There were some small rounding errors in when clamping with
1.0001 and 0.9999 scaling, solve these and add a testcase for drm
helpers, which can be used to prevent more of these errors in the
future.
The new handling in drm_rect_clip_scaled rounds scaling towards 1.0x but
because the rounding error is only 1 pixel with the new math, we won't
ever get in a situation where we go from 1.001 to .999
Maarten Lankhorst (5):
drm/rect: Round above 1 << 16 upwards to correct scale calculation
functions.
drm/rect: Handle rounding errors in drm_rect_clip_scaled, v3.
drm/i915: Do not adjust scale when out of bounds, v2.
drm/selftests: Rename the Kconfig option to CONFIG_DRM_DEBUG_SELFTEST
drm/selftests: Add drm helper selftest
drivers/gpu/drm/Kconfig | 9 +-
drivers/gpu/drm/Makefile | 2 +-
drivers/gpu/drm/drm_atomic_helper.c | 2 +-
drivers/gpu/drm/drm_rect.c | 62 ++++-
drivers/gpu/drm/i915/intel_sprite.c | 144 +++-------
drivers/gpu/drm/selftests/Makefile | 2 +-
.../gpu/drm/selftests/drm_helper_selftests.h | 9 +
drivers/gpu/drm/selftests/test-drm-helper.c | 247 ++++++++++++++++++
include/drm/drm_rect.h | 3 +-
9 files changed, 348 insertions(+), 132 deletions(-)
create mode 100644 drivers/gpu/drm/selftests/drm_helper_selftests.h
create mode 100644 drivers/gpu/drm/selftests/test-drm-helper.c
--
2.17.0
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
