On Wed, Sep 21, 2016 at 2:39 PM, Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx> wrote: >> > @@ -82,7 +82,7 @@ int amdgpu_align_pitch(struct amdgpu_device *adev, int >> > width, int bpp, bool tile >> > aligned += pitch_mask; >> > aligned &= ~pitch_mask; >> > >> > - return aligned; >> > + return aligned * cpp; >> >> Now you multiply by cpp after the rounding. > > That's right, but I don't think that's a problem, as all bpp values returned > by drm_fb_get_bpp_depth() are multiple of 8 bits. Before we have ALIGN(width * cpp, pitch_mask + 1). With your patch we have ALIGN(width, pitch_mask + 1) * cpp. In short we overalign, and instead of e.g. aligning to 256bytes we now align to 256*4 (for xrgb8888). -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
