[Intel-gfx] WAs in init_clock_gating?

Siluvery, Arun arun.siluvery at linux.intel.com
Thu Jul 24 14:41:54 CEST 2014


On 24/07/2014 13:33, Daniel Vetter wrote:
> On Thu, Jul 24, 2014 at 11:43:11AM +0100, Siluvery, Arun wrote:
>> On 07/07/2014 22:24, Daniel Vetter wrote:
>>> On Mon, Jul 7, 2014 at 11:16 PM, Jesse Barnes <jbarnes at virtuousgeek.org> wrote:
>>>> I don't think it's unreasonable to use a macro that checks a global
>>>> list for whether to apply a given WA.  They'll be scattered all over,
>>>> but at least it'll be easy to see:
>>>>    1) whether we implement a given workaround
>>>> and
>>>>    2) which platforms & steppings it applies to based on the table.
>>>
>>> Oh, I agree it's not unreasonable. But I'm kinda begging for the
>>> simple solution since months (years?) and haven't gotten it, while
>>> still getting a steady stream of bug reports and issues. So I've
>>> readjusted my expectations ;-)
>>>
>>> If someone delivers the real deal I'll certainly won't reject it.
>>> -Daniel
>>>
>>
>> I am moving bdw workarounds from clock_gating fn to render ring init fn and
>> testing this before and after gpu reset.
>
> Testing = with an igt? Because I'll ask for this ;-)

Yes, triggering gpu reset with igt, at the moment the test fails because 
of this register.
>
>> One of the workaround is to disable STC optimization, reg CACHE_MODE_1 bit6
>> set to 1. I observed that some times after boot this gets reset to 0
>> (default value) even after applying workarounds; other than workarounds no
>> one else seems to write to this function.
>> Any ideas about this behaviour?
>
> gpu init tends to do this, since clock_gating is run before that.

thanks, I will take a look.

regards
Arun

> -Daniel
>




More information about the Intel-gfx mailing list