dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
* [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS
@ 2024-02-26 21:10 Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil Harry Wentland
                   ` (44 more replies)
  0 siblings, 45 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, Harry Wentland, Ville Syrjala, Pekka Paalanen,
	Simon Ser, Melissa Wen, Jonas Ådahl, Sebastian Wick,
	Shashank Sharma, Alexander Goins, Joshua Ashton,
	Michel Dänzer, Aleix Pol, Xaver Hugl, Victoria Brekenfeld,
	Sima, Uma Shankar, Naseer Ahmed, Christopher Braga,
	Abhinav Kumar, Arthur Grillo, Hector Martin, Liviu Dudau,
	Sasha McIntosh, Chaitanya Kumar Borah

This is an RFC set for a color pipeline API, along with a sample
implementation in VKMS. All the key API bits are here. VKMS now
supports two named transfer function colorops and two matrix
colorops. We have IGT tests that check all four of these colorops
with a pixel-by-pixel comparison that checks that these colorops
do what we expect them to do with a +/- 1 8 bpc code point margin.

The big new change with v4 is the addition of an amdgpu color
pipeline, for all AMD GPUs with DCN 3 and newer. Amdgpu now support
the following:

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

We are working on two more ops for amdgpu, the HDR multiplier
and the 3DLUT, which will give us this:

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

This, essentially 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

After this we'd like to also add the following ops:
- Scaler (Informational only)
- Color Encoding, to replace drm_plane's COLOR_ENCODING
- Color Range, to replace drm_plane's COLOR_RANGE

This patchset is grouped as follows:
 - Patches 1-3: couple general patches/fixes
 - Patches 4-7: introduce kunit to VKMS
 - Patch 7: description of motivation and details behind the
            Color Pipeline API. If you're reading nothing else
            but are interested in the topic I highly recommend
            you take a look at this.
 - Patches 7-27: DRM core and VKMS changes for color pipeline API
 - Patches 28-40: DRM core and amdgpu changes for color pipeline API

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 here but
haven't had a chance to look at. These will (hopefully) be
addressed in future iterations, either in VKMS or amdgpu:
 - Clear documentation for each drm_colorop_type
 - Add custom LUT colorops to VKMS
 - Add pre-blending 3DLUT
 - How to support HW which can't bypass entire pipeline?
 - Add ability to create colorops that don't have BYPASS
 - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
 - read-only scaling colorop which defines scaling taps and position
 - read-only color format colorop to define supported color formats
   for a pipeline
 - named matrices, for things like converting YUV to RGB

IGT tests can be found at
https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/merge_requests/1

IGT patches are also being sent to the igt-dev mailing list.

If you prefer a gitlab MR for review you can find it at
https://gitlab.freedesktop.org/hwentland/linux/-/merge_requests/5

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 (10):
  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

Harry Wentland (32):
  drm: Don't treat 0 as -1 in drm_fixp2int_ceil
  drm: Add helper for conversion from signed-magnitude
  drm: Correctly round for fixp2int_round
  drm/vkms: Round fixp2int conversion in lerp_u16
  drm/vkms: Create separate Kconfig file for VKMS
  drm/vkms: Add kunit tests for VKMS LUT handling
  drm/vkms: Avoid reading beyond LUT array
  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: Add NEXT to colorop state print
  drm/vkms: Add enumerated 1D curve colorop
  drm/vkms: Add kunit tests for linear and sRGB LUTs
  drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
  drm/colorop: Add 3x4 CTM type
  drm/vkms: Pull apply_colorop out of pre_blend_color_transform
  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

 Documentation/gpu/rfc/color_pipeline.rst      | 360 ++++++++
 drivers/gpu/drm/Kconfig                       |  14 +-
 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   | 352 ++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 160 ++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
 .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
 drivers/gpu/drm/drm_atomic.c                  | 161 +++-
 drivers/gpu/drm/drm_atomic_helper.c           |  12 +
 drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
 drivers/gpu/drm/drm_atomic_uapi.c             | 158 ++++
 drivers/gpu/drm/drm_colorop.c                 | 411 +++++++++
 drivers/gpu/drm/drm_ioctl.c                   |   7 +
 drivers/gpu/drm/drm_mode_config.c             |   7 +
 drivers/gpu/drm/drm_plane.c                   |  52 ++
 drivers/gpu/drm/tests/Makefile                |   3 +-
 drivers/gpu/drm/tests/drm_fixp_test.c         |  69 ++
 drivers/gpu/drm/vkms/Kconfig                  |  20 +
 drivers/gpu/drm/vkms/Makefile                 |   4 +-
 drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
 drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 449 ++++++++++
 drivers/gpu/drm/vkms/vkms_colorop.c           | 100 +++
 drivers/gpu/drm/vkms/vkms_composer.c          | 135 ++-
 drivers/gpu/drm/vkms/vkms_drv.h               |   8 +
 drivers/gpu/drm/vkms/vkms_luts.c              | 802 ++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
 drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
 include/drm/drm_atomic.h                      | 122 +++
 include/drm/drm_atomic_uapi.h                 |   3 +
 include/drm/drm_colorop.h                     | 301 +++++++
 include/drm/drm_file.h                        |   7 +
 include/drm/drm_fixed.h                       |  35 +-
 include/drm/drm_mode_config.h                 |  18 +
 include/drm/drm_plane.h                       |  13 +
 include/uapi/drm/drm.h                        |  16 +
 include/uapi/drm/drm_mode.h                   |  14 +
 38 files changed, 3882 insertions(+), 30 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/Kconfig
 create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
 create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
 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
 create mode 100644 include/drm/drm_colorop.h

--
2.44.0


^ permalink raw reply	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-14 13:41   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 02/42] drm: Add helper for conversion from signed-magnitude Harry Wentland
                   ` (43 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland, Simon Ser, Melissa Wen

Unit testing this in VKMS shows that passing 0 into
this function returns -1, which is highly counter-
intuitive. Fix it by checking whether the input is
>= 0 instead of > 0.

Fixes: 64566b5e767f9 ("drm: Add drm_fixp_from_fraction and drm_fixp2int_ceil")
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Simon Ser <contact@emersion.fr>
Reviewed-by: Melissa Wen <mwen@igalia.com>
---
 include/drm/drm_fixed.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/drm/drm_fixed.h b/include/drm/drm_fixed.h
index 6ea339d5de08..0c9f917a4d4b 100644
--- a/include/drm/drm_fixed.h
+++ b/include/drm/drm_fixed.h
@@ -95,7 +95,7 @@ static inline int drm_fixp2int_round(s64 a)
 
 static inline int drm_fixp2int_ceil(s64 a)
 {
-	if (a > 0)
+	if (a >= 0)
 		return drm_fixp2int(a + DRM_FIXED_ALMOST_ONE);
 	else
 		return drm_fixp2int(a - DRM_FIXED_ALMOST_ONE);
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 02/42] drm: Add helper for conversion from signed-magnitude
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-14 13:16   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 03/42] drm: Correctly round for fixp2int_round Harry Wentland
                   ` (42 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

Signed-off-by: Harry Wentland <harry.wentland@amd.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 0c9f917a4d4b..cb842ba80ddd 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;
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 03/42] drm: Correctly round for fixp2int_round
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 02/42] drm: Add helper for conversion from signed-magnitude Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-11 13:11   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 04/42] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
                   ` (41 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

A value of 0x80000000 and higher should round up, and
below should round down. VKMS Kunit tests for lerp_u16
showed that this is not the case. Fix it.

1 << (DRM_FIXED_POINT_HALF - 1) =
1 << 15 = 0x8000

This is not 0.5, but 0.00000762939453125.

Instead of some smart math use a simple if/else to
round up or down. This helps people like me to understand
what the function does.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 include/drm/drm_fixed.h | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/include/drm/drm_fixed.h b/include/drm/drm_fixed.h
index cb842ba80ddd..8ee549f68537 100644
--- a/include/drm/drm_fixed.h
+++ b/include/drm/drm_fixed.h
@@ -77,6 +77,8 @@ static inline u32 dfixed_div(fixed20_12 A, fixed20_12 B)
 #define DRM_FIXED_DIGITS_MASK	(~DRM_FIXED_DECIMAL_MASK)
 #define DRM_FIXED_EPSILON	1LL
 #define DRM_FIXED_ALMOST_ONE	(DRM_FIXED_ONE - DRM_FIXED_EPSILON)
+#define DRM_FIXED_FRACTIONAL	0xffffffffll
+#define DRM_FIXED_HALF		0x80000000ll
 
 /**
  * @drm_sm2fixp
@@ -106,11 +108,6 @@ static inline int drm_fixp2int(s64 a)
 	return ((s64)a) >> DRM_FIXED_POINT;
 }
 
-static inline int drm_fixp2int_round(s64 a)
-{
-	return drm_fixp2int(a + (1 << (DRM_FIXED_POINT_HALF - 1)));
-}
-
 static inline int drm_fixp2int_ceil(s64 a)
 {
 	if (a >= 0)
@@ -119,6 +116,14 @@ static inline int drm_fixp2int_ceil(s64 a)
 		return drm_fixp2int(a - DRM_FIXED_ALMOST_ONE);
 }
 
+static inline int drm_fixp2int_round(s64 a)
+{
+	if ((a & DRM_FIXED_FRACTIONAL) < DRM_FIXED_HALF)
+		return drm_fixp2int(a);
+	else
+		return drm_fixp2int_ceil(a);
+}
+
 static inline unsigned drm_fixp_msbset(s64 a)
 {
 	unsigned shift, sign = (a >> 63) & 1;
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 04/42] drm/vkms: Round fixp2int conversion in lerp_u16
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (2 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 03/42] drm: Correctly round for fixp2int_round Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 05/42] drm/vkms: Create separate Kconfig file for VKMS Harry Wentland
                   ` (40 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

fixp2int always rounds down, fixp2int_ceil rounds up. We need
the new fixp2int_round.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/vkms/vkms_composer.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 3c99fb8b54e2..e70cd473e3be 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -98,7 +98,7 @@ static u16 lerp_u16(u16 a, u16 b, s64 t)
 
 	s64 delta = drm_fixp_mul(b_fp - a_fp,  t);
 
-	return drm_fixp2int(a_fp + delta);
+	return drm_fixp2int_round(a_fp + delta);
 }
 
 static s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 05/42] drm/vkms: Create separate Kconfig file for VKMS
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (3 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 04/42] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-12 18:49   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
                   ` (39 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland, Simon Ser

This aligns with most other DRM drivers and will allow
us to add new VKMS config options without polluting
the DRM Kconfig.

v3:
 - Change SPDX to GPL-2.0-only to match DRM KConfig
   SPDX (Simon)

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Simon Ser <contact@emersion.fr>
---
 drivers/gpu/drm/Kconfig      | 14 +-------------
 drivers/gpu/drm/vkms/Kconfig | 15 +++++++++++++++
 2 files changed, 16 insertions(+), 13 deletions(-)
 create mode 100644 drivers/gpu/drm/vkms/Kconfig

diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index 2520db0b776e..e3ea8793cb8a 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -289,19 +289,7 @@ config DRM_VGEM
 	  as used by Mesa's software renderer for enhanced performance.
 	  If M is selected the module will be called vgem.
 
-config DRM_VKMS
-	tristate "Virtual KMS (EXPERIMENTAL)"
-	depends on DRM && MMU
-	select DRM_KMS_HELPER
-	select DRM_GEM_SHMEM_HELPER
-	select CRC32
-	default n
-	help
-	  Virtual Kernel Mode-Setting (VKMS) is used for testing or for
-	  running GPU in a headless machines. Choose this option to get
-	  a VKMS.
-
-	  If M is selected the module will be called vkms.
+source "drivers/gpu/drm/vkms/Kconfig"
 
 source "drivers/gpu/drm/exynos/Kconfig"
 
diff --git a/drivers/gpu/drm/vkms/Kconfig b/drivers/gpu/drm/vkms/Kconfig
new file mode 100644
index 000000000000..b9ecdebecb0b
--- /dev/null
+++ b/drivers/gpu/drm/vkms/Kconfig
@@ -0,0 +1,15 @@
+# SPDX-License-Identifier: GPL-2.0-only
+
+config DRM_VKMS
+	tristate "Virtual KMS (EXPERIMENTAL)"
+	depends on DRM && MMU
+	select DRM_KMS_HELPER
+	select DRM_GEM_SHMEM_HELPER
+	select CRC32
+	default n
+	help
+	  Virtual Kernel Mode-Setting (VKMS) is used for testing or for
+	  running GPU in a headless machines. Choose this option to get
+	  a VKMS.
+
+	  If M is selected the module will be called vkms.
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (4 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 05/42] drm/vkms: Create separate Kconfig file for VKMS Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-27 12:14   ` Arthur Grillo
  2024-03-11 13:45   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 07/42] drm/vkms: Avoid reading beyond LUT array Harry Wentland
                   ` (38 subsequent siblings)
  44 siblings, 2 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland, Arthur Grillo

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

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

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Cc: Arthur Grillo <arthurgrillo@riseup.net>
---
 drivers/gpu/drm/vkms/Kconfig                  |   5 +
 drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
 drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 163 ++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_composer.c          |   8 +-
 4 files changed, 178 insertions(+), 2 deletions(-)
 create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
 create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c

diff --git a/drivers/gpu/drm/vkms/Kconfig b/drivers/gpu/drm/vkms/Kconfig
index b9ecdebecb0b..c1f8b343ff0e 100644
--- a/drivers/gpu/drm/vkms/Kconfig
+++ b/drivers/gpu/drm/vkms/Kconfig
@@ -13,3 +13,8 @@ config DRM_VKMS
 	  a VKMS.
 
 	  If M is selected the module will be called vkms.
+
+config DRM_VKMS_KUNIT_TESTS
+	tristate "Tests for VKMS" if !KUNIT_ALL_TESTS
+	depends on DRM_VKMS && KUNIT
+	default KUNIT_ALL_TESTS
diff --git a/drivers/gpu/drm/vkms/tests/.kunitconfig b/drivers/gpu/drm/vkms/tests/.kunitconfig
new file mode 100644
index 000000000000..70e378228cbd
--- /dev/null
+++ b/drivers/gpu/drm/vkms/tests/.kunitconfig
@@ -0,0 +1,4 @@
+CONFIG_KUNIT=y
+CONFIG_DRM=y
+CONFIG_DRM_VKMS=y
+CONFIG_DRM_VKMS_KUNIT_TESTS=y
diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
new file mode 100644
index 000000000000..fc73e48aa57c
--- /dev/null
+++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
@@ -0,0 +1,163 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#include <kunit/test.h>
+
+#include <drm/drm_fixed.h>
+
+#define TEST_LUT_SIZE 16
+
+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 },
+};
+
+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)
+{
+	int i;
+
+	KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&test_linear_lut, test_linear_array[0].red)), 0);
+
+	for (i = 0; i < TEST_LUT_SIZE; i++)
+		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&test_linear_lut, test_linear_array[i].red)), 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_LICENSE("GPL");
\ No newline at end of file
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index e70cd473e3be..d178f2a400f6 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -91,7 +91,7 @@ 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)
+u16 lerp_u16(u16 a, u16 b, s64 t)
 {
 	s64 a_fp = drm_int2fixp(a);
 	s64 b_fp = drm_int2fixp(b);
@@ -101,7 +101,7 @@ static u16 lerp_u16(u16 a, u16 b, s64 t)
 	return drm_fixp2int_round(a_fp + delta);
 }
 
-static s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
+s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
 {
 	s64 color_channel_fp = drm_int2fixp(channel_value);
 
@@ -429,3 +429,7 @@ int vkms_set_crc_source(struct drm_crtc *crtc, const char *src_name)
 
 	return ret;
 }
+
+#ifdef CONFIG_DRM_VKMS_KUNIT_TESTS
+#include "tests/vkms_color_tests.c"
+#endif
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 07/42] drm/vkms: Avoid reading beyond LUT array
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (5 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-12 18:54   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 08/42] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
                   ` (37 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx
  Cc: wayland-devel, Harry Wentland, Arthur Grillo, Melissa Wen

When the floor LUT index (drm_fixp2int(lut_index) is the last
index of the array the ceil LUT index will point to an entry
beyond the array. Make sure we guard against it and use the
value of the floor LUT index.

v3:
 - Drop bits from commit description that didn't contribute
   anything of value

Fixes: db1f254f2cfaf ("drm/vkms: Add support to 1D gamma LUT")
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Cc: Arthur Grillo <arthurgrillo@riseup.net>
Reviewed-by: Melissa Wen <mwen@igalia.com>
---
 drivers/gpu/drm/vkms/vkms_composer.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index d178f2a400f6..b90e446d5954 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -123,6 +123,8 @@ static u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 chan
 				      enum lut_channel channel)
 {
 	s64 lut_index = get_lut_index(lut, channel_value);
+	u16 *floor_lut_value, *ceil_lut_value;
+	u16 floor_channel_value, ceil_channel_value;
 
 	/*
 	 * This checks if `struct drm_color_lut` has any gap added by the compiler
@@ -130,11 +132,15 @@ static u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 chan
 	 */
 	static_assert(sizeof(struct drm_color_lut) == sizeof(__u16) * 4);
 
-	u16 *floor_lut_value = (__u16 *)&lut->base[drm_fixp2int(lut_index)];
-	u16 *ceil_lut_value = (__u16 *)&lut->base[drm_fixp2int_ceil(lut_index)];
+	floor_lut_value = (__u16 *)&lut->base[drm_fixp2int(lut_index)];
+	if (drm_fixp2int(lut_index) == (lut->lut_length - 1))
+		/* We're at the end of the LUT array, use same value for ceil and floor */
+		ceil_lut_value = floor_lut_value;
+	else
+		ceil_lut_value = (__u16 *)&lut->base[drm_fixp2int_ceil(lut_index)];
 
-	u16 floor_channel_value = floor_lut_value[channel];
-	u16 ceil_channel_value = ceil_lut_value[channel];
+	floor_channel_value = floor_lut_value[channel];
+	ceil_channel_value = ceil_lut_value[channel];
 
 	return lerp_u16(floor_channel_value, ceil_channel_value,
 			lut_index & DRM_FIXED_DECIMAL_MASK);
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 08/42] drm/doc/rfc: Describe why prescriptive color pipeline is needed
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (6 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 07/42] drm/vkms: Avoid reading beyond LUT array Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-11 15:11   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 09/42] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
                   ` (36 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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)

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 Documentation/gpu/rfc/color_pipeline.rst | 360 +++++++++++++++++++++++
 1 file changed, 360 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..6c653e17054a
--- /dev/null
+++ b/Documentation/gpu/rfc/color_pipeline.rst
@@ -0,0 +1,360 @@
+========================
+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.
+
+
+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, definint 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. While setting other properties might fail atomic check,
+setting the BYPASS property to true should never fail. 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}
+    ├─ "LUT_1D_SIZE": immutable range = 4096
+    ├─ "LUT_1D": 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}
+    ├─ "LUT_3D_SIZE": immutable range = 17
+    ├─ "LUT_3D": blob
+    └─ "NEXT": immutable color operation ID = 73
+
+
+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
+    ├─ "MATRIX_3_4": 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
+    ├─ "LUT_1D_SIZE": immutable range = 4096
+    ├─ "LUT_1D_DATA": blob
+    └─ "NEXT" (immutable) = 48
+
+    Color operation 48
+    ├─ "TYPE" (immutable) = 3D LUT
+    ├─ "LUT_3D_SIZE" (immutable) = 17
+    ├─ "LUT_3D_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
+    ├─ "LUT_3D_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
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 09/42] drm/colorop: Introduce new drm_colorop mode object
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (7 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 08/42] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-05-21 15:28   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 10/42] drm/colorop: Add TYPE property Harry Wentland
                   ` (35 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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)

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 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       | 152 ++++++++++++++++++++++++++
 drivers/gpu/drm/drm_mode_config.c   |   7 ++
 include/drm/drm_atomic.h            |  82 +++++++++++++++
 include/drm/drm_atomic_uapi.h       |   1 +
 include/drm/drm_colorop.h           | 158 ++++++++++++++++++++++++++++
 include/drm/drm_mode_config.h       |  18 ++++
 include/drm/drm_plane.h             |   2 +
 include/uapi/drm/drm_mode.h         |   1 +
 12 files changed, 552 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 104b42df2e95..4b14dcbb6117 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -16,6 +16,7 @@ drm-y := \
 	drm_client.o \
 	drm_client_modeset.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 a91737adf8e7..62e87e6a9653 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"
@@ -108,6 +109,7 @@ 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);
@@ -139,6 +141,10 @@ 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
@@ -250,6 +256,20 @@ 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;
 
@@ -571,6 +591,56 @@ 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_existing_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 39ef0a6addeb..282aa1f7b9a6 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -3022,6 +3022,8 @@ 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;
@@ -3099,6 +3101,16 @@ 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;
+	}
+
 	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 29d4940188d4..1f9b6dfa8ca7 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -34,6 +34,7 @@
 #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>
@@ -642,6 +643,26 @@ 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)
@@ -904,6 +925,16 @@ 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;
@@ -1084,6 +1115,23 @@ 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..a295ab96aee1
--- /dev/null
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -0,0 +1,152 @@
+/*
+ * 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"
+
+/* TODO big colorop doc, including properties, etc. */
+
+/* Init Helpers */
+
+int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
+		     struct drm_plane *plane)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	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->plane = plane;
+
+	list_add_tail(&colorop->head, &config->colorop_list);
+	colorop->index = config->num_colorop++;
+
+	/* add properties */
+	return ret;
+}
+EXPORT_SYMBOL(drm_colorop_init);
+
+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_destroy_state - default state destroy hook
+ * @colorop: drm colorop
+ * @state: colorop state object to release
+ *
+ * Default colorop state destroy hook for drivers which don't have their own
+ * subclassed colorop state structure.
+ */
+void drm_colorop_destroy_state(struct drm_colorop *colorop,
+			       struct drm_colorop_state *state)
+{
+	kfree(state);
+}
+EXPORT_SYMBOL(drm_colorop_destroy_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;
+}
+
+/**
+ * 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)
+{
+	kfree(colorop->state);
+	colorop->state = kzalloc(sizeof(*colorop->state), GFP_KERNEL);
+
+	if (colorop->state)
+		__drm_colorop_reset(colorop, colorop->state);
+}
+EXPORT_SYMBOL(drm_colorop_reset);
diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
index 8525ef851540..30c6fb10353b 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"
@@ -182,11 +183,15 @@ 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);
@@ -413,6 +418,7 @@ 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);
@@ -434,6 +440,7 @@ 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 cf8e1220a4ac..634b2827765f 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;
@@ -398,6 +404,7 @@ struct drm_atomic_state {
 	 * states.
 	 */
 	bool duplicated : 1;
+	struct __drm_colorops_state *colorops;
 	struct __drm_planes_state *planes;
 	struct __drm_crtcs_state *crtcs;
 	int num_connector;
@@ -501,6 +508,9 @@ 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);
@@ -630,6 +640,55 @@ drm_atomic_get_new_plane_state(const struct drm_atomic_state *state,
 	return state->planes[drm_plane_index(plane)].new_state;
 }
 
+
+/**
+ * drm_atomic_get_existing_colorop_state - get colorop state, if it exists
+ * @state: global atomic state object
+ * @colorop: colorop to grab
+ *
+ * This function returns the colorop state for the given colorop, or NULL
+ * if the colorop is not part of the global atomic state.
+ *
+ * This function is deprecated, @drm_atomic_get_old_colorop_state or
+ * @drm_atomic_get_new_colorop_state should be used instead.
+ */
+static inline struct drm_colorop_state *
+drm_atomic_get_existing_colorop_state(struct drm_atomic_state *state,
+				    struct drm_colorop *colorop)
+{
+	return state->colorops[drm_colorop_index(colorop)].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
@@ -877,6 +936,29 @@ 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..e611f830f986
--- /dev/null
+++ b/include/drm/drm_colorop.h
@@ -0,0 +1,158 @@
+/*
+ * 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;
+
+	/* colorop properties */
+
+	/** @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.
+ */
+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 plane.
+	 */
+	unsigned index;
+
+	/* TODO do we need a separate mutex or will we tag along with the plane mutex? */
+
+	/** @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. Either by
+	 * going through the &struct drm_atomic_state pointers, see
+	 * for_each_oldnew_plane_in_state(), for_each_old_plane_in_state() and
+	 * for_each_new_plane_in_state(). Or through careful ordering of atomic
+	 * commit operations as implemented in the atomic helpers, see
+	 * &struct drm_crtc_commit.
+	 *
+	 * TODO keep, remove, or rewrite above plane references?
+	 */
+	struct drm_colorop_state *state;
+
+	/* colorop properties */
+
+	/** @properties: property tracking for this plane */
+	struct drm_object_properties properties;
+
+};
+
+#define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
+
+/**
+ * drm_crtc_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;
+}
+
+int drm_colorop_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);
+
+void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
+				      struct drm_colorop_state *state);
+void drm_colorop_destroy_state(struct drm_colorop *colorop,
+			       struct drm_colorop_state *state);
+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 973119a9176b..492b8c120c80 100644
--- a/include/drm/drm_mode_config.h
+++ b/include/drm/drm_mode_config.h
@@ -505,6 +505,24 @@ struct drm_mode_config {
 	 */
 	struct list_head plane_list;
 
+	/**
+	 * @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;
+
+	/**
+	 * @colorops_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 641fe298052d..9c8c1322f0a6 100644
--- a/include/drm/drm_plane.h
+++ b/include/drm/drm_plane.h
@@ -241,6 +241,8 @@ struct drm_plane_state {
 	 */
 	enum drm_scaling_filter scaling_filter;
 
+	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 7040e7ea80c7..9e8f383935db 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 {
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 10/42] drm/colorop: Add TYPE property
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (8 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 09/42] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-12 15:02   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 11/42] drm/colorop: Add 1D Curve subtype Harry Wentland
                   ` (34 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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

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

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic.c      |  4 +--
 drivers/gpu/drm/drm_atomic_uapi.c |  8 +++++-
 drivers/gpu/drm/drm_colorop.c     | 44 ++++++++++++++++++++++++++++++-
 include/drm/drm_colorop.h         | 17 +++++++++++-
 include/uapi/drm/drm_mode.h       |  4 +++
 5 files changed, 72 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 62e87e6a9653..b400e32c9d39 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -634,8 +634,8 @@ 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 1f9b6dfa8ca7..e3067c095c72 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -660,7 +660,13 @@ 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 a295ab96aee1..3a07a8fe2842 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -32,12 +32,17 @@
 
 /* TODO big colorop doc, including properties, etc. */
 
+static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
+	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
+};
+
 /* Init Helpers */
 
 int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
-		     struct drm_plane *plane)
+		     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);
@@ -46,12 +51,29 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 
 	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;
 }
 EXPORT_SYMBOL(drm_colorop_init);
@@ -150,3 +172,23 @@ void drm_colorop_reset(struct drm_colorop *colorop)
 		__drm_colorop_reset(colorop, colorop->state);
 }
 EXPORT_SYMBOL(drm_colorop_reset);
+
+
+static const char * const colorop_type_name[] = {
+	[DRM_COLOROP_1D_CURVE] = "1D Curve",
+};
+
+/**
+ * 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)
+{
+	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 e611f830f986..cb98c55f8387 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -103,6 +103,21 @@ struct drm_colorop {
 	/** @properties: property tracking for this plane */
 	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)
@@ -127,7 +142,7 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
 }
 
 int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
-		     struct drm_plane *plane);
+		     struct drm_plane *plane, enum drm_colorop_type type);
 
 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 9e8f383935db..0cbd6bef52bc 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -866,6 +866,10 @@ struct drm_color_lut {
 	__u16 reserved;
 };
 
+enum drm_colorop_type {
+	DRM_COLOROP_1D_CURVE
+};
+
 /**
  * struct hdr_metadata_infoframe - HDR Metadata Infoframe Data.
  *
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 11/42] drm/colorop: Add 1D Curve subtype
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (9 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 10/42] drm/colorop: Add TYPE property Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 12/42] drm/colorop: Add BYPASS property Harry Wentland
                   ` (33 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland, Alex Hung

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
Co-developed-by: Alex Hung <alex.hung@amd.com>
---
 drivers/gpu/drm/drm_atomic_uapi.c | 18 +++++---
 drivers/gpu/drm/drm_colorop.c     | 71 +++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         | 24 +++++++++++
 3 files changed, 108 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index e3067c095c72..fdd540cfe24f 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -648,11 +648,17 @@ 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
@@ -662,6 +668,8 @@ 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 3a07a8fe2842..f4740b6115d1 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -36,6 +36,12 @@ 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 */
 
 int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
@@ -78,6 +84,56 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 }
 EXPORT_SYMBOL(drm_colorop_init);
 
+int drm_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_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++;
+	}
+
+	/* 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, 0);
+	drm_colorop_reset(colorop);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_colorop_curve_1d_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
@@ -192,3 +248,18 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type)
 
 	return colorop_type_name[type];
 }
+
+/**
+ * drm_get_colorop_curve_1d_type_name - return a string for 1D curve type
+ * @range: 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)
+{
+	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 cb98c55f8387..539db900f16f 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -30,6 +30,12 @@
 #include <drm/drm_mode.h>
 #include <drm/drm_property.h>
 
+enum drm_colorop_curve_1d_type {
+	DRM_COLOROP_1D_CURVE_SRGB_EOTF,
+	DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+	DRM_COLOROP_1D_CURVE_COUNT
+};
+
 /**
  * struct drm_colorop_state - mutable colorop state
  */
@@ -39,6 +45,13 @@ struct drm_colorop_state {
 
 	/* colorop properties */
 
+	/**
+	 * @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;
 };
@@ -118,6 +131,14 @@ struct drm_colorop {
 	 * this color operation. The type is enum drm_colorop_type.
 	 */
 	struct drm_property *type_property;
+
+	/**
+	 * @curve_1d_type:
+	 *
+	 * 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)
@@ -144,6 +165,9 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
 int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 		     struct drm_plane *plane, enum drm_colorop_type type);
 
+int drm_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);
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 12/42] drm/colorop: Add BYPASS property
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (10 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 11/42] drm/colorop: Add 1D Curve subtype Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 13/42] drm/colorop: Add NEXT property Harry Wentland
                   ` (32 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic_uapi.c |  6 +++++-
 drivers/gpu/drm/drm_colorop.c     | 16 ++++++++++++++++
 include/drm/drm_colorop.h         | 20 ++++++++++++++++++++
 3 files changed, 41 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index fdd540cfe24f..87f00131be11 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -648,7 +648,9 @@ 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,
@@ -668,6 +670,8 @@ 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 f4740b6115d1..29979816a2d1 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -80,6 +80,18 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   colorop->type_property,
 				   colorop->type);
 
+	/* bypass */
+	/* TODO can we reuse the mode_config->active_prop? */
+	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;
 }
 EXPORT_SYMBOL(drm_colorop_init);
@@ -123,6 +135,7 @@ int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *coloro
 	/* 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;
 
@@ -138,6 +151,8 @@ 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 *
@@ -189,6 +204,7 @@ static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
 				      struct drm_colorop *colorop)
 {
 	colorop_state->colorop = colorop;
+	colorop_state->bypass = true;
 }
 
 /**
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 539db900f16f..28aa5c1c309e 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -45,6 +45,14 @@ struct drm_colorop_state {
 
 	/* colorop properties */
 
+	/**
+	 * @bypass:
+	 *
+	 * True if colorop shall be bypassed. False if colorop is
+	 * enabled.
+	 */
+	bool bypass;
+
 	/**
 	 * @curve_1d_type:
 	 *
@@ -132,6 +140,18 @@ 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:
 	 *
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 13/42] drm/colorop: Add NEXT property
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (11 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 12/42] drm/colorop: Add BYPASS property Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 14/42] drm/colorop: Add atomic state print for drm_colorop Harry Wentland
                   ` (31 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_colorop.c | 29 +++++++++++++++++++++++++++++
 include/drm/drm_colorop.h     | 20 ++++++++++++++++++++
 2 files changed, 49 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 29979816a2d1..71c2286333a1 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -59,6 +59,7 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 	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++;
@@ -92,6 +93,15 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   colorop->bypass_property,
 				   1);
 
+	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;
 }
 EXPORT_SYMBOL(drm_colorop_init);
@@ -279,3 +289,22 @@ 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)
+{
+	if (!colorop->next_property)
+		return;
+
+	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 28aa5c1c309e..8060988b5892 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -132,6 +132,14 @@ struct drm_colorop {
 	 */
 	enum drm_colorop_type type;
 
+	/**
+	 * @next:
+	 *
+	 * Read-only
+	 * Pointer to next drm_colorop in pipeline
+	 */
+	struct drm_colorop *next;
+
 	/**
 	 * @type_property:
 	 *
@@ -159,10 +167,20 @@ 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)
 
+
+
+
 /**
  * drm_crtc_find - look up a Colorop object from its ID
  * @dev: DRM device
@@ -213,5 +231,7 @@ 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)
 
+void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
+
 
 #endif /* __DRM_COLOROP_H__ */
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 14/42] drm/colorop: Add atomic state print for drm_colorop
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (12 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 13/42] drm/colorop: Add NEXT property Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 15/42] drm/plane: Add COLOR PIPELINE property Harry Wentland
                   ` (30 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic.c | 25 ++++++++++++++++++++++++-
 include/drm/drm_colorop.h    |  5 +++++
 2 files changed, 29 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index b400e32c9d39..3645c36d63b3 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -783,6 +783,19 @@ 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);
+	drm_printf(p, "\tcurve_1d_type=%s\n", drm_get_colorop_curve_1d_type_name(state->curve_1d_type));
+}
+
 static void drm_atomic_plane_print_state(struct drm_printer *p,
 		const struct drm_plane_state *state)
 {
@@ -804,7 +817,8 @@ 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);
 }
@@ -1840,6 +1854,7 @@ 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;
@@ -1849,6 +1864,14 @@ 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);
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 8060988b5892..fc9a28d138b8 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -231,6 +231,11 @@ 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)
 
+const char *drm_get_color_pipeline_name(struct drm_colorop *colorop);
+
+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);
 
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 15/42] drm/plane: Add COLOR PIPELINE property
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (13 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 14/42] drm/colorop: Add atomic state print for drm_colorop Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 16/42] drm/colorop: Add NEXT to colorop state print Harry Wentland
                   ` (29 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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)

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 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               | 52 +++++++++++++++++++++++
 include/drm/drm_atomic.h                  |  3 ++
 include/drm/drm_atomic_uapi.h             |  2 +
 include/drm/drm_plane.h                   | 11 +++++
 7 files changed, 161 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 3645c36d63b3..27a8805c5fa1 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -1458,6 +1458,52 @@ 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 87f00131be11..86f77a9aa3a8 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -256,6 +256,36 @@ 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
@@ -539,6 +569,16 @@ 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;
+
+		/* set it on drm_plane_state */
+		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,
@@ -621,6 +661,8 @@ 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 672c655c7a8e..a747bd9eb20b 100644
--- a/drivers/gpu/drm/drm_plane.c
+++ b/drivers/gpu/drm/drm_plane.c
@@ -1729,3 +1729,55 @@ int drm_plane_create_scaling_filter_property(struct drm_plane *plane,
 	return 0;
 }
 EXPORT_SYMBOL(drm_plane_create_scaling_filter_property);
+
+#define MAX_COLOR_PIPELINES 5
+
+/**
+ * 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,
+					     struct drm_prop_enum_list *pipelines,
+					     int num_pipelines)
+{
+	struct drm_prop_enum_list all_pipelines[MAX_COLOR_PIPELINES];
+	int len = 0;
+	int i;
+	struct drm_property *prop;
+
+	if (num_pipelines > (MAX_COLOR_PIPELINES - 1))
+		return -EINVAL;
+
+	/* 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))
+		return PTR_ERR(prop);
+
+	drm_object_attach_property(&plane->base, prop, 0);
+	plane->color_pipeline_property = prop;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_create_color_pipeline_property);
\ No newline at end of file
diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index 634b2827765f..2346f19eda9f 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -797,6 +797,9 @@ 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 9c8c1322f0a6..c270bc640df8 100644
--- a/include/drm/drm_plane.h
+++ b/include/drm/drm_plane.h
@@ -766,6 +766,14 @@ 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.
@@ -979,4 +987,7 @@ drm_plane_get_damage_clips(const struct drm_plane_state *state);
 int drm_plane_create_scaling_filter_property(struct drm_plane *plane,
 					     unsigned int supported_filters);
 
+int drm_plane_create_color_pipeline_property(struct drm_plane *plane,
+					     struct drm_prop_enum_list *pipelines,
+					     int num_pipelines);
 #endif
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 16/42] drm/colorop: Add NEXT to colorop state print
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (14 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 15/42] drm/plane: Add COLOR PIPELINE property Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 17/42] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
                   ` (28 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

v3:
 - Read NEXT ID from drm_colorop's next pointer

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic.c | 1 +
 include/drm/drm_colorop.h    | 2 ++
 2 files changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 27a8805c5fa1..d82858dabf06 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -794,6 +794,7 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
 	drm_printf(p, "\ttype=%s\n", drm_get_colorop_type_name(colorop->type));
 	drm_printf(p, "\tbypass=%u\n", state->bypass);
 	drm_printf(p, "\tcurve_1d_type=%s\n", drm_get_colorop_curve_1d_type_name(state->curve_1d_type));
+	drm_printf(p, "\tnext=%d\n", colorop->next ? colorop->next->base.id : 0);
 }
 
 static void drm_atomic_plane_print_state(struct drm_printer *p,
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index fc9a28d138b8..e85ed5aa68d0 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -237,6 +237,8 @@ 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);
+uint32_t drm_colorop_get_next_property(struct drm_colorop *colorop);
+struct drm_colorop *drm_colorop_get_next(struct drm_colorop *colorop);
 
 
 #endif /* __DRM_COLOROP_H__ */
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 17/42] drm/vkms: Add enumerated 1D curve colorop
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (15 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 16/42] drm/colorop: Add NEXT to colorop state print Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-12 15:27   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 18/42] drm/vkms: Add kunit tests for linear and sRGB LUTs Harry Wentland
                   ` (27 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland, Alex Hung

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.

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
---
 drivers/gpu/drm/vkms/Makefile        |   4 +-
 drivers/gpu/drm/vkms/vkms_colorop.c  |  70 +++
 drivers/gpu/drm/vkms/vkms_composer.c |  45 ++
 drivers/gpu/drm/vkms/vkms_drv.h      |   4 +
 drivers/gpu/drm/vkms/vkms_luts.c     | 802 +++++++++++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_luts.h     |  12 +
 drivers/gpu/drm/vkms/vkms_plane.c    |   2 +
 7 files changed, 938 insertions(+), 1 deletion(-)
 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 1b28a6a32948..c38455c46be4 100644
--- a/drivers/gpu/drm/vkms/Makefile
+++ b/drivers/gpu/drm/vkms/Makefile
@@ -6,6 +6,8 @@ vkms-y := \
 	vkms_formats.o \
 	vkms_crtc.o \
 	vkms_composer.o \
-	vkms_writeback.o
+	vkms_writeback.o \
+	vkms_colorop.o \
+	vkms_luts.o
 
 obj-$(CONFIG_DRM_VKMS) += vkms.o
diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
new file mode 100644
index 000000000000..d2db366da6d3
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -0,0 +1,70 @@
+/* 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>
+
+#define MAX_COLOR_PIPELINES 5
+
+static const u64 supported_tfs =
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF);
+
+const int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
+{
+
+	struct drm_colorop *op, *prev_op;
+	struct drm_device *dev = plane->dev;
+	int ret;
+
+	/* 1st op: 1d curve */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs);
+	if (ret)
+		return ret;
+
+	list->type = op->base.id;
+	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", op->base.id);
+
+	prev_op = op;
+
+	/* 2nd op: 1d curve */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
+	return 0;
+}
+
+int vkms_initialize_colorops(struct drm_plane *plane)
+{
+	struct drm_prop_enum_list pipelines[MAX_COLOR_PIPELINES];
+	int len = 0;
+	int ret;
+
+	/* Add color pipeline */
+	ret = vkms_initialize_color_pipeline(plane, &(pipelines[len]));
+	if (ret)
+		return ret;
+	len++;
+
+	/* Create COLOR_PIPELINE property and attach */
+	drm_plane_create_color_pipeline_property(plane, pipelines, len);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index b90e446d5954..9493bdb1ba3f 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -12,6 +12,7 @@
 #include <linux/minmax.h>
 
 #include "vkms_drv.h"
+#include "vkms_luts.h"
 
 static u16 pre_mul_blend_channel(u16 src, u16 dst, u16 alpha)
 {
@@ -163,6 +164,47 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
 	}
 }
 
+static void pre_blend_color_transform(const struct vkms_plane_state *plane_state, struct line_buffer *output_buffer)
+{
+	struct drm_colorop *colorop = plane_state->base.base.color_pipeline;
+
+	while (colorop) {
+		struct drm_colorop_state *colorop_state;
+
+		if (!colorop)
+			return;
+
+		/* TODO this is probably wrong */
+		colorop_state = colorop->state;
+
+		if (!colorop_state)
+			return;
+
+		for (size_t x = 0; x < output_buffer->n_pixels; x++) {
+			struct pixel_argb_u16 *pixel = &output_buffer->pixels[x];
+
+			if (colorop->type == DRM_COLOROP_1D_CURVE &&
+				colorop_state->bypass == false) {
+				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:
+					default:
+						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;
+				}
+			}
+		}
+
+		colorop = colorop->next;
+	}
+}
+
 /**
  * blend - blend the pixels from all planes and compute crc
  * @wb: The writeback frame buffer metadata
@@ -200,6 +242,9 @@ static void blend(struct vkms_writeback_job *wb,
 				continue;
 
 			vkms_compose_row(stage_buffer, plane[i], y_pos);
+
+			pre_blend_color_transform(plane[i], stage_buffer);
+
 			pre_mul_alpha_blend(plane[i]->frame_info, stage_buffer,
 					    output_buffer);
 		}
diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
index 8f5710debb1e..2bcc24c196a2 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.h
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -170,4 +170,8 @@ void vkms_writeback_row(struct vkms_writeback_job *wb, const struct line_buffer
 /* Writeback */
 int vkms_enable_writeback_connector(struct vkms_device *vkmsdev);
 
+/* 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..6553d6d442b4
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_luts.c
@@ -0,0 +1,802 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#include <drm/drm_mode.h>
+
+#include "vkms_drv.h"
+#include "vkms_luts.h"
+
+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..053512a643f7
--- /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_ */
\ No newline at end of file
diff --git a/drivers/gpu/drm/vkms/vkms_plane.c b/drivers/gpu/drm/vkms/vkms_plane.c
index e5c625ab8e3e..8520ee0534d1 100644
--- a/drivers/gpu/drm/vkms/vkms_plane.c
+++ b/drivers/gpu/drm/vkms/vkms_plane.c
@@ -215,5 +215,7 @@ 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;
 }
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 18/42] drm/vkms: Add kunit tests for linear and sRGB LUTs
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (16 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 17/42] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 19/42] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Harry Wentland
                   ` (26 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 37 ++++++++++++++++++-
 1 file changed, 36 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
index fc73e48aa57c..e6ac01dee830 100644
--- a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
+++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
@@ -31,7 +31,6 @@ const struct vkms_color_lut test_linear_lut = {
 	.channel_value2index_ratio = 0xf000fll
 };
 
-
 static void vkms_color_test_get_lut_index(struct kunit *test)
 {
 	int i;
@@ -40,6 +39,19 @@ static void vkms_color_test_get_lut_index(struct kunit *test)
 
 	for (i = 0; i < TEST_LUT_SIZE; i++)
 		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&test_linear_lut, test_linear_array[i].red)), 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)
@@ -148,9 +160,32 @@ 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),
 	{}
 };
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 19/42] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (17 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 18/42] drm/vkms: Add kunit tests for linear and sRGB LUTs Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 20/42] drm/colorop: Add 3x4 CTM type Harry Wentland
                   ` (25 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic_uapi.c | 13 ++++++++++++-
 drivers/gpu/drm/drm_ioctl.c       |  7 +++++++
 include/drm/drm_file.h            |  7 +++++++
 include/uapi/drm/drm.h            | 16 ++++++++++++++++
 4 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 86f77a9aa3a8..ff258b34544e 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -570,6 +570,12 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 	} else if (property == plane->color_range_property) {
 		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;
 		colorop = drm_colorop_find(dev, file_priv, val);
@@ -1179,6 +1185,12 @@ 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;
 
@@ -1191,7 +1203,6 @@ 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 e368fc084c77..da59e37ae228 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 ab230d3af138..63c1d29b8520 100644
--- a/include/drm/drm_file.h
+++ b/include/drm/drm_file.h
@@ -204,6 +204,13 @@ 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 16122819edfe..2d74c49274ee 100644
--- a/include/uapi/drm/drm.h
+++ b/include/uapi/drm/drm.h
@@ -875,6 +875,22 @@ 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 of plane the COLOR_PIPELINE
+ *   property, as well as drm_colorop properties.
+ * - Drivers will ignore these properties
+ *   - COLOR_ENCODING
+ *   - COLOR_RANGE
+ *
+ * The client must enable &DRM_CLIENT_CAP_ATOMIC first.
+ *
+ * This capability is currently in development.
+ */
+#define DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE	7
+
 /* DRM_IOCTL_SET_CLIENT_CAP ioctl argument type */
 struct drm_set_client_cap {
 	__u64 capability;
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 20/42] drm/colorop: Add 3x4 CTM type
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (18 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 19/42] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 21/42] drm/vkms: Pull apply_colorop out of pre_blend_color_transform Harry Wentland
                   ` (24 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic.c      | 14 ++++++++++-
 drivers/gpu/drm/drm_atomic_uapi.c | 29 ++++++++++++++++++++++
 drivers/gpu/drm/drm_colorop.c     | 40 +++++++++++++++++++++++++++++++
 include/drm/drm_colorop.h         | 19 +++++++++++++++
 include/uapi/drm/drm_mode.h       |  9 ++++++-
 5 files changed, 109 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index d82858dabf06..6e736ffa6d7c 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -793,7 +793,19 @@ 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);
-	drm_printf(p, "\tcurve_1d_type=%s\n", drm_get_colorop_curve_1d_type_name(state->curve_1d_type));
+
+	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;
+	case DRM_COLOROP_CTM_3X4:
+		drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
+		break;
+	default:
+		break;
+	}
+
 	drm_printf(p, "\tnext=%d\n", colorop->next ? colorop->next->base.id : 0);
 }
 
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index ff258b34544e..23b248987a7c 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -691,6 +691,30 @@ 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,
@@ -700,6 +724,9 @@ 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",
@@ -722,6 +749,8 @@ 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 71c2286333a1..77777baa1eba 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -34,6 +34,7 @@
 
 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[] = {
@@ -93,6 +94,7 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 				   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)
@@ -157,6 +159,43 @@ int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *coloro
 }
 EXPORT_SYMBOL(drm_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_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
+			     struct drm_plane *plane)
+{
+	int ret;
+
+	ret = drm_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_colorop_ctm_3x4_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
@@ -258,6 +297,7 @@ EXPORT_SYMBOL(drm_colorop_reset);
 
 static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_CURVE] = "1D Curve",
+	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
 };
 
 /**
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index e85ed5aa68d0..4aee29e161d6 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -60,6 +60,14 @@ 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.
+	 */
+	struct drm_property_blob *data;
+
 	/** @state: backpointer to global drm_atomic_state */
 	struct drm_atomic_state *state;
 };
@@ -167,6 +175,17 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @data:
+	 *
+	 * 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
 	 *
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index 0cbd6bef52bc..1a8ea7934fbe 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -851,6 +851,12 @@ 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];
 };
@@ -867,7 +873,8 @@ struct drm_color_lut {
 };
 
 enum drm_colorop_type {
-	DRM_COLOROP_1D_CURVE
+	DRM_COLOROP_1D_CURVE,
+	DRM_COLOROP_CTM_3X4,
 };
 
 /**
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 21/42] drm/vkms: Pull apply_colorop out of pre_blend_color_transform
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (19 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 20/42] drm/colorop: Add 3x4 CTM type Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 22/42] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
                   ` (23 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

The if/switch statement is bound to grow with more types and
subtypes. Pull this out into its own funcion to make things more
manageable and readable.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/vkms/vkms_composer.c | 48 ++++++++++++++++------------
 1 file changed, 28 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 9493bdb1ba3f..25b786b49db0 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -164,6 +164,31 @@ 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)
+{
+	/* TODO is this right? */
+	struct drm_colorop_state *colorop_state = colorop->state;
+
+	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_DEBUG_DRIVER("unkown 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)
 {
 	struct drm_colorop *colorop = plane_state->base.base.color_pipeline;
@@ -180,26 +205,9 @@ static void pre_blend_color_transform(const struct vkms_plane_state *plane_state
 		if (!colorop_state)
 			return;
 
-		for (size_t x = 0; x < output_buffer->n_pixels; x++) {
-			struct pixel_argb_u16 *pixel = &output_buffer->pixels[x];
-
-			if (colorop->type == DRM_COLOROP_1D_CURVE &&
-				colorop_state->bypass == false) {
-				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:
-					default:
-						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;
-				}
-			}
-		}
+		for (size_t x = 0; x < output_buffer->n_pixels; x++)
+			if (!colorop_state->bypass)
+				apply_colorop(&output_buffer->pixels[x], colorop);
 
 		colorop = colorop->next;
 	}
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 22/42] drm/vkms: Use s32 for internal color pipeline precision
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (20 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 21/42] drm/vkms: Pull apply_colorop out of pre_blend_color_transform Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-12 15:50   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 23/42] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
                   ` (22 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/vkms/vkms_composer.c | 57 +++++++++++++++++++++-------
 drivers/gpu/drm/vkms/vkms_drv.h      |  4 ++
 2 files changed, 48 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 25b786b49db0..d2101fa55aa3 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -164,7 +164,7 @@ 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)
 {
 	/* TODO is this right? */
 	struct drm_colorop_state *colorop_state = colorop->state;
@@ -191,25 +191,56 @@ 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 drm_colorop *colorop = plane_state->base.base.color_pipeline;
+	struct drm_colorop *colorop;
+	struct pixel_argb_s32 pixel;
 
-	while (colorop) {
-		struct drm_colorop_state *colorop_state;
+	for (size_t x = 0; x < output_buffer->n_pixels; x++) {
 
-		if (!colorop)
-			return;
+		/*
+		 * 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;
 
-		/* TODO this is probably wrong */
-		colorop_state = colorop->state;
+		colorop = plane_state->base.base.color_pipeline;
+		while (colorop) {
+			struct drm_colorop_state *colorop_state;
 
-		if (!colorop_state)
-			return;
+			if (!colorop)
+				return;
+
+			/* TODO this is probably wrong */
+			colorop_state = colorop->state;
+
+			if (!colorop_state)
+				return;
 
-		for (size_t x = 0; x < output_buffer->n_pixels; x++)
 			if (!colorop_state->bypass)
-				apply_colorop(&output_buffer->pixels[x], colorop);
+				apply_colorop(&pixel, colorop);
 
-		colorop = colorop->next;
+			colorop = colorop->next;
+		}
+
+		/* clamp pixel */
+		pixel.a = max(min(pixel.a, 0xffff), 0x0);
+		pixel.r = max(min(pixel.r, 0xffff), 0x0);
+		pixel.g = max(min(pixel.g, 0xffff), 0x0);
+		pixel.b = max(min(pixel.b, 0xffff), 0x0);
+
+		/* put back to output_buffer */
+		output_buffer->pixels[x].a = pixel.a;
+		output_buffer->pixels[x].r = pixel.r;
+		output_buffer->pixels[x].g = pixel.g;
+		output_buffer->pixels[x].b = pixel.b;
 	}
 }
 
diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
index 2bcc24c196a2..fadb7685a360 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.h
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -36,6 +36,10 @@ struct vkms_frame_info {
 	unsigned int cpp;
 };
 
+struct pixel_argb_s32 {
+	s32 a, r, g, b;
+};
+
 struct pixel_argb_u16 {
 	u16 a, r, g, b;
 };
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 23/42] drm/vkms: add 3x4 matrix in color pipeline
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (21 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 22/42] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-14 15:23   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 24/42] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
                   ` (21 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/vkms/vkms_colorop.c  | 32 +++++++++++++++++++++++++++-
 drivers/gpu/drm/vkms/vkms_composer.c | 27 +++++++++++++++++++++++
 include/drm/drm_colorop.h            |  2 ++
 3 files changed, 60 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
index d2db366da6d3..a0e54b2c1f7a 100644
--- a/drivers/gpu/drm/vkms/vkms_colorop.c
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -35,7 +35,37 @@ const int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pro
 
 	prev_op = op;
 
-	/* 2nd op: 1d curve */
+	/* 2nd op: 3x4 matrix */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_ctm_3x4_init(dev, op, plane);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
+	prev_op = op;
+
+	/* 3rd op: 3x4 matrix */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_ctm_3x4_init(dev, op, plane);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
+	prev_op = op;
+
+	/* 4th op: 1d curve */
 	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!op) {
 		DRM_ERROR("KMS: Failed to allocate colorop\n");
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index d2101fa55aa3..8bbfce651526 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -164,6 +164,30 @@ 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;
+
+	rf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[0]), drm_int2fixp(pixel->r)) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[1]), drm_int2fixp(pixel->g)) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[2]), drm_int2fixp(pixel->b)) +
+	     drm_sm2fixp(matrix->matrix[3]);
+
+	gf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[4]), drm_int2fixp(pixel->r)) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[5]), drm_int2fixp(pixel->g)) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[6]), drm_int2fixp(pixel->b)) +
+	     drm_sm2fixp(matrix->matrix[7]);
+
+	bf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[8]), drm_int2fixp(pixel->r)) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[9]), drm_int2fixp(pixel->g)) +
+	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[10]), drm_int2fixp(pixel->b)) +
+	     drm_sm2fixp(matrix->matrix[11]);
+
+	pixel->r = drm_fixp2int(rf);
+	pixel->g = drm_fixp2int(gf);
+	pixel->b = drm_fixp2int(bf);
+}
+
 static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
 {
 	/* TODO is this right? */
@@ -185,6 +209,9 @@ static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colo
 				DRM_DEBUG_DRIVER("unkown colorop 1D curve type %d\n", 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);
 	}
 
 }
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 4aee29e161d6..8710e550790c 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -224,6 +224,8 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 
 int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
 			      struct drm_plane *plane, u64 supported_tfs);
+int drm_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);
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 24/42] drm/tests: Add a few tests around drm_fixed.h
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (22 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 23/42] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-14 15:36   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 25/42] drm/vkms: Add tests for CTM handling Harry Wentland
                   ` (20 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/tests/Makefile        |  3 +-
 drivers/gpu/drm/tests/drm_fixp_test.c | 69 +++++++++++++++++++++++++++
 2 files changed, 71 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 d6183b3d7688..98468f7908dd 100644
--- a/drivers/gpu/drm/tests/Makefile
+++ b/drivers/gpu/drm/tests/Makefile
@@ -19,6 +19,7 @@ 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..f420f173ff66
--- /dev/null
+++ b/drivers/gpu/drm/tests/drm_fixp_test.c
@@ -0,0 +1,69 @@
+// 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, ((1LL << 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");
\ No newline at end of file
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 25/42] drm/vkms: Add tests for CTM handling
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (23 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 24/42] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-03-14 16:02   ` Pekka Paalanen
  2024-02-26 21:10 ` [RFC PATCH v4 26/42] drm/colorop: pass plane_color_pipeline client cap to atomic check Harry Wentland
                   ` (19 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

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)
 - Rename red, green, blue to Y, U, V where applicable

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 251 ++++++++++++++++++
 drivers/gpu/drm/vkms/vkms_composer.c          |   2 +-
 2 files changed, 252 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
index e6ac01dee830..83d07f7bae37 100644
--- a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
+++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
@@ -3,6 +3,7 @@
 #include <kunit/test.h>
 
 #include <drm/drm_fixed.h>
+#include <drm/drm_mode.h>
 
 #define TEST_LUT_SIZE 16
 
@@ -181,11 +182,261 @@ static void vkms_color_srgb_inv_srgb(struct kunit *test)
 	}
 }
 
+#define FIXPT_HALF        (DRM_FIXED_ONE >> 1)
+#define FIXPT_QUARTER     (DRM_FIXED_ONE >> 2)
+
+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 = 0x7fff;
+	ref.g = 0x3fff;
+	ref.b = 0x3fff;
+
+	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.
+ */
+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 8bbfce651526..2c5715242f91 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -164,7 +164,7 @@ 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)
+void apply_3x4_matrix(struct pixel_argb_s32 *pixel, const struct drm_color_ctm_3x4 *matrix)
 {
 	s64 rf, gf, bf;
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 26/42] drm/colorop: pass plane_color_pipeline client cap to atomic check
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (24 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 25/42] drm/vkms: Add tests for CTM handling Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 27/42] drm/colorop: define a new macro for_each_new_colorop_in_state Harry Wentland
                   ` (18 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic_uapi.c |  1 +
 include/drm/drm_atomic.h          | 17 +++++++++++++++++
 2 files changed, 18 insertions(+)

diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 23b248987a7c..c7c1c614b0d9 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -1571,6 +1571,7 @@ 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 2346f19eda9f..61e6b5553eec 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -404,6 +404,23 @@ struct drm_atomic_state {
 	 * states.
 	 */
 	bool duplicated : 1;
+
+	/**
+	 * @plane_color_pipeline : 1
+	 *
+	 * 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;
+
 	struct __drm_colorops_state *colorops;
 	struct __drm_planes_state *planes;
 	struct __drm_crtcs_state *crtcs;
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 27/42] drm/colorop: define a new macro for_each_new_colorop_in_state
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (25 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 26/42] drm/colorop: pass plane_color_pipeline client cap to atomic check Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 28/42] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Harry Wentland
                   ` (17 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung

From: Alex Hung <alex.hung@amd.com>

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

Signed-off-by: Alex Hung <alex.hung@amd.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 61e6b5553eec..a4c5ff99a515 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -979,6 +979,26 @@ 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
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 28/42] drm/amd/display: Ignore deprecated props when plane_color_pipeline set
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (26 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 27/42] drm/colorop: define a new macro for_each_new_colorop_in_state Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 29/42] drm/amd/display: Add bypass COLOR PIPELINE Harry Wentland
                   ` (16 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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

Signed-off-by: Harry Wentland <harry.wentland@amd.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 f6575d7dee97..47c6fd33fd60 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4948,6 +4948,10 @@ 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;
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 29/42] drm/amd/display: Add bypass COLOR PIPELINE
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (27 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 28/42] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 30/42] drm/amd/display: Skip color pipeline initialization for cursor plane Harry Wentland
                   ` (15 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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

Signed-off-by: Harry Wentland <harry.wentland@amd.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 8a4c40b4c27e..c5c07b4cd6c9 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
@@ -1651,6 +1651,20 @@ 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,
@@ -1744,7 +1758,12 @@ 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);
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 30/42] drm/amd/display: Skip color pipeline initialization for cursor plane
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (28 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 29/42] drm/amd/display: Add bypass COLOR PIPELINE Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
                   ` (14 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung

From: Alex Hung <alex.hung@amd.com>

Signed-off-by: Alex Hung <alex.hung@amd.com>
---
 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 c5c07b4cd6c9..d3f64f586243 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
@@ -1659,6 +1659,9 @@ 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);
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (29 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 30/42] drm/amd/display: Skip color pipeline initialization for cursor plane Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-05-21 15:45   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 32/42] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
                   ` (13 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung, Harry Wentland

From: Alex Hung <alex.hung@amd.com>

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>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 88 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 58 ++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 +++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
 5 files changed, 192 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 9b527bffe11a..3ec759934669 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 +668,19 @@ 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 +1150,59 @@ __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 +1253,25 @@ 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 +1368,8 @@ 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..e8b7fc8bb0f1
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
@@ -0,0 +1,58 @@
+// 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);
+
+int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
+{
+	struct drm_colorop *op;
+	struct drm_device *dev = plane->dev;
+	int ret;
+
+	/* 1D curve - DEGAM TF */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_degam_tfs);
+	if (ret)
+		return ret;
+
+	list->type = op->base.id;
+	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", op->base.id);
+
+	return 0;
+}
\ No newline at end of file
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..f16de6a9fbde
--- /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__*/
\ No newline at end of file
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 d3f64f586243..3adab18332d8 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
@@ -35,6 +35,7 @@
 #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"
 
@@ -1658,10 +1659,19 @@ 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);
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 32/42] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (30 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 33/42] drm/amd/display: Add support for sRGB EOTF in BLND block Harry Wentland
                   ` (12 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung, Harry Wentland

From: Alex Hung <alex.hung@amd.com>

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>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 76 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 20 ++++-
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  1 +
 3 files changed, 96 insertions(+), 1 deletion(-)

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 3ec759934669..8788cfd26abd 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
@@ -1203,6 +1203,70 @@ __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;
+	const struct drm_color_lut *shaper_lut;
+	uint32_t shaper_size;
+
+	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, shaper_lut, shaper_size, 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)
@@ -1258,6 +1322,7 @@ 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 */
@@ -1269,6 +1334,17 @@ 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 e8b7fc8bb0f1..0d1626abf577 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
@@ -34,9 +34,12 @@
 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);
+
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
 {
-	struct drm_colorop *op;
+	struct drm_colorop *op, *prev_op;
 	struct drm_device *dev = plane->dev;
 	int ret;
 
@@ -54,5 +57,20 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 	list->type = op->base.id;
 	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", op->base.id);
 
+	prev_op = op;
+
+	/* 1D curve - SHAPER TF */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_shaper_tfs);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
 	return 0;
 }
\ No newline at end of file
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 f16de6a9fbde..c4b1b187e9bf 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);
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 33/42] drm/amd/display: Add support for sRGB EOTF in BLND block
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (31 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 32/42] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 34/42] drm/colorop: Add PQ 125 EOTF and its inverse Harry Wentland
                   ` (11 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung, Harry Wentland

From: Alex Hung <alex.hung@amd.com>

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>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 77 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 18 +++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  1 +
 3 files changed, 96 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 8788cfd26abd..3e3ae2b58b06 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
@@ -1267,6 +1267,72 @@ __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;
+
+	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)
@@ -1345,6 +1411,17 @@ 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 0d1626abf577..449a2ad6a184 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);
+
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
 {
 	struct drm_colorop *op, *prev_op;
@@ -72,5 +75,20 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	drm_colorop_set_next_property(prev_op, op);
 
+	prev_op = op;
+
+	/* 1D curve - BLND TF */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_blnd_tfs);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
 	return 0;
 }
\ No newline at end of file
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 c4b1b187e9bf..a342c096f406 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);
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 34/42] drm/colorop: Add PQ 125 EOTF and its inverse
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (32 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 33/42] drm/amd/display: Add support for sRGB EOTF in BLND block Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 35/42] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Harry Wentland
                   ` (10 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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.

Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_colorop.c | 2 ++
 include/drm/drm_colorop.h     | 2 ++
 2 files changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 77777baa1eba..781947e42b02 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -40,6 +40,8 @@ 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 8710e550790c..e06d9ea28efd 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -33,6 +33,8 @@
 enum drm_colorop_curve_1d_type {
 	DRM_COLOROP_1D_CURVE_SRGB_EOTF,
 	DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+	DRM_COLOROP_1D_CURVE_PQ_125_EOTF,
+	DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF,
 	DRM_COLOROP_1D_CURVE_COUNT
 };
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 35/42] drm/amd/display: Enable support for PQ 125 EOTF and Inverse
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (33 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 34/42] drm/colorop: Add PQ 125 EOTF and its inverse Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 36/42] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Harry Wentland
                   ` (9 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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: Harry Wentland <harry.wentland@amd.com>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 20 +++++++++++++------
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  8 +++++---
 2 files changed, 19 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 3e3ae2b58b06..20b7eb47388c 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"
@@ -676,6 +677,9 @@ 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;;
 	}
@@ -1158,8 +1162,10 @@ __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) {
@@ -1191,7 +1197,7 @@ __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;
 		}
@@ -1213,8 +1219,10 @@ __set_colorop_in_shaper_1d_curve(struct dc_plane_state *dc_plane_state,
 	const struct drm_color_lut *shaper_lut;
 	uint32_t shaper_size;
 
-	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) {
@@ -1250,7 +1258,7 @@ __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 449a2ad6a184..f99d8e09d89b 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
@@ -32,10 +32,12 @@
 #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);
+const u64 amdgpu_dm_supported_shaper_tfs =
+	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);
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 36/42] drm/colorop: add BT2020/BT709 OETF and Inverse OETF
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (34 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 35/42] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 37/42] drm/amd/display: Add support for BT.709 and BT.2020 TFs Harry Wentland
                   ` (8 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_colorop.c | 2 ++
 include/drm/drm_colorop.h     | 2 ++
 2 files changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 781947e42b02..b10cad5a7208 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -40,6 +40,8 @@ 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_BT2020_INV_OETF] = "BT.2020 Inverse OETF",
+	[DRM_COLOROP_1D_CURVE_BT2020_OETF] = "BT.2020 OETF",
 	[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 e06d9ea28efd..28b3136dabad 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -33,6 +33,8 @@
 enum drm_colorop_curve_1d_type {
 	DRM_COLOROP_1D_CURVE_SRGB_EOTF,
 	DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+	DRM_COLOROP_1D_CURVE_BT2020_INV_OETF,
+	DRM_COLOROP_1D_CURVE_BT2020_OETF,
 	DRM_COLOROP_1D_CURVE_PQ_125_EOTF,
 	DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF,
 	DRM_COLOROP_1D_CURVE_COUNT
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 37/42] drm/amd/display: Add support for BT.709 and BT.2020 TFs
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (35 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 36/42] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 38/42] drm/colorop: Add 1D Curve Custom LUT type Harry Wentland
                   ` (7 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland

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: Harry Wentland <harry.wentland@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c   | 11 ++++++++---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 10 +++++++---
 2 files changed, 15 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 20b7eb47388c..d5d356cf9fc6 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
@@ -677,6 +677,9 @@ 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_BT2020_INV_OETF:
+	case DRM_COLOROP_1D_CURVE_BT2020_OETF:
+		return TRANSFER_FUNCTION_BT709;
 	case DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
 	case DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF:
 		return TRANSFER_FUNCTION_PQ;
@@ -1287,8 +1290,10 @@ __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;
 
-	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) {
@@ -1324,7 +1329,7 @@ __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 f99d8e09d89b..bc66bd4f9fdd 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
@@ -33,14 +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_SRGB_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) |
+	BIT(DRM_COLOROP_1D_CURVE_BT2020_INV_OETF);
 
 int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
 {
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 38/42] drm/colorop: Add 1D Curve Custom LUT type
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (36 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 37/42] drm/amd/display: Add support for BT.709 and BT.2020 TFs Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 39/42] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Harry Wentland
                   ` (6 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung, Harry Wentland

From: Alex Hung <alex.hung@amd.com>

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.

Signed-off-by: Alex Hung <alex.hung@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Co-developed-by: Harry Wentland <harry.wentland@amd.com>
---
 drivers/gpu/drm/drm_atomic.c      |  4 ++++
 drivers/gpu/drm/drm_atomic_uapi.c |  5 +++++
 drivers/gpu/drm/drm_colorop.c     | 36 +++++++++++++++++++++++++++++--
 include/drm/drm_colorop.h         | 16 ++++++++++++++
 include/uapi/drm/drm_mode.h       |  1 +
 5 files changed, 60 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 6e736ffa6d7c..f7d51839ca03 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -799,6 +799,10 @@ 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", state->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 c7c1c614b0d9..6bfe857720cd 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -700,6 +700,9 @@ static int drm_atomic_color_set_data_property(struct drm_colorop *colorop,
 	bool replaced = false;
 
 	switch (colorop->type) {
+	case DRM_COLOROP_1D_LUT:
+		size = state->size * sizeof(struct drm_color_lut);
+		break;
 	case DRM_COLOROP_CTM_3X4:
 		size = sizeof(struct drm_color_ctm_3x4);
 		break;
@@ -749,6 +752,8 @@ 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->size_property) {
+		*val = state->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 b10cad5a7208..4452eaeeb242 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -34,6 +34,7 @@
 
 static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
+	{ DRM_COLOROP_1D_LUT, "1D Curve Custom LUT" },
 	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"}
 };
 
@@ -175,11 +176,41 @@ static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_color
 
 	colorop->data_property = prop;
 	drm_object_attach_property(&colorop->base,
-					colorop->data_property,
-					0);
+				   colorop->data_property,
+				   0);
 
 	return 0;
 }
+int drm_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_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_ATOMIC, "SIZE", lut_size, lut_size);
+	if (!prop)
+		return -ENOMEM;
+
+	colorop->size_property = prop;
+	drm_object_attach_property(&colorop->base, colorop->size_property, 0);
+
+	/* data */
+	ret = drm_colorop_create_data_prop(dev, colorop);
+	if (ret)
+		return ret;
+
+	drm_colorop_reset(colorop);
+	colorop->state->size = lut_size;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_colorop_curve_1d_lut_init);
 
 int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
 			     struct drm_plane *plane)
@@ -301,6 +332,7 @@ EXPORT_SYMBOL(drm_colorop_reset);
 
 static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_CURVE] = "1D Curve",
+	[DRM_COLOROP_1D_LUT] = "1D Curve Custom LUT",
 	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
 };
 
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 28b3136dabad..8adc7ece3bd1 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -64,6 +64,13 @@ struct drm_colorop_state {
 	 */
 	enum drm_colorop_curve_1d_type curve_1d_type;
 
+	/**
+	 * @size:
+	 *
+	 * Number of entries of the custom LUT. This should be read-only.
+	 */
+	uint32_t size;
+
 	/**
 	 * @data:
 	 *
@@ -179,6 +186,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @size_property:
+	 *
+	 * Size property for custom LUT from userspace.
+	 */
+	struct drm_property *size_property;
+
 	/**
 	 * @data:
 	 *
@@ -228,6 +242,8 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
 
 int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
 			      struct drm_plane *plane, u64 supported_tfs);
+int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
+				  struct drm_plane *plane, uint32_t lut_size);
 int drm_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 1a8ea7934fbe..07fd66dc477c 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -874,6 +874,7 @@ struct drm_color_lut {
 
 enum drm_colorop_type {
 	DRM_COLOROP_1D_CURVE,
+	DRM_COLOROP_1D_LUT,
 	DRM_COLOROP_CTM_3X4,
 };
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 39/42] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (37 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 38/42] drm/colorop: Add 1D Curve Custom LUT type Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 40/42] drm/amd/display: add 3x4 matrix colorop Harry Wentland
                   ` (5 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung, Harry Wentland

From: Alex Hung <alex.hung@amd.com>

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>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 170 +++++++++---------
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c |  30 ++++
 2 files changed, 119 insertions(+), 81 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 d5d356cf9fc6..e7b51b29cc04 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
@@ -1212,40 +1212,6 @@ __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;
-	const struct drm_color_lut *shaper_lut;
-	uint32_t shaper_size;
-
-	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, shaper_lut, shaper_size, false);
-	}
-
-	return -EINVAL;
-}
-
 static int
 __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 			      struct dc_plane_state *dc_plane_state,
@@ -1254,64 +1220,61 @@ __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;
+	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;
 
+	/* 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) {
+	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;
+		__set_output_tf(tf, shaper_lut, shaper_size, false);
+	}
+
+	/* 1D LUT - SHAPER LUT */
+	colorop = old_colorop->next;
+	if (!colorop) {
+		drm_dbg(dev, "no Shaper 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;
-
-	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;
-
-	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) {
-		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) {
+	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_state->size)
+			__set_output_tf(tf, shaper_lut, shaper_size, false);
 	}
 
-	return -EINVAL;
+	return 0;
 }
 
 static int
@@ -1322,28 +1285,63 @@ __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_state->size)
+			__set_input_tf(NULL, tf, blend_lut, blend_size);
+	}
+
+	return 0;
 }
 
 static int
@@ -1413,7 +1411,7 @@ 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");
@@ -1424,7 +1422,12 @@ 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");
@@ -1435,6 +1438,11 @@ 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 bc66bd4f9fdd..08480bf61dc5 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 =
@@ -83,6 +84,21 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	prev_op = op;
 
+	/* 1D LUT - SHAPER LUT */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
+	prev_op = op;
+
 	/* 1D curve - BLND TF */
 	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!op) {
@@ -96,5 +112,19 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	drm_colorop_set_next_property(prev_op, op);
 
+	prev_op = op;
+
+	/* 1D LUT - BLND LUT */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
 	return 0;
 }
\ No newline at end of file
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 40/42] drm/amd/display: add 3x4 matrix colorop
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (38 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 39/42] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-26 21:10 ` [RFC PATCH v4 41/42] drm/colorop: Add mutliplier type Harry Wentland
                   ` (4 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung, Harry Wentland

From: Alex Hung <alex.hung@amd.com>

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>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 50 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 15 ++++++
 2 files changed, 65 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 e7b51b29cc04..ef50640b362b 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
@@ -1212,6 +1212,45 @@ __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 (%ld) isn't equal to drm_color_ctm_3x4 (%ld)\n",
+				 blob->length, sizeof(struct drm_color_ctm_3x4));
+	}
+
+	return 0;
+}
+
 static int
 __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
 			      struct dc_plane_state *dc_plane_state,
@@ -1411,6 +1450,17 @@ 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 08480bf61dc5..ba42f1f6b620 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
@@ -69,6 +69,21 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	prev_op = op;
 
+	/* 3x4 matrix */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_ctm_3x4_init(dev, op, plane);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
+	prev_op = op;
+
 	/* 1D curve - SHAPER TF */
 	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!op) {
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 41/42] drm/colorop: Add mutliplier type
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (39 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 40/42] drm/amd/display: add 3x4 matrix colorop Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-05-21 15:38   ` Melissa Wen
  2024-02-26 21:10 ` [RFC PATCH v4 42/42] drm/amd/display: add multiplier colorop Harry Wentland
                   ` (3 subsequent siblings)
  44 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung

From: Alex Hung <alex.hung@amd.com>

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.

Signed-off-by: Alex Hung <alex.hung@amd.com>
---
 drivers/gpu/drm/drm_atomic.c      |  3 +++
 drivers/gpu/drm/drm_atomic_uapi.c |  4 ++++
 drivers/gpu/drm/drm_colorop.c     | 29 +++++++++++++++++++++++++++--
 include/drm/drm_colorop.h         | 16 ++++++++++++++++
 include/uapi/drm/drm_mode.h       |  1 +
 5 files changed, 51 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index f7d51839ca03..af0b6338a55c 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -806,6 +806,9 @@ 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=%u\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 6bfe857720cd..b4ecda563728 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -727,6 +727,8 @@ 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);
@@ -752,6 +754,8 @@ 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->size_property) {
 		*val = state->size;
 	} else if (property == colorop->data_property) {
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 4452eaeeb242..c6cdd743de51 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -35,7 +35,8 @@
 static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
 	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
 	{ DRM_COLOROP_1D_LUT, "1D Curve Custom LUT" },
-	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"}
+	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
+	{ DRM_COLOROP_MULTIPLIER, "Multiplier"},
 };
 
 static const char * const colorop_curve_1d_type_names[] = {
@@ -231,6 +232,29 @@ int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop
 }
 EXPORT_SYMBOL(drm_colorop_ctm_3x4_init);
 
+int drm_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
+			      struct drm_plane *plane)
+{
+	struct drm_property *prop;
+	int ret;
+
+	ret = drm_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_colorop_mult_init);
+
 static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
 							struct drm_colorop_state *state)
 {
@@ -333,7 +357,8 @@ EXPORT_SYMBOL(drm_colorop_reset);
 static const char * const colorop_type_name[] = {
 	[DRM_COLOROP_1D_CURVE] = "1D Curve",
 	[DRM_COLOROP_1D_LUT] = "1D Curve Custom LUT",
-	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
+	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
+	[DRM_COLOROP_MULTIPLIER] = "Multiplier",
 };
 
 /**
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 8adc7ece3bd1..f9f83644cc9f 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -64,6 +64,13 @@ 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;
+
 	/**
 	 * @size:
 	 *
@@ -186,6 +193,13 @@ struct drm_colorop {
 	 */
 	struct drm_property *curve_1d_type_property;
 
+	/**
+	 * @multiplier_property:
+	 *
+	 * Multiplier property for plane gain
+	 */
+	struct drm_property *multiplier_property;
+
 	/**
 	 * @size_property:
 	 *
@@ -246,6 +260,8 @@ int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *co
 				  struct drm_plane *plane, uint32_t lut_size);
 int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
 			     struct drm_plane *plane);
+int drm_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 07fd66dc477c..754fd4c48123 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -876,6 +876,7 @@ enum drm_colorop_type {
 	DRM_COLOROP_1D_CURVE,
 	DRM_COLOROP_1D_LUT,
 	DRM_COLOROP_CTM_3X4,
+	DRM_COLOROP_MULTIPLIER,
 };
 
 /**
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* [RFC PATCH v4 42/42] drm/amd/display: add multiplier colorop
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (40 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 41/42] drm/colorop: Add mutliplier type Harry Wentland
@ 2024-02-26 21:10 ` Harry Wentland
  2024-02-27 10:26 ` [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Joshua Ashton
                   ` (2 subsequent siblings)
  44 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-26 21:10 UTC (permalink / raw)
  To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung

From: Alex Hung <alex.hung@amd.com>

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>
---
 .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 40 +++++++++++++++++++
 .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 15 +++++++
 2 files changed, 55 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 ef50640b362b..b05e4fea8a08 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
@@ -1251,6 +1251,35 @@ __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,
@@ -1461,6 +1490,17 @@ 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 ba42f1f6b620..b739d6cb3e6b 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
@@ -84,6 +84,21 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
 
 	prev_op = op;
 
+	/* Multiplier */
+	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+	if (!op) {
+		DRM_ERROR("KMS: Failed to allocate colorop\n");
+		return -ENOMEM;
+	}
+
+	ret = drm_colorop_mult_init(dev, op, plane);
+	if (ret)
+		return ret;
+
+	drm_colorop_set_next_property(prev_op, op);
+
+	prev_op = op;
+
 	/* 1D curve - SHAPER TF */
 	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
 	if (!op) {
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (41 preceding siblings ...)
  2024-02-26 21:10 ` [RFC PATCH v4 42/42] drm/amd/display: add multiplier colorop Harry Wentland
@ 2024-02-27 10:26 ` Joshua Ashton
  2024-02-27 14:00   ` Harry Wentland
  2024-02-29 10:43 ` Daniel Vetter
  2024-05-21 15:50 ` Melissa Wen
  44 siblings, 1 reply; 69+ messages in thread
From: Joshua Ashton @ 2024-02-27 10:26 UTC (permalink / raw)
  To: Harry Wentland, dri-devel, amd-gfx
  Cc: wayland-devel, Ville Syrjala, Pekka Paalanen, Simon Ser,
	Melissa Wen, Jonas Ådahl, Sebastian Wick, Shashank Sharma,
	Alexander Goins, Michel Dänzer, Aleix Pol, Xaver Hugl,
	Victoria Brekenfeld, Sima, Uma Shankar, Naseer Ahmed,
	Christopher Braga, Abhinav Kumar, Arthur Grillo, Hector Martin,
	Liviu Dudau, Sasha McIntosh, Chaitanya Kumar Borah



On 2/26/24 21:10, Harry Wentland wrote:
> This is an RFC set for a color pipeline API, along with a sample
> implementation in VKMS. All the key API bits are here. VKMS now
> supports two named transfer function colorops and two matrix
> colorops. We have IGT tests that check all four of these colorops
> with a pixel-by-pixel comparison that checks that these colorops
> do what we expect them to do with a +/- 1 8 bpc code point margin.
> 
> The big new change with v4 is the addition of an amdgpu color
> pipeline, for all AMD GPUs with DCN 3 and newer. Amdgpu now support
> the following:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 1D Curve EOTF
> 7. 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.
> 
> We are working on two more ops for amdgpu, the HDR multiplier
> and the 3DLUT, which will give us this:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. HDR Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D Curve EOTF
> 8. 1D LUT
> 
> This, essentially 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
> 
> After this we'd like to also add the following ops:
> - Scaler (Informational only)

Why informational only? Having NEAREST and in general custom taps should 
be possible on AMDGPU right?

We don't have to solve this now, but I just want to make sure that we 
aren't locking this to info only.

Thanks

- Joshie 🐸✨

> - Color Encoding, to replace drm_plane's COLOR_ENCODING
> - Color Range, to replace drm_plane's COLOR_RANGE
> 
> This patchset is grouped as follows:
>   - Patches 1-3: couple general patches/fixes
>   - Patches 4-7: introduce kunit to VKMS
>   - Patch 7: description of motivation and details behind the
>              Color Pipeline API. If you're reading nothing else
>              but are interested in the topic I highly recommend
>              you take a look at this.
>   - Patches 7-27: DRM core and VKMS changes for color pipeline API
>   - Patches 28-40: DRM core and amdgpu changes for color pipeline API
> 
> 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 here but
> haven't had a chance to look at. These will (hopefully) be
> addressed in future iterations, either in VKMS or amdgpu:
>   - Clear documentation for each drm_colorop_type
>   - Add custom LUT colorops to VKMS
>   - Add pre-blending 3DLUT
>   - How to support HW which can't bypass entire pipeline?
>   - Add ability to create colorops that don't have BYPASS
>   - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>   - read-only scaling colorop which defines scaling taps and position
>   - read-only color format colorop to define supported color formats
>     for a pipeline
>   - named matrices, for things like converting YUV to RGB
> 
> IGT tests can be found at
> https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/merge_requests/1
> 
> IGT patches are also being sent to the igt-dev mailing list.
> 
> If you prefer a gitlab MR for review you can find it at
> https://gitlab.freedesktop.org/hwentland/linux/-/merge_requests/5
> 
> 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 (10):
>    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
> 
> Harry Wentland (32):
>    drm: Don't treat 0 as -1 in drm_fixp2int_ceil
>    drm: Add helper for conversion from signed-magnitude
>    drm: Correctly round for fixp2int_round
>    drm/vkms: Round fixp2int conversion in lerp_u16
>    drm/vkms: Create separate Kconfig file for VKMS
>    drm/vkms: Add kunit tests for VKMS LUT handling
>    drm/vkms: Avoid reading beyond LUT array
>    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: Add NEXT to colorop state print
>    drm/vkms: Add enumerated 1D curve colorop
>    drm/vkms: Add kunit tests for linear and sRGB LUTs
>    drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
>    drm/colorop: Add 3x4 CTM type
>    drm/vkms: Pull apply_colorop out of pre_blend_color_transform
>    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
> 
>   Documentation/gpu/rfc/color_pipeline.rst      | 360 ++++++++
>   drivers/gpu/drm/Kconfig                       |  14 +-
>   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   | 352 ++++++++
>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 160 ++++
>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
>   .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
>   drivers/gpu/drm/drm_atomic.c                  | 161 +++-
>   drivers/gpu/drm/drm_atomic_helper.c           |  12 +
>   drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
>   drivers/gpu/drm/drm_atomic_uapi.c             | 158 ++++
>   drivers/gpu/drm/drm_colorop.c                 | 411 +++++++++
>   drivers/gpu/drm/drm_ioctl.c                   |   7 +
>   drivers/gpu/drm/drm_mode_config.c             |   7 +
>   drivers/gpu/drm/drm_plane.c                   |  52 ++
>   drivers/gpu/drm/tests/Makefile                |   3 +-
>   drivers/gpu/drm/tests/drm_fixp_test.c         |  69 ++
>   drivers/gpu/drm/vkms/Kconfig                  |  20 +
>   drivers/gpu/drm/vkms/Makefile                 |   4 +-
>   drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>   drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 449 ++++++++++
>   drivers/gpu/drm/vkms/vkms_colorop.c           | 100 +++
>   drivers/gpu/drm/vkms/vkms_composer.c          | 135 ++-
>   drivers/gpu/drm/vkms/vkms_drv.h               |   8 +
>   drivers/gpu/drm/vkms/vkms_luts.c              | 802 ++++++++++++++++++
>   drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
>   drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
>   include/drm/drm_atomic.h                      | 122 +++
>   include/drm/drm_atomic_uapi.h                 |   3 +
>   include/drm/drm_colorop.h                     | 301 +++++++
>   include/drm/drm_file.h                        |   7 +
>   include/drm/drm_fixed.h                       |  35 +-
>   include/drm/drm_mode_config.h                 |  18 +
>   include/drm/drm_plane.h                       |  13 +
>   include/uapi/drm/drm.h                        |  16 +
>   include/uapi/drm/drm_mode.h                   |  14 +
>   38 files changed, 3882 insertions(+), 30 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/Kconfig
>   create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>   create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>   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
>   create mode 100644 include/drm/drm_colorop.h
> 
> --
> 2.44.0
> 


^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling
  2024-02-26 21:10 ` [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
@ 2024-02-27 12:14   ` Arthur Grillo
  2024-02-27 14:09     ` Harry Wentland
  2024-03-11 13:45   ` Pekka Paalanen
  1 sibling, 1 reply; 69+ messages in thread
From: Arthur Grillo @ 2024-02-27 12:14 UTC (permalink / raw)
  To: Harry Wentland, dri-devel, amd-gfx; +Cc: wayland-devel



On 26/02/24 18:10, Harry Wentland wrote:
> 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
> 
> v4:
>  - Test the critical points of the lerp function (Pekka)
> 
> v3:
>  - Use include way of testing static functions (Arthur)
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> ---
>  drivers/gpu/drm/vkms/Kconfig                  |   5 +
>  drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>  drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 163 ++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_composer.c          |   8 +-
>  4 files changed, 178 insertions(+), 2 deletions(-)
>  create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>  create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> 
> diff --git a/drivers/gpu/drm/vkms/Kconfig b/drivers/gpu/drm/vkms/Kconfig
> index b9ecdebecb0b..c1f8b343ff0e 100644
> --- a/drivers/gpu/drm/vkms/Kconfig
> +++ b/drivers/gpu/drm/vkms/Kconfig
> @@ -13,3 +13,8 @@ config DRM_VKMS
>  	  a VKMS.
>  
>  	  If M is selected the module will be called vkms.
> +
> +config DRM_VKMS_KUNIT_TESTS
> +	tristate "Tests for VKMS" if !KUNIT_ALL_TESTS
> +	depends on DRM_VKMS && KUNIT
> +	default KUNIT_ALL_TESTS
> diff --git a/drivers/gpu/drm/vkms/tests/.kunitconfig b/drivers/gpu/drm/vkms/tests/.kunitconfig
> new file mode 100644
> index 000000000000..70e378228cbd
> --- /dev/null
> +++ b/drivers/gpu/drm/vkms/tests/.kunitconfig
> @@ -0,0 +1,4 @@
> +CONFIG_KUNIT=y
> +CONFIG_DRM=y
> +CONFIG_DRM_VKMS=y
> +CONFIG_DRM_VKMS_KUNIT_TESTS=y
> diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> new file mode 100644
> index 000000000000..fc73e48aa57c
> --- /dev/null
> +++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> @@ -0,0 +1,163 @@
> +/* SPDX-License-Identifier: GPL-2.0+ */
> +
> +#include <kunit/test.h>
> +
> +#include <drm/drm_fixed.h>
> +
> +#define TEST_LUT_SIZE 16
> +
> +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 },
> +};
> +
> +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)
> +{
> +	int i;
> +
> +	KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&test_linear_lut, test_linear_array[0].red)), 0);
> +
> +	for (i = 0; i < TEST_LUT_SIZE; i++)
> +		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&test_linear_lut, test_linear_array[i].red)), 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_LICENSE("GPL");
> \ No newline at end of file
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index e70cd473e3be..d178f2a400f6 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -91,7 +91,7 @@ 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)
> +u16 lerp_u16(u16 a, u16 b, s64 t)

You don't need to do that, as already said on [1].

>  {
>  	s64 a_fp = drm_int2fixp(a);
>  	s64 b_fp = drm_int2fixp(b);
> @@ -101,7 +101,7 @@ static u16 lerp_u16(u16 a, u16 b, s64 t)
>  	return drm_fixp2int_round(a_fp + delta);
>  }
>  
> -static s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
> +s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)

Same here.

[1]: https://lore.kernel.org/dri-devel/4df0cb7c-e738-4927-89ed-51a0e8bcac77@riseup.net/

Best Regards,
~Arthur Grillo

>  {
>  	s64 color_channel_fp = drm_int2fixp(channel_value);
>  
> @@ -429,3 +429,7 @@ int vkms_set_crc_source(struct drm_crtc *crtc, const char *src_name)
>  
>  	return ret;
>  }
> +
> +#ifdef CONFIG_DRM_VKMS_KUNIT_TESTS
> +#include "tests/vkms_color_tests.c"
> +#endif

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS
  2024-02-27 10:26 ` [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Joshua Ashton
@ 2024-02-27 14:00   ` Harry Wentland
  0 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-27 14:00 UTC (permalink / raw)
  To: Joshua Ashton, dri-devel, amd-gfx
  Cc: wayland-devel, Ville Syrjala, Pekka Paalanen, Simon Ser,
	Melissa Wen, Jonas Ådahl, Sebastian Wick, Shashank Sharma,
	Alexander Goins, Michel Dänzer, Aleix Pol, Xaver Hugl,
	Victoria Brekenfeld, Sima, Uma Shankar, Naseer Ahmed,
	Christopher Braga, Abhinav Kumar, Arthur Grillo, Hector Martin,
	Liviu Dudau, Sasha McIntosh, Chaitanya Kumar Borah



On 2024-02-27 05:26, Joshua Ashton wrote:
> 
> 
> On 2/26/24 21:10, Harry Wentland wrote:
>> This is an RFC set for a color pipeline API, along with a sample
>> implementation in VKMS. All the key API bits are here. VKMS now
>> supports two named transfer function colorops and two matrix
>> colorops. We have IGT tests that check all four of these colorops
>> with a pixel-by-pixel comparison that checks that these colorops
>> do what we expect them to do with a +/- 1 8 bpc code point margin.
>>
>> The big new change with v4 is the addition of an amdgpu color
>> pipeline, for all AMD GPUs with DCN 3 and newer. Amdgpu now support
>> the following:
>>
>> 1. 1D Curve EOTF
>> 2. 3x4 CTM
>> 3. Multiplier
>> 4. 1D Curve Inverse EOTF
>> 5. 1D LUT
>> 6. 1D Curve EOTF
>> 7. 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.
>>
>> We are working on two more ops for amdgpu, the HDR multiplier
>> and the 3DLUT, which will give us this:
>>
>> 1. 1D Curve EOTF
>> 2. 3x4 CTM
>> 3. HDR Multiplier
>> 4. 1D Curve Inverse EOTF
>> 5. 1D LUT
>> 6. 3D LUT
>> 7. 1D Curve EOTF
>> 8. 1D LUT
>>
>> This, essentially 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
>>
>> After this we'd like to also add the following ops:
>> - Scaler (Informational only)
> 
> Why informational only? Having NEAREST and in general custom taps should be possible on AMDGPU right?
> 
> We don't have to solve this now, but I just want to make sure that we aren't locking this to info only.
> 

No, this isn't locking it to information only. We could allow for NEAREST or
even custom taps in the future. Just don't want to open that debate now if
we don't have a good reason to.

Harry

> Thanks
> 
> - Joshie 🐸✨
> 
>> - Color Encoding, to replace drm_plane's COLOR_ENCODING
>> - Color Range, to replace drm_plane's COLOR_RANGE
>>
>> This patchset is grouped as follows:
>>   - Patches 1-3: couple general patches/fixes
>>   - Patches 4-7: introduce kunit to VKMS
>>   - Patch 7: description of motivation and details behind the
>>              Color Pipeline API. If you're reading nothing else
>>              but are interested in the topic I highly recommend
>>              you take a look at this.
>>   - Patches 7-27: DRM core and VKMS changes for color pipeline API
>>   - Patches 28-40: DRM core and amdgpu changes for color pipeline API
>>
>> 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 here but
>> haven't had a chance to look at. These will (hopefully) be
>> addressed in future iterations, either in VKMS or amdgpu:
>>   - Clear documentation for each drm_colorop_type
>>   - Add custom LUT colorops to VKMS
>>   - Add pre-blending 3DLUT
>>   - How to support HW which can't bypass entire pipeline?
>>   - Add ability to create colorops that don't have BYPASS
>>   - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>>   - read-only scaling colorop which defines scaling taps and position
>>   - read-only color format colorop to define supported color formats
>>     for a pipeline
>>   - named matrices, for things like converting YUV to RGB
>>
>> IGT tests can be found at
>> https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/merge_requests/1
>>
>> IGT patches are also being sent to the igt-dev mailing list.
>>
>> If you prefer a gitlab MR for review you can find it at
>> https://gitlab.freedesktop.org/hwentland/linux/-/merge_requests/5
>>
>> 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 (10):
>>    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
>>
>> Harry Wentland (32):
>>    drm: Don't treat 0 as -1 in drm_fixp2int_ceil
>>    drm: Add helper for conversion from signed-magnitude
>>    drm: Correctly round for fixp2int_round
>>    drm/vkms: Round fixp2int conversion in lerp_u16
>>    drm/vkms: Create separate Kconfig file for VKMS
>>    drm/vkms: Add kunit tests for VKMS LUT handling
>>    drm/vkms: Avoid reading beyond LUT array
>>    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: Add NEXT to colorop state print
>>    drm/vkms: Add enumerated 1D curve colorop
>>    drm/vkms: Add kunit tests for linear and sRGB LUTs
>>    drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
>>    drm/colorop: Add 3x4 CTM type
>>    drm/vkms: Pull apply_colorop out of pre_blend_color_transform
>>    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
>>
>>   Documentation/gpu/rfc/color_pipeline.rst      | 360 ++++++++
>>   drivers/gpu/drm/Kconfig                       |  14 +-
>>   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   | 352 ++++++++
>>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 160 ++++
>>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
>>   .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
>>   drivers/gpu/drm/drm_atomic.c                  | 161 +++-
>>   drivers/gpu/drm/drm_atomic_helper.c           |  12 +
>>   drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
>>   drivers/gpu/drm/drm_atomic_uapi.c             | 158 ++++
>>   drivers/gpu/drm/drm_colorop.c                 | 411 +++++++++
>>   drivers/gpu/drm/drm_ioctl.c                   |   7 +
>>   drivers/gpu/drm/drm_mode_config.c             |   7 +
>>   drivers/gpu/drm/drm_plane.c                   |  52 ++
>>   drivers/gpu/drm/tests/Makefile                |   3 +-
>>   drivers/gpu/drm/tests/drm_fixp_test.c         |  69 ++
>>   drivers/gpu/drm/vkms/Kconfig                  |  20 +
>>   drivers/gpu/drm/vkms/Makefile                 |   4 +-
>>   drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>>   drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 449 ++++++++++
>>   drivers/gpu/drm/vkms/vkms_colorop.c           | 100 +++
>>   drivers/gpu/drm/vkms/vkms_composer.c          | 135 ++-
>>   drivers/gpu/drm/vkms/vkms_drv.h               |   8 +
>>   drivers/gpu/drm/vkms/vkms_luts.c              | 802 ++++++++++++++++++
>>   drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
>>   drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
>>   include/drm/drm_atomic.h                      | 122 +++
>>   include/drm/drm_atomic_uapi.h                 |   3 +
>>   include/drm/drm_colorop.h                     | 301 +++++++
>>   include/drm/drm_file.h                        |   7 +
>>   include/drm/drm_fixed.h                       |  35 +-
>>   include/drm/drm_mode_config.h                 |  18 +
>>   include/drm/drm_plane.h                       |  13 +
>>   include/uapi/drm/drm.h                        |  16 +
>>   include/uapi/drm/drm_mode.h                   |  14 +
>>   38 files changed, 3882 insertions(+), 30 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/Kconfig
>>   create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>>   create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>>   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
>>   create mode 100644 include/drm/drm_colorop.h
>>
>> -- 
>> 2.44.0
>>
> 


^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling
  2024-02-27 12:14   ` Arthur Grillo
@ 2024-02-27 14:09     ` Harry Wentland
  0 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-02-27 14:09 UTC (permalink / raw)
  To: Arthur Grillo, dri-devel, amd-gfx; +Cc: wayland-devel



On 2024-02-27 07:14, Arthur Grillo wrote:
> 
> 
> On 26/02/24 18:10, Harry Wentland wrote:
>> 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
>>
>> v4:
>>  - Test the critical points of the lerp function (Pekka)
>>
>> v3:
>>  - Use include way of testing static functions (Arthur)
>>
>> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
>> Cc: Arthur Grillo <arthurgrillo@riseup.net>
>> ---
>>  drivers/gpu/drm/vkms/Kconfig                  |   5 +
>>  drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>>  drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 163 ++++++++++++++++++
>>  drivers/gpu/drm/vkms/vkms_composer.c          |   8 +-
>>  4 files changed, 178 insertions(+), 2 deletions(-)
>>  create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>>  create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>>
>> diff --git a/drivers/gpu/drm/vkms/Kconfig b/drivers/gpu/drm/vkms/Kconfig
>> index b9ecdebecb0b..c1f8b343ff0e 100644
>> --- a/drivers/gpu/drm/vkms/Kconfig
>> +++ b/drivers/gpu/drm/vkms/Kconfig
>> @@ -13,3 +13,8 @@ config DRM_VKMS
>>  	  a VKMS.
>>  
>>  	  If M is selected the module will be called vkms.
>> +
>> +config DRM_VKMS_KUNIT_TESTS
>> +	tristate "Tests for VKMS" if !KUNIT_ALL_TESTS
>> +	depends on DRM_VKMS && KUNIT
>> +	default KUNIT_ALL_TESTS
>> diff --git a/drivers/gpu/drm/vkms/tests/.kunitconfig b/drivers/gpu/drm/vkms/tests/.kunitconfig
>> new file mode 100644
>> index 000000000000..70e378228cbd
>> --- /dev/null
>> +++ b/drivers/gpu/drm/vkms/tests/.kunitconfig
>> @@ -0,0 +1,4 @@
>> +CONFIG_KUNIT=y
>> +CONFIG_DRM=y
>> +CONFIG_DRM_VKMS=y
>> +CONFIG_DRM_VKMS_KUNIT_TESTS=y
>> diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>> new file mode 100644
>> index 000000000000..fc73e48aa57c
>> --- /dev/null
>> +++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>> @@ -0,0 +1,163 @@
>> +/* SPDX-License-Identifier: GPL-2.0+ */
>> +
>> +#include <kunit/test.h>
>> +
>> +#include <drm/drm_fixed.h>
>> +
>> +#define TEST_LUT_SIZE 16
>> +
>> +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 },
>> +};
>> +
>> +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)
>> +{
>> +	int i;
>> +
>> +	KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&test_linear_lut, test_linear_array[0].red)), 0);
>> +
>> +	for (i = 0; i < TEST_LUT_SIZE; i++)
>> +		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&test_linear_lut, test_linear_array[i].red)), 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_LICENSE("GPL");
>> \ No newline at end of file
>> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
>> index e70cd473e3be..d178f2a400f6 100644
>> --- a/drivers/gpu/drm/vkms/vkms_composer.c
>> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
>> @@ -91,7 +91,7 @@ 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)
>> +u16 lerp_u16(u16 a, u16 b, s64 t)
> 
> You don't need to do that, as already said on [1].
> 
>>  {
>>  	s64 a_fp = drm_int2fixp(a);
>>  	s64 b_fp = drm_int2fixp(b);
>> @@ -101,7 +101,7 @@ static u16 lerp_u16(u16 a, u16 b, s64 t)
>>  	return drm_fixp2int_round(a_fp + delta);
>>  }
>>  
>> -static s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
>> +s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
> 
> Same here.
> 
> [1]: https://lore.kernel.org/dri-devel/4df0cb7c-e738-4927-89ed-51a0e8bcac77@riseup.net/
> 

Oops, apologies. Somehow missed this when spinning the set.
static will be back in the next version.

Harry

> Best Regards,
> ~Arthur Grillo
> 
>>  {
>>  	s64 color_channel_fp = drm_int2fixp(channel_value);
>>  
>> @@ -429,3 +429,7 @@ int vkms_set_crc_source(struct drm_crtc *crtc, const char *src_name)
>>  
>>  	return ret;
>>  }
>> +
>> +#ifdef CONFIG_DRM_VKMS_KUNIT_TESTS
>> +#include "tests/vkms_color_tests.c"
>> +#endif


^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (42 preceding siblings ...)
  2024-02-27 10:26 ` [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Joshua Ashton
@ 2024-02-29 10:43 ` Daniel Vetter
  2024-02-29 15:31   ` Joshua Ashton
  2024-05-21 15:50 ` Melissa Wen
  44 siblings, 1 reply; 69+ messages in thread
From: Daniel Vetter @ 2024-02-29 10:43 UTC (permalink / raw)
  To: Harry Wentland
  Cc: dri-devel, amd-gfx, wayland-devel, Ville Syrjala, Pekka Paalanen,
	Simon Ser, Melissa Wen, Jonas Ådahl, Sebastian Wick,
	Shashank Sharma, Alexander Goins, Joshua Ashton,
	Michel Dänzer, Aleix Pol, Xaver Hugl, Victoria Brekenfeld,
	Sima, Uma Shankar, Naseer Ahmed, Christopher Braga,
	Abhinav Kumar, Arthur Grillo, Hector Martin, Liviu Dudau,
	Sasha McIntosh, Chaitanya Kumar Borah

On Mon, Feb 26, 2024 at 04:10:14PM -0500, Harry Wentland wrote:
> This is an RFC set for a color pipeline API, along with a sample
> implementation in VKMS. All the key API bits are here. VKMS now
> supports two named transfer function colorops and two matrix
> colorops. We have IGT tests that check all four of these colorops
> with a pixel-by-pixel comparison that checks that these colorops
> do what we expect them to do with a +/- 1 8 bpc code point margin.

So vkms is definitely great to make sure the igts are generic enough and
somewhat useful, but ... does steam run on vkms too? I think that would be
a really good test to show that the api we have here is actually useful
for compositors in a cross-driver way, and not just a neat idea that
doesn't survive harsh reality.

And yes I realize that's probably going to be a bunch of work, but I feel
like the color pipeline discussion has dragged around enough in
hypotheticals and concerns that I think it would really help a lot.

Thoughts?
-Sima

> 
> The big new change with v4 is the addition of an amdgpu color
> pipeline, for all AMD GPUs with DCN 3 and newer. Amdgpu now support
> the following:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 1D Curve EOTF
> 7. 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.
> 
> We are working on two more ops for amdgpu, the HDR multiplier
> and the 3DLUT, which will give us this:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. HDR Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D Curve EOTF
> 8. 1D LUT
> 
> This, essentially 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
> 
> After this we'd like to also add the following ops:
> - Scaler (Informational only)
> - Color Encoding, to replace drm_plane's COLOR_ENCODING
> - Color Range, to replace drm_plane's COLOR_RANGE
> 
> This patchset is grouped as follows:
>  - Patches 1-3: couple general patches/fixes
>  - Patches 4-7: introduce kunit to VKMS
>  - Patch 7: description of motivation and details behind the
>             Color Pipeline API. If you're reading nothing else
>             but are interested in the topic I highly recommend
>             you take a look at this.
>  - Patches 7-27: DRM core and VKMS changes for color pipeline API
>  - Patches 28-40: DRM core and amdgpu changes for color pipeline API
> 
> 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 here but
> haven't had a chance to look at. These will (hopefully) be
> addressed in future iterations, either in VKMS or amdgpu:
>  - Clear documentation for each drm_colorop_type
>  - Add custom LUT colorops to VKMS
>  - Add pre-blending 3DLUT
>  - How to support HW which can't bypass entire pipeline?
>  - Add ability to create colorops that don't have BYPASS
>  - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>  - read-only scaling colorop which defines scaling taps and position
>  - read-only color format colorop to define supported color formats
>    for a pipeline
>  - named matrices, for things like converting YUV to RGB
> 
> IGT tests can be found at
> https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/merge_requests/1
> 
> IGT patches are also being sent to the igt-dev mailing list.
> 
> If you prefer a gitlab MR for review you can find it at
> https://gitlab.freedesktop.org/hwentland/linux/-/merge_requests/5
> 
> 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 (10):
>   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
> 
> Harry Wentland (32):
>   drm: Don't treat 0 as -1 in drm_fixp2int_ceil
>   drm: Add helper for conversion from signed-magnitude
>   drm: Correctly round for fixp2int_round
>   drm/vkms: Round fixp2int conversion in lerp_u16
>   drm/vkms: Create separate Kconfig file for VKMS
>   drm/vkms: Add kunit tests for VKMS LUT handling
>   drm/vkms: Avoid reading beyond LUT array
>   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: Add NEXT to colorop state print
>   drm/vkms: Add enumerated 1D curve colorop
>   drm/vkms: Add kunit tests for linear and sRGB LUTs
>   drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
>   drm/colorop: Add 3x4 CTM type
>   drm/vkms: Pull apply_colorop out of pre_blend_color_transform
>   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
> 
>  Documentation/gpu/rfc/color_pipeline.rst      | 360 ++++++++
>  drivers/gpu/drm/Kconfig                       |  14 +-
>  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   | 352 ++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 160 ++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
>  drivers/gpu/drm/drm_atomic.c                  | 161 +++-
>  drivers/gpu/drm/drm_atomic_helper.c           |  12 +
>  drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
>  drivers/gpu/drm/drm_atomic_uapi.c             | 158 ++++
>  drivers/gpu/drm/drm_colorop.c                 | 411 +++++++++
>  drivers/gpu/drm/drm_ioctl.c                   |   7 +
>  drivers/gpu/drm/drm_mode_config.c             |   7 +
>  drivers/gpu/drm/drm_plane.c                   |  52 ++
>  drivers/gpu/drm/tests/Makefile                |   3 +-
>  drivers/gpu/drm/tests/drm_fixp_test.c         |  69 ++
>  drivers/gpu/drm/vkms/Kconfig                  |  20 +
>  drivers/gpu/drm/vkms/Makefile                 |   4 +-
>  drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>  drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 449 ++++++++++
>  drivers/gpu/drm/vkms/vkms_colorop.c           | 100 +++
>  drivers/gpu/drm/vkms/vkms_composer.c          | 135 ++-
>  drivers/gpu/drm/vkms/vkms_drv.h               |   8 +
>  drivers/gpu/drm/vkms/vkms_luts.c              | 802 ++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
>  drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
>  include/drm/drm_atomic.h                      | 122 +++
>  include/drm/drm_atomic_uapi.h                 |   3 +
>  include/drm/drm_colorop.h                     | 301 +++++++
>  include/drm/drm_file.h                        |   7 +
>  include/drm/drm_fixed.h                       |  35 +-
>  include/drm/drm_mode_config.h                 |  18 +
>  include/drm/drm_plane.h                       |  13 +
>  include/uapi/drm/drm.h                        |  16 +
>  include/uapi/drm/drm_mode.h                   |  14 +
>  38 files changed, 3882 insertions(+), 30 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/Kconfig
>  create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>  create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>  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
>  create mode 100644 include/drm/drm_colorop.h
> 
> --
> 2.44.0
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS
  2024-02-29 10:43 ` Daniel Vetter
@ 2024-02-29 15:31   ` Joshua Ashton
  0 siblings, 0 replies; 69+ messages in thread
From: Joshua Ashton @ 2024-02-29 15:31 UTC (permalink / raw)
  To: Daniel Vetter, Harry Wentland
  Cc: dri-devel, amd-gfx, wayland-devel, Ville Syrjala, Pekka Paalanen,
	Simon Ser, Melissa Wen, Jonas Ådahl, Sebastian Wick,
	Shashank Sharma, Alexander Goins, Michel Dänzer, Aleix Pol,
	Xaver Hugl, Victoria Brekenfeld, Uma Shankar, Naseer Ahmed,
	Christopher Braga, Abhinav Kumar, Arthur Grillo, Hector Martin,
	Liviu Dudau, Sasha McIntosh, Chaitanya Kumar Borah



On 2/29/24 10:43, Daniel Vetter wrote:
> On Mon, Feb 26, 2024 at 04:10:14PM -0500, Harry Wentland wrote:
>> This is an RFC set for a color pipeline API, along with a sample
>> implementation in VKMS. All the key API bits are here. VKMS now
>> supports two named transfer function colorops and two matrix
>> colorops. We have IGT tests that check all four of these colorops
>> with a pixel-by-pixel comparison that checks that these colorops
>> do what we expect them to do with a +/- 1 8 bpc code point margin.
> 
> So vkms is definitely great to make sure the igts are generic enough and
> somewhat useful, but ... does steam run on vkms too? I think that would be
> a really good test to show that the api we have here is actually useful
> for compositors in a cross-driver way, and not just a neat idea that
> doesn't survive harsh reality.
> 
> And yes I realize that's probably going to be a bunch of work, but I feel
> like the color pipeline discussion has dragged around enough in
> hypotheticals and concerns that I think it would really help a lot.

I don't think we have ever tested Steam/Gamescope on vkms.

The last time I tried stuff there, there was all the problems with the 
ttm page table tail thing for virtio stuff that made using Steam games + 
Gamescope unfeasable because Vulkan + bindless, but I have heard those 
are solved now?

I will have to try it again at the weekend and see where it's at.
I am willing to place my bets that some part of the stack will fall over 
relating to modifiers somehow... =P

But yes, testing there would be good too, as we have the full Steam Deck 
OLED HDR color pipeline implemented in shader-based composition 
validated as being 1:1 on a suite of HDR and SDR test images.
(That *will* definitely rely on 3D LUTs being tetrahedrally interpolated 
though)

I'll have a look at this at the weekend and also see about getting a 
Gamescope branch that uses the new wip colorop stuff.

- Joshie 🐸✨

> 
> Thoughts?
> -Sima
> 
>>
>> The big new change with v4 is the addition of an amdgpu color
>> pipeline, for all AMD GPUs with DCN 3 and newer. Amdgpu now support
>> the following:
>>
>> 1. 1D Curve EOTF
>> 2. 3x4 CTM
>> 3. Multiplier
>> 4. 1D Curve Inverse EOTF
>> 5. 1D LUT
>> 6. 1D Curve EOTF
>> 7. 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.
>>
>> We are working on two more ops for amdgpu, the HDR multiplier
>> and the 3DLUT, which will give us this:
>>
>> 1. 1D Curve EOTF
>> 2. 3x4 CTM
>> 3. HDR Multiplier
>> 4. 1D Curve Inverse EOTF
>> 5. 1D LUT
>> 6. 3D LUT
>> 7. 1D Curve EOTF
>> 8. 1D LUT
>>
>> This, essentially 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
>>
>> After this we'd like to also add the following ops:
>> - Scaler (Informational only)
>> - Color Encoding, to replace drm_plane's COLOR_ENCODING
>> - Color Range, to replace drm_plane's COLOR_RANGE
>>
>> This patchset is grouped as follows:
>>   - Patches 1-3: couple general patches/fixes
>>   - Patches 4-7: introduce kunit to VKMS
>>   - Patch 7: description of motivation and details behind the
>>              Color Pipeline API. If you're reading nothing else
>>              but are interested in the topic I highly recommend
>>              you take a look at this.
>>   - Patches 7-27: DRM core and VKMS changes for color pipeline API
>>   - Patches 28-40: DRM core and amdgpu changes for color pipeline API
>>
>> 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 here but
>> haven't had a chance to look at. These will (hopefully) be
>> addressed in future iterations, either in VKMS or amdgpu:
>>   - Clear documentation for each drm_colorop_type
>>   - Add custom LUT colorops to VKMS
>>   - Add pre-blending 3DLUT
>>   - How to support HW which can't bypass entire pipeline?
>>   - Add ability to create colorops that don't have BYPASS
>>   - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>>   - read-only scaling colorop which defines scaling taps and position
>>   - read-only color format colorop to define supported color formats
>>     for a pipeline
>>   - named matrices, for things like converting YUV to RGB
>>
>> IGT tests can be found at
>> https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/merge_requests/1
>>
>> IGT patches are also being sent to the igt-dev mailing list.
>>
>> If you prefer a gitlab MR for review you can find it at
>> https://gitlab.freedesktop.org/hwentland/linux/-/merge_requests/5
>>
>> 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 (10):
>>    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
>>
>> Harry Wentland (32):
>>    drm: Don't treat 0 as -1 in drm_fixp2int_ceil
>>    drm: Add helper for conversion from signed-magnitude
>>    drm: Correctly round for fixp2int_round
>>    drm/vkms: Round fixp2int conversion in lerp_u16
>>    drm/vkms: Create separate Kconfig file for VKMS
>>    drm/vkms: Add kunit tests for VKMS LUT handling
>>    drm/vkms: Avoid reading beyond LUT array
>>    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: Add NEXT to colorop state print
>>    drm/vkms: Add enumerated 1D curve colorop
>>    drm/vkms: Add kunit tests for linear and sRGB LUTs
>>    drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
>>    drm/colorop: Add 3x4 CTM type
>>    drm/vkms: Pull apply_colorop out of pre_blend_color_transform
>>    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
>>
>>   Documentation/gpu/rfc/color_pipeline.rst      | 360 ++++++++
>>   drivers/gpu/drm/Kconfig                       |  14 +-
>>   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   | 352 ++++++++
>>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 160 ++++
>>   .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
>>   .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
>>   drivers/gpu/drm/drm_atomic.c                  | 161 +++-
>>   drivers/gpu/drm/drm_atomic_helper.c           |  12 +
>>   drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
>>   drivers/gpu/drm/drm_atomic_uapi.c             | 158 ++++
>>   drivers/gpu/drm/drm_colorop.c                 | 411 +++++++++
>>   drivers/gpu/drm/drm_ioctl.c                   |   7 +
>>   drivers/gpu/drm/drm_mode_config.c             |   7 +
>>   drivers/gpu/drm/drm_plane.c                   |  52 ++
>>   drivers/gpu/drm/tests/Makefile                |   3 +-
>>   drivers/gpu/drm/tests/drm_fixp_test.c         |  69 ++
>>   drivers/gpu/drm/vkms/Kconfig                  |  20 +
>>   drivers/gpu/drm/vkms/Makefile                 |   4 +-
>>   drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>>   drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 449 ++++++++++
>>   drivers/gpu/drm/vkms/vkms_colorop.c           | 100 +++
>>   drivers/gpu/drm/vkms/vkms_composer.c          | 135 ++-
>>   drivers/gpu/drm/vkms/vkms_drv.h               |   8 +
>>   drivers/gpu/drm/vkms/vkms_luts.c              | 802 ++++++++++++++++++
>>   drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
>>   drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
>>   include/drm/drm_atomic.h                      | 122 +++
>>   include/drm/drm_atomic_uapi.h                 |   3 +
>>   include/drm/drm_colorop.h                     | 301 +++++++
>>   include/drm/drm_file.h                        |   7 +
>>   include/drm/drm_fixed.h                       |  35 +-
>>   include/drm/drm_mode_config.h                 |  18 +
>>   include/drm/drm_plane.h                       |  13 +
>>   include/uapi/drm/drm.h                        |  16 +
>>   include/uapi/drm/drm_mode.h                   |  14 +
>>   38 files changed, 3882 insertions(+), 30 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/Kconfig
>>   create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>>   create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>>   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
>>   create mode 100644 include/drm/drm_colorop.h
>>
>> --
>> 2.44.0
>>
> 


^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 03/42] drm: Correctly round for fixp2int_round
  2024-02-26 21:10 ` [RFC PATCH v4 03/42] drm: Correctly round for fixp2int_round Harry Wentland
@ 2024-03-11 13:11   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-11 13:11 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 2219 bytes --]

On Mon, 26 Feb 2024 16:10:17 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> A value of 0x80000000 and higher should round up, and
> below should round down. VKMS Kunit tests for lerp_u16
> showed that this is not the case. Fix it.
> 
> 1 << (DRM_FIXED_POINT_HALF - 1) =
> 1 << 15 = 0x8000
> 
> This is not 0.5, but 0.00000762939453125.
> 
> Instead of some smart math use a simple if/else to
> round up or down. This helps people like me to understand
> what the function does.
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  include/drm/drm_fixed.h | 15 ++++++++++-----
>  1 file changed, 10 insertions(+), 5 deletions(-)
> 
> diff --git a/include/drm/drm_fixed.h b/include/drm/drm_fixed.h
> index cb842ba80ddd..8ee549f68537 100644
> --- a/include/drm/drm_fixed.h
> +++ b/include/drm/drm_fixed.h
> @@ -77,6 +77,8 @@ static inline u32 dfixed_div(fixed20_12 A, fixed20_12 B)
>  #define DRM_FIXED_DIGITS_MASK	(~DRM_FIXED_DECIMAL_MASK)
>  #define DRM_FIXED_EPSILON	1LL
>  #define DRM_FIXED_ALMOST_ONE	(DRM_FIXED_ONE - DRM_FIXED_EPSILON)
> +#define DRM_FIXED_FRACTIONAL	0xffffffffll
> +#define DRM_FIXED_HALF		0x80000000ll
>  
>  /**
>   * @drm_sm2fixp
> @@ -106,11 +108,6 @@ static inline int drm_fixp2int(s64 a)
>  	return ((s64)a) >> DRM_FIXED_POINT;
>  }
>  
> -static inline int drm_fixp2int_round(s64 a)
> -{
> -	return drm_fixp2int(a + (1 << (DRM_FIXED_POINT_HALF - 1)));
> -}
> -
>  static inline int drm_fixp2int_ceil(s64 a)
>  {
>  	if (a >= 0)
> @@ -119,6 +116,14 @@ static inline int drm_fixp2int_ceil(s64 a)
>  		return drm_fixp2int(a - DRM_FIXED_ALMOST_ONE);
>  }
>  
> +static inline int drm_fixp2int_round(s64 a)
> +{
> +	if ((a & DRM_FIXED_FRACTIONAL) < DRM_FIXED_HALF)
> +		return drm_fixp2int(a);

So, if we take -epsilon (which is -1 in raw s64 value, the largest
possible value less than zero), this would return -1.0? That does not
sound right.

However, the "add 0.5 and truncate" trick always works, so I'd
recommend sticking that.


Thanks,
pq

> +	else
> +		return drm_fixp2int_ceil(a);
> +}
> +
>  static inline unsigned drm_fixp_msbset(s64 a)
>  {
>  	unsigned shift, sign = (a >> 63) & 1;


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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling
  2024-02-26 21:10 ` [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
  2024-02-27 12:14   ` Arthur Grillo
@ 2024-03-11 13:45   ` Pekka Paalanen
  1 sibling, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-11 13:45 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Arthur Grillo

[-- Attachment #1: Type: text/plain, Size: 8619 bytes --]

On Mon, 26 Feb 2024 16:10:20 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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
> 
> v4:
>  - Test the critical points of the lerp function (Pekka)
> 
> v3:
>  - Use include way of testing static functions (Arthur)
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> ---
>  drivers/gpu/drm/vkms/Kconfig                  |   5 +
>  drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>  drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 163 ++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_composer.c          |   8 +-
>  4 files changed, 178 insertions(+), 2 deletions(-)
>  create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>  create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> 

Hi Harry,

since fixed point math is hard, please allow me to nitpick so maybe
these patches could be landed ahead of time.

> diff --git a/drivers/gpu/drm/vkms/Kconfig b/drivers/gpu/drm/vkms/Kconfig
> index b9ecdebecb0b..c1f8b343ff0e 100644
> --- a/drivers/gpu/drm/vkms/Kconfig
> +++ b/drivers/gpu/drm/vkms/Kconfig
> @@ -13,3 +13,8 @@ config DRM_VKMS
>  	  a VKMS.
>  
>  	  If M is selected the module will be called vkms.
> +
> +config DRM_VKMS_KUNIT_TESTS
> +	tristate "Tests for VKMS" if !KUNIT_ALL_TESTS
> +	depends on DRM_VKMS && KUNIT
> +	default KUNIT_ALL_TESTS
> diff --git a/drivers/gpu/drm/vkms/tests/.kunitconfig b/drivers/gpu/drm/vkms/tests/.kunitconfig
> new file mode 100644
> index 000000000000..70e378228cbd
> --- /dev/null
> +++ b/drivers/gpu/drm/vkms/tests/.kunitconfig
> @@ -0,0 +1,4 @@
> +CONFIG_KUNIT=y
> +CONFIG_DRM=y
> +CONFIG_DRM_VKMS=y
> +CONFIG_DRM_VKMS_KUNIT_TESTS=y
> diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> new file mode 100644
> index 000000000000..fc73e48aa57c
> --- /dev/null
> +++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> @@ -0,0 +1,163 @@
> +/* SPDX-License-Identifier: GPL-2.0+ */
> +
> +#include <kunit/test.h>
> +
> +#include <drm/drm_fixed.h>
> +
> +#define TEST_LUT_SIZE 16
> +
> +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 },
> +};
> +
> +const struct vkms_color_lut test_linear_lut = {
> +	.base = test_linear_array,
> +	.lut_length = TEST_LUT_SIZE,
> +	.channel_value2index_ratio = 0xf000fll

Where does 0xf000f come from? Could it be computed from DRM_FIXED_ONE
and ARRAY_LENGTH()?

> +};
> +
> +
> +static void vkms_color_test_get_lut_index(struct kunit *test)
> +{
> +	int i;
> +
> +	KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&test_linear_lut, test_linear_array[0].red)), 0);
> +
> +	for (i = 0; i < TEST_LUT_SIZE; i++)
> +		KUNIT_EXPECT_EQ(test, drm_fixp2int_ceil(get_lut_index(&test_linear_lut, test_linear_array[i].red)), i);

Why this instead of

+	for (i = 0; i < TEST_LUT_SIZE; i++)
+		KUNIT_EXPECT_EQ(test, get_lut_index(&test_linear_lut, test_linear_array[i].red), drm_int2fixp(i));

and

+	for (i = 0; i < 0xffff; i++)
+		KUNIT_EXPECT_EQ(test, drm_fixp2int(get_lut_index(&test_linear_lut, i)), i / 0x1111);

?

I think your original form is leaving quite much room for error in
precision.

> +}
> +
> +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);

What about b = 0xfffd also? The smaller number in b, and especially in
b - a, the more precision error the rounding will forgive.

> +}
> +
> +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_LICENSE("GPL");
> \ No newline at end of file
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index e70cd473e3be..d178f2a400f6 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -91,7 +91,7 @@ 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)
> +u16 lerp_u16(u16 a, u16 b, s64 t)
>  {
>  	s64 a_fp = drm_int2fixp(a);
>  	s64 b_fp = drm_int2fixp(b);
> @@ -101,7 +101,7 @@ static u16 lerp_u16(u16 a, u16 b, s64 t)
>  	return drm_fixp2int_round(a_fp + delta);

Given the past, I guess it wouldn't hurt if all drm fixp functions had
a few tests of their own.


Thanks,
pq

>  }
>  
> -static s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
> +s64 get_lut_index(const struct vkms_color_lut *lut, u16 channel_value)
>  {
>  	s64 color_channel_fp = drm_int2fixp(channel_value);
>  
> @@ -429,3 +429,7 @@ int vkms_set_crc_source(struct drm_crtc *crtc, const char *src_name)
>  
>  	return ret;
>  }
> +
> +#ifdef CONFIG_DRM_VKMS_KUNIT_TESTS
> +#include "tests/vkms_color_tests.c"
> +#endif


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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 08/42] drm/doc/rfc: Describe why prescriptive color pipeline is needed
  2024-02-26 21:10 ` [RFC PATCH v4 08/42] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
@ 2024-03-11 15:11   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-11 15:11 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 18853 bytes --]

On Mon, 26 Feb 2024 16:10:22 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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)
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  Documentation/gpu/rfc/color_pipeline.rst | 360 +++++++++++++++++++++++
>  1 file changed, 360 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..6c653e17054a
> --- /dev/null
> +++ b/Documentation/gpu/rfc/color_pipeline.rst
> @@ -0,0 +1,360 @@
> +========================
> +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.
> +
> +
> +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, definint the type of transformation, such as

typo: definint

> +* 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. While setting other properties might fail atomic check,
> +setting the BYPASS property to true should never fail. 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

Since this is an immutable enum, does it really need to list all the
possible values it will never have?

Internally DRM would have the complete list, but KMS tends to not
expose enum values that cannot be picked.

> +    ├─ "BYPASS": bool {true, false}
> +    ├─ "CURVE_1D_TYPE": enum {sRGB EOTF, sRGB inverse EOTF, PQ EOTF, PQ inverse EOTF, …}

A better name for "sRGB EOTF" would be "power 2.2", referring to a
power function with exponent 2.2. I know these are just examples, but
maybe giving an example of something that looks obvious but won't
actually exist should be avoided.

> +    └─ "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}
> +    ├─ "LUT_1D_SIZE": immutable range = 4096
> +    ├─ "LUT_1D": 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}
> +    ├─ "LUT_3D_SIZE": immutable range = 17
> +    ├─ "LUT_3D": blob
> +    └─ "NEXT": immutable color operation ID = 73
> +
> +
> +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

maybe "all color processing is pass-through"?

"Disabled" could read as "nothing comes through".

> +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.

Yes!

Should there be rules about the string names for drivers? Can any
driver pick anything, even if it conflicts with a different kind of
pipeline?

> +
> +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.

I see it's called plane color pipeline, so I presume there will be
another cap for post-blending pipelines. Ok.

> +
> +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
> +    ├─ "MATRIX_3_4": 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

Current value is illegal. ;-)

> +    └─ "NEXT" (immutable) = 47
> +
> +    Color operation 47
> +    ├─ "TYPE" (immutable) = 1D LUT
> +    ├─ "LUT_1D_SIZE": immutable range = 4096
> +    ├─ "LUT_1D_DATA": blob
> +    └─ "NEXT" (immutable) = 48
> +
> +    Color operation 48
> +    ├─ "TYPE" (immutable) = 3D LUT

I see these LUT things do not have BYPASS. They probably should, even
if it is equivalent to setting the data blob to 0.

> +    ├─ "LUT_3D_SIZE" (immutable) = 17
> +    ├─ "LUT_3D_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
> +    ├─ "LUT_3D_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.

...colorops in additional pipelines.

Especially if exposing hw-specific blocks together with generic blocks
in the same pipeline would result in the possibility of unsupported
combinations.

Maybe there should be a general rule, that it must be possible to
enable all colorops in the same pipeline simultaneously?

> +  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

All in all, I think this is good enough to have my

Acked-by: Pekka Paalanen <pekka.paalanen@collabora.com>

in spite of the comments I had. They are just fine tuning.


Thanks,
pq

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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 10/42] drm/colorop: Add TYPE property
  2024-02-26 21:10 ` [RFC PATCH v4 10/42] drm/colorop: Add TYPE property Harry Wentland
@ 2024-03-12 15:02   ` Pekka Paalanen
  2024-03-12 15:15     ` Simon Ser
  0 siblings, 1 reply; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-12 15:02 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 2770 bytes --]

On Mon, 26 Feb 2024 16:10:24 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> Add a read-only TYPE property. The TYPE specifies the colorop
> type, such as enumerated curve, 1D LUT, CTM, 3D LUT, PWL LUT,
> etc.
> 
> 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
> 
> For now we're only introducing an enumerated 1D LUT type to
> illustrate the concept.
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  drivers/gpu/drm/drm_atomic.c      |  4 +--
>  drivers/gpu/drm/drm_atomic_uapi.c |  8 +++++-
>  drivers/gpu/drm/drm_colorop.c     | 44 ++++++++++++++++++++++++++++++-
>  include/drm/drm_colorop.h         | 17 +++++++++++-
>  include/uapi/drm/drm_mode.h       |  4 +++
>  5 files changed, 72 insertions(+), 5 deletions(-)

...

> diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
> index a295ab96aee1..3a07a8fe2842 100644
> --- a/drivers/gpu/drm/drm_colorop.c
> +++ b/drivers/gpu/drm/drm_colorop.c
> @@ -32,12 +32,17 @@
>  
>  /* TODO big colorop doc, including properties, etc. */
>  
> +static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
> +	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
> +};
> +
>  /* Init Helpers */
>  
>  int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
> -		     struct drm_plane *plane)
> +		     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);
> @@ -46,12 +51,29 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
>  
>  	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));

This list here is the list of all values the enum could take, right?
Should it not be just the one value it's going to have? It'll never
change, and it can never be changed.

> +
> +	if (!prop)
> +		return -ENOMEM;
> +
> +	colorop->type_property = prop;
> +
> +	drm_object_attach_property(&colorop->base,
> +				   colorop->type_property,
> +				   colorop->type);
> +
>  	return ret;
>  }
>  EXPORT_SYMBOL(drm_colorop_init);

Thanks,
pq

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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 10/42] drm/colorop: Add TYPE property
  2024-03-12 15:02   ` Pekka Paalanen
@ 2024-03-12 15:15     ` Simon Ser
  2024-03-12 15:55       ` Pekka Paalanen
  0 siblings, 1 reply; 69+ messages in thread
From: Simon Ser @ 2024-03-12 15:15 UTC (permalink / raw)
  To: Pekka Paalanen; +Cc: Harry Wentland, dri-devel, amd-gfx, wayland-devel

On Tuesday, March 12th, 2024 at 16:02, Pekka Paalanen <pekka.paalanen@collabora.com> wrote:

> This list here is the list of all values the enum could take, right?
> Should it not be just the one value it's going to have? It'll never
> change, and it can never be changed.

Listing all possible values is how other properties behave. Example:

    "type" (immutable): enum {Overlay, Primary, Cursor} = Primary

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 17/42] drm/vkms: Add enumerated 1D curve colorop
  2024-02-26 21:10 ` [RFC PATCH v4 17/42] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
@ 2024-03-12 15:27   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-12 15:27 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Alex Hung

[-- Attachment #1: Type: text/plain, Size: 5288 bytes --]

On Mon, 26 Feb 2024 16:10:31 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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.
> 
> 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
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> ---
>  drivers/gpu/drm/vkms/Makefile        |   4 +-
>  drivers/gpu/drm/vkms/vkms_colorop.c  |  70 +++
>  drivers/gpu/drm/vkms/vkms_composer.c |  45 ++
>  drivers/gpu/drm/vkms/vkms_drv.h      |   4 +
>  drivers/gpu/drm/vkms/vkms_luts.c     | 802 +++++++++++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_luts.h     |  12 +
>  drivers/gpu/drm/vkms/vkms_plane.c    |   2 +
>  7 files changed, 938 insertions(+), 1 deletion(-)
>  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/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index b90e446d5954..9493bdb1ba3f 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -12,6 +12,7 @@
>  #include <linux/minmax.h>
>  
>  #include "vkms_drv.h"
> +#include "vkms_luts.h"
>  
>  static u16 pre_mul_blend_channel(u16 src, u16 dst, u16 alpha)
>  {
> @@ -163,6 +164,47 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
>  	}
>  }
>  
> +static void pre_blend_color_transform(const struct vkms_plane_state *plane_state, struct line_buffer *output_buffer)
> +{
> +	struct drm_colorop *colorop = plane_state->base.base.color_pipeline;
> +
> +	while (colorop) {

I think this would be easier to read if you used

	for (; colorop; colorop = colorop->next) {

and

> +		struct drm_colorop_state *colorop_state;
> +
> +		if (!colorop)
> +			return;
> +
> +		/* TODO this is probably wrong */
> +		colorop_state = colorop->state;
> +
> +		if (!colorop_state)
> +			return;

	if (colorop_state->bypass)
		continue;

Something about 'switch (colorop->type)' to pick a function pointer to
call, but hard to see at this point of the series how that would work.

However, you can pick between srgb_inv_eotf and srgb_eotf already here.
Then inside the loop you can just call one set of
apply_lut_to_channel_value() and not need conditionals and avoid
indentation levels.

> +
> +		for (size_t x = 0; x < output_buffer->n_pixels; x++) {
> +			struct pixel_argb_u16 *pixel = &output_buffer->pixels[x];
> +
> +			if (colorop->type == DRM_COLOROP_1D_CURVE &&
> +				colorop_state->bypass == false) {
> +				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:
> +					default:
> +						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;
> +				}
> +			}

else { aaargh_unknown_colorop(); }

> +		}
> +
> +		colorop = colorop->next;
> +	}
> +}

...

> diff --git a/drivers/gpu/drm/vkms/vkms_luts.c b/drivers/gpu/drm/vkms/vkms_luts.c
> new file mode 100644
> index 000000000000..6553d6d442b4
> --- /dev/null
> +++ b/drivers/gpu/drm/vkms/vkms_luts.c
> @@ -0,0 +1,802 @@
> +/* SPDX-License-Identifier: GPL-2.0+ */
> +
> +#include <drm/drm_mode.h>
> +
> +#include "vkms_drv.h"
> +#include "vkms_luts.h"
> +

Here it would be really nice to explain how the tables were generated.

> +static struct drm_color_lut linear_array[LUT_SIZE] = {
> +	{ 0x0, 0x0, 0x0, 0 },

...

> +	{ 0xffff, 0xffff, 0xffff, 0 },
> +};
> +
> +const struct vkms_color_lut linear_eotf = {
> +	.base = linear_array,
> +	.lut_length = LUT_SIZE,

Why not use just 2 table entries for the linear array?

I didn't see linear_eotf used at all? It could also just skip in the
code, not need an array.

> +	.channel_value2index_ratio = 0xff00ffll
> +};


Thanks,
pq

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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 22/42] drm/vkms: Use s32 for internal color pipeline precision
  2024-02-26 21:10 ` [RFC PATCH v4 22/42] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
@ 2024-03-12 15:50   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-12 15:50 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 5415 bytes --]

On Mon, 26 Feb 2024 16:10:36 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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.

I always go through the same thought process:

- putting the pixel iteration loop outermost is going to be horrible
  for performance

- maybe should turn the temporary line buffer to argb_s32 for all of
  VKMS

- that's going to be a lot of memory traffic... maybe your way is not
  horrible for performance

- maybe processing pixel by pixel is good, if you can prepare the
  operation in advance, so you don't need to analyze colorops again and
  again on each pixel

- eh, maybe it's not a gain after all, needs benchmarking

My comment on patch 17 was like the step 0 of this train of thought. I
just always get the same comments in my mind when seeing the same code
again.


Thanks,
pq

> 
> 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
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  drivers/gpu/drm/vkms/vkms_composer.c | 57 +++++++++++++++++++++-------
>  drivers/gpu/drm/vkms/vkms_drv.h      |  4 ++
>  2 files changed, 48 insertions(+), 13 deletions(-)
> 
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index 25b786b49db0..d2101fa55aa3 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -164,7 +164,7 @@ 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)
>  {
>  	/* TODO is this right? */
>  	struct drm_colorop_state *colorop_state = colorop->state;
> @@ -191,25 +191,56 @@ 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 drm_colorop *colorop = plane_state->base.base.color_pipeline;
> +	struct drm_colorop *colorop;
> +	struct pixel_argb_s32 pixel;
>  
> -	while (colorop) {
> -		struct drm_colorop_state *colorop_state;
> +	for (size_t x = 0; x < output_buffer->n_pixels; x++) {
>  
> -		if (!colorop)
> -			return;
> +		/*
> +		 * 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;
>  
> -		/* TODO this is probably wrong */
> -		colorop_state = colorop->state;
> +		colorop = plane_state->base.base.color_pipeline;
> +		while (colorop) {
> +			struct drm_colorop_state *colorop_state;
>  
> -		if (!colorop_state)
> -			return;
> +			if (!colorop)
> +				return;
> +
> +			/* TODO this is probably wrong */
> +			colorop_state = colorop->state;
> +
> +			if (!colorop_state)
> +				return;
>  
> -		for (size_t x = 0; x < output_buffer->n_pixels; x++)
>  			if (!colorop_state->bypass)
> -				apply_colorop(&output_buffer->pixels[x], colorop);
> +				apply_colorop(&pixel, colorop);
>  
> -		colorop = colorop->next;
> +			colorop = colorop->next;
> +		}
> +
> +		/* clamp pixel */
> +		pixel.a = max(min(pixel.a, 0xffff), 0x0);
> +		pixel.r = max(min(pixel.r, 0xffff), 0x0);
> +		pixel.g = max(min(pixel.g, 0xffff), 0x0);
> +		pixel.b = max(min(pixel.b, 0xffff), 0x0);
> +
> +		/* put back to output_buffer */
> +		output_buffer->pixels[x].a = pixel.a;
> +		output_buffer->pixels[x].r = pixel.r;
> +		output_buffer->pixels[x].g = pixel.g;
> +		output_buffer->pixels[x].b = pixel.b;
>  	}
>  }
>  
> diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
> index 2bcc24c196a2..fadb7685a360 100644
> --- a/drivers/gpu/drm/vkms/vkms_drv.h
> +++ b/drivers/gpu/drm/vkms/vkms_drv.h
> @@ -36,6 +36,10 @@ struct vkms_frame_info {
>  	unsigned int cpp;
>  };
>  
> +struct pixel_argb_s32 {
> +	s32 a, r, g, b;
> +};
> +
>  struct pixel_argb_u16 {
>  	u16 a, r, g, b;
>  };


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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 10/42] drm/colorop: Add TYPE property
  2024-03-12 15:15     ` Simon Ser
@ 2024-03-12 15:55       ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-12 15:55 UTC (permalink / raw)
  To: Simon Ser; +Cc: Harry Wentland, dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 941 bytes --]

On Tue, 12 Mar 2024 15:15:13 +0000
Simon Ser <contact@emersion.fr> wrote:

> On Tuesday, March 12th, 2024 at 16:02, Pekka Paalanen <pekka.paalanen@collabora.com> wrote:
> 
> > This list here is the list of all values the enum could take, right?
> > Should it not be just the one value it's going to have? It'll never
> > change, and it can never be changed.  
> 
> Listing all possible values is how other properties behave. Example:
> 
>     "type" (immutable): enum {Overlay, Primary, Cursor} = Primary

Yeah, that's weird, now that you pointed it out.

Userspace does nothing with the knowledge of what this specific
object's property "could" be since it's immutable. Only the kernel
internals and UAPI docs need the full list, and userspace needs to
hardcode the full list in general so it can recognize each value. But
on the property instance on a specific object, I see no use for the
full list.


Thanks,
pq

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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 05/42] drm/vkms: Create separate Kconfig file for VKMS
  2024-02-26 21:10 ` [RFC PATCH v4 05/42] drm/vkms: Create separate Kconfig file for VKMS Harry Wentland
@ 2024-03-12 18:49   ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-03-12 18:49 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Simon Ser

On 02/26, Harry Wentland wrote:
> This aligns with most other DRM drivers and will allow
> us to add new VKMS config options without polluting
> the DRM Kconfig.
> 
> v3:
>  - Change SPDX to GPL-2.0-only to match DRM KConfig
>    SPDX (Simon)
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Reviewed-by: Simon Ser <contact@emersion.fr>

Just to let you know this patch is already upstream. I cherry-picked it
from the last patchset version and applied to drm-misc-next:
https://cgit.freedesktop.org/drm/drm-misc/commit/drivers/gpu/drm?id=ffcc67cd79ff2e93fd0bdb837c99cbab6c59d38c

Thanks again,

Melissa

> ---
>  drivers/gpu/drm/Kconfig      | 14 +-------------
>  drivers/gpu/drm/vkms/Kconfig | 15 +++++++++++++++
>  2 files changed, 16 insertions(+), 13 deletions(-)
>  create mode 100644 drivers/gpu/drm/vkms/Kconfig
> 
> diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
> index 2520db0b776e..e3ea8793cb8a 100644
> --- a/drivers/gpu/drm/Kconfig
> +++ b/drivers/gpu/drm/Kconfig
> @@ -289,19 +289,7 @@ config DRM_VGEM
>  	  as used by Mesa's software renderer for enhanced performance.
>  	  If M is selected the module will be called vgem.
>  
> -config DRM_VKMS
> -	tristate "Virtual KMS (EXPERIMENTAL)"
> -	depends on DRM && MMU
> -	select DRM_KMS_HELPER
> -	select DRM_GEM_SHMEM_HELPER
> -	select CRC32
> -	default n
> -	help
> -	  Virtual Kernel Mode-Setting (VKMS) is used for testing or for
> -	  running GPU in a headless machines. Choose this option to get
> -	  a VKMS.
> -
> -	  If M is selected the module will be called vkms.
> +source "drivers/gpu/drm/vkms/Kconfig"
>  
>  source "drivers/gpu/drm/exynos/Kconfig"
>  
> diff --git a/drivers/gpu/drm/vkms/Kconfig b/drivers/gpu/drm/vkms/Kconfig
> new file mode 100644
> index 000000000000..b9ecdebecb0b
> --- /dev/null
> +++ b/drivers/gpu/drm/vkms/Kconfig
> @@ -0,0 +1,15 @@
> +# SPDX-License-Identifier: GPL-2.0-only
> +
> +config DRM_VKMS
> +	tristate "Virtual KMS (EXPERIMENTAL)"
> +	depends on DRM && MMU
> +	select DRM_KMS_HELPER
> +	select DRM_GEM_SHMEM_HELPER
> +	select CRC32
> +	default n
> +	help
> +	  Virtual Kernel Mode-Setting (VKMS) is used for testing or for
> +	  running GPU in a headless machines. Choose this option to get
> +	  a VKMS.
> +
> +	  If M is selected the module will be called vkms.
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 07/42] drm/vkms: Avoid reading beyond LUT array
  2024-02-26 21:10 ` [RFC PATCH v4 07/42] drm/vkms: Avoid reading beyond LUT array Harry Wentland
@ 2024-03-12 18:54   ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-03-12 18:54 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Arthur Grillo

On 02/26, Harry Wentland wrote:
> When the floor LUT index (drm_fixp2int(lut_index) is the last
> index of the array the ceil LUT index will point to an entry
> beyond the array. Make sure we guard against it and use the
> value of the floor LUT index.
> 
> v3:
>  - Drop bits from commit description that didn't contribute
>    anything of value
> 
> Fixes: db1f254f2cfaf ("drm/vkms: Add support to 1D gamma LUT")
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> Reviewed-by: Melissa Wen <mwen@igalia.com>

Same. Already applied upstream:
https://cgit.freedesktop.org/drm/drm-misc/commit/drivers/gpu/drm?id=2fee84030d12d9fddfa874e4562d71761a129277

I guess you are working on top of asdn branch and it's okay to me.
I'm just mentioning to avoid confusion.

Melissa

> ---
>  drivers/gpu/drm/vkms/vkms_composer.c | 14 ++++++++++----
>  1 file changed, 10 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index d178f2a400f6..b90e446d5954 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -123,6 +123,8 @@ static u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 chan
>  				      enum lut_channel channel)
>  {
>  	s64 lut_index = get_lut_index(lut, channel_value);
> +	u16 *floor_lut_value, *ceil_lut_value;
> +	u16 floor_channel_value, ceil_channel_value;
>  
>  	/*
>  	 * This checks if `struct drm_color_lut` has any gap added by the compiler
> @@ -130,11 +132,15 @@ static u16 apply_lut_to_channel_value(const struct vkms_color_lut *lut, u16 chan
>  	 */
>  	static_assert(sizeof(struct drm_color_lut) == sizeof(__u16) * 4);
>  
> -	u16 *floor_lut_value = (__u16 *)&lut->base[drm_fixp2int(lut_index)];
> -	u16 *ceil_lut_value = (__u16 *)&lut->base[drm_fixp2int_ceil(lut_index)];
> +	floor_lut_value = (__u16 *)&lut->base[drm_fixp2int(lut_index)];
> +	if (drm_fixp2int(lut_index) == (lut->lut_length - 1))
> +		/* We're at the end of the LUT array, use same value for ceil and floor */
> +		ceil_lut_value = floor_lut_value;
> +	else
> +		ceil_lut_value = (__u16 *)&lut->base[drm_fixp2int_ceil(lut_index)];
>  
> -	u16 floor_channel_value = floor_lut_value[channel];
> -	u16 ceil_channel_value = ceil_lut_value[channel];
> +	floor_channel_value = floor_lut_value[channel];
> +	ceil_channel_value = ceil_lut_value[channel];
>  
>  	return lerp_u16(floor_channel_value, ceil_channel_value,
>  			lut_index & DRM_FIXED_DECIMAL_MASK);
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 02/42] drm: Add helper for conversion from signed-magnitude
  2024-02-26 21:10 ` [RFC PATCH v4 02/42] drm: Add helper for conversion from signed-magnitude Harry Wentland
@ 2024-03-14 13:16   ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-03-14 13:16 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

On 02/26, Harry Wentland wrote:
> 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.
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.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 0c9f917a4d4b..cb842ba80ddd 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);
Hi Harry,

Can we have a #define macro for this constant? ^
Other than that, LGTM. You can add my r-b to the next version.

Thanks,

Melissa
> +	} else {
> +		return a;
> +	}
> +
> +}
> +
>  static inline s64 drm_int2fixp(int a)
>  {
>  	return ((s64)a) << DRM_FIXED_POINT;
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil
  2024-02-26 21:10 ` [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil Harry Wentland
@ 2024-03-14 13:41   ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-03-14 13:41 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Simon Ser

On 02/26, Harry Wentland wrote:
> Unit testing this in VKMS shows that passing 0 into
> this function returns -1, which is highly counter-
> intuitive. Fix it by checking whether the input is
> >= 0 instead of > 0.
> 
> Fixes: 64566b5e767f9 ("drm: Add drm_fixp_from_fraction and drm_fixp2int_ceil")
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Reviewed-by: Simon Ser <contact@emersion.fr>
> Reviewed-by: Melissa Wen <mwen@igalia.com>

It was already applied upstream:
https://cgit.freedesktop.org/drm/drm-misc/commit/include?id=cf8837d7204481026335461629b84ac7f4538fa5

Thanks

Melissa
> ---
>  include/drm/drm_fixed.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/drm/drm_fixed.h b/include/drm/drm_fixed.h
> index 6ea339d5de08..0c9f917a4d4b 100644
> --- a/include/drm/drm_fixed.h
> +++ b/include/drm/drm_fixed.h
> @@ -95,7 +95,7 @@ static inline int drm_fixp2int_round(s64 a)
>  
>  static inline int drm_fixp2int_ceil(s64 a)
>  {
> -	if (a > 0)
> +	if (a >= 0)
>  		return drm_fixp2int(a + DRM_FIXED_ALMOST_ONE);
>  	else
>  		return drm_fixp2int(a - DRM_FIXED_ALMOST_ONE);
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 23/42] drm/vkms: add 3x4 matrix in color pipeline
  2024-02-26 21:10 ` [RFC PATCH v4 23/42] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
@ 2024-03-14 15:23   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-14 15:23 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 5242 bytes --]

On Mon, 26 Feb 2024 16:10:37 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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.

You will test also cases where the matrix configuration will not
produce output equal to input, right?

Otherwise one can accidentally pass the matrix test by ignoring all
matrices.

> 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.
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  drivers/gpu/drm/vkms/vkms_colorop.c  | 32 +++++++++++++++++++++++++++-
>  drivers/gpu/drm/vkms/vkms_composer.c | 27 +++++++++++++++++++++++
>  include/drm/drm_colorop.h            |  2 ++
>  3 files changed, 60 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
> index d2db366da6d3..a0e54b2c1f7a 100644
> --- a/drivers/gpu/drm/vkms/vkms_colorop.c
> +++ b/drivers/gpu/drm/vkms/vkms_colorop.c
> @@ -35,7 +35,37 @@ const int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pro
>  
>  	prev_op = op;
>  
> -	/* 2nd op: 1d curve */
> +	/* 2nd op: 3x4 matrix */
> +	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> +	if (!op) {
> +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> +		return -ENOMEM;
> +	}
> +
> +	ret = drm_colorop_ctm_3x4_init(dev, op, plane);
> +	if (ret)
> +		return ret;
> +
> +	drm_colorop_set_next_property(prev_op, op);
> +
> +	prev_op = op;
> +
> +	/* 3rd op: 3x4 matrix */
> +	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> +	if (!op) {
> +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> +		return -ENOMEM;
> +	}
> +
> +	ret = drm_colorop_ctm_3x4_init(dev, op, plane);
> +	if (ret)
> +		return ret;
> +
> +	drm_colorop_set_next_property(prev_op, op);
> +
> +	prev_op = op;
> +
> +	/* 4th op: 1d curve */
>  	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
>  	if (!op) {
>  		DRM_ERROR("KMS: Failed to allocate colorop\n");
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index d2101fa55aa3..8bbfce651526 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -164,6 +164,30 @@ 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;
> +
> +	rf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[0]), drm_int2fixp(pixel->r)) +
> +	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[1]), drm_int2fixp(pixel->g)) +
> +	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[2]), drm_int2fixp(pixel->b)) +
> +	     drm_sm2fixp(matrix->matrix[3]);

It would be nice if the driver had its private data for the matrix
colorop state, so it could convert the matrix only once when userspace
sets it.


Thanks,
pq

> +
> +	gf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[4]), drm_int2fixp(pixel->r)) +
> +	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[5]), drm_int2fixp(pixel->g)) +
> +	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[6]), drm_int2fixp(pixel->b)) +
> +	     drm_sm2fixp(matrix->matrix[7]);
> +
> +	bf = drm_fixp_mul(drm_sm2fixp(matrix->matrix[8]), drm_int2fixp(pixel->r)) +
> +	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[9]), drm_int2fixp(pixel->g)) +
> +	     drm_fixp_mul(drm_sm2fixp(matrix->matrix[10]), drm_int2fixp(pixel->b)) +
> +	     drm_sm2fixp(matrix->matrix[11]);
> +
> +	pixel->r = drm_fixp2int(rf);
> +	pixel->g = drm_fixp2int(gf);
> +	pixel->b = drm_fixp2int(bf);
> +}
> +
>  static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
>  {
>  	/* TODO is this right? */
> @@ -185,6 +209,9 @@ static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colo
>  				DRM_DEBUG_DRIVER("unkown colorop 1D curve type %d\n", 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);
>  	}
>  
>  }
> diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
> index 4aee29e161d6..8710e550790c 100644
> --- a/include/drm/drm_colorop.h
> +++ b/include/drm/drm_colorop.h
> @@ -224,6 +224,8 @@ int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
>  
>  int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
>  			      struct drm_plane *plane, u64 supported_tfs);
> +int drm_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);


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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 24/42] drm/tests: Add a few tests around drm_fixed.h
  2024-02-26 21:10 ` [RFC PATCH v4 24/42] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
@ 2024-03-14 15:36   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-14 15:36 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 4204 bytes --]

On Mon, 26 Feb 2024 16:10:38 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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.
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  drivers/gpu/drm/tests/Makefile        |  3 +-
>  drivers/gpu/drm/tests/drm_fixp_test.c | 69 +++++++++++++++++++++++++++
>  2 files changed, 71 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 d6183b3d7688..98468f7908dd 100644
> --- a/drivers/gpu/drm/tests/Makefile
> +++ b/drivers/gpu/drm/tests/Makefile
> @@ -19,6 +19,7 @@ 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..f420f173ff66
> --- /dev/null
> +++ b/drivers/gpu/drm/tests/drm_fixp_test.c
> @@ -0,0 +1,69 @@
> +// 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, ((1LL << 63) - 1));
> +
> +	/* 1 */
> +	KUNIT_EXPECT_EQ(test, drm_int2fixp(1), drm_sm2fixp(1ull << DRM_FIXED_POINT));

This sounds like confusing two different APIs.

DRM_FIXED_POINT is for the two's complement representation, and sm is
not.

It does not look like there is a #define for the signed-magnitude UAPI
construct, otherwise drm_color_ctm_s31_32_to_qm_n() would be using it.

It's just an accident that DRM_FIXED_POINT has the right value for sm.

Hm, drm_color_ctm_s31_32_to_qm_n() produces two's complement. Why not
use that for implementing drm_sm2fixp()?


Thanks,
pq

> +
> +	/* -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");
> \ No newline at end of file


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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 25/42] drm/vkms: Add tests for CTM handling
  2024-02-26 21:10 ` [RFC PATCH v4 25/42] drm/vkms: Add tests for CTM handling Harry Wentland
@ 2024-03-14 16:02   ` Pekka Paalanen
  0 siblings, 0 replies; 69+ messages in thread
From: Pekka Paalanen @ 2024-03-14 16:02 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

[-- Attachment #1: Type: text/plain, Size: 9376 bytes --]

On Mon, 26 Feb 2024 16:10:39 -0500
Harry Wentland <harry.wentland@amd.com> wrote:

> 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.
> 
> 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)
>  - Rename red, green, blue to Y, U, V where applicable
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 251 ++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_composer.c          |   2 +-
>  2 files changed, 252 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> index e6ac01dee830..83d07f7bae37 100644
> --- a/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> +++ b/drivers/gpu/drm/vkms/tests/vkms_color_tests.c
> @@ -3,6 +3,7 @@
>  #include <kunit/test.h>
>  
>  #include <drm/drm_fixed.h>
> +#include <drm/drm_mode.h>
>  
>  #define TEST_LUT_SIZE 16
>  
> @@ -181,11 +182,261 @@ static void vkms_color_srgb_inv_srgb(struct kunit *test)
>  	}
>  }
>  
> +#define FIXPT_HALF        (DRM_FIXED_ONE >> 1)
> +#define FIXPT_QUARTER     (DRM_FIXED_ONE >> 2)
> +
> +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

These are supposed to be sign-magnitude, not fixed-point, to my
understanding. It just happens that these specific values have the same
bit pattern in both representations.


> +} };
> +
> +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 = 0x7fff;
> +	ref.g = 0x3fff;
> +	ref.b = 0x3fff;
> +
> +	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.

Shouldn't that be sign-magnitude, not fixed point?

I was hoping to get a reference to a spec like BT.709 and a formula for
getting the blow out of it. IGT can change over time, and I don't know
where IGT for that matrix from.

But ok, this is a test with an arbitrary matrix, not necessarily the
BT.709 matrix specifically.

Btw. which BT.709 matrix is this? YUV->RGB, RGB->XYZ, or the inverse of
one of those? Limited or full quantization range?

Judging by the tests, I guess RGB->YUV full range.

> + */
> +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);

For all of these U/V too, you may want to enforce a range. The value
must be approximately 0x100. Something like 0x17 would very wrong.

Hmm, wait...

Neutral color should have U and V neutral, that is, zero, right? So
what is zero in this integer representation?

You don't seem to be testing negative U or V...


Thanks,
pq

> +
> +	/* 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 8bbfce651526..2c5715242f91 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -164,7 +164,7 @@ 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)
> +void apply_3x4_matrix(struct pixel_argb_s32 *pixel, const struct drm_color_ctm_3x4 *matrix)
>  {
>  	s64 rf, gf, bf;
>  


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

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 09/42] drm/colorop: Introduce new drm_colorop mode object
  2024-02-26 21:10 ` [RFC PATCH v4 09/42] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
@ 2024-05-21 15:28   ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-05-21 15:28 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel

On 02/26, Harry Wentland wrote:
> 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.
> 
> 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)
> 
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  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       | 152 ++++++++++++++++++++++++++
>  drivers/gpu/drm/drm_mode_config.c   |   7 ++
>  include/drm/drm_atomic.h            |  82 +++++++++++++++
>  include/drm/drm_atomic_uapi.h       |   1 +
>  include/drm/drm_colorop.h           | 158 ++++++++++++++++++++++++++++
>  include/drm/drm_mode_config.h       |  18 ++++
>  include/drm/drm_plane.h             |   2 +
>  include/uapi/drm/drm_mode.h         |   1 +
>  12 files changed, 552 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 104b42df2e95..4b14dcbb6117 100644
> --- a/drivers/gpu/drm/Makefile
> +++ b/drivers/gpu/drm/Makefile
> @@ -16,6 +16,7 @@ drm-y := \
>  	drm_client.o \
>  	drm_client_modeset.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 a91737adf8e7..62e87e6a9653 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"
> @@ -108,6 +109,7 @@ 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);
> @@ -139,6 +141,10 @@ 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
> @@ -250,6 +256,20 @@ 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;
>  
> @@ -571,6 +591,56 @@ 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_existing_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 39ef0a6addeb..282aa1f7b9a6 100644
> --- a/drivers/gpu/drm/drm_atomic_helper.c
> +++ b/drivers/gpu/drm/drm_atomic_helper.c
> @@ -3022,6 +3022,8 @@ 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;
> @@ -3099,6 +3101,16 @@ 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;
> +	}
> +
>  	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 29d4940188d4..1f9b6dfa8ca7 100644
> --- a/drivers/gpu/drm/drm_atomic_uapi.c
> +++ b/drivers/gpu/drm/drm_atomic_uapi.c
> @@ -34,6 +34,7 @@
>  #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>
> @@ -642,6 +643,26 @@ 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)
> @@ -904,6 +925,16 @@ 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;
> @@ -1084,6 +1115,23 @@ 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..a295ab96aee1
> --- /dev/null
> +++ b/drivers/gpu/drm/drm_colorop.c
> @@ -0,0 +1,152 @@
> +/*
> + * 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"
> +
> +/* TODO big colorop doc, including properties, etc. */
> +
> +/* Init Helpers */
> +
> +int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
> +		     struct drm_plane *plane)
> +{
> +	struct drm_mode_config *config = &dev->mode_config;
> +	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->plane = plane;
> +
> +	list_add_tail(&colorop->head, &config->colorop_list);
> +	colorop->index = config->num_colorop++;
> +
> +	/* add properties */
> +	return ret;
> +}
> +EXPORT_SYMBOL(drm_colorop_init);
> +
> +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_destroy_state - default state destroy hook
> + * @colorop: drm colorop
> + * @state: colorop state object to release
> + *
> + * Default colorop state destroy hook for drivers which don't have their own
> + * subclassed colorop state structure.
> + */
> +void drm_colorop_destroy_state(struct drm_colorop *colorop,
> +			       struct drm_colorop_state *state)
> +{
> +	kfree(state);
> +}
> +EXPORT_SYMBOL(drm_colorop_destroy_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;
> +}
> +
> +/**
> + * 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)
> +{
> +	kfree(colorop->state);
> +	colorop->state = kzalloc(sizeof(*colorop->state), GFP_KERNEL);
> +
> +	if (colorop->state)
> +		__drm_colorop_reset(colorop, colorop->state);
> +}
> +EXPORT_SYMBOL(drm_colorop_reset);
> diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
> index 8525ef851540..30c6fb10353b 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"
> @@ -182,11 +183,15 @@ 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);
> @@ -413,6 +418,7 @@ 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);
> @@ -434,6 +440,7 @@ 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 cf8e1220a4ac..634b2827765f 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;
> @@ -398,6 +404,7 @@ struct drm_atomic_state {
>  	 * states.
>  	 */
>  	bool duplicated : 1;
> +	struct __drm_colorops_state *colorops;
>  	struct __drm_planes_state *planes;
>  	struct __drm_crtcs_state *crtcs;
>  	int num_connector;
> @@ -501,6 +508,9 @@ 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);
> @@ -630,6 +640,55 @@ drm_atomic_get_new_plane_state(const struct drm_atomic_state *state,
>  	return state->planes[drm_plane_index(plane)].new_state;
>  }
>  
> +
> +/**
> + * drm_atomic_get_existing_colorop_state - get colorop state, if it exists
> + * @state: global atomic state object
> + * @colorop: colorop to grab
> + *
> + * This function returns the colorop state for the given colorop, or NULL
> + * if the colorop is not part of the global atomic state.
> + *
> + * This function is deprecated, @drm_atomic_get_old_colorop_state or
> + * @drm_atomic_get_new_colorop_state should be used instead.
> + */
> +static inline struct drm_colorop_state *
> +drm_atomic_get_existing_colorop_state(struct drm_atomic_state *state,
> +				    struct drm_colorop *colorop)
> +{
> +	return state->colorops[drm_colorop_index(colorop)].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
> @@ -877,6 +936,29 @@ 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..e611f830f986
> --- /dev/null
> +++ b/include/drm/drm_colorop.h
> @@ -0,0 +1,158 @@
> +/*
> + * 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;
> +
> +	/* colorop properties */
> +
> +	/** @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.
> + */
> +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 plane.
> +	 */
> +	unsigned index;
> +
> +	/* TODO do we need a separate mutex or will we tag along with the plane mutex? */
> +
> +	/** @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. Either by
> +	 * going through the &struct drm_atomic_state pointers, see
> +	 * for_each_oldnew_plane_in_state(), for_each_old_plane_in_state() and
> +	 * for_each_new_plane_in_state(). Or through careful ordering of atomic
> +	 * commit operations as implemented in the atomic helpers, see
> +	 * &struct drm_crtc_commit.
> +	 *
> +	 * TODO keep, remove, or rewrite above plane references?
> +	 */
> +	struct drm_colorop_state *state;
> +
> +	/* colorop properties */
> +
> +	/** @properties: property tracking for this plane */
> +	struct drm_object_properties properties;
> +
> +};
> +
> +#define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
> +
> +/**
> + * drm_crtc_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;
> +}
> +
> +int drm_colorop_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);
> +
> +void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
> +				      struct drm_colorop_state *state);
> +void drm_colorop_destroy_state(struct drm_colorop *colorop,
> +			       struct drm_colorop_state *state);
> +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 973119a9176b..492b8c120c80 100644
> --- a/include/drm/drm_mode_config.h
> +++ b/include/drm/drm_mode_config.h
> @@ -505,6 +505,24 @@ struct drm_mode_config {
>  	 */
>  	struct list_head plane_list;
>  
> +	/**
> +	 * @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;
> +
> +	/**
> +	 * @colorops_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 641fe298052d..9c8c1322f0a6 100644
> --- a/include/drm/drm_plane.h
> +++ b/include/drm/drm_plane.h
> @@ -241,6 +241,8 @@ struct drm_plane_state {
>  	 */
>  	enum drm_scaling_filter scaling_filter;
>  

Missing kerneldocs for @color_pipeline.

> +	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 7040e7ea80c7..9e8f383935db 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 {
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 41/42] drm/colorop: Add mutliplier type
  2024-02-26 21:10 ` [RFC PATCH v4 41/42] drm/colorop: Add mutliplier type Harry Wentland
@ 2024-05-21 15:38   ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-05-21 15:38 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Alex Hung

On 02/26, Harry Wentland wrote:
> From: Alex Hung <alex.hung@amd.com>
> 
> 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.
> 
> Signed-off-by: Alex Hung <alex.hung@amd.com>
> ---
>  drivers/gpu/drm/drm_atomic.c      |  3 +++
>  drivers/gpu/drm/drm_atomic_uapi.c |  4 ++++
>  drivers/gpu/drm/drm_colorop.c     | 29 +++++++++++++++++++++++++++--
>  include/drm/drm_colorop.h         | 16 ++++++++++++++++
>  include/uapi/drm/drm_mode.h       |  1 +
>  5 files changed, 51 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
> index f7d51839ca03..af0b6338a55c 100644
> --- a/drivers/gpu/drm/drm_atomic.c
> +++ b/drivers/gpu/drm/drm_atomic.c
> @@ -806,6 +806,9 @@ 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=%u\n", state->multiplier);

Compiler complains of unsigned int instead of llu.

> +		break;
>  	default:
>  		break;
>  	}
> diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
> index 6bfe857720cd..b4ecda563728 100644
> --- a/drivers/gpu/drm/drm_atomic_uapi.c
> +++ b/drivers/gpu/drm/drm_atomic_uapi.c
> @@ -727,6 +727,8 @@ 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);
> @@ -752,6 +754,8 @@ 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->size_property) {
>  		*val = state->size;
>  	} else if (property == colorop->data_property) {
> diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
> index 4452eaeeb242..c6cdd743de51 100644
> --- a/drivers/gpu/drm/drm_colorop.c
> +++ b/drivers/gpu/drm/drm_colorop.c
> @@ -35,7 +35,8 @@
>  static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
>  	{ DRM_COLOROP_1D_CURVE, "1D Curve" },
>  	{ DRM_COLOROP_1D_LUT, "1D Curve Custom LUT" },
> -	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"}
> +	{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
> +	{ DRM_COLOROP_MULTIPLIER, "Multiplier"},
>  };
>  
>  static const char * const colorop_curve_1d_type_names[] = {
> @@ -231,6 +232,29 @@ int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop
>  }
>  EXPORT_SYMBOL(drm_colorop_ctm_3x4_init);
>  
> +int drm_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
> +			      struct drm_plane *plane)
> +{
> +	struct drm_property *prop;
> +	int ret;
> +
> +	ret = drm_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_colorop_mult_init);
> +
>  static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
>  							struct drm_colorop_state *state)
>  {
> @@ -333,7 +357,8 @@ EXPORT_SYMBOL(drm_colorop_reset);
>  static const char * const colorop_type_name[] = {
>  	[DRM_COLOROP_1D_CURVE] = "1D Curve",
>  	[DRM_COLOROP_1D_LUT] = "1D Curve Custom LUT",
> -	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
> +	[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
> +	[DRM_COLOROP_MULTIPLIER] = "Multiplier",
>  };
>  
>  /**
> diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
> index 8adc7ece3bd1..f9f83644cc9f 100644
> --- a/include/drm/drm_colorop.h
> +++ b/include/drm/drm_colorop.h
> @@ -64,6 +64,13 @@ 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;
> +
>  	/**
>  	 * @size:
>  	 *
> @@ -186,6 +193,13 @@ struct drm_colorop {
>  	 */
>  	struct drm_property *curve_1d_type_property;
>  
> +	/**
> +	 * @multiplier_property:
> +	 *
> +	 * Multiplier property for plane gain
> +	 */
> +	struct drm_property *multiplier_property;
> +
>  	/**
>  	 * @size_property:
>  	 *
> @@ -246,6 +260,8 @@ int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *co
>  				  struct drm_plane *plane, uint32_t lut_size);
>  int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
>  			     struct drm_plane *plane);
> +int drm_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 07fd66dc477c..754fd4c48123 100644
> --- a/include/uapi/drm/drm_mode.h
> +++ b/include/uapi/drm/drm_mode.h
> @@ -876,6 +876,7 @@ enum drm_colorop_type {
>  	DRM_COLOROP_1D_CURVE,
>  	DRM_COLOROP_1D_LUT,
>  	DRM_COLOROP_CTM_3X4,
> +	DRM_COLOROP_MULTIPLIER,
>  };
>  
>  /**
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2024-02-26 21:10 ` [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
@ 2024-05-21 15:45   ` Melissa Wen
  2024-05-21 15:55     ` Melissa Wen
  0 siblings, 1 reply; 69+ messages in thread
From: Melissa Wen @ 2024-05-21 15:45 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Alex Hung

On 02/26, Harry Wentland wrote:
> From: Alex Hung <alex.hung@amd.com>
> 
> 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>
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> Co-developed-by: Harry Wentland <harry.wentland@amd.com>
> ---
>  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
>  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 88 +++++++++++++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 58 ++++++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 +++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
>  5 files changed, 192 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 9b527bffe11a..3ec759934669 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 +668,19 @@ 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 +1150,59 @@ __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;

For this patch and the next two, it ^ should be:
`&dc_plane_state->in_transfer_func` (same for shape and blend), right?

> +	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 +1253,25 @@ 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 +1368,8 @@ 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..e8b7fc8bb0f1
> --- /dev/null
> +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> @@ -0,0 +1,58 @@
> +// 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);
> +
> +int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
> +{
> +	struct drm_colorop *op;
> +	struct drm_device *dev = plane->dev;
> +	int ret;
> +
> +	/* 1D curve - DEGAM TF */
> +	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> +	if (!op) {
> +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> +		return -ENOMEM;
> +	}
> +
> +	ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_degam_tfs);
> +	if (ret)
> +		return ret;
> +
> +	list->type = op->base.id;
> +	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", op->base.id);
> +
> +	return 0;
> +}
> \ No newline at end of file
> 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..f16de6a9fbde
> --- /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__*/
> \ No newline at end of file
> 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 d3f64f586243..3adab18332d8 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
> @@ -35,6 +35,7 @@
>  #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"
>  
> @@ -1658,10 +1659,19 @@ 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);
>  
> -- 
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS
  2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
                   ` (43 preceding siblings ...)
  2024-02-29 10:43 ` Daniel Vetter
@ 2024-05-21 15:50 ` Melissa Wen
  44 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-05-21 15:50 UTC (permalink / raw)
  To: Harry Wentland
  Cc: dri-devel, amd-gfx, wayland-devel, Ville Syrjala, Pekka Paalanen,
	Simon Ser, Jonas Ådahl, Sebastian Wick, Shashank Sharma,
	Alexander Goins, Joshua Ashton, Michel Dänzer, Aleix Pol,
	Xaver Hugl, Victoria Brekenfeld, Sima, Uma Shankar, Naseer Ahmed,
	Christopher Braga, Abhinav Kumar, Arthur Grillo, Hector Martin,
	Liviu Dudau, Sasha McIntosh, Chaitanya Kumar Borah

On 02/26, Harry Wentland wrote:
> This is an RFC set for a color pipeline API, along with a sample
> implementation in VKMS. All the key API bits are here. VKMS now
> supports two named transfer function colorops and two matrix
> colorops. We have IGT tests that check all four of these colorops
> with a pixel-by-pixel comparison that checks that these colorops
> do what we expect them to do with a +/- 1 8 bpc code point margin.
> 
> The big new change with v4 is the addition of an amdgpu color
> pipeline, for all AMD GPUs with DCN 3 and newer. Amdgpu now support
> the following:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 1D Curve EOTF
> 7. 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.
> 
> We are working on two more ops for amdgpu, the HDR multiplier
> and the 3DLUT, which will give us this:
> 
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. HDR Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D Curve EOTF
> 8. 1D LUT
> 
> This, essentially 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
> 
> After this we'd like to also add the following ops:
> - Scaler (Informational only)
> - Color Encoding, to replace drm_plane's COLOR_ENCODING
> - Color Range, to replace drm_plane's COLOR_RANGE
> 
> This patchset is grouped as follows:
>  - Patches 1-3: couple general patches/fixes
>  - Patches 4-7: introduce kunit to VKMS
>  - Patch 7: description of motivation and details behind the
>             Color Pipeline API. If you're reading nothing else
>             but are interested in the topic I highly recommend
>             you take a look at this.
>  - Patches 7-27: DRM core and VKMS changes for color pipeline API
>  - Patches 28-40: DRM core and amdgpu changes for color pipeline API
> 
> 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 here but
> haven't had a chance to look at. These will (hopefully) be
> addressed in future iterations, either in VKMS or amdgpu:
>  - Clear documentation for each drm_colorop_type
>  - Add custom LUT colorops to VKMS
>  - Add pre-blending 3DLUT
>  - How to support HW which can't bypass entire pipeline?
>  - Add ability to create colorops that don't have BYPASS
>  - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>  - read-only scaling colorop which defines scaling taps and position
>  - read-only color format colorop to define supported color formats
>    for a pipeline
>  - named matrices, for things like converting YUV to RGB
> 
> IGT tests can be found at
> https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/merge_requests/1
> 
> IGT patches are also being sent to the igt-dev mailing list.
> 
> If you prefer a gitlab MR for review you can find it at
> https://gitlab.freedesktop.org/hwentland/linux/-/merge_requests/5
> 
> v4:

Hey,

I'm not sure if this is the latest version, but this is the one I'm
using now. I just pointed out some minor issues that I found when
applying this series for testing, so I won't forget to report. But I'm
still performing tests and validations on my side.

Melissa

>  - 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 (10):
>   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
> 
> Harry Wentland (32):
>   drm: Don't treat 0 as -1 in drm_fixp2int_ceil
>   drm: Add helper for conversion from signed-magnitude
>   drm: Correctly round for fixp2int_round
>   drm/vkms: Round fixp2int conversion in lerp_u16
>   drm/vkms: Create separate Kconfig file for VKMS
>   drm/vkms: Add kunit tests for VKMS LUT handling
>   drm/vkms: Avoid reading beyond LUT array
>   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: Add NEXT to colorop state print
>   drm/vkms: Add enumerated 1D curve colorop
>   drm/vkms: Add kunit tests for linear and sRGB LUTs
>   drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
>   drm/colorop: Add 3x4 CTM type
>   drm/vkms: Pull apply_colorop out of pre_blend_color_transform
>   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
> 
>  Documentation/gpu/rfc/color_pipeline.rst      | 360 ++++++++
>  drivers/gpu/drm/Kconfig                       |  14 +-
>  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   | 352 ++++++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 160 ++++
>  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h |  36 +
>  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   |  32 +
>  drivers/gpu/drm/drm_atomic.c                  | 161 +++-
>  drivers/gpu/drm/drm_atomic_helper.c           |  12 +
>  drivers/gpu/drm/drm_atomic_state_helper.c     |   5 +
>  drivers/gpu/drm/drm_atomic_uapi.c             | 158 ++++
>  drivers/gpu/drm/drm_colorop.c                 | 411 +++++++++
>  drivers/gpu/drm/drm_ioctl.c                   |   7 +
>  drivers/gpu/drm/drm_mode_config.c             |   7 +
>  drivers/gpu/drm/drm_plane.c                   |  52 ++
>  drivers/gpu/drm/tests/Makefile                |   3 +-
>  drivers/gpu/drm/tests/drm_fixp_test.c         |  69 ++
>  drivers/gpu/drm/vkms/Kconfig                  |  20 +
>  drivers/gpu/drm/vkms/Makefile                 |   4 +-
>  drivers/gpu/drm/vkms/tests/.kunitconfig       |   4 +
>  drivers/gpu/drm/vkms/tests/vkms_color_tests.c | 449 ++++++++++
>  drivers/gpu/drm/vkms/vkms_colorop.c           | 100 +++
>  drivers/gpu/drm/vkms/vkms_composer.c          | 135 ++-
>  drivers/gpu/drm/vkms/vkms_drv.h               |   8 +
>  drivers/gpu/drm/vkms/vkms_luts.c              | 802 ++++++++++++++++++
>  drivers/gpu/drm/vkms/vkms_luts.h              |  12 +
>  drivers/gpu/drm/vkms/vkms_plane.c             |   2 +
>  include/drm/drm_atomic.h                      | 122 +++
>  include/drm/drm_atomic_uapi.h                 |   3 +
>  include/drm/drm_colorop.h                     | 301 +++++++
>  include/drm/drm_file.h                        |   7 +
>  include/drm/drm_fixed.h                       |  35 +-
>  include/drm/drm_mode_config.h                 |  18 +
>  include/drm/drm_plane.h                       |  13 +
>  include/uapi/drm/drm.h                        |  16 +
>  include/uapi/drm/drm_mode.h                   |  14 +
>  38 files changed, 3882 insertions(+), 30 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/Kconfig
>  create mode 100644 drivers/gpu/drm/vkms/tests/.kunitconfig
>  create mode 100644 drivers/gpu/drm/vkms/tests/vkms_color_tests.c
>  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
>  create mode 100644 include/drm/drm_colorop.h
> 
> --
> 2.44.0
> 

^ permalink raw reply	[flat|nested] 69+ messages in thread

* Re: [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block
  2024-05-21 15:45   ` Melissa Wen
@ 2024-05-21 15:55     ` Melissa Wen
  0 siblings, 0 replies; 69+ messages in thread
From: Melissa Wen @ 2024-05-21 15:55 UTC (permalink / raw)
  To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Alex Hung

On 05/21, Melissa Wen wrote:
> On 02/26, Harry Wentland wrote:
> > From: Alex Hung <alex.hung@amd.com>
> > 
> > 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>
> > Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> > Co-developed-by: Harry Wentland <harry.wentland@amd.com>
> > ---
> >  .../gpu/drm/amd/display/amdgpu_dm/Makefile    |  3 +-
> >  .../amd/display/amdgpu_dm/amdgpu_dm_color.c   | 88 +++++++++++++++++++
> >  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 58 ++++++++++++
> >  .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 34 +++++++
> >  .../amd/display/amdgpu_dm/amdgpu_dm_plane.c   | 10 +++
> >  5 files changed, 192 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 9b527bffe11a..3ec759934669 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 +668,19 @@ 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;;
Nitpick: double `;` 

> > +	}
> > +}
> > +
> >  static void __to_dc_lut3d_color(struct dc_rgb *rgb,
> >  				const struct drm_color_lut lut,
> >  				int bit_precision)
> > @@ -1137,6 +1150,59 @@ __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;
> 
> For this patch and the next two, it ^ should be:
> `&dc_plane_state->in_transfer_func` (same for shape and blend), right?
> 
> > +	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 +1253,25 @@ 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 +1368,8 @@ 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..e8b7fc8bb0f1
> > --- /dev/null
> > +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> > @@ -0,0 +1,58 @@
> > +// 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);
> > +
> > +int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
> > +{
> > +	struct drm_colorop *op;
> > +	struct drm_device *dev = plane->dev;
> > +	int ret;
> > +
> > +	/* 1D curve - DEGAM TF */
> > +	op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
> > +	if (!op) {
> > +		DRM_ERROR("KMS: Failed to allocate colorop\n");
> > +		return -ENOMEM;
> > +	}
> > +
> > +	ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_degam_tfs);
> > +	if (ret)
> > +		return ret;
> > +
> > +	list->type = op->base.id;
> > +	list->name = kasprintf(GFP_KERNEL, "Color Pipeline %d", op->base.id);
> > +
> > +	return 0;
> > +}
> > \ No newline at end of file
> > 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..f16de6a9fbde
> > --- /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__*/
> > \ No newline at end of file
> > 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 d3f64f586243..3adab18332d8 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
> > @@ -35,6 +35,7 @@
> >  #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"
> >  
> > @@ -1658,10 +1659,19 @@ 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);
> >  
> > -- 
> > 2.44.0
> > 

^ permalink raw reply	[flat|nested] 69+ messages in thread

end of thread, other threads:[~2024-05-21 15:56 UTC | newest]

Thread overview: 69+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-02-26 21:10 [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 01/42] drm: Don't treat 0 as -1 in drm_fixp2int_ceil Harry Wentland
2024-03-14 13:41   ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 02/42] drm: Add helper for conversion from signed-magnitude Harry Wentland
2024-03-14 13:16   ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 03/42] drm: Correctly round for fixp2int_round Harry Wentland
2024-03-11 13:11   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 04/42] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 05/42] drm/vkms: Create separate Kconfig file for VKMS Harry Wentland
2024-03-12 18:49   ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 06/42] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
2024-02-27 12:14   ` Arthur Grillo
2024-02-27 14:09     ` Harry Wentland
2024-03-11 13:45   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 07/42] drm/vkms: Avoid reading beyond LUT array Harry Wentland
2024-03-12 18:54   ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 08/42] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
2024-03-11 15:11   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 09/42] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
2024-05-21 15:28   ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 10/42] drm/colorop: Add TYPE property Harry Wentland
2024-03-12 15:02   ` Pekka Paalanen
2024-03-12 15:15     ` Simon Ser
2024-03-12 15:55       ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 11/42] drm/colorop: Add 1D Curve subtype Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 12/42] drm/colorop: Add BYPASS property Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 13/42] drm/colorop: Add NEXT property Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 14/42] drm/colorop: Add atomic state print for drm_colorop Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 15/42] drm/plane: Add COLOR PIPELINE property Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 16/42] drm/colorop: Add NEXT to colorop state print Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 17/42] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
2024-03-12 15:27   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 18/42] drm/vkms: Add kunit tests for linear and sRGB LUTs Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 19/42] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 20/42] drm/colorop: Add 3x4 CTM type Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 21/42] drm/vkms: Pull apply_colorop out of pre_blend_color_transform Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 22/42] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
2024-03-12 15:50   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 23/42] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
2024-03-14 15:23   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 24/42] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
2024-03-14 15:36   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 25/42] drm/vkms: Add tests for CTM handling Harry Wentland
2024-03-14 16:02   ` Pekka Paalanen
2024-02-26 21:10 ` [RFC PATCH v4 26/42] drm/colorop: pass plane_color_pipeline client cap to atomic check Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 27/42] drm/colorop: define a new macro for_each_new_colorop_in_state Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 28/42] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 29/42] drm/amd/display: Add bypass COLOR PIPELINE Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 30/42] drm/amd/display: Skip color pipeline initialization for cursor plane Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 31/42] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
2024-05-21 15:45   ` Melissa Wen
2024-05-21 15:55     ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 32/42] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 33/42] drm/amd/display: Add support for sRGB EOTF in BLND block Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 34/42] drm/colorop: Add PQ 125 EOTF and its inverse Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 35/42] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 36/42] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 37/42] drm/amd/display: Add support for BT.709 and BT.2020 TFs Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 38/42] drm/colorop: Add 1D Curve Custom LUT type Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 39/42] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 40/42] drm/amd/display: add 3x4 matrix colorop Harry Wentland
2024-02-26 21:10 ` [RFC PATCH v4 41/42] drm/colorop: Add mutliplier type Harry Wentland
2024-05-21 15:38   ` Melissa Wen
2024-02-26 21:10 ` [RFC PATCH v4 42/42] drm/amd/display: add multiplier colorop Harry Wentland
2024-02-27 10:26 ` [RFC PATCH v4 00/42] Color Pipeline API w/ VKMS Joshua Ashton
2024-02-27 14:00   ` Harry Wentland
2024-02-29 10:43 ` Daniel Vetter
2024-02-29 15:31   ` Joshua Ashton
2024-05-21 15:50 ` Melissa Wen

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).