Re: [PATCH 1/3] ARM: AM33xx: hwmod: Add RNG module data

Re: [PATCH 1/3] ARM: AM33xx: hwmod: Add RNG module data

[Paul Walmsley]

Hi

On Tue, 6 Aug 2013, Lokesh Vutla wrote:

> Add RNG hwmod data for AM33xx SoC.
> 
> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>

I tried to queue this one, but it caused AM33xx boots to hang.  
Thought it might be related to the patch ordering -- seems to me that 
the clock change patch needed to come before this patch -- but if I 
recall correctly, both BeagleBone and BeagleBone-LT still hung.  

Also, this patch created a sparse warning.

Could you take another look at this patch?  Maybe try basing off of this 
tag:

git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending.git tags/for-v3.12/omap-prcm-hwmod

Fix the sparse problem, and see if the patch boots for you?

thanks


- Paul

Re: [PATCH 1/3] ARM: AM33xx: hwmod: Add RNG module data

[Lokesh Vutla]

Hi Paul,
On Thursday 29 August 2013 12:18 AM, Paul Walmsley wrote:
> Hi
> 
> On Tue, 6 Aug 2013, Lokesh Vutla wrote:
> 
>> Add RNG hwmod data for AM33xx SoC.
>>
>> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
> 
> I tried to queue this one, but it caused AM33xx boots to hang.  
> Thought it might be related to the patch ordering -- seems to me that 
> the clock change patch needed to come before this patch -- but if I 
> recall correctly, both BeagleBone and BeagleBone-LT still hung. 
This is because register address space is not passed in hwmod_data as it
is passed from dt. So we need to include 3/3 patch in this series to boot.

Ill fix this sparse warning and post a v2.

PS: Probe for this driver will be failed even after this. Support for this version
of rng ip is already posted and pushed into cryptodev-2.6 tree.

Thanks and regards,
Lokesh
> 
> Also, this patch created a sparse warning.
> 
> Could you take another look at this patch?  Maybe try basing off of this 
> tag:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending.git tags/for-v3.12/omap-prcm-hwmod
> 
> Fix the sparse problem, and see if the patch boots for you?
> 
> thanks
> 
> 
> - Paul
> 

Re: [PATCH 1/3] ARM: AM33xx: hwmod: Add RNG module data

[Paul Walmsley]

Hi

On Thu, 29 Aug 2013, Lokesh Vutla wrote:

> On Thursday 29 August 2013 12:18 AM, Paul Walmsley wrote:
> > Hi
> > 
> > On Tue, 6 Aug 2013, Lokesh Vutla wrote:
> > 
> >> Add RNG hwmod data for AM33xx SoC.
> >>
> >> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
> > 
> > I tried to queue this one, but it caused AM33xx boots to hang.  
> > Thought it might be related to the patch ordering -- seems to me that 
> > the clock change patch needed to come before this patch -- but if I 
> > recall correctly, both BeagleBone and BeagleBone-LT still hung. 
> This is because register address space is not passed in hwmod_data as it
> is passed from dt. So we need to include 3/3 patch in this series to boot.

So this patch has a dependency on a DT data patch?  If so then please 
upstream that through Benoit, and put in the patch description on this 
patch that there's a dependency on that patch.  If you post those together 
in a series, please make sure that the DT patch comes first so it doesn't 
break 'git bisect'. 

I've already queued the clock change patch for upstream.

> PS: Probe for this driver will be failed even after this. Support for 
> this version of rng ip is already posted and pushed into cryptodev-2.6 
> tree.

I don't quite understand, could you please rephrase?

- Paul

Re: [PATCH 1/3] ARM: AM33xx: hwmod: Add RNG module data

[Lokesh Vutla]

Hi Paul,

On Thursday 29 August 2013 05:00 PM, Paul Walmsley wrote:
> Hi
> 
> On Thu, 29 Aug 2013, Lokesh Vutla wrote:
> 
>> On Thursday 29 August 2013 12:18 AM, Paul Walmsley wrote:
>>> Hi
>>>
>>> On Tue, 6 Aug 2013, Lokesh Vutla wrote:
>>>
>>>> Add RNG hwmod data for AM33xx SoC.
>>>>
>>>> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
>>>
>>> I tried to queue this one, but it caused AM33xx boots to hang.  
>>> Thought it might be related to the patch ordering -- seems to me that 
>>> the clock change patch needed to come before this patch -- but if I 
>>> recall correctly, both BeagleBone and BeagleBone-LT still hung. 
>> This is because register address space is not passed in hwmod_data as it
>> is passed from dt. So we need to include 3/3 patch in this series to boot.
> 
> So this patch has a dependency on a DT data patch?  If so then please 
> upstream that through Benoit, and put in the patch description on this 
> patch that there's a dependency on that patch.  If you post those together 
> in a series, please make sure that the DT patch comes first so it doesn't 
> break 'git bisect'.
Ok, Ill repost these two patches and make sure that DT patch comes first
so that 'git bisect' will not break..
 
> 
> I've already queued the clock change patch for upstream.
> 
>> PS: Probe for this driver will be failed even after this. Support for 
>> this version of rng ip is already posted and pushed into cryptodev-2.6 
>> tree.
> 
> I don't quite understand, could you please rephrase?
I just meant that with these patches RNG IP as such is not functional
without the driver changes. 

Thanks and Regards,
Lokesh
> 
> - Paul
>