[PATCH v1] davinci: Add additional JTAG code for AM-1808 and OMAP-L138 Rev 2.0 SoCs

Michael Williamson [off-list ref] writes:

On 12/22/2010 05:06 PM, Kevin Hilman wrote:

quoted

Michael Williamson [off-list ref] writes:

quoted

The JTAG variant code for Rev-2.0 silicon of the OMAP-L138 has changed.
In addition, the variant code for the AM-1808 SoC appears to match
the Rev-2.0 code for the OMAP-L138.  Add an additional entry to support
these chips.

Tested on?  

I'm assuming on the Mity platforms, right?

I tested this using an AM-1808 (456 MHz speedgrade) on a MityARM-1808
configured SoM.  I don't have a rev-2.0 silicon OMAP-L138 to try, but
according to the Rev B datasheet the variant will change as described.
It doesn't break the Rev 1 OMAP-L138 silicon MityDSP-L138F SoM.

There was a thread about this in this list (link below) and I also 
have some background on an e2e link at TI (link below).  

I had forgotten that I pulled this patch from Sudhakar Rajashekhara on the arago 
omapl PSP kernel (link to patch below).  It's really his work.

I can resubmit (with tested-by and correct authorship) 

Yes please.  Also add the links to the threads mentioned above.

Thanks,

Kevin

unless someone at TI
is prepping the same patch?  I sent it because the AM-1808 fails to boot without
it.  Thanks.

Y

-Mike

davinci thread:
http://linux.davincidsp.com/pipermail/davinci-linux-open-source/2010-November/021224.html

e2e thread:
http://e2e.ti.com/support/embedded/f/354/p/67290/248486.aspx

arago kernel patch:
http://arago-project.org/git/projects/?p=linux-omapl1.git;a=commit;h=6157618435e313a444cdf059702bd34036a6e2b7

quoted

Kevin

quoted

Signed-off-by: Michael Williamson <redacted>
---
This is against davinci-linux tree.
Changes since v0, removed whitespace fixups not related to patch.

 arch/arm/mach-davinci/da850.c |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mach-davinci/da850.c b/arch/arm/mach-davinci/da850.c
index 78b5ae2..2d0bba5 100644
--- a/arch/arm/mach-davinci/da850.c
+++ b/arch/arm/mach-davinci/da850.c

@@ -764,6 +764,13 @@ static struct davinci_id da850_ids[] = {
 		.cpu_id		= DAVINCI_CPU_ID_DA850,
 		.name		= "da850/omap-l138",
 	},
+	{
+		.variant	= 0x1,
+		.part_no	= 0xb7d1,
+		.manufacturer	= 0x017,	/* 0x02f >> 1 */
+		.cpu_id		= DAVINCI_CPU_ID_DA850,
+		.name		= "da850/omap-l138/am18xx",
+	},
 };
 
 static struct davinci_timer_instance da850_timer_instance[4] = {

Hi Mike,

On Thu, Dec 23, 2010 at 05:41:54, Kevin Hilman wrote:

Michael Williamson [off-list ref] writes:

quoted

On 12/22/2010 05:06 PM, Kevin Hilman wrote:

quoted

Michael Williamson [off-list ref] writes:

quoted

The JTAG variant code for Rev-2.0 silicon of the OMAP-L138 has changed.
In addition, the variant code for the AM-1808 SoC appears to match
the Rev-2.0 code for the OMAP-L138.  Add an additional entry to support
these chips.

Tested on?  

I'm assuming on the Mity platforms, right?

I tested this using an AM-1808 (456 MHz speedgrade) on a MityARM-1808
configured SoM.  I don't have a rev-2.0 silicon OMAP-L138 to try, but
according to the Rev B datasheet the variant will change as described.
It doesn't break the Rev 1 OMAP-L138 silicon MityDSP-L138F SoM.

There was a thread about this in this list (link below) and I also 
have some background on an e2e link at TI (link below).  

I had forgotten that I pulled this patch from Sudhakar Rajashekhara on the arago 
omapl PSP kernel (link to patch below).  It's really his work.

I can resubmit (with tested-by and correct authorship) 

Yes please.  Also add the links to the threads mentioned above.

Also, please add
Reported-by: Nicolas Luna <redacted>

Thanks,
Sudhakar