On Fri, Dec 09, 2016 at 12:34:55PM +0100, Arkadiusz Hiler wrote:
> On Fri, Dec 09, 2016 at 11:10:03AM +0000, Chris Wilson wrote:
> > On Fri, Dec 09, 2016 at 11:56:10AM +0100, Arkadiusz Hiler wrote:
> > > On Thu, Dec 08, 2016 at 03:02:14PM -0800, anushasr wrote:
> > > > -static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
> > > > +u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
> > > > {
> > > > u32 wopcm_size = GUC_WOPCM_TOP;
> > > >
> > > > @@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
> > > > if (err)
> > > > goto fail;
> > > >
> > > > + intel_huc_load(dev_priv);
> >
> > We don't need error handling? That would simplify a lot of our code!
> > -Chris
>
> With this patch series on this specific piece of code - not really.
> HuC support it intorduce is _best-eforrty_.
>
> If the function would report error we would not act on it in anyway
> other than logging the fail (which the function already does for us).
>
> As Anusha discussed here, there will be some code reorganization due to
> introduction of i915.enable_huc and deprecation of enable_guc_loading.
> Once we want to have enable_huc=2, there is a reason to change the
> signature and report errors.
Ergh, I've got signatures of auth and load mixed up. But most of the
statement above still semi-holds.
--
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
