[Intel-gfx] [PATCH 3/3] drm/i915: Start using comparative INTEL_PCH_TYPE

[Lucas De Marchi]

On Tue, Mar 05, 2019 at 10:38:46AM -0800, Rodrigo Vivi wrote:
>On Tue, Mar 05, 2019 at 09:42:22AM -0800, Lucas De Marchi wrote:
>> On Tue, Mar 05, 2019 at 07:16:47PM +0200, Jani Nikula wrote:
>> > On Tue, 05 Mar 2019, Lucas De Marchi <lucas.demarchi at intel.com> wrote:
>> > > On Tue, Mar 05, 2019 at 04:10:04PM +0200, Jani Nikula wrote:
>> > > > On Mon, 04 Mar 2019, Rodrigo Vivi <rodrigo.vivi at intel.com> wrote:
>> > > > > diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
>> > > > > index e1a051c0fbfe..acd2336bb214 100644
>> > > > > --- a/drivers/gpu/drm/i915/intel_dp.c
>> > > > > +++ b/drivers/gpu/drm/i915/intel_dp.c
>> > > > > @@ -949,8 +949,8 @@ static void intel_pps_get_registers(struct intel_dp *intel_dp,
>> > > > >  	regs->pp_stat = PP_STATUS(pps_idx);
>> > > > >  	regs->pp_on = PP_ON_DELAYS(pps_idx);
>> > > > >  	regs->pp_off = PP_OFF_DELAYS(pps_idx);
>> > > > > -	if (!IS_GEN9_LP(dev_priv) && !HAS_PCH_CNP(dev_priv) &&
>> > > > > -	    !HAS_PCH_ICP(dev_priv))
>> > > > > +	if (INTEL_PCH_TYPE(dev_priv) > PCH_NONE &&
>> > > > > +	    INTEL_PCH_TYPE(dev_priv) < PCH_CNP)
>> > > >
>> > > > This is not right, starts to require PCH.
>> > >
>> > > But in those cases INTEL_PCH_TYPE() will return either NONE or NOP.
>> >
>> > Exactly. Non-PCH platforms before CNP should match, but won't.
>>
>> yeah, right. I misread the !IS_GEN9_LP().
>
>ouch... indeed.
>probably this explains failure on ci for bsw and byt
>
>options:
>
>1. if (!IS_GEN9_LP(dev_priv) && !(INTEL_PCH_TYPE(dev_priv) >= 10))

10? I think you meant PCH_CNP

>
>2. if (INTEL_GEN(dev_priv) <= 8 || IS_GEN9_BC(dev_priv))
>
>3. other ideas?

"all PCHs before CNP, excluding GEN9_LP":

	if (INTEL_PCH_TYPE(dev_priv) < PCH_CNP && !IS_GEN9_LP(dev_priv))


Lucas De Marchi

>
>>
>> Lucas De Marchi
>>
>> >
>> > BR,
>> > Jani.
>> >
>> >
>> > --
>> > Jani Nikula, Intel Open Source Graphics Center