[Intel-gfx] [PATCH] drm/i915: don't pwrite tiled objects through the gtt

Daniel Vetter daniel at ffwll.ch
Fri Apr 13 16:26:31 CEST 2012


On Fri, Apr 13, 2012 at 03:21:42PM +0100, Chris Wilson wrote:
> On Fri, 13 Apr 2012 15:51:51 +0200, Daniel Vetter <daniel.vetter at ffwll.ch> wrote:
> > ... we will botch up the bit17 swizzling. Furthermore tiled pwrite is
> > a (now) unused slowpath, so no one really cares.
> > 
> > This fixes the last swizzling issues I have with i-g-t on my bit17
> > swizzling i915G. No regression, it's been broken since the dawn of
> > gem, but it's nice for regression tracking when really _all_ i-g-t
> > tests work.
> > 
> > Signed-Off-by: Daniel Vetter <daniel.vetter at ffwll.ch>
> > ---
> >  drivers/gpu/drm/i915/i915_gem.c |    1 +
> >  1 files changed, 1 insertions(+), 0 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> > index 71934dd..9415c07 100644
> > --- a/drivers/gpu/drm/i915/i915_gem.c
> > +++ b/drivers/gpu/drm/i915/i915_gem.c
> > @@ -876,6 +876,7 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
> >  
> >  	if (obj->gtt_space &&
> >  	    obj->cache_level == I915_CACHE_NONE &&
> > +	    obj->tiling_mode == I915_TILING_NONE &&
> >  	    obj->map_and_fenceable &&
> >  	    obj->base.write_domain != I915_GEM_DOMAIN_CPU) {
> >  		ret = i915_gem_gtt_pwrite_fast(dev, obj, args, file);
> 
> static inline int i915_gem_object_swizzle(struct drm_i915_gem_object *obj)
> {
> 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
> 	switch (obj->tiling_mode) {
> 	default:
> 	case I915_TILING_NONE: return I915_BIT_6_SIZZLE_NONE;
> 	case I915_TILING_X: return dev_priv->mm.bit_6_swizzle_x;
> 	case I915_TILING_Y: return dev_priv->mm.bit_6_swizzle_y;
> 	}
> }
> 
> And share with i915_gem_tiling.c. Or is that overkill?

The function we might want here already exists as
i915_gem_object_needs_bit17_swizzle but I've figured that's overkill. If
you do tiled pwrites, you're doing it wrong. No point jumping through
hoops for it.
-Daniel
-- 
Daniel Vetter
Mail: daniel at ffwll.ch
Mobile: +41 (0)79 365 57 48



More information about the Intel-gfx mailing list