lkml.org 
[lkml]   [2015]   [Feb]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] ARM: OMAP1: PM: fix some build warnings on 1510-only Kconfigs

On 02/12/2015 11:26 AM, Jon Hunter wrote:
>
> On 02/11/2015 09:14 PM, Tony Lindgren wrote:
>> * Paul Walmsley <paul@pwsan.com> [150211 13:03]:
>>> On Wed, 11 Feb 2015, Tony Lindgren wrote:
>>>
>>>> * Paul Walmsley <paul@pwsan.com> [150210 18:28]:
>>>>> On Tue, 10 Feb 2015, Jon Hunter wrote:
>>>>>> On 07/02/2015 00:23, Paul Walmsley wrote:
>>>>>
>>>>>> Unfortunately, there is not a single TRM for the omap5910 but individual
>>>>>> documents for each chapter in the original TRM. Check out the "OMAP5910
>>>>>> Dual-Core Processor Timer Reference Guide" and possibly the "OMAP5910
>>>>>> Dual-Core Processor Clock Generation and System Reset Management
>>>>>> Reference Guide"
>>>>>>
>>>>>> The omap15xx/5910 did have a 32k timer but as you can see it appears it
>>>>>> was never supported by the kernel for this device (not sure why). I do
>>>>>> recall that there is some errata regarding the 32k timer, if you look at
>>>>>> the omap5910 errata document and search for 32k you should find it.
>>>>>
>>>>> OK thanks for the context. I probably am not going to investigate adding
>>>>> support for this timer on OMAP1510/5910 - am primarily trying to avoid
>>>>> causing a regression on the existing platforms.
>>>>
>>>> At least I've never seen the 32KiHz timer registers in any 15xx
>>>> documentation. Jon are you sure you're not mixing up 5910 (15xx)
>>>> and 5912 (16xx)?
>>>
>>> It's documented in the OMAP5910 Timer Reference Guide (SPRU682A) Section 3
>>> "32-kHz Timer", at the link Jon mentioned. Have not checked the errata
>>> that Jon mentioned though.
>>
>> Interesting. Looks like it's the same as on 16xx at 0xfffb9000.
>> AFAIK that never worked on 15xx. Or maybe the issue was that 15xx
>> is missing the constantly running 32KiHz counter making the timer
>> unusable from PM point of view as the clockevent alone is not enough.
>>
>>> Regarding the patch: I'd suggest keeping the compilation warning fixes
>>> (which was the original purpose of the patch) from anything that changes
>>> the logic too much. That way if there's an error in the patch that
>>> changes the logic and it needs to be reverted, it won't also revert the
>>> warning fixes.
>>
>> Makes sense to me.
>
> Yes that's fine with me as well, I don't wish to over complicate
> matters. I have a couple minor comments though and will respond to the
> latest patch rev.

Actually, nevermind the latest version is fine with me. Jon


\
 
 \ /
  Last update: 2015-02-12 13:41    [W:0.092 / U:0.064 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site