[Intel-gfx] [PATCH 2/7] drm/i915: Use the gt in HAS_ENGINE

Chris Wilson chris at chris-wilson.co.uk
Fri Jun 26 18:03:30 UTC 2020


Quoting Daniele Ceraolo Spurio (2020-06-26 18:45:56)
> 
> 
> On 6/26/20 7:35 AM, Chris Wilson wrote:
> > Quoting Daniele Ceraolo Spurio (2020-06-26 00:42:07)
> >> diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
> >> index 26cae4846c82..ddefc52f6e09 100644
> >> --- a/drivers/gpu/drm/i915/gvt/handlers.c
> >> +++ b/drivers/gpu/drm/i915/gvt/handlers.c
> >> @@ -1867,7 +1867,7 @@ static int csfe_chicken1_mmio_write(struct intel_vgpu *vgpu,
> >>          MMIO_F(prefix(BLT_RING_BASE), s, f, am, rm, d, r, w); \
> >>          MMIO_F(prefix(GEN6_BSD_RING_BASE), s, f, am, rm, d, r, w); \
> >>          MMIO_F(prefix(VEBOX_RING_BASE), s, f, am, rm, d, r, w); \
> >> -       if (HAS_ENGINE(dev_priv, VCS1)) \
> >> +       if (HAS_ENGINE(&dev_priv->gt, VCS1)) \
> > 
> > Implicit param! It can switch to gvt->gt for all callsites, killing the
> > dev_priv locals.
> 
> I've switched this to gvt->gt, but unfortunately the locals will have to 
> stay because some of the display register definitions still use dev_priv 
> implicitly to access the display mmio base and the 
> pipe_offsets/trans_offsets arrays.

Curses. Maybe one day.
-Chris


More information about the Intel-gfx mailing list