[PATCH] Add McSPI devices found in AM33xx SoC to dtsi

[PATCH] Add McSPI devices found in AM33xx SoC to dtsi

[Phil Pokorny]

We're using the BeagleBone for a project and we need the SPI devices
available and configured with all available chip selects.

Please accept the following patch to add these to the base .dtsi file
so they can then be enabled in the beaglebone specific .dts with
minimal fuss.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: patch
Type: application/octet-stream
Size: 1129 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20121008/88a2e9d8/attachment.obj>

[PATCH] Add McSPI devices found in AM33xx SoC to dtsi

[Philip, Avinash]

On Tue, Oct 09, 2012 at 10:40:58, Phil Pokorny wrote:
> We're using the BeagleBone for a project and we need the SPI devices
> available and configured with all available chip selects.
> 
> Please accept the following patch to add these to the base .dtsi file
> so they can then be enabled in the beaglebone specific .dts with
> minimal fuss.

Patch for adding MCSPI in am33xx.dtsi already been sent by me.
https://lkml.org/lkml/2012/9/18/32

Thanks
Avinash

> 

[PATCH] Add McSPI devices found in AM33xx SoC to dtsi

[Phil Pokorny]

On Oct 9, 2012, at 3:36 AM, "Philip, Avinash" <avinashphilip@ti.com>
wrote:

> On Tue, Oct 09, 2012 at 10:40:58, Phil Pokorny wrote:
>> We're using the BeagleBone for a project and we need the SPI devices
>> available and configured with all available chip selects.
>>
>> Please accept the following patch to add these to the base .dtsi file
>> so they can then be enabled in the beaglebone specific .dts with
>> minimal fuss.
>
> Patch for adding MCSPI in am33xx.dtsi already been sent by me.
> https://lkml.org/lkml/2012/9/18/32

Thank you for responding.

Was that accepted?  It didn't seem to be in the Linux-omap git tree I
cloned. I need to check the Torvalds mainline kernel.

I believe the number of chip selects on spi1 is incorrect. There is
only one (1) CS shown in the hardware documentation for AM33xx.

Is OMAP4-McSPI the best match for the hardware?  The mcspi driver
appears to differ only in the register offset used.  A hard coded zero
value for omap2 and a named constant for omap4. The hardware manual
for the AM33xx shows the register map starting at zero.

Thank you,
Phil P.
>

[PATCH] Add McSPI devices found in AM33xx SoC to dtsi

[Philip, Avinash]

On Tue, Oct 09, 2012 at 19:54:28, Phil Pokorny wrote:
> On Oct 9, 2012, at 3:36 AM, "Philip, Avinash" <avinashphilip@ti.com>
> wrote:
> 
> > On Tue, Oct 09, 2012 at 10:40:58, Phil Pokorny wrote:
> >> We're using the BeagleBone for a project and we need the SPI devices
> >> available and configured with all available chip selects.
> >>
> >> Please accept the following patch to add these to the base .dtsi file
> >> so they can then be enabled in the beaglebone specific .dts with
> >> minimal fuss.
> >
> > Patch for adding MCSPI in am33xx.dtsi already been sent by me.
> > https://lkml.org/lkml/2012/9/18/32
> 
> Thank you for responding.
> 
> Was that accepted?  It didn't seem to be in the Linux-omap git tree I
> cloned. I need to check the Torvalds mainline kernel.

It's not accepted.
> 
> I believe the number of chip selects on spi1 is incorrect. There is
> only one (1) CS shown in the hardware documentation for AM33xx.

No, number of chip select for SPI 1 is 2. AM33xx TRM says

TRM section
24.1.2 Unsupported McSPI Features

Feature 			Reason
Chip selects 2 and 3	Not pinned out

Also you can see soc pin outs.

> 
> Is OMAP4-McSPI the best match for the hardware?

Yes.

> The mcspi driver appears to differ only in the register offset used.
> A hard coded zero value for omap2 and a named constant for omap4.
> The hardware manual for the AM33xx shows the register map
>starting at zero.

You can see register offset for am33xx (it is defined at offset 0x100).
Hence offset is required in am33xx.

Thanks
Avinash

> Thank you,
> Phil P.
> >
> 

[PATCH] Add McSPI devices found in AM33xx SoC to dtsi

[Hebbar, Gururaja]

On Tue, Oct 09, 2012 at 19:54:28, Phil Pokorny wrote:
> On Oct 9, 2012, at 3:36 AM, "Philip, Avinash" <avinashphilip@ti.com>
> wrote:
> 
> > On Tue, Oct 09, 2012 at 10:40:58, Phil Pokorny wrote:
> >> We're using the BeagleBone for a project and we need the SPI devices
> >> available and configured with all available chip selects.
> >>
> >> Please accept the following patch to add these to the base .dtsi file
> >> so they can then be enabled in the beaglebone specific .dts with
> >> minimal fuss.
> >
> > Patch for adding MCSPI in am33xx.dtsi already been sent by me.
> > https://lkml.org/lkml/2012/9/18/32
> 
> Thank you for responding.
> 
> Was that accepted?  It didn't seem to be in the Linux-omap git tree I
> cloned. I need to check the Torvalds mainline kernel.
> 
> I believe the number of chip selects on spi1 is incorrect. There is
> only one (1) CS shown in the hardware documentation for AM33xx.

Which Hardware document are you referring to? TRM shows 2 chip
selects for McSPI1

> 
> Is OMAP4-McSPI the best match for the hardware?  The mcspi driver
> appears to differ only in the register offset used.  A hard coded zero
> value for omap2 and a named constant for omap4. The hardware manual
> for the AM33xx shows the register map starting at zero.
> 
> Thank you,
> Phil P.
> >
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 


Regards, 
Gururaja