On Tue, Apr 28, 2020 at 10:55:49AM +0300, Jani Nikula wrote:
> On Thu, 23 Apr 2020, Imre Deak <imre.deak@xxxxxxxxx> wrote:
> > Using an AUX channel which by default belongs to a non-TypeC PHY won't
> > work on a TypeC PHY, since - as a side-effect besides providing an AUX
> > channel - the AUX channel power well affects power manangement specific
> > to the TypeC subsystem. Using a TypeC AUX channel on a non-TypeC PHY
> > would probably also cause problems, so for simplicity prevent both.
> >
> > This fixes at least an ICL-Y machine in CI, which has a buggy VBT
> > setting AUX-B as an alternative channel for port C.
>
> Is it a production machine?
Yes.
> Not happy about adding stuff for pre-pro machines with buggy VBT.
> It'll bite us later. It always has.
If there is a buggy VBT with this issue it will cause a problem
somewhere down the pipeline which is difficult to track down, power well
timeouts, machine hangs etc. I would like to catch this early and avoid
having to spend time debugging these other issues.
> Also, hate to see VBT code call into intel_display.c (intel_phy_is_tc).
That's the way to determine if a port/PHY is TypeC on a platform or not.
--Imre
>
> BR,
> Jani.
>
>
> >
> > Signed-off-by: Imre Deak <imre.deak@xxxxxxxxx>
> > ---
> > drivers/gpu/drm/i915/display/intel_bios.c | 84 +++++++++++++++--------
> > 1 file changed, 57 insertions(+), 27 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_bios.c b/drivers/gpu/drm/i915/display/intel_bios.c
> > index 839124647202..10d463723d12 100644
> > --- a/drivers/gpu/drm/i915/display/intel_bios.c
> > +++ b/drivers/gpu/drm/i915/display/intel_bios.c
> > @@ -1538,11 +1538,38 @@ static enum port get_port_by_aux_ch(struct drm_i915_private *i915, u8 aux_ch)
> > return PORT_NONE;
> > }
> >
> > +static enum aux_ch
> > +intel_bios_port_info_aux_ch(const struct ddi_vbt_port_info *info)
> > +{
> > + switch (info->alternate_aux_channel) {
> > + case DP_AUX_A:
> > + return AUX_CH_A;
> > + case DP_AUX_B:
> > + return AUX_CH_B;
> > + case DP_AUX_C:
> > + return AUX_CH_C;
> > + case DP_AUX_D:
> > + return AUX_CH_D;
> > + case DP_AUX_E:
> > + return AUX_CH_E;
> > + case DP_AUX_F:
> > + return AUX_CH_F;
> > + case DP_AUX_G:
> > + return AUX_CH_G;
> > + default:
> > + MISSING_CASE(info->alternate_aux_channel);
> > + return AUX_CH_A;
> > + }
> > +}
> > +
> > static void sanitize_aux_ch(struct drm_i915_private *dev_priv,
> > enum port port)
> > {
> > struct ddi_vbt_port_info *info = &dev_priv->vbt.ddi_port_info[port];
> > enum port p;
> > + enum aux_ch aux_ch;
> > + bool aux_is_tc;
> > + bool phy_is_tc;
> >
> > if (!info->alternate_aux_channel)
> > return;
> > @@ -1571,6 +1598,35 @@ static void sanitize_aux_ch(struct drm_i915_private *dev_priv,
> >
> > info->supports_dp = false;
> > info->alternate_aux_channel = 0;
> > +
> > + return;
> > + }
> > +
> > + aux_ch = intel_bios_port_info_aux_ch(info);
> > + /* The AUX CH -> default port is a 1:1 mapping. */
> > + aux_is_tc = intel_phy_is_tc(dev_priv,
> > + intel_port_to_phy(dev_priv,
> > + (enum port)aux_ch));
> > + phy_is_tc = intel_phy_is_tc(dev_priv,
> > + intel_port_to_phy(dev_priv, port));
> > + if (aux_is_tc != phy_is_tc) {
> > + /*
> > + * Using an AUX channel which by default belongs to a TypeC
> > + * PHY can't be used for non-TypeC PHYs and vice-versa. The
> > + * reason is that TypeC AUX power wells can only be enabled in
> > + * the current TypeC mode of the PHY and have an effect on power
> > + * management specific to the TypeC subsystem.
> > + */
> > + drm_dbg_kms(&dev_priv->drm,
> > + "Port %c on a %s PHY is trying to use the %s AUX CH %c, "
> > + "disabling DP support on this port.\n",
> > + port_name(port),
> > + phy_is_tc ? "TypeC" : "non-TypeC",
> > + aux_is_tc ? "TypeC" : "non-TypeC",
> > + aux_ch_name(aux_ch));
> > +
> > + info->supports_dp = false;
> > + info->alternate_aux_channel = 0;
> > }
> > }
> >
> > @@ -2595,33 +2651,7 @@ enum aux_ch intel_bios_port_aux_ch(struct drm_i915_private *dev_priv,
> > return aux_ch;
> > }
> >
> > - switch (info->alternate_aux_channel) {
> > - case DP_AUX_A:
> > - aux_ch = AUX_CH_A;
> > - break;
> > - case DP_AUX_B:
> > - aux_ch = AUX_CH_B;
> > - break;
> > - case DP_AUX_C:
> > - aux_ch = AUX_CH_C;
> > - break;
> > - case DP_AUX_D:
> > - aux_ch = AUX_CH_D;
> > - break;
> > - case DP_AUX_E:
> > - aux_ch = AUX_CH_E;
> > - break;
> > - case DP_AUX_F:
> > - aux_ch = AUX_CH_F;
> > - break;
> > - case DP_AUX_G:
> > - aux_ch = AUX_CH_G;
> > - break;
> > - default:
> > - MISSING_CASE(info->alternate_aux_channel);
> > - aux_ch = AUX_CH_A;
> > - break;
> > - }
> > + aux_ch = intel_bios_port_info_aux_ch(info);
> >
> > drm_dbg_kms(&dev_priv->drm, "using AUX %c for port %c (VBT)\n",
> > aux_ch_name(aux_ch), port_name(port));
>
> --
> Jani Nikula, Intel Open Source Graphics Center
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
