SDHI/TMIO driver side, not SoC side.
> >>Now, I'm working/sending sh_modbile_sdhi driver fixup patches for R-Car H2,
> >>but it need more time (= there is merge timing issue)
> >
> >Thanks for supporting Ben regarding this SDHI issue.
> >
> >Would it be possible for you to share a list of patches needed to get
> >SDHI working? If some parts are missing then please post them so the
> >patches are available on public lists.
> >
> >I would like to make it possible for Ben to test your patch stack if
> >he happens to have time.
> 
> Thanks, I think I have all the previous ones that where pushed to the
> linux-sh list. If there's a git tree I could pull from that would make
> our job easier.

A good idea, sorry for not already having that in place.

Morimoto-san, could you help me to create a topic branch for R-Car Gen2
SDHI by sending me a list of patches and/or brances in other people's
trees?

> We have got SDHI0 and SDHI2 working with device tree, but we're
> currently seeing less than a megabyte a second when direcly dd-ing
> blocks from the card. Is this a noted issue? We got ~11MiB without
> DMA on the 3.4-ltsi series.
> 
> As a note, I will be away from the 20th December untill Janurary 6th
> and it is likely the rest of the team here will be away from 21st.

I think it is likely that most of Japan will be on vacation from
the 28th to the 6th. I intend to be but I don't speak for others.

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Kuninori Morimoto]

Hi Ben, Magnus, Simon

> Morimoto-san, could you help me to create a topic branch for R-Car Gen2
> SDHI by sending me a list of patches and/or brances in other people's
> trees?

First of all, I need to say sorry to Ben.
The SDHI resource size fix was correct, but only for CH0/CH1.

CH0 0x200
CH1 0x200
CH2 0x100
CH3 0x100

are correct size.
Ben, could you please send v2 patch ?

I will send R-Car H2 Lager SDHI support patch (as RFC) soon.

Best regards
---
Kuninori Morimoto

[PATCH] ARM: shmobile: r8a7790 - fix shdi resource sies

[Ben Dooks]

On 16/12/13 04:18, Kuninori Morimoto wrote:
>
> Hi Ben, Magnus, Simon
>
>> Morimoto-san, could you help me to create a topic branch for R-Car Gen2
>> SDHI by sending me a list of patches and/or brances in other people's
>> trees?
>
> First of all, I need to say sorry to Ben.
> The SDHI resource size fix was correct, but only for CH0/CH1.
>
> CH0 0x200
> CH1 0x200
> CH2 0x100
> CH3 0x100
>
> are correct size.
> Ben, could you please send v2 patch ?

Yes, I will sort out posting a branch later. I will do a new test with
Simon's latest devel branch to check.

-- 
Ben Dooks				http://www.codethink.co.uk/
Senior Engineer				Codethink - Providing Genius