Il 04/08/23 09:28, AngeloGioacchino Del Regno ha scritto:
Changes in v10:
- Moved snippet from patch [7/15] to patch [6/15] as that was
intended to be there instead; fixes build issue for patch [6/15]
as pointed out by the kernel text robot (oops, sorry!)
Hello CK,
is there anything wrong about this series?
All commits do work on their own, there are no regressions as we tested this
on multiple Chromebooks (MT8173, MT8183, MT8192, MT8195) and a MT6795 Helio
X10 smartphone, and kept testing for months now.
Can you please pick it?
Thanks,
Angelo
Changes in v9:
- As per previous conversation with CK Hu, added a commit that
de-commonizes the gamma setting function that was used in
both DISP_AAL and DISP_GAMMA, now each of them have their
own .gamma_set() callback (mtk_disp_gamma_set_common() has
been removed).
- Added a change to use bitfield macros in mtk_disp_aal.c
- Added a change to compress of_device_id entries in mtk_disp_aal.c
- Tested again on MT6795, MT8173, MT8186, MT8192, MT8195
Changes in v8:
- Changed lut_size to be a mtk_disp_gamma_set_common() function
parameter to pass lut size from AAL
Changes in v7:
- Added check for NULL dev for AAL-gamma case
- Added get_lut_size callback for AAL-gamma
- Added comment to clarify SoC 10/12 bits support and old vs new
register layout as suggested by Alexandre M.
Changes in v6:
- Fixed smatch warning in patch 11/11, ref.:
https://lore.kernel.org/all/202306101458.lRXHEE0Z-lkp@xxxxxxxxx/
Changes in v5:
- Removed incorrect comment on default LUT size and bits
- Removed useless check for num_lut_banks
- Added comment about CMDQ implementation on patch 5
- Evaluated passing lut size/bits from AAL, idea discarded as
the implementation would be rather tricky while bringing no
benefits.
Changes in v4:
- Fixed assignment typo appeared in v3
Changes in v3:
- Fixed issues due to variables renaming during cleanup (oops)
- This is actually the right series, since v2 was taken from the
wrong kernel tree.... :-)
Changes in v2:
- Added explicit inclusion of linux/bitfield.h in patch [06/11]
This series adds support for GAMMA IP requiring and/or supporting
a 12-bits LUT using a slightly different register layout and programming
sequence for multiple LUT banks: this IP version is currently found
on a number of SoCs, not only including the Chromebook/IoT oriented
Kompanio 1200/1380 MT8195/MT8195T, but also Smartphone chips such as
the Dimensity 9200 (MT6985) and others.
This series was tested on MT8195, MT8192, MT8173, MT6795:
* MT6795, MT8192, MT8173: No regression, works fine.
* MT8195: Color correction is finally working!
AngeloGioacchino Del Regno (15):
drm/mediatek: gamma: Reduce indentation in mtk_gamma_set_common()
drm/mediatek: gamma: Support SoC specific LUT size
drm/mediatek: gamma: Improve and simplify HW LUT calculation
drm/mediatek: gamma: Enable the Gamma LUT table only after programming
drm/mediatek: gamma: Use bitfield macros
drm/mediatek: aal: Use bitfield macros
drm/mediatek: De-commonize disp_aal/disp_gamma gamma_set functions
drm/mediatek: gamma: Support specifying number of bits per LUT
component
drm/mediatek: gamma: Support multi-bank gamma LUT
drm/mediatek: gamma: Add support for 12-bit LUT and MT8195
drm/mediatek: gamma: Make sure relay mode is disabled
drm/mediatek: gamma: Program gamma LUT type for descending or rising
drm/mediatek: aal: Rewrite kerneldoc for struct mtk_disp_aal
drm/mediatek: gamma: Add kerneldoc for struct mtk_disp_gamma
drm/mediatek: aal: Compress of_device_id entries and add sentinel
Jason-JH.Lin (1):
drm/mediatek: gamma: Adjust mtk_drm_gamma_set_common parameters
drivers/gpu/drm/mediatek/mtk_disp_aal.c | 84 ++++++--
drivers/gpu/drm/mediatek/mtk_disp_drv.h | 3 +-
drivers/gpu/drm/mediatek/mtk_disp_gamma.c | 203 ++++++++++++++++----
drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 8 +-
drivers/gpu/drm/mediatek/mtk_drm_crtc.h | 1 -
drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c | 2 +
drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h | 9 +
7 files changed, 256 insertions(+), 54 deletions(-)
