dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
* [PATCH V9 00/43] Color Pipeline API w/ VKMS
@ 2025-04-30  1:10 Alex Hung
  2025-04-30  1:10 ` [PATCH V9 01/43] drm: Add helper for conversion from signed-magnitude Alex Hung
                   ` (44 more replies)
  0 siblings, 45 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo

This is an RFC set for a color pipeline API, along with implementations
in VKMS and amdgpu. It is tested with a set of IGT tests that can be
found at [1]. The IGT tests run a pixel-by-pixel comparison with an
allowable delta variation as the goal for these transformations is
perceptual correctness, not complete pixel accuracy.

v5 of this patchset fleshed out documentation for colorops and the
various defines that are being introduced.

v6 addresses a few comments from various reviewers.

v7 simplifies 3D LUT and addresses more comments from various reviewers.

v8 fixes typo and errors and address comments from reviewers.

v9 refactors cleanup functions, fixes typo and errors, and addresses
   comments from reviewers.

VKMS supports two named transfer function colorops and two matrix
colorops.

Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:

1. 1D Curve EOTF
2. 3x4 CTM
3. Multiplier
4. 1D Curve Inverse EOTF
5. 1D LUT
6. 3D LUT
7. 1D Curve EOTF
8. 1D LUT

The supported curves for the 1D Curve type are:
- sRGB EOTF and its inverse
- PQ EOTF, scaled to [0.0, 125.0] and its inverse
- BT.2020/BT.709 OETF and its inverse

Note that the 1st and 5th colorops take the EOTF or Inverse
OETF while the 3rd colorop takes the Inverse EOTF or OETF.

The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
exists for other drivers to describe their own 3D LUT capability.

This mirrors the color pipeline used by gamescope and presented by
Melissa Wen, with the exception of the DEGAM LUT, which is not currently
used. See [1]
https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf

At this point we're hoping to see gamescope and kwin implementations
take shape. The existing pipeline should be enough to satisfy the
gamescope use-cases on the drm_plane.

In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
support to the color pipeline. I have sketched these out already but
don't have it all hooked up yet. This should not hinder adoption of this
API for gaming use-cases.

We'll also want to advertise IN_FORMATS on a color pipeline as some
color pipelines won't be able to work for all IN_FORMATS on a plane.
Again, I have a sketch but no full implementation yet. This is not
currently required by the AMD color pipeline and could be added after
the merge of this set.

VKMS patches could still be improved in a few ways, though the
payoff might be limited and I would rather focus on other work
at the moment. The most obvious thing to improve would be to
eliminate the hard-coded LUTs for identity, and sRGB, and replace
them with fixed-point math instead.

There are plenty of things that I would like to see, but they could
be added after the merge of this patchset:
 - COLOR_ENCODING and COLOR_RANGE
 - IN_FORMATS for a color pipeline
 - Is it possible to support HW which can't bypass entire pipeline?
 - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
 - read-only scaling colorop which defines scaling taps and position
 - named matrices, for things like converting YUV to RGB
 - Add custom LUT colorops to VKMS

IGT tests can be found at [1] or on the igt-dev mailing list. There
have been no updates since v5 and rebase on latest main is straight-
forward.

A kernel branch can be found at [2].

[1] https://gitlab.freedesktop.org/alex.hung/igt-gpu-tools/-/tree/amd-color-pipeline-v8
[2] https://gitlab.freedesktop.org/alex.hung/linux/-/tree/amd-color-pipeline-v9

v9:
 - Update RFC documents for 3DLUT and fallback behaviours (Simon Ser)
 - Specify colorop function names by _plane_ (Chaitanya Kumar Borah)
 - Remove redundant comments (Simon Ser)
 - Fix typo in commit description (Shengyu Qu)
 - Move destroy and cleanup functions earlier (Simon Ser)
 - Move DRM_COLOROP_1D_CURVE_BT2020_* from middle to end (Simon Ser)
 - Chagne "bool allow_bypass" to "uint32_t flags" for better extensibility (Simon Ser)
 - Return a value in __set_dm_plane_colorop_3dlut

v8:
 - Change VKMS config names (Louis Chauvet)
 - Remove deprecated function "drm_atomic_get_existing_colorop_state" (Louis Chauvet)
 - Remove null check in drm_colorop_set_next_property (Simon Ser)
 - Remove MAX_COLOR_PIPELINES in drm (Simon Ser)
 - Update kernel docs and documents for DRM_COLOROP_3D_LUT (Simon Ser)
 - Add comments for dmr_color_lut (Louis Chauvet)
 - Fix typos and replace DRM_ERROR and DRM_WARN_ONCE by drm_err drm_WARN_ONCE (Louis Chauvet)
 - Fix incorrect conditions in __set_colorop_in_tf_1d_curve (Leo Li)
 - Add DRM_MODE_PROP_ATOMIC to drm_property_create_range (Simon Ser)
 - Change "1D Curve Custom LUT" to "1D LUT" (Simon Ser)
 - Return error when __set_output_tf fails (Leo Li)
 - Return -EINVAL when drm_color_ctm_3x4's size mismatches (Leo Li)
 - Set initialized to 0 and return when drm_lut3d_size is 0 (Harry Wentland)
 - Rework tf->type = TF_TYPE_BYPASS for shaper (Harry Wentland & Leo Li)

v7:
 - Simplify 3D LUT by removing lut_3d_modes and update doc accordingly (Simon Ser)
 - Add destroy / cleanup functions when color pipeline initialization
   fails (Louis Chauvet)
 - Fix checkpatch errors and warning (spaces, long lines, {}, and etc)
 - Change lut's size from drm_colorop_state->size to drm_colorop->lut_size
   and from size_property to lut_size_property
 - Update some kernel updates and commit messagesa (Louis Chauvet)
 - Squash "drm/colorop: Add atomic state print for drm_colorop" and
   "drm/colorop: Add NEXT to colorop state print" (Simon Ser)
 - Add missing MODULE_DESCRIPTION (Jeff Johnson)

v6:
 - Eliminate need to include test as .c file (Louis Chauvet)
 - some doc updates
 - cleanup if colorop alloc or init fails in VKMS and amdgpu (Louis Chauvet)
 - couple other small improvements in VKMS, such as rounding (Louis Chauvet)
 - take ref for DATA blob in duplicate_state func & fix refcount issues (Xaver Hugl)

v5:
 - amdgpu 3D LUT
 - Don't require BYPASS
 - update RFC docs and add to TOC tree
 - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
 - add amdgpu color pipeline doc
 - define SIZE property similar to drm_crtc's GAMMA_SIZE
 - various minor fixes and cleanups

v4:
 - Add amdgpu color pipeline (WIP)
 - Don't block setting of deprecated properties, instead pass client cap
   to atomic check so drivers can ignore these props
 - Drop IOCTL definitions (Pekka)
 - Use enum property for colorop TYPE (Pekka)
 - A few cleanups to the docs (Pekka)
 - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
 - Add missing function declarations (Chaitanya Kumar Borah)
 - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
 - Add helper for creation of pipeline drm_plane property (Pekka)
 - Always create Bypass pipeline (Pekka)
 - A bunch of changes to VKMS kunit tests (Pekka)
 - Fix index in CTM doc (Pekka)

v3:
 - Abandon IOCTLs and discover colorops as clients iterate the pipeline
 - Remove need for libdrm
 - Add color_pipeline client cap and make mutually exclusive with
   COLOR_RANGE and COLOR_ENCODING properties
 - add CTM colorop to VKMS
 - Use include way for kunit testing static functions (Arthur)
 - Make TYPE a range property
 - Move enum drm_colorop_type to uapi header
 - and a bunch of smaller bits that are highlighted in the relevant commit
   description

v2:
 - Rebased on drm-misc-next
 - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
 - Incorporate feedback in color_pipeline.rst doc
 - Add support for sRGB inverse EOTF
 - Add 2nd enumerated TF colorop to VKMS
 - Fix LUTs and some issues with applying LUTs in VKMS

Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
Cc: Simon Ser <contact@emersion.fr>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Melissa Wen <mwen@igalia.com>
Cc: Jonas Ådahl <jadahl@redhat.com>
Cc: Sebastian Wick <sebastian.wick@redhat.com>
Cc: Shashank Sharma <shashank.sharma@amd.com>
Cc: Alexander Goins <agoins@nvidia.com>
Cc: Joshua Ashton <joshua@froggi.es>
Cc: Michel Dänzer <mdaenzer@redhat.com>
Cc: Aleix Pol <aleixpol@kde.org>
Cc: Xaver Hugl <xaver.hugl@gmail.com>
Cc: Victoria Brekenfeld <victoria@system76.com>
Cc: Sima <daniel@ffwll.ch>
Cc: Uma Shankar <uma.shankar@intel.com>
Cc: Naseer Ahmed <quic_naseer@quicinc.com>
Cc: Christopher Braga <quic_cbraga@quicinc.com>
Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
Cc: Arthur Grillo <arthurgrillo@riseup.net>
Cc: Hector Martin <marcan@marcan.st>
Cc: Liviu Dudau <Liviu.Dudau@arm.com>
Cc: Sasha McIntosh <sashamcintosh@google.com>
Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>

Alex Hung (14):
  drm/colorop: Add destroy functions for color pipeline
  drm/colorop: define a new macro for_each_new_colorop_in_state
  drm/amd/display: Skip color pipeline initialization for cursor plane
  drm/amd/display: Add support for sRGB EOTF in DEGAM block
  drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
  drm/amd/display: Add support for sRGB EOTF in BLND block
  drm/colorop: Add 1D Curve Custom LUT type
  drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
  drm/amd/display: add 3x4 matrix colorop
  drm/colorop: Add mutliplier type
  drm/amd/display: add multiplier colorop
  drm/amd/display: Swap matrix and multiplier
  drm/colorop: Add 3D LUT support to color pipeline
  drm/amd/display: add 3D LUT colorop

Harry Wentland (29):
  drm: Add helper for conversion from signed-magnitude
  drm/vkms: Add kunit tests for VKMS LUT handling
  drm/doc/rfc: Describe why prescriptive color pipeline is needed
  drm/colorop: Introduce new drm_colorop mode object
  drm/colorop: Add TYPE property
  drm/colorop: Add 1D Curve subtype
  drm/colorop: Add BYPASS property
  drm/colorop: Add NEXT property
  drm/colorop: Add atomic state print for drm_colorop
  drm/plane: Add COLOR PIPELINE property
  drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
  Documentation/gpu: document drm_colorop
  drm/vkms: Add enumerated 1D curve colorop
  drm/vkms: Add kunit tests for linear and sRGB LUTs
  drm/colorop: Add 3x4 CTM type
  drm/vkms: Use s32 for internal color pipeline precision
  drm/vkms: add 3x4 matrix in color pipeline
  drm/tests: Add a few tests around drm_fixed.h
  drm/vkms: Add tests for CTM handling
  drm/colorop: pass plane_color_pipeline client cap to atomic check
  drm/amd/display: Ignore deprecated props when plane_color_pipeline set
  drm/amd/display: Add bypass COLOR PIPELINE
  drm/colorop: Add PQ 125 EOTF and its inverse
  drm/amd/display: Enable support for PQ 125 EOTF and Inverse
  drm/colorop: add BT2020/BT709 OETF and Inverse OETF
  drm/amd/display: Add support for BT.709 and BT.2020 TFs
  drm/colorop: Define LUT_1D interpolation
  drm/colorop: allow non-bypass colorops
  drm/amd/display: Add AMD color pipeline doc

 Documentation/gpu/drm-kms.rst                 |  15 +
 Documentation/gpu/rfc/color_pipeline.rst      | 378 ++++++++
 Documentation/gpu/rfc/index.rst               |   3 +
 drivers/gpu/drm/Makefile                      |   1 +
 .../gpu/drm/amd/display/amdgpu_dm/Makefile    |   3 +-
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |   4 +
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 573 ++++++++++++-
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 207 +++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
 .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
 drivers/gpu/drm/drm_atomic.c                  | 170 +++-
 drivers/gpu/drm/drm_atomic_helper.c           |  12 +
 drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
 drivers/gpu/drm/drm_atomic_uapi.c             | 175 ++++
 drivers/gpu/drm/drm_colorop.c                 | 598 +++++++++++++
 drivers/gpu/drm/drm_ioctl.c                   |   7 +
 drivers/gpu/drm/drm_mode_config.c             |   7 +
 drivers/gpu/drm/drm_plane.c                   |  59 ++
 drivers/gpu/drm/tests/Makefile                |   3 +-
 drivers/gpu/drm/tests/drm_fixp_test.c         |  71 ++
 drivers/gpu/drm/vkms/Makefile                 |   4 +-
 drivers/gpu/drm/vkms/tests/Makefile           |   2 +-
 drivers/gpu/drm/vkms/tests/vkms_color_test.c  | 459 ++++++++++
 drivers/gpu/drm/vkms/vkms_colorop.c           | 115 +++
 drivers/gpu/drm/vkms/vkms_composer.c          | 134 ++-
 drivers/gpu/drm/vkms/vkms_composer.h          |  27 +
 drivers/gpu/drm/vkms/vkms_drv.h               |   7 +
 drivers/gpu/drm/vkms/vkms_luts.c              | 808 ++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
 drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
 include/drm/drm_atomic.h                      | 111 +++
 include/drm/drm_atomic_uapi.h                 |   3 +
 include/drm/drm_colorop.h                     | 442 ++++++++++
 include/drm/drm_file.h                        |   7 +
 include/drm/drm_fixed.h                       |  18 +
 include/drm/drm_mode_config.h                 |  18 +
 include/drm/drm_plane.h                       |  19 +
 include/uapi/drm/amdgpu_drm.h                 |   9 -
 include/uapi/drm/drm.h                        |  15 +
 include/uapi/drm/drm_mode.h                   | 121 +++
 40 files changed, 4642 insertions(+), 50 deletions(-)
 create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
 create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
 create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
 create mode 100644 drivers/gpu/drm/drm_colorop.c
 create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
 create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_test.c
 create mode 100644 drivers/gpu/drm/vkms/vkms_colorop.c
 create mode 100644 drivers/gpu/drm/vkms/vkms_composer.h
 create mode 100644 drivers/gpu/drm/vkms/vkms_luts.c
 create mode 100644 drivers/gpu/drm/vkms/vkms_luts.h
 create mode 100644 include/drm/drm_colorop.h

-- 
2.43.0


^ permalink raw reply	[flat|nested] 82+ messages in thread

* [PATCH V9 01/43] drm: Add helper for conversion from signed-magnitude
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 02/43] drm/vkms: Add kunit tests for VKMS LUT handling Alex Hung
                   ` (43 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

CTM values are defined as signed-magnitude values. Add
a helper that converts from CTM signed-magnitude fixed
point value to the twos-complement value used by
drm_fixed.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
 include/drm/drm_fixed.h | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/include/drm/drm_fixed.h b/include/drm/drm_fixed.h
index 1922188f00e8..0b44f2f294ce 100644
--- a/include/drm/drm_fixed.h
+++ b/include/drm/drm_fixed.h
@@ -78,6 +78,24 @@ static inline u32 dfixed_div(fixed20_12 A, fixed20_12 B)
 #define DRM_FIXED_EPSILON	1LL
 #define DRM_FIXED_ALMOST_ONE	(DRM_FIXED_ONE - DRM_FIXED_EPSILON)
 
+/**
+ * @drm_sm2fixp
+ *
+ * Convert a 1.31.32 signed-magnitude fixed point to 32.32
+ * 2s-complement fixed point
+ *
+ * @return s64 2s-complement fixed point
+ */
+static inline s64 drm_sm2fixp(__u64 a)
+{
+	if ((a & (1LL << 63))) {
+		return -(a & 0x7fffffffffffffffll);
+	} else {
+		return a;
+	}
+
+}
+
 static inline s64 drm_int2fixp(int a)
 {
 	return ((s64)a) << DRM_FIXED_POINT;
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 02/43] drm/vkms: Add kunit tests for VKMS LUT handling
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
  2025-04-30  1:10 ` [PATCH V9 01/43] drm: Add helper for conversion from signed-magnitude Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 03/43] drm/doc/rfc: Describe why prescriptive color pipeline is needed Alex Hung
                   ` (42 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Debugging LUT math is much easier when we can unit test
it. Add kunit functionality to VKMS and add tests for
 - get_lut_index
 - lerp_u16

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Cc: Arthur Grillo <arthurgrillo@riseup.net>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v8:
 - Update config names (Louis Chauvet)

v7:
 - Fix checkpatch warnings and errors (Louis Chauvet)
  - Change SPDX-License-Identifier: GPL-2.0+ from /* */ to //
  - Fix checkpatch errors and warnings (new line at EOF, redundant spaces, and long lines)
  - Add static to const struct vkms_color_lut test_linear_lut
 - Add "MODULE_DESCRIPTION" (Jeff Johnson)


v6:
 - Eliminate need to include test as .c file (Louis Chauvet)

v5:
 - Bring back static for lerp_u16 and get_lut_index (Arthur)

v4:
 - Test the critical points of the lerp function (Pekka)

v3:
 - Use include way of testing static functions (Arthur)

 drivers/gpu/drm/vkms/tests/Makefile          |   2 +-
 drivers/gpu/drm/vkms/tests/vkms_color_test.c | 172 +++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_composer.c         |   8 +-
 drivers/gpu/drm/vkms/vkms_composer.h         |  13 ++
 4 files changed, 192 insertions(+), 3 deletions(-)
 create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_test.c
 create mode 100644 drivers/gpu/drm/vkms/vkms_composer.h

diff --git a/drivers/gpu/drm/vkms/tests/Makefile b/drivers/gpu/drm/vkms/tests/Makefile
index 9ded37b67a46..e92f7143e540 100644
--- a/drivers/gpu/drm/vkms/tests/Makefile
+++ b/drivers/gpu/drm/vkms/tests/Makefile
@@ -1,3 +1,3 @@
 # SPDX-License-Identifier: GPL-2.0-only
 
-obj-$(CONFIG_DRM_VKMS_KUNIT_TEST) += vkms_config_test.o
+obj-$(CONFIG_DRM_VKMS_KUNIT_TEST) += vkms_config_test.o vkms_color_test.o
diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_test.c b/drivers/gpu/drm/vkms/tests/vkms_color_test.c
new file mode 100644
index 000000000000..affca56cac7b
--- /dev/null
+++ b/drivers/gpu/drm/vkms/tests/vkms_color_test.c
@@ -0,0 +1,172 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <kunit/test.h>
+
+#include <drm/drm_fixed.h>
+#include <drm/drm_mode.h>
+#include "../vkms_drv.h"
+#include "../vkms_composer.h"
+
+#define TEST_LUT_SIZE 16
+
+MODULE_IMPORT_NS("EXPORTED_FOR_KUNIT_TESTING");
+
+static struct drm_color_lut test_linear_array[TEST_LUT_SIZE] = {
+	{ 0x0, 0x0, 0x0, 0 },
+	{ 0x1111, 0x1111, 0x1111, 0 },
+	{ 0x2222, 0x2222, 0x2222, 0 },
+	{ 0x3333, 0x3333, 0x3333, 0 },
+	{ 0x4444, 0x4444, 0x4444, 0 },
+	{ 0x5555, 0x5555, 0x5555, 0 },
+	{ 0x6666, 0x6666, 0x6666, 0 },
+	{ 0x7777, 0x7777, 0x7777, 0 },
+	{ 0x8888, 0x8888, 0x8888, 0 },
+	{ 0x9999, 0x9999, 0x9999, 0 },
+	{ 0xaaaa, 0xaaaa, 0xaaaa, 0 },
+	{ 0xbbbb, 0xbbbb, 0xbbbb, 0 },
+	{ 0xcccc, 0xcccc, 0xcccc, 0 },
+	{ 0xdddd, 0xdddd, 0xdddd, 0 },
+	{ 0xeeee, 0xeeee, 0xeeee, 0 },
+	{ 0xffff, 0xffff, 0xffff, 0 },
+};
+
+static const struct vkms_color_lut test_linear_lut = {
+	.base = test_linear_array,
+	.lut_length = TEST_LUT_SIZE,
+	.channel_value2index_ratio = 0xf000fll
+};
+
+
+static void vkms_color_test_get_lut_index(struct kunit *test)
+{
+	s64 lut_index;
+	int i;
+
+	lut_index = get_lut_index(&test_linear_lut, test_linear_array[0].red);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int(lut_index), 0);
+
+	for (i = 0; i < TEST_LUT_SIZE; i++) {
+		lut_index = get_lut_index(&test_linear_lut, test_linear_array[i].red);
+		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(lut_index), i);
+	}
+}
+
+static void vkms_color_test_lerp(struct kunit *test)
+{
+	/*** half-way round down ***/
+	s64 t = 0x80000000 - 1;
+
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x10, t), 0x8);
+
+	/* odd a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0x10, t), 0x8);
+
+	/* odd b */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0xf, t), 0x8);
+
+	/* b = a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x10, t), 0x10);
+
+	/* b = a + 1 */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x11, t), 0x10);
+
+	/*** half-way round up ***/
+	t = 0x80000000;
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x10, t), 0x8);
+
+	/* odd a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0x10, t), 0x9);
+
+	/* odd b */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0xf, t), 0x8);
+
+	/* b = a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x10, t), 0x10);
+
+	/* b = a + 1 */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x11, t), 0x11);
+
+	/*** t = 0.0 ***/
+	t = 0x0;
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x10, t), 0x0);
+
+	/* odd a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0x10, t), 0x1);
+
+	/* odd b */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0xf, t), 0x1);
+
+	/* b = a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x10, t), 0x10);
+
+	/* b = a + 1 */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x11, t), 0x10);
+
+	/*** t = 1.0 ***/
+	t = 0x100000000;
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x10, t), 0x10);
+
+	/* odd a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0x10, t), 0x10);
+
+	/* odd b */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0xf, t), 0xf);
+
+	/* b = a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x10, t), 0x10);
+
+	/* b = a + 1 */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x11, t), 0x11);
+
+	/*** t = 0.0 + 1 ***/
+	t = 0x0 + 1;
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x10, t), 0x0);
+
+	/* odd a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0x10, t), 0x1);
+
+	/* odd b */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0xf, t), 0x1);
+
+	/* b = a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x10, t), 0x10);
+
+	/* b = a + 1 */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x11, t), 0x10);
+
+	/*** t = 1.0 - 1 ***/
+	t = 0x100000000 - 1;
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x10, t), 0x10);
+
+	/* odd a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0x10, t), 0x10);
+
+	/* odd b */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x1, 0xf, t), 0xf);
+
+	/* b = a */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x10, t), 0x10);
+
+	/* b = a + 1 */
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x10, 0x11, t), 0x11);
+
+	/*** t chosen to verify the flipping point of result a (or b) to a+1 (or b-1) ***/
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x1, 0x80000000 - 1), 0x0);
+	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x1, 0x80000000), 0x1);
+}
+
+static struct kunit_case vkms_color_test_cases[] = {
+	KUNIT_CASE(vkms_color_test_get_lut_index),
+	KUNIT_CASE(vkms_color_test_lerp),
+	{}
+};
+
+static struct kunit_suite vkms_color_test_suite = {
+	.name = "vkms-color",
+	.test_cases = vkms_color_test_cases,
+};
+
+kunit_test_suite(vkms_color_test_suite);
+
+MODULE_DESCRIPTION("Kunit test for VKMS LUT handling");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index fa269d279e25..b0dc95f971d8 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -12,6 +12,8 @@
 #include <linux/minmax.h>
 
 #include "vkms_drv.h"
+#include <kunit/visibility.h>
+#include "vkms_composer.h"
 
 static u16 pre_mul_blend_channel(u16 src, u16 dst, u16 alpha)
 {
@@ -60,7 +62,7 @@ static void fill_background(const struct pixel_argb_u16 *background_color,
 }
 
 // lerp(a, b, t) = a + (b - a) * t
-static u16 lerp_u16(u16 a, u16 b, s64 t)
+VISIBLE_IF_KUNIT u16 lerp_u16(u16 a, u16 b, s64 t)
 {
 	s64 a_fp = drm_int2fixp(a);
 	s64 b_fp = drm_int2fixp(b);
@@ -69,13 +71,15 @@ static u16 lerp_u16(u16 a, u16 b, s64 t)
 
 	return drm_fixp2int_round(a_fp + delta);
 }
+EXPORT_SYMBOL_IF_KUNIT(lerp_u16);
 
-static s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
+VISIBLE_IF_KUNIT s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
 {
 	s64 color_channel_fp = drm_int2fixp(channel_value);
 
 	return drm_fixp_mul(color_channel_fp, lut->channel_value2index_ratio);
 }
+EXPORT_SYMBOL_IF_KUNIT(get_lut_index);
 
 /*
  * This enum is related to the positions of the variables inside
diff --git a/drivers/gpu/drm/vkms/vkms_composer.h b/drivers/gpu/drm/vkms/vkms_composer.h
new file mode 100644
index 000000000000..9316a053e7d7
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_composer.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#ifndef _VKMS_COMPOSER_H_
+#define _VKMS_COMPOSER_H_
+
+#include <kunit/visibility.h>
+
+#if IS_ENABLED(CONFIG_KUNIT)
+u16 lerp_u16(u16 a, u16 b, s64 t);
+s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value);
+#endif
+
+#endif /* _VKMS_COMPOSER_H_ */
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 03/43] drm/doc/rfc: Describe why prescriptive color pipeline is needed
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
  2025-04-30  1:10 ` [PATCH V9 01/43] drm: Add helper for conversion from signed-magnitude Alex Hung
  2025-04-30  1:10 ` [PATCH V9 02/43] drm/vkms: Add kunit tests for VKMS LUT handling Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-05-15  9:30   ` Simon Ser
  2025-04-30  1:10 ` [PATCH V9 04/43] drm/colorop: Introduce new drm_colorop mode object Alex Hung
                   ` (41 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Add documentation for color pipeline API.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update documents according to new 3DLUT changes (Simon Ser)
 - Spell out the behaviours when fallback to shaders/CPU (Simon Ser)

v8:
 - Fix typo "definint" -> "defining"

v7:
 - Add a commit messages

v5:
 - Don't require BYPASS to succeed (Sebastian)
 - use DATA for 1D and 3D LUT types (Sebastian)
 - update 3DLUT ops to use 3DLUT_MODES and 3DLUT_MODE_INDEX
 - Add section on drm_colorop extensibility
 - Add color_pipeline.rst to RFC toc tree

v4:
 - Drop IOCTL docs since we dropped the IOCTLs (Pekka)
 - Clarify reading and setting of COLOR_PIPELINE prop (Pekka)
 - Add blurb about not requiring to reject a pipeline due to
   incompatible ops, as long as op can be bypassed (Pekka)
 - Dropped informational strings (such as input CSC) as they're
   not actually intended to be advertised (Pekka)

v3:
 - Describe DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE (Sebastian)
 - Ask for clear documentation of colorop behavior (Sebastian)

v2:
 - Update colorop visualizations to match reality (Sebastian, Alex Hung)
 - Updated wording (Pekka)
 - Change BYPASS wording to make it non-mandatory (Sebastian)
 - Drop cover-letter-like paragraph from COLOR_PIPELINE Plane Property
   section (Pekka)
 - Use PQ EOTF instead of its inverse in Pipeline Programming example (Melissa)
 - Add "Driver Implementer's Guide" section (Pekka)
 - Add "Driver Forward/Backward Compatibility" section (Sebastian, Pekka)

 Documentation/gpu/rfc/color_pipeline.rst | 378 +++++++++++++++++++++++
 Documentation/gpu/rfc/index.rst          |   3 +
 2 files changed, 381 insertions(+)
 create mode 100644 Documentation/gpu/rfc/color_pipeline.rst

diff --git a/Documentation/gpu/rfc/color_pipeline.rst b/Documentation/gpu/rfc/color_pipeline.rst
new file mode 100644
index 000000000000..cd1cc2d0f988
--- /dev/null
+++ b/Documentation/gpu/rfc/color_pipeline.rst
@@ -0,0 +1,378 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+========================
+Linux Color Pipeline API
+========================
+
+What problem are we solving?
+============================
+
+We would like to support pre-, and post-blending complex color
+transformations in display controller hardware in order to allow for
+HW-supported HDR use-cases, as well as to provide support to
+color-managed applications, such as video or image editors.
+
+It is possible to support an HDR output on HW supporting the Colorspace
+and HDR Metadata drm_connector properties, but that requires the
+compositor or application to render and compose the content into one
+final buffer intended for display. Doing so is costly.
+
+Most modern display HW offers various 1D LUTs, 3D LUTs, matrices, and other
+operations to support color transformations. These operations are often
+implemented in fixed-function HW and therefore much more power efficient than
+performing similar operations via shaders or CPU.
+
+We would like to make use of this HW functionality to support complex color
+transformations with no, or minimal CPU or shader load. The switch between HW
+fixed-function blocks and shaders/CPU must be seamless with no visible
+difference when fallback to shaders/CPU is neceesary at any time.
+
+
+How are other OSes solving this problem?
+========================================
+
+The most widely supported use-cases regard HDR content, whether video or
+gaming.
+
+Most OSes will specify the source content format (color gamut, encoding transfer
+function, and other metadata, such as max and average light levels) to a driver.
+Drivers will then program their fixed-function HW accordingly to map from a
+source content buffer's space to a display's space.
+
+When fixed-function HW is not available the compositor will assemble a shader to
+ask the GPU to perform the transformation from the source content format to the
+display's format.
+
+A compositor's mapping function and a driver's mapping function are usually
+entirely separate concepts. On OSes where a HW vendor has no insight into
+closed-source compositor code such a vendor will tune their color management
+code to visually match the compositor's. On other OSes, where both mapping
+functions are open to an implementer they will ensure both mappings match.
+
+This results in mapping algorithm lock-in, meaning that no-one alone can
+experiment with or introduce new mapping algorithms and achieve
+consistent results regardless of which implementation path is taken.
+
+Why is Linux different?
+=======================
+
+Unlike other OSes, where there is one compositor for one or more drivers, on
+Linux we have a many-to-many relationship. Many compositors; many drivers.
+In addition each compositor vendor or community has their own view of how
+color management should be done. This is what makes Linux so beautiful.
+
+This means that a HW vendor can now no longer tune their driver to one
+compositor, as tuning it to one could make it look fairly different from
+another compositor's color mapping.
+
+We need a better solution.
+
+
+Descriptive API
+===============
+
+An API that describes the source and destination colorspaces is a descriptive
+API. It describes the input and output color spaces but does not describe
+how precisely they should be mapped. Such a mapping includes many minute
+design decision that can greatly affect the look of the final result.
+
+It is not feasible to describe such mapping with enough detail to ensure the
+same result from each implementation. In fact, these mappings are a very active
+research area.
+
+
+Prescriptive API
+================
+
+A prescriptive API describes not the source and destination colorspaces. It
+instead prescribes a recipe for how to manipulate pixel values to arrive at the
+desired outcome.
+
+This recipe is generally an ordered list of straight-forward operations,
+with clear mathematical definitions, such as 1D LUTs, 3D LUTs, matrices,
+or other operations that can be described in a precise manner.
+
+
+The Color Pipeline API
+======================
+
+HW color management pipelines can significantly differ between HW
+vendors in terms of availability, ordering, and capabilities of HW
+blocks. This makes a common definition of color management blocks and
+their ordering nigh impossible. Instead we are defining an API that
+allows user space to discover the HW capabilities in a generic manner,
+agnostic of specific drivers and hardware.
+
+
+drm_colorop Object
+==================
+
+To support the definition of color pipelines we define the DRM core
+object type drm_colorop. Individual drm_colorop objects will be chained
+via the NEXT property of a drm_colorop to constitute a color pipeline.
+Each drm_colorop object is unique, i.e., even if multiple color
+pipelines have the same operation they won't share the same drm_colorop
+object to describe that operation.
+
+Note that drivers are not expected to map drm_colorop objects statically
+to specific HW blocks. The mapping of drm_colorop objects is entirely a
+driver-internal detail and can be as dynamic or static as a driver needs
+it to be. See more in the Driver Implementation Guide section below.
+
+Each drm_colorop has three core properties:
+
+TYPE: An enumeration property, defining the type of transformation, such as
+* enumerated curve
+* custom (uniform) 1D LUT
+* 3x3 matrix
+* 3x4 matrix
+* 3D LUT
+* etc.
+
+Depending on the type of transformation other properties will describe
+more details.
+
+BYPASS: A boolean property that can be used to easily put a block into
+bypass mode. The BYPASS property is not mandatory for a colorop, as long
+as the entire pipeline can get bypassed by setting the COLOR_PIPELINE on
+a plane to '0'.
+
+NEXT: The ID of the next drm_colorop in a color pipeline, or 0 if this
+drm_colorop is the last in the chain.
+
+An example of a drm_colorop object might look like one of these::
+
+    /* 1D enumerated curve */
+    Color operation 42
+    ├─ "TYPE": immutable enum {1D enumerated curve, 1D LUT, 3x3 matrix, 3x4 matrix, 3D LUT, etc.} = 1D enumerated curve
+    ├─ "BYPASS": bool {true, false}
+    ├─ "CURVE_1D_TYPE": enum {sRGB EOTF, sRGB inverse EOTF, PQ EOTF, PQ inverse EOTF, …}
+    └─ "NEXT": immutable color operation ID = 43
+
+    /* custom 4k entry 1D LUT */
+    Color operation 52
+    ├─ "TYPE": immutable enum {1D enumerated curve, 1D LUT, 3x3 matrix, 3x4 matrix, 3D LUT, etc.} = 1D LUT
+    ├─ "BYPASS": bool {true, false}
+    ├─ "SIZE": immutable range = 4096
+    ├─ "DATA": blob
+    └─ "NEXT": immutable color operation ID = 0
+
+    /* 17^3 3D LUT */
+    Color operation 72
+    ├─ "TYPE": immutable enum {1D enumerated curve, 1D LUT, 3x3 matrix, 3x4 matrix, 3D LUT, etc.} = 3D LUT
+    ├─ "BYPASS": bool {true, false}
+    ├─ "SIZE": immutable range = 17
+    ├─ "DATA": blob
+    └─ "NEXT": immutable color operation ID = 73
+
+drm_colorop extensibility
+-------------------------
+
+Unlike existing DRM core objects, like &drm_plane, drm_colorop is not
+extensible. This simplifies implementations and keeps all functionality
+for managing &drm_colorop objects in the DRM core.
+
+If there is a need one may introduce a simple &drm_colorop_funcs
+function table in the future, for example to support an IN_FORMATS
+property on a &drm_colorop.
+
+If a driver requires the ability to create a driver-specific colorop
+object they will need to add &drm_colorop func table support with
+support for the usual functions, like destroy, atomic_duplicate_state,
+and atomic_destroy_state.
+
+
+COLOR_PIPELINE Plane Property
+=============================
+
+Color Pipelines are created by a driver and advertised via a new
+COLOR_PIPELINE enum property on each plane. Values of the property
+always include object id 0, which is the default and means all color
+processing is disabled. Additional values will be the object IDs of the
+first drm_colorop in a pipeline. A driver can create and advertise none,
+one, or more possible color pipelines. A DRM client will select a color
+pipeline by setting the COLOR PIPELINE to the respective value.
+
+NOTE: Many DRM clients will set enumeration properties via the string
+value, often hard-coding it. Since this enumeration is generated based
+on the colorop object IDs it is important to perform the Color Pipeline
+Discovery, described below, instead of hard-coding color pipeline
+assignment. Drivers might generate the enum strings dynamically.
+Hard-coded strings might only work for specific drivers on a specific
+pieces of HW. Color Pipeline Discovery can work universally, as long as
+drivers implement the required color operations.
+
+The COLOR_PIPELINE property is only exposed when the
+DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE is set. Drivers shall ignore any
+existing pre-blend color operations when this cap is set, such as
+COLOR_RANGE and COLOR_ENCODING. If drivers want to support COLOR_RANGE
+or COLOR_ENCODING functionality when the color pipeline client cap is
+set, they are expected to expose colorops in the pipeline to allow for
+the appropriate color transformation.
+
+Setting of the COLOR_PIPELINE plane property or drm_colorop properties
+is only allowed for userspace that sets this client cap.
+
+An example of a COLOR_PIPELINE property on a plane might look like this::
+
+    Plane 10
+    ├─ "TYPE": immutable enum {Overlay, Primary, Cursor} = Primary
+    ├─ …
+    └─ "COLOR_PIPELINE": enum {0, 42, 52} = 0
+
+
+Color Pipeline Discovery
+========================
+
+A DRM client wanting color management on a drm_plane will:
+
+1. Get the COLOR_PIPELINE property of the plane
+2. iterate all COLOR_PIPELINE enum values
+3. for each enum value walk the color pipeline (via the NEXT pointers)
+   and see if the available color operations are suitable for the
+   desired color management operations
+
+If userspace encounters an unknown or unsuitable color operation during
+discovery it does not need to reject the entire color pipeline outright,
+as long as the unknown or unsuitable colorop has a "BYPASS" property.
+Drivers will ensure that a bypassed block does not have any effect.
+
+An example of chained properties to define an AMD pre-blending color
+pipeline might look like this::
+
+    Plane 10
+    ├─ "TYPE" (immutable) = Primary
+    └─ "COLOR_PIPELINE": enum {0, 44} = 0
+
+    Color operation 44
+    ├─ "TYPE" (immutable) = 1D enumerated curve
+    ├─ "BYPASS": bool
+    ├─ "CURVE_1D_TYPE": enum {sRGB EOTF, PQ EOTF} = sRGB EOTF
+    └─ "NEXT" (immutable) = 45
+
+    Color operation 45
+    ├─ "TYPE" (immutable) = 3x4 Matrix
+    ├─ "BYPASS": bool
+    ├─ "DATA": blob
+    └─ "NEXT" (immutable) = 46
+
+    Color operation 46
+    ├─ "TYPE" (immutable) = 1D enumerated curve
+    ├─ "BYPASS": bool
+    ├─ "CURVE_1D_TYPE": enum {sRGB Inverse EOTF, PQ Inverse EOTF} = sRGB EOTF
+    └─ "NEXT" (immutable) = 47
+
+    Color operation 47
+    ├─ "TYPE" (immutable) = 1D LUT
+    ├─ "SIZE": immutable range = 4096
+    ├─ "DATA": blob
+    └─ "NEXT" (immutable) = 48
+
+    Color operation 48
+    ├─ "TYPE" (immutable) = 3D LUT
+    ├─ "DATA": blob
+    └─ "NEXT" (immutable) = 49
+
+    Color operation 49
+    ├─ "TYPE" (immutable) = 1D enumerated curve
+    ├─ "BYPASS": bool
+    ├─ "CURVE_1D_TYPE": enum {sRGB EOTF, PQ EOTF} = sRGB EOTF
+    └─ "NEXT" (immutable) = 0
+
+
+Color Pipeline Programming
+==========================
+
+Once a DRM client has found a suitable pipeline it will:
+
+1. Set the COLOR_PIPELINE enum value to the one pointing at the first
+   drm_colorop object of the desired pipeline
+2. Set the properties for all drm_colorop objects in the pipeline to the
+   desired values, setting BYPASS to true for unused drm_colorop blocks,
+   and false for enabled drm_colorop blocks
+3. Perform (TEST_ONLY or not) atomic commit with all the other KMS
+   states it wishes to change
+
+To configure the pipeline for an HDR10 PQ plane and blending in linear
+space, a compositor might perform an atomic commit with the following
+property values::
+
+    Plane 10
+    └─ "COLOR_PIPELINE" = 42
+
+    Color operation 42
+    └─ "BYPASS" = true
+
+    Color operation 44
+    └─ "BYPASS" = true
+
+    Color operation 45
+    └─ "BYPASS" = true
+
+    Color operation 46
+    └─ "BYPASS" = true
+
+    Color operation 47
+    ├─ "DATA" = Gamut mapping + tone mapping + night mode
+    └─ "BYPASS" = false
+
+    Color operation 48
+    ├─ "CURVE_1D_TYPE" = PQ EOTF
+    └─ "BYPASS" = false
+
+
+Driver Implementer's Guide
+==========================
+
+What does this all mean for driver implementations? As noted above the
+colorops can map to HW directly but don't need to do so. Here are some
+suggestions on how to think about creating your color pipelines:
+
+- Try to expose pipelines that use already defined colorops, even if
+  your hardware pipeline is split differently. This allows existing
+  userspace to immediately take advantage of the hardware.
+
+- Additionally, try to expose your actual hardware blocks as colorops.
+  Define new colorop types where you believe it can offer significant
+  benefits if userspace learns to program them.
+
+- Avoid defining new colorops for compound operations with very narrow
+  scope. If you have a hardware block for a special operation that
+  cannot be split further, you can expose that as a new colorop type.
+  However, try to not define colorops for "use cases", especially if
+  they require you to combine multiple hardware blocks.
+
+- Design new colorops as prescriptive, not descriptive; by the
+  mathematical formula, not by the assumed input and output.
+
+A defined colorop type must be deterministic. The exact behavior of the
+colorop must be documented entirely, whether via a mathematical formula
+or some other description. Its operation can depend only on its
+properties and input and nothing else, allowed error tolerance
+notwithstanding.
+
+
+Driver Forward/Backward Compatibility
+=====================================
+
+As this is uAPI drivers can't regress color pipelines that have been
+introduced for a given HW generation. New HW generations are free to
+abandon color pipelines advertised for previous generations.
+Nevertheless, it can be beneficial to carry support for existing color
+pipelines forward as those will likely already have support in DRM
+clients.
+
+Introducing new colorops to a pipeline is fine, as long as they can be
+bypassed or are purely informational. DRM clients implementing support
+for the pipeline can always skip unknown properties as long as they can
+be confident that doing so will not cause unexpected results.
+
+If a new colorop doesn't fall into one of the above categories
+(bypassable or informational) the modified pipeline would be unusable
+for user space. In this case a new pipeline should be defined.
+
+
+References
+==========
+
+1. https://lore.kernel.org/dri-devel/QMers3awXvNCQlyhWdTtsPwkp5ie9bze_hD5nAccFW7a_RXlWjYB7MoUW_8CKLT2bSQwIXVi5H6VULYIxCdgvryZoAoJnC5lZgyK1QWn488=@emersion.fr/
\ No newline at end of file
diff --git a/Documentation/gpu/rfc/index.rst b/Documentation/gpu/rfc/index.rst
index 396e535377fb..ef19b0ba2a3e 100644
--- a/Documentation/gpu/rfc/index.rst
+++ b/Documentation/gpu/rfc/index.rst
@@ -35,3 +35,6 @@ host such documentation:
 .. toctree::
 
     i915_vm_bind.rst
+
+.. toctree::
+    color_pipeline.rst
\ No newline at end of file
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 04/43] drm/colorop: Introduce new drm_colorop mode object
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (2 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 03/43] drm/doc/rfc: Describe why prescriptive color pipeline is needed Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 05/43] drm/colorop: Add TYPE property Alex Hung
                   ` (40 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

This patches introduces a new drm_colorop mode object. This
object represents color transformations and can be used to
define color pipelines.

We also introduce the drm_colorop_state here, as well as
various helpers and state tracking bits.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v8:
 - Remove drm_atomic_get_existing_colorop_state (Louis Chauvet)

v7:
 - Fix checkpatch warnings and errors
  - Add a tab to for_each_oldnew_colorop_in_state definition
  - Change unsigned index to unsigned int index
  - Fix a checkpatch warning - a new line after variable declaration

v6:
 - Comment that properties validity depends on type (Louis Chauvet)

v5:
 - Add comment to drm_atomic_state.colorops
 - Replace a misplaced 'plane' with 'colorop' in comment
 - Fix colorop_list kernel doc
 - Add kernel doc for color_pipeline
 - drop unused drm_colorop_destroy_state
 - drop drm_colorop_init, to be introduced in later patch
   when used
 - Add kernel docs
 - Drop TODOs

v4:
 - Drop IOCTL definitions (Pekka)
 - add missing declaration (Chaitanya Kumar Borah)

v3:
 - Drop TODO for lock (it's handled in drm_modeset_drop_locks)
   (Melissa)
 - Don't get plane state when getting colorop state
 - Make some functions static (kernel test robot)

 drivers/gpu/drm/Makefile            |   1 +
 drivers/gpu/drm/drm_atomic.c        |  70 ++++++++++++
 drivers/gpu/drm/drm_atomic_helper.c |  12 ++
 drivers/gpu/drm/drm_atomic_uapi.c   |  48 ++++++++
 drivers/gpu/drm/drm_colorop.c       | 104 +++++++++++++++++
 drivers/gpu/drm/drm_mode_config.c   |   7 ++
 include/drm/drm_atomic.h            |  70 ++++++++++++
 include/drm/drm_atomic_uapi.h       |   1 +
 include/drm/drm_colorop.h           | 171 ++++++++++++++++++++++++++++
 include/drm/drm_mode_config.h       |  18 +++
 include/drm/drm_plane.h             |   8 ++
 include/uapi/drm/drm_mode.h         |   1 +
 12 files changed, 511 insertions(+)
 create mode 100644 drivers/gpu/drm/drm_colorop.c
 create mode 100644 include/drm/drm_colorop.h

diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index b5d5561bbe5f..1e1dad810032 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -41,6 +41,7 @@ drm-y := \
 	drm_bridge.o \
 	drm_cache.o \
 	drm_color_mgmt.o \
+	drm_colorop.o \
 	drm_connector.o \
 	drm_crtc.o \
 	drm_displayid.o \
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 0138cf0b8b63..20d6393ec72c 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -42,6 +42,7 @@
 #include <drm/drm_mode.h>
 #include <drm/drm_print.h>
 #include <drm/drm_writeback.h>
+#include <drm/drm_colorop.h>
 
 #include "drm_crtc_internal.h"
 #include "drm_internal.h"
@@ -107,6 +108,7 @@ void drm_atomic_state_default_release(struct drm_atomic_state *state)
 	kfree(state->connectors);
 	kfree(state->crtcs);
 	kfree(state->planes);
+	kfree(state->colorops);
 	kfree(state->private_objs);
 }
 EXPORT_SYMBOL(drm_atomic_state_default_release);
@@ -138,6 +140,10 @@ drm_atomic_state_init(struct drm_device *dev, struct drm_atomic_state *state)
 				sizeof(*state->planes), GFP_KERNEL);
 	if (!state->planes)
 		goto fail;
+	state->colorops = kcalloc(dev->mode_config.num_colorop,
+				  sizeof(*state->colorops), GFP_KERNEL);
+	if (!state->colorops)
+		goto fail;
 
 	/*
 	 * Because drm_atomic_state can be committed asynchronously we need our
@@ -249,6 +255,20 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 		state->planes[i].new_state = NULL;
 	}
 
+	for (i = 0; i < config->num_colorop; i++) {
+		struct drm_colorop *colorop = state->colorops[i].ptr;
+
+		if (!colorop)
+			continue;
+
+		drm_colorop_atomic_destroy_state(colorop,
+						 state->colorops[i].state);
+		state->colorops[i].ptr = NULL;
+		state->colorops[i].state = NULL;
+		state->colorops[i].old_state = NULL;
+		state->colorops[i].new_state = NULL;
+	}
+
 	for (i = 0; i < state->num_private_objs; i++) {
 		struct drm_private_obj *obj = state->private_objs[i].ptr;
 
@@ -568,6 +588,56 @@ drm_atomic_get_plane_state(struct drm_atomic_state *state,
 }
 EXPORT_SYMBOL(drm_atomic_get_plane_state);
 
+
+/**
+ * drm_atomic_get_colorop_state - get colorop state
+ * @state: global atomic state object
+ * @colorop: colorop to get state object for
+ *
+ * This function returns the colorop state for the given colorop, allocating it
+ * if needed. It will also grab the relevant plane lock to make sure that the
+ * state is consistent.
+ *
+ * Returns:
+ *
+ * Either the allocated state or the error code encoded into the pointer. When
+ * the error is EDEADLK then the w/w mutex code has detected a deadlock and the
+ * entire atomic sequence must be restarted. All other errors are fatal.
+ */
+struct drm_colorop_state *
+drm_atomic_get_colorop_state(struct drm_atomic_state *state,
+			     struct drm_colorop *colorop)
+{
+	int ret, index = drm_colorop_index(colorop);
+	struct drm_colorop_state *colorop_state;
+
+	WARN_ON(!state->acquire_ctx);
+
+	colorop_state = drm_atomic_get_new_colorop_state(state, colorop);
+	if (colorop_state)
+		return colorop_state;
+
+	ret = drm_modeset_lock(&colorop->plane->mutex, state->acquire_ctx);
+	if (ret)
+		return ERR_PTR(ret);
+
+	colorop_state = drm_atomic_helper_colorop_duplicate_state(colorop);
+	if (!colorop_state)
+		return ERR_PTR(-ENOMEM);
+
+	state->colorops[index].state = colorop_state;
+	state->colorops[index].ptr = colorop;
+	state->colorops[index].old_state = colorop->state;
+	state->colorops[index].new_state = colorop_state;
+	colorop_state->state = state;
+
+	drm_dbg_atomic(colorop->dev, "Added [COLOROP:%d] %p state to %p\n",
+		       colorop->base.id, colorop_state, state);
+
+	return colorop_state;
+}
+EXPORT_SYMBOL(drm_atomic_get_colorop_state);
+
 static bool
 plane_switching_crtc(const struct drm_plane_state *old_plane_state,
 		     const struct drm_plane_state *new_plane_state)
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
index ee64ca1b1bec..b584faabb72a 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -3060,6 +3060,8 @@ int drm_atomic_helper_swap_state(struct drm_atomic_state *state,
 	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
 	struct drm_plane *plane;
 	struct drm_plane_state *old_plane_state, *new_plane_state;
+	struct drm_colorop *colorop;
+	struct drm_colorop_state *old_colorop_state, *new_colorop_state;
 	struct drm_crtc_commit *commit;
 	struct drm_private_obj *obj;
 	struct drm_private_state *old_obj_state, *new_obj_state;
@@ -3137,6 +3139,16 @@ int drm_atomic_helper_swap_state(struct drm_atomic_state *state,
 		}
 	}
 
+	for_each_oldnew_colorop_in_state(state, colorop, old_colorop_state, new_colorop_state, i) {
+		WARN_ON(colorop->state != old_colorop_state);
+
+		old_colorop_state->state = state;
+		new_colorop_state->state = NULL;
+
+		state->colorops[i].state = old_colorop_state;
+		colorop->state = new_colorop_state;
+	}
+
 	drm_panic_lock(state->dev, flags);
 	for_each_oldnew_plane_in_state(state, plane, old_plane_state, new_plane_state, i) {
 		WARN_ON(plane->state != old_plane_state);
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index c2726af6698e..10e625c17f90 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -35,6 +35,7 @@
 #include <drm/drm_drv.h>
 #include <drm/drm_writeback.h>
 #include <drm/drm_vblank.h>
+#include <drm/drm_colorop.h>
 
 #include <linux/dma-fence.h>
 #include <linux/uaccess.h>
@@ -643,6 +644,26 @@ drm_atomic_plane_get_property(struct drm_plane *plane,
 	return 0;
 }
 
+
+static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
+		struct drm_colorop_state *state, struct drm_file *file_priv,
+		struct drm_property *property, uint64_t val)
+{
+	drm_dbg_atomic(colorop->dev,
+			"[COLOROP:%d] unknown property [PROP:%d:%s]]\n",
+			colorop->base.id,
+			property->base.id, property->name);
+	return -EINVAL;
+}
+
+static int
+drm_atomic_colorop_get_property(struct drm_colorop *colorop,
+		const struct drm_colorop_state *state,
+		struct drm_property *property, uint64_t *val)
+{
+	return -EINVAL;
+}
+
 static int drm_atomic_set_writeback_fb_for_connector(
 		struct drm_connector_state *conn_state,
 		struct drm_framebuffer *fb)
@@ -909,6 +930,16 @@ int drm_atomic_get_property(struct drm_mode_object *obj,
 				plane->state, property, val);
 		break;
 	}
+	case DRM_MODE_OBJECT_COLOROP: {
+		struct drm_colorop *colorop = obj_to_colorop(obj);
+
+		if (colorop->plane)
+			WARN_ON(!drm_modeset_is_locked(&colorop->plane->mutex));
+
+		ret = drm_atomic_colorop_get_property(colorop,
+				colorop->state, property, val);
+		break;
+	}
 	default:
 		drm_dbg_atomic(dev, "[OBJECT:%d] has no properties\n", obj->id);
 		ret = -EINVAL;
@@ -1105,6 +1136,23 @@ int drm_atomic_set_property(struct drm_atomic_state *state,
 		ret = drm_atomic_plane_set_property(plane,
 				plane_state, file_priv,
 				prop, prop_value);
+
+		break;
+	}
+	case DRM_MODE_OBJECT_COLOROP: {
+		struct drm_colorop *colorop = obj_to_colorop(obj);
+		struct drm_colorop_state *colorop_state;
+
+		colorop_state = drm_atomic_get_colorop_state(state, colorop);
+		if (IS_ERR(colorop_state)) {
+			ret = PTR_ERR(colorop_state);
+			break;
+		}
+
+		ret = drm_atomic_colorop_set_property(colorop,
+				colorop_state, file_priv,
+				prop, prop_value);
+
 		break;
 	}
 	default:
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
new file mode 100644
index 000000000000..d215e22c9d20
--- /dev/null
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -0,0 +1,104 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright (C) 2023 Advanced Micro Devices, Inc. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include <drm/drm_colorop.h>
+#include <drm/drm_print.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_plane.h>
+
+#include "drm_crtc_internal.h"
+
+static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
+							struct drm_colorop_state *state)
+{
+	memcpy(state, colorop->state, sizeof(*state));
+}
+
+struct drm_colorop_state *
+drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop)
+{
+	struct drm_colorop_state *state;
+
+	if (WARN_ON(!colorop->state))
+		return NULL;
+
+	state = kmalloc(sizeof(*state), GFP_KERNEL);
+	if (state)
+		__drm_atomic_helper_colorop_duplicate_state(colorop, state);
+
+	return state;
+}
+
+
+void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
+				      struct drm_colorop_state *state)
+{
+	kfree(state);
+}
+
+/**
+ * __drm_colorop_state_reset - resets colorop state to default values
+ * @colorop_state: atomic colorop state, must not be NULL
+ * @colorop: colorop object, must not be NULL
+ *
+ * Initializes the newly allocated @colorop_state with default
+ * values. This is useful for drivers that subclass the CRTC state.
+ */
+static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
+				      struct drm_colorop *colorop)
+{
+	colorop_state->colorop = colorop;
+}
+
+/**
+ * __drm_colorop_reset - reset state on colorop
+ * @colorop: drm colorop
+ * @colorop_state: colorop state to assign
+ *
+ * Initializes the newly allocated @colorop_state and assigns it to
+ * the &drm_crtc->state pointer of @colorop, usually required when
+ * initializing the drivers or when called from the &drm_colorop_funcs.reset
+ * hook.
+ *
+ * This is useful for drivers that subclass the colorop state.
+ */
+static void __drm_colorop_reset(struct drm_colorop *colorop,
+				struct drm_colorop_state *colorop_state)
+{
+	if (colorop_state)
+		__drm_colorop_state_reset(colorop_state, colorop);
+
+	colorop->state = colorop_state;
+}
+
+void drm_colorop_reset(struct drm_colorop *colorop)
+{
+	kfree(colorop->state);
+	colorop->state = kzalloc(sizeof(*colorop->state), GFP_KERNEL);
+
+	if (colorop->state)
+		__drm_colorop_reset(colorop, colorop->state);
+}
diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
index 8642a2fb25a9..369a66bad9da 100644
--- a/drivers/gpu/drm/drm_mode_config.c
+++ b/drivers/gpu/drm/drm_mode_config.c
@@ -29,6 +29,7 @@
 #include <drm/drm_managed.h>
 #include <drm/drm_mode_config.h>
 #include <drm/drm_print.h>
+#include <drm/drm_colorop.h>
 #include <linux/dma-resv.h>
 
 #include "drm_crtc_internal.h"
@@ -191,11 +192,15 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 void drm_mode_config_reset(struct drm_device *dev)
 {
 	struct drm_crtc *crtc;
+	struct drm_colorop *colorop;
 	struct drm_plane *plane;
 	struct drm_encoder *encoder;
 	struct drm_connector *connector;
 	struct drm_connector_list_iter conn_iter;
 
+	drm_for_each_colorop(colorop, dev)
+		drm_colorop_reset(colorop);
+
 	drm_for_each_plane(plane, dev)
 		if (plane->funcs->reset)
 			plane->funcs->reset(plane);
@@ -429,6 +434,7 @@ int drmm_mode_config_init(struct drm_device *dev)
 	INIT_LIST_HEAD(&dev->mode_config.property_list);
 	INIT_LIST_HEAD(&dev->mode_config.property_blob_list);
 	INIT_LIST_HEAD(&dev->mode_config.plane_list);
+	INIT_LIST_HEAD(&dev->mode_config.colorop_list);
 	INIT_LIST_HEAD(&dev->mode_config.privobj_list);
 	idr_init_base(&dev->mode_config.object_idr, 1);
 	idr_init_base(&dev->mode_config.tile_idr, 1);
@@ -450,6 +456,7 @@ int drmm_mode_config_init(struct drm_device *dev)
 	dev->mode_config.num_crtc = 0;
 	dev->mode_config.num_encoder = 0;
 	dev->mode_config.num_total_plane = 0;
+	dev->mode_config.num_colorop = 0;
 
 	if (IS_ENABLED(CONFIG_LOCKDEP)) {
 		struct drm_modeset_acquire_ctx modeset_ctx;
diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index 38636a593c9d..92dde4fb21d2 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -30,6 +30,7 @@
 
 #include <drm/drm_crtc.h>
 #include <drm/drm_util.h>
+#include <drm/drm_colorop.h>
 
 /**
  * struct drm_crtc_commit - track modeset commits on a CRTC
@@ -157,6 +158,11 @@ struct drm_crtc_commit {
 	bool abort_completion;
 };
 
+struct __drm_colorops_state {
+	struct drm_colorop *ptr;
+	struct drm_colorop_state *state, *old_state, *new_state;
+};
+
 struct __drm_planes_state {
 	struct drm_plane *ptr;
 	struct drm_plane_state *state, *old_state, *new_state;
@@ -458,6 +464,14 @@ struct drm_atomic_state {
 	 */
 	bool duplicated : 1;
 
+	/**
+	 * @colorops:
+	 *
+	 * Pointer to array of @drm_colorop and @drm_colorop_state part of this
+	 * update.
+	 */
+	struct __drm_colorops_state *colorops;
+
 	/**
 	 * @planes:
 	 *
@@ -599,6 +613,9 @@ drm_atomic_get_crtc_state(struct drm_atomic_state *state,
 struct drm_plane_state * __must_check
 drm_atomic_get_plane_state(struct drm_atomic_state *state,
 			   struct drm_plane *plane);
+struct drm_colorop_state *
+drm_atomic_get_colorop_state(struct drm_atomic_state *state,
+			     struct drm_colorop *colorop);
 struct drm_connector_state * __must_check
 drm_atomic_get_connector_state(struct drm_atomic_state *state,
 			       struct drm_connector *connector);
@@ -731,6 +748,36 @@ drm_atomic_get_new_plane_state(const struct drm_atomic_state *state,
 	return state->planes[drm_plane_index(plane)].new_state;
 }
 
+/**
+ * drm_atomic_get_old_colorop_state - get colorop state, if it exists
+ * @state: global atomic state object
+ * @colorop: colorop to grab
+ *
+ * This function returns the old colorop state for the given colorop, or
+ * NULL if the colorop is not part of the global atomic state.
+ */
+static inline struct drm_colorop_state *
+drm_atomic_get_old_colorop_state(struct drm_atomic_state *state,
+			       struct drm_colorop *colorop)
+{
+	return state->colorops[drm_colorop_index(colorop)].old_state;
+}
+
+/**
+ * drm_atomic_get_new_colorop_state - get colorop state, if it exists
+ * @state: global atomic state object
+ * @colorop: colorop to grab
+ *
+ * This function returns the new colorop state for the given colorop, or
+ * NULL if the colorop is not part of the global atomic state.
+ */
+static inline struct drm_colorop_state *
+drm_atomic_get_new_colorop_state(struct drm_atomic_state *state,
+			       struct drm_colorop *colorop)
+{
+	return state->colorops[drm_colorop_index(colorop)].new_state;
+}
+
 /**
  * drm_atomic_get_existing_connector_state - get connector state, if it exists
  * @state: global atomic state object
@@ -978,6 +1025,29 @@ void drm_state_dump(struct drm_device *dev, struct drm_printer *p);
 			     (new_crtc_state) = (__state)->crtcs[__i].new_state, \
 			     (void)(new_crtc_state) /* Only to avoid unused-but-set-variable warning */, 1))
 
+/**
+ * for_each_oldnew_colorop_in_state - iterate over all colorops in an atomic update
+ * @__state: &struct drm_atomic_state pointer
+ * @colorop: &struct drm_colorop iteration cursor
+ * @old_colorop_state: &struct drm_colorop_state iteration cursor for the old state
+ * @new_colorop_state: &struct drm_colorop_state iteration cursor for the new state
+ * @__i: int iteration cursor, for macro-internal use
+ *
+ * This iterates over all colorops in an atomic update, tracking both old and
+ * new state. This is useful in places where the state delta needs to be
+ * considered, for example in atomic check functions.
+ */
+#define for_each_oldnew_colorop_in_state(__state, colorop, old_colorop_state, new_colorop_state, __i) \
+	for ((__i) = 0;							\
+	     (__i) < (__state)->dev->mode_config.num_colorop;		\
+	     (__i)++)							\
+		for_each_if ((__state)->colorops[__i].ptr &&		\
+			     ((colorop) = (__state)->colorops[__i].ptr,	\
+			      (void)(colorop) /* Only to avoid unused-but-set-variable warning */, \
+			      (old_colorop_state) = (__state)->colorops[__i].old_state,\
+			      (new_colorop_state) = (__state)->colorops[__i].new_state, 1))
+
+
 /**
  * for_each_oldnew_plane_in_state - iterate over all planes in an atomic update
  * @__state: &struct drm_atomic_state pointer
diff --git a/include/drm/drm_atomic_uapi.h b/include/drm/drm_atomic_uapi.h
index 4c6d39d7bdb2..70a115d523cd 100644
--- a/include/drm/drm_atomic_uapi.h
+++ b/include/drm/drm_atomic_uapi.h
@@ -37,6 +37,7 @@ struct drm_crtc;
 struct drm_connector_state;
 struct dma_fence;
 struct drm_framebuffer;
+struct drm_colorop;
 
 int __must_check
 drm_atomic_set_mode_for_crtc(struct drm_crtc_state *state,
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
new file mode 100644
index 000000000000..008665d2d960
--- /dev/null
+++ b/include/drm/drm_colorop.h
@@ -0,0 +1,171 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright (C) 2023 Advanced Micro Devices, Inc. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DRM_COLOROP_H__
+#define __DRM_COLOROP_H__
+
+#include <drm/drm_mode_object.h>
+#include <drm/drm_mode.h>
+#include <drm/drm_property.h>
+
+/**
+ * struct drm_colorop_state - mutable colorop state
+ */
+struct drm_colorop_state {
+	/** @colorop: backpointer to the colorop */
+	struct drm_colorop *colorop;
+
+	/*
+	 * Color properties
+	 *
+	 * The following fields are not always valid, their usage depends
+	 * on the colorop type. See their associated comment for more
+	 * information.
+	 */
+
+	/** @state: backpointer to global drm_atomic_state */
+	struct drm_atomic_state *state;
+};
+
+/**
+ * struct drm_colorop - DRM color operation control structure
+ *
+ * A colorop represents one color operation. They can be chained via
+ * the 'next' pointer to build a color pipeline.
+ *
+ * Since colorops cannot stand-alone and are used to describe colorop
+ * operations on a plane they don't have their own locking mechanism but
+ * are locked and programmed along with their associated &drm_plane.
+ *
+ */
+struct drm_colorop {
+	/** @dev: parent DRM device */
+	struct drm_device *dev;
+
+	/**
+	 * @head:
+	 *
+	 * List of all colorops on @dev, linked from &drm_mode_config.colorop_list.
+	 * Invariant over the lifetime of @dev and therefore does not need
+	 * locking.
+	 */
+	struct list_head head;
+
+	/**
+	 * @index: Position inside the mode_config.list, can be used as an array
+	 * index. It is invariant over the lifetime of the colorop.
+	 */
+	unsigned int index;
+
+	/** @base: base mode object */
+	struct drm_mode_object base;
+
+	/**
+	 * @plane:
+	 *
+	 * The plane on which the colorop sits. A drm_colorop is always unique
+	 * to a plane.
+	 */
+	struct drm_plane *plane;
+
+	/**
+	 * @state:
+	 *
+	 * Current atomic state for this colorop.
+	 *
+	 * This is protected by @mutex. Note that nonblocking atomic commits
+	 * access the current colorop state without taking locks.
+	 */
+	struct drm_colorop_state *state;
+
+	/*
+	 * Color properties
+	 *
+	 * The following fields are not always valid, their usage depends
+	 * on the colorop type. See their associated comment for more
+	 * information.
+	 */
+
+	/** @properties: property tracking for this colorop */
+	struct drm_object_properties properties;
+
+};
+
+#define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
+
+/**
+ * drm_colorop_find - look up a Colorop object from its ID
+ * @dev: DRM device
+ * @file_priv: drm file to check for lease against.
+ * @id: &drm_mode_object ID
+ *
+ * This can be used to look up a Colorop from its userspace ID. Only used by
+ * drivers for legacy IOCTLs and interface, nowadays extensions to the KMS
+ * userspace interface should be done using &drm_property.
+ */
+static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
+		struct drm_file *file_priv,
+		uint32_t id)
+{
+	struct drm_mode_object *mo;
+
+	mo = drm_mode_object_find(dev, file_priv, id, DRM_MODE_OBJECT_COLOROP);
+	return mo ? obj_to_colorop(mo) : NULL;
+}
+
+struct drm_colorop_state *
+drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
+
+void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
+				      struct drm_colorop_state *state);
+
+/**
+ * drm_colorop_reset - reset colorop atomic state
+ * @colorop: drm colorop
+ *
+ * Resets the atomic state for @colorop by freeing the state pointer (which might
+ * be NULL, e.g. at driver load time) and allocating a new empty state object.
+ */
+void drm_colorop_reset(struct drm_colorop *colorop);
+
+/**
+ * drm_colorop_index - find the index of a registered colorop
+ * @colorop: colorop to find index for
+ *
+ * Given a registered colorop, return the index of that colorop within a DRM
+ * device's list of colorops.
+ */
+static inline unsigned int drm_colorop_index(const struct drm_colorop *colorop)
+{
+	return colorop->index;
+}
+
+
+#define drm_for_each_colorop(colorop, dev) \
+	list_for_each_entry(colorop, &(dev)->mode_config.colorop_list, head)
+
+
+#endif /* __DRM_COLOROP_H__ */
diff --git a/include/drm/drm_mode_config.h b/include/drm/drm_mode_config.h
index 4b8f0370b79b..0226a45541d7 100644
--- a/include/drm/drm_mode_config.h
+++ b/include/drm/drm_mode_config.h
@@ -499,6 +499,24 @@ struct drm_mode_config {
 	 */
 	struct raw_spinlock panic_lock;
 
+	/**
+	 * @num_colorop:
+	 *
+	 * Number of colorop objects on this device.
+	 * This is invariant over the lifetime of a device and hence doesn't
+	 * need any locks.
+	 */
+	int num_colorop;
+
+	/**
+	 * @colorop_list:
+	 *
+	 * List of colorop objects linked with &drm_colorop.head. This is
+	 * invariant over the lifetime of a device and hence doesn't need any
+	 * locks.
+	 */
+	struct list_head colorop_list;
+
 	/**
 	 * @num_crtc:
 	 *
diff --git a/include/drm/drm_plane.h b/include/drm/drm_plane.h
index dd718c62ac31..2a9b902623c0 100644
--- a/include/drm/drm_plane.h
+++ b/include/drm/drm_plane.h
@@ -243,6 +243,14 @@ struct drm_plane_state {
 	 */
 	enum drm_scaling_filter scaling_filter;
 
+	/**
+	 * @color_pipeline:
+	 *
+	 * The first colorop of the active color pipeline, or NULL, if no
+	 * color pipeline is active.
+	 */
+	struct drm_colorop *color_pipeline;
+
 	/**
 	 * @commit: Tracks the pending commit to prevent use-after-free conditions,
 	 * and for async plane updates.
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index c082810c08a8..d49cded6c568 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -629,6 +629,7 @@ struct drm_mode_connector_set_property {
 #define DRM_MODE_OBJECT_FB 0xfbfbfbfb
 #define DRM_MODE_OBJECT_BLOB 0xbbbbbbbb
 #define DRM_MODE_OBJECT_PLANE 0xeeeeeeee
+#define DRM_MODE_OBJECT_COLOROP 0xfafafafa
 #define DRM_MODE_OBJECT_ANY 0
 
 struct drm_mode_obj_get_properties {
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 05/43] drm/colorop: Add TYPE property
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (3 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 04/43] drm/colorop: Introduce new drm_colorop mode object Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 06/43] drm/colorop: Add 1D Curve subtype Alex Hung
                   ` (39 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Add a read-only TYPE property. The TYPE specifies the colorop
type, such as enumerated curve, 1D LUT, CTM, 3D LUT, PWL LUT,
etc.

For now we're only introducing an enumerated 1D LUT type to
illustrate the concept.

Reviewed-by: Simon Ser <contact@emersion.fr>
Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v6:
 - fix doc typo (Alex Hung)

v5:
 - Add drm_get_colorop_type_name in header
 - Add kernel docs

v4:
 - Use enum property for TYPE (Pekka)

v3:
 - Make TYPE a range property
 - Move enum drm_colorop_type to uapi header
 - Fix drm_get_colorop_type_name description

 drivers/gpu/drm/drm_atomic.c      |  4 ++--
 drivers/gpu/drm/drm_atomic_uapi.c |  8 +++++++-
 drivers/gpu/drm/drm_colorop.c     | 12 ++++++++++++
 include/drm/drm_colorop.h         | 23 +++++++++++++++++++++++
 include/uapi/drm/drm_mode.h       | 19 +++++++++++++++++++
 5 files changed, 63 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 20d6393ec72c..886dc8b28abc 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -631,8 +631,8 @@ drm_atomic_get_colorop_state(struct drm_atomic_state *state,
 	state->colorops[index].new_state = colorop_state;
 	colorop_state->state = state;
 
-	drm_dbg_atomic(colorop->dev, "Added [COLOROP:%d] %p state to %p\n",
-		       colorop->base.id, colorop_state, state);
+	drm_dbg_atomic(colorop->dev, "Added [COLOROP:%d:%d] %p state to %p\n",
+		       colorop->base.id, colorop->type, colorop_state, state);
 
 	return colorop_state;
 }
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 10e625c17f90..358a95065459 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -661,7 +661,13 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 		const struct drm_colorop_state *state,
 		struct drm_property *property, uint64_t *val)
 {
-	return -EINVAL;
+	if (property == colorop->type_property) {
+		*val = colorop->type;
+	} else {
+		return -EINVAL;
+	}
+
+	return 0;
 }
 
 static int drm_atomic_set_writeback_fb_for_connector(
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index d215e22c9d20..1459a28c7e7b 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -102,3 +102,15 @@ void drm_colorop_reset(struct drm_colorop *colorop)
 	if (colorop->state)
 		__drm_colorop_reset(colorop, colorop->state);
 }
+
+static const char * const colorop_type_name[] = {
+	[DRM_COLOROP_1D_CURVE] = "1D Curve",
+};
+
+const char *drm_get_colorop_type_name(enum drm_colorop_type type)
+{
+	if (WARN_ON(type >= ARRAY_SIZE(colorop_type_name)))
+		return "unknown";
+
+	return colorop_type_name[type];
+}
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 008665d2d960..9c9698545f63 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -112,6 +112,21 @@ struct drm_colorop {
 	/** @properties: property tracking for this colorop */
 	struct drm_object_properties properties;
 
+	/**
+	 * @type:
+	 *
+	 * Read-only
+	 * Type of color operation
+	 */
+	enum drm_colorop_type type;
+
+	/**
+	 * @type_property:
+	 *
+	 * Read-only "TYPE" property for specifying the type of
+	 * this color operation. The type is enum drm_colorop_type.
+	 */
+	struct drm_property *type_property;
 };
 
 #define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
@@ -167,5 +182,13 @@ static inline unsigned int drm_colorop_index(const struct drm_colorop *colorop)
 #define drm_for_each_colorop(colorop, dev) \
 	list_for_each_entry(colorop, &(dev)->mode_config.colorop_list, head)
 
+/**
+ * drm_get_colorop_type_name - return a string for colorop type
+ * @type: colorop type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_type_name(enum drm_colorop_type type);
 
 #endif /* __DRM_COLOROP_H__ */
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index d49cded6c568..ea6d88f683cd 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -858,6 +858,25 @@ struct drm_color_lut {
 	__u16 reserved;
 };
 
+/**
+ * enum drm_colorop_type - Type of color operation
+ *
+ * drm_colorops can be of many different types. Each type behaves differently
+ * and defines a different set of properties. This enum defines all types and
+ * gives a high-level description.
+ */
+enum drm_colorop_type {
+	/**
+	 * @DRM_COLOROP_1D_CURVE:
+	 *
+	 * enum string "1D Curve"
+	 *
+	 * A 1D curve that is being applied to all color channels. The
+	 * curve is specified via the CURVE_1D_TYPE colorop property.
+	 */
+	DRM_COLOROP_1D_CURVE
+};
+
 /**
  * struct drm_plane_size_hint - Plane size hints
  * @width: The width of the plane in pixel
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 06/43] drm/colorop: Add 1D Curve subtype
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (4 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 05/43] drm/colorop: Add TYPE property Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 07/43] drm/colorop: Add BYPASS property Alex Hung
                   ` (38 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Add a new drm_colorop with DRM_COLOROP_1D_CURVE with two subtypes:
DRM_COLOROP_1D_CURVE_SRGB_EOTF and DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF.

Reviewed-by: Simon Ser <contact@emersion.fr>
Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Co-developed-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9: Specify function names by _plane_ (Chaitanya Kumar Borah)

v5:
 - Add drm_get_colorop_curve_1d_type_name in header
 - Add drm_colorop_init
 - Set default curve
 - Add kernel docs

v4:
 - Use drm_colorop_curve_1d_type_enum_list to get name (Pekka)
 - Create separate init function for 1D curve
 - Pass supported TFs into 1D curve init function

 drivers/gpu/drm/drm_atomic_uapi.c |  18 ++--
 drivers/gpu/drm/drm_colorop.c     | 134 ++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         |  63 ++++++++++++++
 3 files changed, 210 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 358a95065459..995f5ce87efc 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -649,11 +649,17 @@ static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
 		struct drm_colorop_state *state, struct drm_file *file_priv,
 		struct drm_property *property, uint64_t val)
 {
-	drm_dbg_atomic(colorop->dev,
-			"[COLOROP:%d] unknown property [PROP:%d:%s]]\n",
-			colorop->base.id,
-			property->base.id, property->name);
-	return -EINVAL;
+	if (property == colorop->curve_1d_type_property) {
+		state->curve_1d_type = val;
+	} else {
+		drm_dbg_atomic(colorop->dev,
+			       "[COLOROP:%d:%d] unknown property [PROP:%d:%s]]\n",
+			       colorop->base.id, colorop->type,
+			       property->base.id, property->name);
+		return -EINVAL;
+	}
+
+	return 0;
 }
 
 static int
@@ -663,6 +669,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 {
 	if (property == colorop->type_property) {
 		*val = colorop->type;
+	} else if (property == colorop->curve_1d_type_property) {
+		*val = state->curve_1d_type;
 	} else {
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 1459a28c7e7b..6fbc3c284d33 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -31,6 +31,123 @@
 
 #include "drm_crtc_internal.h"
 
+static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
+	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
+};
+
+static const char * const colorop_curve_1d_type_names[] = {
+	[DRM_COLOROP_1D_CURVE_SRGB_EOTF] = "sRGB EOTF",
+	[DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF] = "sRGB Inverse EOTF",
+};
+
+
+/* Init Helpers */
+
+static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
+			    struct drm_plane *plane, enum drm_colorop_type type)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_property *prop;
+	int ret = 0;
+
+	ret = drm_mode_object_add(dev, &colorop->base, DRM_MODE_OBJECT_COLOROP);
+	if (ret)
+		return ret;
+
+	colorop->base.properties = &colorop->properties;
+	colorop->dev = dev;
+	colorop->type = type;
+	colorop->plane = plane;
+
+	list_add_tail(&colorop->head, &config->colorop_list);
+	colorop->index = config->num_colorop++;
+
+	/* add properties */
+
+	/* type */
+	prop = drm_property_create_enum(dev,
+					DRM_MODE_PROP_IMMUTABLE,
+					"TYPE", drm_colorop_type_enum_list,
+					ARRAY_SIZE(drm_colorop_type_enum_list));
+
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->type_property = prop;
+
+	drm_object_attach_property(&colorop->base,
+				   colorop->type_property,
+				   colorop->type);
+
+	return ret;
+}
+
+/**
+ * drm_plane_colorop_curve_1d_init - Initialize a DRM_COLOROP_1D_CURVE
+ *
+ * @dev: DRM device
+ * @colorop: The drm_colorop object to initialize
+ * @plane: The associated drm_plane
+ * @supported_tfs: A bitfield of supported drm_plane_colorop_curve_1d_init enum values,
+ *                 created using BIT(curve_type) and combined with the OR '|'
+ *                 operator.
+ * @return zero on success, -E value on failure
+ */
+int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
+				    struct drm_plane *plane, u64 supported_tfs)
+{
+	struct drm_prop_enum_list enum_list[DRM_COLOROP_1D_CURVE_COUNT];
+	int i, len;
+
+	struct drm_property *prop;
+	int ret;
+
+	if (!supported_tfs) {
+		drm_err(dev,
+			"No supported TFs for new 1D curve colorop on [PLANE:%d:%s]\n",
+			plane->base.id, plane->name);
+		return -EINVAL;
+	}
+
+	if ((supported_tfs & -BIT(DRM_COLOROP_1D_CURVE_COUNT)) != 0) {
+		drm_err(dev, "Unknown TF provided on [PLANE:%d:%s]\n",
+			plane->base.id, plane->name);
+		return -EINVAL;
+	}
+
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_CURVE);
+	if (ret)
+		return ret;
+
+	len = 0;
+	for (i = 0; i < DRM_COLOROP_1D_CURVE_COUNT; i++) {
+		if ((supported_tfs & BIT(i)) == 0)
+			continue;
+
+		enum_list[len].type = i;
+		enum_list[len].name = colorop_curve_1d_type_names[i];
+		len++;
+	}
+
+	if (WARN_ON(len <= 0))
+		return -EINVAL;
+
+
+	/* initialize 1D curve only attribute */
+	prop = drm_property_create_enum(dev, DRM_MODE_PROP_ATOMIC, "CURVE_1D_TYPE",
+					enum_list, len);
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->curve_1d_type_property = prop;
+	drm_object_attach_property(&colorop->base, colorop->curve_1d_type_property,
+				   enum_list[0].type);
+	drm_colorop_reset(colorop);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_colorop_curve_1d_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
@@ -70,7 +187,16 @@ void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
 static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
 				      struct drm_colorop *colorop)
 {
+	u64 val;
+
 	colorop_state->colorop = colorop;
+
+	if (colorop->curve_1d_type_property) {
+		drm_object_property_get_default_value(&colorop->base,
+						colorop->curve_1d_type_property,
+						&val);
+		colorop_state->curve_1d_type = val;
+	}
 }
 
 /**
@@ -114,3 +240,11 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type)
 
 	return colorop_type_name[type];
 }
+
+const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type)
+{
+	if (WARN_ON(type >= ARRAY_SIZE(colorop_curve_1d_type_names)))
+		return "unknown";
+
+	return colorop_curve_1d_type_names[type];
+}
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 9c9698545f63..fa167e642e0d 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -31,6 +31,42 @@
 #include <drm/drm_mode.h>
 #include <drm/drm_property.h>
 
+
+/**
+ * enum drm_colorop_curve_1d_type - type of 1D curve
+ *
+ * Describes a 1D curve to be applied by the DRM_COLOROP_1D_CURVE colorop.
+ */
+enum drm_colorop_curve_1d_type {
+	/**
+	 * @DRM_COLOROP_1D_CURVE_SRGB_EOTF:
+	 *
+	 * enum string "sRGB EOTF"
+	 *
+	 * sRGB piece-wise electro-optical transfer function. Transfer
+	 * characteristics as defined by IEC 61966-2-1 sRGB. Equivalent
+	 * to H.273 TransferCharacteristics code point 13 with
+	 * MatrixCoefficients set to 0.
+	 */
+	DRM_COLOROP_1D_CURVE_SRGB_EOTF,
+
+	/**
+	 * @DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
+	 *
+	 * enum string "sRGB Inverse EOTF"
+	 *
+	 * The inverse of &DRM_COLOROP_1D_CURVE_SRGB_EOTF
+	 */
+	DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+
+	/**
+	 * @DRM_COLOROP_1D_CURVE_COUNT:
+	 *
+	 * enum value denoting the size of the enum
+	 */
+	DRM_COLOROP_1D_CURVE_COUNT
+};
+
 /**
  * struct drm_colorop_state - mutable colorop state
  */
@@ -46,6 +82,13 @@ struct drm_colorop_state {
 	 * information.
 	 */
 
+	/**
+	 * @curve_1d_type:
+	 *
+	 * Type of 1D curve.
+	 */
+	enum drm_colorop_curve_1d_type curve_1d_type;
+
 	/** @state: backpointer to global drm_atomic_state */
 	struct drm_atomic_state *state;
 };
@@ -127,6 +170,14 @@ struct drm_colorop {
 	 * this color operation. The type is enum drm_colorop_type.
 	 */
 	struct drm_property *type_property;
+
+	/**
+	 * @curve_1d_type_property:
+	 *
+	 * Sub-type for DRM_COLOROP_1D_CURVE type.
+	 */
+	struct drm_property *curve_1d_type_property;
+
 };
 
 #define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
@@ -151,6 +202,9 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
 	return mo ? obj_to_colorop(mo) : NULL;
 }
 
+int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
+				    struct drm_plane *plane, u64 supported_tfs);
+
 struct drm_colorop_state *
 drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
 
@@ -191,4 +245,13 @@ static inline unsigned int drm_colorop_index(const struct drm_colorop *colorop)
  */
 const char *drm_get_colorop_type_name(enum drm_colorop_type type);
 
+/**
+ * drm_get_colorop_curve_1d_type_name - return a string for 1D curve type
+ * @type: 1d curve type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
+
 #endif /* __DRM_COLOROP_H__ */
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 07/43] drm/colorop: Add BYPASS property
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (5 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 06/43] drm/colorop: Add 1D Curve subtype Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 08/43] drm/colorop: Add NEXT property Alex Hung
                   ` (37 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

We want to be able to bypass each colorop at all times.
Introduce a new BYPASS boolean property for this.

Reviewed-by: Simon Ser <contact@emersion.fr>
Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v6:
 - clarify that bypass is only valid if BYPASS prop exists (Louis Chauvet)

 drivers/gpu/drm/drm_atomic_uapi.c |  6 +++++-
 drivers/gpu/drm/drm_colorop.c     | 15 +++++++++++++++
 include/drm/drm_colorop.h         | 22 ++++++++++++++++++++++
 3 files changed, 42 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 995f5ce87efc..e5da00fdbca0 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -649,7 +649,9 @@ static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
 		struct drm_colorop_state *state, struct drm_file *file_priv,
 		struct drm_property *property, uint64_t val)
 {
-	if (property == colorop->curve_1d_type_property) {
+	if (property == colorop->bypass_property) {
+		state->bypass = val;
+	} else if (property == colorop->curve_1d_type_property) {
 		state->curve_1d_type = val;
 	} else {
 		drm_dbg_atomic(colorop->dev,
@@ -669,6 +671,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 {
 	if (property == colorop->type_property) {
 		*val = colorop->type;
+	} else if (property == colorop->bypass_property) {
+		*val = state->bypass;
 	} else if (property == colorop->curve_1d_type_property) {
 		*val = state->curve_1d_type;
 	} else {
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 6fbc3c284d33..11c37916c758 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -79,6 +79,17 @@ static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *co
 				   colorop->type_property,
 				   colorop->type);
 
+	/* bypass */
+	prop = drm_property_create_bool(dev, DRM_MODE_PROP_ATOMIC,
+					"BYPASS");
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->bypass_property = prop;
+	drm_object_attach_property(&colorop->base,
+				   colorop->bypass_property,
+				   1);
+
 	return ret;
 }
 
@@ -136,6 +147,7 @@ int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *
 	/* initialize 1D curve only attribute */
 	prop = drm_property_create_enum(dev, DRM_MODE_PROP_ATOMIC, "CURVE_1D_TYPE",
 					enum_list, len);
+
 	if (!prop)
 		return -ENOMEM;
 
@@ -152,6 +164,8 @@ static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colo
 							struct drm_colorop_state *state)
 {
 	memcpy(state, colorop->state, sizeof(*state));
+
+	state->bypass = true;
 }
 
 struct drm_colorop_state *
@@ -190,6 +204,7 @@ static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
 	u64 val;
 
 	colorop_state->colorop = colorop;
+	colorop_state->bypass = true;
 
 	if (colorop->curve_1d_type_property) {
 		drm_object_property_get_default_value(&colorop->base,
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index fa167e642e0d..49e6564c17ba 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -82,6 +82,16 @@ struct drm_colorop_state {
 	 * information.
 	 */
 
+
+	/**
+	 * @bypass:
+	 *
+	 * When the property BYPASS exists on this colorop, this stores
+	 * the requested bypass state: true if colorop shall be bypassed,
+	 * false if colorop is enabled.
+	 */
+	bool bypass;
+
 	/**
 	 * @curve_1d_type:
 	 *
@@ -171,6 +181,18 @@ struct drm_colorop {
 	 */
 	struct drm_property *type_property;
 
+	/**
+	 * @bypass_property:
+	 *
+	 * Boolean property to control enablement of the color
+	 * operation. Setting bypass to "true" shall always be supported
+	 * in order to allow compositors to quickly fall back to
+	 * alternate methods of color processing. This is important
+	 * since setting color operations can fail due to unique
+	 * HW constraints.
+	 */
+	struct drm_property *bypass_property;
+
 	/**
 	 * @curve_1d_type_property:
 	 *
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 08/43] drm/colorop: Add NEXT property
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (6 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 07/43] drm/colorop: Add BYPASS property Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 09/43] drm/colorop: Add atomic state print for drm_colorop Alex Hung
                   ` (36 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

We'll construct color pipelines out of drm_colorop by
chaining them via the NEXT pointer. NEXT will point to
the next drm_colorop in the pipeline, or by 0 if we're
at the end of the pipeline.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v8:
 - Remove null check "colorop->next_property" in drm_colorop_set_next_property

v5:
 - move next comment here from Add 3x4 CTM patch (Sebastian)
 - Add kernel doc

v4:
 - Allow setting of NEXT property to NULL (Chaitanya Kumar Borah)

v3:
 - Add next pointer to colorop to be used by drivers
   and in DRM core

 drivers/gpu/drm/drm_colorop.c | 27 +++++++++++++++++++++++++++
 include/drm/drm_colorop.h     | 17 +++++++++++++++++
 2 files changed, 44 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 11c37916c758..43cdd68a69a2 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -58,6 +58,7 @@ static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *co
 	colorop->dev = dev;
 	colorop->type = type;
 	colorop->plane = plane;
+	colorop->next = NULL;
 
 	list_add_tail(&colorop->head, &config->colorop_list);
 	colorop->index = config->num_colorop++;
@@ -90,6 +91,16 @@ static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *co
 				   colorop->bypass_property,
 				   1);
 
+	/* next */
+	prop = drm_property_create_object(dev, DRM_MODE_PROP_IMMUTABLE | DRM_MODE_PROP_ATOMIC,
+					  "NEXT", DRM_MODE_OBJECT_COLOROP);
+	if (!prop)
+		return -ENOMEM;
+	colorop->next_property = prop;
+	drm_object_attach_property(&colorop->base,
+				   colorop->next_property,
+				   0);
+
 	return ret;
 }
 
@@ -263,3 +274,19 @@ const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type ty
 
 	return colorop_curve_1d_type_names[type];
 }
+
+/**
+ * drm_colorop_set_next_property - sets the next pointer
+ * @colorop: drm colorop
+ * @next: next colorop
+ *
+ * Should be used when constructing the color pipeline
+ */
+void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next)
+{
+	drm_object_property_set_value(&colorop->base,
+				      colorop->next_property,
+				      next ? next->base.id : 0);
+	colorop->next = next;
+}
+EXPORT_SYMBOL(drm_colorop_set_next_property);
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 49e6564c17ba..791ab940c158 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -173,6 +173,14 @@ struct drm_colorop {
 	 */
 	enum drm_colorop_type type;
 
+	/**
+	 * @next:
+	 *
+	 * Read-only
+	 * Pointer to next drm_colorop in pipeline
+	 */
+	struct drm_colorop *next;
+
 	/**
 	 * @type_property:
 	 *
@@ -200,6 +208,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @next_property:
+	 *
+	 * Read-only property to next colorop in the pipeline
+	 */
+	struct drm_property *next_property;
+
 };
 
 #define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
@@ -276,4 +291,6 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type);
  */
 const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
 
+void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
+
 #endif /* __DRM_COLOROP_H__ */
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 09/43] drm/colorop: Add atomic state print for drm_colorop
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (7 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 08/43] drm/colorop: Add NEXT property Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 10/43] drm/plane: Add COLOR PIPELINE property Alex Hung
                   ` (35 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Print atomic state for drm_colorop in debugfs

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v8:
 - Add switch statement to print colorop type (Simon Ser)

v7:
 - Add a commit messages
 - Squash "drm/colorop: Add NEXT to colorop state print" (Simon Ser)

 drivers/gpu/drm/drm_atomic.c | 35 ++++++++++++++++++++++++++++++++++-
 1 file changed, 34 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 886dc8b28abc..dc679c8b904e 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -777,6 +777,29 @@ static int drm_atomic_plane_check(const struct drm_plane_state *old_plane_state,
 	return 0;
 }
 
+
+
+static void drm_atomic_colorop_print_state(struct drm_printer *p,
+		const struct drm_colorop_state *state)
+{
+	struct drm_colorop *colorop = state->colorop;
+
+	drm_printf(p, "colorop[%u]:\n", colorop->base.id);
+	drm_printf(p, "\ttype=%s\n", drm_get_colorop_type_name(colorop->type));
+	drm_printf(p, "\tbypass=%u\n", state->bypass);
+
+	switch (colorop->type) {
+	case DRM_COLOROP_1D_CURVE:
+		drm_printf(p, "\tcurve_1d_type=%s\n",
+			   drm_get_colorop_curve_1d_type_name(state->curve_1d_type));
+		break;
+	default:
+		break;
+	}
+
+	drm_printf(p, "\tnext=%d\n", colorop->next ? colorop->next->base.id : 0);
+}
+
 static void drm_atomic_plane_print_state(struct drm_printer *p,
 		const struct drm_plane_state *state)
 {
@@ -798,7 +821,8 @@ static void drm_atomic_plane_print_state(struct drm_printer *p,
 	drm_printf(p, "\tcolor-range=%s\n",
 		   drm_get_color_range_name(state->color_range));
 	drm_printf(p, "\tcolor_mgmt_changed=%d\n", state->color_mgmt_changed);
-
+	drm_printf(p, "\tcolor-pipeline=%d\n",
+		   state->color_pipeline ? state->color_pipeline->base.id : 0);
 	if (plane->funcs->atomic_print_state)
 		plane->funcs->atomic_print_state(p, state);
 }
@@ -1903,6 +1927,7 @@ static void __drm_state_dump(struct drm_device *dev, struct drm_printer *p,
 			     bool take_locks)
 {
 	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_colorop *colorop;
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
 	struct drm_connector *connector;
@@ -1912,6 +1937,14 @@ static void __drm_state_dump(struct drm_device *dev, struct drm_printer *p,
 	if (!drm_drv_uses_atomic_modeset(dev))
 		return;
 
+	list_for_each_entry(colorop, &config->colorop_list, head) {
+		if (take_locks)
+			drm_modeset_lock(&colorop->plane->mutex, NULL);
+		drm_atomic_colorop_print_state(p, colorop->state);
+		if (take_locks)
+			drm_modeset_unlock(&colorop->plane->mutex);
+	}
+
 	list_for_each_entry(plane, &config->plane_list, head) {
 		if (take_locks)
 			drm_modeset_lock(&plane->mutex, NULL);
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 10/43] drm/plane: Add COLOR PIPELINE property
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (8 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 09/43] drm/colorop: Add atomic state print for drm_colorop Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 11/43] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Alex Hung
                   ` (34 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

We're adding a new enum COLOR PIPELINE property. This
property will have entries for each COLOR PIPELINE by
referencing the DRM object ID of the first drm_colorop
of the pipeline. 0 disables the entire COLOR PIPELINE.

Userspace can use this to discover the available color
pipelines, as well as set the desired one. The color
pipelines are programmed via properties on the actual
drm_colorop objects.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Remove redundant comments (Simon Ser)

v8:
 - Use upper case in drm_dbg_atomic (Simon Ser)
 - Remove MAX_COLOR_PIPELINES (Simon Ser)
 - Make pipelines const in drm_plane_create_color_pipeline_property (Simon Ser)

v7:
 - Fix a checkpatch warning - a new line after variable declaration

v4:
 - Add pipeline property creation helper (Pekka)
 - Fix function comment for
   drm_atomic_set_colorop_for_plane (Pekka)
 - Always create Bypass pipeline (Pekka)
 - Add missing function declaration (Chaitanya Kumar Borah)

 drivers/gpu/drm/drm_atomic.c              | 46 ++++++++++++++++++
 drivers/gpu/drm/drm_atomic_state_helper.c |  5 ++
 drivers/gpu/drm/drm_atomic_uapi.c         | 42 ++++++++++++++++
 drivers/gpu/drm/drm_plane.c               | 59 +++++++++++++++++++++++
 include/drm/drm_atomic.h                  |  3 ++
 include/drm/drm_atomic_uapi.h             |  2 +
 include/drm/drm_plane.h                   | 11 +++++
 7 files changed, 168 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index dc679c8b904e..7b042c38d50d 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -1531,6 +1531,52 @@ drm_atomic_add_affected_planes(struct drm_atomic_state *state,
 }
 EXPORT_SYMBOL(drm_atomic_add_affected_planes);
 
+/**
+ * drm_atomic_add_affected_colorops - add colorops for plane
+ * @state: atomic state
+ * @plane: DRM plane
+ *
+ * This function walks the current configuration and adds all colorops
+ * currently used by @plane to the atomic configuration @state. This is useful
+ * when an atomic commit also needs to check all currently enabled colorop on
+ * @plane, e.g. when changing the mode. It's also useful when re-enabling a plane
+ * to avoid special code to force-enable all colorops.
+ *
+ * Since acquiring a colorop state will always also acquire the w/w mutex of the
+ * current plane for that colorop (if there is any) adding all the colorop states for
+ * a plane will not reduce parallelism of atomic updates.
+ *
+ * Returns:
+ * 0 on success or can fail with -EDEADLK or -ENOMEM. When the error is EDEADLK
+ * then the w/w mutex code has detected a deadlock and the entire atomic
+ * sequence must be restarted. All other errors are fatal.
+ */
+int
+drm_atomic_add_affected_colorops(struct drm_atomic_state *state,
+				 struct drm_plane *plane)
+{
+	struct drm_colorop *colorop;
+	struct drm_colorop_state *colorop_state;
+
+	WARN_ON(!drm_atomic_get_new_plane_state(state, plane));
+
+	drm_dbg_atomic(plane->dev,
+		       "Adding all current colorops for [PLANE:%d:%s] to %p\n",
+		       plane->base.id, plane->name, state);
+
+	drm_for_each_colorop(colorop, plane->dev) {
+		if (colorop->plane != plane)
+			continue;
+
+		colorop_state = drm_atomic_get_colorop_state(state, colorop);
+		if (IS_ERR(colorop_state))
+			return PTR_ERR(colorop_state);
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_atomic_add_affected_colorops);
+
 /**
  * drm_atomic_check_only - check whether a given config would work
  * @state: atomic configuration to check
diff --git a/drivers/gpu/drm/drm_atomic_state_helper.c b/drivers/gpu/drm/drm_atomic_state_helper.c
index 519228eb1095..d1dd082b1286 100644
--- a/drivers/gpu/drm/drm_atomic_state_helper.c
+++ b/drivers/gpu/drm/drm_atomic_state_helper.c
@@ -267,6 +267,11 @@ void __drm_atomic_helper_plane_state_reset(struct drm_plane_state *plane_state,
 			plane_state->color_range = val;
 	}
 
+	if (plane->color_pipeline_property) {
+		/* default is always NULL, i.e., bypass */
+		plane_state->color_pipeline = NULL;
+	}
+
 	if (plane->zpos_property) {
 		if (!drm_object_property_get_default_value(&plane->base,
 							   plane->zpos_property,
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index e5da00fdbca0..0b2b3242c976 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -257,6 +257,36 @@ drm_atomic_set_fb_for_plane(struct drm_plane_state *plane_state,
 }
 EXPORT_SYMBOL(drm_atomic_set_fb_for_plane);
 
+
+/**
+ * drm_atomic_set_colorop_for_plane - set colorop for plane
+ * @plane_state: atomic state object for the plane
+ * @colorop: colorop to use for the plane
+ *
+ * Helper function to select the color pipeline on a plane by setting
+ * it to the first drm_colorop element of the pipeline.
+ */
+void
+drm_atomic_set_colorop_for_plane(struct drm_plane_state *plane_state,
+				 struct drm_colorop *colorop)
+{
+	struct drm_plane *plane = plane_state->plane;
+
+	if (colorop)
+		drm_dbg_atomic(plane->dev,
+			       "Set [COLOROP:%d] for [PLANE:%d:%s] state %p\n",
+			       colorop->base.id, plane->base.id, plane->name,
+			       plane_state);
+	else
+		drm_dbg_atomic(plane->dev,
+			       "Set [NOCOLOROP] for [PLANE:%d:%s] state %p\n",
+			       plane->base.id, plane->name, plane_state);
+
+	plane_state->color_pipeline = colorop;
+}
+EXPORT_SYMBOL(drm_atomic_set_colorop_for_plane);
+
+
 /**
  * drm_atomic_set_crtc_for_connector - set CRTC for connector
  * @conn_state: atomic state object for the connector
@@ -540,6 +570,16 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 		state->color_encoding = val;
 	} else if (property == plane->color_range_property) {
 		state->color_range = val;
+	} else if (property == plane->color_pipeline_property) {
+		/* find DRM colorop object */
+		struct drm_colorop *colorop = NULL;
+
+		colorop = drm_colorop_find(dev, file_priv, val);
+
+		if (val && !colorop)
+			return -EACCES;
+
+		drm_atomic_set_colorop_for_plane(state, colorop);
 	} else if (property == config->prop_fb_damage_clips) {
 		ret = drm_property_replace_blob_from_id(dev,
 					&state->fb_damage_clips,
@@ -622,6 +662,8 @@ drm_atomic_plane_get_property(struct drm_plane *plane,
 		*val = state->color_encoding;
 	} else if (property == plane->color_range_property) {
 		*val = state->color_range;
+	} else if (property == plane->color_pipeline_property) {
+		*val = (state->color_pipeline) ? state->color_pipeline->base.id : 0;
 	} else if (property == config->prop_fb_damage_clips) {
 		*val = (state->fb_damage_clips) ?
 			state->fb_damage_clips->base.id : 0;
diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
index a28b22fdd7a4..694db8fb6555 100644
--- a/drivers/gpu/drm/drm_plane.c
+++ b/drivers/gpu/drm/drm_plane.c
@@ -1793,3 +1793,62 @@ int drm_plane_add_size_hints_property(struct drm_plane *plane,
 	return 0;
 }
 EXPORT_SYMBOL(drm_plane_add_size_hints_property);
+
+/**
+ * drm_plane_create_color_pipeline_property - create a new color pipeline
+ * property
+ *
+ * @plane: drm plane
+ * @pipelines: list of pipelines
+ * @num_pipelines: number of pipelines
+ *
+ * Create the COLOR_PIPELINE plane property to specific color pipelines on
+ * the plane.
+ *
+ * RETURNS:
+ * Zero for success or -errno
+ */
+int drm_plane_create_color_pipeline_property(struct drm_plane *plane,
+					     const struct drm_prop_enum_list *pipelines,
+					     int num_pipelines)
+{
+	struct drm_prop_enum_list *all_pipelines;
+	struct drm_property *prop;
+	int len = 0;
+	int i;
+
+	all_pipelines = kcalloc(num_pipelines + 1,
+				sizeof(*all_pipelines),
+				GFP_KERNEL);
+
+	if (!all_pipelines) {
+		drm_err(plane->dev, "failed to allocate color pipeline\n");
+		return -ENOMEM;
+	}
+
+	/* Create default Bypass color pipeline */
+	all_pipelines[len].type = 0;
+	all_pipelines[len].name = "Bypass";
+	len++;
+
+	/* Add all other color pipelines */
+	for (i = 0; i < num_pipelines; i++, len++) {
+		all_pipelines[len].type = pipelines[i].type;
+		all_pipelines[len].name = pipelines[i].name;
+	}
+
+	prop = drm_property_create_enum(plane->dev, DRM_MODE_PROP_ATOMIC,
+					"COLOR_PIPELINE",
+					all_pipelines, len);
+	if (IS_ERR(prop)) {
+		kfree(all_pipelines);
+		return PTR_ERR(prop);
+	}
+
+	drm_object_attach_property(&plane->base, prop, 0);
+	plane->color_pipeline_property = prop;
+
+	kfree(all_pipelines);
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_create_color_pipeline_property);
diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index 92dde4fb21d2..7904a38caa77 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -886,6 +886,9 @@ drm_atomic_add_affected_connectors(struct drm_atomic_state *state,
 int __must_check
 drm_atomic_add_affected_planes(struct drm_atomic_state *state,
 			       struct drm_crtc *crtc);
+int __must_check
+drm_atomic_add_affected_colorops(struct drm_atomic_state *state,
+				 struct drm_plane *plane);
 
 int __must_check drm_atomic_check_only(struct drm_atomic_state *state);
 int __must_check drm_atomic_commit(struct drm_atomic_state *state);
diff --git a/include/drm/drm_atomic_uapi.h b/include/drm/drm_atomic_uapi.h
index 70a115d523cd..436315523326 100644
--- a/include/drm/drm_atomic_uapi.h
+++ b/include/drm/drm_atomic_uapi.h
@@ -50,6 +50,8 @@ drm_atomic_set_crtc_for_plane(struct drm_plane_state *plane_state,
 			      struct drm_crtc *crtc);
 void drm_atomic_set_fb_for_plane(struct drm_plane_state *plane_state,
 				 struct drm_framebuffer *fb);
+void drm_atomic_set_colorop_for_plane(struct drm_plane_state *plane_state,
+				      struct drm_colorop *colorop);
 int __must_check
 drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
 				  struct drm_crtc *crtc);
diff --git a/include/drm/drm_plane.h b/include/drm/drm_plane.h
index 2a9b902623c0..3c195909394e 100644
--- a/include/drm/drm_plane.h
+++ b/include/drm/drm_plane.h
@@ -774,6 +774,14 @@ struct drm_plane {
 	 */
 	struct drm_property *color_range_property;
 
+	/**
+	 * @color_pipeline_property:
+	 *
+	 * Optional "COLOR_PIPELINE" enum property for specifying
+	 * a color pipeline to use on the plane.
+	 */
+	struct drm_property *color_pipeline_property;
+
 	/**
 	 * @scaling_filter_property: property to apply a particular filter while
 	 * scaling.
@@ -997,4 +1005,7 @@ int drm_plane_add_size_hints_property(struct drm_plane *plane,
 				      const struct drm_plane_size_hint *hints,
 				      int num_hints);
 
+int drm_plane_create_color_pipeline_property(struct drm_plane *plane,
+					     const struct drm_prop_enum_list *pipelines,
+					     int num_pipelines);
 #endif
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 11/43] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (9 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 10/43] drm/plane: Add COLOR PIPELINE property Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 12/43] Documentation/gpu: document drm_colorop Alex Hung
                   ` (33 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

With the introduction of the pre-blending color pipeline we
can no longer have color operations that don't have a clear
position in the color pipeline. We deprecate all existing
plane properties. For upstream drivers those are:
 - COLOR_ENCODING
 - COLOR_RANGE

Drivers are expected to ignore these properties when
programming the HW. DRM clients that register with
DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE will not be allowed to
set the COLOR_ENCODING and COLOR_RANGE properties.

Setting of the COLOR_PIPELINE plane property or drm_colorop
properties is only allowed for userspace that sets this
client cap.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Fix typo in commit description (Shengyu Qu)

v8:
 - Disallow setting of COLOR_RANGE and COLOR_ENCODING when
   DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE is set

v5:
 - Fix kernel docs

v4:
 - Don't block setting of COLOR_RANGE and COLOR_ENCODING
   when client cap is set

 drivers/gpu/drm/drm_atomic_uapi.c | 23 ++++++++++++++++++++++-
 drivers/gpu/drm/drm_ioctl.c       |  7 +++++++
 include/drm/drm_file.h            |  7 +++++++
 include/uapi/drm/drm.h            | 15 +++++++++++++++
 4 files changed, 51 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 0b2b3242c976..c0bcaec249de 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -567,10 +567,26 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 	} else if (property == plane->zpos_property) {
 		state->zpos = val;
 	} else if (property == plane->color_encoding_property) {
+		if (file_priv->plane_color_pipeline) {
+			drm_dbg_atomic(dev,
+				       "Setting COLOR_ENCODING plane property not permitted with DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE client cap\n");
+			return -EINVAL;
+		}
 		state->color_encoding = val;
 	} else if (property == plane->color_range_property) {
+		if (file_priv->plane_color_pipeline) {
+			drm_dbg_atomic(dev,
+				       "Setting COLOR_RANGE plane property not permitted with DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE client cap\n");
+			return -EINVAL;
+		}
 		state->color_range = val;
 	} else if (property == plane->color_pipeline_property) {
+		if (!file_priv->plane_color_pipeline) {
+			drm_dbg_atomic(dev,
+				       "Setting COLOR_PIPELINE plane property not permitted unless DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE is set\n");
+			return -EINVAL;
+		}
+
 		/* find DRM colorop object */
 		struct drm_colorop *colorop = NULL;
 
@@ -1200,6 +1216,12 @@ int drm_atomic_set_property(struct drm_atomic_state *state,
 		break;
 	}
 	case DRM_MODE_OBJECT_COLOROP: {
+		if (!file_priv->plane_color_pipeline) {
+			drm_dbg_atomic(prop->dev,
+				       "[OBJECT:%d] is a colorop but DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE not set\n",
+				       obj->id);
+			ret = -EINVAL;
+		}
 		struct drm_colorop *colorop = obj_to_colorop(obj);
 		struct drm_colorop_state *colorop_state;
 
@@ -1212,7 +1234,6 @@ int drm_atomic_set_property(struct drm_atomic_state *state,
 		ret = drm_atomic_colorop_set_property(colorop,
 				colorop_state, file_priv,
 				prop, prop_value);
-
 		break;
 	}
 	default:
diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c
index f593dc569d31..5c89c586da7c 100644
--- a/drivers/gpu/drm/drm_ioctl.c
+++ b/drivers/gpu/drm/drm_ioctl.c
@@ -373,6 +373,13 @@ drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 			return -EINVAL;
 		file_priv->supports_virtualized_cursor_plane = req->value;
 		break;
+	case DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE:
+		if (!file_priv->atomic)
+			return -EINVAL;
+		if (req->value > 1)
+			return -EINVAL;
+		file_priv->plane_color_pipeline = req->value;
+		break;
 	default:
 		return -EINVAL;
 	}
diff --git a/include/drm/drm_file.h b/include/drm/drm_file.h
index 94d365b22505..86929ca667aa 100644
--- a/include/drm/drm_file.h
+++ b/include/drm/drm_file.h
@@ -206,6 +206,13 @@ struct drm_file {
 	 */
 	bool writeback_connectors;
 
+	/**
+	 * @plane_color_pipeline:
+	 *
+	 * True if client understands plane color pipelines
+	 */
+	bool plane_color_pipeline;
+
 	/**
 	 * @was_master:
 	 *
diff --git a/include/uapi/drm/drm.h b/include/uapi/drm/drm.h
index e63a71d3c607..80c35ff7f21e 100644
--- a/include/uapi/drm/drm.h
+++ b/include/uapi/drm/drm.h
@@ -875,6 +875,21 @@ struct drm_get_cap {
  */
 #define DRM_CLIENT_CAP_CURSOR_PLANE_HOTSPOT	6
 
+/**
+ * DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
+ *
+ * If set to 1 the DRM core will allow setting the COLOR_PIPELINE
+ * property on a &drm_plane, as well as drm_colorop properties.
+ *
+ * Setting of these plane properties will be rejected when this client
+ * cap is set:
+ * - COLOR_ENCODING
+ * - COLOR_RANGE
+ *
+ * The client must enable &DRM_CLIENT_CAP_ATOMIC first.
+ */
+#define DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE	7
+
 /* DRM_IOCTL_SET_CLIENT_CAP ioctl argument type */
 struct drm_set_client_cap {
 	__u64 capability;
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 12/43] Documentation/gpu: document drm_colorop
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (10 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 11/43] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 13/43] drm/colorop: Add destroy functions for color pipeline Alex Hung
                   ` (32 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Add kernel doc for drm_colorop objects.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v8:
 - Move this after "drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE" (Simon Ser)
 - Update DOC overview (Simon Ser)

v7:
 - Add a commit messages

v5:
 - Drop TODO

 Documentation/gpu/drm-kms.rst | 15 +++++++++++++++
 drivers/gpu/drm/drm_colorop.c | 31 +++++++++++++++++++++++++++++++
 2 files changed, 46 insertions(+)

diff --git a/Documentation/gpu/drm-kms.rst b/Documentation/gpu/drm-kms.rst
index abfe220764e1..2292e65f044c 100644
--- a/Documentation/gpu/drm-kms.rst
+++ b/Documentation/gpu/drm-kms.rst
@@ -413,6 +413,21 @@ Plane Panic Functions Reference
 .. kernel-doc:: drivers/gpu/drm/drm_panic.c
    :export:
 
+Colorop Abstraction
+===================
+
+.. kernel-doc:: drivers/gpu/drm/drm_colorop.c
+   :doc: overview
+
+Colorop Functions Reference
+---------------------------
+
+.. kernel-doc:: include/drm/drm_colorop.h
+   :internal:
+
+.. kernel-doc:: drivers/gpu/drm/drm_colorop.c
+   :export:
+
 Display Modes Function Reference
 ================================
 
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 43cdd68a69a2..7b3ecf7ddd11 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -31,6 +31,37 @@
 
 #include "drm_crtc_internal.h"
 
+/**
+ * DOC: overview
+ *
+ * When userspace signals the &DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE it
+ * should use the COLOR_PIPELINE plane property and associated colorops
+ * for any color operation on the &drm_plane. Setting of all old color
+ * properties, such as COLOR_ENCODING and COLOR_RANGE, will be rejected
+ * and the values of the properties will be ignored.
+ *
+ * Colorops are only advertised and valid for atomic drivers and atomic
+ * userspace that signals the &DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
+ * client cap.
+ *
+ * A colorop represents a single color operation. Colorops are chained
+ * via the NEXT property and make up color pipelines. Color pipelines
+ * are advertised and selected via the COLOR_PIPELINE &drm_plane
+ * property.
+ *
+ * A colorop will be of a certain type, advertised by the read-only TYPE
+ * property. Each type of colorop will advertise a different set of
+ * properties and is programmed in a different manner. Types can be
+ * enumerated 1D curves, 1D LUTs, 3D LUTs, matrices, etc. See the
+ * &drm_colorop_type documentation for information on each type.
+ *
+ * If a colorop advertises the BYPASS property it can be bypassed.
+ *
+ * Information about colorop and color pipeline design decisions can be
+ * found at rfc/color_pipeline.rst, but note that this document will
+ * grow stale over time.
+ */
+
 static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
 };
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 13/43] drm/colorop: Add destroy functions for color pipeline
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (11 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 12/43] Documentation/gpu: document drm_colorop Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-05-15  9:34   ` Simon Ser
  2025-04-30  1:10 ` [PATCH V9 14/43] drm/vkms: Add enumerated 1D curve colorop Alex Hung
                   ` (31 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

The functions are to clean up color pipeline when a device driver
fails to create its color pipeline.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v9:
 - Move from from latest commit to here, and drm_colorop_pipeline_destroy
   is called in respective commits.


 drivers/gpu/drm/drm_colorop.c | 36 +++++++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h     |  2 ++
 2 files changed, 38 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 7b3ecf7ddd11..57a8c1063fdd 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -135,6 +135,42 @@ static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *co
 	return ret;
 }
 
+/**
+ * drm_colorop_cleanup - Cleanup a drm_colorop object in color_pipeline
+ *
+ * @colorop: The drm_colorop object to be cleaned
+ */
+static void drm_colorop_cleanup(struct drm_colorop *colorop)
+{
+	struct drm_device *dev = colorop->dev;
+	struct drm_mode_config *config = &dev->mode_config;
+
+	list_del(&colorop->head);
+	config->num_colorop--;
+
+	kfree(colorop->state);
+}
+
+/**
+ * drm_colorop_pipeline_destroy - Helper for color pipeline destruction
+ *
+ * @plane: - The drm_plane structure containing the color_pipeline
+ *
+ * Provides a default color pipeline destroy handler for a planes.
+ */
+void drm_colorop_pipeline_destroy(struct drm_plane *plane)
+{
+	struct drm_device *dev = plane->dev;
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_colorop *colorop, *next;
+
+	list_for_each_entry_safe(colorop, next, &config->colorop_list, head) {
+		drm_colorop_cleanup(colorop);
+		kfree(colorop);
+	}
+}
+EXPORT_SYMBOL(drm_colorop_pipeline_destroy);
+
 /**
  * drm_plane_colorop_curve_1d_init - Initialize a DRM_COLOROP_1D_CURVE
  *
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 791ab940c158..8e4d79ba1eff 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -239,6 +239,8 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
 	return mo ? obj_to_colorop(mo) : NULL;
 }
 
+void drm_colorop_pipeline_destroy(struct drm_plane *plane);
+
 int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
 				    struct drm_plane *plane, u64 supported_tfs);
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 14/43] drm/vkms: Add enumerated 1D curve colorop
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (12 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 13/43] drm/colorop: Add destroy functions for color pipeline Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs Alex Hung
                   ` (30 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

This patch introduces a VKMS color pipeline that includes two
drm_colorops for named transfer functions. For now the only ones
supported are sRGB EOTF, sRGB Inverse EOTF, and a Linear TF.
We will expand this in the future but I don't want to do so
without accompanying IGT tests.

We introduce a new vkms_luts.c file that hard-codes sRGB EOTF,
sRGB Inverse EOTF, and a linear EOTF LUT. These have been
generated with 256 entries each as IGT is currently testing
only 8 bpc surfaces. We will likely need higher precision
but I'm reluctant to make that change without clear indication
that we need it. We'll revisit and, if necessary, regenerate
the LUTs when we have IGT tests for higher precision buffers.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v8:
 - Replace DRM_ERROR by drm_err (Louis Chauvet)
 - Replace DRM_WARN_ONCE by drm_WARN_ONCE (Louis Chauvet)
 - Fix conflicts with upstream VKMS (Louis Chauvet)
 - Add comments for drm_color_lut linear_array (Louis Chauvet)

v7:
 - Fix checkpatch warnings (Louis Chauvet)
  - Change kzalloc(sizeof(struct drm_colorop) ...) to kzalloc(sizeof(*ops[i]) ...)
  - Remove if (ops[i]) before kfree(ops[i])
  - Fix styles by adding and removing spaces (new lines, tabs and so on)

v6:
 - drop 'len' var (Louis Chauvet)
 - cleanup if colorop alloc or init fails (Louis Chauvet)
 - switch loop in pre_blend_transform (Louis Chauvet)
 - drop extraneous if (colorop) inside while (colorop) (Louis Chauvet)

v5:
 - Squash with "Pull apply_colorop out of pre_blend_color_transform"
   (Sebastian)
 - Fix warnings
 - Fix include
 - Drop TODOs

v4:
 - Drop _tf_ from color_pipeline init function
 - Pass supported TFs into colorop init
 - Create bypass pipeline in DRM helper (Pekka)

v2:
 - Add commit description
 - Fix sRGB EOTF LUT definition
 - Add linear and sRGB inverse EOTF LUTs

 drivers/gpu/drm/vkms/Makefile        |   4 +-
 drivers/gpu/drm/vkms/vkms_colorop.c  |  81 +++
 drivers/gpu/drm/vkms/vkms_composer.c |  51 +-
 drivers/gpu/drm/vkms/vkms_drv.h      |   3 +
 drivers/gpu/drm/vkms/vkms_luts.c     | 808 +++++++++++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_luts.h     |  12 +
 drivers/gpu/drm/vkms/vkms_plane.c    |   2 +
 7 files changed, 959 insertions(+), 2 deletions(-)
 create mode 100644 drivers/gpu/drm/vkms/vkms_colorop.c
 create mode 100644 drivers/gpu/drm/vkms/vkms_luts.c
 create mode 100644 drivers/gpu/drm/vkms/vkms_luts.h

diff --git a/drivers/gpu/drm/vkms/Makefile b/drivers/gpu/drm/vkms/Makefile
index d657865e573f..0b8936674f69 100644
--- a/drivers/gpu/drm/vkms/Makefile
+++ b/drivers/gpu/drm/vkms/Makefile
@@ -8,7 +8,9 @@ vkms-y := \
 	vkms_composer.o \
 	vkms_writeback.o \
 	vkms_connector.o \
-	vkms_config.o
+	vkms_config.o \
+	vkms_colorop.o \
+	vkms_luts.o
 
 obj-$(CONFIG_DRM_VKMS) += vkms.o
 obj-$(CONFIG_DRM_VKMS_KUNIT_TEST) += tests/
diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
new file mode 100644
index 000000000000..2fb74b2bd001
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -0,0 +1,81 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <linux/slab.h>
+#include <drm/drm_colorop.h>
+#include <drm/drm_print.h>
+#include <drm/drm_property.h>
+#include <drm/drm_plane.h>
+
+#include "vkms_drv.h"
+
+static const u64 supported_tfs =
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF);
+
+#define MAX_COLOR_PIPELINE_OPS 2
+
+static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
+{
+	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
+	struct drm_device *dev = plane->dev;
+	int ret;
+	int i = 0;
+
+	memset(ops, 0, sizeof(ops));
+
+	/* 1st op: 1d curve */
+	ops[i] = kzalloc(sizeof(*ops[i]), GFP_KERNEL);
+	if (!ops[i]) {
+		drm_err(dev, "KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs);
+	if (ret)
+		goto cleanup;
+
+	list->type = ops[i]->base.id;
+	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", ops[i]->base.id);
+
+	i++;
+
+	/* 2nd op: 1d curve */
+	ops[i] = kzalloc(sizeof(*ops[i]), GFP_KERNEL);
+	if (!ops[i]) {
+		drm_err(dev, "KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i - 1], ops[i]);
+
+	return 0;
+
+cleanup:
+	drm_colorop_pipeline_destroy(plane);
+
+	return ret;
+}
+
+int vkms_initialize_colorops(struct drm_plane *plane)
+{
+	struct drm_prop_enum_list pipeline;
+	int ret;
+
+	/* Add color pipeline */
+	ret = vkms_initialize_color_pipeline(plane, &pipeline);
+	if (ret)
+		return ret;
+
+	/* Create COLOR_PIPELINE property and attach */
+	ret = drm_plane_create_color_pipeline_property(plane, &pipeline, 1);
+	if (ret)
+		return ret;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index b0dc95f971d8..3ab829556db2 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -14,6 +14,7 @@
 #include "vkms_drv.h"
 #include <kunit/visibility.h>
 #include "vkms_composer.h"
+#include "vkms_luts.h"
 
 static u16 pre_mul_blend_channel(u16 src, u16 dst, u16 alpha)
 {
@@ -136,6 +137,54 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
 	}
 }
 
+static void apply_colorop(struct pixel_argb_u16 *pixel, struct drm_colorop *colorop)
+{
+	struct drm_colorop_state *colorop_state = colorop->state;
+	struct drm_device *dev = colorop->dev;
+
+	if (colorop->type == DRM_COLOROP_1D_CURVE) {
+		switch (colorop_state->curve_1d_type) {
+		case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
+			pixel->r = apply_lut_to_channel_value(&srgb_inv_eotf, pixel->r, LUT_RED);
+			pixel->g = apply_lut_to_channel_value(&srgb_inv_eotf, pixel->g, LUT_GREEN);
+			pixel->b = apply_lut_to_channel_value(&srgb_inv_eotf, pixel->b, LUT_BLUE);
+			break;
+		case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
+			pixel->r = apply_lut_to_channel_value(&srgb_eotf, pixel->r, LUT_RED);
+			pixel->g = apply_lut_to_channel_value(&srgb_eotf, pixel->g, LUT_GREEN);
+			pixel->b = apply_lut_to_channel_value(&srgb_eotf, pixel->b, LUT_BLUE);
+			break;
+		default:
+			drm_WARN_ONCE(dev, true,
+				      "unknown colorop 1D curve type %d\n",
+				      colorop_state->curve_1d_type);
+			break;
+		}
+	}
+}
+
+static void pre_blend_color_transform(const struct vkms_plane_state *plane_state,
+				      struct line_buffer *output_buffer)
+{
+	for (size_t x = 0; x < output_buffer->n_pixels; x++) {
+		struct drm_colorop *colorop = plane_state->base.base.color_pipeline;
+
+		while (colorop) {
+			struct drm_colorop_state *colorop_state;
+
+			colorop_state = colorop->state;
+
+			if (!colorop_state)
+				return;
+
+			if (!colorop_state->bypass)
+				apply_colorop(&output_buffer->pixels[x], colorop);
+
+			colorop = colorop->next;
+		}
+	}
+}
+
 /**
  * direction_for_rotation() - Get the correct reading direction for a given rotation
  *
@@ -351,7 +400,7 @@ static void blend_line(struct vkms_plane_state *current_plane, int y,
 	 */
 	current_plane->pixel_read_line(current_plane, src_x_start, src_y_start, direction,
 				       pixel_count, &stage_buffer->pixels[dst_x_start]);
-
+	pre_blend_color_transform(current_plane, stage_buffer);
 	pre_mul_alpha_blend(stage_buffer, output_buffer,
 			    dst_x_start, pixel_count);
 }
diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
index a74a7fc3a056..2e192b1e5be8 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.h
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -265,4 +265,7 @@ void vkms_writeback_row(struct vkms_writeback_job *wb, const struct line_buffer
 /* Writeback */
 int vkms_enable_writeback_connector(struct vkms_device *vkmsdev, struct vkms_output *vkms_out);
 
+/* Colorops */
+int vkms_initialize_colorops(struct drm_plane *plane);
+
 #endif /* _VKMS_DRV_H_ */
diff --git a/drivers/gpu/drm/vkms/vkms_luts.c b/drivers/gpu/drm/vkms/vkms_luts.c
new file mode 100644
index 000000000000..7e8d3070deea
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_luts.c
@@ -0,0 +1,808 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <drm/drm_mode.h>
+
+#include "vkms_drv.h"
+#include "vkms_luts.h"
+
+/*
+ * These luts were generated with a LUT generated based on
+ * skia's transfer function code. The LUT generator can be
+ * found at
+ * https://gitlab.freedesktop.org/hwentland/lutgen
+ */
+
+static struct drm_color_lut linear_array[LUT_SIZE] = {
+	{ 0x0, 0x0, 0x0, 0 },
+	{ 0x101, 0x101, 0x101, 0 },
+	{ 0x202, 0x202, 0x202, 0 },
+	{ 0x303, 0x303, 0x303, 0 },
+	{ 0x404, 0x404, 0x404, 0 },
+	{ 0x505, 0x505, 0x505, 0 },
+	{ 0x606, 0x606, 0x606, 0 },
+	{ 0x707, 0x707, 0x707, 0 },
+	{ 0x808, 0x808, 0x808, 0 },
+	{ 0x909, 0x909, 0x909, 0 },
+	{ 0xa0a, 0xa0a, 0xa0a, 0 },
+	{ 0xb0b, 0xb0b, 0xb0b, 0 },
+	{ 0xc0c, 0xc0c, 0xc0c, 0 },
+	{ 0xd0d, 0xd0d, 0xd0d, 0 },
+	{ 0xe0e, 0xe0e, 0xe0e, 0 },
+	{ 0xf0f, 0xf0f, 0xf0f, 0 },
+	{ 0x1010, 0x1010, 0x1010, 0 },
+	{ 0x1111, 0x1111, 0x1111, 0 },
+	{ 0x1212, 0x1212, 0x1212, 0 },
+	{ 0x1313, 0x1313, 0x1313, 0 },
+	{ 0x1414, 0x1414, 0x1414, 0 },
+	{ 0x1515, 0x1515, 0x1515, 0 },
+	{ 0x1616, 0x1616, 0x1616, 0 },
+	{ 0x1717, 0x1717, 0x1717, 0 },
+	{ 0x1818, 0x1818, 0x1818, 0 },
+	{ 0x1919, 0x1919, 0x1919, 0 },
+	{ 0x1a1a, 0x1a1a, 0x1a1a, 0 },
+	{ 0x1b1b, 0x1b1b, 0x1b1b, 0 },
+	{ 0x1c1c, 0x1c1c, 0x1c1c, 0 },
+	{ 0x1d1d, 0x1d1d, 0x1d1d, 0 },
+	{ 0x1e1e, 0x1e1e, 0x1e1e, 0 },
+	{ 0x1f1f, 0x1f1f, 0x1f1f, 0 },
+	{ 0x2020, 0x2020, 0x2020, 0 },
+	{ 0x2121, 0x2121, 0x2121, 0 },
+	{ 0x2222, 0x2222, 0x2222, 0 },
+	{ 0x2323, 0x2323, 0x2323, 0 },
+	{ 0x2424, 0x2424, 0x2424, 0 },
+	{ 0x2525, 0x2525, 0x2525, 0 },
+	{ 0x2626, 0x2626, 0x2626, 0 },
+	{ 0x2727, 0x2727, 0x2727, 0 },
+	{ 0x2828, 0x2828, 0x2828, 0 },
+	{ 0x2929, 0x2929, 0x2929, 0 },
+	{ 0x2a2a, 0x2a2a, 0x2a2a, 0 },
+	{ 0x2b2b, 0x2b2b, 0x2b2b, 0 },
+	{ 0x2c2c, 0x2c2c, 0x2c2c, 0 },
+	{ 0x2d2d, 0x2d2d, 0x2d2d, 0 },
+	{ 0x2e2e, 0x2e2e, 0x2e2e, 0 },
+	{ 0x2f2f, 0x2f2f, 0x2f2f, 0 },
+	{ 0x3030, 0x3030, 0x3030, 0 },
+	{ 0x3131, 0x3131, 0x3131, 0 },
+	{ 0x3232, 0x3232, 0x3232, 0 },
+	{ 0x3333, 0x3333, 0x3333, 0 },
+	{ 0x3434, 0x3434, 0x3434, 0 },
+	{ 0x3535, 0x3535, 0x3535, 0 },
+	{ 0x3636, 0x3636, 0x3636, 0 },
+	{ 0x3737, 0x3737, 0x3737, 0 },
+	{ 0x3838, 0x3838, 0x3838, 0 },
+	{ 0x3939, 0x3939, 0x3939, 0 },
+	{ 0x3a3a, 0x3a3a, 0x3a3a, 0 },
+	{ 0x3b3b, 0x3b3b, 0x3b3b, 0 },
+	{ 0x3c3c, 0x3c3c, 0x3c3c, 0 },
+	{ 0x3d3d, 0x3d3d, 0x3d3d, 0 },
+	{ 0x3e3e, 0x3e3e, 0x3e3e, 0 },
+	{ 0x3f3f, 0x3f3f, 0x3f3f, 0 },
+	{ 0x4040, 0x4040, 0x4040, 0 },
+	{ 0x4141, 0x4141, 0x4141, 0 },
+	{ 0x4242, 0x4242, 0x4242, 0 },
+	{ 0x4343, 0x4343, 0x4343, 0 },
+	{ 0x4444, 0x4444, 0x4444, 0 },
+	{ 0x4545, 0x4545, 0x4545, 0 },
+	{ 0x4646, 0x4646, 0x4646, 0 },
+	{ 0x4747, 0x4747, 0x4747, 0 },
+	{ 0x4848, 0x4848, 0x4848, 0 },
+	{ 0x4949, 0x4949, 0x4949, 0 },
+	{ 0x4a4a, 0x4a4a, 0x4a4a, 0 },
+	{ 0x4b4b, 0x4b4b, 0x4b4b, 0 },
+	{ 0x4c4c, 0x4c4c, 0x4c4c, 0 },
+	{ 0x4d4d, 0x4d4d, 0x4d4d, 0 },
+	{ 0x4e4e, 0x4e4e, 0x4e4e, 0 },
+	{ 0x4f4f, 0x4f4f, 0x4f4f, 0 },
+	{ 0x5050, 0x5050, 0x5050, 0 },
+	{ 0x5151, 0x5151, 0x5151, 0 },
+	{ 0x5252, 0x5252, 0x5252, 0 },
+	{ 0x5353, 0x5353, 0x5353, 0 },
+	{ 0x5454, 0x5454, 0x5454, 0 },
+	{ 0x5555, 0x5555, 0x5555, 0 },
+	{ 0x5656, 0x5656, 0x5656, 0 },
+	{ 0x5757, 0x5757, 0x5757, 0 },
+	{ 0x5858, 0x5858, 0x5858, 0 },
+	{ 0x5959, 0x5959, 0x5959, 0 },
+	{ 0x5a5a, 0x5a5a, 0x5a5a, 0 },
+	{ 0x5b5b, 0x5b5b, 0x5b5b, 0 },
+	{ 0x5c5c, 0x5c5c, 0x5c5c, 0 },
+	{ 0x5d5d, 0x5d5d, 0x5d5d, 0 },
+	{ 0x5e5e, 0x5e5e, 0x5e5e, 0 },
+	{ 0x5f5f, 0x5f5f, 0x5f5f, 0 },
+	{ 0x6060, 0x6060, 0x6060, 0 },
+	{ 0x6161, 0x6161, 0x6161, 0 },
+	{ 0x6262, 0x6262, 0x6262, 0 },
+	{ 0x6363, 0x6363, 0x6363, 0 },
+	{ 0x6464, 0x6464, 0x6464, 0 },
+	{ 0x6565, 0x6565, 0x6565, 0 },
+	{ 0x6666, 0x6666, 0x6666, 0 },
+	{ 0x6767, 0x6767, 0x6767, 0 },
+	{ 0x6868, 0x6868, 0x6868, 0 },
+	{ 0x6969, 0x6969, 0x6969, 0 },
+	{ 0x6a6a, 0x6a6a, 0x6a6a, 0 },
+	{ 0x6b6b, 0x6b6b, 0x6b6b, 0 },
+	{ 0x6c6c, 0x6c6c, 0x6c6c, 0 },
+	{ 0x6d6d, 0x6d6d, 0x6d6d, 0 },
+	{ 0x6e6e, 0x6e6e, 0x6e6e, 0 },
+	{ 0x6f6f, 0x6f6f, 0x6f6f, 0 },
+	{ 0x7070, 0x7070, 0x7070, 0 },
+	{ 0x7171, 0x7171, 0x7171, 0 },
+	{ 0x7272, 0x7272, 0x7272, 0 },
+	{ 0x7373, 0x7373, 0x7373, 0 },
+	{ 0x7474, 0x7474, 0x7474, 0 },
+	{ 0x7575, 0x7575, 0x7575, 0 },
+	{ 0x7676, 0x7676, 0x7676, 0 },
+	{ 0x7777, 0x7777, 0x7777, 0 },
+	{ 0x7878, 0x7878, 0x7878, 0 },
+	{ 0x7979, 0x7979, 0x7979, 0 },
+	{ 0x7a7a, 0x7a7a, 0x7a7a, 0 },
+	{ 0x7b7b, 0x7b7b, 0x7b7b, 0 },
+	{ 0x7c7c, 0x7c7c, 0x7c7c, 0 },
+	{ 0x7d7d, 0x7d7d, 0x7d7d, 0 },
+	{ 0x7e7e, 0x7e7e, 0x7e7e, 0 },
+	{ 0x7f7f, 0x7f7f, 0x7f7f, 0 },
+	{ 0x8080, 0x8080, 0x8080, 0 },
+	{ 0x8181, 0x8181, 0x8181, 0 },
+	{ 0x8282, 0x8282, 0x8282, 0 },
+	{ 0x8383, 0x8383, 0x8383, 0 },
+	{ 0x8484, 0x8484, 0x8484, 0 },
+	{ 0x8585, 0x8585, 0x8585, 0 },
+	{ 0x8686, 0x8686, 0x8686, 0 },
+	{ 0x8787, 0x8787, 0x8787, 0 },
+	{ 0x8888, 0x8888, 0x8888, 0 },
+	{ 0x8989, 0x8989, 0x8989, 0 },
+	{ 0x8a8a, 0x8a8a, 0x8a8a, 0 },
+	{ 0x8b8b, 0x8b8b, 0x8b8b, 0 },
+	{ 0x8c8c, 0x8c8c, 0x8c8c, 0 },
+	{ 0x8d8d, 0x8d8d, 0x8d8d, 0 },
+	{ 0x8e8e, 0x8e8e, 0x8e8e, 0 },
+	{ 0x8f8f, 0x8f8f, 0x8f8f, 0 },
+	{ 0x9090, 0x9090, 0x9090, 0 },
+	{ 0x9191, 0x9191, 0x9191, 0 },
+	{ 0x9292, 0x9292, 0x9292, 0 },
+	{ 0x9393, 0x9393, 0x9393, 0 },
+	{ 0x9494, 0x9494, 0x9494, 0 },
+	{ 0x9595, 0x9595, 0x9595, 0 },
+	{ 0x9696, 0x9696, 0x9696, 0 },
+	{ 0x9797, 0x9797, 0x9797, 0 },
+	{ 0x9898, 0x9898, 0x9898, 0 },
+	{ 0x9999, 0x9999, 0x9999, 0 },
+	{ 0x9a9a, 0x9a9a, 0x9a9a, 0 },
+	{ 0x9b9b, 0x9b9b, 0x9b9b, 0 },
+	{ 0x9c9c, 0x9c9c, 0x9c9c, 0 },
+	{ 0x9d9d, 0x9d9d, 0x9d9d, 0 },
+	{ 0x9e9e, 0x9e9e, 0x9e9e, 0 },
+	{ 0x9f9f, 0x9f9f, 0x9f9f, 0 },
+	{ 0xa0a0, 0xa0a0, 0xa0a0, 0 },
+	{ 0xa1a1, 0xa1a1, 0xa1a1, 0 },
+	{ 0xa2a2, 0xa2a2, 0xa2a2, 0 },
+	{ 0xa3a3, 0xa3a3, 0xa3a3, 0 },
+	{ 0xa4a4, 0xa4a4, 0xa4a4, 0 },
+	{ 0xa5a5, 0xa5a5, 0xa5a5, 0 },
+	{ 0xa6a6, 0xa6a6, 0xa6a6, 0 },
+	{ 0xa7a7, 0xa7a7, 0xa7a7, 0 },
+	{ 0xa8a8, 0xa8a8, 0xa8a8, 0 },
+	{ 0xa9a9, 0xa9a9, 0xa9a9, 0 },
+	{ 0xaaaa, 0xaaaa, 0xaaaa, 0 },
+	{ 0xabab, 0xabab, 0xabab, 0 },
+	{ 0xacac, 0xacac, 0xacac, 0 },
+	{ 0xadad, 0xadad, 0xadad, 0 },
+	{ 0xaeae, 0xaeae, 0xaeae, 0 },
+	{ 0xafaf, 0xafaf, 0xafaf, 0 },
+	{ 0xb0b0, 0xb0b0, 0xb0b0, 0 },
+	{ 0xb1b1, 0xb1b1, 0xb1b1, 0 },
+	{ 0xb2b2, 0xb2b2, 0xb2b2, 0 },
+	{ 0xb3b3, 0xb3b3, 0xb3b3, 0 },
+	{ 0xb4b4, 0xb4b4, 0xb4b4, 0 },
+	{ 0xb5b5, 0xb5b5, 0xb5b5, 0 },
+	{ 0xb6b6, 0xb6b6, 0xb6b6, 0 },
+	{ 0xb7b7, 0xb7b7, 0xb7b7, 0 },
+	{ 0xb8b8, 0xb8b8, 0xb8b8, 0 },
+	{ 0xb9b9, 0xb9b9, 0xb9b9, 0 },
+	{ 0xbaba, 0xbaba, 0xbaba, 0 },
+	{ 0xbbbb, 0xbbbb, 0xbbbb, 0 },
+	{ 0xbcbc, 0xbcbc, 0xbcbc, 0 },
+	{ 0xbdbd, 0xbdbd, 0xbdbd, 0 },
+	{ 0xbebe, 0xbebe, 0xbebe, 0 },
+	{ 0xbfbf, 0xbfbf, 0xbfbf, 0 },
+	{ 0xc0c0, 0xc0c0, 0xc0c0, 0 },
+	{ 0xc1c1, 0xc1c1, 0xc1c1, 0 },
+	{ 0xc2c2, 0xc2c2, 0xc2c2, 0 },
+	{ 0xc3c3, 0xc3c3, 0xc3c3, 0 },
+	{ 0xc4c4, 0xc4c4, 0xc4c4, 0 },
+	{ 0xc5c5, 0xc5c5, 0xc5c5, 0 },
+	{ 0xc6c6, 0xc6c6, 0xc6c6, 0 },
+	{ 0xc7c7, 0xc7c7, 0xc7c7, 0 },
+	{ 0xc8c8, 0xc8c8, 0xc8c8, 0 },
+	{ 0xc9c9, 0xc9c9, 0xc9c9, 0 },
+	{ 0xcaca, 0xcaca, 0xcaca, 0 },
+	{ 0xcbcb, 0xcbcb, 0xcbcb, 0 },
+	{ 0xcccc, 0xcccc, 0xcccc, 0 },
+	{ 0xcdcd, 0xcdcd, 0xcdcd, 0 },
+	{ 0xcece, 0xcece, 0xcece, 0 },
+	{ 0xcfcf, 0xcfcf, 0xcfcf, 0 },
+	{ 0xd0d0, 0xd0d0, 0xd0d0, 0 },
+	{ 0xd1d1, 0xd1d1, 0xd1d1, 0 },
+	{ 0xd2d2, 0xd2d2, 0xd2d2, 0 },
+	{ 0xd3d3, 0xd3d3, 0xd3d3, 0 },
+	{ 0xd4d4, 0xd4d4, 0xd4d4, 0 },
+	{ 0xd5d5, 0xd5d5, 0xd5d5, 0 },
+	{ 0xd6d6, 0xd6d6, 0xd6d6, 0 },
+	{ 0xd7d7, 0xd7d7, 0xd7d7, 0 },
+	{ 0xd8d8, 0xd8d8, 0xd8d8, 0 },
+	{ 0xd9d9, 0xd9d9, 0xd9d9, 0 },
+	{ 0xdada, 0xdada, 0xdada, 0 },
+	{ 0xdbdb, 0xdbdb, 0xdbdb, 0 },
+	{ 0xdcdc, 0xdcdc, 0xdcdc, 0 },
+	{ 0xdddd, 0xdddd, 0xdddd, 0 },
+	{ 0xdede, 0xdede, 0xdede, 0 },
+	{ 0xdfdf, 0xdfdf, 0xdfdf, 0 },
+	{ 0xe0e0, 0xe0e0, 0xe0e0, 0 },
+	{ 0xe1e1, 0xe1e1, 0xe1e1, 0 },
+	{ 0xe2e2, 0xe2e2, 0xe2e2, 0 },
+	{ 0xe3e3, 0xe3e3, 0xe3e3, 0 },
+	{ 0xe4e4, 0xe4e4, 0xe4e4, 0 },
+	{ 0xe5e5, 0xe5e5, 0xe5e5, 0 },
+	{ 0xe6e6, 0xe6e6, 0xe6e6, 0 },
+	{ 0xe7e7, 0xe7e7, 0xe7e7, 0 },
+	{ 0xe8e8, 0xe8e8, 0xe8e8, 0 },
+	{ 0xe9e9, 0xe9e9, 0xe9e9, 0 },
+	{ 0xeaea, 0xeaea, 0xeaea, 0 },
+	{ 0xebeb, 0xebeb, 0xebeb, 0 },
+	{ 0xecec, 0xecec, 0xecec, 0 },
+	{ 0xeded, 0xeded, 0xeded, 0 },
+	{ 0xeeee, 0xeeee, 0xeeee, 0 },
+	{ 0xefef, 0xefef, 0xefef, 0 },
+	{ 0xf0f0, 0xf0f0, 0xf0f0, 0 },
+	{ 0xf1f1, 0xf1f1, 0xf1f1, 0 },
+	{ 0xf2f2, 0xf2f2, 0xf2f2, 0 },
+	{ 0xf3f3, 0xf3f3, 0xf3f3, 0 },
+	{ 0xf4f4, 0xf4f4, 0xf4f4, 0 },
+	{ 0xf5f5, 0xf5f5, 0xf5f5, 0 },
+	{ 0xf6f6, 0xf6f6, 0xf6f6, 0 },
+	{ 0xf7f7, 0xf7f7, 0xf7f7, 0 },
+	{ 0xf8f8, 0xf8f8, 0xf8f8, 0 },
+	{ 0xf9f9, 0xf9f9, 0xf9f9, 0 },
+	{ 0xfafa, 0xfafa, 0xfafa, 0 },
+	{ 0xfbfb, 0xfbfb, 0xfbfb, 0 },
+	{ 0xfcfc, 0xfcfc, 0xfcfc, 0 },
+	{ 0xfdfd, 0xfdfd, 0xfdfd, 0 },
+	{ 0xfefe, 0xfefe, 0xfefe, 0 },
+	{ 0xffff, 0xffff, 0xffff, 0 },
+};
+
+const struct vkms_color_lut linear_eotf = {
+	.base = linear_array,
+	.lut_length = LUT_SIZE,
+	.channel_value2index_ratio = 0xff00ffll
+};
+
+static struct drm_color_lut srgb_array[LUT_SIZE] = {
+	{ 0x0, 0x0, 0x0, 0 },
+	{ 0x13, 0x13, 0x13, 0 },
+	{ 0x27, 0x27, 0x27, 0 },
+	{ 0x3b, 0x3b, 0x3b, 0 },
+	{ 0x4f, 0x4f, 0x4f, 0 },
+	{ 0x63, 0x63, 0x63, 0 },
+	{ 0x77, 0x77, 0x77, 0 },
+	{ 0x8b, 0x8b, 0x8b, 0 },
+	{ 0x9f, 0x9f, 0x9f, 0 },
+	{ 0xb3, 0xb3, 0xb3, 0 },
+	{ 0xc6, 0xc6, 0xc6, 0 },
+	{ 0xdb, 0xdb, 0xdb, 0 },
+	{ 0xf0, 0xf0, 0xf0, 0 },
+	{ 0x107, 0x107, 0x107, 0 },
+	{ 0x11f, 0x11f, 0x11f, 0 },
+	{ 0x139, 0x139, 0x139, 0 },
+	{ 0x153, 0x153, 0x153, 0 },
+	{ 0x16f, 0x16f, 0x16f, 0 },
+	{ 0x18c, 0x18c, 0x18c, 0 },
+	{ 0x1aa, 0x1aa, 0x1aa, 0 },
+	{ 0x1ca, 0x1ca, 0x1ca, 0 },
+	{ 0x1eb, 0x1eb, 0x1eb, 0 },
+	{ 0x20d, 0x20d, 0x20d, 0 },
+	{ 0x231, 0x231, 0x231, 0 },
+	{ 0x256, 0x256, 0x256, 0 },
+	{ 0x27d, 0x27d, 0x27d, 0 },
+	{ 0x2a4, 0x2a4, 0x2a4, 0 },
+	{ 0x2ce, 0x2ce, 0x2ce, 0 },
+	{ 0x2f9, 0x2f9, 0x2f9, 0 },
+	{ 0x325, 0x325, 0x325, 0 },
+	{ 0x352, 0x352, 0x352, 0 },
+	{ 0x381, 0x381, 0x381, 0 },
+	{ 0x3b2, 0x3b2, 0x3b2, 0 },
+	{ 0x3e4, 0x3e4, 0x3e4, 0 },
+	{ 0x418, 0x418, 0x418, 0 },
+	{ 0x44d, 0x44d, 0x44d, 0 },
+	{ 0x484, 0x484, 0x484, 0 },
+	{ 0x4bc, 0x4bc, 0x4bc, 0 },
+	{ 0x4f6, 0x4f6, 0x4f6, 0 },
+	{ 0x531, 0x531, 0x531, 0 },
+	{ 0x56e, 0x56e, 0x56e, 0 },
+	{ 0x5ad, 0x5ad, 0x5ad, 0 },
+	{ 0x5ed, 0x5ed, 0x5ed, 0 },
+	{ 0x62f, 0x62f, 0x62f, 0 },
+	{ 0x672, 0x672, 0x672, 0 },
+	{ 0x6b7, 0x6b7, 0x6b7, 0 },
+	{ 0x6fe, 0x6fe, 0x6fe, 0 },
+	{ 0x746, 0x746, 0x746, 0 },
+	{ 0x791, 0x791, 0x791, 0 },
+	{ 0x7dc, 0x7dc, 0x7dc, 0 },
+	{ 0x82a, 0x82a, 0x82a, 0 },
+	{ 0x879, 0x879, 0x879, 0 },
+	{ 0x8ca, 0x8ca, 0x8ca, 0 },
+	{ 0x91d, 0x91d, 0x91d, 0 },
+	{ 0x971, 0x971, 0x971, 0 },
+	{ 0x9c7, 0x9c7, 0x9c7, 0 },
+	{ 0xa1f, 0xa1f, 0xa1f, 0 },
+	{ 0xa79, 0xa79, 0xa79, 0 },
+	{ 0xad4, 0xad4, 0xad4, 0 },
+	{ 0xb32, 0xb32, 0xb32, 0 },
+	{ 0xb91, 0xb91, 0xb91, 0 },
+	{ 0xbf2, 0xbf2, 0xbf2, 0 },
+	{ 0xc54, 0xc54, 0xc54, 0 },
+	{ 0xcb9, 0xcb9, 0xcb9, 0 },
+	{ 0xd1f, 0xd1f, 0xd1f, 0 },
+	{ 0xd88, 0xd88, 0xd88, 0 },
+	{ 0xdf2, 0xdf2, 0xdf2, 0 },
+	{ 0xe5e, 0xe5e, 0xe5e, 0 },
+	{ 0xecc, 0xecc, 0xecc, 0 },
+	{ 0xf3c, 0xf3c, 0xf3c, 0 },
+	{ 0xfad, 0xfad, 0xfad, 0 },
+	{ 0x1021, 0x1021, 0x1021, 0 },
+	{ 0x1096, 0x1096, 0x1096, 0 },
+	{ 0x110e, 0x110e, 0x110e, 0 },
+	{ 0x1187, 0x1187, 0x1187, 0 },
+	{ 0x1203, 0x1203, 0x1203, 0 },
+	{ 0x1280, 0x1280, 0x1280, 0 },
+	{ 0x12ff, 0x12ff, 0x12ff, 0 },
+	{ 0x1380, 0x1380, 0x1380, 0 },
+	{ 0x1404, 0x1404, 0x1404, 0 },
+	{ 0x1489, 0x1489, 0x1489, 0 },
+	{ 0x1510, 0x1510, 0x1510, 0 },
+	{ 0x1599, 0x1599, 0x1599, 0 },
+	{ 0x1624, 0x1624, 0x1624, 0 },
+	{ 0x16b2, 0x16b2, 0x16b2, 0 },
+	{ 0x1741, 0x1741, 0x1741, 0 },
+	{ 0x17d2, 0x17d2, 0x17d2, 0 },
+	{ 0x1865, 0x1865, 0x1865, 0 },
+	{ 0x18fb, 0x18fb, 0x18fb, 0 },
+	{ 0x1992, 0x1992, 0x1992, 0 },
+	{ 0x1a2c, 0x1a2c, 0x1a2c, 0 },
+	{ 0x1ac8, 0x1ac8, 0x1ac8, 0 },
+	{ 0x1b65, 0x1b65, 0x1b65, 0 },
+	{ 0x1c05, 0x1c05, 0x1c05, 0 },
+	{ 0x1ca7, 0x1ca7, 0x1ca7, 0 },
+	{ 0x1d4b, 0x1d4b, 0x1d4b, 0 },
+	{ 0x1df1, 0x1df1, 0x1df1, 0 },
+	{ 0x1e99, 0x1e99, 0x1e99, 0 },
+	{ 0x1f44, 0x1f44, 0x1f44, 0 },
+	{ 0x1ff0, 0x1ff0, 0x1ff0, 0 },
+	{ 0x209f, 0x209f, 0x209f, 0 },
+	{ 0x2150, 0x2150, 0x2150, 0 },
+	{ 0x2203, 0x2203, 0x2203, 0 },
+	{ 0x22b8, 0x22b8, 0x22b8, 0 },
+	{ 0x2370, 0x2370, 0x2370, 0 },
+	{ 0x2429, 0x2429, 0x2429, 0 },
+	{ 0x24e5, 0x24e5, 0x24e5, 0 },
+	{ 0x25a3, 0x25a3, 0x25a3, 0 },
+	{ 0x2663, 0x2663, 0x2663, 0 },
+	{ 0x2726, 0x2726, 0x2726, 0 },
+	{ 0x27ea, 0x27ea, 0x27ea, 0 },
+	{ 0x28b1, 0x28b1, 0x28b1, 0 },
+	{ 0x297a, 0x297a, 0x297a, 0 },
+	{ 0x2a45, 0x2a45, 0x2a45, 0 },
+	{ 0x2b13, 0x2b13, 0x2b13, 0 },
+	{ 0x2be3, 0x2be3, 0x2be3, 0 },
+	{ 0x2cb5, 0x2cb5, 0x2cb5, 0 },
+	{ 0x2d89, 0x2d89, 0x2d89, 0 },
+	{ 0x2e60, 0x2e60, 0x2e60, 0 },
+	{ 0x2f39, 0x2f39, 0x2f39, 0 },
+	{ 0x3014, 0x3014, 0x3014, 0 },
+	{ 0x30f2, 0x30f2, 0x30f2, 0 },
+	{ 0x31d2, 0x31d2, 0x31d2, 0 },
+	{ 0x32b4, 0x32b4, 0x32b4, 0 },
+	{ 0x3398, 0x3398, 0x3398, 0 },
+	{ 0x347f, 0x347f, 0x347f, 0 },
+	{ 0x3569, 0x3569, 0x3569, 0 },
+	{ 0x3654, 0x3654, 0x3654, 0 },
+	{ 0x3742, 0x3742, 0x3742, 0 },
+	{ 0x3832, 0x3832, 0x3832, 0 },
+	{ 0x3925, 0x3925, 0x3925, 0 },
+	{ 0x3a1a, 0x3a1a, 0x3a1a, 0 },
+	{ 0x3b11, 0x3b11, 0x3b11, 0 },
+	{ 0x3c0b, 0x3c0b, 0x3c0b, 0 },
+	{ 0x3d07, 0x3d07, 0x3d07, 0 },
+	{ 0x3e05, 0x3e05, 0x3e05, 0 },
+	{ 0x3f06, 0x3f06, 0x3f06, 0 },
+	{ 0x400a, 0x400a, 0x400a, 0 },
+	{ 0x410f, 0x410f, 0x410f, 0 },
+	{ 0x4218, 0x4218, 0x4218, 0 },
+	{ 0x4322, 0x4322, 0x4322, 0 },
+	{ 0x442f, 0x442f, 0x442f, 0 },
+	{ 0x453f, 0x453f, 0x453f, 0 },
+	{ 0x4650, 0x4650, 0x4650, 0 },
+	{ 0x4765, 0x4765, 0x4765, 0 },
+	{ 0x487c, 0x487c, 0x487c, 0 },
+	{ 0x4995, 0x4995, 0x4995, 0 },
+	{ 0x4ab1, 0x4ab1, 0x4ab1, 0 },
+	{ 0x4bcf, 0x4bcf, 0x4bcf, 0 },
+	{ 0x4cf0, 0x4cf0, 0x4cf0, 0 },
+	{ 0x4e13, 0x4e13, 0x4e13, 0 },
+	{ 0x4f39, 0x4f39, 0x4f39, 0 },
+	{ 0x5061, 0x5061, 0x5061, 0 },
+	{ 0x518b, 0x518b, 0x518b, 0 },
+	{ 0x52b9, 0x52b9, 0x52b9, 0 },
+	{ 0x53e8, 0x53e8, 0x53e8, 0 },
+	{ 0x551b, 0x551b, 0x551b, 0 },
+	{ 0x5650, 0x5650, 0x5650, 0 },
+	{ 0x5787, 0x5787, 0x5787, 0 },
+	{ 0x58c1, 0x58c1, 0x58c1, 0 },
+	{ 0x59fd, 0x59fd, 0x59fd, 0 },
+	{ 0x5b3c, 0x5b3c, 0x5b3c, 0 },
+	{ 0x5c7e, 0x5c7e, 0x5c7e, 0 },
+	{ 0x5dc2, 0x5dc2, 0x5dc2, 0 },
+	{ 0x5f09, 0x5f09, 0x5f09, 0 },
+	{ 0x6052, 0x6052, 0x6052, 0 },
+	{ 0x619e, 0x619e, 0x619e, 0 },
+	{ 0x62ec, 0x62ec, 0x62ec, 0 },
+	{ 0x643d, 0x643d, 0x643d, 0 },
+	{ 0x6591, 0x6591, 0x6591, 0 },
+	{ 0x66e7, 0x66e7, 0x66e7, 0 },
+	{ 0x6840, 0x6840, 0x6840, 0 },
+	{ 0x699b, 0x699b, 0x699b, 0 },
+	{ 0x6afa, 0x6afa, 0x6afa, 0 },
+	{ 0x6c5a, 0x6c5a, 0x6c5a, 0 },
+	{ 0x6dbe, 0x6dbe, 0x6dbe, 0 },
+	{ 0x6f24, 0x6f24, 0x6f24, 0 },
+	{ 0x708c, 0x708c, 0x708c, 0 },
+	{ 0x71f8, 0x71f8, 0x71f8, 0 },
+	{ 0x7366, 0x7366, 0x7366, 0 },
+	{ 0x74d6, 0x74d6, 0x74d6, 0 },
+	{ 0x764a, 0x764a, 0x764a, 0 },
+	{ 0x77c0, 0x77c0, 0x77c0, 0 },
+	{ 0x7938, 0x7938, 0x7938, 0 },
+	{ 0x7ab4, 0x7ab4, 0x7ab4, 0 },
+	{ 0x7c32, 0x7c32, 0x7c32, 0 },
+	{ 0x7db3, 0x7db3, 0x7db3, 0 },
+	{ 0x7f36, 0x7f36, 0x7f36, 0 },
+	{ 0x80bc, 0x80bc, 0x80bc, 0 },
+	{ 0x8245, 0x8245, 0x8245, 0 },
+	{ 0x83d1, 0x83d1, 0x83d1, 0 },
+	{ 0x855f, 0x855f, 0x855f, 0 },
+	{ 0x86f0, 0x86f0, 0x86f0, 0 },
+	{ 0x8884, 0x8884, 0x8884, 0 },
+	{ 0x8a1a, 0x8a1a, 0x8a1a, 0 },
+	{ 0x8bb4, 0x8bb4, 0x8bb4, 0 },
+	{ 0x8d50, 0x8d50, 0x8d50, 0 },
+	{ 0x8eee, 0x8eee, 0x8eee, 0 },
+	{ 0x9090, 0x9090, 0x9090, 0 },
+	{ 0x9234, 0x9234, 0x9234, 0 },
+	{ 0x93db, 0x93db, 0x93db, 0 },
+	{ 0x9585, 0x9585, 0x9585, 0 },
+	{ 0x9732, 0x9732, 0x9732, 0 },
+	{ 0x98e1, 0x98e1, 0x98e1, 0 },
+	{ 0x9a93, 0x9a93, 0x9a93, 0 },
+	{ 0x9c48, 0x9c48, 0x9c48, 0 },
+	{ 0x9e00, 0x9e00, 0x9e00, 0 },
+	{ 0x9fbb, 0x9fbb, 0x9fbb, 0 },
+	{ 0xa178, 0xa178, 0xa178, 0 },
+	{ 0xa338, 0xa338, 0xa338, 0 },
+	{ 0xa4fb, 0xa4fb, 0xa4fb, 0 },
+	{ 0xa6c1, 0xa6c1, 0xa6c1, 0 },
+	{ 0xa88a, 0xa88a, 0xa88a, 0 },
+	{ 0xaa56, 0xaa56, 0xaa56, 0 },
+	{ 0xac24, 0xac24, 0xac24, 0 },
+	{ 0xadf5, 0xadf5, 0xadf5, 0 },
+	{ 0xafc9, 0xafc9, 0xafc9, 0 },
+	{ 0xb1a0, 0xb1a0, 0xb1a0, 0 },
+	{ 0xb37a, 0xb37a, 0xb37a, 0 },
+	{ 0xb557, 0xb557, 0xb557, 0 },
+	{ 0xb736, 0xb736, 0xb736, 0 },
+	{ 0xb919, 0xb919, 0xb919, 0 },
+	{ 0xbafe, 0xbafe, 0xbafe, 0 },
+	{ 0xbce6, 0xbce6, 0xbce6, 0 },
+	{ 0xbed2, 0xbed2, 0xbed2, 0 },
+	{ 0xc0c0, 0xc0c0, 0xc0c0, 0 },
+	{ 0xc2b0, 0xc2b0, 0xc2b0, 0 },
+	{ 0xc4a4, 0xc4a4, 0xc4a4, 0 },
+	{ 0xc69b, 0xc69b, 0xc69b, 0 },
+	{ 0xc895, 0xc895, 0xc895, 0 },
+	{ 0xca91, 0xca91, 0xca91, 0 },
+	{ 0xcc91, 0xcc91, 0xcc91, 0 },
+	{ 0xce93, 0xce93, 0xce93, 0 },
+	{ 0xd098, 0xd098, 0xd098, 0 },
+	{ 0xd2a1, 0xd2a1, 0xd2a1, 0 },
+	{ 0xd4ac, 0xd4ac, 0xd4ac, 0 },
+	{ 0xd6ba, 0xd6ba, 0xd6ba, 0 },
+	{ 0xd8cb, 0xd8cb, 0xd8cb, 0 },
+	{ 0xdadf, 0xdadf, 0xdadf, 0 },
+	{ 0xdcf7, 0xdcf7, 0xdcf7, 0 },
+	{ 0xdf11, 0xdf11, 0xdf11, 0 },
+	{ 0xe12e, 0xe12e, 0xe12e, 0 },
+	{ 0xe34e, 0xe34e, 0xe34e, 0 },
+	{ 0xe571, 0xe571, 0xe571, 0 },
+	{ 0xe796, 0xe796, 0xe796, 0 },
+	{ 0xe9bf, 0xe9bf, 0xe9bf, 0 },
+	{ 0xebeb, 0xebeb, 0xebeb, 0 },
+	{ 0xee1a, 0xee1a, 0xee1a, 0 },
+	{ 0xf04c, 0xf04c, 0xf04c, 0 },
+	{ 0xf281, 0xf281, 0xf281, 0 },
+	{ 0xf4b9, 0xf4b9, 0xf4b9, 0 },
+	{ 0xf6f4, 0xf6f4, 0xf6f4, 0 },
+	{ 0xf932, 0xf932, 0xf932, 0 },
+	{ 0xfb73, 0xfb73, 0xfb73, 0 },
+	{ 0xfdb7, 0xfdb7, 0xfdb7, 0 },
+	{ 0xffff, 0xffff, 0xffff, 0 },
+};
+
+const struct vkms_color_lut srgb_eotf = {
+	.base = srgb_array,
+	.lut_length = LUT_SIZE,
+	.channel_value2index_ratio = 0xff00ffll
+};
+
+static struct drm_color_lut srgb_inv_array[LUT_SIZE] = {
+	{ 0x0, 0x0, 0x0, 0 },
+	{ 0xcc2, 0xcc2, 0xcc2, 0 },
+	{ 0x15be, 0x15be, 0x15be, 0 },
+	{ 0x1c56, 0x1c56, 0x1c56, 0 },
+	{ 0x21bd, 0x21bd, 0x21bd, 0 },
+	{ 0x2666, 0x2666, 0x2666, 0 },
+	{ 0x2a8a, 0x2a8a, 0x2a8a, 0 },
+	{ 0x2e4c, 0x2e4c, 0x2e4c, 0 },
+	{ 0x31c0, 0x31c0, 0x31c0, 0 },
+	{ 0x34f6, 0x34f6, 0x34f6, 0 },
+	{ 0x37f9, 0x37f9, 0x37f9, 0 },
+	{ 0x3acf, 0x3acf, 0x3acf, 0 },
+	{ 0x3d80, 0x3d80, 0x3d80, 0 },
+	{ 0x4010, 0x4010, 0x4010, 0 },
+	{ 0x4284, 0x4284, 0x4284, 0 },
+	{ 0x44dd, 0x44dd, 0x44dd, 0 },
+	{ 0x4720, 0x4720, 0x4720, 0 },
+	{ 0x494e, 0x494e, 0x494e, 0 },
+	{ 0x4b69, 0x4b69, 0x4b69, 0 },
+	{ 0x4d73, 0x4d73, 0x4d73, 0 },
+	{ 0x4f6e, 0x4f6e, 0x4f6e, 0 },
+	{ 0x5159, 0x5159, 0x5159, 0 },
+	{ 0x5337, 0x5337, 0x5337, 0 },
+	{ 0x5509, 0x5509, 0x5509, 0 },
+	{ 0x56cf, 0x56cf, 0x56cf, 0 },
+	{ 0x588a, 0x588a, 0x588a, 0 },
+	{ 0x5a3b, 0x5a3b, 0x5a3b, 0 },
+	{ 0x5be2, 0x5be2, 0x5be2, 0 },
+	{ 0x5d80, 0x5d80, 0x5d80, 0 },
+	{ 0x5f16, 0x5f16, 0x5f16, 0 },
+	{ 0x60a4, 0x60a4, 0x60a4, 0 },
+	{ 0x6229, 0x6229, 0x6229, 0 },
+	{ 0x63a8, 0x63a8, 0x63a8, 0 },
+	{ 0x6520, 0x6520, 0x6520, 0 },
+	{ 0x6691, 0x6691, 0x6691, 0 },
+	{ 0x67fc, 0x67fc, 0x67fc, 0 },
+	{ 0x6961, 0x6961, 0x6961, 0 },
+	{ 0x6ac0, 0x6ac0, 0x6ac0, 0 },
+	{ 0x6c19, 0x6c19, 0x6c19, 0 },
+	{ 0x6d6e, 0x6d6e, 0x6d6e, 0 },
+	{ 0x6ebd, 0x6ebd, 0x6ebd, 0 },
+	{ 0x7008, 0x7008, 0x7008, 0 },
+	{ 0x714d, 0x714d, 0x714d, 0 },
+	{ 0x728f, 0x728f, 0x728f, 0 },
+	{ 0x73cc, 0x73cc, 0x73cc, 0 },
+	{ 0x7504, 0x7504, 0x7504, 0 },
+	{ 0x7639, 0x7639, 0x7639, 0 },
+	{ 0x776a, 0x776a, 0x776a, 0 },
+	{ 0x7897, 0x7897, 0x7897, 0 },
+	{ 0x79c1, 0x79c1, 0x79c1, 0 },
+	{ 0x7ae7, 0x7ae7, 0x7ae7, 0 },
+	{ 0x7c09, 0x7c09, 0x7c09, 0 },
+	{ 0x7d28, 0x7d28, 0x7d28, 0 },
+	{ 0x7e44, 0x7e44, 0x7e44, 0 },
+	{ 0x7f5d, 0x7f5d, 0x7f5d, 0 },
+	{ 0x8073, 0x8073, 0x8073, 0 },
+	{ 0x8186, 0x8186, 0x8186, 0 },
+	{ 0x8296, 0x8296, 0x8296, 0 },
+	{ 0x83a4, 0x83a4, 0x83a4, 0 },
+	{ 0x84ae, 0x84ae, 0x84ae, 0 },
+	{ 0x85b6, 0x85b6, 0x85b6, 0 },
+	{ 0x86bc, 0x86bc, 0x86bc, 0 },
+	{ 0x87bf, 0x87bf, 0x87bf, 0 },
+	{ 0x88bf, 0x88bf, 0x88bf, 0 },
+	{ 0x89be, 0x89be, 0x89be, 0 },
+	{ 0x8ab9, 0x8ab9, 0x8ab9, 0 },
+	{ 0x8bb3, 0x8bb3, 0x8bb3, 0 },
+	{ 0x8cab, 0x8cab, 0x8cab, 0 },
+	{ 0x8da0, 0x8da0, 0x8da0, 0 },
+	{ 0x8e93, 0x8e93, 0x8e93, 0 },
+	{ 0x8f84, 0x8f84, 0x8f84, 0 },
+	{ 0x9073, 0x9073, 0x9073, 0 },
+	{ 0x9161, 0x9161, 0x9161, 0 },
+	{ 0x924c, 0x924c, 0x924c, 0 },
+	{ 0x9335, 0x9335, 0x9335, 0 },
+	{ 0x941d, 0x941d, 0x941d, 0 },
+	{ 0x9503, 0x9503, 0x9503, 0 },
+	{ 0x95e7, 0x95e7, 0x95e7, 0 },
+	{ 0x96c9, 0x96c9, 0x96c9, 0 },
+	{ 0x97aa, 0x97aa, 0x97aa, 0 },
+	{ 0x9889, 0x9889, 0x9889, 0 },
+	{ 0x9966, 0x9966, 0x9966, 0 },
+	{ 0x9a42, 0x9a42, 0x9a42, 0 },
+	{ 0x9b1c, 0x9b1c, 0x9b1c, 0 },
+	{ 0x9bf5, 0x9bf5, 0x9bf5, 0 },
+	{ 0x9ccc, 0x9ccc, 0x9ccc, 0 },
+	{ 0x9da1, 0x9da1, 0x9da1, 0 },
+	{ 0x9e76, 0x9e76, 0x9e76, 0 },
+	{ 0x9f49, 0x9f49, 0x9f49, 0 },
+	{ 0xa01a, 0xa01a, 0xa01a, 0 },
+	{ 0xa0ea, 0xa0ea, 0xa0ea, 0 },
+	{ 0xa1b9, 0xa1b9, 0xa1b9, 0 },
+	{ 0xa286, 0xa286, 0xa286, 0 },
+	{ 0xa352, 0xa352, 0xa352, 0 },
+	{ 0xa41d, 0xa41d, 0xa41d, 0 },
+	{ 0xa4e7, 0xa4e7, 0xa4e7, 0 },
+	{ 0xa5af, 0xa5af, 0xa5af, 0 },
+	{ 0xa676, 0xa676, 0xa676, 0 },
+	{ 0xa73c, 0xa73c, 0xa73c, 0 },
+	{ 0xa801, 0xa801, 0xa801, 0 },
+	{ 0xa8c5, 0xa8c5, 0xa8c5, 0 },
+	{ 0xa987, 0xa987, 0xa987, 0 },
+	{ 0xaa48, 0xaa48, 0xaa48, 0 },
+	{ 0xab09, 0xab09, 0xab09, 0 },
+	{ 0xabc8, 0xabc8, 0xabc8, 0 },
+	{ 0xac86, 0xac86, 0xac86, 0 },
+	{ 0xad43, 0xad43, 0xad43, 0 },
+	{ 0xadff, 0xadff, 0xadff, 0 },
+	{ 0xaeba, 0xaeba, 0xaeba, 0 },
+	{ 0xaf74, 0xaf74, 0xaf74, 0 },
+	{ 0xb02d, 0xb02d, 0xb02d, 0 },
+	{ 0xb0e5, 0xb0e5, 0xb0e5, 0 },
+	{ 0xb19c, 0xb19c, 0xb19c, 0 },
+	{ 0xb252, 0xb252, 0xb252, 0 },
+	{ 0xb307, 0xb307, 0xb307, 0 },
+	{ 0xb3bb, 0xb3bb, 0xb3bb, 0 },
+	{ 0xb46f, 0xb46f, 0xb46f, 0 },
+	{ 0xb521, 0xb521, 0xb521, 0 },
+	{ 0xb5d3, 0xb5d3, 0xb5d3, 0 },
+	{ 0xb683, 0xb683, 0xb683, 0 },
+	{ 0xb733, 0xb733, 0xb733, 0 },
+	{ 0xb7e2, 0xb7e2, 0xb7e2, 0 },
+	{ 0xb890, 0xb890, 0xb890, 0 },
+	{ 0xb93d, 0xb93d, 0xb93d, 0 },
+	{ 0xb9ea, 0xb9ea, 0xb9ea, 0 },
+	{ 0xba96, 0xba96, 0xba96, 0 },
+	{ 0xbb40, 0xbb40, 0xbb40, 0 },
+	{ 0xbbea, 0xbbea, 0xbbea, 0 },
+	{ 0xbc94, 0xbc94, 0xbc94, 0 },
+	{ 0xbd3c, 0xbd3c, 0xbd3c, 0 },
+	{ 0xbde4, 0xbde4, 0xbde4, 0 },
+	{ 0xbe8b, 0xbe8b, 0xbe8b, 0 },
+	{ 0xbf31, 0xbf31, 0xbf31, 0 },
+	{ 0xbfd7, 0xbfd7, 0xbfd7, 0 },
+	{ 0xc07b, 0xc07b, 0xc07b, 0 },
+	{ 0xc120, 0xc120, 0xc120, 0 },
+	{ 0xc1c3, 0xc1c3, 0xc1c3, 0 },
+	{ 0xc266, 0xc266, 0xc266, 0 },
+	{ 0xc308, 0xc308, 0xc308, 0 },
+	{ 0xc3a9, 0xc3a9, 0xc3a9, 0 },
+	{ 0xc449, 0xc449, 0xc449, 0 },
+	{ 0xc4e9, 0xc4e9, 0xc4e9, 0 },
+	{ 0xc589, 0xc589, 0xc589, 0 },
+	{ 0xc627, 0xc627, 0xc627, 0 },
+	{ 0xc6c5, 0xc6c5, 0xc6c5, 0 },
+	{ 0xc763, 0xc763, 0xc763, 0 },
+	{ 0xc7ff, 0xc7ff, 0xc7ff, 0 },
+	{ 0xc89b, 0xc89b, 0xc89b, 0 },
+	{ 0xc937, 0xc937, 0xc937, 0 },
+	{ 0xc9d2, 0xc9d2, 0xc9d2, 0 },
+	{ 0xca6c, 0xca6c, 0xca6c, 0 },
+	{ 0xcb06, 0xcb06, 0xcb06, 0 },
+	{ 0xcb9f, 0xcb9f, 0xcb9f, 0 },
+	{ 0xcc37, 0xcc37, 0xcc37, 0 },
+	{ 0xcccf, 0xcccf, 0xcccf, 0 },
+	{ 0xcd66, 0xcd66, 0xcd66, 0 },
+	{ 0xcdfd, 0xcdfd, 0xcdfd, 0 },
+	{ 0xce93, 0xce93, 0xce93, 0 },
+	{ 0xcf29, 0xcf29, 0xcf29, 0 },
+	{ 0xcfbe, 0xcfbe, 0xcfbe, 0 },
+	{ 0xd053, 0xd053, 0xd053, 0 },
+	{ 0xd0e7, 0xd0e7, 0xd0e7, 0 },
+	{ 0xd17a, 0xd17a, 0xd17a, 0 },
+	{ 0xd20d, 0xd20d, 0xd20d, 0 },
+	{ 0xd2a0, 0xd2a0, 0xd2a0, 0 },
+	{ 0xd331, 0xd331, 0xd331, 0 },
+	{ 0xd3c3, 0xd3c3, 0xd3c3, 0 },
+	{ 0xd454, 0xd454, 0xd454, 0 },
+	{ 0xd4e4, 0xd4e4, 0xd4e4, 0 },
+	{ 0xd574, 0xd574, 0xd574, 0 },
+	{ 0xd603, 0xd603, 0xd603, 0 },
+	{ 0xd692, 0xd692, 0xd692, 0 },
+	{ 0xd720, 0xd720, 0xd720, 0 },
+	{ 0xd7ae, 0xd7ae, 0xd7ae, 0 },
+	{ 0xd83c, 0xd83c, 0xd83c, 0 },
+	{ 0xd8c9, 0xd8c9, 0xd8c9, 0 },
+	{ 0xd955, 0xd955, 0xd955, 0 },
+	{ 0xd9e1, 0xd9e1, 0xd9e1, 0 },
+	{ 0xda6d, 0xda6d, 0xda6d, 0 },
+	{ 0xdaf8, 0xdaf8, 0xdaf8, 0 },
+	{ 0xdb83, 0xdb83, 0xdb83, 0 },
+	{ 0xdc0d, 0xdc0d, 0xdc0d, 0 },
+	{ 0xdc97, 0xdc97, 0xdc97, 0 },
+	{ 0xdd20, 0xdd20, 0xdd20, 0 },
+	{ 0xdda9, 0xdda9, 0xdda9, 0 },
+	{ 0xde31, 0xde31, 0xde31, 0 },
+	{ 0xdeb9, 0xdeb9, 0xdeb9, 0 },
+	{ 0xdf41, 0xdf41, 0xdf41, 0 },
+	{ 0xdfc8, 0xdfc8, 0xdfc8, 0 },
+	{ 0xe04f, 0xe04f, 0xe04f, 0 },
+	{ 0xe0d5, 0xe0d5, 0xe0d5, 0 },
+	{ 0xe15b, 0xe15b, 0xe15b, 0 },
+	{ 0xe1e0, 0xe1e0, 0xe1e0, 0 },
+	{ 0xe266, 0xe266, 0xe266, 0 },
+	{ 0xe2ea, 0xe2ea, 0xe2ea, 0 },
+	{ 0xe36f, 0xe36f, 0xe36f, 0 },
+	{ 0xe3f3, 0xe3f3, 0xe3f3, 0 },
+	{ 0xe476, 0xe476, 0xe476, 0 },
+	{ 0xe4f9, 0xe4f9, 0xe4f9, 0 },
+	{ 0xe57c, 0xe57c, 0xe57c, 0 },
+	{ 0xe5fe, 0xe5fe, 0xe5fe, 0 },
+	{ 0xe680, 0xe680, 0xe680, 0 },
+	{ 0xe702, 0xe702, 0xe702, 0 },
+	{ 0xe783, 0xe783, 0xe783, 0 },
+	{ 0xe804, 0xe804, 0xe804, 0 },
+	{ 0xe884, 0xe884, 0xe884, 0 },
+	{ 0xe905, 0xe905, 0xe905, 0 },
+	{ 0xe984, 0xe984, 0xe984, 0 },
+	{ 0xea04, 0xea04, 0xea04, 0 },
+	{ 0xea83, 0xea83, 0xea83, 0 },
+	{ 0xeb02, 0xeb02, 0xeb02, 0 },
+	{ 0xeb80, 0xeb80, 0xeb80, 0 },
+	{ 0xebfe, 0xebfe, 0xebfe, 0 },
+	{ 0xec7b, 0xec7b, 0xec7b, 0 },
+	{ 0xecf9, 0xecf9, 0xecf9, 0 },
+	{ 0xed76, 0xed76, 0xed76, 0 },
+	{ 0xedf2, 0xedf2, 0xedf2, 0 },
+	{ 0xee6f, 0xee6f, 0xee6f, 0 },
+	{ 0xeeeb, 0xeeeb, 0xeeeb, 0 },
+	{ 0xef66, 0xef66, 0xef66, 0 },
+	{ 0xefe2, 0xefe2, 0xefe2, 0 },
+	{ 0xf05d, 0xf05d, 0xf05d, 0 },
+	{ 0xf0d7, 0xf0d7, 0xf0d7, 0 },
+	{ 0xf152, 0xf152, 0xf152, 0 },
+	{ 0xf1cc, 0xf1cc, 0xf1cc, 0 },
+	{ 0xf245, 0xf245, 0xf245, 0 },
+	{ 0xf2bf, 0xf2bf, 0xf2bf, 0 },
+	{ 0xf338, 0xf338, 0xf338, 0 },
+	{ 0xf3b0, 0xf3b0, 0xf3b0, 0 },
+	{ 0xf429, 0xf429, 0xf429, 0 },
+	{ 0xf4a1, 0xf4a1, 0xf4a1, 0 },
+	{ 0xf519, 0xf519, 0xf519, 0 },
+	{ 0xf590, 0xf590, 0xf590, 0 },
+	{ 0xf608, 0xf608, 0xf608, 0 },
+	{ 0xf67e, 0xf67e, 0xf67e, 0 },
+	{ 0xf6f5, 0xf6f5, 0xf6f5, 0 },
+	{ 0xf76b, 0xf76b, 0xf76b, 0 },
+	{ 0xf7e1, 0xf7e1, 0xf7e1, 0 },
+	{ 0xf857, 0xf857, 0xf857, 0 },
+	{ 0xf8cd, 0xf8cd, 0xf8cd, 0 },
+	{ 0xf942, 0xf942, 0xf942, 0 },
+	{ 0xf9b7, 0xf9b7, 0xf9b7, 0 },
+	{ 0xfa2b, 0xfa2b, 0xfa2b, 0 },
+	{ 0xfaa0, 0xfaa0, 0xfaa0, 0 },
+	{ 0xfb14, 0xfb14, 0xfb14, 0 },
+	{ 0xfb88, 0xfb88, 0xfb88, 0 },
+	{ 0xfbfb, 0xfbfb, 0xfbfb, 0 },
+	{ 0xfc6e, 0xfc6e, 0xfc6e, 0 },
+	{ 0xfce1, 0xfce1, 0xfce1, 0 },
+	{ 0xfd54, 0xfd54, 0xfd54, 0 },
+	{ 0xfdc6, 0xfdc6, 0xfdc6, 0 },
+	{ 0xfe39, 0xfe39, 0xfe39, 0 },
+	{ 0xfeaa, 0xfeaa, 0xfeaa, 0 },
+	{ 0xff1c, 0xff1c, 0xff1c, 0 },
+	{ 0xff8d, 0xff8d, 0xff8d, 0 },
+	{ 0xffff, 0xffff, 0xffff, 0 },
+};
+
+const struct vkms_color_lut srgb_inv_eotf = {
+	.base = srgb_inv_array,
+	.lut_length = LUT_SIZE,
+	.channel_value2index_ratio = 0xff00ffll
+};
diff --git a/drivers/gpu/drm/vkms/vkms_luts.h b/drivers/gpu/drm/vkms/vkms_luts.h
new file mode 100644
index 000000000000..925a4a7b84e2
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_luts.h
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#ifndef _VKMS_LUTS_H_
+#define _VKMS_LUTS_H_
+
+#define LUT_SIZE 256
+
+extern const struct vkms_color_lut linear_eotf;
+extern const struct vkms_color_lut srgb_eotf;
+extern const struct vkms_color_lut srgb_inv_eotf;
+
+#endif /* _VKMS_LUTS_H_ */
diff --git a/drivers/gpu/drm/vkms/vkms_plane.c b/drivers/gpu/drm/vkms/vkms_plane.c
index e34f8c7f83c3..b39c4fe471f2 100644
--- a/drivers/gpu/drm/vkms/vkms_plane.c
+++ b/drivers/gpu/drm/vkms/vkms_plane.c
@@ -205,5 +205,7 @@ struct vkms_plane *vkms_plane_init(struct vkms_device *vkmsdev,
 	drm_plane_create_rotation_property(&plane->base, DRM_MODE_ROTATE_0,
 					   DRM_MODE_ROTATE_MASK | DRM_MODE_REFLECT_MASK);
 
+	vkms_initialize_colorops(&plane->base);
+
 	return plane;
 }
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (13 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 14/43] drm/vkms: Add enumerated 1D curve colorop Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-05-07 10:03   ` kernel test robot
  2025-04-30  1:10 ` [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type Alex Hung
                   ` (29 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Two tests are added to VKMS LUT handling:
- linear
- inv_srgb

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v7:
 - Fix checkpatch warnings (Louis Chauvet)
  - Adde a commit messages
  - Fix code styles by adding and removing spaces (new lines, tabs and so on)

 drivers/gpu/drm/vkms/tests/vkms_color_test.c | 39 +++++++++++++++++++-
 drivers/gpu/drm/vkms/vkms_composer.c         | 17 ++-------
 drivers/gpu/drm/vkms/vkms_composer.h         | 13 +++++++
 3 files changed, 55 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_test.c b/drivers/gpu/drm/vkms/tests/vkms_color_test.c
index affca56cac7b..02fbe19c744a 100644
--- a/drivers/gpu/drm/vkms/tests/vkms_color_test.c
+++ b/drivers/gpu/drm/vkms/tests/vkms_color_test.c
@@ -6,6 +6,7 @@
 #include <drm/drm_mode.h>
 #include "../vkms_drv.h"
 #include "../vkms_composer.h"
+#include "../vkms_luts.h"
 
 #define TEST_LUT_SIZE 16
 
@@ -36,7 +37,6 @@ static const struct vkms_color_lut test_linear_lut = {
 	.channel_value2index_ratio = 0xf000fll
 };
 
-
 static void vkms_color_test_get_lut_index(struct kunit *test)
 {
 	s64 lut_index;
@@ -49,6 +49,19 @@ static void vkms_color_test_get_lut_index(struct kunit *test)
 		lut_index = get_lut_index(&test_linear_lut, test_linear_array[i].red);
 		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(lut_index), i);
 	}
+
+	KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&srgb_eotf, 0x0)), 0x0);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_eotf, 0x0)), 0x0);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_eotf, 0x101)), 0x1);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_eotf, 0x202)), 0x2);
+
+	KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&srgb_inv_eotf, 0x0)), 0x0);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_inv_eotf, 0x0)), 0x0);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_inv_eotf, 0x101)), 0x1);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_inv_eotf, 0x202)), 0x2);
+
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_eotf, 0xfefe)), 0xfe);
+	KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&srgb_eotf, 0xffff)), 0xff);
 }
 
 static void vkms_color_test_lerp(struct kunit *test)
@@ -155,9 +168,33 @@ static void vkms_color_test_lerp(struct kunit *test)
 	KUNIT_EXPECT_EQ(test, lerp_u16(0x0, 0x1, 0x80000000), 0x1);
 }
 
+static void vkms_color_test_linear(struct kunit *test)
+{
+	for (int i = 0; i < LUT_SIZE; i++) {
+		int linear = apply_lut_to_channel_value(&linear_eotf, i * 0x101, LUT_RED);
+
+		KUNIT_EXPECT_EQ(test, DIV_ROUND_CLOSEST(linear, 0x101), i);
+	}
+}
+
+static void vkms_color_srgb_inv_srgb(struct kunit *test)
+{
+	u16 srgb, final;
+
+	for (int i = 0; i < LUT_SIZE; i++) {
+		srgb = apply_lut_to_channel_value(&srgb_eotf, i * 0x101, LUT_RED);
+		final = apply_lut_to_channel_value(&srgb_inv_eotf, srgb, LUT_RED);
+
+		KUNIT_EXPECT_GE(test, final / 0x101, i - 1);
+		KUNIT_EXPECT_LE(test, final / 0x101, i + 1);
+	}
+}
+
 static struct kunit_case vkms_color_test_cases[] = {
 	KUNIT_CASE(vkms_color_test_get_lut_index),
 	KUNIT_CASE(vkms_color_test_lerp),
+	KUNIT_CASE(vkms_color_test_linear),
+	KUNIT_CASE(vkms_color_srgb_inv_srgb),
 	{}
 };
 
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 3ab829556db2..72890ffea13a 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -82,19 +82,8 @@ VISIBLE_IF_KUNIT s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel
 }
 EXPORT_SYMBOL_IF_KUNIT(get_lut_index);
 
-/*
- * This enum is related to the positions of the variables inside
- * `struct drm_color_lut`, so the order of both needs to be the same.
- */
-enum lut_channel {
-	LUT_RED = 0,
-	LUT_GREEN,
-	LUT_BLUE,
-	LUT_RESERVED
-};
-
-static u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 channel_value,
-				      enum lut_channel channel)
+VISIBLE_IF_KUNIT u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 channel_value,
+						enum lut_channel channel)
 {
 	s64 lut_index = get_lut_index(lut, channel_value);
 	u16 *floor_lut_value, *ceil_lut_value;
@@ -119,6 +108,8 @@ static u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 chan
 	return lerp_u16(floor_channel_value, ceil_channel_value,
 			lut_index & DRM_FIXED_DECIMAL_MASK);
 }
+EXPORT_SYMBOL_IF_KUNIT(apply_lut_to_channel_value);
+
 
 static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buffer *output_buffer)
 {
diff --git a/drivers/gpu/drm/vkms/vkms_composer.h b/drivers/gpu/drm/vkms/vkms_composer.h
index 9316a053e7d7..67ae09913460 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.h
+++ b/drivers/gpu/drm/vkms/vkms_composer.h
@@ -5,9 +5,22 @@
 
 #include <kunit/visibility.h>
 
+/*
+ * This enum is related to the positions of the variables inside
+ * `struct drm_color_lut`, so the order of both needs to be the same.
+ */
+enum lut_channel {
+	LUT_RED = 0,
+	LUT_GREEN,
+	LUT_BLUE,
+	LUT_RESERVED
+};
+
 #if IS_ENABLED(CONFIG_KUNIT)
 u16 lerp_u16(u16 a, u16 b, s64 t);
 s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value);
+u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 channel_value,
+			       enum lut_channel channel);
 #endif
 
 #endif /* _VKMS_COMPOSER_H_ */
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (14 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-06-16 11:30   ` Borah, Chaitanya Kumar
  2025-04-30  1:10 ` [PATCH V9 17/43] drm/vkms: Use s32 for internal color pipeline precision Alex Hung
                   ` (28 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

This type is used to support a 3x4 matrix in colorops. A 3x4
matrix uses the last column as a "bias" column. Some HW exposes
support for 3x4. The calculation looks like:

 out   matrix    in
 |R|   |0  1  2  3 |   | R |
 |G| = |4  5  6  7 | x | G |
 |B|   |8  9  10 11|   | B |
                       |1.0|

This is also the first colorop where we need a blob property to
program the property. For that we'll introduce a new DATA
property that can be used by all colorop TYPEs requiring a
blob. The way a DATA blob is read depends on the TYPE of
the colorop.

We only create the DATA property for property types that
need it.

Reviewed-by: Simon Ser <contact@emersion.fr>
Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v9:
 - Merge cleanup code for colorop->state->data in drm_colorop_cleanup (Simon Ser)
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v6:
 - take ref for DATA blob in duplicate_state func (Xaver Hugl)

v5:
 - Add function signature for init (Sebastian)
 - Fix kernel-doc

v4:
 - Create helper function for creating 3x4 CTM colorop
 - Fix CTM indexes in comment (Pekka)

 drivers/gpu/drm/drm_atomic.c      |  3 ++
 drivers/gpu/drm/drm_atomic_uapi.c | 30 ++++++++++++++++++++
 drivers/gpu/drm/drm_colorop.c     | 47 +++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         | 24 ++++++++++++++++
 include/uapi/drm/amdgpu_drm.h     |  9 ------
 include/uapi/drm/drm_mode.h       | 32 ++++++++++++++++++++-
 6 files changed, 135 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 7b042c38d50d..809bd856d378 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -793,6 +793,9 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 		drm_printf(p, "\tcurve_1d_type=%s\n",
 			   drm_get_colorop_curve_1d_type_name(state->curve_1d_type));
 		break;
+	case DRM_COLOROP_CTM_3X4:
+		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
+		break;
 	default:
 		break;
 	}
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index c0bcaec249de..be73cb9f502e 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -702,6 +702,31 @@ drm_atomic_plane_get_property(struct drm_plane *plane,
 	return 0;
 }
 
+static int drm_atomic_color_set_data_property(struct drm_colorop *colorop,
+					      struct drm_colorop_state *state,
+					      struct drm_property *property,
+					      uint64_t val)
+{
+	ssize_t elem_size = -1;
+	ssize_t size = -1;
+	bool replaced = false;
+
+	switch (colorop->type) {
+	case DRM_COLOROP_CTM_3X4:
+		size = sizeof(struct drm_color_ctm_3x4);
+		break;
+	default:
+		/* should never get here */
+		return -EINVAL;
+	}
+
+	return drm_property_replace_blob_from_id(colorop->dev,
+						 &state->data,
+						 val,
+						 size,
+						 elem_size,
+						 &replaced);
+}
 
 static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
 		struct drm_colorop_state *state, struct drm_file *file_priv,
@@ -711,6 +736,9 @@ static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
 		state->bypass = val;
 	} else if (property == colorop->curve_1d_type_property) {
 		state->curve_1d_type = val;
+	} else if (property == colorop->data_property) {
+		return drm_atomic_color_set_data_property(colorop, state,
+							  property, val);
 	} else {
 		drm_dbg_atomic(colorop->dev,
 			       "[COLOROP:%d:%d] unknown property [PROP:%d:%s]]\n",
@@ -733,6 +761,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 		*val = state->bypass;
 	} else if (property == colorop->curve_1d_type_property) {
 		*val = state->curve_1d_type;
+	} else if (property == colorop->data_property) {
+		*val = (state->data) ? state->data->base.id : 0;
 	} else {
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 57a8c1063fdd..65351aaa7771 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -64,6 +64,7 @@
 
 static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
+	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
 };
 
 static const char * const colorop_curve_1d_type_names[] = {
@@ -148,6 +149,11 @@ static void drm_colorop_cleanup(struct drm_colorop *colorop)
 	list_del(&colorop->head);
 	config->num_colorop--;
 
+	if (colorop->state && colorop->state->data) {
+		drm_property_blob_put(colorop->state->data);
+		colorop->state->data = NULL;
+	}
+
 	kfree(colorop->state);
 }
 
@@ -238,11 +244,51 @@ int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *
 }
 EXPORT_SYMBOL(drm_plane_colorop_curve_1d_init);
 
+static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_colorop *colorop)
+{
+	struct drm_property *prop;
+
+	/* data */
+	prop = drm_property_create(dev, DRM_MODE_PROP_ATOMIC | DRM_MODE_PROP_BLOB,
+				   "DATA", 0);
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->data_property = prop;
+	drm_object_attach_property(&colorop->base,
+				   colorop->data_property,
+				   0);
+
+	return 0;
+}
+
+int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
+				   struct drm_plane *plane)
+{
+	int ret;
+
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_CTM_3X4);
+	if (ret)
+		return ret;
+
+	ret = drm_colorop_create_data_prop(dev, colorop);
+	if (ret)
+		return ret;
+
+	drm_colorop_reset(colorop);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_colorop_ctm_3x4_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
 	memcpy(state, colorop->state, sizeof(*state));
 
+	if (state->data)
+		drm_property_blob_get(state->data);
+
 	state->bypass = true;
 }
 
@@ -324,6 +370,7 @@ void drm_colorop_reset(struct drm_colorop *colorop)
 
 static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_CURVE] = "1D Curve",
+	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
 };
 
 const char *drm_get_colorop_type_name(enum drm_colorop_type type)
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 8e4d79ba1eff..5f0cddc3922f 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -99,6 +99,17 @@ struct drm_colorop_state {
 	 */
 	enum drm_colorop_curve_1d_type curve_1d_type;
 
+	/**
+	 * @data:
+	 *
+	 * Data blob for any TYPE that requires such a blob. The
+	 * interpretation of the blob is TYPE-specific.
+	 *
+	 * See the &drm_colorop_type documentation for how blob is laid
+	 * out.
+	 */
+	struct drm_property_blob *data;
+
 	/** @state: backpointer to global drm_atomic_state */
 	struct drm_atomic_state *state;
 };
@@ -208,6 +219,17 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @data_property:
+	 *
+	 * blob property for any TYPE that requires a blob of data,
+	 * such as 1DLUT, CTM, 3DLUT, etc.
+	 *
+	 * The way this blob is interpreted depends on the TYPE of
+	 * this
+	 */
+	struct drm_property *data_property;
+
 	/**
 	 * @next_property:
 	 *
@@ -243,6 +265,8 @@ void drm_colorop_pipeline_destroy(struct drm_plane *plane);
 
 int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
 				    struct drm_plane *plane, u64 supported_tfs);
+int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
+				   struct drm_plane *plane);
 
 struct drm_colorop_state *
 drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index 25d5c6e90a99..96fabf9c9827 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -1300,15 +1300,6 @@ struct drm_amdgpu_info_gpuvm_fault {
 #define AMDGPU_FAMILY_GC_11_5_0			150 /* GC 11.5.0 */
 #define AMDGPU_FAMILY_GC_12_0_0			152 /* GC 12.0.0 */
 
-/* FIXME wrong namespace! */
-struct drm_color_ctm_3x4 {
-	/*
-	 * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude
-	 * (not two's complement!) format.
-	 */
-	__u64 matrix[12];
-};
-
 #if defined(__cplusplus)
 }
 #endif
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index ea6d88f683cd..651bdf48b766 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -847,6 +847,20 @@ struct drm_color_ctm {
 	__u64 matrix[9];
 };
 
+struct drm_color_ctm_3x4 {
+	/*
+	 * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude
+	 * (not two's complement!) format.
+	 *
+	 * out   matrix          in
+	 * |R|   |0  1  2  3 |   | R |
+	 * |G| = |4  5  6  7 | x | G |
+	 * |B|   |8  9  10 11|   | B |
+	 *                       |1.0|
+	 */
+	__u64 matrix[12];
+};
+
 struct drm_color_lut {
 	/*
 	 * Values are mapped linearly to 0.0 - 1.0 range, with 0x0 == 0.0 and
@@ -874,7 +888,23 @@ enum drm_colorop_type {
 	 * A 1D curve that is being applied to all color channels. The
 	 * curve is specified via the CURVE_1D_TYPE colorop property.
 	 */
-	DRM_COLOROP_1D_CURVE
+	DRM_COLOROP_1D_CURVE,
+
+	/**
+	 * @DRM_COLOROP_CTM_3X4:
+	 *
+	 * enum string "3x4 Matrix"
+	 *
+	 * A 3x4 matrix. Its values are specified via the
+	 * &drm_color_ctm_3x4 struct provided via the DATA property.
+	 *
+	 * The DATA blob is a float[12]:
+	 * out   matrix          in
+	 * | R |   | 0  1  2  3  |   | R |
+	 * | G | = | 4  5  6  7  | x | G |
+	 * | B |   | 8  9  10 12 |   | B |
+	 */
+	DRM_COLOROP_CTM_3X4,
 };
 
 /**
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 17/43] drm/vkms: Use s32 for internal color pipeline precision
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (15 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 18/43] drm/vkms: add 3x4 matrix in color pipeline Alex Hung
                   ` (27 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Certain operations require us to preserve values below 0.0 and
above 1.0 (0x0 and 0xffff respectively in 16 bpc unorm). One
such operation is a BT709 encoding operation followed by its
decoding operation, or the reverse.

We'll use s32 values as intermediate in and outputs of our
color operations, for the operations where it matters.

For now this won't apply to LUT operations. We might want to
update those to work on s32 as well, but it's unclear how
that should work for unorm LUT definitions. We'll revisit
that once we add LUT + CTM tests.

In order to allow for this we'll also invert the nesting of our
colorop processing loops. We now use the pixel iteration loop
on the outside and the colorop iteration on the inside.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v7:
 - Fix checkpatch warnings
  - Add a commit messages
  - Fix code styles by adding and removing spaces (new lines, tabs and so on)

v6:
 - use clamp_val instead of manual clamping (Louis Chauvet)

v4:
 - Clarify that we're packing 16-bit UNORM into s32, not
   converting values to a different representation (Pekka)

v3:
 - Use new colorop->next pointer

 drivers/gpu/drm/vkms/vkms_composer.c | 27 +++++++++++++++++++++++++--
 drivers/gpu/drm/vkms/vkms_drv.h      |  4 ++++
 2 files changed, 29 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 72890ffea13a..36f0caabb67b 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -128,7 +128,7 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
 	}
 }
 
-static void apply_colorop(struct pixel_argb_u16 *pixel, struct drm_colorop *colorop)
+static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
 {
 	struct drm_colorop_state *colorop_state = colorop->state;
 	struct drm_device *dev = colorop->dev;
@@ -157,9 +157,26 @@ static void apply_colorop(struct pixel_argb_u16 *pixel, struct drm_colorop *colo
 static void pre_blend_color_transform(const struct vkms_plane_state *plane_state,
 				      struct line_buffer *output_buffer)
 {
+	struct pixel_argb_s32 pixel;
+
 	for (size_t x = 0; x < output_buffer->n_pixels; x++) {
 		struct drm_colorop *colorop = plane_state->base.base.color_pipeline;
 
+		/*
+		 * Some operations, such as applying a BT709 encoding matrix,
+		 * followed by a decoding matrix, require that we preserve
+		 * values above 1.0 and below 0.0 until the end of the pipeline.
+		 *
+		 * Pack the 16-bit UNORM values into s32 to give us head-room to
+		 * avoid clipping until we're at the end of the pipeline. Clip
+		 * intentionally at the end of the pipeline before packing
+		 * UNORM values back into u16.
+		 */
+		pixel.a = output_buffer->pixels[x].a;
+		pixel.r = output_buffer->pixels[x].r;
+		pixel.g = output_buffer->pixels[x].g;
+		pixel.b = output_buffer->pixels[x].b;
+
 		while (colorop) {
 			struct drm_colorop_state *colorop_state;
 
@@ -169,10 +186,16 @@ static void pre_blend_color_transform(const struct vkms_plane_state *plane_state
 				return;
 
 			if (!colorop_state->bypass)
-				apply_colorop(&output_buffer->pixels[x], colorop);
+				apply_colorop(&pixel, colorop);
 
 			colorop = colorop->next;
 		}
+
+		/* clamp values */
+		output_buffer->pixels[x].a = clamp_val(pixel.a, 0, 0xffff);
+		output_buffer->pixels[x].r = clamp_val(pixel.r, 0, 0xffff);
+		output_buffer->pixels[x].g = clamp_val(pixel.g, 0, 0xffff);
+		output_buffer->pixels[x].b = clamp_val(pixel.b, 0, 0xffff);
 	}
 }
 
diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
index 2e192b1e5be8..5b340ffbf7d0 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.h
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -45,6 +45,10 @@ struct vkms_frame_info {
 	unsigned int rotation;
 };
 
+struct pixel_argb_s32 {
+	s32 a, r, g, b;
+};
+
 struct pixel_argb_u16 {
 	u16 a, r, g, b;
 };
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 18/43] drm/vkms: add 3x4 matrix in color pipeline
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (16 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 17/43] drm/vkms: Use s32 for internal color pipeline precision Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 19/43] drm/tests: Add a few tests around drm_fixed.h Alex Hung
                   ` (26 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

We add two 3x4 matrices into the VKMS color pipeline. The reason
we're adding matrices is so that we can test that application
of a matrix and its inverse yields an output equal to the input
image.

One complication with the matrix implementation has to do with
the fact that the matrix entries are in signed-magnitude fixed
point, whereas the drm_fixed.h implementation uses 2s-complement.
The latter one is the one that we want for easy addition and
subtraction, so we convert all entries to 2s-complement.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v8:
 - Replace DRM_ERROR with drm_err (Louis Chauvet)

v7:
 - Fix checkpatch warnings
  - Change kzalloc(sizeof(struct drm_colorop) ...) to kzalloc(sizeof(*ops[i]) ...)
  - Change i-1to i - 1
  - Add a new line at EOF

v6:
 - pre-compute colors (Louis Chauvet)
 - round matrix output (Louis Chauvet)

 drivers/gpu/drm/vkms/vkms_colorop.c  | 34 +++++++++++++++++++++++++++-
 drivers/gpu/drm/vkms/vkms_composer.c | 33 +++++++++++++++++++++++++++
 2 files changed, 66 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
index 2fb74b2bd001..70be35804ea9 100644
--- a/drivers/gpu/drm/vkms/vkms_colorop.c
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -12,7 +12,7 @@ static const u64 supported_tfs =
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF);
 
-#define MAX_COLOR_PIPELINE_OPS 2
+#define MAX_COLOR_PIPELINE_OPS 4
 
 static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
 {
@@ -48,6 +48,38 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i - 1], ops[i]);
+
+	i++;
+
+	/* 3rd op: 3x4 matrix */
+	ops[i] = kzalloc(sizeof(*ops[i]), GFP_KERNEL);
+	if (!ops[i]) {
+		drm_err(dev, "KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i - 1], ops[i]);
+
+	i++;
+
+	/* 4th op: 1d curve */
+	ops[i] = kzalloc(sizeof(*ops[i]), GFP_KERNEL);
+	if (!ops[i]) {
+		drm_err(dev, "KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
 	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs);
 	if (ret)
 		goto cleanup;
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 36f0caabb67b..8fe16d26d680 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -128,6 +128,35 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
 	}
 }
 
+static void apply_3x4_matrix(struct pixel_argb_s32 *pixel, const struct drm_color_ctm_3x4 *matrix)
+{
+	s64 rf, gf, bf;
+	s64 r, g, b;
+
+	r = drm_int2fixp(pixel->r);
+	g = drm_int2fixp(pixel->g);
+	b = drm_int2fixp(pixel->b);
+
+	rf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[0]), r) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[1]), g) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[2]), b) +
+	     drm_sm2fixp(matrix->matrix[3]);
+
+	gf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[4]), r) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[5]), g) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[6]), b) +
+	     drm_sm2fixp(matrix->matrix[7]);
+
+	bf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[8]), r) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[9]), g) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[10]), b) +
+	     drm_sm2fixp(matrix->matrix[11]);
+
+	pixel->r = drm_fixp2int_round(rf);
+	pixel->g = drm_fixp2int_round(gf);
+	pixel->b = drm_fixp2int_round(bf);
+}
+
 static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
 {
 	struct drm_colorop_state *colorop_state = colorop->state;
@@ -151,6 +180,10 @@ static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colo
 				      colorop_state->curve_1d_type);
 			break;
 		}
+	} else if (colorop->type == DRM_COLOROP_CTM_3X4) {
+		if (colorop_state->data)
+			apply_3x4_matrix(pixel,
+					 (struct drm_color_ctm_3x4 *) colorop_state->data->data);
 	}
 }
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 19/43] drm/tests: Add a few tests around drm_fixed.h
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (17 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 18/43] drm/vkms: add 3x4 matrix in color pipeline Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 20/43] drm/vkms: Add tests for CTM handling Alex Hung
                   ` (25 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

While working on the CTM implementation of VKMS I had to ascertain
myself of a few assumptions. One of those is whether drm_fixed.h
treats its numbers using signed-magnitude or twos-complement. It is
twos-complement.

In order to make someone else's day easier I am adding the
drm_test_int2fixp test that validates the above assumption.

I am also adding a test for the new sm2fixp function that converts
from a signed-magnitude fixed point to the twos-complement fixed
point.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v7:
 - Fix checkpatch warnings (Louis Chauvet)

v6:
 - add missing MODULE_DESCRIPTION (Jeff Johnson)
 - fix buffer overflow


 drivers/gpu/drm/tests/Makefile        |  3 +-
 drivers/gpu/drm/tests/drm_fixp_test.c | 71 +++++++++++++++++++++++++++
 2 files changed, 73 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c

diff --git a/drivers/gpu/drm/tests/Makefile b/drivers/gpu/drm/tests/Makefile
index 3afd6587df08..91437cf34d92 100644
--- a/drivers/gpu/drm/tests/Makefile
+++ b/drivers/gpu/drm/tests/Makefile
@@ -23,6 +23,7 @@ obj-$(CONFIG_DRM_KUNIT_TEST) += \
 	drm_modes_test.o \
 	drm_plane_helper_test.o \
 	drm_probe_helper_test.o \
-	drm_rect_test.o
+	drm_rect_test.o	\
+	drm_fixp_test.o
 
 CFLAGS_drm_mm_test.o := $(DISABLE_STRUCTLEAK_PLUGIN)
diff --git a/drivers/gpu/drm/tests/drm_fixp_test.c b/drivers/gpu/drm/tests/drm_fixp_test.c
new file mode 100644
index 000000000000..de91177af213
--- /dev/null
+++ b/drivers/gpu/drm/tests/drm_fixp_test.c
@@ -0,0 +1,71 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright 2022 Advanced Micro Devices, Inc.
+ */
+
+#include <kunit/test.h>
+#include <drm/drm_fixed.h>
+
+static void drm_test_sm2fixp(struct kunit *test)
+{
+	KUNIT_EXPECT_EQ(test, 0x7fffffffffffffffll, ((1ull << 63) - 1));
+
+	/* 1 */
+	KUNIT_EXPECT_EQ(test, drm_int2fixp(1), drm_sm2fixp(1ull << DRM_FIXED_POINT));
+
+	/* -1 */
+	KUNIT_EXPECT_EQ(test, drm_int2fixp(-1),
+			drm_sm2fixp((1ull << 63) | (1ull << DRM_FIXED_POINT)));
+
+	/* 0.5 */
+	KUNIT_EXPECT_EQ(test, drm_fixp_from_fraction(1, 2),
+			drm_sm2fixp(1ull << (DRM_FIXED_POINT - 1)));
+
+	/* -0.5 */
+	KUNIT_EXPECT_EQ(test, drm_fixp_from_fraction(-1, 2),
+			drm_sm2fixp((1ull << 63) | (1ull << (DRM_FIXED_POINT - 1))));
+}
+
+static void drm_test_int2fixp(struct kunit *test)
+{
+	/* 1 */
+	KUNIT_EXPECT_EQ(test, 1ll << 32, drm_int2fixp(1));
+
+	/* -1 */
+	KUNIT_EXPECT_EQ(test, -(1ll << 32), drm_int2fixp(-1));
+
+	/* 1 + (-1) = 0 */
+	KUNIT_EXPECT_EQ(test, 0, drm_int2fixp(1) + drm_int2fixp(-1));
+
+	/* 1 / 2 */
+	KUNIT_EXPECT_EQ(test, 1ll << 31, drm_fixp_from_fraction(1, 2));
+
+	/* -0.5 */
+	KUNIT_EXPECT_EQ(test, -(1ll << 31), drm_fixp_from_fraction(-1, 2));
+
+	/* (1 / 2) + (-1) = 0.5 */
+	KUNIT_EXPECT_EQ(test, 1ll << 31, drm_fixp_from_fraction(-1, 2) + drm_int2fixp(1));
+
+	/* (1 / 2) - 1) = 0.5 */
+	KUNIT_EXPECT_EQ(test, -(1ll << 31), drm_fixp_from_fraction(1, 2) + drm_int2fixp(-1));
+
+	/* (1 / 2) - 1) = 0.5 */
+	KUNIT_EXPECT_EQ(test, -(1ll << 31), drm_fixp_from_fraction(1, 2) - drm_int2fixp(1));
+}
+
+static struct kunit_case drm_fixp_tests[] = {
+	KUNIT_CASE(drm_test_int2fixp),
+	KUNIT_CASE(drm_test_sm2fixp),
+	{ }
+};
+
+static struct kunit_suite drm_rect_test_suite = {
+	.name = "drm_fixp",
+	.test_cases = drm_fixp_tests,
+};
+
+kunit_test_suite(drm_rect_test_suite);
+
+MODULE_AUTHOR("AMD");
+MODULE_LICENSE("GPL and additional rights");
+MODULE_DESCRIPTION("Unit tests for drm_fixed.h");
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 20/43] drm/vkms: Add tests for CTM handling
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (18 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 19/43] drm/tests: Add a few tests around drm_fixed.h Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 21/43] drm/colorop: pass plane_color_pipeline client cap to atomic check Alex Hung
                   ` (24 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

A whole slew of tests for CTM handling that greatly helped in
debugging the CTM code. The extent of tests might seem a bit
silly but they're fast and might someday help save someone
else's day when debugging this.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v7:
 - Fix a checkpatch a long-line warning (Louis Chauvet)

v6:
 - update reference values since we're now rounding

v5:
 - Make apply_3x4_matrix static

v4:
 - Comment on origin of bt709_enc matrix (Pekka)
 - Use full opaque alpha (Pekka)
 - Add additional check for Y < 0xffff (Pekka)
 - Remove unused code (Pekka)

 drivers/gpu/drm/vkms/tests/vkms_color_test.c | 250 +++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_composer.c         |   4 +-
 drivers/gpu/drm/vkms/vkms_composer.h         |   1 +
 3 files changed, 254 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_test.c b/drivers/gpu/drm/vkms/tests/vkms_color_test.c
index 02fbe19c744a..a9c3af5025b3 100644
--- a/drivers/gpu/drm/vkms/tests/vkms_color_test.c
+++ b/drivers/gpu/drm/vkms/tests/vkms_color_test.c
@@ -190,11 +190,261 @@ static void vkms_color_srgb_inv_srgb(struct kunit *test)
 	}
 }
 
+#define FIXPT_HALF        (DRM_FIXED_ONE >> 1)
+#define FIXPT_QUARTER     (DRM_FIXED_ONE >> 2)
+
+static const struct drm_color_ctm_3x4 test_matrix_3x4_50_desat = { {
+	FIXPT_HALF, FIXPT_QUARTER, FIXPT_QUARTER, 0,
+	FIXPT_QUARTER, FIXPT_HALF, FIXPT_QUARTER, 0,
+	FIXPT_QUARTER, FIXPT_QUARTER, FIXPT_HALF, 0
+} };
+
+static void vkms_color_ctm_3x4_50_desat(struct kunit *test)
+{
+	struct pixel_argb_s32 ref, out;
+
+	/* full white */
+	ref.a = 0xffff;
+	ref.r = 0xffff;
+	ref.g = 0xffff;
+	ref.b = 0xffff;
+
+	memcpy(&out, &ref, sizeof(out));
+	apply_3x4_matrix(&out, &test_matrix_3x4_50_desat);
+
+	KUNIT_EXPECT_MEMEQ(test, &ref, &out, sizeof(out));
+
+	/* full black */
+	ref.a = 0xffff;
+	ref.r = 0x0;
+	ref.g = 0x0;
+	ref.b = 0x0;
+
+	memcpy(&out, &ref, sizeof(out));
+	apply_3x4_matrix(&out, &test_matrix_3x4_50_desat);
+
+	KUNIT_EXPECT_MEMEQ(test, &ref, &out, sizeof(out));
+
+	/* 50% grey */
+	ref.a = 0xffff;
+	ref.r = 0x8000;
+	ref.g = 0x8000;
+	ref.b = 0x8000;
+
+	memcpy(&out, &ref, sizeof(out));
+	apply_3x4_matrix(&out, &test_matrix_3x4_50_desat);
+
+	KUNIT_EXPECT_MEMEQ(test, &ref, &out, sizeof(out));
+
+	/* full red to 50% desat */
+	ref.a = 0xffff;
+	ref.r = 0x8000;
+	ref.g = 0x4000;
+	ref.b = 0x4000;
+
+	out.a = 0xffff;
+	out.r = 0xffff;
+	out.g = 0x0;
+	out.b = 0x0;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_50_desat);
+
+	KUNIT_EXPECT_MEMEQ(test, &ref, &out, sizeof(out));
+}
+
+/*
+ * BT.709 encoding matrix
+ *
+ * Values printed from within IGT when converting
+ * igt_matrix_3x4_bt709_enc to the fixed-point format expected
+ * by DRM/KMS.
+ */
+static const struct drm_color_ctm_3x4 test_matrix_3x4_bt709_enc = { {
+	0x00000000366cf400ull, 0x00000000b7175900ull, 0x0000000127bb300ull, 0,
+	0x800000001993b3a0ull, 0x800000005609fe80ull, 0x000000006f9db200ull, 0,
+	0x000000009d70a400ull, 0x800000008f011100ull, 0x800000000e6f9330ull, 0
+} };
+
+static void vkms_color_ctm_3x4_bt709(struct kunit *test)
+{
+	struct pixel_argb_s32 out;
+
+	/* full white to bt709 */
+	out.a = 0xffff;
+	out.r = 0xffff;
+	out.g = 0xffff;
+	out.b = 0xffff;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 255 */
+	KUNIT_EXPECT_GT(test, out.r, 0xfe00);
+	KUNIT_EXPECT_LT(test, out.r, 0x10000);
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x0100);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x0100);
+
+	/* full black to bt709 */
+	out.a = 0xffff;
+	out.r = 0x0;
+	out.g = 0x0;
+	out.b = 0x0;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 0 */
+	KUNIT_EXPECT_LT(test, out.r, 0x100);
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x0100);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x0100);
+
+	/* gray to bt709 */
+	out.a = 0xffff;
+	out.r = 0x7fff;
+	out.g = 0x7fff;
+	out.b = 0x7fff;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 127 */
+	KUNIT_EXPECT_GT(test, out.r, 0x7e00);
+	KUNIT_EXPECT_LT(test, out.r, 0x8000);
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x0100);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x0100);
+
+	/* == red 255 - bt709 enc == */
+	out.a = 0xffff;
+	out.r = 0xffff;
+	out.g = 0x0;
+	out.b = 0x0;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 54 */
+	KUNIT_EXPECT_GT(test, out.r, 0x3500);
+	KUNIT_EXPECT_LT(test, out.r, 0x3700);
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x0100);
+
+	/* V 157 */
+	KUNIT_EXPECT_GT(test, out.b, 0x9C00);
+	KUNIT_EXPECT_LT(test, out.b, 0x9E00);
+
+
+	/* == green 255 - bt709 enc == */
+	out.a = 0xffff;
+	out.r = 0x0;
+	out.g = 0xffff;
+	out.b = 0x0;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 182 */
+	KUNIT_EXPECT_GT(test, out.r, 0xB500);
+	KUNIT_EXPECT_LT(test, out.r, 0xB780); /* laxed by half*/
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x0100);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x0100);
+
+	/* == blue 255 - bt709 enc == */
+	out.a = 0xffff;
+	out.r = 0x0;
+	out.g = 0x0;
+	out.b = 0xffff;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 18 */
+	KUNIT_EXPECT_GT(test, out.r, 0x1100);
+	KUNIT_EXPECT_LT(test, out.r, 0x1300);
+
+	/* U 111 */
+	KUNIT_EXPECT_GT(test, out.g, 0x6E00);
+	KUNIT_EXPECT_LT(test, out.g, 0x7000);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x0100);
+
+	/* == red 140 - bt709 enc == */
+	out.a = 0xffff;
+	out.r = 0x8c8c;
+	out.g = 0x0;
+	out.b = 0x0;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 30 */
+	KUNIT_EXPECT_GT(test, out.r, 0x1D00);
+	KUNIT_EXPECT_LT(test, out.r, 0x1F00);
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x100);
+
+	/* V 87 */
+	KUNIT_EXPECT_GT(test, out.b, 0x5600);
+	KUNIT_EXPECT_LT(test, out.b, 0x5800);
+
+	/* == green 140 - bt709 enc == */
+	out.a = 0xffff;
+	out.r = 0x0;
+	out.g = 0x8c8c;
+	out.b = 0x0;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 30 */
+	KUNIT_EXPECT_GT(test, out.r, 0x6400);
+	KUNIT_EXPECT_LT(test, out.r, 0x6600);
+
+	/* U 0 */
+	KUNIT_EXPECT_LT(test, out.g, 0x100);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x100);
+
+
+	/* == blue 140 - bt709 enc == */
+	out.a = 0xffff;
+	out.r = 0x0;
+	out.g = 0x0;
+	out.b = 0x8c8c;
+
+	apply_3x4_matrix(&out, &test_matrix_3x4_bt709_enc);
+
+	/* Y 30 */
+	KUNIT_EXPECT_GT(test, out.r, 0x900);
+	KUNIT_EXPECT_LT(test, out.r, 0xB00);
+
+	/* U 61 */
+	KUNIT_EXPECT_GT(test, out.g, 0x3C00);
+	KUNIT_EXPECT_LT(test, out.g, 0x3E00);
+
+	/* V 0 */
+	KUNIT_EXPECT_LT(test, out.b, 0x100);
+
+}
+
 static struct kunit_case vkms_color_test_cases[] = {
 	KUNIT_CASE(vkms_color_test_get_lut_index),
 	KUNIT_CASE(vkms_color_test_lerp),
 	KUNIT_CASE(vkms_color_test_linear),
 	KUNIT_CASE(vkms_color_srgb_inv_srgb),
+	KUNIT_CASE(vkms_color_ctm_3x4_50_desat),
+	KUNIT_CASE(vkms_color_ctm_3x4_bt709),
 	{}
 };
 
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 8fe16d26d680..54fd00b49360 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -128,7 +128,8 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
 	}
 }
 
-static void apply_3x4_matrix(struct pixel_argb_s32 *pixel, const struct drm_color_ctm_3x4 *matrix)
+VISIBLE_IF_KUNIT void apply_3x4_matrix(struct pixel_argb_s32 *pixel,
+				       const struct drm_color_ctm_3x4 *matrix)
 {
 	s64 rf, gf, bf;
 	s64 r, g, b;
@@ -156,6 +157,7 @@ static void apply_3x4_matrix(struct pixel_argb_s32 *pixel, const struct drm_colo
 	pixel->g = drm_fixp2int_round(gf);
 	pixel->b = drm_fixp2int_round(bf);
 }
+EXPORT_SYMBOL_IF_KUNIT(apply_3x4_matrix);
 
 static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
 {
diff --git a/drivers/gpu/drm/vkms/vkms_composer.h b/drivers/gpu/drm/vkms/vkms_composer.h
index 67ae09913460..afef65a5c3ba 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.h
+++ b/drivers/gpu/drm/vkms/vkms_composer.h
@@ -21,6 +21,7 @@ u16 lerp_u16(u16 a, u16 b, s64 t);
 s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value);
 u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 channel_value,
 			       enum lut_channel channel);
+void apply_3x4_matrix(struct pixel_argb_s32 *pixel, const struct drm_color_ctm_3x4 *matrix);
 #endif
 
 #endif /* _VKMS_COMPOSER_H_ */
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 21/43] drm/colorop: pass plane_color_pipeline client cap to atomic check
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (19 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 20/43] drm/vkms: Add tests for CTM handling Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 22/43] drm/colorop: define a new macro for_each_new_colorop_in_state Alex Hung
                   ` (23 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Drivers will need to know whether an atomic check/commit
originated from a client with DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
so they can ignore deprecated properties, like COLOR_ENCODING
and COLOR_RANGE.

Pass the plane_color_pipeline bit to drm_atomic_state.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v5:
 - Fix kernel docs

 drivers/gpu/drm/drm_atomic_uapi.c |  1 +
 include/drm/drm_atomic.h          | 18 ++++++++++++++++++
 2 files changed, 19 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index be73cb9f502e..0de0fdd8af5b 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -1603,6 +1603,7 @@ int drm_mode_atomic_ioctl(struct drm_device *dev,
 	drm_modeset_acquire_init(&ctx, DRM_MODESET_ACQUIRE_INTERRUPTIBLE);
 	state->acquire_ctx = &ctx;
 	state->allow_modeset = !!(arg->flags & DRM_MODE_ATOMIC_ALLOW_MODESET);
+	state->plane_color_pipeline = file_priv->plane_color_pipeline;
 
 retry:
 	copied_objs = 0;
diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index 7904a38caa77..4a30ed8ab1a8 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -464,6 +464,24 @@ struct drm_atomic_state {
 	 */
 	bool duplicated : 1;
 
+	/**
+	 * @plane_color_pipeline:
+	 *
+	 * Indicates whether this atomic state originated with a client that
+	 * set the DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE.
+	 *
+	 * Drivers and helper functions should use this to ignore legacy
+	 * properties that are incompatible with the drm_plane COLOR_PIPELINE
+	 * behavior, such as:
+	 *
+	 *  - COLOR_RANGE
+	 *  - COLOR_ENCODING
+	 *
+	 * or any other driver-specific properties that might affect pixel
+	 * values.
+	 */
+	bool plane_color_pipeline : 1;
+
 	/**
 	 * @colorops:
 	 *
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 22/43] drm/colorop: define a new macro for_each_new_colorop_in_state
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (20 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 21/43] drm/colorop: pass plane_color_pipeline client cap to atomic check Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-05-12 18:32   ` Melissa Wen
  2025-04-30  1:10 ` [PATCH V9 23/43] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Alex Hung
                   ` (22 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Create a new macro for_each_new_colorop_in_state to access new
drm_colorop_state updated from uapi.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
 include/drm/drm_atomic.h | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index 4a30ed8ab1a8..4b3bd459a1eb 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -1069,6 +1069,26 @@ void drm_state_dump(struct drm_device *dev, struct drm_printer *p);
 			      (new_colorop_state) = (__state)->colorops[__i].new_state, 1))
 
 
+/**
+ * for_each_new_colorop_in_state - iterate over all colorops in an atomic update
+ * @__state: &struct drm_atomic_state pointer
+ * @colorop: &struct drm_colorop iteration cursor
+ * @new_colorop_state: &struct drm_colorop_state iteration cursor for the new state
+ * @__i: int iteration cursor, for macro-internal use
+ *
+ * This iterates over all colorops in an atomic update, tracking new state. This is
+ * useful is useful in places where the state delta needs to be considered, for
+ * example in atomic check functions.
+ */
+#define for_each_new_colorop_in_state(__state, colorop, new_colorop_state, __i) \
+	for ((__i) = 0;							\
+	     (__i) < (__state)->dev->mode_config.num_colorop;	\
+	     (__i)++)							\
+		for_each_if ((__state)->colorops[__i].ptr &&		\
+			     ((colorop) = (__state)->colorops[__i].ptr,	\
+			      (void)(colorop) /* Only to avoid unused-but-set-variable warning */, \
+			      (new_colorop_state) = (__state)->colorops[__i].new_state, 1))
+
 /**
  * for_each_oldnew_plane_in_state - iterate over all planes in an atomic update
  * @__state: &struct drm_atomic_state pointer
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 23/43] drm/amd/display: Ignore deprecated props when plane_color_pipeline set
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (21 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 22/43] drm/colorop: define a new macro for_each_new_colorop_in_state Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 24/43] drm/amd/display: Add bypass COLOR PIPELINE Alex Hung
                   ` (21 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

When the plane_color_pipeline bit is set we should ignore
deprecated properties, such as COLOR_RANGE and COLOR_ENCODING.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 9fed4471405f..4d3f24313905 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -5739,6 +5739,10 @@ fill_plane_color_attributes(const struct drm_plane_state *plane_state,
 
 	*color_space = COLOR_SPACE_SRGB;
 
+	/* Ignore properties when DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE is set */
+	if (plane_state->state && plane_state->state->plane_color_pipeline)
+		return 0;
+
 	/* DRM color properties only affect non-RGB formats. */
 	if (format < SURFACE_PIXEL_FORMAT_VIDEO_BEGIN)
 		return 0;
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 24/43] drm/amd/display: Add bypass COLOR PIPELINE
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (22 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 23/43] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-05-12 18:41   ` Melissa Wen
  2025-04-30  1:10 ` [PATCH V9 25/43] drm/amd/display: Skip color pipeline initialization for cursor plane Alex Hung
                   ` (20 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Add the default Bypass pipeline and ensure it passes the
kms_colorop test plane-XR30-XR30-bypass.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
index 3e0f45f1711c..fb6f07603050 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
@@ -1784,6 +1784,20 @@ dm_atomic_plane_get_property(struct drm_plane *plane,
 }
 #endif
 
+#define MAX_COLOR_PIPELINES 5
+
+static int
+dm_plane_init_colorops(struct drm_plane *plane)
+{
+	struct drm_prop_enum_list pipelines[MAX_COLOR_PIPELINES];
+	int len = 0;
+
+	/* Create COLOR_PIPELINE property and attach */
+	drm_plane_create_color_pipeline_property(plane, pipelines, len);
+
+	return 0;
+}
+
 static const struct drm_plane_funcs dm_plane_funcs = {
 	.update_plane	= drm_atomic_helper_update_plane,
 	.disable_plane	= drm_atomic_helper_disable_plane,
@@ -1890,7 +1904,12 @@ int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm,
 
 #ifdef AMD_PRIVATE_COLOR
 	dm_atomic_plane_attach_color_mgmt_properties(dm, plane);
+#else
+	res = dm_plane_init_colorops(plane);
+	if (res)
+		return res;
 #endif
+
 	/* Create (reset) the plane state */
 	if (plane->funcs->reset)
 		plane->funcs->reset(plane);
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 25/43] drm/amd/display: Skip color pipeline initialization for cursor plane
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (23 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 24/43] drm/amd/display: Add bypass COLOR PIPELINE Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block Alex Hung
                   ` (19 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

cursor plane does not need to have color pipeline.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v7:
 - Add a commit messages

 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
index fb6f07603050..feb4b99bf629 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
@@ -1792,6 +1792,9 @@ dm_plane_init_colorops(struct drm_plane *plane)
 	struct drm_prop_enum_list pipelines[MAX_COLOR_PIPELINES];
 	int len = 0;
 
+	if (plane->type == DRM_PLANE_TYPE_CURSOR)
+		return 0;
+
 	/* Create COLOR_PIPELINE property and attach */
 	drm_plane_create_color_pipeline_property(plane, pipelines, len);
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (24 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 25/43] drm/amd/display: Skip color pipeline initialization for cursor plane Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-05-12 18:50   ` Melissa Wen
  2025-04-30  1:10 ` [PATCH V9 27/43] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Alex Hung
                   ` (18 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Expose one 1D curve colorop with support for
DRM_COLOROP_1D_CURVE_SRGB_EOTF and program HW to perform
the sRGB transform when the colorop is not in bypass.

With this change the following IGT test passes:
kms_colorop --run plane-XR30-XR30-srgb_eotf

The color pipeline now consists of a single colorop:
1. 1D curve colorop w/ sRGB EOTF

Signed-off-by: Alex Hung <alex.hung@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)
 - Update replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)

v8:
 - Fix incorrect && by || in __set_colorop_in_tf_1d_curve (Leo Li)

v7:
 - Fix checkpatch warnings
  - Change switch "{ }" position
  - Delete double ";"
  - Delete "{ }" for single-line if-statement
  - Add a new line at EOF
  - Change SPDX-License-Identifier: GPL-2.0+ from // to /* */

v6:
 - cleanup if colorop alloc or init fails

 .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 86 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 69 +++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 ++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
 5 files changed, 201 insertions(+), 1 deletion(-)
 create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
 create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
index ab2a97e354da..46158d67ab12 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
@@ -38,7 +38,8 @@ AMDGPUDM = \
 	amdgpu_dm_pp_smu.o \
 	amdgpu_dm_psr.o \
 	amdgpu_dm_replay.o \
-	amdgpu_dm_wb.o
+	amdgpu_dm_wb.o \
+	amdgpu_dm_colorop.o
 
 ifdef CONFIG_DRM_AMD_DC_FP
 AMDGPUDM += dc_fpu.o
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index ebabfe3a512f..0b513ab5050f 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -668,6 +668,18 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf)
 	}
 }
 
+static enum dc_transfer_func_predefined
+amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
+{
+	switch (tf) {
+	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
+	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
+		return TRANSFER_FUNCTION_SRGB;
+	default:
+		return TRANSFER_FUNCTION_LINEAR;
+	}
+}
+
 static void __to_dc_lut3d_color(struct dc_rgb *rgb,
 				const struct drm_color_lut lut,
 				int bit_precision)
@@ -1137,6 +1149,59 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state,
 	return 0;
 }
 
+static int
+__set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
+		       struct drm_colorop_state *colorop_state)
+{
+	struct dc_transfer_func *tf = &dc_plane_state->in_transfer_func;
+	struct drm_colorop *colorop = colorop_state->colorop;
+	struct drm_device *drm = colorop->dev;
+
+	if (colorop->type != DRM_COLOROP_1D_CURVE ||
+	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
+		return -EINVAL;
+
+	if (colorop_state->bypass) {
+		tf->type = TF_TYPE_BYPASS;
+		tf->tf = TRANSFER_FUNCTION_LINEAR;
+		return 0;
+	}
+
+	drm_dbg(drm, "Degamma colorop with ID: %d\n", colorop->base.id);
+
+	tf->type = TF_TYPE_PREDEFINED;
+	tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
+
+	return 0;
+}
+
+static int
+__set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
+			       struct dc_plane_state *dc_plane_state,
+			       struct drm_colorop *colorop)
+{
+	struct drm_colorop *old_colorop;
+	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+	struct drm_atomic_state *state = plane_state->state;
+	int i = 0;
+
+	old_colorop = colorop;
+
+	/* 1st op: 1d curve - degamma */
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
+
+	if (!colorop_state)
+		return -EINVAL;
+
+	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);
+}
+
 static int
 amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state,
 				     struct dc_plane_state *dc_plane_state)
@@ -1187,6 +1252,24 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state,
 	return 0;
 }
 
+static int
+amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
+		       struct dc_plane_state *dc_plane_state)
+{
+	struct drm_colorop *colorop = plane_state->color_pipeline;
+	int ret;
+
+	/* 1D Curve - DEGAM TF */
+	if (!colorop)
+		return -EINVAL;
+
+	ret = __set_dm_plane_colorop_degamma(plane_state, dc_plane_state, colorop);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
 /**
  * amdgpu_dm_update_plane_color_mgmt: Maps DRM color management to DC plane.
  * @crtc: amdgpu_dm crtc state
@@ -1283,5 +1366,8 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc,
 		dc_plane_state->input_csc_color_matrix.enable_adjustment = false;
 	}
 
+	if (!amdgpu_dm_plane_set_colorop_properties(plane_state, dc_plane_state))
+		return 0;
+
 	return amdgpu_dm_plane_set_color_properties(plane_state, dc_plane_state);
 }
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
new file mode 100644
index 000000000000..bc9980aad7d8
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -0,0 +1,69 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright 2023 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include <drm/drm_print.h>
+#include <drm/drm_plane.h>
+#include <drm/drm_property.h>
+#include <drm/drm_colorop.h>
+
+#include "amdgpu_dm_colorop.h"
+
+const u64 amdgpu_dm_supported_degam_tfs =
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF);
+
+#define MAX_COLOR_PIPELINE_OPS 10
+
+int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
+{
+	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
+	struct drm_device *dev = plane->dev;
+	int ret;
+	int i = 0;
+
+	memset(ops, 0, sizeof(ops));
+
+	/* 1D curve - DEGAM TF */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_degam_tfs);
+	if (ret)
+		goto cleanup;
+
+	list->type = ops[i]->base.id;
+	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", ops[i]->base.id);
+
+	return 0;
+
+cleanup:
+	drm_colorop_pipeline_destroy(plane);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
new file mode 100644
index 000000000000..3324e2a66079
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright 2023 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __AMDGPU_DM_COLOROP_H__
+#define __AMDGPU_DM_COLOROP_H__
+
+extern const u64 amdgpu_dm_supported_degam_tfs;
+
+int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list);
+
+#endif /* __AMDGPU_DM_COLOROP_H__*/
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
index feb4b99bf629..742ed56d1e58 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
@@ -37,6 +37,7 @@
 #include "amdgpu_display.h"
 #include "amdgpu_dm_trace.h"
 #include "amdgpu_dm_plane.h"
+#include "amdgpu_dm_colorop.h"
 #include "gc/gc_11_0_0_offset.h"
 #include "gc/gc_11_0_0_sh_mask.h"
 
@@ -1791,10 +1792,19 @@ dm_plane_init_colorops(struct drm_plane *plane)
 {
 	struct drm_prop_enum_list pipelines[MAX_COLOR_PIPELINES];
 	int len = 0;
+	int ret;
 
 	if (plane->type == DRM_PLANE_TYPE_CURSOR)
 		return 0;
 
+	/* initialize default pipeline */
+	ret = amdgpu_dm_initialize_default_pipeline(plane, &(pipelines[len]));
+	if (ret) {
+		DRM_ERROR("Failed to create color pipeline for plane %d: %d\n", plane->base.id, ret);
+		return ret;
+	}
+	len++;
+
 	/* Create COLOR_PIPELINE property and attach */
 	drm_plane_create_color_pipeline_property(plane, pipelines, len);
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 27/43] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (25 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 28/43] drm/amd/display: Add support for sRGB EOTF in BLND block Alex Hung
                   ` (17 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Expose a 2nd curve colorop with support for
DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF and program HW to
perform the sRGB Inverse EOTF on the shaper block
when the colorop is not in bypass.

With this change the follow IGT tests pass:
kms_colorop --run plane-XR30-XR30-srgb_inv_eotf
kms_colorop --run plane-XR30-XR30-srgb_eotf-srgb_inv_eotf

The color pipeline now consists of the following colorops:
1. 1D curve colorop w/ sRGB EOTF support
2. 1D curve colorop w/ sRGB Inverse EOTF support

Signed-off-by: Alex Hung <alex.hung@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v6:
 - don't pass uninitialized variable into __set_output_tf

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 74 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 19 +++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  1 +
 3 files changed, 94 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 0b513ab5050f..4662d88885de 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1202,6 +1202,68 @@ __set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
 	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);
 }
 
+static int
+__set_colorop_in_shaper_1d_curve(struct dc_plane_state *dc_plane_state,
+		       struct drm_colorop_state *colorop_state)
+{
+	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
+	struct drm_colorop *colorop = colorop_state->colorop;
+	struct drm_device *drm = colorop->dev;
+
+	if (colorop->type != DRM_COLOROP_1D_CURVE &&
+	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF)
+		return -EINVAL;
+
+	if (colorop_state->bypass) {
+		tf->type = TF_TYPE_BYPASS;
+		tf->tf = TRANSFER_FUNCTION_LINEAR;
+		return 0;
+	}
+
+	drm_dbg(drm, "Shaper colorop with ID: %d\n", colorop->base.id);
+
+	if (colorop->type == DRM_COLOROP_1D_CURVE) {
+		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+		tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
+		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+		return __set_output_tf(tf, 0, 0, false);
+	}
+
+	return -EINVAL;
+}
+
+static int
+__set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
+			      struct dc_plane_state *dc_plane_state,
+			      struct drm_colorop *colorop)
+{
+	struct drm_colorop *old_colorop;
+	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+	struct drm_atomic_state *state = plane_state->state;
+	int i = 0;
+
+	old_colorop = colorop;
+
+	/* 2nd op: 1d curve - shaper */
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+
+		if (new_colorop_state->colorop == old_colorop) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
+
+	if (!colorop_state)
+		return -EINVAL;
+
+	return __set_colorop_in_shaper_1d_curve(dc_plane_state, colorop_state);
+}
+
 static int
 amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state,
 				     struct dc_plane_state *dc_plane_state)
@@ -1257,6 +1319,7 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 		       struct dc_plane_state *dc_plane_state)
 {
 	struct drm_colorop *colorop = plane_state->color_pipeline;
+	struct drm_device *dev = plane_state->plane->dev;
 	int ret;
 
 	/* 1D Curve - DEGAM TF */
@@ -1267,6 +1330,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
+	/* 1D Curve - SHAPER TF */
+	colorop = colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no Shaper TF colorop found\n");
+		return -EINVAL;
+	}
+
+	ret = __set_dm_plane_colorop_shaper(plane_state, dc_plane_state, colorop);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index bc9980aad7d8..5597a693ed47 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -34,6 +34,9 @@
 const u64 amdgpu_dm_supported_degam_tfs =
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF);
 
+const u64 amdgpu_dm_supported_shaper_tfs =
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF);
+
 #define MAX_COLOR_PIPELINE_OPS 10
 
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
@@ -60,6 +63,22 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 	list->type = ops[i]->base.id;
 	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", ops[i]->base.id);
 
+	i++;
+
+	/* 1D curve - SHAPER TF */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_shaper_tfs);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
+
 	return 0;
 
 cleanup:
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
index 3324e2a66079..71cd27994528 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
@@ -28,6 +28,7 @@
 #define __AMDGPU_DM_COLOROP_H__
 
 extern const u64 amdgpu_dm_supported_degam_tfs;
+extern const u64 amdgpu_dm_supported_shaper_tfs;
 
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list);
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 28/43] drm/amd/display: Add support for sRGB EOTF in BLND block
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (26 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 27/43] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:10 ` [PATCH V9 29/43] drm/colorop: Add PQ 125 EOTF and its inverse Alex Hung
                   ` (16 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Expose a 3rd 1D curve colorop, with support for
DRM_COLOROP_1D_CURVE_SRGB_EOTF and program the BLND block
to perform the sRGB transform when the colorop is not in
bypass

With this change the following IGT test passes:
kms_colorop --run plane-XR30-XR30-srgb_eotf-srgb_inv_eotf-srgb_eotf

The color pipeline now consists of the following colorops:
1. 1D curve colorop w/ sRGB EOTF support
2. 1D curve colorop w/ sRGB Inverse EOTF support
3. 1D curve colorop w/ sRGB EOTF support

Signed-off-by: Alex Hung <alex.hung@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v7:
 - Initialized uint32_t blend_size = 0 by default (kernel test robot)

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 77 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 19 +++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  1 +
 3 files changed, 97 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 4662d88885de..7ce95ddf3851 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1264,6 +1264,72 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 	return __set_colorop_in_shaper_1d_curve(dc_plane_state, colorop_state);
 }
 
+
+static int
+__set_colorop_1d_curve_blend_tf_lut(struct dc_plane_state *dc_plane_state,
+				  struct drm_colorop_state *colorop_state)
+{
+
+	struct dc_transfer_func *tf = &dc_plane_state->blend_tf;
+	struct drm_colorop *colorop = colorop_state->colorop;
+	struct drm_device *drm = colorop->dev;
+	const struct drm_color_lut *blend_lut;
+	uint32_t blend_size = 0;
+
+	if (colorop->type != DRM_COLOROP_1D_CURVE &&
+	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
+		return -EINVAL;
+
+	if (colorop_state->bypass) {
+		tf->type = TF_TYPE_BYPASS;
+		tf->tf = TRANSFER_FUNCTION_LINEAR;
+		return 0;
+	}
+
+	drm_dbg(drm, "Blend colorop with ID: %d\n", colorop->base.id);
+
+	if (colorop->type == DRM_COLOROP_1D_CURVE) {
+		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+		tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
+		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+		return __set_input_tf(NULL, tf, blend_lut, blend_size);
+	}
+
+	return -EINVAL;
+}
+
+static int
+__set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
+			     struct dc_plane_state *dc_plane_state,
+			     struct drm_colorop *colorop)
+{
+	struct drm_colorop *old_colorop;
+	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+	struct drm_atomic_state *state = plane_state->state;
+	int i = 0;
+
+	old_colorop = colorop;
+
+	/* 3nd op: 1d curve - blend */
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+
+		if (new_colorop_state->colorop == old_colorop) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
+
+	if (!colorop_state)
+		return -EINVAL;
+
+	return __set_colorop_1d_curve_blend_tf_lut(dc_plane_state, colorop_state);
+}
+
 static int
 amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state,
 				     struct dc_plane_state *dc_plane_state)
@@ -1341,6 +1407,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
+	/* 1D Curve - BLND TF */
+	colorop = colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no Blend TF colorop found\n");
+		return -EINVAL;
+	}
+
+	ret = __set_dm_plane_colorop_blend(plane_state, dc_plane_state, colorop);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 5597a693ed47..5ea24da088b6 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -37,6 +37,9 @@ const u64 amdgpu_dm_supported_degam_tfs =
 const u64 amdgpu_dm_supported_shaper_tfs =
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF);
 
+const u64 amdgpu_dm_supported_blnd_tfs =
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF);
+
 #define MAX_COLOR_PIPELINE_OPS 10
 
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
@@ -79,6 +82,22 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	drm_colorop_set_next_property(ops[i-1], ops[i]);
 
+	i++;
+
+	/* 1D curve - BLND TF */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_blnd_tfs);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
+
 	return 0;
 
 cleanup:
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
index 71cd27994528..2e1617ffc8ee 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
@@ -29,6 +29,7 @@
 
 extern const u64 amdgpu_dm_supported_degam_tfs;
 extern const u64 amdgpu_dm_supported_shaper_tfs;
+extern const u64 amdgpu_dm_supported_blnd_tfs;
 
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list);
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 29/43] drm/colorop: Add PQ 125 EOTF and its inverse
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (27 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 28/43] drm/amd/display: Add support for sRGB EOTF in BLND block Alex Hung
@ 2025-04-30  1:10 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 30/43] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Alex Hung
                   ` (15 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

The PQ function defines a mapping of code values to nits (cd/m^2).
The max code value maps to 10,000 nits.

Windows DWM's canonical composition color space (CCCS)  defaults
to composing SDR contents to 80 nits and uses a float value of
1.0 to represent this. For this reason AMD HW hard-codes a PQ
function that is scaled by 125, yielding 80 nit PQ values for
1.0 and 10,000 nits at 125.0.

This patch introduces this scaled PQ EOTF and its inverse as
1D curve types.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
 drivers/gpu/drm/drm_colorop.c |  2 ++
 include/drm/drm_colorop.h     | 24 ++++++++++++++++++++++++
 2 files changed, 26 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 65351aaa7771..9ddb3dccaf72 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -70,6 +70,8 @@ static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 static const char * const colorop_curve_1d_type_names[] = {
 	[DRM_COLOROP_1D_CURVE_SRGB_EOTF] = "sRGB EOTF",
 	[DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF] = "sRGB Inverse EOTF",
+	[DRM_COLOROP_1D_CURVE_PQ_125_EOTF] = "PQ 125 EOTF",
+	[DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF] = "PQ 125 Inverse EOTF",
 };
 
 
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 5f0cddc3922f..40d70d4a34a0 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -59,6 +59,30 @@ enum drm_colorop_curve_1d_type {
 	 */
 	DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
 
+	/**
+	 * @DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
+	 *
+	 * enum string "PQ 125 EOTF"
+	 *
+	 * The PQ transfer function, scaled by 125.0f, so that 10,000
+	 * nits correspond to 125.0f.
+	 *
+	 * Transfer characteristics of the PQ function as defined by
+	 * SMPTE ST 2084 (2014) for 10-, 12-, 14-, and 16-bit systems
+	 * and Rec. ITU-R BT.2100-2 perceptual quantization (PQ) system,
+	 * represented by H.273 TransferCharacteristics code point 16.
+	 */
+	DRM_COLOROP_1D_CURVE_PQ_125_EOTF,
+
+	/**
+	 * @DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF:
+	 *
+	 * enum string "PQ 125 Inverse EOTF"
+	 *
+	 * The inverse of DRM_COLOROP_1D_CURVE_PQ_125_EOTF.
+	 */
+	DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF,
+
 	/**
 	 * @DRM_COLOROP_1D_CURVE_COUNT:
 	 *
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 30/43] drm/amd/display: Enable support for PQ 125 EOTF and Inverse
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (28 preceding siblings ...)
  2025-04-30  1:10 ` [PATCH V9 29/43] drm/colorop: Add PQ 125 EOTF and its inverse Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 31/43] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Alex Hung
                   ` (14 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

This patchset enables support for the PQ_125 EOTF and its inverse
on all existing plane 1D curve colorops, i.e., on DEGAM, SHAPER,
and BLND blocks.

With this patchset the following IGT subtests are passing:
kms_colorop --run plane-XR30-XR30-pq_125_eotf
kms_colorop --run plane-XR30-XR30-pq_125_inv_eotf
kms_colorop --run plane-XR30-XR30-pq_125_eotf-pq_125_inv_eotf
kms_colorop --run plane-XR30-XR30-pq_125_eotf-pq_125_inv_eotf-pq_125_eotf

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
v8: 
 - Move BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) in amdgpu_dm_supported_blnd_tfs
   from "drm/amd/display: Add support for BT.709 and BT.2020 TFs" (Leo Li)

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 20 +++++++++++++------
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  9 ++++++---
 2 files changed, 20 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 7ce95ddf3851..63044e0296cb 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -25,6 +25,7 @@
 #include "amdgpu.h"
 #include "amdgpu_mode.h"
 #include "amdgpu_dm.h"
+#include "amdgpu_dm_colorop.h"
 #include "dc.h"
 #include "modules/color/color_gamma.h"
 #include "basics/conversion.h"
@@ -675,6 +676,9 @@ amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
 	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
 	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
 		return TRANSFER_FUNCTION_SRGB;
+	case DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
+	case DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF:
+		return TRANSFER_FUNCTION_PQ;
 	default:
 		return TRANSFER_FUNCTION_LINEAR;
 	}
@@ -1157,8 +1161,10 @@ __set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
 	struct drm_colorop *colorop = colorop_state->colorop;
 	struct drm_device *drm = colorop->dev;
 
-	if (colorop->type != DRM_COLOROP_1D_CURVE ||
-	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
+	if (colorop->type != DRM_COLOROP_1D_CURVE)
+		return -EINVAL;
+
+	if (!(BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_degam_tfs))
 		return -EINVAL;
 
 	if (colorop_state->bypass) {
@@ -1190,7 +1196,7 @@ __set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
 	/* 1st op: 1d curve - degamma */
 	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
 		if (new_colorop_state->colorop == old_colorop &&
-		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
+		    (BIT(new_colorop_state->curve_1d_type) & amdgpu_dm_supported_degam_tfs)) {
 			colorop_state = new_colorop_state;
 			break;
 		}
@@ -1210,8 +1216,10 @@ __set_colorop_in_shaper_1d_curve(struct dc_plane_state *dc_plane_state,
 	struct drm_colorop *colorop = colorop_state->colorop;
 	struct drm_device *drm = colorop->dev;
 
-	if (colorop->type != DRM_COLOROP_1D_CURVE &&
-	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF)
+	if (colorop->type != DRM_COLOROP_1D_CURVE)
+		return -EINVAL;
+
+	if (!(BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_shaper_tfs))
 		return -EINVAL;
 
 	if (colorop_state->bypass) {
@@ -1247,7 +1255,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 	/* 2nd op: 1d curve - shaper */
 	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
 		if (new_colorop_state->colorop == old_colorop &&
-		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF) {
+		    (BIT(new_colorop_state->curve_1d_type) & amdgpu_dm_supported_shaper_tfs)) {
 			colorop_state = new_colorop_state;
 			break;
 		}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 5ea24da088b6..aa753b0d6f13 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -32,13 +32,16 @@
 #include "amdgpu_dm_colorop.h"
 
 const u64 amdgpu_dm_supported_degam_tfs =
-	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF);
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF);
 
 const u64 amdgpu_dm_supported_shaper_tfs =
-	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF);
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF);
 
 const u64 amdgpu_dm_supported_blnd_tfs =
-	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF);
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF);
 
 #define MAX_COLOR_PIPELINE_OPS 10
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 31/43] drm/colorop: add BT2020/BT709 OETF and Inverse OETF
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (29 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 30/43] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-15  9:35   ` Simon Ser
  2025-04-30  1:11 ` [PATCH V9 32/43] drm/amd/display: Add support for BT.709 and BT.2020 TFs Alex Hung
                   ` (13 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

The BT.709 and BT.2020 OETFs are the same, the only difference
being that the BT.2020 variant is defined with more precision
for 10 and 12-bit per color encodings.

Both are used as encoding functions for video content, and are
therefore defined as OETF (opto-electronic transfer function)
instead of as EOTF (electro-optical transfer function).

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Move DRM_COLOROP_1D_CURVE_BT2020_* from middle to end (Simon Ser)

 drivers/gpu/drm/drm_colorop.c |  2 ++
 include/drm/drm_colorop.h     | 23 +++++++++++++++++++++++
 2 files changed, 25 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 9ddb3dccaf72..3ae4a93f0200 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -72,6 +72,8 @@ static const char * const colorop_curve_1d_type_names[] = {
 	[DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF] = "sRGB Inverse EOTF",
 	[DRM_COLOROP_1D_CURVE_PQ_125_EOTF] = "PQ 125 EOTF",
 	[DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF] = "PQ 125 Inverse EOTF",
+	[DRM_COLOROP_1D_CURVE_BT2020_INV_OETF] = "BT.2020 Inverse OETF",
+	[DRM_COLOROP_1D_CURVE_BT2020_OETF] = "BT.2020 OETF",
 };
 
 
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 40d70d4a34a0..766caedea2f3 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -83,6 +83,29 @@ enum drm_colorop_curve_1d_type {
 	 */
 	DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF,
 
+	/**
+	 * @DRM_COLOROP_1D_CURVE_BT2020_INV_OETF:
+	 *
+	 * enum string "BT.2020 Inverse OETF"
+	 *
+	 * The inverse of &DRM_COLOROP_1D_CURVE_BT2020_OETF
+	 */
+	DRM_COLOROP_1D_CURVE_BT2020_INV_OETF,
+
+	/**
+	 * @DRM_COLOROP_1D_CURVE_BT2020_OETF:
+	 *
+	 * enum string "BT.2020 OETF"
+	 *
+	 * The BT.2020/BT.709 transfer function. The BT.709 and BT.2020
+	 * transfer functions are the same, the only difference is that
+	 * BT.2020 is defined with more precision for 10 and 12-bit
+	 * encodings.
+	 *
+	 *
+	 */
+	DRM_COLOROP_1D_CURVE_BT2020_OETF,
+
 	/**
 	 * @DRM_COLOROP_1D_CURVE_COUNT:
 	 *
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 32/43] drm/amd/display: Add support for BT.709 and BT.2020 TFs
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (30 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 31/43] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-12 19:37   ` Melissa Wen
  2025-04-30  1:11 ` [PATCH V9 33/43] drm/colorop: Add 1D Curve Custom LUT type Alex Hung
                   ` (12 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

This adds support for the BT.709/BT.2020 transfer functions
on all current 1D curve plane colorops, i.e., on DEGAM, SHAPER,
and BLND blocks.

With this change the following IGT subtests pass:
kms_colorop --run plane-XR30-XR30-bt2020_inv_oetf
kms_colorop --run plane-XR30-XR30-bt2020_oetf

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Move DRM_COLOROP_1D_CURVE_BT2020_* from middle to end

v8: 
 - Move BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) in amdgpu_dm_supported_blnd_tfs
   to "drm/amd/display: Enable support for PQ 125 EOTF and Inverse" (Leo Li)

 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c   | 11 ++++++++---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  9 ++++++---
 2 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 63044e0296cb..f645f9ded95f 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -679,6 +679,9 @@ amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
 	case DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
 	case DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF:
 		return TRANSFER_FUNCTION_PQ;
+	case DRM_COLOROP_1D_CURVE_BT2020_INV_OETF:
+	case DRM_COLOROP_1D_CURVE_BT2020_OETF:
+		return TRANSFER_FUNCTION_BT709;
 	default:
 		return TRANSFER_FUNCTION_LINEAR;
 	}
@@ -1284,8 +1287,10 @@ __set_colorop_1d_curve_blend_tf_lut(struct dc_plane_state *dc_plane_state,
 	const struct drm_color_lut *blend_lut;
 	uint32_t blend_size = 0;
 
-	if (colorop->type != DRM_COLOROP_1D_CURVE &&
-	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
+	if (colorop->type != DRM_COLOROP_1D_CURVE)
+		return -EINVAL;
+
+	if (!(BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs))
 		return -EINVAL;
 
 	if (colorop_state->bypass) {
@@ -1321,7 +1326,7 @@ __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
 	/* 3nd op: 1d curve - blend */
 	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
 		if (new_colorop_state->colorop == old_colorop &&
-		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
+		    (BIT(new_colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs)) {
 			colorop_state = new_colorop_state;
 			break;
 		}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index aa753b0d6f13..b989e1ca19e2 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -33,15 +33,18 @@
 
 const u64 amdgpu_dm_supported_degam_tfs =
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
-	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF);
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_BT2020_INV_OETF);
 
 const u64 amdgpu_dm_supported_shaper_tfs =
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF) |
-	BIT(DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF);
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_BT2020_OETF);
 
 const u64 amdgpu_dm_supported_blnd_tfs =
 	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
-	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF);
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_BT2020_INV_OETF);
 
 #define MAX_COLOR_PIPELINE_OPS 10
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 33/43] drm/colorop: Add 1D Curve Custom LUT type
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (31 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 32/43] drm/amd/display: Add support for BT.709 and BT.2020 TFs Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 34/43] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Alex Hung
                   ` (11 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

We've previously introduced DRM_COLOROP_1D_CURVE for
pre-defined 1D curves. But we also have HW that supports
custom curves and userspace needs the ability to pass
custom curves, aka LUTs.

This patch introduces a new colorop type, called
DRM_COLOROP_1D_LUT that provides a SIZE property which
is used by a driver to advertise the supported SIZE
of the LUT, as well as a DATA property which userspace
uses to set the LUT.

DATA and size function in the same way as current drm_crtc
GAMMA and DEGAMMA LUTs.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v8:
 - Add DRM_MODE_PROP_ATOMIC to drm_property_create_range (Simon Ser)
 - Change "1D Curve Custom LUT" to "1D LUT" (Simon Ser)

v7:
 - Change "size" to "lut_size" (this affects multiple following commits)
 - Move "lut_size" from drm_colorop_state to drm_colorop
 - Modify other files accordingly (i.e. from drm_colorop_state->size
   to drm_colorop->lut_size)

v5:
 - Add kernel doc
 - Define SIZE in similar manner to GAMMA_SIZE on drm_crtc (Melissa)

 drivers/gpu/drm/drm_atomic.c      |  4 +++
 drivers/gpu/drm/drm_atomic_uapi.c |  5 ++++
 drivers/gpu/drm/drm_colorop.c     | 43 +++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         | 16 ++++++++++++
 include/uapi/drm/drm_mode.h       | 14 ++++++++++
 5 files changed, 82 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 809bd856d378..1dda90554e46 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -793,6 +793,10 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 		drm_printf(p, "\tcurve_1d_type=%s\n",
 			   drm_get_colorop_curve_1d_type_name(state->curve_1d_type));
 		break;
+	case DRM_COLOROP_1D_LUT:
+		drm_printf(p, "\tsize=%d\n", colorop->lut_size);
+		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
+		break;
 	case DRM_COLOROP_CTM_3X4:
 		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
 		break;
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 0de0fdd8af5b..0cd250db3981 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -712,6 +712,9 @@ static int drm_atomic_color_set_data_property(struct drm_colorop *colorop,
 	bool replaced = false;
 
 	switch (colorop->type) {
+	case DRM_COLOROP_1D_LUT:
+		size = colorop->lut_size * sizeof(struct drm_color_lut);
+		break;
 	case DRM_COLOROP_CTM_3X4:
 		size = sizeof(struct drm_color_ctm_3x4);
 		break;
@@ -761,6 +764,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 		*val = state->bypass;
 	} else if (property == colorop->curve_1d_type_property) {
 		*val = state->curve_1d_type;
+	} else if (property == colorop->lut_size_property) {
+		*val = colorop->lut_size;
 	} else if (property == colorop->data_property) {
 		*val = (state->data) ? state->data->base.id : 0;
 	} else {
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 3ae4a93f0200..6c87f5b9f7f9 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -64,6 +64,7 @@
 
 static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
+	{ DRM_COLOROP_1D_LUT, "1D LUT" },
 	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
 };
 
@@ -266,6 +267,47 @@ static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_color
 	return 0;
 }
 
+/**
+ * drm_plane_colorop_curve_1d_lut_init - Initialize a DRM_COLOROP_1D_LUT
+ *
+ * @dev: DRM device
+ * @colorop: The drm_colorop object to initialize
+ * @plane: The associated drm_plane
+ * @lut_size: LUT size supported by driver
+ * @return zero on success, -E value on failure
+ */
+int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
+					struct drm_plane *plane, uint32_t lut_size)
+{
+	struct drm_property *prop;
+	int ret;
+
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_LUT);
+	if (ret)
+		return ret;
+
+	/* initialize 1D LUT only attribute */
+	/* LUT size */
+	prop = drm_property_create_range(dev, DRM_MODE_PROP_IMMUTABLE | DRM_MODE_PROP_ATOMIC,
+					 "SIZE", 0, UINT_MAX);
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->lut_size_property = prop;
+	drm_object_attach_property(&colorop->base, colorop->lut_size_property, lut_size);
+	colorop->lut_size = lut_size;
+
+	/* data */
+	ret = drm_colorop_create_data_prop(dev, colorop);
+	if (ret)
+		return ret;
+
+	drm_colorop_reset(colorop);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_colorop_curve_1d_lut_init);
+
 int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   struct drm_plane *plane)
 {
@@ -374,6 +416,7 @@ void drm_colorop_reset(struct drm_colorop *colorop)
 
 static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_CURVE] = "1D Curve",
+	[DRM_COLOROP_1D_LUT] = "1D LUT",
 	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
 };
 
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 766caedea2f3..52e965577bd8 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -259,6 +259,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *bypass_property;
 
+	/**
+	 * @lut_size:
+	 *
+	 * Number of entries of the custom LUT. This should be read-only.
+	 */
+	uint32_t lut_size;
+
 	/**
 	 * @curve_1d_type_property:
 	 *
@@ -266,6 +273,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @lut_size_property:
+	 *
+	 * Size property for custom LUT from userspace.
+	 */
+	struct drm_property *lut_size_property;
+
 	/**
 	 * @data_property:
 	 *
@@ -312,6 +326,8 @@ void drm_colorop_pipeline_destroy(struct drm_plane *plane);
 
 int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
 				    struct drm_plane *plane, u64 supported_tfs);
+int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
+					struct drm_plane *plane, uint32_t lut_size);
 int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   struct drm_plane *plane);
 
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index 651bdf48b766..dde250dd7a51 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -890,6 +890,20 @@ enum drm_colorop_type {
 	 */
 	DRM_COLOROP_1D_CURVE,
 
+	/**
+	 * @DRM_COLOROP_1D_LUT:
+	 *
+	 * enum string "1D LUT"
+	 *
+	 * A simple 1D LUT of uniformly spaced &drm_color_lut entries,
+	 * packed into a blob via the DATA property. The driver's
+	 * expected LUT size is advertised via the SIZE property.
+	 *
+	 * The DATA blob is an array of struct drm_color_lut with size
+	 * of "lut_size".
+	 */
+	DRM_COLOROP_1D_LUT,
+
 	/**
 	 * @DRM_COLOROP_CTM_3X4:
 	 *
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 34/43] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (32 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 33/43] drm/colorop: Add 1D Curve Custom LUT type Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop Alex Hung
                   ` (10 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

This patch adds colorops for custom 1D LUTs in the SHAPER and
BLND HW blocks.

With this change the following IGT tests pass:
kms_colorop --run plane-XR30-XR30-srgb_inv_eotf_lut
kms_colorop --run plane-XR30-XR30-srgb_inv_eotf_lut-srgb_eotf_lut

The color pipeline now consists of the following colorops:
1. 1D curve colorop
2. 1D curve colorop
3. 1D LUT
4. 1D curve colorop
5. 1D LUT

The 1D curve colorops support sRGB, BT2020, and PQ scaled to 125.0.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v8:
 - Return error when __set_output_tf fails (Leo Li)
 - 

v7:
 - Initialize uint32_t blend_size = 0 by default (kernel test robot)
 - Modify state->size to colorop->lut_size

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 173 ++++++++++--------
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  32 ++++
 2 files changed, 126 insertions(+), 79 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index f645f9ded95f..cf73c794ac5a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1211,38 +1211,6 @@ __set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
 	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);
 }
 
-static int
-__set_colorop_in_shaper_1d_curve(struct dc_plane_state *dc_plane_state,
-		       struct drm_colorop_state *colorop_state)
-{
-	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
-	struct drm_colorop *colorop = colorop_state->colorop;
-	struct drm_device *drm = colorop->dev;
-
-	if (colorop->type != DRM_COLOROP_1D_CURVE)
-		return -EINVAL;
-
-	if (!(BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_shaper_tfs))
-		return -EINVAL;
-
-	if (colorop_state->bypass) {
-		tf->type = TF_TYPE_BYPASS;
-		tf->tf = TRANSFER_FUNCTION_LINEAR;
-		return 0;
-	}
-
-	drm_dbg(drm, "Shaper colorop with ID: %d\n", colorop->base.id);
-
-	if (colorop->type == DRM_COLOROP_1D_CURVE) {
-		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
-		tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
-		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
-		return __set_output_tf(tf, 0, 0, false);
-	}
-
-	return -EINVAL;
-}
-
 static int
 __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 			      struct dc_plane_state *dc_plane_state,
@@ -1251,64 +1219,66 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 	struct drm_colorop *old_colorop;
 	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
 	struct drm_atomic_state *state = plane_state->state;
-	int i = 0;
+	enum dc_transfer_func_predefined default_tf = TRANSFER_FUNCTION_LINEAR;
+	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
+	const struct drm_color_lut *shaper_lut;
+	struct drm_device *dev = colorop->dev;
+	uint32_t shaper_size;
+	int i = 0, ret = 0;
 
+	/* 1D Curve - SHAPER TF */
 	old_colorop = colorop;
-
-	/* 2nd op: 1d curve - shaper */
 	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
 		if (new_colorop_state->colorop == old_colorop &&
 		    (BIT(new_colorop_state->curve_1d_type) & amdgpu_dm_supported_shaper_tfs)) {
 			colorop_state = new_colorop_state;
 			break;
 		}
-
-		if (new_colorop_state->colorop == old_colorop) {
-			colorop_state = new_colorop_state;
-			break;
-		}
 	}
 
-	if (!colorop_state)
-		return -EINVAL;
-
-	return __set_colorop_in_shaper_1d_curve(dc_plane_state, colorop_state);
-}
-
-
-static int
-__set_colorop_1d_curve_blend_tf_lut(struct dc_plane_state *dc_plane_state,
-				  struct drm_colorop_state *colorop_state)
-{
-
-	struct dc_transfer_func *tf = &dc_plane_state->blend_tf;
-	struct drm_colorop *colorop = colorop_state->colorop;
-	struct drm_device *drm = colorop->dev;
-	const struct drm_color_lut *blend_lut;
-	uint32_t blend_size = 0;
-
-	if (colorop->type != DRM_COLOROP_1D_CURVE)
-		return -EINVAL;
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_1D_CURVE) {
+		drm_dbg(dev, "Shaper TF colorop with ID: %d\n", colorop->base.id);
+		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+		tf->tf = default_tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
+		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+		ret = __set_output_tf(tf, 0, 0, false);
+		if (ret)
+			return ret;
+	}
 
-	if (!(BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs))
+	/* 1D LUT - SHAPER LUT */
+	colorop = old_colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no Shaper LUT colorop found\n");
 		return -EINVAL;
-
-	if (colorop_state->bypass) {
-		tf->type = TF_TYPE_BYPASS;
-		tf->tf = TRANSFER_FUNCTION_LINEAR;
-		return 0;
 	}
 
-	drm_dbg(drm, "Blend colorop with ID: %d\n", colorop->base.id);
+	old_colorop = colorop;
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->colorop->type == DRM_COLOROP_1D_LUT) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
 
-	if (colorop->type == DRM_COLOROP_1D_CURVE) {
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_1D_LUT) {
+		drm_dbg(dev, "Shaper LUT colorop with ID: %d\n", colorop->base.id);
 		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
-		tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
+		tf->tf = default_tf;
 		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
-		return __set_input_tf(NULL, tf, blend_lut, blend_size);
+		shaper_lut = __extract_blob_lut(colorop_state->data, &shaper_size);
+		shaper_size = shaper_lut != NULL ? shaper_size : 0;
+
+		/* Custom LUT size must be the same as supported size */
+		if (shaper_size == colorop->lut_size) {
+			ret = __set_output_tf(tf, shaper_lut, shaper_size, false);
+			if (ret)
+				return ret;
+		}
 	}
 
-	return -EINVAL;
+	return 0;
 }
 
 static int
@@ -1319,28 +1289,63 @@ __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
 	struct drm_colorop *old_colorop;
 	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
 	struct drm_atomic_state *state = plane_state->state;
+	enum dc_transfer_func_predefined default_tf = TRANSFER_FUNCTION_LINEAR;
+	struct dc_transfer_func *tf = &dc_plane_state->blend_tf;
+	const struct drm_color_lut *blend_lut;
+	struct drm_device *dev = colorop->dev;
+	uint32_t blend_size;
 	int i = 0;
 
+	/* 1D Curve - BLND TF */
 	old_colorop = colorop;
-
-	/* 3nd op: 1d curve - blend */
 	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
 		if (new_colorop_state->colorop == old_colorop &&
 		    (BIT(new_colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs)) {
 			colorop_state = new_colorop_state;
 			break;
 		}
+	}
 
-		if (new_colorop_state->colorop == old_colorop) {
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_1D_CURVE &&
+	    (BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs)) {
+		drm_dbg(dev, "Blend TF colorop with ID: %d\n", colorop->base.id);
+		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+		tf->tf = default_tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
+		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+		__set_input_tf(NULL, tf, blend_lut, blend_size);
+	}
+
+	/* 1D Curve - BLND LUT */
+	colorop = old_colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no Blend LUT colorop found\n");
+		return -EINVAL;
+	}
+
+	old_colorop = colorop;
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->colorop->type == DRM_COLOROP_1D_LUT) {
 			colorop_state = new_colorop_state;
 			break;
 		}
 	}
 
-	if (!colorop_state)
-		return -EINVAL;
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_1D_LUT &&
+	    (BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs)) {
+		drm_dbg(dev, "Blend LUT colorop with ID: %d\n", colorop->base.id);
+		tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+		tf->tf = default_tf;
+		tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+		blend_lut = __extract_blob_lut(colorop_state->data, &blend_size);
+		blend_size = blend_lut != NULL ? blend_size : 0;
 
-	return __set_colorop_1d_curve_blend_tf_lut(dc_plane_state, colorop_state);
+		/* Custom LUT size must be the same as supported size */
+		if (blend_size == colorop->lut_size)
+			__set_input_tf(NULL, tf, blend_lut, blend_size);
+	}
+
+	return 0;
 }
 
 static int
@@ -1409,7 +1414,7 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
-	/* 1D Curve - SHAPER TF */
+	/* 1D Curve & LUT - SHAPER TF & LUT */
 	colorop = colorop->next;
 	if (!colorop) {
 		drm_dbg(dev, "no Shaper TF colorop found\n");
@@ -1420,7 +1425,12 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
-	/* 1D Curve - BLND TF */
+	/* Shaper LUT colorop is already handled, just skip here */
+	colorop = colorop->next;
+	if (!colorop)
+		return -EINVAL;
+
+	/* 1D Curve & LUT - BLND TF & LUT */
 	colorop = colorop->next;
 	if (!colorop) {
 		drm_dbg(dev, "no Blend TF colorop found\n");
@@ -1431,6 +1441,11 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
+	/* BLND LUT colorop is already handled, just skip here */
+	colorop = colorop->next;
+	if (!colorop)
+		return -EINVAL;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index b989e1ca19e2..1a03f7ee2fbb 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -29,6 +29,7 @@
 #include <drm/drm_property.h>
 #include <drm/drm_colorop.h>
 
+#include "amdgpu.h"
 #include "amdgpu_dm_colorop.h"
 
 const u64 amdgpu_dm_supported_degam_tfs =
@@ -90,6 +91,22 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	i++;
 
+	/* 1D LUT - SHAPER LUT */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
+
+	i++;
+
 	/* 1D curve - BLND TF */
 	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!ops[i]) {
@@ -104,6 +121,21 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	drm_colorop_set_next_property(ops[i-1], ops[i]);
 
+	i++;
+
+	/* 1D LUT - BLND LUT */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
 	return 0;
 
 cleanup:
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (33 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 34/43] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-07 12:08   ` kernel test robot
  2025-04-30  1:11 ` [PATCH V9 36/43] drm/colorop: Add mutliplier type Alex Hung
                   ` (9 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

This adds support for a 3x4 color transformation matrix.

With this change the following IGT tests pass:
kms_colorop --run plane-XR30-XR30-ctm_3x4_50_desat
kms_colorop --run plane-XR30-XR30-ctm_3x4_overdrive
kms_colorop --run plane-XR30-XR30-ctm_3x4_oversaturate
kms_colorop --run plane-XR30-XR30-ctm_3x4_bt709_enc
kms_colorop --run plane-XR30-XR30-ctm_3x4_bt709_dec

The color pipeline now consists of the following colorops:
1. 1D curve colorop
2. 3x4 CTM
3. 1D curve colorop
4. 1D LUT
5. 1D curve colorop
6. 1D LUT

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v8:
 - Return -EINVAL when drm_color_ctm_3x4's size mismatches (Leo Li)

v7:
 - Change %lu to %zu for sizeof() when In drm_warn

v6:
 - fix warnings in dbg prints

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 52 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 16 ++++++
 2 files changed, 68 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index cf73c794ac5a..268d4837759e 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1211,6 +1211,47 @@ __set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
 	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);
 }
 
+static int
+__set_dm_plane_colorop_3x4_matrix(struct drm_plane_state *plane_state,
+				  struct dc_plane_state *dc_plane_state,
+				  struct drm_colorop *colorop)
+{
+	struct drm_colorop *old_colorop;
+	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+	struct drm_atomic_state *state = plane_state->state;
+	const struct drm_device *dev = colorop->dev;
+	const struct drm_property_blob *blob;
+	struct drm_color_ctm_3x4 *ctm = NULL;
+	int i = 0;
+
+	/* 3x4 matrix */
+	old_colorop = colorop;
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->colorop->type == DRM_COLOROP_CTM_3X4) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
+
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_CTM_3X4) {
+		drm_dbg(dev, "3x4 matrix colorop with ID: %d\n", colorop->base.id);
+		blob = colorop_state->data;
+		if (blob->length == sizeof(struct drm_color_ctm_3x4)) {
+			ctm = blob ? (struct drm_color_ctm_3x4 *) blob->data : NULL;
+			__drm_ctm_3x4_to_dc_matrix(ctm, dc_plane_state->gamut_remap_matrix.matrix);
+			dc_plane_state->gamut_remap_matrix.enable_remap = true;
+			dc_plane_state->input_csc_color_matrix.enable_adjustment = false;
+		} else {
+			drm_warn(dev, "blob->length (%lu) isn't equal to drm_color_ctm_3x4 (%zu)\n",
+				 blob->length, sizeof(struct drm_color_ctm_3x4));
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
 static int
 __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 			      struct dc_plane_state *dc_plane_state,
@@ -1414,6 +1455,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
+	/* 3x4 matrix */
+	colorop = colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no 3x4 matrix colorop found\n");
+		return -EINVAL;
+	}
+
+	ret = __set_dm_plane_colorop_3x4_matrix(plane_state, dc_plane_state, colorop);
+	if (ret)
+		return ret;
+
 	/* 1D Curve & LUT - SHAPER TF & LUT */
 	colorop = colorop->next;
 	if (!colorop) {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 1a03f7ee2fbb..39f0b383700b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -75,6 +75,22 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	i++;
 
+	/* 3x4 matrix */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
+
+	i++;
+
 	/* 1D curve - SHAPER TF */
 	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!ops[i]) {
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 36/43] drm/colorop: Add mutliplier type
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (34 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-12 20:02   ` Melissa Wen
  2025-04-30  1:11 ` [PATCH V9 37/43] drm/amd/display: add multiplier colorop Alex Hung
                   ` (8 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

This introduces a new drm_colorop_type: DRM_COLOROP_MULTIPLIER.

It's a simple multiplier to all pixel values. The value is
specified via a S31.32 fixed point provided via the
"MULTIPLIER" property.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v7:
 - Modify size_property to lut_size_property

 drivers/gpu/drm/drm_atomic.c      |  3 +++
 drivers/gpu/drm/drm_atomic_uapi.c |  4 ++++
 drivers/gpu/drm/drm_colorop.c     | 33 +++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         | 16 +++++++++++++++
 include/uapi/drm/drm_mode.h       | 11 +++++++++++
 5 files changed, 67 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 1dda90554e46..907ca790689f 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -800,6 +800,9 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 	case DRM_COLOROP_CTM_3X4:
 		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
 		break;
+	case DRM_COLOROP_MULTIPLIER:
+		drm_printf(p, "\tmultiplier=%llu\n", state->multiplier);
+		break;
 	default:
 		break;
 	}
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 0cd250db3981..a7f1d75bb4ea 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -739,6 +739,8 @@ static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
 		state->bypass = val;
 	} else if (property == colorop->curve_1d_type_property) {
 		state->curve_1d_type = val;
+	} else if (property == colorop->multiplier_property) {
+		state->multiplier = val;
 	} else if (property == colorop->data_property) {
 		return drm_atomic_color_set_data_property(colorop, state,
 							  property, val);
@@ -764,6 +766,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 		*val = state->bypass;
 	} else if (property == colorop->curve_1d_type_property) {
 		*val = state->curve_1d_type;
+	} else if (property == colorop->multiplier_property) {
+		*val = state->multiplier;
 	} else if (property == colorop->lut_size_property) {
 		*val = colorop->lut_size;
 	} else if (property == colorop->data_property) {
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 6c87f5b9f7f9..a3cf62c5e263 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -66,6 +66,7 @@ static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
 	{ DRM_COLOROP_1D_LUT, "1D LUT" },
 	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
+	{ DRM_COLOROP_MULTIPLIER, "Multiplier"},
 };
 
 static const char * const colorop_curve_1d_type_names[] = {
@@ -327,6 +328,37 @@ int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *c
 }
 EXPORT_SYMBOL(drm_plane_colorop_ctm_3x4_init);
 
+/**
+ * drm_plane_colorop_mult_init - Initialize a DRM_COLOROP_MULTIPLIER
+ *
+ * @dev: DRM device
+ * @colorop: The drm_colorop object to initialize
+ * @plane: The associated drm_plane
+ * @return zero on success, -E value on failure
+ */
+int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
+				struct drm_plane *plane)
+{
+	struct drm_property *prop;
+	int ret;
+
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_MULTIPLIER);
+	if (ret)
+		return ret;
+
+	prop = drm_property_create_range(dev, DRM_MODE_PROP_ATOMIC, "MULTIPLIER", 0, U64_MAX);
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->multiplier_property = prop;
+	drm_object_attach_property(&colorop->base, colorop->multiplier_property, 0);
+
+	drm_colorop_reset(colorop);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_colorop_mult_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
@@ -418,6 +450,7 @@ static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_CURVE] = "1D Curve",
 	[DRM_COLOROP_1D_LUT] = "1D LUT",
 	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
+	[DRM_COLOROP_MULTIPLIER] = "Multiplier",
 };
 
 const char *drm_get_colorop_type_name(enum drm_colorop_type type)
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 52e965577bd8..888184aef7a0 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -146,6 +146,13 @@ struct drm_colorop_state {
 	 */
 	enum drm_colorop_curve_1d_type curve_1d_type;
 
+	/**
+	 * @multiplier:
+	 *
+	 * Multiplier to 'gain' the plane. Format is S31.32 sign-magnitude.
+	 */
+	uint64_t multiplier;
+
 	/**
 	 * @data:
 	 *
@@ -273,6 +280,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @multiplier_property:
+	 *
+	 * Multiplier property for plane gain
+	 */
+	struct drm_property *multiplier_property;
+
 	/**
 	 * @lut_size_property:
 	 *
@@ -330,6 +344,8 @@ int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_color
 					struct drm_plane *plane, uint32_t lut_size);
 int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   struct drm_plane *plane);
+int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
+				struct drm_plane *plane);
 
 struct drm_colorop_state *
 drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index dde250dd7a51..fa5e1ddaf4f0 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -919,6 +919,17 @@ enum drm_colorop_type {
 	 * | B |   | 8  9  10 12 |   | B |
 	 */
 	DRM_COLOROP_CTM_3X4,
+
+	/**
+	 * @DRM_COLOROP_MULTIPLIER:
+	 *
+	 * enum string "Multiplier"
+	 *
+	 * A simple multiplier, applied to all color values. The
+	 * multiplier is specified as a S31.32 via the MULTIPLIER
+	 * property.
+	 */
+	DRM_COLOROP_MULTIPLIER,
 };
 
 /**
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 37/43] drm/amd/display: add multiplier colorop
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (35 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 36/43] drm/colorop: Add mutliplier type Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 38/43] drm/amd/display: Swap matrix and multiplier Alex Hung
                   ` (7 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

This adds support for a multiplier. This multiplier is
programmed via the HDR Multiplier in DCN.

With this change the following IGT tests pass:
kms_colorop --run plane-XR30-XR30-multiply_125
kms_colorop --run plane-XR30-XR30-multiply_inv_125

The color pipeline now consists of the following colorops:
1. 1D curve colorop
2. 3x4 CTM
3. Multiplier
4. 1D curve colorop
5. 1D LUT
6. 1D curve colorop
7. 1D LUT

Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 40 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 16 ++++++++
 2 files changed, 56 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 268d4837759e..0044ea112ef8 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1252,6 +1252,35 @@ __set_dm_plane_colorop_3x4_matrix(struct drm_plane_state *plane_state,
 	return 0;
 }
 
+static int
+__set_dm_plane_colorop_multiplier(struct drm_plane_state *plane_state,
+				  struct dc_plane_state *dc_plane_state,
+				  struct drm_colorop *colorop)
+{
+	struct drm_colorop *old_colorop;
+	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+	struct drm_atomic_state *state = plane_state->state;
+	const struct drm_device *dev = colorop->dev;
+	int i = 0;
+
+	/* Multiplier */
+	old_colorop = colorop;
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->colorop->type == DRM_COLOROP_MULTIPLIER) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
+
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_MULTIPLIER) {
+		drm_dbg(dev, "Multiplier colorop with ID: %d\n", colorop->base.id);
+		dc_plane_state->hdr_mult = amdgpu_dm_fixpt_from_s3132(colorop_state->multiplier);
+	}
+
+	return 0;
+}
+
 static int
 __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 			      struct dc_plane_state *dc_plane_state,
@@ -1466,6 +1495,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
+	/* Multiplier */
+	colorop = colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no multiplier colorop found\n");
+		return -EINVAL;
+	}
+
+	ret = __set_dm_plane_colorop_multiplier(plane_state, dc_plane_state, colorop);
+	if (ret)
+		return ret;
+
 	/* 1D Curve & LUT - SHAPER TF & LUT */
 	colorop = colorop->next;
 	if (!colorop) {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 39f0b383700b..efd4799165a8 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -91,6 +91,22 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	i++;
 
+	/* Multiplier */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_mult_init(dev, ops[i], plane);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
+
+	i++;
+
 	/* 1D curve - SHAPER TF */
 	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!ops[i]) {
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 38/43] drm/amd/display: Swap matrix and multiplier
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (36 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 37/43] drm/amd/display: add multiplier colorop Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 39/43] drm/colorop: Define LUT_1D interpolation Alex Hung
                   ` (6 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Swap the order of matrix and multiplier as designed in hardware.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c  | 12 ++++++------
 .../drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c    |  8 ++++----
 2 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 0044ea112ef8..313716f2003f 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1484,25 +1484,25 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (ret)
 		return ret;
 
-	/* 3x4 matrix */
+	/* Multiplier */
 	colorop = colorop->next;
 	if (!colorop) {
-		drm_dbg(dev, "no 3x4 matrix colorop found\n");
+		drm_dbg(dev, "no multiplier colorop found\n");
 		return -EINVAL;
 	}
 
-	ret = __set_dm_plane_colorop_3x4_matrix(plane_state, dc_plane_state, colorop);
+	ret = __set_dm_plane_colorop_multiplier(plane_state, dc_plane_state, colorop);
 	if (ret)
 		return ret;
 
-	/* Multiplier */
+	/* 3x4 matrix */
 	colorop = colorop->next;
 	if (!colorop) {
-		drm_dbg(dev, "no multiplier colorop found\n");
+		drm_dbg(dev, "no 3x4 matrix colorop found\n");
 		return -EINVAL;
 	}
 
-	ret = __set_dm_plane_colorop_multiplier(plane_state, dc_plane_state, colorop);
+	ret = __set_dm_plane_colorop_3x4_matrix(plane_state, dc_plane_state, colorop);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index efd4799165a8..bce461ecdad4 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -75,7 +75,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	i++;
 
-	/* 3x4 matrix */
+	/* Multiplier */
 	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!ops[i]) {
 		DRM_ERROR("KMS: Failed to allocate colorop\n");
@@ -83,7 +83,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	ret = drm_plane_colorop_mult_init(dev, ops[i], plane);
 	if (ret)
 		goto cleanup;
 
@@ -91,7 +91,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	i++;
 
-	/* Multiplier */
+	/* 3x4 matrix */
 	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!ops[i]) {
 		DRM_ERROR("KMS: Failed to allocate colorop\n");
@@ -99,7 +99,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_mult_init(dev, ops[i], plane);
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
 	if (ret)
 		goto cleanup;
 
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 39/43] drm/colorop: Define LUT_1D interpolation
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (37 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 38/43] drm/amd/display: Swap matrix and multiplier Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 40/43] drm/colorop: allow non-bypass colorops Alex Hung
                   ` (5 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

We want to make sure userspace is aware of the 1D LUT
interpolation. While linear interpolation is common it
might not be supported on all HW. Give driver implementers
a way to specify their interpolation.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v7:
 - Fix a checkpatch long-line warning
 - Modify state->size to colorop->lut_size

 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  6 ++--
 drivers/gpu/drm/drm_atomic.c                  |  2 ++
 drivers/gpu/drm/drm_colorop.c                 | 36 ++++++++++++++++++-
 include/drm/drm_colorop.h                     | 19 +++++++++-
 include/uapi/drm/drm_mode.h                   | 13 +++++++
 5 files changed, 72 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index bce461ecdad4..ff11cb0603c8 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -131,7 +131,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES);
+	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES,
+						  DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
 	if (ret)
 		goto cleanup;
 
@@ -163,7 +164,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES);
+	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES,
+						  DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
 	if (ret)
 		goto cleanup;
 
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 907ca790689f..90df638f25d5 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -795,6 +795,8 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 		break;
 	case DRM_COLOROP_1D_LUT:
 		drm_printf(p, "\tsize=%d\n", colorop->lut_size);
+		drm_printf(p, "\tinterpolation=%s\n",
+			   drm_get_colorop_lut1d_interpolation_name(colorop->lut1d_interpolation));
 		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
 		break;
 	case DRM_COLOROP_CTM_3X4:
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index a3cf62c5e263..125b3e507d6c 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -78,6 +78,9 @@ static const char * const colorop_curve_1d_type_names[] = {
 	[DRM_COLOROP_1D_CURVE_BT2020_OETF] = "BT.2020 OETF",
 };
 
+static const struct drm_prop_enum_list drm_colorop_lut1d_interpolation_list[] = {
+	{ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR, "Linear" },
+};
 
 /* Init Helpers */
 
@@ -275,10 +278,12 @@ static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_color
  * @colorop: The drm_colorop object to initialize
  * @plane: The associated drm_plane
  * @lut_size: LUT size supported by driver
+ * @lut1d_interpolation: 1D LUT interpolation type
  * @return zero on success, -E value on failure
  */
 int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
-					struct drm_plane *plane, uint32_t lut_size)
+					struct drm_plane *plane, uint32_t lut_size,
+					enum drm_colorop_lut1d_interpolation_type lut1d_interpolation)
 {
 	struct drm_property *prop;
 	int ret;
@@ -298,6 +303,17 @@ int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_color
 	drm_object_attach_property(&colorop->base, colorop->lut_size_property, lut_size);
 	colorop->lut_size = lut_size;
 
+	/* Interpolation */
+	prop = drm_property_create_enum(dev, DRM_MODE_PROP_IMMUTABLE, "LUT1D_INTERPOLATION",
+					drm_colorop_lut1d_interpolation_list,
+					ARRAY_SIZE(drm_colorop_lut1d_interpolation_list));
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->lut1d_interpolation_property = prop;
+	drm_object_attach_property(&colorop->base, prop, lut1d_interpolation);
+	colorop->lut1d_interpolation = lut1d_interpolation;
+
 	/* data */
 	ret = drm_colorop_create_data_prop(dev, colorop);
 	if (ret)
@@ -452,6 +468,9 @@ static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
 	[DRM_COLOROP_MULTIPLIER] = "Multiplier",
 };
+static const char * const colorop_lut1d_interpolation_name[] = {
+	[DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR] = "Linear",
+};
 
 const char *drm_get_colorop_type_name(enum drm_colorop_type type)
 {
@@ -469,6 +488,21 @@ const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type ty
 	return colorop_curve_1d_type_names[type];
 }
 
+/**
+ * drm_get_colorop_lut1d_interpolation_name: return a string for interpolation type
+ * @type: interpolation type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_interpolation_type type)
+{
+	if (WARN_ON(type >= ARRAY_SIZE(colorop_lut1d_interpolation_name)))
+		return "unknown";
+
+	return colorop_lut1d_interpolation_name[type];
+}
+
 /**
  * drm_colorop_set_next_property - sets the next pointer
  * @colorop: drm colorop
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 888184aef7a0..00dbde7e1a92 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -273,6 +273,21 @@ struct drm_colorop {
 	 */
 	uint32_t lut_size;
 
+	/**
+	 * @lut1d_interpolation:
+	 *
+	 * Read-only
+	 * Interpolation for DRM_COLOROP_1D_LUT
+	 */
+	enum drm_colorop_lut1d_interpolation_type lut1d_interpolation;
+
+	/**
+	 * @lut1d_interpolation_property:
+	 *
+	 * Read-only property for DRM_COLOROP_1D_LUT interpolation
+	 */
+	struct drm_property *lut1d_interpolation_property;
+
 	/**
 	 * @curve_1d_type_property:
 	 *
@@ -341,7 +356,8 @@ void drm_colorop_pipeline_destroy(struct drm_plane *plane);
 int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
 				    struct drm_plane *plane, u64 supported_tfs);
 int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
-					struct drm_plane *plane, uint32_t lut_size);
+					struct drm_plane *plane, uint32_t lut_size,
+					enum drm_colorop_lut1d_interpolation_type lut1d_interpolation);
 int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   struct drm_plane *plane);
 int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
@@ -395,6 +411,7 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type);
  * const pointer and hence is threadsafe.
  */
 const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
+const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_interpolation_type type);
 
 void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
 
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index fa5e1ddaf4f0..d76c8ffe5408 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -932,6 +932,19 @@ enum drm_colorop_type {
 	DRM_COLOROP_MULTIPLIER,
 };
 
+/**
+ * enum drm_colorop_lut1d_interpolation_type - type of interpolation for 1D LUTs
+ */
+enum drm_colorop_lut1d_interpolation_type {
+	/**
+	 * @DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR:
+	 *
+	 * Linear interpolation. Values between points of the LUT will be
+	 * linearly interpolated.
+	 */
+	DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR,
+};
+
 /**
  * struct drm_plane_size_hint - Plane size hints
  * @width: The width of the plane in pixel
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 40/43] drm/colorop: allow non-bypass colorops
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (38 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 39/43] drm/colorop: Define LUT_1D interpolation Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-15  9:37   ` Simon Ser
  2025-04-30  1:11 ` [PATCH V9 41/43] drm/colorop: Add 3D LUT support to color pipeline Alex Hung
                   ` (4 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Not all HW will be able to do bypass on all color
operations. Introduce an 32 bits 'flags' for all colorop
init functions and DRM_COLOROP_FLAG_ALLOW_BYPASS for creating
the BYPASS property when it's true.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)
 - Chagne "bool allow_bypass" to "uint32_t flags" for better extensibility (Simon Ser)

 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 22 +++++++---
 drivers/gpu/drm/drm_atomic.c                  |  3 +-
 drivers/gpu/drm/drm_colorop.c                 | 44 +++++++++++--------
 drivers/gpu/drm/vkms/vkms_colorop.c           | 10 +++--
 include/drm/drm_colorop.h                     | 11 +++--
 5 files changed, 55 insertions(+), 35 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index ff11cb0603c8..10b3e3906461 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -66,7 +66,9 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_degam_tfs);
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane,
+					      amdgpu_dm_supported_degam_tfs,
+					      DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -83,7 +85,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_mult_init(dev, ops[i], plane);
+	ret = drm_plane_colorop_mult_init(dev, ops[i], plane, DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -99,7 +101,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane, DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -115,7 +117,9 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_shaper_tfs);
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane,
+					      amdgpu_dm_supported_shaper_tfs,
+					      DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -132,7 +136,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 	}
 
 	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES,
-						  DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
+						  DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR,
+						  DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -148,7 +153,9 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_blnd_tfs);
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane,
+					      amdgpu_dm_supported_blnd_tfs,
+					      DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -165,7 +172,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 	}
 
 	ret = drm_plane_colorop_curve_1d_lut_init(dev, ops[i], plane, MAX_COLOR_LUT_ENTRIES,
-						  DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
+						  DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR,
+						  DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 90df638f25d5..41dd6a90e13b 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -786,7 +786,8 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 
 	drm_printf(p, "colorop[%u]:\n", colorop->base.id);
 	drm_printf(p, "\ttype=%s\n", drm_get_colorop_type_name(colorop->type));
-	drm_printf(p, "\tbypass=%u\n", state->bypass);
+	if (colorop->bypass_property)
+		drm_printf(p, "\tbypass=%u\n", state->bypass);
 
 	switch (colorop->type) {
 	case DRM_COLOROP_1D_CURVE:
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 125b3e507d6c..38bbe57db348 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -85,7 +85,7 @@ static const struct drm_prop_enum_list drm_colorop_lut1d_interpolation_list[] =
 /* Init Helpers */
 
 static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
-			    struct drm_plane *plane, enum drm_colorop_type type)
+			    struct drm_plane *plane, enum drm_colorop_type type, uint32_t flags)
 {
 	struct drm_mode_config *config = &dev->mode_config;
 	struct drm_property *prop;
@@ -121,16 +121,18 @@ static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *co
 				   colorop->type_property,
 				   colorop->type);
 
-	/* bypass */
-	prop = drm_property_create_bool(dev, DRM_MODE_PROP_ATOMIC,
-					"BYPASS");
-	if (!prop)
-		return -ENOMEM;
-
-	colorop->bypass_property = prop;
-	drm_object_attach_property(&colorop->base,
-				   colorop->bypass_property,
-				   1);
+	if (flags & DRM_COLOROP_FLAG_ALLOW_BYPASS) {
+		/* bypass */
+		prop = drm_property_create_bool(dev, DRM_MODE_PROP_ATOMIC,
+						"BYPASS");
+		if (!prop)
+			return -ENOMEM;
+
+		colorop->bypass_property = prop;
+		drm_object_attach_property(&colorop->base,
+					colorop->bypass_property,
+					1);
+	}
 
 	/* next */
 	prop = drm_property_create_object(dev, DRM_MODE_PROP_IMMUTABLE | DRM_MODE_PROP_ATOMIC,
@@ -195,10 +197,11 @@ EXPORT_SYMBOL(drm_colorop_pipeline_destroy);
  * @supported_tfs: A bitfield of supported drm_plane_colorop_curve_1d_init enum values,
  *                 created using BIT(curve_type) and combined with the OR '|'
  *                 operator.
+ * @flags: bitmask of misc, see DRM_COLOROP_FLAGS_* defines.
  * @return zero on success, -E value on failure
  */
 int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
-				    struct drm_plane *plane, u64 supported_tfs)
+				    struct drm_plane *plane, u64 supported_tfs, uint32_t flags)
 {
 	struct drm_prop_enum_list enum_list[DRM_COLOROP_1D_CURVE_COUNT];
 	int i, len;
@@ -219,7 +222,7 @@ int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *
 		return -EINVAL;
 	}
 
-	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_CURVE);
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_CURVE, flags);
 	if (ret)
 		return ret;
 
@@ -279,16 +282,18 @@ static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_color
  * @plane: The associated drm_plane
  * @lut_size: LUT size supported by driver
  * @lut1d_interpolation: 1D LUT interpolation type
+ * @flags: bitmask of misc, see DRM_COLOROP_FLAGS_* defines.
  * @return zero on success, -E value on failure
  */
 int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
 					struct drm_plane *plane, uint32_t lut_size,
-					enum drm_colorop_lut1d_interpolation_type lut1d_interpolation)
+					enum drm_colorop_lut1d_interpolation_type lut1d_interpolation,
+					uint32_t flags)
 {
 	struct drm_property *prop;
 	int ret;
 
-	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_LUT);
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_LUT, flags);
 	if (ret)
 		return ret;
 
@@ -326,11 +331,11 @@ int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_color
 EXPORT_SYMBOL(drm_plane_colorop_curve_1d_lut_init);
 
 int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
-				   struct drm_plane *plane)
+				   struct drm_plane *plane, uint32_t flags)
 {
 	int ret;
 
-	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_CTM_3X4);
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_CTM_3X4, flags);
 	if (ret)
 		return ret;
 
@@ -350,15 +355,16 @@ EXPORT_SYMBOL(drm_plane_colorop_ctm_3x4_init);
  * @dev: DRM device
  * @colorop: The drm_colorop object to initialize
  * @plane: The associated drm_plane
+ * @flags: bitmask of misc, see DRM_COLOROP_FLAGS_* defines.
  * @return zero on success, -E value on failure
  */
 int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
-				struct drm_plane *plane)
+				struct drm_plane *plane, uint32_t flags)
 {
 	struct drm_property *prop;
 	int ret;
 
-	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_MULTIPLIER);
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_MULTIPLIER, flags);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
index 70be35804ea9..f61834d2ea30 100644
--- a/drivers/gpu/drm/vkms/vkms_colorop.c
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -31,7 +31,8 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs);
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs,
+					      DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -48,7 +49,7 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane, DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -64,7 +65,7 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane);
+	ret = drm_plane_colorop_ctm_3x4_init(dev, ops[i], plane, DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
@@ -80,7 +81,8 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
 		goto cleanup;
 	}
 
-	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs);
+	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, supported_tfs,
+					      DRM_COLOROP_FLAG_ALLOW_BYPASS);
 	if (ret)
 		goto cleanup;
 
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 00dbde7e1a92..e79472739d11 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -31,6 +31,8 @@
 #include <drm/drm_mode.h>
 #include <drm/drm_property.h>
 
+/* DRM colorop flags */
+#define DRM_COLOROP_FLAG_ALLOW_BYPASS	(1<<0)	/* Allow bypass on the drm_colorop */
 
 /**
  * enum drm_colorop_curve_1d_type - type of 1D curve
@@ -354,14 +356,15 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
 void drm_colorop_pipeline_destroy(struct drm_plane *plane);
 
 int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
-				    struct drm_plane *plane, u64 supported_tfs);
+				    struct drm_plane *plane, u64 supported_tfs, uint32_t flags);
 int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
 					struct drm_plane *plane, uint32_t lut_size,
-					enum drm_colorop_lut1d_interpolation_type lut1d_interpolation);
+					enum drm_colorop_lut1d_interpolation_type lut1d_interpolation,
+					uint32_t flags);
 int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
-				   struct drm_plane *plane);
+				   struct drm_plane *plane, uint32_t flags);
 int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
-				struct drm_plane *plane);
+				struct drm_plane *plane, uint32_t flags);
 
 struct drm_colorop_state *
 drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 41/43] drm/colorop: Add 3D LUT support to color pipeline
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (39 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 40/43] drm/colorop: allow non-bypass colorops Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-04-30  1:11 ` [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop Alex Hung
                   ` (3 subsequent siblings)
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

It is to be used to enable HDR by allowing userpace to create and pass
3D LUTs to kernel and hardware.

new drm_colorop_type: DRM_COLOROP_3D_LUT.

Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Update function names by _plane_ (Chaitanya Kumar Borah)

v8:
 - Fix typo in subject (Simon Ser)
 - Update documentation for DRM_COLOROP_3D_LUT (Simon Ser)
 - Delete empty lines (Simon Ser)

v7:
 - Simplify 3D LUT by removing lut_3d_modes and related functions (Simon Ser)

 drivers/gpu/drm/drm_atomic.c      |  6 +++
 drivers/gpu/drm/drm_atomic_uapi.c |  6 +++
 drivers/gpu/drm/drm_colorop.c     | 72 +++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         | 21 +++++++++
 include/uapi/drm/drm_mode.h       | 33 ++++++++++++++
 5 files changed, 138 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 41dd6a90e13b..35640154c70a 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -806,6 +806,12 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 	case DRM_COLOROP_MULTIPLIER:
 		drm_printf(p, "\tmultiplier=%llu\n", state->multiplier);
 		break;
+	case DRM_COLOROP_3D_LUT:
+		drm_printf(p, "\tsize=%d\n", colorop->lut_size);
+		drm_printf(p, "\tinterpolation=%s\n",
+			   drm_get_colorop_lut3d_interpolation_name(colorop->lut3d_interpolation));
+		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
+		break;
 	default:
 		break;
 	}
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index a7f1d75bb4ea..497dba5ffe14 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -718,6 +718,10 @@ static int drm_atomic_color_set_data_property(struct drm_colorop *colorop,
 	case DRM_COLOROP_CTM_3X4:
 		size = sizeof(struct drm_color_ctm_3x4);
 		break;
+	case DRM_COLOROP_3D_LUT:
+		size = colorop->lut_size * colorop->lut_size * colorop->lut_size *
+		       sizeof(struct drm_color_lut);
+		break;
 	default:
 		/* should never get here */
 		return -EINVAL;
@@ -770,6 +774,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
 		*val = state->multiplier;
 	} else if (property == colorop->lut_size_property) {
 		*val = colorop->lut_size;
+	} else if (property == colorop->lut3d_interpolation_property) {
+		*val = colorop->lut3d_interpolation;
 	} else if (property == colorop->data_property) {
 		*val = (state->data) ? state->data->base.id : 0;
 	} else {
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 38bbe57db348..528de1829350 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -67,6 +67,7 @@ static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_LUT, "1D LUT" },
 	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
 	{ DRM_COLOROP_MULTIPLIER, "Multiplier"},
+	{ DRM_COLOROP_3D_LUT, "3D LUT"},
 };
 
 static const char * const colorop_curve_1d_type_names[] = {
@@ -82,6 +83,11 @@ static const struct drm_prop_enum_list drm_colorop_lut1d_interpolation_list[] =
 	{ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR, "Linear" },
 };
 
+
+static const struct drm_prop_enum_list drm_colorop_lut3d_interpolation_list[] = {
+	{ DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL, "Tetrahedral" },
+};
+
 /* Init Helpers */
 
 static int drm_plane_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
@@ -381,6 +387,51 @@ int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colo
 }
 EXPORT_SYMBOL(drm_plane_colorop_mult_init);
 
+int drm_plane_colorop_3dlut_init(struct drm_device *dev, struct drm_colorop *colorop,
+				 struct drm_plane *plane,
+				 uint32_t lut_size,
+				 enum drm_colorop_lut3d_interpolation_type interpolation,
+				 uint32_t flags)
+{
+	struct drm_property *prop;
+	int ret;
+
+	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_3D_LUT, flags);
+	if (ret)
+		return ret;
+
+	/* LUT size */
+	prop = drm_property_create_range(dev, DRM_MODE_PROP_IMMUTABLE  | DRM_MODE_PROP_ATOMIC,
+					 "SIZE", 0, UINT_MAX);
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->lut_size_property = prop;
+	drm_object_attach_property(&colorop->base, colorop->lut_size_property, lut_size);
+	colorop->lut_size = lut_size;
+
+	/* interpolation */
+	prop = drm_property_create_enum(dev, DRM_MODE_PROP_IMMUTABLE, "LUT3D_INTERPOLATION",
+					drm_colorop_lut3d_interpolation_list,
+					ARRAY_SIZE(drm_colorop_lut3d_interpolation_list));
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->lut3d_interpolation_property = prop;
+	drm_object_attach_property(&colorop->base, prop, interpolation);
+	colorop->lut3d_interpolation = interpolation;
+
+	/* data */
+	ret = drm_colorop_create_data_prop(dev, colorop);
+	if (ret)
+		return ret;
+
+	drm_colorop_reset(colorop);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_colorop_3dlut_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
@@ -473,7 +524,13 @@ static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_LUT] = "1D LUT",
 	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
 	[DRM_COLOROP_MULTIPLIER] = "Multiplier",
+	[DRM_COLOROP_3D_LUT] = "3D LUT",
 };
+
+static const char * const colorop_lu3d_interpolation_name[] = {
+	[DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL] = "Tetrahedral",
+};
+
 static const char * const colorop_lut1d_interpolation_name[] = {
 	[DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR] = "Linear",
 };
@@ -509,6 +566,21 @@ const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_inte
 	return colorop_lut1d_interpolation_name[type];
 }
 
+/**
+ * drm_get_colorop_lut3d_interpolation_name - return a string for interpolation type
+ * @type: interpolation type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_lut3d_interpolation_name(enum drm_colorop_lut3d_interpolation_type type)
+{
+	if (WARN_ON(type >= ARRAY_SIZE(colorop_lu3d_interpolation_name)))
+		return "unknown";
+
+	return colorop_lu3d_interpolation_name[type];
+}
+
 /**
  * drm_colorop_set_next_property - sets the next pointer
  * @colorop: drm colorop
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index e79472739d11..d5aebe46970f 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -283,6 +283,14 @@ struct drm_colorop {
 	 */
 	enum drm_colorop_lut1d_interpolation_type lut1d_interpolation;
 
+	/**
+	 * @lut3d_interpolation:
+	 *
+	 * Read-only
+	 * Interpolation for DRM_COLOROP_3D_LUT
+	 */
+	enum drm_colorop_lut3d_interpolation_type lut3d_interpolation;
+
 	/**
 	 * @lut1d_interpolation_property:
 	 *
@@ -311,6 +319,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *lut_size_property;
 
+	/**
+	 * @lut3d_interpolation_property:
+	 *
+	 * Read-only property for DRM_COLOROP_3D_LUT interpolation
+	 */
+	struct drm_property *lut3d_interpolation_property;
+
 	/**
 	 * @data_property:
 	 *
@@ -365,6 +380,11 @@ int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *c
 				   struct drm_plane *plane, uint32_t flags);
 int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
 				struct drm_plane *plane, uint32_t flags);
+int drm_plane_colorop_3dlut_init(struct drm_device *dev, struct drm_colorop *colorop,
+				 struct drm_plane *plane,
+				 uint32_t lut_size,
+				 enum drm_colorop_lut3d_interpolation_type interpolation,
+				 uint32_t flags);
 
 struct drm_colorop_state *
 drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
@@ -415,6 +435,7 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type);
  */
 const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
 const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_interpolation_type type);
+const char *drm_get_colorop_lut3d_interpolation_name(enum drm_colorop_lut3d_interpolation_type type);
 
 void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
 
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index d76c8ffe5408..88fafbdeb2a2 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -930,6 +930,39 @@ enum drm_colorop_type {
 	 * property.
 	 */
 	DRM_COLOROP_MULTIPLIER,
+
+	/**
+	 * @DRM_COLOROP_3D_LUT:
+	 *
+	 * enum string "3D LUT"
+	 *
+	 * A 3D LUT of &drm_color_lut entries,
+	 * packed into a blob via the DATA property. The driver's expected
+	 * LUT size is advertised via the SIZE property, i.e., a 3D LUT with
+	 * 17x17x17 entries will have SIZE set to 17.
+	 *
+	 * The DATA blob is a 3D array of struct drm_color_lut with dimension
+	 * length of "lut_size".
+	 * The LUT elements are traversed like so:
+	 *
+	 *   for R in range 0..n
+	 *     for G in range 0..n
+	 *       for B in range 0..n
+	 *         color = lut3d[R][G][B]
+	 */
+	DRM_COLOROP_3D_LUT,
+};
+
+/**
+ * enum drm_colorop_lut3d_interpolation_type - type of 3DLUT interpolation
+ */
+enum drm_colorop_lut3d_interpolation_type {
+	/**
+	 * @DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL:
+	 *
+	 * Tetrahedral 3DLUT interpolation
+	 */
+	DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
 };
 
 /**
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (40 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 41/43] drm/colorop: Add 3D LUT support to color pipeline Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-13  0:52   ` Melissa Wen
  2025-04-30  1:11 ` [PATCH V9 43/43] drm/amd/display: Add AMD color pipeline doc Alex Hung
                   ` (2 subsequent siblings)
  44 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

This adds support for a 3D LUT.

The color pipeline now consists of the following colorops:
1. 1D curve colorop
2. Multiplier
3. 3x4 CTM
4. 1D curve colorop
5. 1D LUT
6. 3D LUT
7. 1D curve colorop
8. 1D LUT

Signed-off-by: Alex Hung <alex.hung@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
V9:
 - Return a value in __set_dm_plane_colorop_3dlut

v8:
 - Set initialized to 0 and return when drm_lut3d_size is 0 (Harry Wentland)
 - Rework tf->type = TF_TYPE_BYPASS for shaper (Harry Wentland & Leo Li)

v7:
 - Simplify 3D LUT according to drm_colorop changes (Simon Ser)

 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 94 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 20 ++++
 2 files changed, 114 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 313716f2003f..dfdd3f557570 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -1293,6 +1293,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
 	const struct drm_color_lut *shaper_lut;
 	struct drm_device *dev = colorop->dev;
+	bool enabled = false;
 	uint32_t shaper_size;
 	int i = 0, ret = 0;
 
@@ -1314,6 +1315,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 		ret = __set_output_tf(tf, 0, 0, false);
 		if (ret)
 			return ret;
+		enabled = true;
 	}
 
 	/* 1D LUT - SHAPER LUT */
@@ -1345,12 +1347,93 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 			ret = __set_output_tf(tf, shaper_lut, shaper_size, false);
 			if (ret)
 				return ret;
+			enabled = true;
 		}
 	}
 
+	if (!enabled)
+		tf->type = TF_TYPE_BYPASS;
+
 	return 0;
 }
 
+/* __set_colorop_3dlut - set DRM 3D LUT to DC stream
+ * @drm_lut3d: user 3D LUT
+ * @drm_lut3d_size: size of 3D LUT
+ * @lut3d: DC 3D LUT
+ *
+ * Map user 3D LUT data to DC 3D LUT and all necessary bits to program it
+ * on DCN accordingly.
+ */
+static void __set_colorop_3dlut(const struct drm_color_lut *drm_lut3d,
+				uint32_t drm_lut3d_size,
+				struct dc_3dlut *lut)
+{
+	if (!drm_lut3d_size) {
+		lut->state.bits.initialized = 0;
+		return;
+	}
+
+	/* Only supports 17x17x17 3D LUT (12-bit) now */
+	lut->lut_3d.use_12bits = true;
+	lut->lut_3d.use_tetrahedral_9 = false;
+
+	lut->state.bits.initialized = 1;
+	__drm_3dlut_to_dc_3dlut(drm_lut3d, drm_lut3d_size, &lut->lut_3d,
+				lut->lut_3d.use_tetrahedral_9, 12);
+
+}
+
+static int
+__set_dm_plane_colorop_3dlut(struct drm_plane_state *plane_state,
+			     struct dc_plane_state *dc_plane_state,
+			     struct drm_colorop *colorop)
+{
+	struct drm_colorop *old_colorop;
+	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
+	struct drm_atomic_state *state = plane_state->state;
+	const struct amdgpu_device *adev = drm_to_adev(colorop->dev);
+	const struct drm_device *dev = colorop->dev;
+	const struct drm_color_lut *lut3d;
+	uint32_t lut3d_size;
+	int i = 0, ret = 0;
+
+	/* 3D LUT */
+	old_colorop = colorop;
+	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+		if (new_colorop_state->colorop == old_colorop &&
+		    new_colorop_state->colorop->type == DRM_COLOROP_3D_LUT) {
+			colorop_state = new_colorop_state;
+			break;
+		}
+	}
+
+	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_3D_LUT) {
+		if (!adev->dm.dc->caps.color.dpp.hw_3d_lut) {
+			drm_dbg(dev, "3D LUT is not supported by hardware\n");
+			return -EINVAL;
+		}
+
+		drm_dbg(dev, "3D LUT colorop with ID: %d\n", colorop->base.id);
+		lut3d = __extract_blob_lut(colorop_state->data, &lut3d_size);
+		lut3d_size = lut3d != NULL ? lut3d_size : 0;
+		__set_colorop_3dlut(lut3d, lut3d_size, &dc_plane_state->lut3d_func);
+
+		/* 3D LUT requires shaper. If shaper colorop is bypassed, enable shaper curve
+		 * with TRANSFER_FUNCTION_LINEAR
+		 */
+		if (tf->type == TF_TYPE_BYPASS) {
+			tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+			tf->tf = TRANSFER_FUNCTION_LINEAR;
+			tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+			ret = __set_output_tf(tf, NULL, 0, false);
+		}
+	}
+
+	return ret;
+}
+
 static int
 __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
 			     struct dc_plane_state *dc_plane_state,
@@ -1522,6 +1605,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
 	if (!colorop)
 		return -EINVAL;
 
+	/* 3D LUT */
+	colorop = colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no 3D LUT colorop found\n");
+		return -EINVAL;
+	}
+
+	ret = __set_dm_plane_colorop_3dlut(plane_state, dc_plane_state, colorop);
+	if (ret)
+		return ret;
+
 	/* 1D Curve & LUT - BLND TF & LUT */
 	colorop = colorop->next;
 	if (!colorop) {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 10b3e3906461..e90774294971 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -49,6 +49,8 @@ const u64 amdgpu_dm_supported_blnd_tfs =
 
 #define MAX_COLOR_PIPELINE_OPS 10
 
+#define LUT3D_SIZE		17
+
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
 {
 	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
@@ -145,6 +147,24 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	i++;
 
+	/* 3D LUT */
+	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!ops[i]) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	ret = drm_plane_colorop_3dlut_init(dev, ops[i], plane, LUT3D_SIZE,
+				     DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
+				     DRM_COLOROP_FLAG_ALLOW_BYPASS);
+	if (ret)
+		goto cleanup;
+
+	drm_colorop_set_next_property(ops[i-1], ops[i]);
+
+	i++;
+
 	/* 1D curve - BLND TF */
 	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!ops[i]) {
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* [PATCH V9 43/43] drm/amd/display: Add AMD color pipeline doc
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (41 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop Alex Hung
@ 2025-04-30  1:11 ` Alex Hung
  2025-05-13 15:59 ` [PATCH V9 00/43] Color Pipeline API w/ VKMS Melissa Wen
  2025-05-15  9:45 ` Simon Ser
  44 siblings, 0 replies; 82+ messages in thread
From: Alex Hung @ 2025-04-30  1:11 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, alex.hung, leo.liu, ville.syrjala,
	pekka.paalanen, contact, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

From: Harry Wentland <harry.wentland@amd.com>

Add kernel doc for AMD color pipeline.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 122 +++++++++++++++---
 1 file changed, 102 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index dfdd3f557570..ef30ea707077 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -33,6 +33,32 @@
 /**
  * DOC: overview
  *
+ * We have three types of color management in the AMD display driver.
+ * 1. the legacy &drm_crtc DEGAMMA, CTM, and GAMMA properties
+ * 2. AMD driver private color management on &drm_plane and &drm_crtc
+ * 3. AMD plane color pipeline
+ *
+ * The CRTC properties are the original color management. When they were
+ * implemented per-plane color management was not a thing yet. Because
+ * of that we could get away with plumbing the DEGAMMA and CTM
+ * properties to pre-blending HW functions. This is incompatible with
+ * per-plane color management, such as via the AMD private properties or
+ * the new drm_plane color pipeline. The only compatible CRTC property
+ * with per-plane color management is the GAMMA property as it is
+ * applied post-blending.
+ *
+ * The AMD driver private color management properties are only exposed
+ * when the kernel is built explicitly with -DAMD_PRIVATE_COLOR. They
+ * are temporary building blocks on the path to full-fledged &drm_plane
+ * and &drm_crtc color pipelines and lay the driver's groundwork for the
+ * color pipelines.
+ *
+ * The AMD plane color pipeline describes AMD's &drm_colorops via the
+ * &drm_plane's COLOR_PIPELINE property.
+ *
+ * drm_crtc Properties
+ * -------------------
+ *
  * The DC interface to HW gives us the following color management blocks
  * per pipe (surface):
  *
@@ -43,33 +69,89 @@
  * - Surface regamma LUT (normalized)
  * - Output CSC (normalized)
  *
- * But these aren't a direct mapping to DRM color properties. The current DRM
- * interface exposes CRTC degamma, CRTC CTM and CRTC regamma while our hardware
- * is essentially giving:
+ * But these aren't a direct mapping to DRM color properties. The
+ * current DRM interface exposes CRTC degamma, CRTC CTM and CRTC regamma
+ * while our hardware is essentially giving:
  *
  * Plane CTM -> Plane degamma -> Plane CTM -> Plane regamma -> Plane CTM
  *
- * The input gamma LUT block isn't really applicable here since it operates
- * on the actual input data itself rather than the HW fp representation. The
- * input and output CSC blocks are technically available to use as part of
- * the DC interface but are typically used internally by DC for conversions
- * between color spaces. These could be blended together with user
- * adjustments in the future but for now these should remain untouched.
+ * The input gamma LUT block isn't really applicable here since it
+ * operates on the actual input data itself rather than the HW fp
+ * representation. The input and output CSC blocks are technically
+ * available to use as part of the DC interface but are typically used
+ * internally by DC for conversions between color spaces. These could be
+ * blended together with user adjustments in the future but for now
+ * these should remain untouched.
+ *
+ * The pipe blending also happens after these blocks so we don't
+ * actually support any CRTC props with correct blending with multiple
+ * planes - but we can still support CRTC color management properties in
+ * DM in most single plane cases correctly with clever management of the
+ * DC interface in DM.
+ *
+ * As per DRM documentation, blocks should be in hardware bypass when
+ * their respective property is set to NULL. A linear DGM/RGM LUT should
+ * also considered as putting the respective block into bypass mode.
+ *
+ * This means that the following configuration is assumed to be the
+ * default:
+ *
+ * Plane DGM Bypass -> Plane CTM Bypass -> Plane RGM Bypass -> ... CRTC
+ * DGM Bypass -> CRTC CTM Bypass -> CRTC RGM Bypass
+ *
+ * AMD Private Color Management on drm_plane
+ * -----------------------------------------
+ *
+ * The AMD private color management properties on a &drm_plane are:
+ *
+ * - AMD_PLANE_DEGAMMA_LUT
+ * - AMD_PLANE_DEGAMMA_LUT_SIZE
+ * - AMD_PLANE_DEGAMMA_TF
+ * - AMD_PLANE_HDR_MULT
+ * - AMD_PLANE_CTM
+ * - AMD_PLANE_SHAPER_LUT
+ * - AMD_PLANE_SHAPER_LUT_SIZE
+ * - AMD_PLANE_SHAPER_TF
+ * - AMD_PLANE_LUT3D
+ * - AMD_PLANE_LUT3D_SIZE
+ * - AMD_PLANE_BLEND_LUT
+ * - AMD_PLANE_BLEND_LUT_SIZE
+ * - AMD_PLANE_BLEND_TF
+ *
+ * The AMD private color management property on a &drm_crtc is:
+ *
+ * - AMD_CRTC_REGAMMA_TF
+ *
+ * Use of these properties is discouraged.
+ *
+ * AMD plane color pipeline
+ * ------------------------
+ *
+ * The AMD &drm_plane color pipeline is advertised for DCN generations
+ * 3.0 and newer. It exposes these elements in this order:
+ *
+ * 1. 1D curve colorop
+ * 2. Multiplier
+ * 3. 3x4 CTM
+ * 4. 1D curve colorop
+ * 5. 1D LUT
+ * 6. 3D LUT
+ * 7. 1D curve colorop
+ * 8. 1D LUT
+ *
+ * The multiplier (#2) is a simple multiplier that is applied to all
+ * channels.
+ *
+ * The 3x4 CTM (#3) is a simple 3x4 matrix.
  *
- * The pipe blending also happens after these blocks so we don't actually
- * support any CRTC props with correct blending with multiple planes - but we
- * can still support CRTC color management properties in DM in most single
- * plane cases correctly with clever management of the DC interface in DM.
+ * #1, and #7 are non-linear to linear curves. #4 is a linear to
+ * non-linear curve. They support sRGB, PQ, and BT.709/BT.2020 EOTFs or
+ * their inverse.
  *
- * As per DRM documentation, blocks should be in hardware bypass when their
- * respective property is set to NULL. A linear DGM/RGM LUT should also
- * considered as putting the respective block into bypass mode.
+ * The 1D LUTs (#5 and #8) are plain 4096 entry LUTs.
  *
- * This means that the following
- * configuration is assumed to be the default:
+ * The 3DLUT (#6) is a tetrahedrally interpolated 17 cube LUT.
  *
- * Plane DGM Bypass -> Plane CTM Bypass -> Plane RGM Bypass -> ...
- * CRTC DGM Bypass -> CRTC CTM Bypass -> CRTC RGM Bypass
  */
 
 #define MAX_DRM_LUT_VALUE 0xFFFF
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs
  2025-04-30  1:10 ` [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs Alex Hung
@ 2025-05-07 10:03   ` kernel test robot
  0 siblings, 0 replies; 82+ messages in thread
From: kernel test robot @ 2025-05-07 10:03 UTC (permalink / raw)
  To: Alex Hung, dri-devel, amd-gfx
  Cc: oe-kbuild-all, wayland-devel, harry.wentland, alex.hung, leo.liu,
	ville.syrjala, pekka.paalanen, contact, mwen, jadahl,
	sebastian.wick, shashank.sharma, agoins, joshua, mdaenzer,
	aleixpol, xaver.hugl, victoria, daniel, uma.shankar, quic_naseer,
	quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Hi Alex,

kernel test robot noticed the following build errors:

[auto build test ERROR on drm-exynos/exynos-drm-next]
[also build test ERROR on next-20250507]
[cannot apply to linus/master v6.15-rc5]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Alex-Hung/drm-Add-helper-for-conversion-from-signed-magnitude/20250430-092409
base:   https://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos.git exynos-drm-next
patch link:    https://lore.kernel.org/r/20250430011115.223996-16-alex.hung%40amd.com
patch subject: [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs
config: x86_64-randconfig-073-20250502 (https://download.01.org/0day-ci/archive/20250507/202505071750.LFUxqdSH-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250507/202505071750.LFUxqdSH-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202505071750.LFUxqdSH-lkp@intel.com/

All errors (new ones prefixed by >>, old ones prefixed by <<):

WARNING: modpost: missing MODULE_DESCRIPTION() in mm/kasan/kasan_test.o
WARNING: modpost: missing MODULE_DESCRIPTION() in lib/tests/slub_kunit.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/fpga/tests/fpga-mgr-test.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/fpga/tests/fpga-bridge-test.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/fpga/tests/fpga-region-test.o
>> ERROR: modpost: "srgb_eotf" [drivers/gpu/drm/vkms/tests/vkms_color_test.ko] undefined!
>> ERROR: modpost: "srgb_inv_eotf" [drivers/gpu/drm/vkms/tests/vkms_color_test.ko] undefined!
>> ERROR: modpost: "linear_eotf" [drivers/gpu/drm/vkms/tests/vkms_color_test.ko] undefined!

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop
  2025-04-30  1:11 ` [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop Alex Hung
@ 2025-05-07 12:08   ` kernel test robot
  0 siblings, 0 replies; 82+ messages in thread
From: kernel test robot @ 2025-05-07 12:08 UTC (permalink / raw)
  To: Alex Hung, dri-devel, amd-gfx
  Cc: oe-kbuild-all, wayland-devel, harry.wentland, alex.hung, leo.liu,
	ville.syrjala, pekka.paalanen, contact, mwen, jadahl,
	sebastian.wick, shashank.sharma, agoins, joshua, mdaenzer,
	aleixpol, xaver.hugl, victoria, daniel, uma.shankar, quic_naseer,
	quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Hi Alex,

kernel test robot noticed the following build warnings:

[auto build test WARNING on drm-exynos/exynos-drm-next]
[also build test WARNING on next-20250507]
[cannot apply to linus/master drm/drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-misc/drm-misc-next drm-tip/drm-tip v6.15-rc5]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Alex-Hung/drm-Add-helper-for-conversion-from-signed-magnitude/20250430-092409
base:   https://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos.git exynos-drm-next
patch link:    https://lore.kernel.org/r/20250430011115.223996-36-alex.hung%40amd.com
patch subject: [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop
config: parisc-randconfig-r122-20250501 (https://download.01.org/0day-ci/archive/20250507/202505071913.raR4oCjA-lkp@intel.com/config)
compiler: hppa-linux-gcc (GCC) 8.5.0
reproduce: (https://download.01.org/0day-ci/archive/20250507/202505071913.raR4oCjA-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202505071913.raR4oCjA-lkp@intel.com/

All warnings (new ones prefixed by >>):

   In file included from include/linux/device.h:15,
                    from include/drm/drm_print.h:31,
                    from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_ring.h:29,
                    from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_ctx.h:29,
                    from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:43,
                    from drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:25:
   drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c: In function '__set_dm_plane_colorop_3x4_matrix':
>> drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:41:22: warning: format '%lu' expects argument of type 'long unsigned int', but argument 3 has type 'size_t' {aka 'const unsigned int'} [-Wformat=]
    #define dev_fmt(fmt) "amdgpu: " fmt
                         ^~~~~~~~~~
   include/linux/dev_printk.h:110:16: note: in definition of macro 'dev_printk_index_wrap'
      _p_func(dev, fmt, ##__VA_ARGS__);   \
                   ^~~
   include/linux/dev_printk.h:156:54: note: in expansion of macro 'dev_fmt'
     dev_printk_index_wrap(_dev_warn, KERN_WARNING, dev, dev_fmt(fmt), ##__VA_ARGS__)
                                                         ^~~~~~~
   include/drm/drm_print.h:595:2: note: in expansion of macro 'dev_warn'
     dev_##level##type(__drm_to_dev(drm), "[drm] " fmt, ##__VA_ARGS__)
     ^~~~
   include/drm/drm_print.h:605:2: note: in expansion of macro '__drm_printk'
     __drm_printk((drm), warn,, fmt, ##__VA_ARGS__)
     ^~~~~~~~~~~~
   drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1246:4: note: in expansion of macro 'drm_warn'
       drm_warn(dev, "blob->length (%lu) isn't equal to drm_color_ctm_3x4 (%zu)\n",
       ^~~~~~~~
   drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1246:35: note: format string is defined here
       drm_warn(dev, "blob->length (%lu) isn't equal to drm_color_ctm_3x4 (%zu)\n",
                                    ~~^
                                    %u


vim +41 drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h

539489fc91ea77 Aurabindo Pillai 2020-04-08  40  
539489fc91ea77 Aurabindo Pillai 2020-04-08 @41  #define dev_fmt(fmt) "amdgpu: " fmt
539489fc91ea77 Aurabindo Pillai 2020-04-08  42  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 22/43] drm/colorop: define a new macro for_each_new_colorop_in_state
  2025-04-30  1:10 ` [PATCH V9 22/43] drm/colorop: define a new macro for_each_new_colorop_in_state Alex Hung
@ 2025-05-12 18:32   ` Melissa Wen
  0 siblings, 0 replies; 82+ messages in thread
From: Melissa Wen @ 2025-05-12 18:32 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 04/29, Alex Hung wrote:
> Create a new macro for_each_new_colorop_in_state to access new
> drm_colorop_state updated from uapi.
> 
> Reviewed-by: Simon Ser <contact@emersion.fr>
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
>  include/drm/drm_atomic.h | 20 ++++++++++++++++++++
>  1 file changed, 20 insertions(+)
> 
> diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
> index 4a30ed8ab1a8..4b3bd459a1eb 100644
> --- a/include/drm/drm_atomic.h
> +++ b/include/drm/drm_atomic.h
> @@ -1069,6 +1069,26 @@ void drm_state_dump(struct drm_device *dev, struct drm_printer *p);
>  			      (new_colorop_state) = (__state)->colorops[__i].new_state, 1))
>  
>  
> +/**
> + * for_each_new_colorop_in_state - iterate over all colorops in an atomic update
> + * @__state: &struct drm_atomic_state pointer
> + * @colorop: &struct drm_colorop iteration cursor
> + * @new_colorop_state: &struct drm_colorop_state iteration cursor for the new state
> + * @__i: int iteration cursor, for macro-internal use
> + *
> + * This iterates over all colorops in an atomic update, tracking new state. This is
> + * useful is useful in places where the state delta needs to be considered, for

nit: "is useful" duplication

> + * example in atomic check functions.
> + */
> +#define for_each_new_colorop_in_state(__state, colorop, new_colorop_state, __i) \
> +	for ((__i) = 0;							\
> +	     (__i) < (__state)->dev->mode_config.num_colorop;	\
> +	     (__i)++)							\
> +		for_each_if ((__state)->colorops[__i].ptr &&		\
> +			     ((colorop) = (__state)->colorops[__i].ptr,	\
> +			      (void)(colorop) /* Only to avoid unused-but-set-variable warning */, \
> +			      (new_colorop_state) = (__state)->colorops[__i].new_state, 1))
> +
>  /**
>   * for_each_oldnew_plane_in_state - iterate over all planes in an atomic update
>   * @__state: &struct drm_atomic_state pointer
> -- 
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 24/43] drm/amd/display: Add bypass COLOR PIPELINE
  2025-04-30  1:10 ` [PATCH V9 24/43] drm/amd/display: Add bypass COLOR PIPELINE Alex Hung
@ 2025-05-12 18:41   ` Melissa Wen
  0 siblings, 0 replies; 82+ messages in thread
From: Melissa Wen @ 2025-05-12 18:41 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 04/29, Alex Hung wrote:
> From: Harry Wentland <harry.wentland@amd.com>
> 
> Add the default Bypass pipeline and ensure it passes the
> kms_colorop test plane-XR30-XR30-bypass.
> 
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 19 +++++++++++++++++++
>  1 file changed, 19 insertions(+)
> 
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
> index 3e0f45f1711c..fb6f07603050 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
> @@ -1784,6 +1784,20 @@ dm_atomic_plane_get_property(struct drm_plane *plane,
>  }
>  #endif
> 
> +#define MAX_COLOR_PIPELINES 5
> +
> +static int
> +dm_plane_init_colorops(struct drm_plane *plane)
> +{
> +	struct drm_prop_enum_list pipelines[MAX_COLOR_PIPELINES];
> +	int len = 0;
> +
> +	/* Create COLOR_PIPELINE property and attach */
> +	drm_plane_create_color_pipeline_property(plane, pipelines, len);
> +
> +	return 0;
> +}
> +

This function is unused if AMD_PRIVATE_COLOR (warning).
I think you can add an #else just before the define MAX_COLOR_PIPELINES
(replacing the above #endif there, and move the #endif to here to solve
it.

Melissa

>  static const struct drm_plane_funcs dm_plane_funcs = {
>  	.update_plane	= drm_atomic_helper_update_plane,
>  	.disable_plane	= drm_atomic_helper_disable_plane,
> @@ -1890,7 +1904,12 @@ int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm,
>  
>  #ifdef AMD_PRIVATE_COLOR
>  	dm_atomic_plane_attach_color_mgmt_properties(dm, plane);
> +#else
> +	res = dm_plane_init_colorops(plane);
> +	if (res)
> +		return res;
>  #endif
> +
>  	/* Create (reset) the plane state */
>  	if (plane->funcs->reset)
>  		plane->funcs->reset(plane);
> -- 
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2025-04-30  1:10 ` [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block Alex Hung
@ 2025-05-12 18:50   ` Melissa Wen
  2025-05-13  7:42     ` Pekka Paalanen
  0 siblings, 1 reply; 82+ messages in thread
From: Melissa Wen @ 2025-05-12 18:50 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 04/29, Alex Hung wrote:
> Expose one 1D curve colorop with support for
> DRM_COLOROP_1D_CURVE_SRGB_EOTF and program HW to perform
> the sRGB transform when the colorop is not in bypass.
> 
> With this change the following IGT test passes:
> kms_colorop --run plane-XR30-XR30-srgb_eotf
> 
> The color pipeline now consists of a single colorop:
> 1. 1D curve colorop w/ sRGB EOTF
> 
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Co-developed-by: Harry Wentland <harry.wentland@amd.com>
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
> V9:
>  - Update function names by _plane_ (Chaitanya Kumar Borah)
>  - Update replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)
> 
> v8:
>  - Fix incorrect && by || in __set_colorop_in_tf_1d_curve (Leo Li)
> 
> v7:
>  - Fix checkpatch warnings
>   - Change switch "{ }" position
>   - Delete double ";"
>   - Delete "{ }" for single-line if-statement
>   - Add a new line at EOF
>   - Change SPDX-License-Identifier: GPL-2.0+ from // to /* */
> 
> v6:
>  - cleanup if colorop alloc or init fails
> 
>  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
>  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 86 +++++++++++++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 69 +++++++++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 ++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
>  5 files changed, 201 insertions(+), 1 deletion(-)
>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> 
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> index ab2a97e354da..46158d67ab12 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> @@ -38,7 +38,8 @@ AMDGPUDM = \
>  	amdgpu_dm_pp_smu.o \
>  	amdgpu_dm_psr.o \
>  	amdgpu_dm_replay.o \
> -	amdgpu_dm_wb.o
> +	amdgpu_dm_wb.o \
> +	amdgpu_dm_colorop.o
>  
>  ifdef CONFIG_DRM_AMD_DC_FP
>  AMDGPUDM += dc_fpu.o
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> index ebabfe3a512f..0b513ab5050f 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> @@ -668,6 +668,18 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf)
>  	}
>  }
>  
> +static enum dc_transfer_func_predefined
> +amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
> +{
> +	switch (tf) {
> +	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
> +	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
> +		return TRANSFER_FUNCTION_SRGB;
> +	default:
> +		return TRANSFER_FUNCTION_LINEAR;
> +	}
> +}
> +
>  static void __to_dc_lut3d_color(struct dc_rgb *rgb,
>  				const struct drm_color_lut lut,
>  				int bit_precision)
> @@ -1137,6 +1149,59 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state,
>  	return 0;
>  }
>  
> +static int
> +__set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
> +		       struct drm_colorop_state *colorop_state)
> +{
> +	struct dc_transfer_func *tf = &dc_plane_state->in_transfer_func;
> +	struct drm_colorop *colorop = colorop_state->colorop;
> +	struct drm_device *drm = colorop->dev;
> +
> +	if (colorop->type != DRM_COLOROP_1D_CURVE ||
> +	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
> +		return -EINVAL;
> +
> +	if (colorop_state->bypass) {
> +		tf->type = TF_TYPE_BYPASS;
> +		tf->tf = TRANSFER_FUNCTION_LINEAR;
> +		return 0;
> +	}
> +
> +	drm_dbg(drm, "Degamma colorop with ID: %d\n", colorop->base.id);
> +
> +	tf->type = TF_TYPE_PREDEFINED;
> +	tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
> +
> +	return 0;
> +}
> +
> +static int
> +__set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
> +			       struct dc_plane_state *dc_plane_state,
> +			       struct drm_colorop *colorop)
> +{
> +	struct drm_colorop *old_colorop;
> +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
> +	struct drm_atomic_state *state = plane_state->state;
> +	int i = 0;
> +
> +	old_colorop = colorop;
> +
> +	/* 1st op: 1d curve - degamma */
> +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
> +		if (new_colorop_state->colorop == old_colorop &&
> +		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
> +			colorop_state = new_colorop_state;
> +			break;
> +		}
> +	}
> +
> +	if (!colorop_state)
> +		return -EINVAL;
> +
> +	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);

I wonder what will happen if plane degamma isn't set, but CRTC degamma
LUT or legacy CRTC regamma LUT (with its implicity sRGB degamma) is used
together with other plane color ops.

I can imagine the mess, so I think CRTC degamma LUT and legacy CRTC
regamma LUT should be somehow entirely disabled (or rejected) if plane
color pipeline is in use.

Melissa

> +}
> +
>  static int
>  amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state,
>  				     struct dc_plane_state *dc_plane_state)
> @@ -1187,6 +1252,24 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state,
>  	return 0;
>  }
>  
> +static int
> +amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
> +		       struct dc_plane_state *dc_plane_state)
> +{
> +	struct drm_colorop *colorop = plane_state->color_pipeline;
> +	int ret;
> +
> +	/* 1D Curve - DEGAM TF */
> +	if (!colorop)
> +		return -EINVAL;
> +
> +	ret = __set_dm_plane_colorop_degamma(plane_state, dc_plane_state, colorop);
> +	if (ret)
> +		return ret;
> +
> +	return 0;
> +}
> +
>  /**
>   * amdgpu_dm_update_plane_color_mgmt: Maps DRM color management to DC plane.
>   * @crtc: amdgpu_dm crtc state
> @@ -1283,5 +1366,8 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc,
>  		dc_plane_state->input_csc_color_matrix.enable_adjustment = false;
>  	}
>  
> +	if (!amdgpu_dm_plane_set_colorop_properties(plane_state, dc_plane_state))
> +		return 0;
> +
>  	return amdgpu_dm_plane_set_color_properties(plane_state, dc_plane_state);
>  }
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> new file mode 100644
> index 000000000000..bc9980aad7d8
> --- /dev/null
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> @@ -0,0 +1,69 @@
> +// SPDX-License-Identifier: MIT
> +/*
> + * Copyright 2023 Advanced Micro Devices, Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a
> + * copy of this software and associated documentation files (the "Software"),
> + * to deal in the Software without restriction, including without limitation
> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> + * and/or sell copies of the Software, and to permit persons to whom the
> + * Software is furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
> + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
> + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
> + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
> + * OTHER DEALINGS IN THE SOFTWARE.
> + *
> + * Authors: AMD
> + *
> + */
> +
> +#include <drm/drm_print.h>
> +#include <drm/drm_plane.h>
> +#include <drm/drm_property.h>
> +#include <drm/drm_colorop.h>
> +
> +#include "amdgpu_dm_colorop.h"
> +
> +const u64 amdgpu_dm_supported_degam_tfs =
> +	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF);
> +
> +#define MAX_COLOR_PIPELINE_OPS 10
> +
> +int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
> +{
> +	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
> +	struct drm_device *dev = plane->dev;
> +	int ret;
> +	int i = 0;
> +
> +	memset(ops, 0, sizeof(ops));
> +
> +	/* 1D curve - DEGAM TF */
> +	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> +	if (!ops[i]) {
> +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> +		ret = -ENOMEM;
> +		goto cleanup;
> +	}
> +
> +	ret = drm_plane_colorop_curve_1d_init(dev, ops[i], plane, amdgpu_dm_supported_degam_tfs);
> +	if (ret)
> +		goto cleanup;
> +
> +	list->type = ops[i]->base.id;
> +	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", ops[i]->base.id);
> +
> +	return 0;
> +
> +cleanup:
> +	drm_colorop_pipeline_destroy(plane);
> +
> +	return ret;
> +}
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> new file mode 100644
> index 000000000000..3324e2a66079
> --- /dev/null
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> @@ -0,0 +1,34 @@
> +/* SPDX-License-Identifier: MIT */
> +/*
> + * Copyright 2023 Advanced Micro Devices, Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a
> + * copy of this software and associated documentation files (the "Software"),
> + * to deal in the Software without restriction, including without limitation
> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> + * and/or sell copies of the Software, and to permit persons to whom the
> + * Software is furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
> + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
> + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
> + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
> + * OTHER DEALINGS IN THE SOFTWARE.
> + *
> + * Authors: AMD
> + *
> + */
> +
> +#ifndef __AMDGPU_DM_COLOROP_H__
> +#define __AMDGPU_DM_COLOROP_H__
> +
> +extern const u64 amdgpu_dm_supported_degam_tfs;
> +
> +int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list);
> +
> +#endif /* __AMDGPU_DM_COLOROP_H__*/
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
> index feb4b99bf629..742ed56d1e58 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
> @@ -37,6 +37,7 @@
>  #include "amdgpu_display.h"
>  #include "amdgpu_dm_trace.h"
>  #include "amdgpu_dm_plane.h"
> +#include "amdgpu_dm_colorop.h"
>  #include "gc/gc_11_0_0_offset.h"
>  #include "gc/gc_11_0_0_sh_mask.h"
>  
> @@ -1791,10 +1792,19 @@ dm_plane_init_colorops(struct drm_plane *plane)
>  {
>  	struct drm_prop_enum_list pipelines[MAX_COLOR_PIPELINES];
>  	int len = 0;
> +	int ret;
>  
>  	if (plane->type == DRM_PLANE_TYPE_CURSOR)
>  		return 0;
>  
> +	/* initialize default pipeline */
> +	ret = amdgpu_dm_initialize_default_pipeline(plane, &(pipelines[len]));
> +	if (ret) {
> +		DRM_ERROR("Failed to create color pipeline for plane %d: %d\n", plane->base.id, ret);
> +		return ret;
> +	}
> +	len++;
> +
>  	/* Create COLOR_PIPELINE property and attach */
>  	drm_plane_create_color_pipeline_property(plane, pipelines, len);
>  
> -- 
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 32/43] drm/amd/display: Add support for BT.709 and BT.2020 TFs
  2025-04-30  1:11 ` [PATCH V9 32/43] drm/amd/display: Add support for BT.709 and BT.2020 TFs Alex Hung
@ 2025-05-12 19:37   ` Melissa Wen
  0 siblings, 0 replies; 82+ messages in thread
From: Melissa Wen @ 2025-05-12 19:37 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 04/29, Alex Hung wrote:
> From: Harry Wentland <harry.wentland@amd.com>
> 
> This adds support for the BT.709/BT.2020 transfer functions
> on all current 1D curve plane colorops, i.e., on DEGAM, SHAPER,
> and BLND blocks.
> 
> With this change the following IGT subtests pass:
> kms_colorop --run plane-XR30-XR30-bt2020_inv_oetf
> kms_colorop --run plane-XR30-XR30-bt2020_oetf
> 
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
> V9:
>  - Move DRM_COLOROP_1D_CURVE_BT2020_* from middle to end
> 
> v8: 
>  - Move BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) in amdgpu_dm_supported_blnd_tfs
>    to "drm/amd/display: Enable support for PQ 125 EOTF and Inverse" (Leo Li)
> 
>  .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c   | 11 ++++++++---
>  .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  9 ++++++---
>  2 files changed, 14 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> index 63044e0296cb..f645f9ded95f 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> @@ -679,6 +679,9 @@ amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
>  	case DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
>  	case DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF:
>  		return TRANSFER_FUNCTION_PQ;
> +	case DRM_COLOROP_1D_CURVE_BT2020_INV_OETF:
> +	case DRM_COLOROP_1D_CURVE_BT2020_OETF:
> +		return TRANSFER_FUNCTION_BT709;
>  	default:
>  		return TRANSFER_FUNCTION_LINEAR;
>  	}
> @@ -1284,8 +1287,10 @@ __set_colorop_1d_curve_blend_tf_lut(struct dc_plane_state *dc_plane_state,
>  	const struct drm_color_lut *blend_lut;
>  	uint32_t blend_size = 0;
>  
> -	if (colorop->type != DRM_COLOROP_1D_CURVE &&
> -	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
> +	if (colorop->type != DRM_COLOROP_1D_CURVE)
> +		return -EINVAL;
> +
> +	if (!(BIT(colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs))
>  		return -EINVAL;
>  
>  	if (colorop_state->bypass) {
> @@ -1321,7 +1326,7 @@ __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
>  	/* 3nd op: 1d curve - blend */
>  	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
>  		if (new_colorop_state->colorop == old_colorop &&
> -		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
> +		    (BIT(new_colorop_state->curve_1d_type) & amdgpu_dm_supported_blnd_tfs)) {

nitpick: I think these changes to check amdgpu_dm_supported_*_tfs should
happen earlier, together with degamma and shaper curves in the patch 30.

Melissa

>  			colorop_state = new_colorop_state;
>  			break;
>  		}
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> index aa753b0d6f13..b989e1ca19e2 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> @@ -33,15 +33,18 @@
>  
>  const u64 amdgpu_dm_supported_degam_tfs =
>  	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
> -	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF);
> +	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) |
> +	BIT(DRM_COLOROP_1D_CURVE_BT2020_INV_OETF);
>  
>  const u64 amdgpu_dm_supported_shaper_tfs =
>  	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF) |
> -	BIT(DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF);
> +	BIT(DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF) |
> +	BIT(DRM_COLOROP_1D_CURVE_BT2020_OETF);
>  
>  const u64 amdgpu_dm_supported_blnd_tfs =
>  	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
> -	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF);
> +	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) |
> +	BIT(DRM_COLOROP_1D_CURVE_BT2020_INV_OETF);
>  
>  #define MAX_COLOR_PIPELINE_OPS 10
>  
> -- 
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 36/43] drm/colorop: Add mutliplier type
  2025-04-30  1:11 ` [PATCH V9 36/43] drm/colorop: Add mutliplier type Alex Hung
@ 2025-05-12 20:02   ` Melissa Wen
  0 siblings, 0 replies; 82+ messages in thread
From: Melissa Wen @ 2025-05-12 20:02 UTC (permalink / raw)
  To: Alex Hung, dri-devel, amd-gfx
  Cc: wayland-devel, harry.wentland, leo.liu, ville.syrjala,
	pekka.paalanen, contact, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, xaver.hugl, victoria, daniel,
	uma.shankar, quic_naseer, quic_cbraga, quic_abhinavk, marcan,
	Liviu.Dudau, sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Daniel Stone



On 29/04/2025 22:11, Alex Hung wrote:

Typo in the commit msg: mutliplier -> multiplier
> This introduces a new drm_colorop_type: DRM_COLOROP_MULTIPLIER.
>
> It's a simple multiplier to all pixel values. The value is
> specified via a S31.32 fixed point provided via the
> "MULTIPLIER" property.
>
> Reviewed-by: Simon Ser <contact@emersion.fr>
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
> V9:
>   - Update function names by _plane_ (Chaitanya Kumar Borah)
>
> v7:
>   - Modify size_property to lut_size_property
>
>   drivers/gpu/drm/drm_atomic.c      |  3 +++
>   drivers/gpu/drm/drm_atomic_uapi.c |  4 ++++
>   drivers/gpu/drm/drm_colorop.c     | 33 +++++++++++++++++++++++++++++++
>   include/drm/drm_colorop.h         | 16 +++++++++++++++
>   include/uapi/drm/drm_mode.h       | 11 +++++++++++
>   5 files changed, 67 insertions(+)
>
> diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
> index 1dda90554e46..907ca790689f 100644
> --- a/drivers/gpu/drm/drm_atomic.c
> +++ b/drivers/gpu/drm/drm_atomic.c
> @@ -800,6 +800,9 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
>   	case DRM_COLOROP_CTM_3X4:
>   		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
>   		break;
> +	case DRM_COLOROP_MULTIPLIER:
> +		drm_printf(p, "\tmultiplier=%llu\n", state->multiplier);
> +		break;
>   	default:
>   		break;
>   	}
> diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
> index 0cd250db3981..a7f1d75bb4ea 100644
> --- a/drivers/gpu/drm/drm_atomic_uapi.c
> +++ b/drivers/gpu/drm/drm_atomic_uapi.c
> @@ -739,6 +739,8 @@ static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
>   		state->bypass = val;
>   	} else if (property == colorop->curve_1d_type_property) {
>   		state->curve_1d_type = val;
> +	} else if (property == colorop->multiplier_property) {
> +		state->multiplier = val;
>   	} else if (property == colorop->data_property) {
>   		return drm_atomic_color_set_data_property(colorop, state,
>   							  property, val);
> @@ -764,6 +766,8 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
>   		*val = state->bypass;
>   	} else if (property == colorop->curve_1d_type_property) {
>   		*val = state->curve_1d_type;
> +	} else if (property == colorop->multiplier_property) {
> +		*val = state->multiplier;
>   	} else if (property == colorop->lut_size_property) {
>   		*val = colorop->lut_size;
>   	} else if (property == colorop->data_property) {
> diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
> index 6c87f5b9f7f9..a3cf62c5e263 100644
> --- a/drivers/gpu/drm/drm_colorop.c
> +++ b/drivers/gpu/drm/drm_colorop.c
> @@ -66,6 +66,7 @@ static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
>   	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
>   	{ DRM_COLOROP_1D_LUT, "1D LUT" },
>   	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
> +	{ DRM_COLOROP_MULTIPLIER, "Multiplier"},
>   };
>   
>   static const char * const colorop_curve_1d_type_names[] = {
> @@ -327,6 +328,37 @@ int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *c
>   }
>   EXPORT_SYMBOL(drm_plane_colorop_ctm_3x4_init);
>   
> +/**
> + * drm_plane_colorop_mult_init - Initialize a DRM_COLOROP_MULTIPLIER
> + *
> + * @dev: DRM device
> + * @colorop: The drm_colorop object to initialize
> + * @plane: The associated drm_plane
> + * @return zero on success, -E value on failure
> + */
> +int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
> +				struct drm_plane *plane)
> +{
> +	struct drm_property *prop;
> +	int ret;
> +
> +	ret = drm_plane_colorop_init(dev, colorop, plane, DRM_COLOROP_MULTIPLIER);
> +	if (ret)
> +		return ret;
> +
> +	prop = drm_property_create_range(dev, DRM_MODE_PROP_ATOMIC, "MULTIPLIER", 0, U64_MAX);
> +	if (!prop)
> +		return -ENOMEM;
> +
> +	colorop->multiplier_property = prop;
> +	drm_object_attach_property(&colorop->base, colorop->multiplier_property, 0);
> +
> +	drm_colorop_reset(colorop);
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL(drm_plane_colorop_mult_init);
> +
>   static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
>   							struct drm_colorop_state *state)
>   {
> @@ -418,6 +450,7 @@ static const char * const colorop_type_name[] = {
>   	[DRM_COLOROP_1D_CURVE] = "1D Curve",
>   	[DRM_COLOROP_1D_LUT] = "1D LUT",
>   	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
> +	[DRM_COLOROP_MULTIPLIER] = "Multiplier",
>   };
>   
>   const char *drm_get_colorop_type_name(enum drm_colorop_type type)
> diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
> index 52e965577bd8..888184aef7a0 100644
> --- a/include/drm/drm_colorop.h
> +++ b/include/drm/drm_colorop.h
> @@ -146,6 +146,13 @@ struct drm_colorop_state {
>   	 */
>   	enum drm_colorop_curve_1d_type curve_1d_type;
>   
> +	/**
> +	 * @multiplier:
> +	 *
> +	 * Multiplier to 'gain' the plane. Format is S31.32 sign-magnitude.
> +	 */
> +	uint64_t multiplier;
> +
>   	/**
>   	 * @data:
>   	 *
> @@ -273,6 +280,13 @@ struct drm_colorop {
>   	 */
>   	struct drm_property *curve_1d_type_property;
>   
> +	/**
> +	 * @multiplier_property:
> +	 *
> +	 * Multiplier property for plane gain
> +	 */
> +	struct drm_property *multiplier_property;
> +
>   	/**
>   	 * @lut_size_property:
>   	 *
> @@ -330,6 +344,8 @@ int drm_plane_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_color
>   					struct drm_plane *plane, uint32_t lut_size);
>   int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
>   				   struct drm_plane *plane);
> +int drm_plane_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
> +				struct drm_plane *plane);
>   
>   struct drm_colorop_state *
>   drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
> diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
> index dde250dd7a51..fa5e1ddaf4f0 100644
> --- a/include/uapi/drm/drm_mode.h
> +++ b/include/uapi/drm/drm_mode.h
> @@ -919,6 +919,17 @@ enum drm_colorop_type {
>   	 * | B |   | 8  9  10 12 |   | B |
>   	 */
>   	DRM_COLOROP_CTM_3X4,
> +
> +	/**
> +	 * @DRM_COLOROP_MULTIPLIER:
> +	 *
> +	 * enum string "Multiplier"
> +	 *
> +	 * A simple multiplier, applied to all color values. The
> +	 * multiplier is specified as a S31.32 via the MULTIPLIER
> +	 * property.
> +	 */
> +	DRM_COLOROP_MULTIPLIER,
>   };
>   
>   /**


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop
  2025-04-30  1:11 ` [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop Alex Hung
@ 2025-05-13  0:52   ` Melissa Wen
  2025-05-13  3:29     ` Alex Hung
  0 siblings, 1 reply; 82+ messages in thread
From: Melissa Wen @ 2025-05-13  0:52 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 04/29, Alex Hung wrote:
> This adds support for a 3D LUT.
> 
> The color pipeline now consists of the following colorops:
> 1. 1D curve colorop
> 2. Multiplier
> 3. 3x4 CTM
> 4. 1D curve colorop
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D curve colorop
> 8. 1D LUT
> 
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
> V9:
>  - Return a value in __set_dm_plane_colorop_3dlut
> 
> v8:
>  - Set initialized to 0 and return when drm_lut3d_size is 0 (Harry Wentland)
>  - Rework tf->type = TF_TYPE_BYPASS for shaper (Harry Wentland & Leo Li)
> 
> v7:
>  - Simplify 3D LUT according to drm_colorop changes (Simon Ser)
> 
>  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 94 +++++++++++++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 20 ++++
>  2 files changed, 114 insertions(+)
> 
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> index 313716f2003f..dfdd3f557570 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> @@ -1293,6 +1293,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
>  	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
>  	const struct drm_color_lut *shaper_lut;
>  	struct drm_device *dev = colorop->dev;
> +	bool enabled = false;
>  	uint32_t shaper_size;
>  	int i = 0, ret = 0;
>  
> @@ -1314,6 +1315,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
>  		ret = __set_output_tf(tf, 0, 0, false);
>  		if (ret)
>  			return ret;
> +		enabled = true;
>  	}
>  
>  	/* 1D LUT - SHAPER LUT */
> @@ -1345,12 +1347,93 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
>  			ret = __set_output_tf(tf, shaper_lut, shaper_size, false);
>  			if (ret)
>  				return ret;
> +			enabled = true;
>  		}
>  	}
>  
> +	if (!enabled)
> +		tf->type = TF_TYPE_BYPASS;
> +
>  	return 0;
>  }
>  
> +/* __set_colorop_3dlut - set DRM 3D LUT to DC stream
> + * @drm_lut3d: user 3D LUT
> + * @drm_lut3d_size: size of 3D LUT
> + * @lut3d: DC 3D LUT
> + *
> + * Map user 3D LUT data to DC 3D LUT and all necessary bits to program it
> + * on DCN accordingly.
> + */
> +static void __set_colorop_3dlut(const struct drm_color_lut *drm_lut3d,
> +				uint32_t drm_lut3d_size,
> +				struct dc_3dlut *lut)
> +{
> +	if (!drm_lut3d_size) {
> +		lut->state.bits.initialized = 0;
> +		return;
> +	}
> +
> +	/* Only supports 17x17x17 3D LUT (12-bit) now */
> +	lut->lut_3d.use_12bits = true;
> +	lut->lut_3d.use_tetrahedral_9 = false;
> +
> +	lut->state.bits.initialized = 1;
> +	__drm_3dlut_to_dc_3dlut(drm_lut3d, drm_lut3d_size, &lut->lut_3d,
> +				lut->lut_3d.use_tetrahedral_9, 12);
> +
> +}
> +
> +static int
> +__set_dm_plane_colorop_3dlut(struct drm_plane_state *plane_state,
> +			     struct dc_plane_state *dc_plane_state,
> +			     struct drm_colorop *colorop)
> +{
> +	struct drm_colorop *old_colorop;
> +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
> +	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
> +	struct drm_atomic_state *state = plane_state->state;
> +	const struct amdgpu_device *adev = drm_to_adev(colorop->dev);
> +	const struct drm_device *dev = colorop->dev;
> +	const struct drm_color_lut *lut3d;
> +	uint32_t lut3d_size;
> +	int i = 0, ret = 0;
> +
> +	/* 3D LUT */
> +	old_colorop = colorop;
> +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
> +		if (new_colorop_state->colorop == old_colorop &&
> +		    new_colorop_state->colorop->type == DRM_COLOROP_3D_LUT) {
> +			colorop_state = new_colorop_state;
> +			break;
> +		}
> +	}
> +
> +	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_3D_LUT) {
> +		if (!adev->dm.dc->caps.color.dpp.hw_3d_lut) {

I wonder if this check is no longer accurate in DCN versions with MCM
(MPC only) 3D LUT caps, such as DCN 3.2 and DCN 4.01.

Also, looking back those patches that introduced shaper and blnd tf and
luts, I don't see similar validation, but IIRC shaper caps directly
depends on 3d lut, for example. IIRC something around blnd func caps
also changed in the above-mentioned DCN versions.

Melissa

> +			drm_dbg(dev, "3D LUT is not supported by hardware\n");
> +			return -EINVAL;
> +		}
> +
> +		drm_dbg(dev, "3D LUT colorop with ID: %d\n", colorop->base.id);
> +		lut3d = __extract_blob_lut(colorop_state->data, &lut3d_size);
> +		lut3d_size = lut3d != NULL ? lut3d_size : 0;
> +		__set_colorop_3dlut(lut3d, lut3d_size, &dc_plane_state->lut3d_func);
> +
> +		/* 3D LUT requires shaper. If shaper colorop is bypassed, enable shaper curve
> +		 * with TRANSFER_FUNCTION_LINEAR
> +		 */
> +		if (tf->type == TF_TYPE_BYPASS) {
> +			tf->type = TF_TYPE_DISTRIBUTED_POINTS;
> +			tf->tf = TRANSFER_FUNCTION_LINEAR;
> +			tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
> +			ret = __set_output_tf(tf, NULL, 0, false);
> +		}
> +	}
> +
> +	return ret;
> +}
> +
>  static int
>  __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
>  			     struct dc_plane_state *dc_plane_state,
> @@ -1522,6 +1605,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
>  	if (!colorop)
>  		return -EINVAL;
>  
> +	/* 3D LUT */
> +	colorop = colorop->next;
> +	if (!colorop) {
> +		drm_dbg(dev, "no 3D LUT colorop found\n");
> +		return -EINVAL;
> +	}
> +
> +	ret = __set_dm_plane_colorop_3dlut(plane_state, dc_plane_state, colorop);
> +	if (ret)
> +		return ret;
> +
>  	/* 1D Curve & LUT - BLND TF & LUT */
>  	colorop = colorop->next;
>  	if (!colorop) {
> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> index 10b3e3906461..e90774294971 100644
> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> @@ -49,6 +49,8 @@ const u64 amdgpu_dm_supported_blnd_tfs =
>  
>  #define MAX_COLOR_PIPELINE_OPS 10
>  
> +#define LUT3D_SIZE		17
> +
>  int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
>  {
>  	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
> @@ -145,6 +147,24 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
>  
>  	i++;
>  
> +	/* 3D LUT */
> +	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> +	if (!ops[i]) {
> +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> +		ret = -ENOMEM;
> +		goto cleanup;
> +	}
> +
> +	ret = drm_plane_colorop_3dlut_init(dev, ops[i], plane, LUT3D_SIZE,
> +				     DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
> +				     DRM_COLOROP_FLAG_ALLOW_BYPASS);
> +	if (ret)
> +		goto cleanup;
> +
> +	drm_colorop_set_next_property(ops[i-1], ops[i]);
> +
> +	i++;
> +
>  	/* 1D curve - BLND TF */
>  	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
>  	if (!ops[i]) {
> -- 
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop
  2025-05-13  0:52   ` Melissa Wen
@ 2025-05-13  3:29     ` Alex Hung
  2025-05-13 15:52       ` Melissa Wen
  0 siblings, 1 reply; 82+ messages in thread
From: Alex Hung @ 2025-05-13  3:29 UTC (permalink / raw)
  To: Melissa Wen
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone



On 5/12/25 18:52, Melissa Wen wrote:
> On 04/29, Alex Hung wrote:
>> This adds support for a 3D LUT.
>>
>> The color pipeline now consists of the following colorops:
>> 1. 1D curve colorop
>> 2. Multiplier
>> 3. 3x4 CTM
>> 4. 1D curve colorop
>> 5. 1D LUT
>> 6. 3D LUT
>> 7. 1D curve colorop
>> 8. 1D LUT
>>
>> Signed-off-by: Alex Hung <alex.hung@amd.com>
>> Reviewed-by: Daniel Stone <daniels@collabora.com>
>> ---
>> V9:
>>   - Return a value in __set_dm_plane_colorop_3dlut
>>
>> v8:
>>   - Set initialized to 0 and return when drm_lut3d_size is 0 (Harry Wentland)
>>   - Rework tf->type = TF_TYPE_BYPASS for shaper (Harry Wentland & Leo Li)
>>
>> v7:
>>   - Simplify 3D LUT according to drm_colorop changes (Simon Ser)
>>
>>   .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 94 +++++++++++++++++++
>>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 20 ++++
>>   2 files changed, 114 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
>> index 313716f2003f..dfdd3f557570 100644
>> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
>> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
>> @@ -1293,6 +1293,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
>>   	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
>>   	const struct drm_color_lut *shaper_lut;
>>   	struct drm_device *dev = colorop->dev;
>> +	bool enabled = false;
>>   	uint32_t shaper_size;
>>   	int i = 0, ret = 0;
>>   
>> @@ -1314,6 +1315,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
>>   		ret = __set_output_tf(tf, 0, 0, false);
>>   		if (ret)
>>   			return ret;
>> +		enabled = true;
>>   	}
>>   
>>   	/* 1D LUT - SHAPER LUT */
>> @@ -1345,12 +1347,93 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
>>   			ret = __set_output_tf(tf, shaper_lut, shaper_size, false);
>>   			if (ret)
>>   				return ret;
>> +			enabled = true;
>>   		}
>>   	}
>>   
>> +	if (!enabled)
>> +		tf->type = TF_TYPE_BYPASS;
>> +
>>   	return 0;
>>   }
>>   
>> +/* __set_colorop_3dlut - set DRM 3D LUT to DC stream
>> + * @drm_lut3d: user 3D LUT
>> + * @drm_lut3d_size: size of 3D LUT
>> + * @lut3d: DC 3D LUT
>> + *
>> + * Map user 3D LUT data to DC 3D LUT and all necessary bits to program it
>> + * on DCN accordingly.
>> + */
>> +static void __set_colorop_3dlut(const struct drm_color_lut *drm_lut3d,
>> +				uint32_t drm_lut3d_size,
>> +				struct dc_3dlut *lut)
>> +{
>> +	if (!drm_lut3d_size) {
>> +		lut->state.bits.initialized = 0;
>> +		return;
>> +	}
>> +
>> +	/* Only supports 17x17x17 3D LUT (12-bit) now */
>> +	lut->lut_3d.use_12bits = true;
>> +	lut->lut_3d.use_tetrahedral_9 = false;
>> +
>> +	lut->state.bits.initialized = 1;
>> +	__drm_3dlut_to_dc_3dlut(drm_lut3d, drm_lut3d_size, &lut->lut_3d,
>> +				lut->lut_3d.use_tetrahedral_9, 12);
>> +
>> +}
>> +
>> +static int
>> +__set_dm_plane_colorop_3dlut(struct drm_plane_state *plane_state,
>> +			     struct dc_plane_state *dc_plane_state,
>> +			     struct drm_colorop *colorop)
>> +{
>> +	struct drm_colorop *old_colorop;
>> +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
>> +	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
>> +	struct drm_atomic_state *state = plane_state->state;
>> +	const struct amdgpu_device *adev = drm_to_adev(colorop->dev);
>> +	const struct drm_device *dev = colorop->dev;
>> +	const struct drm_color_lut *lut3d;
>> +	uint32_t lut3d_size;
>> +	int i = 0, ret = 0;
>> +
>> +	/* 3D LUT */
>> +	old_colorop = colorop;
>> +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
>> +		if (new_colorop_state->colorop == old_colorop &&
>> +		    new_colorop_state->colorop->type == DRM_COLOROP_3D_LUT) {
>> +			colorop_state = new_colorop_state;
>> +			break;
>> +		}
>> +	}
>> +
>> +	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_3D_LUT) {
>> +		if (!adev->dm.dc->caps.color.dpp.hw_3d_lut) {
> 
> I wonder if this check is no longer accurate in DCN versions with MCM
> (MPC only) 3D LUT caps, such as DCN 3.2 and DCN 4.01.

The current goal is to validate on specific fixed platforms. We will add 
more platform-specific implemenation when we enable this on new DCN 
hardware.

In some case, new hardware may have a different color pipeline. For 
example, there can be no DRM_COLOROP_3D_LUT colorop in a color pipeline, 
and it is not necessary to check DCN versions. In other cases, we will 
need to check different DCN versions or different flags for sure.

> 
> Also, looking back those patches that introduced shaper and blnd tf and
> luts, I don't see similar validation, but IIRC shaper caps directly
> depends on 3d lut, for example. IIRC something around blnd func caps
> also changed in the above-mentioned DCN versions.
> 
> Melissa
 > >> +			drm_dbg(dev, "3D LUT is not supported by hardware\n");
>> +			return -EINVAL;
>> +		}
>> +
>> +		drm_dbg(dev, "3D LUT colorop with ID: %d\n", colorop->base.id);
>> +		lut3d = __extract_blob_lut(colorop_state->data, &lut3d_size);
>> +		lut3d_size = lut3d != NULL ? lut3d_size : 0;
>> +		__set_colorop_3dlut(lut3d, lut3d_size, &dc_plane_state->lut3d_func);
>> +
>> +		/* 3D LUT requires shaper. If shaper colorop is bypassed, enable shaper curve
>> +		 * with TRANSFER_FUNCTION_LINEAR
>> +		 */
>> +		if (tf->type == TF_TYPE_BYPASS) {
>> +			tf->type = TF_TYPE_DISTRIBUTED_POINTS;
>> +			tf->tf = TRANSFER_FUNCTION_LINEAR;
>> +			tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
>> +			ret = __set_output_tf(tf, NULL, 0, false);
>> +		}

shaper / 3dlut dependency is checked here.

>> +	}
>> +
>> +	return ret;
>> +}
>> +
>>   static int
>>   __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
>>   			     struct dc_plane_state *dc_plane_state,
>> @@ -1522,6 +1605,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
>>   	if (!colorop)
>>   		return -EINVAL;
>>   
>> +	/* 3D LUT */
>> +	colorop = colorop->next;
>> +	if (!colorop) {
>> +		drm_dbg(dev, "no 3D LUT colorop found\n");
>> +		return -EINVAL;
>> +	}
>> +
>> +	ret = __set_dm_plane_colorop_3dlut(plane_state, dc_plane_state, colorop);
>> +	if (ret)
>> +		return ret;
>> +
>>   	/* 1D Curve & LUT - BLND TF & LUT */
>>   	colorop = colorop->next;
>>   	if (!colorop) {
>> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>> index 10b3e3906461..e90774294971 100644
>> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>> @@ -49,6 +49,8 @@ const u64 amdgpu_dm_supported_blnd_tfs =
>>   
>>   #define MAX_COLOR_PIPELINE_OPS 10
>>   
>> +#define LUT3D_SIZE		17
>> +
>>   int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
>>   {
>>   	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
>> @@ -145,6 +147,24 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
>>   
>>   	i++;
>>   
>> +	/* 3D LUT */
>> +	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
>> +	if (!ops[i]) {
>> +		DRM_ERROR("KMS: Failed to allocate colorop\n");
>> +		ret = -ENOMEM;
>> +		goto cleanup;
>> +	}
>> +
>> +	ret = drm_plane_colorop_3dlut_init(dev, ops[i], plane, LUT3D_SIZE,
>> +				     DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
>> +				     DRM_COLOROP_FLAG_ALLOW_BYPASS);
>> +	if (ret)
>> +		goto cleanup;
>> +
>> +	drm_colorop_set_next_property(ops[i-1], ops[i]);
>> +
>> +	i++;
>> +
>>   	/* 1D curve - BLND TF */
>>   	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
>>   	if (!ops[i]) {
>> -- 
>> 2.43.0
>>


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2025-05-12 18:50   ` Melissa Wen
@ 2025-05-13  7:42     ` Pekka Paalanen
  2025-05-13 15:36       ` Melissa Wen
  0 siblings, 1 reply; 82+ messages in thread
From: Pekka Paalanen @ 2025-05-13  7:42 UTC (permalink / raw)
  To: Melissa Wen
  Cc: Alex Hung, dri-devel, amd-gfx, wayland-devel, harry.wentland,
	leo.liu, ville.syrjala, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

[-- Attachment #1: Type: text/plain, Size: 6397 bytes --]

On Mon, 12 May 2025 15:50:17 -0300
Melissa Wen <mwen@igalia.com> wrote:

> On 04/29, Alex Hung wrote:
> > Expose one 1D curve colorop with support for
> > DRM_COLOROP_1D_CURVE_SRGB_EOTF and program HW to perform
> > the sRGB transform when the colorop is not in bypass.
> > 
> > With this change the following IGT test passes:
> > kms_colorop --run plane-XR30-XR30-srgb_eotf
> > 
> > The color pipeline now consists of a single colorop:
> > 1. 1D curve colorop w/ sRGB EOTF
> > 
> > Signed-off-by: Alex Hung <alex.hung@amd.com>
> > Co-developed-by: Harry Wentland <harry.wentland@amd.com>
> > Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> > Reviewed-by: Daniel Stone <daniels@collabora.com>
> > ---
> > V9:
> >  - Update function names by _plane_ (Chaitanya Kumar Borah)
> >  - Update replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)
> > 
> > v8:
> >  - Fix incorrect && by || in __set_colorop_in_tf_1d_curve (Leo Li)
> > 
> > v7:
> >  - Fix checkpatch warnings
> >   - Change switch "{ }" position
> >   - Delete double ";"
> >   - Delete "{ }" for single-line if-statement
> >   - Add a new line at EOF
> >   - Change SPDX-License-Identifier: GPL-2.0+ from // to /* */
> > 
> > v6:
> >  - cleanup if colorop alloc or init fails
> > 
> >  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
> >  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 86 +++++++++++++++++++
> >  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 69 +++++++++++++++
> >  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 ++++++++
> >  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
> >  5 files changed, 201 insertions(+), 1 deletion(-)
> >  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> >  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> > 
> > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> > index ab2a97e354da..46158d67ab12 100644
> > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> > @@ -38,7 +38,8 @@ AMDGPUDM = \
> >  	amdgpu_dm_pp_smu.o \
> >  	amdgpu_dm_psr.o \
> >  	amdgpu_dm_replay.o \
> > -	amdgpu_dm_wb.o
> > +	amdgpu_dm_wb.o \
> > +	amdgpu_dm_colorop.o
> >  
> >  ifdef CONFIG_DRM_AMD_DC_FP
> >  AMDGPUDM += dc_fpu.o
> > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > index ebabfe3a512f..0b513ab5050f 100644
> > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > @@ -668,6 +668,18 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf)
> >  	}
> >  }
> >  
> > +static enum dc_transfer_func_predefined
> > +amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
> > +{
> > +	switch (tf) {
> > +	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
> > +	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
> > +		return TRANSFER_FUNCTION_SRGB;
> > +	default:
> > +		return TRANSFER_FUNCTION_LINEAR;
> > +	}
> > +}
> > +
> >  static void __to_dc_lut3d_color(struct dc_rgb *rgb,
> >  				const struct drm_color_lut lut,
> >  				int bit_precision)
> > @@ -1137,6 +1149,59 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state,
> >  	return 0;
> >  }
> >  
> > +static int
> > +__set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
> > +		       struct drm_colorop_state *colorop_state)
> > +{
> > +	struct dc_transfer_func *tf = &dc_plane_state->in_transfer_func;
> > +	struct drm_colorop *colorop = colorop_state->colorop;
> > +	struct drm_device *drm = colorop->dev;
> > +
> > +	if (colorop->type != DRM_COLOROP_1D_CURVE ||
> > +	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
> > +		return -EINVAL;
> > +
> > +	if (colorop_state->bypass) {
> > +		tf->type = TF_TYPE_BYPASS;
> > +		tf->tf = TRANSFER_FUNCTION_LINEAR;
> > +		return 0;
> > +	}
> > +
> > +	drm_dbg(drm, "Degamma colorop with ID: %d\n", colorop->base.id);
> > +
> > +	tf->type = TF_TYPE_PREDEFINED;
> > +	tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
> > +
> > +	return 0;
> > +}
> > +
> > +static int
> > +__set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
> > +			       struct dc_plane_state *dc_plane_state,
> > +			       struct drm_colorop *colorop)
> > +{
> > +	struct drm_colorop *old_colorop;
> > +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
> > +	struct drm_atomic_state *state = plane_state->state;
> > +	int i = 0;
> > +
> > +	old_colorop = colorop;
> > +
> > +	/* 1st op: 1d curve - degamma */
> > +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
> > +		if (new_colorop_state->colorop == old_colorop &&
> > +		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
> > +			colorop_state = new_colorop_state;
> > +			break;
> > +		}
> > +	}
> > +
> > +	if (!colorop_state)
> > +		return -EINVAL;
> > +
> > +	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);  
> 
> I wonder what will happen if plane degamma isn't set, but CRTC degamma
> LUT or legacy CRTC regamma LUT (with its implicity sRGB degamma) is used
> together with other plane color ops.
> 
> I can imagine the mess, so I think CRTC degamma LUT and legacy CRTC
> regamma LUT should be somehow entirely disabled (or rejected) if plane
> color pipeline is in use.

Hi Melissa,

if using a plane color pipeline means that a CRTC LUT cannot be used, it
will severely limit the usefulness of the whole KMS color processing. In
Weston's case it would prohibit *all* KMS off-loading when color
management is in use.

Weston chooses to do composition and blending in an optical space. This
means that plane color pipelines are required to convert incoming
pixels into the optical space, and a CRTC LUT (a CRTC color pipeline in
the future) is required to convert from the optical space to the
monitor signalling (electrical space).

I don't know what "with its implicity sRGB degamma" means, but there
cannot be any implicit curves at all. The driver has no knowledge of
how the framebuffer pixels are encoded, nor about what the blending
space should be.


Thanks,
pq

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2025-05-13  7:42     ` Pekka Paalanen
@ 2025-05-13 15:36       ` Melissa Wen
  2025-05-13 20:39         ` Harry Wentland
  0 siblings, 1 reply; 82+ messages in thread
From: Melissa Wen @ 2025-05-13 15:36 UTC (permalink / raw)
  To: Pekka Paalanen
  Cc: Alex Hung, dri-devel, amd-gfx, wayland-devel, harry.wentland,
	leo.liu, ville.syrjala, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 05/13, Pekka Paalanen wrote:
> On Mon, 12 May 2025 15:50:17 -0300
> Melissa Wen <mwen@igalia.com> wrote:
> 
> > On 04/29, Alex Hung wrote:
> > > Expose one 1D curve colorop with support for
> > > DRM_COLOROP_1D_CURVE_SRGB_EOTF and program HW to perform
> > > the sRGB transform when the colorop is not in bypass.
> > > 
> > > With this change the following IGT test passes:
> > > kms_colorop --run plane-XR30-XR30-srgb_eotf
> > > 
> > > The color pipeline now consists of a single colorop:
> > > 1. 1D curve colorop w/ sRGB EOTF
> > > 
> > > Signed-off-by: Alex Hung <alex.hung@amd.com>
> > > Co-developed-by: Harry Wentland <harry.wentland@amd.com>
> > > Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> > > Reviewed-by: Daniel Stone <daniels@collabora.com>
> > > ---
> > > V9:
> > >  - Update function names by _plane_ (Chaitanya Kumar Borah)
> > >  - Update replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)
> > > 
> > > v8:
> > >  - Fix incorrect && by || in __set_colorop_in_tf_1d_curve (Leo Li)
> > > 
> > > v7:
> > >  - Fix checkpatch warnings
> > >   - Change switch "{ }" position
> > >   - Delete double ";"
> > >   - Delete "{ }" for single-line if-statement
> > >   - Add a new line at EOF
> > >   - Change SPDX-License-Identifier: GPL-2.0+ from // to /* */
> > > 
> > > v6:
> > >  - cleanup if colorop alloc or init fails
> > > 
> > >  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
> > >  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 86 +++++++++++++++++++
> > >  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 69 +++++++++++++++
> > >  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 ++++++++
> > >  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
> > >  5 files changed, 201 insertions(+), 1 deletion(-)
> > >  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> > >  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> > > 
> > > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> > > index ab2a97e354da..46158d67ab12 100644
> > > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> > > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> > > @@ -38,7 +38,8 @@ AMDGPUDM = \
> > >  	amdgpu_dm_pp_smu.o \
> > >  	amdgpu_dm_psr.o \
> > >  	amdgpu_dm_replay.o \
> > > -	amdgpu_dm_wb.o
> > > +	amdgpu_dm_wb.o \
> > > +	amdgpu_dm_colorop.o
> > >  
> > >  ifdef CONFIG_DRM_AMD_DC_FP
> > >  AMDGPUDM += dc_fpu.o
> > > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > > index ebabfe3a512f..0b513ab5050f 100644
> > > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > > @@ -668,6 +668,18 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf)
> > >  	}
> > >  }
> > >  
> > > +static enum dc_transfer_func_predefined
> > > +amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
> > > +{
> > > +	switch (tf) {
> > > +	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
> > > +	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
> > > +		return TRANSFER_FUNCTION_SRGB;
> > > +	default:
> > > +		return TRANSFER_FUNCTION_LINEAR;
> > > +	}
> > > +}
> > > +
> > >  static void __to_dc_lut3d_color(struct dc_rgb *rgb,
> > >  				const struct drm_color_lut lut,
> > >  				int bit_precision)
> > > @@ -1137,6 +1149,59 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state,
> > >  	return 0;
> > >  }
> > >  
> > > +static int
> > > +__set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
> > > +		       struct drm_colorop_state *colorop_state)
> > > +{
> > > +	struct dc_transfer_func *tf = &dc_plane_state->in_transfer_func;
> > > +	struct drm_colorop *colorop = colorop_state->colorop;
> > > +	struct drm_device *drm = colorop->dev;
> > > +
> > > +	if (colorop->type != DRM_COLOROP_1D_CURVE ||
> > > +	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
> > > +		return -EINVAL;
> > > +
> > > +	if (colorop_state->bypass) {
> > > +		tf->type = TF_TYPE_BYPASS;
> > > +		tf->tf = TRANSFER_FUNCTION_LINEAR;
> > > +		return 0;
> > > +	}
> > > +
> > > +	drm_dbg(drm, "Degamma colorop with ID: %d\n", colorop->base.id);
> > > +
> > > +	tf->type = TF_TYPE_PREDEFINED;
> > > +	tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static int
> > > +__set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
> > > +			       struct dc_plane_state *dc_plane_state,
> > > +			       struct drm_colorop *colorop)
> > > +{
> > > +	struct drm_colorop *old_colorop;
> > > +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
> > > +	struct drm_atomic_state *state = plane_state->state;
> > > +	int i = 0;
> > > +
> > > +	old_colorop = colorop;
> > > +
> > > +	/* 1st op: 1d curve - degamma */
> > > +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
> > > +		if (new_colorop_state->colorop == old_colorop &&
> > > +		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
> > > +			colorop_state = new_colorop_state;
> > > +			break;
> > > +		}
> > > +	}
> > > +
> > > +	if (!colorop_state)
> > > +		return -EINVAL;
> > > +
> > > +	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);  
> > 
> > I wonder what will happen if plane degamma isn't set, but CRTC degamma
> > LUT or legacy CRTC regamma LUT (with its implicity sRGB degamma) is used
> > together with other plane color ops.
> > 
> > I can imagine the mess, so I think CRTC degamma LUT and legacy CRTC
> > regamma LUT should be somehow entirely disabled (or rejected) if plane
> > color pipeline is in use.
> 
> Hi Melissa,
> 
> if using a plane color pipeline means that a CRTC LUT cannot be used, it
> will severely limit the usefulness of the whole KMS color processing. In
> Weston's case it would prohibit *all* KMS off-loading when color
> management is in use.
> 
> Weston chooses to do composition and blending in an optical space. This
> means that plane color pipelines are required to convert incoming
> pixels into the optical space, and a CRTC LUT (a CRTC color pipeline in
> the future) is required to convert from the optical space to the
> monitor signalling (electrical space).

Hi Pekka,

IIRC, Weston needs one post-blending 1D LUT and with my suggestion the
atomic CRTC regamma LUT works fine and can be this 1D LUT.
So, instead of an atomic post-blending/CRTC color pipeline with:

[blending] -> CRTC 1D LUT -> CRTC CTM -> CRTC 1D LUT

when plane color pipeline is in use, the driver accepts only:

[blending] -> CRTC CTM -> CRTC 1D LUT

If AMD driver continues accepting/exposing CRTC degamma LUT plus plane
color pipeline, and userpace wants something like:

Plane shaper LUT -> Plane 3D LUT -> Plane Blnd LUT -> [blending] -> **CRTC** degamma LUT

I understand that this weird sequence is what will actually happen:

**CRTC** degamma LUT -> Plane shaper LUT -> Plane 3D LUT -> Plane Blnd LUT -> [blending]

Because userspace doesn't care if this is a "degamma" or "regamma" LUT
and they will probably pick the first 1D LUT in the post-blending color
pipeline, which currently means CRTC degamma LUT. So, better if it takes
the CRTC regamma LUT that is actually a post-blending 1D LUT. Where I
expected this result:

Plane shaper LUT -> PLane 3D LUT -> PLane Blnd LUT -> [blending] -> CRTC regamma LUT

You can vizualize better this degamma issue in this diagram:
https://raw.githubusercontent.com/melissawen/melissawen.github.io/master/img/xdc-2023-colors-talk/rainbow-treasure-xdc-2023-17.png

In the current driver-specific implementation, driver rejects atomic commits
in case of collision, i.e., Plane degamma LUT + CRTC degamma LUT, but
IIRC, the driver doesn't handle it if it doesn't clash.

> 
> I don't know what "with its implicity sRGB degamma" means, but there
> cannot be any implicit curves at all. The driver has no knowledge of
> how the framebuffer pixels are encoded, nor about what the blending
> space should be.

It targets the legacy CRTC regamma LUT (256 size), not the atomic CRTC
color pipeline.

Melissa

> 
> 
> Thanks,
> pq



^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop
  2025-05-13  3:29     ` Alex Hung
@ 2025-05-13 15:52       ` Melissa Wen
  0 siblings, 0 replies; 82+ messages in thread
From: Melissa Wen @ 2025-05-13 15:52 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

On 05/12, Alex Hung wrote:
> 
> 
> On 5/12/25 18:52, Melissa Wen wrote:
> > On 04/29, Alex Hung wrote:
> > > This adds support for a 3D LUT.
> > > 
> > > The color pipeline now consists of the following colorops:
> > > 1. 1D curve colorop
> > > 2. Multiplier
> > > 3. 3x4 CTM
> > > 4. 1D curve colorop
> > > 5. 1D LUT
> > > 6. 3D LUT
> > > 7. 1D curve colorop
> > > 8. 1D LUT
> > > 
> > > Signed-off-by: Alex Hung <alex.hung@amd.com>
> > > Reviewed-by: Daniel Stone <daniels@collabora.com>
> > > ---
> > > V9:
> > >   - Return a value in __set_dm_plane_colorop_3dlut
> > > 
> > > v8:
> > >   - Set initialized to 0 and return when drm_lut3d_size is 0 (Harry Wentland)
> > >   - Rework tf->type = TF_TYPE_BYPASS for shaper (Harry Wentland & Leo Li)
> > > 
> > > v7:
> > >   - Simplify 3D LUT according to drm_colorop changes (Simon Ser)
> > > 
> > >   .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 94 +++++++++++++++++++
> > >   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 20 ++++
> > >   2 files changed, 114 insertions(+)
> > > 
> > > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > > index 313716f2003f..dfdd3f557570 100644
> > > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> > > @@ -1293,6 +1293,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
> > >   	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
> > >   	const struct drm_color_lut *shaper_lut;
> > >   	struct drm_device *dev = colorop->dev;
> > > +	bool enabled = false;
> > >   	uint32_t shaper_size;
> > >   	int i = 0, ret = 0;
> > > @@ -1314,6 +1315,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
> > >   		ret = __set_output_tf(tf, 0, 0, false);
> > >   		if (ret)
> > >   			return ret;
> > > +		enabled = true;
> > >   	}
> > >   	/* 1D LUT - SHAPER LUT */
> > > @@ -1345,12 +1347,93 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
> > >   			ret = __set_output_tf(tf, shaper_lut, shaper_size, false);
> > >   			if (ret)
> > >   				return ret;
> > > +			enabled = true;
> > >   		}
> > >   	}
> > > +	if (!enabled)
> > > +		tf->type = TF_TYPE_BYPASS;
> > > +
> > >   	return 0;
> > >   }
> > > +/* __set_colorop_3dlut - set DRM 3D LUT to DC stream
> > > + * @drm_lut3d: user 3D LUT
> > > + * @drm_lut3d_size: size of 3D LUT
> > > + * @lut3d: DC 3D LUT
> > > + *
> > > + * Map user 3D LUT data to DC 3D LUT and all necessary bits to program it
> > > + * on DCN accordingly.
> > > + */
> > > +static void __set_colorop_3dlut(const struct drm_color_lut *drm_lut3d,
> > > +				uint32_t drm_lut3d_size,
> > > +				struct dc_3dlut *lut)
> > > +{
> > > +	if (!drm_lut3d_size) {
> > > +		lut->state.bits.initialized = 0;
> > > +		return;
> > > +	}
> > > +
> > > +	/* Only supports 17x17x17 3D LUT (12-bit) now */
> > > +	lut->lut_3d.use_12bits = true;
> > > +	lut->lut_3d.use_tetrahedral_9 = false;
> > > +
> > > +	lut->state.bits.initialized = 1;
> > > +	__drm_3dlut_to_dc_3dlut(drm_lut3d, drm_lut3d_size, &lut->lut_3d,
> > > +				lut->lut_3d.use_tetrahedral_9, 12);
> > > +
> > > +}
> > > +
> > > +static int
> > > +__set_dm_plane_colorop_3dlut(struct drm_plane_state *plane_state,
> > > +			     struct dc_plane_state *dc_plane_state,
> > > +			     struct drm_colorop *colorop)
> > > +{
> > > +	struct drm_colorop *old_colorop;
> > > +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
> > > +	struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
> > > +	struct drm_atomic_state *state = plane_state->state;
> > > +	const struct amdgpu_device *adev = drm_to_adev(colorop->dev);
> > > +	const struct drm_device *dev = colorop->dev;
> > > +	const struct drm_color_lut *lut3d;
> > > +	uint32_t lut3d_size;
> > > +	int i = 0, ret = 0;
> > > +
> > > +	/* 3D LUT */
> > > +	old_colorop = colorop;
> > > +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
> > > +		if (new_colorop_state->colorop == old_colorop &&
> > > +		    new_colorop_state->colorop->type == DRM_COLOROP_3D_LUT) {
> > > +			colorop_state = new_colorop_state;
> > > +			break;
> > > +		}
> > > +	}
> > > +
> > > +	if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_3D_LUT) {
> > > +		if (!adev->dm.dc->caps.color.dpp.hw_3d_lut) {
> > 
> > I wonder if this check is no longer accurate in DCN versions with MCM
> > (MPC only) 3D LUT caps, such as DCN 3.2 and DCN 4.01.
> 
> The current goal is to validate on specific fixed platforms. We will add
> more platform-specific implemenation when we enable this on new DCN
> hardware.
> 
> In some case, new hardware may have a different color pipeline. For example,
> there can be no DRM_COLOROP_3D_LUT colorop in a color pipeline, and it is
> not necessary to check DCN versions. In other cases, we will need to check
> different DCN versions or different flags for sure.

Hi Alex,

So, if this implementation doesn't handle these differences, which
specific DCN version this plane color pipeline targets?

Cover letter says DCN 3 or newer, but once this patch lands, it will
expose color the same plane color caps on any DCN versions if the
management of these differences are not managed.

> 
> > 
> > Also, looking back those patches that introduced shaper and blnd tf and
> > luts, I don't see similar validation, but IIRC shaper caps directly
> > depends on 3d lut, for example. IIRC something around blnd func caps
> > also changed in the above-mentioned DCN versions.
> > 
> > Melissa
> > >> +			drm_dbg(dev, "3D LUT is not supported by hardware\n");
> > > +			return -EINVAL;
> > > +		}
> > > +
> > > +		drm_dbg(dev, "3D LUT colorop with ID: %d\n", colorop->base.id);
> > > +		lut3d = __extract_blob_lut(colorop_state->data, &lut3d_size);
> > > +		lut3d_size = lut3d != NULL ? lut3d_size : 0;
> > > +		__set_colorop_3dlut(lut3d, lut3d_size, &dc_plane_state->lut3d_func);
> > > +
> > > +		/* 3D LUT requires shaper. If shaper colorop is bypassed, enable shaper curve
> > > +		 * with TRANSFER_FUNCTION_LINEAR
> > > +		 */
> > > +		if (tf->type == TF_TYPE_BYPASS) {
> > > +			tf->type = TF_TYPE_DISTRIBUTED_POINTS;
> > > +			tf->tf = TRANSFER_FUNCTION_LINEAR;
> > > +			tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
> > > +			ret = __set_output_tf(tf, NULL, 0, false);
> > > +		}
> 
> shaper / 3dlut dependency is checked here.

Right. It handles the dependency when 3D LUT is available, but what I
expected was that, if `!adev->dm.dc->caps.color.dpp.hw_3dlut`:
- driver should not expose shaper curve/LUT too; and
- driver rejects any attempt to program shaper already in that patch
  that introduced the shaper caps.

Melissa

> 
> > > +	}
> > > +
> > > +	return ret;
> > > +}
> > > +
> > >   static int
> > >   __set_dm_plane_colorop_blend(struct drm_plane_state *plane_state,
> > >   			     struct dc_plane_state *dc_plane_state,
> > > @@ -1522,6 +1605,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
> > >   	if (!colorop)
> > >   		return -EINVAL;
> > > +	/* 3D LUT */
> > > +	colorop = colorop->next;
> > > +	if (!colorop) {
> > > +		drm_dbg(dev, "no 3D LUT colorop found\n");
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	ret = __set_dm_plane_colorop_3dlut(plane_state, dc_plane_state, colorop);
> > > +	if (ret)
> > > +		return ret;
> > > +
> > >   	/* 1D Curve & LUT - BLND TF & LUT */
> > >   	colorop = colorop->next;
> > >   	if (!colorop) {
> > > diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> > > index 10b3e3906461..e90774294971 100644
> > > --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> > > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> > > @@ -49,6 +49,8 @@ const u64 amdgpu_dm_supported_blnd_tfs =
> > >   #define MAX_COLOR_PIPELINE_OPS 10
> > > +#define LUT3D_SIZE		17
> > > +
> > >   int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
> > >   {
> > >   	struct drm_colorop *ops[MAX_COLOR_PIPELINE_OPS];
> > > @@ -145,6 +147,24 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
> > >   	i++;
> > > +	/* 3D LUT */
> > > +	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> > > +	if (!ops[i]) {
> > > +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> > > +		ret = -ENOMEM;
> > > +		goto cleanup;
> > > +	}
> > > +
> > > +	ret = drm_plane_colorop_3dlut_init(dev, ops[i], plane, LUT3D_SIZE,
> > > +				     DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
> > > +				     DRM_COLOROP_FLAG_ALLOW_BYPASS);
> > > +	if (ret)
> > > +		goto cleanup;
> > > +
> > > +	drm_colorop_set_next_property(ops[i-1], ops[i]);
> > > +
> > > +	i++;
> > > +
> > >   	/* 1D curve - BLND TF */
> > >   	ops[i] = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> > >   	if (!ops[i]) {
> > > -- 
> > > 2.43.0
> > > 
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (42 preceding siblings ...)
  2025-04-30  1:11 ` [PATCH V9 43/43] drm/amd/display: Add AMD color pipeline doc Alex Hung
@ 2025-05-13 15:59 ` Melissa Wen
  2025-05-15  9:45 ` Simon Ser
  44 siblings, 0 replies; 82+ messages in thread
From: Melissa Wen @ 2025-05-13 15:59 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo

On 04/29, Alex Hung wrote:
> This is an RFC set for a color pipeline API, along with implementations
> in VKMS and amdgpu. It is tested with a set of IGT tests that can be
> found at [1]. The IGT tests run a pixel-by-pixel comparison with an
> allowable delta variation as the goal for these transformations is
> perceptual correctness, not complete pixel accuracy.

Hi,

Thanks for this work. I found it very solid and well documented.
The drm colorop implementation looks good to me, but I still have some
questions about AMD specific bits.

Patches 1, 3-13, 16, 21-23, 25, 35-41 and 43 are:
Reviewed-by: Melissa Wen <mwen@igalia.com>

And I dropped some comments and questions to some AMD-specific patches.

> 
> v5 of this patchset fleshed out documentation for colorops and the
> various defines that are being introduced.
> 
> v6 addresses a few comments from various reviewers.
> 
> v7 simplifies 3D LUT and addresses more comments from various reviewers.
> 
> v8 fixes typo and errors and address comments from reviewers.
> 
> v9 refactors cleanup functions, fixes typo and errors, and addresses
>    comments from reviewers.
> 
> VKMS supports two named transfer function colorops and two matrix
> colorops.
> 
> Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D Curve EOTF
> 8. 1D LUT
> 
> The supported curves for the 1D Curve type are:
> - sRGB EOTF and its inverse
> - PQ EOTF, scaled to [0.0, 125.0] and its inverse
> - BT.2020/BT.709 OETF and its inverse
> 
> Note that the 1st and 5th colorops take the EOTF or Inverse
> OETF while the 3rd colorop takes the Inverse EOTF or OETF.
> 
> The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
> exists for other drivers to describe their own 3D LUT capability.
> 
> This mirrors the color pipeline used by gamescope and presented by
> Melissa Wen, with the exception of the DEGAM LUT, which is not currently
> used. See [1]
> https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
> 
> At this point we're hoping to see gamescope and kwin implementations
> take shape. The existing pipeline should be enough to satisfy the
> gamescope use-cases on the drm_plane.
> 
> In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
> support to the color pipeline. I have sketched these out already but
> don't have it all hooked up yet. This should not hinder adoption of this
> API for gaming use-cases.
> 
> We'll also want to advertise IN_FORMATS on a color pipeline as some
> color pipelines won't be able to work for all IN_FORMATS on a plane.
> Again, I have a sketch but no full implementation yet. This is not
> currently required by the AMD color pipeline and could be added after
> the merge of this set.
> 
> VKMS patches could still be improved in a few ways, though the
> payoff might be limited and I would rather focus on other work
> at the moment. The most obvious thing to improve would be to
> eliminate the hard-coded LUTs for identity, and sRGB, and replace
> them with fixed-point math instead.
> 
> There are plenty of things that I would like to see, but they could
> be added after the merge of this patchset:
>  - COLOR_ENCODING and COLOR_RANGE
>  - IN_FORMATS for a color pipeline
>  - Is it possible to support HW which can't bypass entire pipeline?
>  - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>  - read-only scaling colorop which defines scaling taps and position
>  - named matrices, for things like converting YUV to RGB
>  - Add custom LUT colorops to VKMS
> 
> IGT tests can be found at [1] or on the igt-dev mailing list. There
> have been no updates since v5 and rebase on latest main is straight-
> forward.
> 
> A kernel branch can be found at [2].
> 
> [1] https://gitlab.freedesktop.org/alex.hung/igt-gpu-tools/-/tree/amd-color-pipeline-v8
> [2] https://gitlab.freedesktop.org/alex.hung/linux/-/tree/amd-color-pipeline-v9
> 
> v9:
>  - Update RFC documents for 3DLUT and fallback behaviours (Simon Ser)
>  - Specify colorop function names by _plane_ (Chaitanya Kumar Borah)
>  - Remove redundant comments (Simon Ser)
>  - Fix typo in commit description (Shengyu Qu)
>  - Move destroy and cleanup functions earlier (Simon Ser)
>  - Move DRM_COLOROP_1D_CURVE_BT2020_* from middle to end (Simon Ser)
>  - Chagne "bool allow_bypass" to "uint32_t flags" for better extensibility (Simon Ser)
>  - Return a value in __set_dm_plane_colorop_3dlut
> 
> v8:
>  - Change VKMS config names (Louis Chauvet)
>  - Remove deprecated function "drm_atomic_get_existing_colorop_state" (Louis Chauvet)
>  - Remove null check in drm_colorop_set_next_property (Simon Ser)
>  - Remove MAX_COLOR_PIPELINES in drm (Simon Ser)
>  - Update kernel docs and documents for DRM_COLOROP_3D_LUT (Simon Ser)
>  - Add comments for dmr_color_lut (Louis Chauvet)
>  - Fix typos and replace DRM_ERROR and DRM_WARN_ONCE by drm_err drm_WARN_ONCE (Louis Chauvet)
>  - Fix incorrect conditions in __set_colorop_in_tf_1d_curve (Leo Li)
>  - Add DRM_MODE_PROP_ATOMIC to drm_property_create_range (Simon Ser)
>  - Change "1D Curve Custom LUT" to "1D LUT" (Simon Ser)
>  - Return error when __set_output_tf fails (Leo Li)
>  - Return -EINVAL when drm_color_ctm_3x4's size mismatches (Leo Li)
>  - Set initialized to 0 and return when drm_lut3d_size is 0 (Harry Wentland)
>  - Rework tf->type = TF_TYPE_BYPASS for shaper (Harry Wentland & Leo Li)
> 
> v7:
>  - Simplify 3D LUT by removing lut_3d_modes and update doc accordingly (Simon Ser)
>  - Add destroy / cleanup functions when color pipeline initialization
>    fails (Louis Chauvet)
>  - Fix checkpatch errors and warning (spaces, long lines, {}, and etc)
>  - Change lut's size from drm_colorop_state->size to drm_colorop->lut_size
>    and from size_property to lut_size_property
>  - Update some kernel updates and commit messagesa (Louis Chauvet)
>  - Squash "drm/colorop: Add atomic state print for drm_colorop" and
>    "drm/colorop: Add NEXT to colorop state print" (Simon Ser)
>  - Add missing MODULE_DESCRIPTION (Jeff Johnson)
> 
> v6:
>  - Eliminate need to include test as .c file (Louis Chauvet)
>  - some doc updates
>  - cleanup if colorop alloc or init fails in VKMS and amdgpu (Louis Chauvet)
>  - couple other small improvements in VKMS, such as rounding (Louis Chauvet)
>  - take ref for DATA blob in duplicate_state func & fix refcount issues (Xaver Hugl)
> 
> v5:
>  - amdgpu 3D LUT
>  - Don't require BYPASS
>  - update RFC docs and add to TOC tree
>  - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
>  - add amdgpu color pipeline doc
>  - define SIZE property similar to drm_crtc's GAMMA_SIZE
>  - various minor fixes and cleanups
> 
> v4:
>  - Add amdgpu color pipeline (WIP)
>  - Don't block setting of deprecated properties, instead pass client cap
>    to atomic check so drivers can ignore these props
>  - Drop IOCTL definitions (Pekka)
>  - Use enum property for colorop TYPE (Pekka)
>  - A few cleanups to the docs (Pekka)
>  - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
>  - Add missing function declarations (Chaitanya Kumar Borah)
>  - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
>  - Add helper for creation of pipeline drm_plane property (Pekka)
>  - Always create Bypass pipeline (Pekka)
>  - A bunch of changes to VKMS kunit tests (Pekka)
>  - Fix index in CTM doc (Pekka)
> 
> v3:
>  - Abandon IOCTLs and discover colorops as clients iterate the pipeline
>  - Remove need for libdrm
>  - Add color_pipeline client cap and make mutually exclusive with
>    COLOR_RANGE and COLOR_ENCODING properties
>  - add CTM colorop to VKMS
>  - Use include way for kunit testing static functions (Arthur)
>  - Make TYPE a range property
>  - Move enum drm_colorop_type to uapi header
>  - and a bunch of smaller bits that are highlighted in the relevant commit
>    description
> 
> v2:
>  - Rebased on drm-misc-next
>  - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
>  - Incorporate feedback in color_pipeline.rst doc
>  - Add support for sRGB inverse EOTF
>  - Add 2nd enumerated TF colorop to VKMS
>  - Fix LUTs and some issues with applying LUTs in VKMS
> 
> Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
> Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
> Cc: Simon Ser <contact@emersion.fr>
> Cc: Harry Wentland <harry.wentland@amd.com>
> Cc: Melissa Wen <mwen@igalia.com>
> Cc: Jonas Ådahl <jadahl@redhat.com>
> Cc: Sebastian Wick <sebastian.wick@redhat.com>
> Cc: Shashank Sharma <shashank.sharma@amd.com>
> Cc: Alexander Goins <agoins@nvidia.com>
> Cc: Joshua Ashton <joshua@froggi.es>
> Cc: Michel Dänzer <mdaenzer@redhat.com>
> Cc: Aleix Pol <aleixpol@kde.org>
> Cc: Xaver Hugl <xaver.hugl@gmail.com>
> Cc: Victoria Brekenfeld <victoria@system76.com>
> Cc: Sima <daniel@ffwll.ch>
> Cc: Uma Shankar <uma.shankar@intel.com>
> Cc: Naseer Ahmed <quic_naseer@quicinc.com>
> Cc: Christopher Braga <quic_cbraga@quicinc.com>
> Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> Cc: Hector Martin <marcan@marcan.st>
> Cc: Liviu Dudau <Liviu.Dudau@arm.com>
> Cc: Sasha McIntosh <sashamcintosh@google.com>
> Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
> 
> Alex Hung (14):
>   drm/colorop: Add destroy functions for color pipeline
>   drm/colorop: define a new macro for_each_new_colorop_in_state
>   drm/amd/display: Skip color pipeline initialization for cursor plane
>   drm/amd/display: Add support for sRGB EOTF in DEGAM block
>   drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
>   drm/amd/display: Add support for sRGB EOTF in BLND block
>   drm/colorop: Add 1D Curve Custom LUT type
>   drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
>   drm/amd/display: add 3x4 matrix colorop
>   drm/colorop: Add mutliplier type
>   drm/amd/display: add multiplier colorop
>   drm/amd/display: Swap matrix and multiplier
>   drm/colorop: Add 3D LUT support to color pipeline
>   drm/amd/display: add 3D LUT colorop
> 
> Harry Wentland (29):
>   drm: Add helper for conversion from signed-magnitude
>   drm/vkms: Add kunit tests for VKMS LUT handling
>   drm/doc/rfc: Describe why prescriptive color pipeline is needed
>   drm/colorop: Introduce new drm_colorop mode object
>   drm/colorop: Add TYPE property
>   drm/colorop: Add 1D Curve subtype
>   drm/colorop: Add BYPASS property
>   drm/colorop: Add NEXT property
>   drm/colorop: Add atomic state print for drm_colorop
>   drm/plane: Add COLOR PIPELINE property
>   drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
>   Documentation/gpu: document drm_colorop
>   drm/vkms: Add enumerated 1D curve colorop
>   drm/vkms: Add kunit tests for linear and sRGB LUTs
>   drm/colorop: Add 3x4 CTM type
>   drm/vkms: Use s32 for internal color pipeline precision
>   drm/vkms: add 3x4 matrix in color pipeline
>   drm/tests: Add a few tests around drm_fixed.h
>   drm/vkms: Add tests for CTM handling
>   drm/colorop: pass plane_color_pipeline client cap to atomic check
>   drm/amd/display: Ignore deprecated props when plane_color_pipeline set
>   drm/amd/display: Add bypass COLOR PIPELINE
>   drm/colorop: Add PQ 125 EOTF and its inverse
>   drm/amd/display: Enable support for PQ 125 EOTF and Inverse
>   drm/colorop: add BT2020/BT709 OETF and Inverse OETF
>   drm/amd/display: Add support for BT.709 and BT.2020 TFs
>   drm/colorop: Define LUT_1D interpolation
>   drm/colorop: allow non-bypass colorops
>   drm/amd/display: Add AMD color pipeline doc
> 
>  Documentation/gpu/drm-kms.rst                 |  15 +
>  Documentation/gpu/rfc/color_pipeline.rst      | 378 ++++++++
>  Documentation/gpu/rfc/index.rst               |   3 +
>  drivers/gpu/drm/Makefile                      |   1 +
>  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |   3 +-
>  .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |   4 +
>  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 573 ++++++++++++-
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 207 +++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
>  drivers/gpu/drm/drm_atomic.c                  | 170 +++-
>  drivers/gpu/drm/drm_atomic_helper.c           |  12 +
>  drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
>  drivers/gpu/drm/drm_atomic_uapi.c             | 175 ++++
>  drivers/gpu/drm/drm_colorop.c                 | 598 +++++++++++++
>  drivers/gpu/drm/drm_ioctl.c                   |   7 +
>  drivers/gpu/drm/drm_mode_config.c             |   7 +
>  drivers/gpu/drm/drm_plane.c                   |  59 ++
>  drivers/gpu/drm/tests/Makefile                |   3 +-
>  drivers/gpu/drm/tests/drm_fixp_test.c         |  71 ++
>  drivers/gpu/drm/vkms/Makefile                 |   4 +-
>  drivers/gpu/drm/vkms/tests/Makefile           |   2 +-
>  drivers/gpu/drm/vkms/tests/vkms_color_test.c  | 459 ++++++++++
>  drivers/gpu/drm/vkms/vkms_colorop.c           | 115 +++
>  drivers/gpu/drm/vkms/vkms_composer.c          | 134 ++-
>  drivers/gpu/drm/vkms/vkms_composer.h          |  27 +
>  drivers/gpu/drm/vkms/vkms_drv.h               |   7 +
>  drivers/gpu/drm/vkms/vkms_luts.c              | 808 ++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
>  drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
>  include/drm/drm_atomic.h                      | 111 +++
>  include/drm/drm_atomic_uapi.h                 |   3 +
>  include/drm/drm_colorop.h                     | 442 ++++++++++
>  include/drm/drm_file.h                        |   7 +
>  include/drm/drm_fixed.h                       |  18 +
>  include/drm/drm_mode_config.h                 |  18 +
>  include/drm/drm_plane.h                       |  19 +
>  include/uapi/drm/amdgpu_drm.h                 |   9 -
>  include/uapi/drm/drm.h                        |  15 +
>  include/uapi/drm/drm_mode.h                   | 121 +++
>  40 files changed, 4642 insertions(+), 50 deletions(-)
>  create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
>  create mode 100644 drivers/gpu/drm/drm_colorop.c
>  create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
>  create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_test.c
>  create mode 100644 drivers/gpu/drm/vkms/vkms_colorop.c
>  create mode 100644 drivers/gpu/drm/vkms/vkms_composer.h
>  create mode 100644 drivers/gpu/drm/vkms/vkms_luts.c
>  create mode 100644 drivers/gpu/drm/vkms/vkms_luts.h
>  create mode 100644 include/drm/drm_colorop.h
> 
> -- 
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2025-05-13 15:36       ` Melissa Wen
@ 2025-05-13 20:39         ` Harry Wentland
  2025-05-15 10:57           ` Pekka Paalanen
  0 siblings, 1 reply; 82+ messages in thread
From: Harry Wentland @ 2025-05-13 20:39 UTC (permalink / raw)
  To: Melissa Wen, Pekka Paalanen
  Cc: Alex Hung, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, contact, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, xaver.hugl, victoria, daniel,
	uma.shankar, quic_naseer, quic_cbraga, quic_abhinavk, marcan,
	Liviu.Dudau, sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Daniel Stone



On 2025-05-13 11:36, Melissa Wen wrote:
> On 05/13, Pekka Paalanen wrote:
>> On Mon, 12 May 2025 15:50:17 -0300
>> Melissa Wen <mwen@igalia.com> wrote:
>>
>>> On 04/29, Alex Hung wrote:
>>>> Expose one 1D curve colorop with support for
>>>> DRM_COLOROP_1D_CURVE_SRGB_EOTF and program HW to perform
>>>> the sRGB transform when the colorop is not in bypass.
>>>>
>>>> With this change the following IGT test passes:
>>>> kms_colorop --run plane-XR30-XR30-srgb_eotf
>>>>
>>>> The color pipeline now consists of a single colorop:
>>>> 1. 1D curve colorop w/ sRGB EOTF
>>>>
>>>> Signed-off-by: Alex Hung <alex.hung@amd.com>
>>>> Co-developed-by: Harry Wentland <harry.wentland@amd.com>
>>>> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
>>>> Reviewed-by: Daniel Stone <daniels@collabora.com>
>>>> ---
>>>> V9:
>>>>  - Update function names by _plane_ (Chaitanya Kumar Borah)
>>>>  - Update replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)
>>>>
>>>> v8:
>>>>  - Fix incorrect && by || in __set_colorop_in_tf_1d_curve (Leo Li)
>>>>
>>>> v7:
>>>>  - Fix checkpatch warnings
>>>>   - Change switch "{ }" position
>>>>   - Delete double ";"
>>>>   - Delete "{ }" for single-line if-statement
>>>>   - Add a new line at EOF
>>>>   - Change SPDX-License-Identifier: GPL-2.0+ from // to /* */
>>>>
>>>> v6:
>>>>  - cleanup if colorop alloc or init fails
>>>>
>>>>  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
>>>>  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 86 +++++++++++++++++++
>>>>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 69 +++++++++++++++
>>>>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 ++++++++
>>>>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
>>>>  5 files changed, 201 insertions(+), 1 deletion(-)
>>>>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>>>>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
>>>>
>>>> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
>>>> index ab2a97e354da..46158d67ab12 100644
>>>> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
>>>> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
>>>> @@ -38,7 +38,8 @@ AMDGPUDM = \
>>>>  	amdgpu_dm_pp_smu.o \
>>>>  	amdgpu_dm_psr.o \
>>>>  	amdgpu_dm_replay.o \
>>>> -	amdgpu_dm_wb.o
>>>> +	amdgpu_dm_wb.o \
>>>> +	amdgpu_dm_colorop.o
>>>>  
>>>>  ifdef CONFIG_DRM_AMD_DC_FP
>>>>  AMDGPUDM += dc_fpu.o
>>>> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
>>>> index ebabfe3a512f..0b513ab5050f 100644
>>>> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
>>>> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
>>>> @@ -668,6 +668,18 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf)
>>>>  	}
>>>>  }
>>>>  
>>>> +static enum dc_transfer_func_predefined
>>>> +amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
>>>> +{
>>>> +	switch (tf) {
>>>> +	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
>>>> +	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
>>>> +		return TRANSFER_FUNCTION_SRGB;
>>>> +	default:
>>>> +		return TRANSFER_FUNCTION_LINEAR;
>>>> +	}
>>>> +}
>>>> +
>>>>  static void __to_dc_lut3d_color(struct dc_rgb *rgb,
>>>>  				const struct drm_color_lut lut,
>>>>  				int bit_precision)
>>>> @@ -1137,6 +1149,59 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state,
>>>>  	return 0;
>>>>  }
>>>>  
>>>> +static int
>>>> +__set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
>>>> +		       struct drm_colorop_state *colorop_state)
>>>> +{
>>>> +	struct dc_transfer_func *tf = &dc_plane_state->in_transfer_func;
>>>> +	struct drm_colorop *colorop = colorop_state->colorop;
>>>> +	struct drm_device *drm = colorop->dev;
>>>> +
>>>> +	if (colorop->type != DRM_COLOROP_1D_CURVE ||
>>>> +	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
>>>> +		return -EINVAL;
>>>> +
>>>> +	if (colorop_state->bypass) {
>>>> +		tf->type = TF_TYPE_BYPASS;
>>>> +		tf->tf = TRANSFER_FUNCTION_LINEAR;
>>>> +		return 0;
>>>> +	}
>>>> +
>>>> +	drm_dbg(drm, "Degamma colorop with ID: %d\n", colorop->base.id);
>>>> +
>>>> +	tf->type = TF_TYPE_PREDEFINED;
>>>> +	tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
>>>> +
>>>> +	return 0;
>>>> +}
>>>> +
>>>> +static int
>>>> +__set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
>>>> +			       struct dc_plane_state *dc_plane_state,
>>>> +			       struct drm_colorop *colorop)
>>>> +{
>>>> +	struct drm_colorop *old_colorop;
>>>> +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
>>>> +	struct drm_atomic_state *state = plane_state->state;
>>>> +	int i = 0;
>>>> +
>>>> +	old_colorop = colorop;
>>>> +
>>>> +	/* 1st op: 1d curve - degamma */
>>>> +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
>>>> +		if (new_colorop_state->colorop == old_colorop &&
>>>> +		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
>>>> +			colorop_state = new_colorop_state;
>>>> +			break;
>>>> +		}
>>>> +	}
>>>> +
>>>> +	if (!colorop_state)
>>>> +		return -EINVAL;
>>>> +
>>>> +	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);  
>>>
>>> I wonder what will happen if plane degamma isn't set, but CRTC degamma
>>> LUT or legacy CRTC regamma LUT (with its implicity sRGB degamma) is used
>>> together with other plane color ops.
>>>
>>> I can imagine the mess, so I think CRTC degamma LUT and legacy CRTC
>>> regamma LUT should be somehow entirely disabled (or rejected) if plane
>>> color pipeline is in use.
>>
>> Hi Melissa,
>>
>> if using a plane color pipeline means that a CRTC LUT cannot be used, it
>> will severely limit the usefulness of the whole KMS color processing. In
>> Weston's case it would prohibit *all* KMS off-loading when color
>> management is in use.
>>
>> Weston chooses to do composition and blending in an optical space. This
>> means that plane color pipelines are required to convert incoming
>> pixels into the optical space, and a CRTC LUT (a CRTC color pipeline in
>> the future) is required to convert from the optical space to the
>> monitor signalling (electrical space).
> 
> Hi Pekka,
> 
> IIRC, Weston needs one post-blending 1D LUT and with my suggestion the
> atomic CRTC regamma LUT works fine and can be this 1D LUT.
> So, instead of an atomic post-blending/CRTC color pipeline with:
> 
> [blending] -> CRTC 1D LUT -> CRTC CTM -> CRTC 1D LUT
> 
> when plane color pipeline is in use, the driver accepts only:
> 
> [blending] -> CRTC CTM -> CRTC 1D LUT
> 
> If AMD driver continues accepting/exposing CRTC degamma LUT plus plane
> color pipeline, and userpace wants something like:
> 
> Plane shaper LUT -> Plane 3D LUT -> Plane Blnd LUT -> [blending] -> **CRTC** degamma LUT
> 
> I understand that this weird sequence is what will actually happen:
> 
> **CRTC** degamma LUT -> Plane shaper LUT -> Plane 3D LUT -> Plane Blnd LUT -> [blending]
> 

Right, and this is quirky behavior by the AMD driver because
historically we only had the CRTC LUTs and had to use the CRTC
degamma LUT to program degamma, which technically needs to happen
pre-blending, i.e., on the plane.

The degamma LUT never really made sense on a crtc, but here
we are.

> Because userspace doesn't care if this is a "degamma" or "regamma" LUT
> and they will probably pick the first 1D LUT in the post-blending color
> pipeline, which currently means CRTC degamma LUT. So, better if it takes

Ideally a userspace implementation would be the correct LUT
but not sure that would always happen.

> the CRTC regamma LUT that is actually a post-blending 1D LUT. Where I
> expected this result:
> 
> Plane shaper LUT -> PLane 3D LUT -> PLane Blnd LUT -> [blending] -> CRTC regamma LUT
> 
> You can vizualize better this degamma issue in this diagram:
> https://raw.githubusercontent.com/melissawen/melissawen.github.io/master/img/xdc-2023-colors-talk/rainbow-treasure-xdc-2023-17.png
> 
> In the current driver-specific implementation, driver rejects atomic commits
> in case of collision, i.e., Plane degamma LUT + CRTC degamma LUT, but
> IIRC, the driver doesn't handle it if it doesn't clash.
> 

Yeah, I agree the most straight-forward way to avoid problems
going forward is to not advertise the CRTC degamma LUT for HW
generations where we'll advertise the plane pipeline.

I talked to Alex to make that change.

Harry

>>
>> I don't know what "with its implicity sRGB degamma" means, but there
>> cannot be any implicit curves at all. The driver has no knowledge of
>> how the framebuffer pixels are encoded, nor about what the blending
>> space should be.
> 
> It targets the legacy CRTC regamma LUT (256 size), not the atomic CRTC
> color pipeline.
> 
> Melissa
> 
>>
>>
>> Thanks,
>> pq
> 
> 


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 03/43] drm/doc/rfc: Describe why prescriptive color pipeline is needed
  2025-04-30  1:10 ` [PATCH V9 03/43] drm/doc/rfc: Describe why prescriptive color pipeline is needed Alex Hung
@ 2025-05-15  9:30   ` Simon Ser
  0 siblings, 0 replies; 82+ messages in thread
From: Simon Ser @ 2025-05-15  9:30 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Reviewed-by: Simon Ser <contact@emersion.fr>

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 13/43] drm/colorop: Add destroy functions for color pipeline
  2025-04-30  1:10 ` [PATCH V9 13/43] drm/colorop: Add destroy functions for color pipeline Alex Hung
@ 2025-05-15  9:34   ` Simon Ser
  0 siblings, 0 replies; 82+ messages in thread
From: Simon Ser @ 2025-05-15  9:34 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Reviewed-by: Simon Ser <contact@emersion.fr>

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 31/43] drm/colorop: add BT2020/BT709 OETF and Inverse OETF
  2025-04-30  1:11 ` [PATCH V9 31/43] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Alex Hung
@ 2025-05-15  9:35   ` Simon Ser
  0 siblings, 0 replies; 82+ messages in thread
From: Simon Ser @ 2025-05-15  9:35 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Reviewed-by: Simon Ser <contact@emersion.fr>

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 40/43] drm/colorop: allow non-bypass colorops
  2025-04-30  1:11 ` [PATCH V9 40/43] drm/colorop: allow non-bypass colorops Alex Hung
@ 2025-05-15  9:37   ` Simon Ser
  0 siblings, 0 replies; 82+ messages in thread
From: Simon Ser @ 2025-05-15  9:37 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

Reviewed-by: Simon Ser <contact@emersion.fr>

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
                   ` (43 preceding siblings ...)
  2025-05-13 15:59 ` [PATCH V9 00/43] Color Pipeline API w/ VKMS Melissa Wen
@ 2025-05-15  9:45 ` Simon Ser
  2025-05-15 14:10   ` Harry Wentland
  44 siblings, 1 reply; 82+ messages in thread
From: Simon Ser @ 2025-05-15  9:45 UTC (permalink / raw)
  To: Alex Hung
  Cc: dri-devel, amd-gfx, wayland-devel, harry.wentland, leo.liu,
	ville.syrjala, pekka.paalanen, mwen, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo

I've reviewed all of the core DRM patches :)

Have there been updates from user-space implementations?

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2025-05-13 20:39         ` Harry Wentland
@ 2025-05-15 10:57           ` Pekka Paalanen
  0 siblings, 0 replies; 82+ messages in thread
From: Pekka Paalanen @ 2025-05-15 10:57 UTC (permalink / raw)
  To: Melissa Wen
  Cc: Harry Wentland, Alex Hung, dri-devel, amd-gfx, wayland-devel,
	leo.liu, ville.syrjala, contact, jadahl, sebastian.wick,
	shashank.sharma, agoins, joshua, mdaenzer, aleixpol, xaver.hugl,
	victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Daniel Stone

[-- Attachment #1: Type: text/plain, Size: 10470 bytes --]

On Tue, 13 May 2025 16:39:51 -0400
Harry Wentland <harry.wentland@amd.com> wrote:

> On 2025-05-13 11:36, Melissa Wen wrote:
> > On 05/13, Pekka Paalanen wrote:  
> >> On Mon, 12 May 2025 15:50:17 -0300
> >> Melissa Wen <mwen@igalia.com> wrote:
> >>  
> >>> On 04/29, Alex Hung wrote:  
> >>>> Expose one 1D curve colorop with support for
> >>>> DRM_COLOROP_1D_CURVE_SRGB_EOTF and program HW to perform
> >>>> the sRGB transform when the colorop is not in bypass.
> >>>>
> >>>> With this change the following IGT test passes:
> >>>> kms_colorop --run plane-XR30-XR30-srgb_eotf
> >>>>
> >>>> The color pipeline now consists of a single colorop:
> >>>> 1. 1D curve colorop w/ sRGB EOTF
> >>>>
> >>>> Signed-off-by: Alex Hung <alex.hung@amd.com>
> >>>> Co-developed-by: Harry Wentland <harry.wentland@amd.com>
> >>>> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> >>>> Reviewed-by: Daniel Stone <daniels@collabora.com>
> >>>> ---
> >>>> V9:
> >>>>  - Update function names by _plane_ (Chaitanya Kumar Borah)
> >>>>  - Update replace cleanup code by drm_colorop_pipeline_destroy (Simon Ser)
> >>>>
> >>>> v8:
> >>>>  - Fix incorrect && by || in __set_colorop_in_tf_1d_curve (Leo Li)
> >>>>
> >>>> v7:
> >>>>  - Fix checkpatch warnings
> >>>>   - Change switch "{ }" position
> >>>>   - Delete double ";"
> >>>>   - Delete "{ }" for single-line if-statement
> >>>>   - Add a new line at EOF
> >>>>   - Change SPDX-License-Identifier: GPL-2.0+ from // to /* */
> >>>>
> >>>> v6:
> >>>>  - cleanup if colorop alloc or init fails
> >>>>
> >>>>  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
> >>>>  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 86 +++++++++++++++++++
> >>>>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 69 +++++++++++++++
> >>>>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 ++++++++
> >>>>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
> >>>>  5 files changed, 201 insertions(+), 1 deletion(-)
> >>>>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> >>>>  create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> >>>>
> >>>> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> >>>> index ab2a97e354da..46158d67ab12 100644
> >>>> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> >>>> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
> >>>> @@ -38,7 +38,8 @@ AMDGPUDM = \
> >>>>  	amdgpu_dm_pp_smu.o \
> >>>>  	amdgpu_dm_psr.o \
> >>>>  	amdgpu_dm_replay.o \
> >>>> -	amdgpu_dm_wb.o
> >>>> +	amdgpu_dm_wb.o \
> >>>> +	amdgpu_dm_colorop.o
> >>>>  
> >>>>  ifdef CONFIG_DRM_AMD_DC_FP
> >>>>  AMDGPUDM += dc_fpu.o
> >>>> diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> >>>> index ebabfe3a512f..0b513ab5050f 100644
> >>>> --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> >>>> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
> >>>> @@ -668,6 +668,18 @@ amdgpu_tf_to_dc_tf(enum amdgpu_transfer_function tf)
> >>>>  	}
> >>>>  }
> >>>>  
> >>>> +static enum dc_transfer_func_predefined
> >>>> +amdgpu_colorop_tf_to_dc_tf(enum drm_colorop_curve_1d_type tf)
> >>>> +{
> >>>> +	switch (tf) {
> >>>> +	case DRM_COLOROP_1D_CURVE_SRGB_EOTF:
> >>>> +	case DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
> >>>> +		return TRANSFER_FUNCTION_SRGB;
> >>>> +	default:
> >>>> +		return TRANSFER_FUNCTION_LINEAR;
> >>>> +	}
> >>>> +}
> >>>> +
> >>>>  static void __to_dc_lut3d_color(struct dc_rgb *rgb,
> >>>>  				const struct drm_color_lut lut,
> >>>>  				int bit_precision)
> >>>> @@ -1137,6 +1149,59 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state,
> >>>>  	return 0;
> >>>>  }
> >>>>  
> >>>> +static int
> >>>> +__set_colorop_in_tf_1d_curve(struct dc_plane_state *dc_plane_state,
> >>>> +		       struct drm_colorop_state *colorop_state)
> >>>> +{
> >>>> +	struct dc_transfer_func *tf = &dc_plane_state->in_transfer_func;
> >>>> +	struct drm_colorop *colorop = colorop_state->colorop;
> >>>> +	struct drm_device *drm = colorop->dev;
> >>>> +
> >>>> +	if (colorop->type != DRM_COLOROP_1D_CURVE ||
> >>>> +	    colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_EOTF)
> >>>> +		return -EINVAL;
> >>>> +
> >>>> +	if (colorop_state->bypass) {
> >>>> +		tf->type = TF_TYPE_BYPASS;
> >>>> +		tf->tf = TRANSFER_FUNCTION_LINEAR;
> >>>> +		return 0;
> >>>> +	}
> >>>> +
> >>>> +	drm_dbg(drm, "Degamma colorop with ID: %d\n", colorop->base.id);
> >>>> +
> >>>> +	tf->type = TF_TYPE_PREDEFINED;
> >>>> +	tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
> >>>> +
> >>>> +	return 0;
> >>>> +}
> >>>> +
> >>>> +static int
> >>>> +__set_dm_plane_colorop_degamma(struct drm_plane_state *plane_state,
> >>>> +			       struct dc_plane_state *dc_plane_state,
> >>>> +			       struct drm_colorop *colorop)
> >>>> +{
> >>>> +	struct drm_colorop *old_colorop;
> >>>> +	struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
> >>>> +	struct drm_atomic_state *state = plane_state->state;
> >>>> +	int i = 0;
> >>>> +
> >>>> +	old_colorop = colorop;
> >>>> +
> >>>> +	/* 1st op: 1d curve - degamma */
> >>>> +	for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
> >>>> +		if (new_colorop_state->colorop == old_colorop &&
> >>>> +		    new_colorop_state->curve_1d_type == DRM_COLOROP_1D_CURVE_SRGB_EOTF) {
> >>>> +			colorop_state = new_colorop_state;
> >>>> +			break;
> >>>> +		}
> >>>> +	}
> >>>> +
> >>>> +	if (!colorop_state)
> >>>> +		return -EINVAL;
> >>>> +
> >>>> +	return __set_colorop_in_tf_1d_curve(dc_plane_state, colorop_state);    
> >>>
> >>> I wonder what will happen if plane degamma isn't set, but CRTC degamma
> >>> LUT or legacy CRTC regamma LUT (with its implicity sRGB degamma) is used
> >>> together with other plane color ops.
> >>>
> >>> I can imagine the mess, so I think CRTC degamma LUT and legacy CRTC
> >>> regamma LUT should be somehow entirely disabled (or rejected) if plane
> >>> color pipeline is in use.  
> >>
> >> Hi Melissa,
> >>
> >> if using a plane color pipeline means that a CRTC LUT cannot be used, it
> >> will severely limit the usefulness of the whole KMS color processing. In
> >> Weston's case it would prohibit *all* KMS off-loading when color
> >> management is in use.
> >>
> >> Weston chooses to do composition and blending in an optical space. This
> >> means that plane color pipelines are required to convert incoming
> >> pixels into the optical space, and a CRTC LUT (a CRTC color pipeline in
> >> the future) is required to convert from the optical space to the
> >> monitor signalling (electrical space).  
> > 
> > Hi Pekka,
> > 
> > IIRC, Weston needs one post-blending 1D LUT and with my suggestion the
> > atomic CRTC regamma LUT works fine and can be this 1D LUT.
> > So, instead of an atomic post-blending/CRTC color pipeline with:
> > 
> > [blending] -> CRTC 1D LUT -> CRTC CTM -> CRTC 1D LUT
> > 
> > when plane color pipeline is in use, the driver accepts only:
> > 
> > [blending] -> CRTC CTM -> CRTC 1D LUT
> > 

Hi Melissa,

sorry, that is fine. I misunderstood disabling both degamma LUT and
regamma LUT as no LUT stage being left available on a CRTC. After all,
only GAMMA_LUT and DEGAMMA_LUT exist upstream.

> > If AMD driver continues accepting/exposing CRTC degamma LUT plus plane
> > color pipeline, and userpace wants something like:
> > 
> > Plane shaper LUT -> Plane 3D LUT -> Plane Blnd LUT -> [blending] -> **CRTC** degamma LUT
> > 
> > I understand that this weird sequence is what will actually happen:
> > 
> > **CRTC** degamma LUT -> Plane shaper LUT -> Plane 3D LUT -> Plane Blnd LUT -> [blending]
> >   
> 
> Right, and this is quirky behavior by the AMD driver because
> historically we only had the CRTC LUTs and had to use the CRTC
> degamma LUT to program degamma, which technically needs to happen
> pre-blending, i.e., on the plane.
> 
> The degamma LUT never really made sense on a crtc, but here
> we are.

Huh. I always assumed that CRTC DEGAMMA_LUT was something in the
hardware after blending and that someone just wanted to expose that
hardware feature. I'd never have guessed that it might have applied
before blending on purpose.

The position of the CRTC CTM before or after blending does not matter.
The positions are mathematically equivalent I think.

> > Because userspace doesn't care if this is a "degamma" or "regamma" LUT
> > and they will probably pick the first 1D LUT in the post-blending color
> > pipeline, which currently means CRTC degamma LUT. So, better if it takes  
> 
> Ideally a userspace implementation would be the correct LUT
> but not sure that would always happen.
> 

FWIW, people who rely on ICC profile VCGT might be quite precise about
how the VCGT is applied. I've been essentially called ignorant for
applying VCGT anywhere else than in "the" hardware LUT.

> > the CRTC regamma LUT that is actually a post-blending 1D LUT. Where I
> > expected this result:
> > 
> > Plane shaper LUT -> PLane 3D LUT -> PLane Blnd LUT -> [blending] -> CRTC regamma LUT
> > 
> > You can vizualize better this degamma issue in this diagram:
> > https://raw.githubusercontent.com/melissawen/melissawen.github.io/master/img/xdc-2023-colors-talk/rainbow-treasure-xdc-2023-17.png
> > 
> > In the current driver-specific implementation, driver rejects atomic commits
> > in case of collision, i.e., Plane degamma LUT + CRTC degamma LUT, but
> > IIRC, the driver doesn't handle it if it doesn't clash.
> >   
> 
> Yeah, I agree the most straight-forward way to avoid problems
> going forward is to not advertise the CRTC degamma LUT for HW
> generations where we'll advertise the plane pipeline.
> 
> I talked to Alex to make that change.

Sounds good to me.


Thanks,
pq

> >> I don't know what "with its implicity sRGB degamma" means, but there
> >> cannot be any implicit curves at all. The driver has no knowledge of
> >> how the framebuffer pixels are encoded, nor about what the blending
> >> space should be.  
> > 
> > It targets the legacy CRTC regamma LUT (256 size), not the atomic CRTC
> > color pipeline.
> > 
> > Melissa
> >   
> >>
> >>
> >> Thanks,
> >> pq  
> > 
> >   
> 


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15  9:45 ` Simon Ser
@ 2025-05-15 14:10   ` Harry Wentland
  2025-05-15 14:15     ` Simon Ser
  2025-05-15 17:19     ` Daniel Stone
  0 siblings, 2 replies; 82+ messages in thread
From: Harry Wentland @ 2025-05-15 14:10 UTC (permalink / raw)
  To: Simon Ser, Alex Hung, Misyl Toad, Xaver Hugl, Ribeiro
  Cc: dri-devel, amd-gfx, wayland-devel, leo.liu, ville.syrjala,
	pekka.paalanen, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, xaver.hugl, victoria, daniel,
	uma.shankar, quic_naseer, quic_cbraga, quic_abhinavk, marcan,
	Liviu.Dudau, sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo



On 2025-05-15 05:45, Simon Ser wrote:
> I've reviewed all of the core DRM patches :)
> 
> Have there been updates from user-space implementations?

I know Leandro has been working on Weston to make use of
this and last year Xaver had a prototype in kwin.

Ideally we'd have gamescope adopt it and switch off the
"temporary" AMD private properties and show that it
behaves the same without breaking things.

I added/moved the relevant folks into the "To" field.

Harry

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15 14:10   ` Harry Wentland
@ 2025-05-15 14:15     ` Simon Ser
  2025-05-15 17:19     ` Daniel Stone
  1 sibling, 0 replies; 82+ messages in thread
From: Simon Ser @ 2025-05-15 14:15 UTC (permalink / raw)
  To: Harry Wentland
  Cc: Alex Hung, Misyl Toad, xaver.hugl, Ribeiro, dri-devel, amd-gfx,
	wayland-devel, leo.liu, ville.syrjala, pekka.paalanen, mwen,
	jadahl, sebastian.wick, shashank.sharma, agoins, joshua, mdaenzer,
	aleixpol, victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo

On Thursday, May 15th, 2025 at 16:11, Harry Wentland <harry.wentland@amd.com> wrote:

> > Have there been updates from user-space implementations?
> 
> I know Leandro has been working on Weston to make use of
> this and last year Xaver had a prototype in kwin.

Cool!

> Ideally we'd have gamescope adopt it and switch off the
> "temporary" AMD private properties and show that it
> behaves the same without breaking things.

Right, I've started working on this a long time ago [1], but didn't
manage to finish. I will not be able to continue this work, feel free to
continue the work, pick apart useful bits or start from scratch, as you
prefer.

Ideally I'd like to add support in libliftoff/wlroots as well.

[1]: https://github.com/ValveSoftware/gamescope/pull/1309

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15 14:10   ` Harry Wentland
  2025-05-15 14:15     ` Simon Ser
@ 2025-05-15 17:19     ` Daniel Stone
  2025-05-15 18:02       ` Harry Wentland
  1 sibling, 1 reply; 82+ messages in thread
From: Daniel Stone @ 2025-05-15 17:19 UTC (permalink / raw)
  To: Harry Wentland
  Cc: Simon Ser, Alex Hung, Misyl Toad, Xaver Hugl, Ribeiro, dri-devel,
	amd-gfx, wayland-devel, leo.liu, ville.syrjala, pekka.paalanen,
	mwen, jadahl, sebastian.wick, shashank.sharma, agoins, joshua,
	mdaenzer, aleixpol, victoria, daniel, uma.shankar, quic_naseer,
	quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo

Hi,

On Thu, 15 May 2025 at 15:11, Harry Wentland <harry.wentland@amd.com> wrote:
> On 2025-05-15 05:45, Simon Ser wrote:
> > I've reviewed all of the core DRM patches :)
> >
> > Have there been updates from user-space implementations?
>
> I know Leandro has been working on Weston to make use of
> this and last year Xaver had a prototype in kwin.

Yeah, the Weston patches are marching on. We've still been doing a
little bit of cleanup and prep work in the background to land them,
but we also can't land them until the kernel lands. None of that work
is material to the uAPI though: as said previously, the uAPI looks
completely solid and it's something we can definitely beneficially use
in Weston. (Even if we do need the obvious follow-ons for
post-blending as well ...)

Cheers,
Daniel

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15 17:19     ` Daniel Stone
@ 2025-05-15 18:02       ` Harry Wentland
  2025-05-15 18:39         ` Daniel Stone
  0 siblings, 1 reply; 82+ messages in thread
From: Harry Wentland @ 2025-05-15 18:02 UTC (permalink / raw)
  To: Daniel Stone
  Cc: Simon Ser, Alex Hung, Misyl Toad, Xaver Hugl, Ribeiro, dri-devel,
	amd-gfx, wayland-devel, leo.liu, ville.syrjala, pekka.paalanen,
	mwen, jadahl, sebastian.wick, shashank.sharma, agoins, joshua,
	mdaenzer, aleixpol, victoria, daniel, uma.shankar, quic_naseer,
	quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo



On 2025-05-15 13:19, Daniel Stone wrote:
> Hi,
> 
> On Thu, 15 May 2025 at 15:11, Harry Wentland <harry.wentland@amd.com> wrote:
>> On 2025-05-15 05:45, Simon Ser wrote:
>>> I've reviewed all of the core DRM patches :)
>>>
>>> Have there been updates from user-space implementations?
>>
>> I know Leandro has been working on Weston to make use of
>> this and last year Xaver had a prototype in kwin.
> 
> Yeah, the Weston patches are marching on. We've still been doing a
> little bit of cleanup and prep work in the background to land them,
> but we also can't land them until the kernel lands. None of that work
> is material to the uAPI though: as said previously, the uAPI looks
> completely solid and it's something we can definitely beneficially use
> in Weston. (Even if we do need the obvious follow-ons for
> post-blending as well ...)
> 

We can't merge kernel uAPI without canonical userspace that uses it.
To move forward we'll need a userspace to at least publish a branch
that shows the use of this new uAPI.

Do you have a public branch for the Weston work for this?

Harry

> Cheers,
> Daniel


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15 18:02       ` Harry Wentland
@ 2025-05-15 18:39         ` Daniel Stone
  2025-05-15 19:59           ` Leandro Ribeiro
  0 siblings, 1 reply; 82+ messages in thread
From: Daniel Stone @ 2025-05-15 18:39 UTC (permalink / raw)
  To: Harry Wentland
  Cc: Simon Ser, Alex Hung, Misyl Toad, Xaver Hugl, Ribeiro, dri-devel,
	amd-gfx, wayland-devel, leo.liu, ville.syrjala, pekka.paalanen,
	mwen, jadahl, sebastian.wick, shashank.sharma, agoins, joshua,
	mdaenzer, aleixpol, victoria, daniel, uma.shankar, quic_naseer,
	quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo

Hi,

On Thu, 15 May 2025 at 19:02, Harry Wentland <harry.wentland@amd.com> wrote:
> On 2025-05-15 13:19, Daniel Stone wrote:
> > Yeah, the Weston patches are marching on. We've still been doing a
> > little bit of cleanup and prep work in the background to land them,
> > but we also can't land them until the kernel lands. None of that work
> > is material to the uAPI though: as said previously, the uAPI looks
> > completely solid and it's something we can definitely beneficially use
> > in Weston. (Even if we do need the obvious follow-ons for
> > post-blending as well ...)
>
> We can't merge kernel uAPI without canonical userspace that uses it.
> To move forward we'll need a userspace to at least publish a branch
> that shows the use of this new uAPI.
>
> Do you have a public branch for the Weston work for this?

Yeah, https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/1702
has been around for a little while now. There are some driver bugs
that Leandro commented on, but they don't seem material to the uAPI as
such?

Cheers,
Daniel

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15 18:39         ` Daniel Stone
@ 2025-05-15 19:59           ` Leandro Ribeiro
  2025-05-17 11:51             ` Xaver Hugl
  0 siblings, 1 reply; 82+ messages in thread
From: Leandro Ribeiro @ 2025-05-15 19:59 UTC (permalink / raw)
  To: Daniel Stone, Harry Wentland
  Cc: Simon Ser, Alex Hung, Misyl Toad, Xaver Hugl, dri-devel, amd-gfx,
	wayland-devel, leo.liu, ville.syrjala, pekka.paalanen, mwen,
	jadahl, sebastian.wick, shashank.sharma, agoins, joshua, mdaenzer,
	aleixpol, victoria, daniel, uma.shankar, quic_naseer, quic_cbraga,
	quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo



On 5/15/25 15:39, Daniel Stone wrote:
> Hi,
> 
> On Thu, 15 May 2025 at 19:02, Harry Wentland <harry.wentland@amd.com> wrote:
>> On 2025-05-15 13:19, Daniel Stone wrote:
>>> Yeah, the Weston patches are marching on. We've still been doing a
>>> little bit of cleanup and prep work in the background to land them,
>>> but we also can't land them until the kernel lands. None of that work
>>> is material to the uAPI though: as said previously, the uAPI looks
>>> completely solid and it's something we can definitely beneficially use
>>> in Weston. (Even if we do need the obvious follow-ons for
>>> post-blending as well ...)
>>
>> We can't merge kernel uAPI without canonical userspace that uses it.
>> To move forward we'll need a userspace to at least publish a branch
>> that shows the use of this new uAPI.
>>
>> Do you have a public branch for the Weston work for this?
> 
> Yeah, https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/1702
> has been around for a little while now. There are some driver bugs
> that Leandro commented on, but they don't seem material to the uAPI as
> such?

Hello,

Yes, there's nothing related to the API that is blocking us. It seemed
very flexible and easy to use. The bugs that I've spotted are probably
internal to AMD driver.

I'd say that the Weston patches are converging nicely, we just need time
to get them fully reviewed. We had a few preparation MR's to land
before !1702, and now there's only one left (!1617).

Thanks,
Leandro
> 
> Cheers,
> Daniel


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-15 19:59           ` Leandro Ribeiro
@ 2025-05-17 11:51             ` Xaver Hugl
  2025-05-21 19:48               ` Harry Wentland
  0 siblings, 1 reply; 82+ messages in thread
From: Xaver Hugl @ 2025-05-17 11:51 UTC (permalink / raw)
  To: Leandro Ribeiro
  Cc: Daniel Stone, Harry Wentland, Simon Ser, Alex Hung, Misyl Toad,
	dri-devel, amd-gfx, wayland-devel, leo.liu, ville.syrjala,
	pekka.paalanen, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo

Am Do., 15. Mai 2025 um 22:00 Uhr schrieb Leandro Ribeiro
<leandro.ribeiro@collabora.com>:
>
>
>
> On 5/15/25 15:39, Daniel Stone wrote:
> > Hi,
> >
> > On Thu, 15 May 2025 at 19:02, Harry Wentland <harry.wentland@amd.com> wrote:
> >> On 2025-05-15 13:19, Daniel Stone wrote:
> >>> Yeah, the Weston patches are marching on. We've still been doing a
> >>> little bit of cleanup and prep work in the background to land them,
> >>> but we also can't land them until the kernel lands. None of that work
> >>> is material to the uAPI though: as said previously, the uAPI looks
> >>> completely solid and it's something we can definitely beneficially use
> >>> in Weston. (Even if we do need the obvious follow-ons for
> >>> post-blending as well ...)
> >>
> >> We can't merge kernel uAPI without canonical userspace that uses it.
> >> To move forward we'll need a userspace to at least publish a branch
> >> that shows the use of this new uAPI.
> >>
> >> Do you have a public branch for the Weston work for this?
> >
> > Yeah, https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/1702
> > has been around for a little while now. There are some driver bugs
> > that Leandro commented on, but they don't seem material to the uAPI as
> > such?
>
> Hello,
>
> Yes, there's nothing related to the API that is blocking us. It seemed
> very flexible and easy to use. The bugs that I've spotted are probably
> internal to AMD driver.
>
> I'd say that the Weston patches are converging nicely, we just need time
> to get them fully reviewed. We had a few preparation MR's to land
> before !1702, and now there's only one left (!1617).

I also updated the KWin MR
(https://invent.kde.org/plasma/kwin/-/merge_requests/6600), it can now
use all the available properties and I think it's ready. I found two
issues with the kernel patches though:
- while attempting to set COLOR_ENCODING and COLOR_RANGE results in
the atomic commit being rejected, the existing values still get
applied if you use YCbCr-type buffers. I would've expected the color
pipeline to operate on the YUV values in that case - and leave
conversion to RGB up to the compositor adding the relevant matrix to
the pipeline
- the interpolation mode drm properties for 1D and 3D LUTs are
immutable, I think they shouldn't be - to make it less annoying if in
the future we decide to add modes that userspace can set

Other than that, I agree that it's ready to go.

> Thanks,
> Leandro
> >
> > Cheers,
> > Daniel
>

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-17 11:51             ` Xaver Hugl
@ 2025-05-21 19:48               ` Harry Wentland
  2025-05-22  7:57                 ` Pekka Paalanen
  0 siblings, 1 reply; 82+ messages in thread
From: Harry Wentland @ 2025-05-21 19:48 UTC (permalink / raw)
  To: Xaver Hugl, Leandro Ribeiro
  Cc: Daniel Stone, Simon Ser, Alex Hung, Misyl Toad, dri-devel,
	amd-gfx, wayland-devel, leo.liu, ville.syrjala, pekka.paalanen,
	mwen, jadahl, sebastian.wick, shashank.sharma, agoins, joshua,
	mdaenzer, aleixpol, victoria, daniel, uma.shankar, quic_naseer,
	quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau, sashamcintosh,
	chaitanya.kumar.borah, louis.chauvet, Arthur Grillo



On 2025-05-17 07:51, Xaver Hugl wrote:
> Am Do., 15. Mai 2025 um 22:00 Uhr schrieb Leandro Ribeiro
> <leandro.ribeiro@collabora.com>:
>>
>>
>>
>> On 5/15/25 15:39, Daniel Stone wrote:
>>> Hi,
>>>
>>> On Thu, 15 May 2025 at 19:02, Harry Wentland <harry.wentland@amd.com> wrote:
>>>> On 2025-05-15 13:19, Daniel Stone wrote:
>>>>> Yeah, the Weston patches are marching on. We've still been doing a
>>>>> little bit of cleanup and prep work in the background to land them,
>>>>> but we also can't land them until the kernel lands. None of that work
>>>>> is material to the uAPI though: as said previously, the uAPI looks
>>>>> completely solid and it's something we can definitely beneficially use
>>>>> in Weston. (Even if we do need the obvious follow-ons for
>>>>> post-blending as well ...)
>>>>
>>>> We can't merge kernel uAPI without canonical userspace that uses it.
>>>> To move forward we'll need a userspace to at least publish a branch
>>>> that shows the use of this new uAPI.
>>>>
>>>> Do you have a public branch for the Weston work for this?
>>>
>>> Yeah, https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/1702
>>> has been around for a little while now. There are some driver bugs
>>> that Leandro commented on, but they don't seem material to the uAPI as
>>> such?
>>
>> Hello,
>>
>> Yes, there's nothing related to the API that is blocking us. It seemed
>> very flexible and easy to use. The bugs that I've spotted are probably
>> internal to AMD driver.
>>
>> I'd say that the Weston patches are converging nicely, we just need time
>> to get them fully reviewed. We had a few preparation MR's to land
>> before !1702, and now there's only one left (!1617).
> 
> I also updated the KWin MR
> (https://invent.kde.org/plasma/kwin/-/merge_requests/6600), it can now
> use all the available properties and I think it's ready. I found two
> issues with the kernel patches though:
> - while attempting to set COLOR_ENCODING and COLOR_RANGE results in
> the atomic commit being rejected, the existing values still get
> applied if you use YCbCr-type buffers. I would've expected the color
> pipeline to operate on the YUV values in that case - and leave
> conversion to RGB up to the compositor adding the relevant matrix to
> the pipeline

AMD HW always operates on RGB values, so there'll always be an
implicit conversion of YCbCr-type buffers to RGB. What we should
do is reject YCbCr-type buffers with the color pipeline until we
implement support for COLOR_ENCODING and COLOR_RANGE as a new
CSC colorop.

> - the interpolation mode drm properties for 1D and 3D LUTs are
> immutable, I think they shouldn't be - to make it less annoying if in
> the future we decide to add modes that userspace can set
> 

Makes sense to me.

Harry

> Other than that, I agree that it's ready to go.
> 
>> Thanks,
>> Leandro
>>>
>>> Cheers,
>>> Daniel
>>


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-21 19:48               ` Harry Wentland
@ 2025-05-22  7:57                 ` Pekka Paalanen
  2025-05-22 13:28                   ` Harry Wentland
  0 siblings, 1 reply; 82+ messages in thread
From: Pekka Paalanen @ 2025-05-22  7:57 UTC (permalink / raw)
  To: Harry Wentland
  Cc: Xaver Hugl, Leandro Ribeiro, Daniel Stone, Simon Ser, Alex Hung,
	Misyl Toad, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo

[-- Attachment #1: Type: text/plain, Size: 3853 bytes --]

On Wed, 21 May 2025 15:48:00 -0400
Harry Wentland <harry.wentland@amd.com> wrote:

> On 2025-05-17 07:51, Xaver Hugl wrote:
> > Am Do., 15. Mai 2025 um 22:00 Uhr schrieb Leandro Ribeiro
> > <leandro.ribeiro@collabora.com>:  
> >>
> >>
> >>
> >> On 5/15/25 15:39, Daniel Stone wrote:  
> >>> Hi,
> >>>
> >>> On Thu, 15 May 2025 at 19:02, Harry Wentland <harry.wentland@amd.com> wrote:  
> >>>> On 2025-05-15 13:19, Daniel Stone wrote:  
> >>>>> Yeah, the Weston patches are marching on. We've still been doing a
> >>>>> little bit of cleanup and prep work in the background to land them,
> >>>>> but we also can't land them until the kernel lands. None of that work
> >>>>> is material to the uAPI though: as said previously, the uAPI looks
> >>>>> completely solid and it's something we can definitely beneficially use
> >>>>> in Weston. (Even if we do need the obvious follow-ons for
> >>>>> post-blending as well ...)  
> >>>>
> >>>> We can't merge kernel uAPI without canonical userspace that uses it.
> >>>> To move forward we'll need a userspace to at least publish a branch
> >>>> that shows the use of this new uAPI.
> >>>>
> >>>> Do you have a public branch for the Weston work for this?  
> >>>
> >>> Yeah, https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/1702
> >>> has been around for a little while now. There are some driver bugs
> >>> that Leandro commented on, but they don't seem material to the uAPI as
> >>> such?  
> >>
> >> Hello,
> >>
> >> Yes, there's nothing related to the API that is blocking us. It seemed
> >> very flexible and easy to use. The bugs that I've spotted are probably
> >> internal to AMD driver.
> >>
> >> I'd say that the Weston patches are converging nicely, we just need time
> >> to get them fully reviewed. We had a few preparation MR's to land
> >> before !1702, and now there's only one left (!1617).  
> > 
> > I also updated the KWin MR
> > (https://invent.kde.org/plasma/kwin/-/merge_requests/6600), it can now
> > use all the available properties and I think it's ready. I found two
> > issues with the kernel patches though:
> > - while attempting to set COLOR_ENCODING and COLOR_RANGE results in
> > the atomic commit being rejected, the existing values still get
> > applied if you use YCbCr-type buffers. I would've expected the color
> > pipeline to operate on the YUV values in that case - and leave
> > conversion to RGB up to the compositor adding the relevant matrix to
> > the pipeline  
> 
> AMD HW always operates on RGB values, so there'll always be an
> implicit conversion of YCbCr-type buffers to RGB.

Surely the transformation from YCbCr to RGB can be one that maps each
of Y, Cb and Cr channels to G, B and R ranges [0.0, 1.0]?

Of course, that's not identity transformation (Cb and Cr have domain
[-0.5, 0.5] or something like that) and R, B may not handle negative
values.

But if R and B channels can be negative too, then identity mapping
would be usable.

There will need to be a definition of how YCbCr enters the color
pipeline: the order of the channels and their domains. The rest can be
just more colorops.

We should be careful to allow limited range YUV formats to be
interpreted as full range to avoid clipping the sub-black and
super-white signal ranges. The color pipeline can be configured to deal
with those ranges as necessary.

> What we should
> do is reject YCbCr-type buffers with the color pipeline until we
> implement support for COLOR_ENCODING and COLOR_RANGE as a new
> CSC colorop.

Rejecting is fine, but is implementing COLOR_ENCODING and COLOR_RANGE
really a good idea instead of making the color pipelines handle them?

Wasn't the original plan to hide all such legacy plane properties when
userspace signals color pipeline support?


Thanks,
pq

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-22  7:57                 ` Pekka Paalanen
@ 2025-05-22 13:28                   ` Harry Wentland
  2025-05-22 13:49                     ` Simon Ser
  0 siblings, 1 reply; 82+ messages in thread
From: Harry Wentland @ 2025-05-22 13:28 UTC (permalink / raw)
  To: Pekka Paalanen
  Cc: Xaver Hugl, Leandro Ribeiro, Daniel Stone, Simon Ser, Alex Hung,
	Misyl Toad, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo



On 2025-05-22 03:57, Pekka Paalanen wrote:
> On Wed, 21 May 2025 15:48:00 -0400
> Harry Wentland <harry.wentland@amd.com> wrote:
> 
>> On 2025-05-17 07:51, Xaver Hugl wrote:
>>> Am Do., 15. Mai 2025 um 22:00 Uhr schrieb Leandro Ribeiro
>>> <leandro.ribeiro@collabora.com>:  
>>>>
>>>>
>>>>
>>>> On 5/15/25 15:39, Daniel Stone wrote:  
>>>>> Hi,
>>>>>
>>>>> On Thu, 15 May 2025 at 19:02, Harry Wentland <harry.wentland@amd.com> wrote:  
>>>>>> On 2025-05-15 13:19, Daniel Stone wrote:  
>>>>>>> Yeah, the Weston patches are marching on. We've still been doing a
>>>>>>> little bit of cleanup and prep work in the background to land them,
>>>>>>> but we also can't land them until the kernel lands. None of that work
>>>>>>> is material to the uAPI though: as said previously, the uAPI looks
>>>>>>> completely solid and it's something we can definitely beneficially use
>>>>>>> in Weston. (Even if we do need the obvious follow-ons for
>>>>>>> post-blending as well ...)  
>>>>>>
>>>>>> We can't merge kernel uAPI without canonical userspace that uses it.
>>>>>> To move forward we'll need a userspace to at least publish a branch
>>>>>> that shows the use of this new uAPI.
>>>>>>
>>>>>> Do you have a public branch for the Weston work for this?  
>>>>>
>>>>> Yeah, https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/1702
>>>>> has been around for a little while now. There are some driver bugs
>>>>> that Leandro commented on, but they don't seem material to the uAPI as
>>>>> such?  
>>>>
>>>> Hello,
>>>>
>>>> Yes, there's nothing related to the API that is blocking us. It seemed
>>>> very flexible and easy to use. The bugs that I've spotted are probably
>>>> internal to AMD driver.
>>>>
>>>> I'd say that the Weston patches are converging nicely, we just need time
>>>> to get them fully reviewed. We had a few preparation MR's to land
>>>> before !1702, and now there's only one left (!1617).  
>>>
>>> I also updated the KWin MR
>>> (https://invent.kde.org/plasma/kwin/-/merge_requests/6600), it can now
>>> use all the available properties and I think it's ready. I found two
>>> issues with the kernel patches though:
>>> - while attempting to set COLOR_ENCODING and COLOR_RANGE results in
>>> the atomic commit being rejected, the existing values still get
>>> applied if you use YCbCr-type buffers. I would've expected the color
>>> pipeline to operate on the YUV values in that case - and leave
>>> conversion to RGB up to the compositor adding the relevant matrix to
>>> the pipeline  
>>
>> AMD HW always operates on RGB values, so there'll always be an
>> implicit conversion of YCbCr-type buffers to RGB.
> 
> Surely the transformation from YCbCr to RGB can be one that maps each
> of Y, Cb and Cr channels to G, B and R ranges [0.0, 1.0]?
> 

Probably, but that's not a HW config that's been explored, while the
one performing a color-space conversion has been used for years.

> Of course, that's not identity transformation (Cb and Cr have domain
> [-0.5, 0.5] or something like that) and R, B may not handle negative
> values.
> 
> But if R and B channels can be negative too, then identity mapping
> would be usable.
> 
> There will need to be a definition of how YCbCr enters the color
> pipeline: the order of the channels and their domains. The rest can be
> just more colorops.
> 
> We should be careful to allow limited range YUV formats to be
> interpreted as full range to avoid clipping the sub-black and
> super-white signal ranges. The color pipeline can be configured to deal
> with those ranges as necessary.
> 
>> What we should
>> do is reject YCbCr-type buffers with the color pipeline until we
>> implement support for COLOR_ENCODING and COLOR_RANGE as a new
>> CSC colorop.
> 
> Rejecting is fine, but is implementing COLOR_ENCODING and COLOR_RANGE
> really a good idea instead of making the color pipelines handle them?
> 
> Wasn't the original plan to hide all such legacy plane properties when
> userspace signals color pipeline support?
> 

It still is. But handling the color-space conversion via a new
colorop with two properties: COLOR_ENCODING and COLOR_RANGE
seemed to be the most straight-forward way of dealing with this.

I'm open to other suggestions and we can explore them.

Harry

> 
> Thanks,
> pq


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-22 13:28                   ` Harry Wentland
@ 2025-05-22 13:49                     ` Simon Ser
  2025-05-22 13:54                       ` Harry Wentland
  0 siblings, 1 reply; 82+ messages in thread
From: Simon Ser @ 2025-05-22 13:49 UTC (permalink / raw)
  To: Harry Wentland
  Cc: Pekka Paalanen, Xaver Hugl, Leandro Ribeiro, Daniel Stone,
	Alex Hung, Misyl Toad, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo

On Thursday, May 22nd, 2025 at 15:28, Harry Wentland <harry.wentland@amd.com> wrote:

> > > What we should
> > > do is reject YCbCr-type buffers with the color pipeline until we
> > > implement support for COLOR_ENCODING and COLOR_RANGE as a new
> > > CSC colorop.
> > 
> > Rejecting is fine, but is implementing COLOR_ENCODING and COLOR_RANGE
> > really a good idea instead of making the color pipelines handle them?
> > 
> > Wasn't the original plan to hide all such legacy plane properties when
> > userspace signals color pipeline support?
> 
> It still is. But handling the color-space conversion via a new
> colorop with two properties: COLOR_ENCODING and COLOR_RANGE
> seemed to be the most straight-forward way of dealing with this.

I think everybody in this thread is suggesting that the old plane
COLOR_ENCODING and COLOR_RANGE properties are rejected and will never
be accepted again with the color pipeline. I think Harry is just saying
that a new colorop can be introduced with encoding/range properties, but
this is completely separate from the old props?

Simon

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-22 13:49                     ` Simon Ser
@ 2025-05-22 13:54                       ` Harry Wentland
  2025-05-22 15:27                         ` Pekka Paalanen
  0 siblings, 1 reply; 82+ messages in thread
From: Harry Wentland @ 2025-05-22 13:54 UTC (permalink / raw)
  To: Simon Ser
  Cc: Pekka Paalanen, Xaver Hugl, Leandro Ribeiro, Daniel Stone,
	Alex Hung, Misyl Toad, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo



On 2025-05-22 09:49, Simon Ser wrote:
> On Thursday, May 22nd, 2025 at 15:28, Harry Wentland <harry.wentland@amd.com> wrote:
> 
>>>> What we should
>>>> do is reject YCbCr-type buffers with the color pipeline until we
>>>> implement support for COLOR_ENCODING and COLOR_RANGE as a new
>>>> CSC colorop.
>>>
>>> Rejecting is fine, but is implementing COLOR_ENCODING and COLOR_RANGE
>>> really a good idea instead of making the color pipelines handle them?
>>>
>>> Wasn't the original plan to hide all such legacy plane properties when
>>> userspace signals color pipeline support?
>>
>> It still is. But handling the color-space conversion via a new
>> colorop with two properties: COLOR_ENCODING and COLOR_RANGE
>> seemed to be the most straight-forward way of dealing with this.
> 
> I think everybody in this thread is suggesting that the old plane
> COLOR_ENCODING and COLOR_RANGE properties are rejected and will never
> be accepted again with the color pipeline. I think Harry is just saying
> that a new colorop can be introduced with encoding/range properties, but
> this is completely separate from the old props?
> 

Correct. Thanks for clarifying.

Harry

> Simon


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-22 13:54                       ` Harry Wentland
@ 2025-05-22 15:27                         ` Pekka Paalanen
  2025-05-22 15:49                           ` Harry Wentland
  0 siblings, 1 reply; 82+ messages in thread
From: Pekka Paalanen @ 2025-05-22 15:27 UTC (permalink / raw)
  To: Harry Wentland
  Cc: Simon Ser, Xaver Hugl, Leandro Ribeiro, Daniel Stone, Alex Hung,
	Misyl Toad, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo

[-- Attachment #1: Type: text/plain, Size: 1450 bytes --]

On Thu, 22 May 2025 09:54:50 -0400
Harry Wentland <harry.wentland@amd.com> wrote:

> On 2025-05-22 09:49, Simon Ser wrote:
> > On Thursday, May 22nd, 2025 at 15:28, Harry Wentland <harry.wentland@amd.com> wrote:
> >   
> >>>> What we should
> >>>> do is reject YCbCr-type buffers with the color pipeline until we
> >>>> implement support for COLOR_ENCODING and COLOR_RANGE as a new
> >>>> CSC colorop.  
> >>>
> >>> Rejecting is fine, but is implementing COLOR_ENCODING and COLOR_RANGE
> >>> really a good idea instead of making the color pipelines handle them?
> >>>
> >>> Wasn't the original plan to hide all such legacy plane properties when
> >>> userspace signals color pipeline support?  
> >>
> >> It still is. But handling the color-space conversion via a new
> >> colorop with two properties: COLOR_ENCODING and COLOR_RANGE
> >> seemed to be the most straight-forward way of dealing with this.  
> > 
> > I think everybody in this thread is suggesting that the old plane
> > COLOR_ENCODING and COLOR_RANGE properties are rejected and will never
> > be accepted again with the color pipeline. I think Harry is just saying
> > that a new colorop can be introduced with encoding/range properties, but
> > this is completely separate from the old props?
> >   
> 
> Correct. Thanks for clarifying.

Oh! Ok, that's cool. I thought we were talking about KMS plane
properties and not a new colorop.


Thanks,
pq

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 00/43] Color Pipeline API w/ VKMS
  2025-05-22 15:27                         ` Pekka Paalanen
@ 2025-05-22 15:49                           ` Harry Wentland
  0 siblings, 0 replies; 82+ messages in thread
From: Harry Wentland @ 2025-05-22 15:49 UTC (permalink / raw)
  To: Pekka Paalanen
  Cc: Simon Ser, Xaver Hugl, Leandro Ribeiro, Daniel Stone, Alex Hung,
	Misyl Toad, dri-devel, amd-gfx, wayland-devel, leo.liu,
	ville.syrjala, mwen, jadahl, sebastian.wick, shashank.sharma,
	agoins, joshua, mdaenzer, aleixpol, victoria, daniel, uma.shankar,
	quic_naseer, quic_cbraga, quic_abhinavk, marcan, Liviu.Dudau,
	sashamcintosh, chaitanya.kumar.borah, louis.chauvet,
	Arthur Grillo



On 2025-05-22 11:27, Pekka Paalanen wrote:
> On Thu, 22 May 2025 09:54:50 -0400
> Harry Wentland <harry.wentland@amd.com> wrote:
> 
>> On 2025-05-22 09:49, Simon Ser wrote:
>>> On Thursday, May 22nd, 2025 at 15:28, Harry Wentland <harry.wentland@amd.com> wrote:
>>>   
>>>>>> What we should
>>>>>> do is reject YCbCr-type buffers with the color pipeline until we
>>>>>> implement support for COLOR_ENCODING and COLOR_RANGE as a new
>>>>>> CSC colorop.  
>>>>>
>>>>> Rejecting is fine, but is implementing COLOR_ENCODING and COLOR_RANGE
>>>>> really a good idea instead of making the color pipelines handle them?
>>>>>
>>>>> Wasn't the original plan to hide all such legacy plane properties when
>>>>> userspace signals color pipeline support?  
>>>>
>>>> It still is. But handling the color-space conversion via a new
>>>> colorop with two properties: COLOR_ENCODING and COLOR_RANGE
>>>> seemed to be the most straight-forward way of dealing with this.  
>>>
>>> I think everybody in this thread is suggesting that the old plane
>>> COLOR_ENCODING and COLOR_RANGE properties are rejected and will never
>>> be accepted again with the color pipeline. I think Harry is just saying
>>> that a new colorop can be introduced with encoding/range properties, but
>>> this is completely separate from the old props?
>>>   
>>
>> Correct. Thanks for clarifying.
> 
> Oh! Ok, that's cool. I thought we were talking about KMS plane
> properties and not a new colorop.
> 

My conciseness/verbosity filter just needed some tuning. :)

Harry

> 
> Thanks,
> pq


^ permalink raw reply	[flat|nested] 82+ messages in thread

* RE: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
  2025-04-30  1:10 ` [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type Alex Hung
@ 2025-06-16 11:30   ` Borah, Chaitanya Kumar
  2025-06-16 13:32     ` Pekka Paalanen
  0 siblings, 1 reply; 82+ messages in thread
From: Borah, Chaitanya Kumar @ 2025-06-16 11:30 UTC (permalink / raw)
  To: Alex Hung, dri-devel@lists.freedesktop.org,
	amd-gfx@lists.freedesktop.org
  Cc: wayland-devel@lists.freedesktop.org, harry.wentland@amd.com,
	leo.liu@amd.com, ville.syrjala@linux.intel.com,
	pekka.paalanen@collabora.com, contact@emersion.fr,
	mwen@igalia.com, jadahl@redhat.com, sebastian.wick@redhat.com,
	shashank.sharma@amd.com, agoins@nvidia.com, joshua@froggi.es,
	mdaenzer@redhat.com, aleixpol@kde.org, xaver.hugl@gmail.com,
	victoria@system76.com, daniel@ffwll.ch, Shankar, Uma,
	quic_naseer@quicinc.com, quic_cbraga@quicinc.com,
	quic_abhinavk@quicinc.com, marcan@marcan.st, Liviu.Dudau@arm.com,
	sashamcintosh@google.com, louis.chauvet@bootlin.com, Daniel Stone



> -----Original Message-----
> From: Alex Hung <alex.hung@amd.com>
> Sent: Wednesday, April 30, 2025 6:41 AM
> To: dri-devel@lists.freedesktop.org; amd-gfx@lists.freedesktop.org
> Cc: wayland-devel@lists.freedesktop.org; harry.wentland@amd.com;
> alex.hung@amd.com; leo.liu@amd.com; ville.syrjala@linux.intel.com;
> pekka.paalanen@collabora.com; contact@emersion.fr; mwen@igalia.com;
> jadahl@redhat.com; sebastian.wick@redhat.com;
> shashank.sharma@amd.com; agoins@nvidia.com; joshua@froggi.es;
> mdaenzer@redhat.com; aleixpol@kde.org; xaver.hugl@gmail.com;
> victoria@system76.com; daniel@ffwll.ch; Shankar, Uma
> <uma.shankar@intel.com>; quic_naseer@quicinc.com;
> quic_cbraga@quicinc.com; quic_abhinavk@quicinc.com; marcan@marcan.st;
> Liviu.Dudau@arm.com; sashamcintosh@google.com; Borah, Chaitanya
> Kumar <chaitanya.kumar.borah@intel.com>; louis.chauvet@bootlin.com;
> Daniel Stone <daniels@collabora.com>
> Subject: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
> 
> From: Harry Wentland <harry.wentland@amd.com>
> 
> This type is used to support a 3x4 matrix in colorops. A 3x4 matrix uses the
> last column as a "bias" column. Some HW exposes support for 3x4. The
> calculation looks like:
> 
>  out   matrix    in
>  |R|   |0  1  2  3 |   | R |
>  |G| = |4  5  6  7 | x | G |
>  |B|   |8  9  10 11|   | B |
>                        |1.0|
> 
> This is also the first colorop where we need a blob property to program the
> property. For that we'll introduce a new DATA property that can be used by all
> colorop TYPEs requiring a blob. The way a DATA blob is read depends on the
> TYPE of the colorop.
> 
> We only create the DATA property for property types that need it.

Is there any value to adding pre-offsets [1] in the uapi? 

 |R/Cr|    | c0 c1 c2 |   ( |R/Cr|   |preoff0| )   |postoff0|
 |G/Y | = | c3 c4 c5 | x ( |G/Y | + |preoff1| ) + |postoff1|
 |B/Cb|   | c6 c7 c8 |   ( |B/Cb|   |preoff2| )   |postoff2|

Handling limited range values is one use case that I can think of.  

[1] https://cgit.freedesktop.org/drm-tip/tree/drivers/gpu/drm/i915/display/intel_color.c#n112

Regards

Chaitanya

> 
> Reviewed-by: Simon Ser <contact@emersion.fr>
> Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Reviewed-by: Daniel Stone <daniels@collabora.com>
> ---
> v9:
>  - Merge cleanup code for colorop->state->data in drm_colorop_cleanup
> (Simon Ser)
>  - Update function names by _plane_ (Chaitanya Kumar Borah)
> 
> v6:
>  - take ref for DATA blob in duplicate_state func (Xaver Hugl)
> 
> v5:
>  - Add function signature for init (Sebastian)
>  - Fix kernel-doc
> 
> v4:
>  - Create helper function for creating 3x4 CTM colorop
>  - Fix CTM indexes in comment (Pekka)
> 
>  drivers/gpu/drm/drm_atomic.c      |  3 ++
>  drivers/gpu/drm/drm_atomic_uapi.c | 30 ++++++++++++++++++++
>  drivers/gpu/drm/drm_colorop.c     | 47 +++++++++++++++++++++++++++++++
>  include/drm/drm_colorop.h         | 24 ++++++++++++++++
>  include/uapi/drm/amdgpu_drm.h     |  9 ------
>  include/uapi/drm/drm_mode.h       | 32 ++++++++++++++++++++-
>  6 files changed, 135 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
> index 7b042c38d50d..809bd856d378 100644
> --- a/drivers/gpu/drm/drm_atomic.c
> +++ b/drivers/gpu/drm/drm_atomic.c
> @@ -793,6 +793,9 @@ static void drm_atomic_colorop_print_state(struct
> drm_printer *p,
>  		drm_printf(p, "\tcurve_1d_type=%s\n",
>  			   drm_get_colorop_curve_1d_type_name(state-
> >curve_1d_type));
>  		break;
> +	case DRM_COLOROP_CTM_3X4:
> +		drm_printf(p, "\tdata blob id=%d\n", state->data ? state-
> >data->base.id : 0);
> +		break;
>  	default:
>  		break;
>  	}
> diff --git a/drivers/gpu/drm/drm_atomic_uapi.c
> b/drivers/gpu/drm/drm_atomic_uapi.c
> index c0bcaec249de..be73cb9f502e 100644
> --- a/drivers/gpu/drm/drm_atomic_uapi.c
> +++ b/drivers/gpu/drm/drm_atomic_uapi.c
> @@ -702,6 +702,31 @@ drm_atomic_plane_get_property(struct drm_plane
> *plane,
>  	return 0;
>  }
> 
> +static int drm_atomic_color_set_data_property(struct drm_colorop
> *colorop,
> +					      struct drm_colorop_state *state,
> +					      struct drm_property *property,
> +					      uint64_t val)
> +{
> +	ssize_t elem_size = -1;
> +	ssize_t size = -1;
> +	bool replaced = false;
> +
> +	switch (colorop->type) {
> +	case DRM_COLOROP_CTM_3X4:
> +		size = sizeof(struct drm_color_ctm_3x4);
> +		break;
> +	default:
> +		/* should never get here */
> +		return -EINVAL;
> +	}
> +
> +	return drm_property_replace_blob_from_id(colorop->dev,
> +						 &state->data,
> +						 val,
> +						 size,
> +						 elem_size,
> +						 &replaced);
> +}
> 
>  static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
>  		struct drm_colorop_state *state, struct drm_file *file_priv,
> @@ -711,6 +736,9 @@ static int drm_atomic_colorop_set_property(struct
> drm_colorop *colorop,
>  		state->bypass = val;
>  	} else if (property == colorop->curve_1d_type_property) {
>  		state->curve_1d_type = val;
> +	} else if (property == colorop->data_property) {
> +		return drm_atomic_color_set_data_property(colorop, state,
> +							  property, val);
>  	} else {
>  		drm_dbg_atomic(colorop->dev,
>  			       "[COLOROP:%d:%d] unknown property
> [PROP:%d:%s]]\n", @@ -733,6 +761,8 @@
> drm_atomic_colorop_get_property(struct drm_colorop *colorop,
>  		*val = state->bypass;
>  	} else if (property == colorop->curve_1d_type_property) {
>  		*val = state->curve_1d_type;
> +	} else if (property == colorop->data_property) {
> +		*val = (state->data) ? state->data->base.id : 0;
>  	} else {
>  		return -EINVAL;
>  	}
> diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
> index 57a8c1063fdd..65351aaa7771 100644
> --- a/drivers/gpu/drm/drm_colorop.c
> +++ b/drivers/gpu/drm/drm_colorop.c
> @@ -64,6 +64,7 @@
> 
>  static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
>  	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
> +	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
>  };
> 
>  static const char * const colorop_curve_1d_type_names[] = { @@ -148,6
> +149,11 @@ static void drm_colorop_cleanup(struct drm_colorop *colorop)
>  	list_del(&colorop->head);
>  	config->num_colorop--;
> 
> +	if (colorop->state && colorop->state->data) {
> +		drm_property_blob_put(colorop->state->data);
> +		colorop->state->data = NULL;
> +	}
> +
>  	kfree(colorop->state);
>  }
> 
> @@ -238,11 +244,51 @@ int drm_plane_colorop_curve_1d_init(struct
> drm_device *dev, struct drm_colorop *  }
> EXPORT_SYMBOL(drm_plane_colorop_curve_1d_init);
> 
> +static int drm_colorop_create_data_prop(struct drm_device *dev, struct
> +drm_colorop *colorop) {
> +	struct drm_property *prop;
> +
> +	/* data */
> +	prop = drm_property_create(dev, DRM_MODE_PROP_ATOMIC |
> DRM_MODE_PROP_BLOB,
> +				   "DATA", 0);
> +	if (!prop)
> +		return -ENOMEM;
> +
> +	colorop->data_property = prop;
> +	drm_object_attach_property(&colorop->base,
> +				   colorop->data_property,
> +				   0);
> +
> +	return 0;
> +}
> +
> +int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct
> drm_colorop *colorop,
> +				   struct drm_plane *plane)
> +{
> +	int ret;
> +
> +	ret = drm_plane_colorop_init(dev, colorop, plane,
> DRM_COLOROP_CTM_3X4);
> +	if (ret)
> +		return ret;
> +
> +	ret = drm_colorop_create_data_prop(dev, colorop);
> +	if (ret)
> +		return ret;
> +
> +	drm_colorop_reset(colorop);
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL(drm_plane_colorop_ctm_3x4_init);
> +
>  static void __drm_atomic_helper_colorop_duplicate_state(struct
> drm_colorop *colorop,
>  							struct
> drm_colorop_state *state)  {
>  	memcpy(state, colorop->state, sizeof(*state));
> 
> +	if (state->data)
> +		drm_property_blob_get(state->data);
> +
>  	state->bypass = true;
>  }
> 
> @@ -324,6 +370,7 @@ void drm_colorop_reset(struct drm_colorop
> *colorop)
> 
>  static const char * const colorop_type_name[] = {
>  	[DRM_COLOROP_1D_CURVE] = "1D Curve",
> +	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
>  };
> 
>  const char *drm_get_colorop_type_name(enum drm_colorop_type type)
> diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h index
> 8e4d79ba1eff..5f0cddc3922f 100644
> --- a/include/drm/drm_colorop.h
> +++ b/include/drm/drm_colorop.h
> @@ -99,6 +99,17 @@ struct drm_colorop_state {
>  	 */
>  	enum drm_colorop_curve_1d_type curve_1d_type;
> 
> +	/**
> +	 * @data:
> +	 *
> +	 * Data blob for any TYPE that requires such a blob. The
> +	 * interpretation of the blob is TYPE-specific.
> +	 *
> +	 * See the &drm_colorop_type documentation for how blob is laid
> +	 * out.
> +	 */
> +	struct drm_property_blob *data;
> +
>  	/** @state: backpointer to global drm_atomic_state */
>  	struct drm_atomic_state *state;
>  };
> @@ -208,6 +219,17 @@ struct drm_colorop {
>  	 */
>  	struct drm_property *curve_1d_type_property;
> 
> +	/**
> +	 * @data_property:
> +	 *
> +	 * blob property for any TYPE that requires a blob of data,
> +	 * such as 1DLUT, CTM, 3DLUT, etc.
> +	 *
> +	 * The way this blob is interpreted depends on the TYPE of
> +	 * this
> +	 */
> +	struct drm_property *data_property;
> +
>  	/**
>  	 * @next_property:
>  	 *
> @@ -243,6 +265,8 @@ void drm_colorop_pipeline_destroy(struct drm_plane
> *plane);
> 
>  int drm_plane_colorop_curve_1d_init(struct drm_device *dev, struct
> drm_colorop *colorop,
>  				    struct drm_plane *plane, u64
> supported_tfs);
> +int drm_plane_colorop_ctm_3x4_init(struct drm_device *dev, struct
> drm_colorop *colorop,
> +				   struct drm_plane *plane);
> 
>  struct drm_colorop_state *
>  drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
> diff --git a/include/uapi/drm/amdgpu_drm.h
> b/include/uapi/drm/amdgpu_drm.h index 25d5c6e90a99..96fabf9c9827
> 100644
> --- a/include/uapi/drm/amdgpu_drm.h
> +++ b/include/uapi/drm/amdgpu_drm.h
> @@ -1300,15 +1300,6 @@ struct drm_amdgpu_info_gpuvm_fault {
>  #define AMDGPU_FAMILY_GC_11_5_0			150 /* GC 11.5.0 */
>  #define AMDGPU_FAMILY_GC_12_0_0			152 /* GC 12.0.0 */
> 
> -/* FIXME wrong namespace! */
> -struct drm_color_ctm_3x4 {
> -	/*
> -	 * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude
> -	 * (not two's complement!) format.
> -	 */
> -	__u64 matrix[12];
> -};
> -
>  #if defined(__cplusplus)
>  }
>  #endif
> diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
> index ea6d88f683cd..651bdf48b766 100644
> --- a/include/uapi/drm/drm_mode.h
> +++ b/include/uapi/drm/drm_mode.h
> @@ -847,6 +847,20 @@ struct drm_color_ctm {
>  	__u64 matrix[9];
>  };
> 
> +struct drm_color_ctm_3x4 {
> +	/*
> +	 * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude
> +	 * (not two's complement!) format.
> +	 *
> +	 * out   matrix          in
> +	 * |R|   |0  1  2  3 |   | R |
> +	 * |G| = |4  5  6  7 | x | G |
> +	 * |B|   |8  9  10 11|   | B |
> +	 *                       |1.0|
> +	 */
> +	__u64 matrix[12];
> +};
> +
>  struct drm_color_lut {
>  	/*
>  	 * Values are mapped linearly to 0.0 - 1.0 range, with 0x0 == 0.0 and
> @@ -874,7 +888,23 @@ enum drm_colorop_type {
>  	 * A 1D curve that is being applied to all color channels. The
>  	 * curve is specified via the CURVE_1D_TYPE colorop property.
>  	 */
> -	DRM_COLOROP_1D_CURVE
> +	DRM_COLOROP_1D_CURVE,
> +
> +	/**
> +	 * @DRM_COLOROP_CTM_3X4:
> +	 *
> +	 * enum string "3x4 Matrix"
> +	 *
> +	 * A 3x4 matrix. Its values are specified via the
> +	 * &drm_color_ctm_3x4 struct provided via the DATA property.
> +	 *
> +	 * The DATA blob is a float[12]:
> +	 * out   matrix          in
> +	 * | R |   | 0  1  2  3  |   | R |
> +	 * | G | = | 4  5  6  7  | x | G |
> +	 * | B |   | 8  9  10 12 |   | B |
> +	 */
> +	DRM_COLOROP_CTM_3X4,
>  };
> 
>  /**
> --
> 2.43.0


^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
  2025-06-16 11:30   ` Borah, Chaitanya Kumar
@ 2025-06-16 13:32     ` Pekka Paalanen
  2025-06-17  5:33       ` Borah, Chaitanya Kumar
  0 siblings, 1 reply; 82+ messages in thread
From: Pekka Paalanen @ 2025-06-16 13:32 UTC (permalink / raw)
  To: Borah, Chaitanya Kumar
  Cc: Alex Hung, dri-devel@lists.freedesktop.org,
	amd-gfx@lists.freedesktop.org,
	wayland-devel@lists.freedesktop.org, harry.wentland@amd.com,
	leo.liu@amd.com, ville.syrjala@linux.intel.com, mwen@igalia.com,
	jadahl@redhat.com, sebastian.wick@redhat.com,
	shashank.sharma@amd.com, agoins@nvidia.com, joshua@froggi.es,
	mdaenzer@redhat.com, aleixpol@kde.org, xaver.hugl@gmail.com,
	victoria@system76.com, daniel@ffwll.ch, Shankar, Uma,
	quic_naseer@quicinc.com, quic_cbraga@quicinc.com,
	quic_abhinavk@quicinc.com, marcan@marcan.st, Liviu.Dudau@arm.com,
	sashamcintosh@google.com, louis.chauvet@bootlin.com, Daniel Stone

[-- Attachment #1: Type: text/plain, Size: 2647 bytes --]

On Mon, 16 Jun 2025 11:30:23 +0000
"Borah, Chaitanya Kumar" <chaitanya.kumar.borah@intel.com> wrote:

> > -----Original Message-----
> > From: Alex Hung <alex.hung@amd.com>
> > Sent: Wednesday, April 30, 2025 6:41 AM
> > To: dri-devel@lists.freedesktop.org; amd-gfx@lists.freedesktop.org
> > Cc: wayland-devel@lists.freedesktop.org; harry.wentland@amd.com;
> > alex.hung@amd.com; leo.liu@amd.com; ville.syrjala@linux.intel.com;
> > pekka.paalanen@collabora.com; contact@emersion.fr; mwen@igalia.com;
> > jadahl@redhat.com; sebastian.wick@redhat.com;
> > shashank.sharma@amd.com; agoins@nvidia.com; joshua@froggi.es;
> > mdaenzer@redhat.com; aleixpol@kde.org; xaver.hugl@gmail.com;
> > victoria@system76.com; daniel@ffwll.ch; Shankar, Uma
> > <uma.shankar@intel.com>; quic_naseer@quicinc.com;
> > quic_cbraga@quicinc.com; quic_abhinavk@quicinc.com; marcan@marcan.st;
> > Liviu.Dudau@arm.com; sashamcintosh@google.com; Borah, Chaitanya
> > Kumar <chaitanya.kumar.borah@intel.com>; louis.chauvet@bootlin.com;
> > Daniel Stone <daniels@collabora.com>
> > Subject: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
> > 
> > From: Harry Wentland <harry.wentland@amd.com>
> > 
> > This type is used to support a 3x4 matrix in colorops. A 3x4 matrix uses the
> > last column as a "bias" column. Some HW exposes support for 3x4. The
> > calculation looks like:
> > 
> >  out   matrix    in
> >  |R|   |0  1  2  3 |   | R |
> >  |G| = |4  5  6  7 | x | G |
> >  |B|   |8  9  10 11|   | B |
> >                        |1.0|
> > 
> > This is also the first colorop where we need a blob property to program the
> > property. For that we'll introduce a new DATA property that can be used by all
> > colorop TYPEs requiring a blob. The way a DATA blob is read depends on the
> > TYPE of the colorop.
> > 
> > We only create the DATA property for property types that need it.  
> 
> Is there any value to adding pre-offsets [1] in the uapi? 
> 
>  |R/Cr|    | c0 c1 c2 |   ( |R/Cr|   |preoff0| )   |postoff0|
>  |G/Y | = | c3 c4 c5 | x ( |G/Y | + |preoff1| ) + |postoff1|
>  |B/Cb|   | c6 c7 c8 |   ( |B/Cb|   |preoff2| )   |postoff2|
> 
> Handling limited range values is one use case that I can think of.  

Hi,

in the mathematical sense, no. A pre-offset can always be converted
into a post-offset by multiplying it with the 3x3 matrix (and adding to
the existing post-offset). This can be pre-computed, no need to do
it separately for every pixel.

For hardware reasons, I have no idea.

> [1] https://cgit.freedesktop.org/drm-tip/tree/drivers/gpu/drm/i915/display/intel_color.c#n112
> 

Thanks,
pq

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 82+ messages in thread

* RE: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
  2025-06-16 13:32     ` Pekka Paalanen
@ 2025-06-17  5:33       ` Borah, Chaitanya Kumar
  2025-06-17 15:25         ` Ville Syrjälä
  0 siblings, 1 reply; 82+ messages in thread
From: Borah, Chaitanya Kumar @ 2025-06-17  5:33 UTC (permalink / raw)
  To: Pekka Paalanen
  Cc: Alex Hung, dri-devel@lists.freedesktop.org,
	amd-gfx@lists.freedesktop.org,
	wayland-devel@lists.freedesktop.org, harry.wentland@amd.com,
	leo.liu@amd.com, ville.syrjala@linux.intel.com, mwen@igalia.com,
	jadahl@redhat.com, sebastian.wick@redhat.com,
	shashank.sharma@amd.com, agoins@nvidia.com, joshua@froggi.es,
	mdaenzer@redhat.com, aleixpol@kde.org, xaver.hugl@gmail.com,
	victoria@system76.com, daniel@ffwll.ch, Shankar, Uma,
	quic_naseer@quicinc.com, quic_cbraga@quicinc.com,
	quic_abhinavk@quicinc.com, marcan@marcan.st, Liviu.Dudau@arm.com,
	sashamcintosh@google.com, louis.chauvet@bootlin.com, Daniel Stone



> -----Original Message-----
> From: Pekka Paalanen <pekka.paalanen@collabora.com>
> Sent: Monday, June 16, 2025 7:02 PM
> To: Borah, Chaitanya Kumar <chaitanya.kumar.borah@intel.com>
> Cc: Alex Hung <alex.hung@amd.com>; dri-devel@lists.freedesktop.org; amd-
> gfx@lists.freedesktop.org; wayland-devel@lists.freedesktop.org;
> harry.wentland@amd.com; leo.liu@amd.com; ville.syrjala@linux.intel.com;
> mwen@igalia.com; jadahl@redhat.com; sebastian.wick@redhat.com;
> shashank.sharma@amd.com; agoins@nvidia.com; joshua@froggi.es;
> mdaenzer@redhat.com; aleixpol@kde.org; xaver.hugl@gmail.com;
> victoria@system76.com; daniel@ffwll.ch; Shankar, Uma
> <uma.shankar@intel.com>; quic_naseer@quicinc.com;
> quic_cbraga@quicinc.com; quic_abhinavk@quicinc.com; marcan@marcan.st;
> Liviu.Dudau@arm.com; sashamcintosh@google.com;
> louis.chauvet@bootlin.com; Daniel Stone <daniels@collabora.com>
> Subject: Re: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
> 
> On Mon, 16 Jun 2025 11:30:23 +0000
> "Borah, Chaitanya Kumar" <chaitanya.kumar.borah@intel.com> wrote:
> 
> > > -----Original Message-----
> > > From: Alex Hung <alex.hung@amd.com>
> > > Sent: Wednesday, April 30, 2025 6:41 AM
> > > To: dri-devel@lists.freedesktop.org; amd-gfx@lists.freedesktop.org
> > > Cc: wayland-devel@lists.freedesktop.org; harry.wentland@amd.com;
> > > alex.hung@amd.com; leo.liu@amd.com; ville.syrjala@linux.intel.com;
> > > pekka.paalanen@collabora.com; contact@emersion.fr; mwen@igalia.com;
> > > jadahl@redhat.com; sebastian.wick@redhat.com;
> > > shashank.sharma@amd.com; agoins@nvidia.com; joshua@froggi.es;
> > > mdaenzer@redhat.com; aleixpol@kde.org; xaver.hugl@gmail.com;
> > > victoria@system76.com; daniel@ffwll.ch; Shankar, Uma
> > > <uma.shankar@intel.com>; quic_naseer@quicinc.com;
> > > quic_cbraga@quicinc.com; quic_abhinavk@quicinc.com;
> > > marcan@marcan.st; Liviu.Dudau@arm.com; sashamcintosh@google.com;
> > > Borah, Chaitanya Kumar <chaitanya.kumar.borah@intel.com>;
> > > louis.chauvet@bootlin.com; Daniel Stone <daniels@collabora.com>
> > > Subject: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
> > >
> > > From: Harry Wentland <harry.wentland@amd.com>
> > >
> > > This type is used to support a 3x4 matrix in colorops. A 3x4 matrix
> > > uses the last column as a "bias" column. Some HW exposes support for
> > > 3x4. The calculation looks like:
> > >
> > >  out   matrix    in
> > >  |R|   |0  1  2  3 |   | R |
> > >  |G| = |4  5  6  7 | x | G |
> > >  |B|   |8  9  10 11|   | B |
> > >                        |1.0|
> > >
> > > This is also the first colorop where we need a blob property to
> > > program the property. For that we'll introduce a new DATA property
> > > that can be used by all colorop TYPEs requiring a blob. The way a
> > > DATA blob is read depends on the TYPE of the colorop.
> > >
> > > We only create the DATA property for property types that need it.
> >
> > Is there any value to adding pre-offsets [1] in the uapi?
> >
> >  |R/Cr|    | c0 c1 c2 |   ( |R/Cr|   |preoff0| )   |postoff0|
> >  |G/Y | = | c3 c4 c5 | x ( |G/Y | + |preoff1| ) + |postoff1|
> >  |B/Cb|   | c6 c7 c8 |   ( |B/Cb|   |preoff2| )   |postoff2|
> >
> > Handling limited range values is one use case that I can think of.
> 
> Hi,
> 
> in the mathematical sense, no. A pre-offset can always be converted into a
> post-offset by multiplying it with the 3x3 matrix (and adding to the existing
> post-offset). This can be pre-computed, no need to do it separately for every
> pixel.
> 
> For hardware reasons, I have no idea.

Thank you for the reply, Pekka. Our hardware does allow programming Pre-offsets separately.
Currently I can't think of a particular advantage of that if mathematically a post-offset does the job but I will keep this thread posted if I find something out.

Regards

Chaitanya

> 
> > [1]
> > https://cgit.freedesktop.org/drm-tip/tree/drivers/gpu/drm/i915/display
> > /intel_color.c#n112
> >
> 
> Thanks,
> pq

^ permalink raw reply	[flat|nested] 82+ messages in thread

* Re: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
  2025-06-17  5:33       ` Borah, Chaitanya Kumar
@ 2025-06-17 15:25         ` Ville Syrjälä
  0 siblings, 0 replies; 82+ messages in thread
From: Ville Syrjälä @ 2025-06-17 15:25 UTC (permalink / raw)
  To: Borah, Chaitanya Kumar
  Cc: Pekka Paalanen, Alex Hung, dri-devel@lists.freedesktop.org,
	amd-gfx@lists.freedesktop.org,
	wayland-devel@lists.freedesktop.org, harry.wentland@amd.com,
	leo.liu@amd.com, mwen@igalia.com, jadahl@redhat.com,
	sebastian.wick@redhat.com, shashank.sharma@amd.com,
	agoins@nvidia.com, joshua@froggi.es, mdaenzer@redhat.com,
	aleixpol@kde.org, xaver.hugl@gmail.com, victoria@system76.com,
	daniel@ffwll.ch, Shankar, Uma, quic_naseer@quicinc.com,
	quic_cbraga@quicinc.com, quic_abhinavk@quicinc.com,
	marcan@marcan.st, Liviu.Dudau@arm.com, sashamcintosh@google.com,
	louis.chauvet@bootlin.com, Daniel Stone

On Tue, Jun 17, 2025 at 05:33:20AM +0000, Borah, Chaitanya Kumar wrote:
> 
> 
> > -----Original Message-----
> > From: Pekka Paalanen <pekka.paalanen@collabora.com>
> > Sent: Monday, June 16, 2025 7:02 PM
> > To: Borah, Chaitanya Kumar <chaitanya.kumar.borah@intel.com>
> > Cc: Alex Hung <alex.hung@amd.com>; dri-devel@lists.freedesktop.org; amd-
> > gfx@lists.freedesktop.org; wayland-devel@lists.freedesktop.org;
> > harry.wentland@amd.com; leo.liu@amd.com; ville.syrjala@linux.intel.com;
> > mwen@igalia.com; jadahl@redhat.com; sebastian.wick@redhat.com;
> > shashank.sharma@amd.com; agoins@nvidia.com; joshua@froggi.es;
> > mdaenzer@redhat.com; aleixpol@kde.org; xaver.hugl@gmail.com;
> > victoria@system76.com; daniel@ffwll.ch; Shankar, Uma
> > <uma.shankar@intel.com>; quic_naseer@quicinc.com;
> > quic_cbraga@quicinc.com; quic_abhinavk@quicinc.com; marcan@marcan.st;
> > Liviu.Dudau@arm.com; sashamcintosh@google.com;
> > louis.chauvet@bootlin.com; Daniel Stone <daniels@collabora.com>
> > Subject: Re: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
> > 
> > On Mon, 16 Jun 2025 11:30:23 +0000
> > "Borah, Chaitanya Kumar" <chaitanya.kumar.borah@intel.com> wrote:
> > 
> > > > -----Original Message-----
> > > > From: Alex Hung <alex.hung@amd.com>
> > > > Sent: Wednesday, April 30, 2025 6:41 AM
> > > > To: dri-devel@lists.freedesktop.org; amd-gfx@lists.freedesktop.org
> > > > Cc: wayland-devel@lists.freedesktop.org; harry.wentland@amd.com;
> > > > alex.hung@amd.com; leo.liu@amd.com; ville.syrjala@linux.intel.com;
> > > > pekka.paalanen@collabora.com; contact@emersion.fr; mwen@igalia.com;
> > > > jadahl@redhat.com; sebastian.wick@redhat.com;
> > > > shashank.sharma@amd.com; agoins@nvidia.com; joshua@froggi.es;
> > > > mdaenzer@redhat.com; aleixpol@kde.org; xaver.hugl@gmail.com;
> > > > victoria@system76.com; daniel@ffwll.ch; Shankar, Uma
> > > > <uma.shankar@intel.com>; quic_naseer@quicinc.com;
> > > > quic_cbraga@quicinc.com; quic_abhinavk@quicinc.com;
> > > > marcan@marcan.st; Liviu.Dudau@arm.com; sashamcintosh@google.com;
> > > > Borah, Chaitanya Kumar <chaitanya.kumar.borah@intel.com>;
> > > > louis.chauvet@bootlin.com; Daniel Stone <daniels@collabora.com>
> > > > Subject: [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type
> > > >
> > > > From: Harry Wentland <harry.wentland@amd.com>
> > > >
> > > > This type is used to support a 3x4 matrix in colorops. A 3x4 matrix
> > > > uses the last column as a "bias" column. Some HW exposes support for
> > > > 3x4. The calculation looks like:
> > > >
> > > >  out   matrix    in
> > > >  |R|   |0  1  2  3 |   | R |
> > > >  |G| = |4  5  6  7 | x | G |
> > > >  |B|   |8  9  10 11|   | B |
> > > >                        |1.0|
> > > >
> > > > This is also the first colorop where we need a blob property to
> > > > program the property. For that we'll introduce a new DATA property
> > > > that can be used by all colorop TYPEs requiring a blob. The way a
> > > > DATA blob is read depends on the TYPE of the colorop.
> > > >
> > > > We only create the DATA property for property types that need it.
> > >
> > > Is there any value to adding pre-offsets [1] in the uapi?
> > >
> > >  |R/Cr|    | c0 c1 c2 |   ( |R/Cr|   |preoff0| )   |postoff0|
> > >  |G/Y | = | c3 c4 c5 | x ( |G/Y | + |preoff1| ) + |postoff1|
> > >  |B/Cb|   | c6 c7 c8 |   ( |B/Cb|   |preoff2| )   |postoff2|
> > >
> > > Handling limited range values is one use case that I can think of.
> > 
> > Hi,
> > 
> > in the mathematical sense, no. A pre-offset can always be converted into a
> > post-offset by multiplying it with the 3x3 matrix (and adding to the existing
> > post-offset). This can be pre-computed, no need to do it separately for every
> > pixel.
> > 
> > For hardware reasons, I have no idea.
> 
> Thank you for the reply, Pekka. Our hardware does allow programming Pre-offsets separately.
> Currently I can't think of a particular advantage of that if mathematically a post-offset does the job but I will keep this thread posted if I find something out.

Just FYI we have three different kinds of hardware:
- 3x3 matrix
- 3x3 matrix + programmable pre offsets + hardcoded post offsets for limited range conversion
- 3x3 matrix + programmable pre and post offsets

-- 
Ville Syrjälä
Intel

^ permalink raw reply	[flat|nested] 82+ messages in thread

end of thread, other threads:[~2025-06-17 15:25 UTC | newest]

Thread overview: 82+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-04-30  1:10 [PATCH V9 00/43] Color Pipeline API w/ VKMS Alex Hung
2025-04-30  1:10 ` [PATCH V9 01/43] drm: Add helper for conversion from signed-magnitude Alex Hung
2025-04-30  1:10 ` [PATCH V9 02/43] drm/vkms: Add kunit tests for VKMS LUT handling Alex Hung
2025-04-30  1:10 ` [PATCH V9 03/43] drm/doc/rfc: Describe why prescriptive color pipeline is needed Alex Hung
2025-05-15  9:30   ` Simon Ser
2025-04-30  1:10 ` [PATCH V9 04/43] drm/colorop: Introduce new drm_colorop mode object Alex Hung
2025-04-30  1:10 ` [PATCH V9 05/43] drm/colorop: Add TYPE property Alex Hung
2025-04-30  1:10 ` [PATCH V9 06/43] drm/colorop: Add 1D Curve subtype Alex Hung
2025-04-30  1:10 ` [PATCH V9 07/43] drm/colorop: Add BYPASS property Alex Hung
2025-04-30  1:10 ` [PATCH V9 08/43] drm/colorop: Add NEXT property Alex Hung
2025-04-30  1:10 ` [PATCH V9 09/43] drm/colorop: Add atomic state print for drm_colorop Alex Hung
2025-04-30  1:10 ` [PATCH V9 10/43] drm/plane: Add COLOR PIPELINE property Alex Hung
2025-04-30  1:10 ` [PATCH V9 11/43] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Alex Hung
2025-04-30  1:10 ` [PATCH V9 12/43] Documentation/gpu: document drm_colorop Alex Hung
2025-04-30  1:10 ` [PATCH V9 13/43] drm/colorop: Add destroy functions for color pipeline Alex Hung
2025-05-15  9:34   ` Simon Ser
2025-04-30  1:10 ` [PATCH V9 14/43] drm/vkms: Add enumerated 1D curve colorop Alex Hung
2025-04-30  1:10 ` [PATCH V9 15/43] drm/vkms: Add kunit tests for linear and sRGB LUTs Alex Hung
2025-05-07 10:03   ` kernel test robot
2025-04-30  1:10 ` [PATCH V9 16/43] drm/colorop: Add 3x4 CTM type Alex Hung
2025-06-16 11:30   ` Borah, Chaitanya Kumar
2025-06-16 13:32     ` Pekka Paalanen
2025-06-17  5:33       ` Borah, Chaitanya Kumar
2025-06-17 15:25         ` Ville Syrjälä
2025-04-30  1:10 ` [PATCH V9 17/43] drm/vkms: Use s32 for internal color pipeline precision Alex Hung
2025-04-30  1:10 ` [PATCH V9 18/43] drm/vkms: add 3x4 matrix in color pipeline Alex Hung
2025-04-30  1:10 ` [PATCH V9 19/43] drm/tests: Add a few tests around drm_fixed.h Alex Hung
2025-04-30  1:10 ` [PATCH V9 20/43] drm/vkms: Add tests for CTM handling Alex Hung
2025-04-30  1:10 ` [PATCH V9 21/43] drm/colorop: pass plane_color_pipeline client cap to atomic check Alex Hung
2025-04-30  1:10 ` [PATCH V9 22/43] drm/colorop: define a new macro for_each_new_colorop_in_state Alex Hung
2025-05-12 18:32   ` Melissa Wen
2025-04-30  1:10 ` [PATCH V9 23/43] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Alex Hung
2025-04-30  1:10 ` [PATCH V9 24/43] drm/amd/display: Add bypass COLOR PIPELINE Alex Hung
2025-05-12 18:41   ` Melissa Wen
2025-04-30  1:10 ` [PATCH V9 25/43] drm/amd/display: Skip color pipeline initialization for cursor plane Alex Hung
2025-04-30  1:10 ` [PATCH V9 26/43] drm/amd/display: Add support for sRGB EOTF in DEGAM block Alex Hung
2025-05-12 18:50   ` Melissa Wen
2025-05-13  7:42     ` Pekka Paalanen
2025-05-13 15:36       ` Melissa Wen
2025-05-13 20:39         ` Harry Wentland
2025-05-15 10:57           ` Pekka Paalanen
2025-04-30  1:10 ` [PATCH V9 27/43] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Alex Hung
2025-04-30  1:10 ` [PATCH V9 28/43] drm/amd/display: Add support for sRGB EOTF in BLND block Alex Hung
2025-04-30  1:10 ` [PATCH V9 29/43] drm/colorop: Add PQ 125 EOTF and its inverse Alex Hung
2025-04-30  1:11 ` [PATCH V9 30/43] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Alex Hung
2025-04-30  1:11 ` [PATCH V9 31/43] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Alex Hung
2025-05-15  9:35   ` Simon Ser
2025-04-30  1:11 ` [PATCH V9 32/43] drm/amd/display: Add support for BT.709 and BT.2020 TFs Alex Hung
2025-05-12 19:37   ` Melissa Wen
2025-04-30  1:11 ` [PATCH V9 33/43] drm/colorop: Add 1D Curve Custom LUT type Alex Hung
2025-04-30  1:11 ` [PATCH V9 34/43] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Alex Hung
2025-04-30  1:11 ` [PATCH V9 35/43] drm/amd/display: add 3x4 matrix colorop Alex Hung
2025-05-07 12:08   ` kernel test robot
2025-04-30  1:11 ` [PATCH V9 36/43] drm/colorop: Add mutliplier type Alex Hung
2025-05-12 20:02   ` Melissa Wen
2025-04-30  1:11 ` [PATCH V9 37/43] drm/amd/display: add multiplier colorop Alex Hung
2025-04-30  1:11 ` [PATCH V9 38/43] drm/amd/display: Swap matrix and multiplier Alex Hung
2025-04-30  1:11 ` [PATCH V9 39/43] drm/colorop: Define LUT_1D interpolation Alex Hung
2025-04-30  1:11 ` [PATCH V9 40/43] drm/colorop: allow non-bypass colorops Alex Hung
2025-05-15  9:37   ` Simon Ser
2025-04-30  1:11 ` [PATCH V9 41/43] drm/colorop: Add 3D LUT support to color pipeline Alex Hung
2025-04-30  1:11 ` [PATCH V9 42/43] drm/amd/display: add 3D LUT colorop Alex Hung
2025-05-13  0:52   ` Melissa Wen
2025-05-13  3:29     ` Alex Hung
2025-05-13 15:52       ` Melissa Wen
2025-04-30  1:11 ` [PATCH V9 43/43] drm/amd/display: Add AMD color pipeline doc Alex Hung
2025-05-13 15:59 ` [PATCH V9 00/43] Color Pipeline API w/ VKMS Melissa Wen
2025-05-15  9:45 ` Simon Ser
2025-05-15 14:10   ` Harry Wentland
2025-05-15 14:15     ` Simon Ser
2025-05-15 17:19     ` Daniel Stone
2025-05-15 18:02       ` Harry Wentland
2025-05-15 18:39         ` Daniel Stone
2025-05-15 19:59           ` Leandro Ribeiro
2025-05-17 11:51             ` Xaver Hugl
2025-05-21 19:48               ` Harry Wentland
2025-05-22  7:57                 ` Pekka Paalanen
2025-05-22 13:28                   ` Harry Wentland
2025-05-22 13:49                     ` Simon Ser
2025-05-22 13:54                       ` Harry Wentland
2025-05-22 15:27                         ` Pekka Paalanen
2025-05-22 15:49                           ` Harry Wentland

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).