On Fri, Nov 22, 2019 at 07:56:21PM +0200, Ville Syrjala wrote: > From: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > > Limit the scaled clip to only clip at most dst_w/h pixels. > This avoids the problem with clip_scaled() not being able > to return negative values. Since new_src_w/h is now properly > bounded we can remove the clamp()s. > > Cc: Benjamin Gaignard <benjamin.gaignard@xxxxxx> > Cc: Maarten Lankhorst <maarten.lankhorst@xxxxxxxxxxxxxxx> > Cc: Daniel Vetter <daniel@xxxxxxxx> > Testcase: igt/kms_selftest/drm_rect_clip_scaled_signed_vs_unsigned > Signed-off-by: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > --- > drivers/gpu/drm/drm_rect.c | 11 +++++++---- > 1 file changed, 7 insertions(+), 4 deletions(-) > > diff --git a/drivers/gpu/drm/drm_rect.c b/drivers/gpu/drm/drm_rect.c > index 818738e83d06..a9c7f90836f3 100644 > --- a/drivers/gpu/drm/drm_rect.c > +++ b/drivers/gpu/drm/drm_rect.c > @@ -59,6 +59,9 @@ static u32 clip_scaled(u32 src, u32 dst, u32 clip) > if (dst == 0) > return 0; > > + /* Only clip what we have. Keeps the result bounded. */ > + clip = min(clip, dst); > + > tmp = mul_u32_u32(src, dst - clip); > > /* > @@ -94,7 +97,7 @@ bool drm_rect_clip_scaled(struct drm_rect *src, struct drm_rect *dst, btw I just noticed that the kerneldoc for drm_rect_clip_scaled() wasn't updated after Maarten's patch :-/ It still talks about hscale and vscale. I'm going to type a patch to fix that. On your patch here: Reviewed-by: Daniel Vetter <daniel.vetter@xxxxxxxx> > u32 new_src_w = clip_scaled(drm_rect_width(src), > drm_rect_width(dst), diff); > > - src->x1 = clamp_t(int64_t, src->x2 - new_src_w, INT_MIN, INT_MAX); > + src->x1 = src->x2 - new_src_w; > dst->x1 = clip->x1; > } > diff = clip->y1 - dst->y1; > @@ -102,7 +105,7 @@ bool drm_rect_clip_scaled(struct drm_rect *src, struct drm_rect *dst, > u32 new_src_h = clip_scaled(drm_rect_height(src), > drm_rect_height(dst), diff); > > - src->y1 = clamp_t(int64_t, src->y2 - new_src_h, INT_MIN, INT_MAX); > + src->y1 = src->y2 - new_src_h; > dst->y1 = clip->y1; > } > diff = dst->x2 - clip->x2; > @@ -110,7 +113,7 @@ bool drm_rect_clip_scaled(struct drm_rect *src, struct drm_rect *dst, > u32 new_src_w = clip_scaled(drm_rect_width(src), > drm_rect_width(dst), diff); > > - src->x2 = clamp_t(int64_t, src->x1 + new_src_w, INT_MIN, INT_MAX); > + src->x2 = src->x1 + new_src_w; > dst->x2 = clip->x2; > } > diff = dst->y2 - clip->y2; > @@ -118,7 +121,7 @@ bool drm_rect_clip_scaled(struct drm_rect *src, struct drm_rect *dst, > u32 new_src_h = clip_scaled(drm_rect_height(src), > drm_rect_height(dst), diff); > > - src->y2 = clamp_t(int64_t, src->y1 + new_src_h, INT_MIN, INT_MAX); > + src->y2 = src->y1 + new_src_h; > dst->y2 = clip->y2; > } > > -- > 2.23.0 > -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
