On Wed, Aug 03, 2016 at 05:00:01PM +0300, Joonas Lahtinen wrote: > On ke, 2016-08-03 at 14:42 +0100, Chris Wilson wrote: > > The ggtt->probe() vfunc is only called once during GGTT initialisation. > > We can keep it in a local for the function, sparing a few bytes in our > > i915_ggtt persistent struct. > > > > Signed-off-by: Chris Wilson <[email protected]> > > --- > > drivers/gpu/drm/i915/i915_gem_gtt.c | 15 ++++++++------- > > drivers/gpu/drm/i915/i915_gem_gtt.h | 2 -- > > 2 files changed, 8 insertions(+), 9 deletions(-) > > > > diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c > > b/drivers/gpu/drm/i915/i915_gem_gtt.c > > index 3fc77776d89b..70178a77d0c8 100644 > > --- a/drivers/gpu/drm/i915/i915_gem_gtt.c > > +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c > > @@ -3215,13 +3215,17 @@ int i915_ggtt_probe_hw(struct drm_device *dev) > > { > > struct drm_i915_private *dev_priv = to_i915(dev); > > struct i915_ggtt *ggtt = &dev_priv->ggtt; > > + int (*probe)(struct i915_ggtt *ggtt); > > int ret; > > > > + ggtt->base.dev = dev; > > + ggtt->base.is_ggtt = true; > > + > > if (INTEL_INFO(dev)->gen <= 5) { > > - ggtt->probe = i915_gmch_probe; > > + probe = i915_gmch_probe; > > ggtt->base.cleanup = i915_gmch_remove; > > } else if (INTEL_INFO(dev)->gen < 8) { > > - ggtt->probe = gen6_gmch_probe; > > + probe = gen6_gmch_probe; > > ggtt->base.cleanup = gen6_gmch_remove; > > > > if (HAS_EDRAM(dev)) > > @@ -3235,14 +3239,11 @@ int i915_ggtt_probe_hw(struct drm_device *dev) > > else > > ggtt->base.pte_encode = snb_pte_encode; > > } else { > > - ggtt->probe = gen8_gmch_probe; > > + probe = gen8_gmch_probe; > > ggtt->base.cleanup = gen6_gmch_remove; > > } > > > > I'm pretty sure I already saw a version of this with no function > pointer at all? I preferred that version.
Yes, I do have a patch to change this around to avoid the vfunc, I was just doing a little drive-by whilst in the vicinity. It's a bigger patch though - but it's already rebased on top of this, so I'll send it along. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
