| Message ID | 20191028163950.9429-1-matthew.auld@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm/i915/blt: fixup block_size rounding | expand |
On Mon, Oct 28, 2019 at 04:39:50PM +0000, Matthew Auld wrote: > There is nothing to say that the obj->base.size is actually a multiple > of the block_size. > > Reported-by: Chris Wilson <chris@chris-wilson.co.uk> > Signed-off-by: Matthew Auld <matthew.auld@intel.com> > Cc: Chris Wilson <chris@chris-wilson.co.uk> > --- > drivers/gpu/drm/i915/gem/i915_gem_object_blt.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c > index 516e61e99212..5597f1345a63 100644 > --- a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c > +++ b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c > @@ -30,7 +30,7 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce, > GEM_BUG_ON(intel_engine_is_virtual(ce->engine)); > intel_engine_pm_get(ce->engine); > > - count = div_u64(vma->size, block_size); > + count = DIV64_U64_ROUND_UP(vma->size, block_size); block_size size look to be u32? > size = (1 + 8 * count) * sizeof(u32); > size = round_up(size, PAGE_SIZE); > pool = intel_engine_get_pool(ce->engine, size); > @@ -214,7 +214,7 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce, > GEM_BUG_ON(intel_engine_is_virtual(ce->engine)); > intel_engine_pm_get(ce->engine); > > - count = div_u64(dst->size, block_size); > + count = DIV64_U64_ROUND_UP(dst->size, block_size); > size = (1 + 11 * count) * sizeof(u32); > size = round_up(size, PAGE_SIZE); > pool = intel_engine_get_pool(ce->engine, size); > -- > 2.20.1 > > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoting Ville Syrjälä (2019-10-28 16:46:46) > On Mon, Oct 28, 2019 at 04:39:50PM +0000, Matthew Auld wrote: > > There is nothing to say that the obj->base.size is actually a multiple > > of the block_size. > > > > Reported-by: Chris Wilson <chris@chris-wilson.co.uk> > > Signed-off-by: Matthew Auld <matthew.auld@intel.com> > > Cc: Chris Wilson <chris@chris-wilson.co.uk> > > --- > > drivers/gpu/drm/i915/gem/i915_gem_object_blt.c | 4 ++-- > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c > > index 516e61e99212..5597f1345a63 100644 > > --- a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c > > +++ b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c > > @@ -30,7 +30,7 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce, > > GEM_BUG_ON(intel_engine_is_virtual(ce->engine)); > > intel_engine_pm_get(ce->engine); > > > > - count = div_u64(vma->size, block_size); > > + count = DIV64_U64_ROUND_UP(vma->size, block_size); > > block_size size look to be u32? And we can control it to be a pot. -Chris
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c index 516e61e99212..5597f1345a63 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c @@ -30,7 +30,7 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce, GEM_BUG_ON(intel_engine_is_virtual(ce->engine)); intel_engine_pm_get(ce->engine); - count = div_u64(vma->size, block_size); + count = DIV64_U64_ROUND_UP(vma->size, block_size); size = (1 + 8 * count) * sizeof(u32); size = round_up(size, PAGE_SIZE); pool = intel_engine_get_pool(ce->engine, size); @@ -214,7 +214,7 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce, GEM_BUG_ON(intel_engine_is_virtual(ce->engine)); intel_engine_pm_get(ce->engine); - count = div_u64(dst->size, block_size); + count = DIV64_U64_ROUND_UP(dst->size, block_size); size = (1 + 11 * count) * sizeof(u32); size = round_up(size, PAGE_SIZE); pool = intel_engine_get_pool(ce->engine, size);
There is nothing to say that the obj->base.size is actually a multiple of the block_size. Reported-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Matthew Auld <matthew.auld@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> --- drivers/gpu/drm/i915/gem/i915_gem_object_blt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)