Re: [PATCH v3 7/9] drm/i915/tgl: Fix dkl link training

From: Imre Deak <imre.deak@intel.com>
To: "Souza, Jose" <jose.souza@intel.com>
Cc: "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>
Subject: Re: [PATCH v3 7/9] drm/i915/tgl: Fix dkl link training
Date: Wed, 25 Sep 2019 18:25:26 +0300	[thread overview]
Message-ID: <20190925152526.GE8958@ideak-desk.fi.intel.com> (raw)
In-Reply-To: <7e992b6db12bef8584203458ed58f96de0e8096e.camel@intel.com>

On Wed, Sep 25, 2019 at 02:59:02AM +0300, Souza, Jose wrote:
> On Tue, 2019-09-24 at 18:58 +0300, Imre Deak wrote:
> > On Mon, Sep 23, 2019 at 12:55:11PM -0700, José Roberto de Souza
> > wrote:
> > > Link training is failling when running link at 2.7GHz and 1.62GHz
> > > and
> > > following BSpec pll algorithm.
> > > 
> > > Comparing the values calculated and the ones from the reference
> > > table
> > > it looks like MG_CLKTOP2_CORECLKCTL1_A_DIVRATIO should not always
> > > set
> > > to 5. For DP ports ICL mg pll algorithm sets it to 10 or 5 based on
> > > div2 value, that matches with dkl hardcoded table.
> > > 
> > > So implementing this way as it proved to work in HW and leaving a
> > > comment so we know why it do not match BSpec.
> > > 
> > > Issue reported on BSpec 49204.
> > > 
> > > Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
> > > ---
> > >  drivers/gpu/drm/i915/display/intel_dpll_mgr.c | 7 ++++++-
> > >  1 file changed, 6 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
> > > b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
> > > index 496df4095a21..3c881d4cf973 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
> > > @@ -2631,7 +2631,12 @@ static bool icl_mg_pll_find_divisors(int
> > > clock_khz, bool is_dp, bool use_ssc,
> > >  
> > >  			if (div2 >= 2) {
> > >  				if (is_dkl) {
> > > -					a_divratio = 5;
> > > +					/*
> > > +					 * Note: a_divratio not
> > > matching TGL
> > > +					 * BSpec algorithm but matching
> > > +					 * hardcoded values and working
> > > on HW
> > > +					 */
> > > +					a_divratio = 10;
> > 
> > Yes, matches the hardcoded values for CLKTOP2_HSCLKCTL/CORECLKCTL1,
> > with
> > the exception that the hardcoded value for CLKTOP2_HSCLKCTL 5p4 is
> > also
> > incorrect wrt. div2 within that (should be 1 but it's 2).
> > 
> > However we don't have any hardcoded values in Bspec for HDMI except
> > for
> > the 5.94Gbps/div2=1 case. So until otherwise proven, for HDMI/div2>=2
> > I
> > wouldn't change the value, rather have the same logic as for ICL that
> > is
> > is_dp ? 10 : 5;
> 
> I was trying to avoid touch any HDMI calculation as I can't test it but
> it looks sane, so I will keep the if is_dkl, replace to "a_divratio =
> is_dp ? 10 : 5;" and add to the comment that this is the supposed value
> for HDMI but it is not tested.

Ok, you could also assign only tlinedrv based on the platform.

> 
> > 
> > >  					tlinedrv = 1;
> > >  				} else {
> > >  					a_divratio = is_dp ? 10 : 5;
> > > -- 
> > > 2.23.0
> > > 
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx@lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx