* [PATCH v5 00/44] Color Pipeline API w/ VKMS
@ 2024-08-19 20:56 Harry Wentland
2024-08-19 20:56 ` [PATCH v5 01/44] drm: Add helper for conversion from signed-magnitude Harry Wentland
` (45 more replies)
0 siblings, 46 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 implementations
in VKMS and amdgpu. It is tested with a set of IGT tests that can be
found at [1]. The IGT tests run a pixel-by-pixel comparison with an
allowable delta variation as the goal for these transformations is
perceptual correctness, not complete pixel accuracy.
v5 of this patchset fleshed out documentation for colorops and the
various defines that are being introduced.
VKMS supports two named transfer function colorops and two matrix
colorops.
Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
1. 1D Curve EOTF
2. 3x4 CTM
3. Multiplier
4. 1D Curve Inverse EOTF
5. 1D LUT
6. 3D LUT
7. 1D Curve EOTF
8. 1D LUT
The supported curves for the 1D Curve type are:
- sRGB EOTF and its inverse
- PQ EOTF, scaled to [0.0, 125.0] and its inverse
- BT.2020/BT.709 OETF and its inverse
Note that the 1st and 5th colorops take the EOTF or Inverse
OETF while the 3rd colorop takes the Inverse EOTF or OETF.
The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
exists for other drivers to describe their own 3D LUT capability.
This mirrors the color pipeline used by gamescope and presented by
Melissa Wen, with the exception of the DEGAM LUT, which is not currently
used. See [1]
https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
At this point we're hoping to see gamescope and kwin implementations
take shape. The existing pipeline should be enough to satisfy the
gamescope use-cases on the drm_plane.
In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
support to the color pipeline. I have sketched these out already but
don't have it all hooked up yet. This should not hinder adoption of this
API for gaming use-cases.
We'll also want to advertise IN_FORMATS on a color pipeline as some
color pipelines won't be able to work for all IN_FORMATS on a plane.
Again, I have a sketch but no full implementation yet. This is not
currently required by the AMD color pipeline and could be added after
the merge of this set.
VKMS patches could still be improved in a few ways, though the
payoff might be limited and I would rather focus on other work
at the moment. The most obvious thing to improve would be to
eliminate the hard-coded LUTs for identity, and sRGB, and replace
them with fixed-point math instead.
There are plenty of things that I would like to see, but they could
be added after the merge of this patchset:
- COLOR_ENCODING and COLOR_RANGE
- IN_FORMATS for a color pipeline
- Is it possible to support HW which can't bypass entire pipeline?
- Can we do a LOAD / COMMIT model for LUTs (and other properties)?
- read-only scaling colorop which defines scaling taps and position
- named matrices, for things like converting YUV to RGB
- Add custom LUT colorops to VKMS
IGT tests can be found at [1] or on the igt-dev mailing list.
A kernel branch can be found at [2].
I've also rebased Uma and Chaitanya's patches for the Intel color
pipeline on top of this to show how I envision them to mesh with
my changes. The relevant branches can be found at [3] for the kernel
and [4] for IGT. There were some rebase conflicts in i915 and I'm
not entirely sure I've resolved all of them correctly, but the branch
compiles and shows my thoughts for the new DRM concepts to support
Intel's pipeline.
[1] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-color-pipeline-v5
[2] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v5
[3] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-intel-color-pipeline-v5
[4] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-intel-color-pipeline-v5
v5:
- amdgpu 3D LUT
- Don't require BYPASS
- update RFC docs and add to TOC tree
- add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
- add amdgpu color pipeline doc
- define SIZE property similar to drm_crtc's GAMMA_SIZE
- various minor fixes and cleanups
v4:
- Add amdgpu color pipeline (WIP)
- Don't block setting of deprecated properties, instead pass client cap
to atomic check so drivers can ignore these props
- Drop IOCTL definitions (Pekka)
- Use enum property for colorop TYPE (Pekka)
- A few cleanups to the docs (Pekka)
- Rework the TYPE enum to name relation to avoid code duplication (Pekka)
- Add missing function declarations (Chaitanya Kumar Borah)
- Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
- Add helper for creation of pipeline drm_plane property (Pekka)
- Always create Bypass pipeline (Pekka)
- A bunch of changes to VKMS kunit tests (Pekka)
- Fix index in CTM doc (Pekka)
v3:
- Abandon IOCTLs and discover colorops as clients iterate the pipeline
- Remove need for libdrm
- Add color_pipeline client cap and make mutually exclusive with
COLOR_RANGE and COLOR_ENCODING properties
- add CTM colorop to VKMS
- Use include way for kunit testing static functions (Arthur)
- Make TYPE a range property
- Move enum drm_colorop_type to uapi header
- and a bunch of smaller bits that are highlighted in the relevant commit
description
v2:
- Rebased on drm-misc-next
- Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
- Incorporate feedback in color_pipeline.rst doc
- Add support for sRGB inverse EOTF
- Add 2nd enumerated TF colorop to VKMS
- Fix LUTs and some issues with applying LUTs in VKMS
Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
Cc: Simon Ser <contact@emersion.fr>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Melissa Wen <mwen@igalia.com>
Cc: Jonas Ådahl <jadahl@redhat.com>
Cc: Sebastian Wick <sebastian.wick@redhat.com>
Cc: Shashank Sharma <shashank.sharma@amd.com>
Cc: Alexander Goins <agoins@nvidia.com>
Cc: Joshua Ashton <joshua@froggi.es>
Cc: Michel Dänzer <mdaenzer@redhat.com>
Cc: Aleix Pol <aleixpol@kde.org>
Cc: Xaver Hugl <xaver.hugl@gmail.com>
Cc: Victoria Brekenfeld <victoria@system76.com>
Cc: Sima <daniel@ffwll.ch>
Cc: Uma Shankar <uma.shankar@intel.com>
Cc: Naseer Ahmed <quic_naseer@quicinc.com>
Cc: Christopher Braga <quic_cbraga@quicinc.com>
Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
Cc: Arthur Grillo <arthurgrillo@riseup.net>
Cc: Hector Martin <marcan@marcan.st>
Cc: Liviu Dudau <Liviu.Dudau@arm.com>
Cc: Sasha McIntosh <sashamcintosh@google.com>
Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
Alex Hung (13):
drm/colorop: define a new macro for_each_new_colorop_in_state
drm/amd/display: Skip color pipeline initialization for cursor plane
drm/amd/display: Add support for sRGB EOTF in DEGAM block
drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
drm/amd/display: Add support for sRGB EOTF in BLND block
drm/colorop: Add 1D Curve Custom LUT type
drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
drm/amd/display: add 3x4 matrix colorop
drm/colorop: Add mutliplier type
drm/amd/display: add multiplier colorop
drm/amd/display: Swap matrix and multiplier
drm/colorop: Add 3D LUT supports to color pipeline
drm/amd/display: add 3D LUT colorop
Harry Wentland (31):
drm: Add helper for conversion from signed-magnitude
drm/vkms: Round fixp2int conversion in lerp_u16
drm/vkms: Add kunit tests for VKMS LUT handling
drm/doc/rfc: Describe why prescriptive color pipeline is needed
drm/colorop: Introduce new drm_colorop mode object
drm/colorop: Add TYPE property
drm/colorop: Add 1D Curve subtype
Documentation/gpu: document drm_colorop
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: Use s32 for internal color pipeline precision
drm/vkms: add 3x4 matrix in color pipeline
drm/tests: Add a few tests around drm_fixed.h
drm/vkms: Add tests for CTM handling
drm/colorop: pass plane_color_pipeline client cap to atomic check
drm/amd/display: Ignore deprecated props when plane_color_pipeline set
drm/amd/display: Add bypass COLOR PIPELINE
drm/colorop: Add PQ 125 EOTF and its inverse
drm/amd/display: Enable support for PQ 125 EOTF and Inverse
drm/colorop: add BT2020/BT709 OETF and Inverse OETF
drm/amd/display: Add support for BT.709 and BT.2020 TFs
drm/colorop: Define LUT_1D interpolation
drm/colorop: allow non-bypass colorops
drm/amd/display: Add AMD color pipeline doc
Documentation/gpu/drm-kms.rst | 15 +
Documentation/gpu/rfc/color_pipeline.rst | 376 ++++++++
Documentation/gpu/rfc/index.rst | 3 +
drivers/gpu/drm/Makefile | 1 +
.../gpu/drm/amd/display/amdgpu_dm/Makefile | 3 +-
.../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +
.../amd/display/amdgpu_dm/amdgpu_dm_color.c | 581 ++++++++++++-
.../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 196 +++++
.../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 36 +
.../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 32 +
drivers/gpu/drm/drm_atomic.c | 184 +++-
drivers/gpu/drm/drm_atomic_helper.c | 12 +
drivers/gpu/drm/drm_atomic_state_helper.c | 5 +
drivers/gpu/drm/drm_atomic_uapi.c | 175 ++++
drivers/gpu/drm/drm_colorop.c | 562 ++++++++++++
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 | 5 +
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 | 102 +++
drivers/gpu/drm/vkms/vkms_composer.c | 115 ++-
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 | 130 +++
include/drm/drm_atomic_uapi.h | 3 +
include/drm/drm_colorop.h | 427 ++++++++++
include/drm/drm_file.h | 7 +
include/drm/drm_fixed.h | 18 +
include/drm/drm_mode_config.h | 18 +
include/drm/drm_plane.h | 19 +
include/uapi/drm/amdgpu_drm.h | 9 -
include/uapi/drm/drm.h | 16 +
include/uapi/drm/drm_mode.h | 156 ++++
40 files changed, 4595 insertions(+), 34 deletions(-)
create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
create mode 100644 drivers/gpu/drm/drm_colorop.c
create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
create mode 100644 drivers/gpu/drm/vkms/tests/.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
--
2.46.0
^ permalink raw reply [flat|nested] 69+ messages in thread
* [PATCH v5 01/44] drm: Add helper for conversion from signed-magnitude
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 02/44] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
` (44 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 ef8bc8d72039..793768c1e776 100644
--- a/include/drm/drm_fixed.h
+++ b/include/drm/drm_fixed.h
@@ -77,6 +77,24 @@ 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;
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 02/44] drm/vkms: Round fixp2int conversion in lerp_u16
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
2024-08-19 20:56 ` [PATCH v5 01/44] drm: Add helper for conversion from signed-magnitude Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
` (43 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 e7441b227b3c..3d6785d081f2 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)
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
2024-08-19 20:56 ` [PATCH v5 01/44] drm: Add helper for conversion from signed-magnitude Harry Wentland
2024-08-19 20:56 ` [PATCH v5 02/44] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-21 1:05 ` Jeff Johnson
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 04/44] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
` (42 subsequent siblings)
45 siblings, 2 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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
v5:
- Bring back static for lerp_u16 and get_lut_index (Arthur)
v4:
- Test the critical points of the lerp function (Pekka)
v3:
- Use include way of testing static functions (Arthur)
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 | 4 +
4 files changed, 176 insertions(+)
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 3d6785d081f2..3ecda70c2b55 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -435,3 +435,7 @@ 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
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 04/44] drm/doc/rfc: Describe why prescriptive color pipeline is needed
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (2 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 05/44] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
` (41 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland
v5:
- Don't require BYPASS to succeed (Sebastian)
- use DATA for 1D and 3D LUT types (Sebastian)
- update 3DLUT ops to use 3DLUT_MODES and 3DLUT_MODE_INDEX
- Add section on drm_colorop extensibility
- Add color_pipeline.rst to RFC toc tree
v4:
- Drop IOCTL docs since we dropped the IOCTLs (Pekka)
- Clarify reading and setting of COLOR_PIPELINE prop (Pekka)
- Add blurb about not requiring to reject a pipeline due to
incompatible ops, as long as op can be bypassed (Pekka)
- Dropped informational strings (such as input CSC) as they're
not actually intended to be advertised (Pekka)
v3:
- Describe DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE (Sebastian)
- Ask for clear documentation of colorop behavior (Sebastian)
v2:
- Update colorop visualizations to match reality (Sebastian, Alex Hung)
- Updated wording (Pekka)
- Change BYPASS wording to make it non-mandatory (Sebastian)
- Drop cover-letter-like paragraph from COLOR_PIPELINE Plane Property
section (Pekka)
- Use PQ EOTF instead of its inverse in Pipeline Programming example (Melissa)
- Add "Driver Implementer's Guide" section (Pekka)
- Add "Driver Forward/Backward Compatibility" section (Sebastian, Pekka)
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
Documentation/gpu/rfc/color_pipeline.rst | 376 +++++++++++++++++++++++
Documentation/gpu/rfc/index.rst | 3 +
2 files changed, 379 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..4135f07b85a7
--- /dev/null
+++ b/Documentation/gpu/rfc/color_pipeline.rst
@@ -0,0 +1,376 @@
+========================
+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. The BYPASS property is not mandatory for a colorop, as long
+as the entire pipeline can get bypassed by setting the COLOR_PIPELINE on
+a plane to '0'.
+
+NEXT: The ID of the next drm_colorop in a color pipeline, or 0 if this
+drm_colorop is the last in the chain.
+
+An example of a drm_colorop object might look like one of these::
+
+ /* 1D enumerated curve */
+ Color operation 42
+ ├─ "TYPE": immutable enum {1D enumerated curve, 1D LUT, 3x3 matrix, 3x4 matrix, 3D LUT, etc.} = 1D enumerated curve
+ ├─ "BYPASS": bool {true, false}
+ ├─ "CURVE_1D_TYPE": enum {sRGB EOTF, sRGB inverse EOTF, PQ EOTF, PQ inverse EOTF, …}
+ └─ "NEXT": immutable color operation ID = 43
+
+ /* custom 4k entry 1D LUT */
+ Color operation 52
+ ├─ "TYPE": immutable enum {1D enumerated curve, 1D LUT, 3x3 matrix, 3x4 matrix, 3D LUT, etc.} = 1D LUT
+ ├─ "BYPASS": bool {true, false}
+ ├─ "SIZE": immutable range = 4096
+ ├─ "DATA": blob
+ └─ "NEXT": immutable color operation ID = 0
+
+ /* 17^3 3D LUT */
+ Color operation 72
+ ├─ "TYPE": immutable enum {1D enumerated curve, 1D LUT, 3x3 matrix, 3x4 matrix, 3D LUT, etc.} = 3D LUT
+ ├─ "BYPASS": bool {true, false}
+ ├─ "3DLUT_MODES": read-only blob of supported 3DLUT modes
+ ├─ "3DLUT_MODE_INDEX": index of selected 3DLUT mode
+ ├─ "DATA": blob
+ └─ "NEXT": immutable color operation ID = 73
+
+drm_colorop extensibility
+-------------------------
+
+Unlike existing DRM core objects, like &drm_plane, drm_colorop is not
+extensible. This simplifies implementations and keeps all functionality
+for managing &drm_colorop objects in the DRM core.
+
+If there is a need one may introduce a simple &drm_colorop_funcs
+function table in the future, for example to support an IN_FORMATS
+property on a &drm_colorop.
+
+If a driver requires the ability to create a driver-specific colorop
+object they will need to add &drm_colorop func table support with
+support for the usual functions, like destroy, atomic_duplicate_state,
+and atomic_destroy_state.
+
+
+COLOR_PIPELINE Plane Property
+=============================
+
+Color Pipelines are created by a driver and advertised via a new
+COLOR_PIPELINE enum property on each plane. Values of the property
+always include object id 0, which is the default and means all color
+processing is disabled. Additional values will be the object IDs of the
+first drm_colorop in a pipeline. A driver can create and advertise none,
+one, or more possible color pipelines. A DRM client will select a color
+pipeline by setting the COLOR PIPELINE to the respective value.
+
+NOTE: Many DRM clients will set enumeration properties via the string
+value, often hard-coding it. Since this enumeration is generated based
+on the colorop object IDs it is important to perform the Color Pipeline
+Discovery, described below, instead of hard-coding color pipeline
+assignment. Drivers might generate the enum strings dynamically.
+Hard-coded strings might only work for specific drivers on a specific
+pieces of HW. Color Pipeline Discovery can work universally, as long as
+drivers implement the required color operations.
+
+The COLOR_PIPELINE property is only exposed when the
+DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE is set. Drivers shall ignore any
+existing pre-blend color operations when this cap is set, such as
+COLOR_RANGE and COLOR_ENCODING. If drivers want to support COLOR_RANGE
+or COLOR_ENCODING functionality when the color pipeline client cap is
+set, they are expected to expose colorops in the pipeline to allow for
+the appropriate color transformation.
+
+Setting of the COLOR_PIPELINE plane property or drm_colorop properties
+is only allowed for userspace that sets this client cap.
+
+An example of a COLOR_PIPELINE property on a plane might look like this::
+
+ Plane 10
+ ├─ "TYPE": immutable enum {Overlay, Primary, Cursor} = Primary
+ ├─ …
+ └─ "COLOR_PIPELINE": enum {0, 42, 52} = 0
+
+
+Color Pipeline Discovery
+========================
+
+A DRM client wanting color management on a drm_plane will:
+
+1. Get the COLOR_PIPELINE property of the plane
+2. iterate all COLOR_PIPELINE enum values
+3. for each enum value walk the color pipeline (via the NEXT pointers)
+ and see if the available color operations are suitable for the
+ desired color management operations
+
+If userspace encounters an unknown or unsuitable color operation during
+discovery it does not need to reject the entire color pipeline outright,
+as long as the unknown or unsuitable colorop has a "BYPASS" property.
+Drivers will ensure that a bypassed block does not have any effect.
+
+An example of chained properties to define an AMD pre-blending color
+pipeline might look like this::
+
+ Plane 10
+ ├─ "TYPE" (immutable) = Primary
+ └─ "COLOR_PIPELINE": enum {0, 44} = 0
+
+ Color operation 44
+ ├─ "TYPE" (immutable) = 1D enumerated curve
+ ├─ "BYPASS": bool
+ ├─ "CURVE_1D_TYPE": enum {sRGB EOTF, PQ EOTF} = sRGB EOTF
+ └─ "NEXT" (immutable) = 45
+
+ Color operation 45
+ ├─ "TYPE" (immutable) = 3x4 Matrix
+ ├─ "BYPASS": bool
+ ├─ "DATA": blob
+ └─ "NEXT" (immutable) = 46
+
+ Color operation 46
+ ├─ "TYPE" (immutable) = 1D enumerated curve
+ ├─ "BYPASS": bool
+ ├─ "CURVE_1D_TYPE": enum {sRGB Inverse EOTF, PQ Inverse EOTF} = sRGB EOTF
+ └─ "NEXT" (immutable) = 47
+
+ Color operation 47
+ ├─ "TYPE" (immutable) = 1D LUT
+ ├─ "SIZE": immutable range = 4096
+ ├─ "DATA": blob
+ └─ "NEXT" (immutable) = 48
+
+ Color operation 48
+ ├─ "TYPE" (immutable) = 3D LUT
+ ├─ "3DLUT_MODE_INDEX": 0
+ ├─ "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
diff --git a/Documentation/gpu/rfc/index.rst b/Documentation/gpu/rfc/index.rst
index 476719771eef..d3726adeca6a 100644
--- a/Documentation/gpu/rfc/index.rst
+++ b/Documentation/gpu/rfc/index.rst
@@ -31,3 +31,6 @@ host such documentation:
.. toctree::
i915_vm_bind.rst
+
+.. toctree::
+ color_pipeline.rst
\ No newline at end of file
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 05/44] drm/colorop: Introduce new drm_colorop mode object
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (3 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 04/44] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 06/44] drm/colorop: Add TYPE property Harry Wentland
` (40 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Add comment to drm_atomic_state.colorops
- Replace a misplaced 'plane' with 'colorop' in comment
- Fix colorop_list kernel doc
- Add kernel doc for color_pipeline
- drop unused drm_colorop_destroy_state
- drop drm_colorop_init, to be introduced in later patch
when used
- Add kernel docs
- Drop TODOs
v4:
- Drop IOCTL definitions (Pekka)
- add missing declaration (Chaitanya Kumar Borah)
v3:
- Drop TODO for lock (it's handled in drm_modeset_drop_locks)
(Melissa)
- Don't get plane state when getting colorop state
- Make some functions static (kernel test robot)
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 | 103 +++++++++++++++++++
drivers/gpu/drm/drm_mode_config.c | 7 ++
include/drm/drm_atomic.h | 89 ++++++++++++++++
include/drm/drm_atomic_uapi.h | 1 +
include/drm/drm_colorop.h | 152 ++++++++++++++++++++++++++++
include/drm/drm_mode_config.h | 18 ++++
include/drm/drm_plane.h | 8 ++
include/uapi/drm/drm_mode.h | 1 +
12 files changed, 510 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 68cc9258ffc4..45b687fadabe 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -44,6 +44,7 @@ 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 6e516c39a372..f82240eccc69 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 fb97b51b38f1..6418ddc184de 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -3024,6 +3024,8 @@ 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;
@@ -3101,6 +3103,16 @@ int drm_atomic_helper_swap_state(struct drm_atomic_state *state,
}
}
+ for_each_oldnew_colorop_in_state(state, colorop, old_colorop_state, new_colorop_state, i) {
+ WARN_ON(colorop->state != old_colorop_state);
+
+ old_colorop_state->state = state;
+ new_colorop_state->state = NULL;
+
+ state->colorops[i].state = old_colorop_state;
+ colorop->state = new_colorop_state;
+ }
+
drm_panic_lock(state->dev, flags);
for_each_oldnew_plane_in_state(state, plane, old_plane_state, new_plane_state, i) {
WARN_ON(plane->state != old_plane_state);
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 7609c798d73d..fce7683e4178 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)
@@ -908,6 +929,16 @@ 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;
@@ -1091,6 +1122,23 @@ 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..e005aba7d28a
--- /dev/null
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -0,0 +1,103 @@
+/*
+ * Copyright (C) 2023 Advanced Micro Devices, Inc. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include <drm/drm_colorop.h>
+#include <drm/drm_print.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_plane.h>
+
+#include "drm_crtc_internal.h"
+
+static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
+ struct drm_colorop_state *state)
+{
+ memcpy(state, colorop->state, sizeof(*state));
+}
+
+struct drm_colorop_state *
+drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop)
+{
+ struct drm_colorop_state *state;
+
+ if (WARN_ON(!colorop->state))
+ return NULL;
+
+ state = kmalloc(sizeof(*state), GFP_KERNEL);
+ if (state)
+ __drm_atomic_helper_colorop_duplicate_state(colorop, state);
+
+ return state;
+}
+
+
+void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
+ struct drm_colorop_state *state)
+{
+ kfree(state);
+}
+
+/**
+ * __drm_colorop_state_reset - resets colorop state to default values
+ * @colorop_state: atomic colorop state, must not be NULL
+ * @colorop: colorop object, must not be NULL
+ *
+ * Initializes the newly allocated @colorop_state with default
+ * values. This is useful for drivers that subclass the CRTC state.
+ */
+static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
+ struct drm_colorop *colorop)
+{
+ colorop_state->colorop = colorop;
+}
+
+/**
+ * __drm_colorop_reset - reset state on colorop
+ * @colorop: drm colorop
+ * @colorop_state: colorop state to assign
+ *
+ * Initializes the newly allocated @colorop_state and assigns it to
+ * the &drm_crtc->state pointer of @colorop, usually required when
+ * initializing the drivers or when called from the &drm_colorop_funcs.reset
+ * hook.
+ *
+ * This is useful for drivers that subclass the colorop state.
+ */
+static void __drm_colorop_reset(struct drm_colorop *colorop,
+ struct drm_colorop_state *colorop_state)
+{
+ if (colorop_state)
+ __drm_colorop_state_reset(colorop_state, colorop);
+
+ colorop->state = colorop_state;
+}
+
+void drm_colorop_reset(struct drm_colorop *colorop)
+{
+ kfree(colorop->state);
+ colorop->state = kzalloc(sizeof(*colorop->state), GFP_KERNEL);
+
+ if (colorop->state)
+ __drm_colorop_reset(colorop, colorop->state);
+}
diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
index 37d2e0a4ef4b..f238a2f049b0 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);
@@ -420,6 +425,7 @@ 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);
@@ -441,6 +447,7 @@ 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 4d7f4c5f2001..d2b44121b7fb 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;
@@ -408,6 +414,14 @@ struct drm_atomic_state {
*/
bool duplicated : 1;
+ /**
+ * @colorops:
+ *
+ * Pointer to array of @drm_colorop and @drm_colorop_state part of this
+ * update.
+ */
+ struct __drm_colorops_state *colorops;
+
/**
* @planes:
*
@@ -549,6 +563,9 @@ 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);
@@ -678,6 +695,55 @@ 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
@@ -925,6 +991,29 @@ 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..1bd658617840
--- /dev/null
+++ b/include/drm/drm_colorop.h
@@ -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
+ *
+ */
+
+#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 colorop.
+ */
+ unsigned index;
+
+ /** @base: base mode object */
+ struct drm_mode_object base;
+
+ /**
+ * @plane:
+ *
+ * The plane on which the colorop sits. A drm_colorop is always unique
+ * to a plane.
+ */
+ struct drm_plane *plane;
+
+ /**
+ * @state:
+ *
+ * Current atomic state for this colorop.
+ *
+ * This is protected by @mutex. Note that nonblocking atomic commits
+ * access the current colorop state without taking locks.
+ */
+ struct drm_colorop_state *state;
+
+ /* colorop properties */
+
+ /** @properties: property tracking for this colorop */
+ struct drm_object_properties properties;
+
+};
+
+#define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
+
+/**
+ * drm_colorop_find - look up a Colorop object from its ID
+ * @dev: DRM device
+ * @file_priv: drm file to check for lease against.
+ * @id: &drm_mode_object ID
+ *
+ * This can be used to look up a Colorop from its userspace ID. Only used by
+ * drivers for legacy IOCTLs and interface, nowadays extensions to the KMS
+ * userspace interface should be done using &drm_property.
+ */
+static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
+ struct drm_file *file_priv,
+ uint32_t id)
+{
+ struct drm_mode_object *mo;
+ mo = drm_mode_object_find(dev, file_priv, id, DRM_MODE_OBJECT_COLOROP);
+ return mo ? obj_to_colorop(mo) : NULL;
+}
+
+struct drm_colorop_state *
+drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
+
+void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
+ struct drm_colorop_state *state);
+
+/**
+ * drm_colorop_reset - reset colorop atomic state
+ * @colorop: drm colorop
+ *
+ * Resets the atomic state for @colorop by freeing the state pointer (which might
+ * be NULL, e.g. at driver load time) and allocating a new empty state object.
+ */
+void drm_colorop_reset(struct drm_colorop *colorop);
+
+/**
+ * drm_colorop_index - find the index of a registered colorop
+ * @colorop: colorop to find index for
+ *
+ * Given a registered colorop, return the index of that colorop within a DRM
+ * device's list of colorops.
+ */
+static inline unsigned int drm_colorop_index(const struct drm_colorop *colorop)
+{
+ return colorop->index;
+}
+
+
+#define drm_for_each_colorop(colorop, dev) \
+ list_for_each_entry(colorop, &(dev)->mode_config.colorop_list, head)
+
+
+#endif /* __DRM_COLOROP_H__ */
diff --git a/include/drm/drm_mode_config.h b/include/drm/drm_mode_config.h
index ab0f167474b1..dfdd80da8da2 100644
--- a/include/drm/drm_mode_config.h
+++ b/include/drm/drm_mode_config.h
@@ -515,6 +515,24 @@ struct drm_mode_config {
*/
struct raw_spinlock panic_lock;
+ /**
+ * @num_colorop:
+ *
+ * Number of colorop objects on this device.
+ * This is invariant over the lifetime of a device and hence doesn't
+ * need any locks.
+ */
+ int num_colorop;
+
+ /**
+ * @colorop_list:
+ *
+ * List of colorop objects linked with &drm_colorop.head. This is
+ * invariant over the lifetime of a device and hence doesn't need any
+ * locks.
+ */
+ struct list_head colorop_list;
+
/**
* @num_crtc:
*
diff --git a/include/drm/drm_plane.h b/include/drm/drm_plane.h
index dd718c62ac31..2a9b902623c0 100644
--- a/include/drm/drm_plane.h
+++ b/include/drm/drm_plane.h
@@ -243,6 +243,14 @@ struct drm_plane_state {
*/
enum drm_scaling_filter scaling_filter;
+ /**
+ * @color_pipeline:
+ *
+ * The first colorop of the active color pipeline, or NULL, if no
+ * color pipeline is active.
+ */
+ struct drm_colorop *color_pipeline;
+
/**
* @commit: Tracks the pending commit to prevent use-after-free conditions,
* and for async plane updates.
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index d390011b89b4..af71a5c84aa0 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 {
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 06/44] drm/colorop: Add TYPE property
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (4 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 05/44] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 07/44] drm/colorop: Add 1D Curve subtype Harry Wentland
` (39 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Add drm_get_colorop_type_name in header
- Add kernel docs
v4:
- Use enum property for TYPE (Pekka)
v3:
- Make TYPE a range property
- Move enum drm_colorop_type to uapi header
- Fix drm_get_colorop_type_name description
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 | 12 ++++++++++++
include/drm/drm_colorop.h | 23 +++++++++++++++++++++++
include/uapi/drm/drm_mode.h | 15 +++++++++++++++
5 files changed, 59 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index f82240eccc69..b09db9835e1f 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 fce7683e4178..f76a83d79993 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 e005aba7d28a..79af42b349a0 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -101,3 +101,15 @@ void drm_colorop_reset(struct drm_colorop *colorop)
if (colorop->state)
__drm_colorop_reset(colorop, colorop->state);
}
+
+static const char * const colorop_type_name[] = {
+ [DRM_COLOROP_1D_CURVE] = "1D Curve",
+};
+
+const char *drm_get_colorop_type_name(enum drm_colorop_type type)
+{
+ if (WARN_ON(type >= ARRAY_SIZE(colorop_type_name)))
+ return "unknown";
+
+ return colorop_type_name[type];
+}
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 1bd658617840..75ec3ec827ac 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -94,6 +94,21 @@ struct drm_colorop {
/** @properties: property tracking for this colorop */
struct drm_object_properties properties;
+ /**
+ * @type:
+ *
+ * Read-only
+ * Type of color operation
+ */
+ enum drm_colorop_type type;
+
+ /**
+ * @type_property:
+ *
+ * Read-only "TYPE" property for specifying the type of
+ * this color operation. The type is enum drm_colorop_type.
+ */
+ struct drm_property *type_property;
};
#define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
@@ -148,5 +163,13 @@ static inline unsigned int drm_colorop_index(const struct drm_colorop *colorop)
#define drm_for_each_colorop(colorop, dev) \
list_for_each_entry(colorop, &(dev)->mode_config.colorop_list, head)
+/**
+ * drm_get_colorop_type_name - return a string for colorop type
+ * @type: colorop type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_type_name(enum drm_colorop_type type);
#endif /* __DRM_COLOROP_H__ */
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index af71a5c84aa0..3dd89ae6515a 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -858,6 +858,21 @@ struct drm_color_lut {
__u16 reserved;
};
+/**
+ * enum drm_colorop_type - Type of color operation
+ *
+ * @DRM_COLOROP_1D_CURVE: A 1D curve that is being applied to all color channels.
+ * The curve is specified via tha CURVE_1D_TYPE colorop
+ * property.
+ *
+ * drm_colorops can be of many different types. Each type behaves differently
+ * and defines a different set of properties. This enum defines all types and
+ * gives a high-level description.
+ */
+enum drm_colorop_type {
+ DRM_COLOROP_1D_CURVE
+};
+
/**
* struct drm_plane_size_hint - Plane size hints
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 07/44] drm/colorop: Add 1D Curve subtype
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (5 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 06/44] drm/colorop: Add TYPE property Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 08/44] Documentation/gpu: document drm_colorop Harry Wentland
` (38 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland, Alex Hung
v5:
- Add drm_get_colorop_curve_1d_type_name in header
- Add drm_colorop_init
- Set default curve
- Add kernel docs
v4:
- Use drm_colorop_curve_1d_type_enum_list to get name (Pekka)
- Create separate init function for 1D curve
- Pass supported TFs into 1D curve init function
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 | 134 ++++++++++++++++++++++++++++++
include/drm/drm_colorop.h | 60 +++++++++++++
3 files changed, 207 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index f76a83d79993..b867bff49f67 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 79af42b349a0..98d22ac163e8 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -30,6 +30,123 @@
#include "drm_crtc_internal.h"
+static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
+ { DRM_COLOROP_1D_CURVE, "1D Curve" },
+};
+
+static const char * const colorop_curve_1d_type_names[] = {
+ [DRM_COLOROP_1D_CURVE_SRGB_EOTF] = "sRGB EOTF",
+ [DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF] = "sRGB Inverse EOTF",
+};
+
+
+/* Init Helpers */
+
+static int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
+ struct drm_plane *plane, enum drm_colorop_type type)
+{
+ struct drm_mode_config *config = &dev->mode_config;
+ struct drm_property *prop;
+ int ret = 0;
+
+ ret = drm_mode_object_add(dev, &colorop->base, DRM_MODE_OBJECT_COLOROP);
+ if (ret)
+ return ret;
+
+ colorop->base.properties = &colorop->properties;
+ colorop->dev = dev;
+ colorop->type = type;
+ colorop->plane = plane;
+
+ list_add_tail(&colorop->head, &config->colorop_list);
+ colorop->index = config->num_colorop++;
+
+ /* add properties */
+
+ /* type */
+ prop = drm_property_create_enum(dev,
+ DRM_MODE_PROP_IMMUTABLE,
+ "TYPE", drm_colorop_type_enum_list,
+ ARRAY_SIZE(drm_colorop_type_enum_list));
+
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->type_property = prop;
+
+ drm_object_attach_property(&colorop->base,
+ colorop->type_property,
+ colorop->type);
+
+ return ret;
+}
+
+/**
+ * drm_colorop_curve_1d_init - Initialize a DRM_COLOROP_1D_CURVE
+ *
+ * @dev: DRM device
+ * @colorop: The drm_colorop object to initialize
+ * @plane: The associated drm_plane
+ * @supported_tfs: A bitfield of supported drm_colorop_curve_1d_init enum values,
+ * created using BIT(curve_type) and combined with the OR '|'
+ * operator.
+ * @return zero on success, -E value on failure
+ */
+int drm_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++;
+ }
+
+ if (WARN_ON(len <= 0))
+ return -EINVAL;
+
+
+ /* initialize 1D curve only attribute */
+ prop = drm_property_create_enum(dev, DRM_MODE_PROP_ATOMIC, "CURVE_1D_TYPE",
+ enum_list, len);
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->curve_1d_type_property = prop;
+ drm_object_attach_property(&colorop->base, colorop->curve_1d_type_property,
+ enum_list[0].type);
+ drm_colorop_reset(colorop);
+
+ return 0;
+}
+EXPORT_SYMBOL(drm_colorop_curve_1d_init);
+
static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
struct drm_colorop_state *state)
{
@@ -69,7 +186,16 @@ void drm_colorop_atomic_destroy_state(struct drm_colorop *colorop,
static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
struct drm_colorop *colorop)
{
+ u64 val;
+
colorop_state->colorop = colorop;
+
+ if (colorop->curve_1d_type_property) {
+ drm_object_property_get_default_value(&colorop->base,
+ colorop->curve_1d_type_property,
+ &val);
+ colorop_state->curve_1d_type = val;
+ }
}
/**
@@ -113,3 +239,11 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type)
return colorop_type_name[type];
}
+
+const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type)
+{
+ if (WARN_ON(type >= ARRAY_SIZE(colorop_curve_1d_type_names)))
+ return "unknown";
+
+ return colorop_curve_1d_type_names[type];
+}
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 75ec3ec827ac..dd6943117461 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -30,6 +30,39 @@
#include <drm/drm_mode.h>
#include <drm/drm_property.h>
+
+/**
+ * enum drm_colorop_curve_1d_type - type of 1D curve
+ *
+ * Describes a 1D curve to be applied by the DRM_COLOROP_1D_CURVE colorop.
+ *
+ */
+enum drm_colorop_curve_1d_type {
+ /**
+ * @DRM_COLOROP_1D_CURVE_SRGB_EOTF:
+ *
+ * sRGB piece-wise electro-optical transfer function. Transfer
+ * characteristics as defined by IEC 61966-2-1 sRGB. Equivalent
+ * to H.273 TransferCharacteristics code point 13 with
+ * MatrixCoefficients set to 0.
+ */
+ DRM_COLOROP_1D_CURVE_SRGB_EOTF,
+
+ /**
+ * @DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF:
+ *
+ * The inverse of &DRM_COLOROP_1D_CURVE_SRGB_EOTF
+ */
+ DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+
+ /**
+ * @DRM_COLOROP_1D_CURVE_COUNT:
+ *
+ * enum value denoting the size of the enum
+ */
+ DRM_COLOROP_1D_CURVE_COUNT
+};
+
/**
* struct drm_colorop_state - mutable colorop state
*/
@@ -39,6 +72,13 @@ 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;
};
@@ -109,6 +149,14 @@ struct drm_colorop {
* this color operation. The type is enum drm_colorop_type.
*/
struct drm_property *type_property;
+
+ /**
+ * @curve_1d_type_property:
+ *
+ * Sub-type for DRM_COLOROP_1D_CURVE type.
+ */
+ struct drm_property *curve_1d_type_property;
+
};
#define obj_to_colorop(x) container_of(x, struct drm_colorop, base)
@@ -132,6 +180,9 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
return mo ? obj_to_colorop(mo) : NULL;
}
+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);
@@ -172,4 +223,13 @@ static inline unsigned int drm_colorop_index(const struct drm_colorop *colorop)
*/
const char *drm_get_colorop_type_name(enum drm_colorop_type type);
+/**
+ * drm_get_colorop_curve_1d_type_name - return a string for 1D curve type
+ * @type: 1d curve type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
+
#endif /* __DRM_COLOROP_H__ */
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 08/44] Documentation/gpu: document drm_colorop
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (6 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 07/44] drm/colorop: Add 1D Curve subtype Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 09/44] drm/colorop: Add BYPASS property Harry Wentland
` (37 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
Documentation/gpu/drm-kms.rst | 15 +++++++++++++++
drivers/gpu/drm/drm_colorop.c | 31 +++++++++++++++++++++++++++++++
2 files changed, 46 insertions(+)
diff --git a/Documentation/gpu/drm-kms.rst b/Documentation/gpu/drm-kms.rst
index abfe220764e1..2292e65f044c 100644
--- a/Documentation/gpu/drm-kms.rst
+++ b/Documentation/gpu/drm-kms.rst
@@ -413,6 +413,21 @@ Plane Panic Functions Reference
.. kernel-doc:: drivers/gpu/drm/drm_panic.c
:export:
+Colorop Abstraction
+===================
+
+.. kernel-doc:: drivers/gpu/drm/drm_colorop.c
+ :doc: overview
+
+Colorop Functions Reference
+---------------------------
+
+.. kernel-doc:: include/drm/drm_colorop.h
+ :internal:
+
+.. kernel-doc:: drivers/gpu/drm/drm_colorop.c
+ :export:
+
Display Modes Function Reference
================================
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 98d22ac163e8..5b844daabd44 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -30,6 +30,37 @@
#include "drm_crtc_internal.h"
+/**
+ * DOC: overview
+ *
+ * A colorop represents a single color operation. Colorops are chained
+ * via the NEXT property and make up color pipelines. Color pipelines
+ * are advertised and selected via the COLOR_PIPELINE &drm_plane
+ * property.
+ *
+ * A colorop will be of a certain type, advertised by the read-only TYPE
+ * property. Each type of colorop will advertise a different set of
+ * properties and is programmed in a different manner. Types can be
+ * enumerated 1D curves, 1D LUTs, 3D LUTs, matrices, etc. See the
+ * &drm_colorop_type documentation for information on each type.
+ *
+ * If a colorop advertises the BYPASS property it can be bypassed.
+ *
+ * Since colorops cannot stand-alone and are used to describe colorop
+ * operations on a plane they don't have their own locking mechanism but
+ * are locked and programmed along with their associated &drm_plane.
+ *
+ * Colorops are only advertised and valid for atomic drivers and atomic
+ * userspace that signals the DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE client
+ * cap. When a driver advertises the COLOR_PIPELINE property on a
+ * &drm_plane and userspace signals the
+ * DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE the driver shall ignore all other
+ * plane color properties, such as COLOR_ENCODING and COLOR_RANGE.
+ *
+ * More information about colorops and color pipelines can be found at
+ * rfc/color_pipeline.rst.
+ */
+
static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
{ DRM_COLOROP_1D_CURVE, "1D Curve" },
};
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 09/44] drm/colorop: Add BYPASS property
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (7 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 08/44] Documentation/gpu: document drm_colorop Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 10/44] drm/colorop: Add NEXT property Harry Wentland
` (36 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Drop TODO
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
drivers/gpu/drm/drm_atomic_uapi.c | 6 +++++-
drivers/gpu/drm/drm_colorop.c | 15 +++++++++++++++
include/drm/drm_colorop.h | 20 ++++++++++++++++++++
3 files changed, 40 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index b867bff49f67..419813be21b0 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 5b844daabd44..054e7f01151b 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -109,6 +109,17 @@ static int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
colorop->type_property,
colorop->type);
+ /* bypass */
+ prop = drm_property_create_bool(dev, DRM_MODE_PROP_ATOMIC,
+ "BYPASS");
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->bypass_property = prop;
+ drm_object_attach_property(&colorop->base,
+ colorop->bypass_property,
+ 1);
+
return ret;
}
@@ -166,6 +177,7 @@ 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;
@@ -182,6 +194,8 @@ 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 *
@@ -220,6 +234,7 @@ static void __drm_colorop_state_reset(struct drm_colorop_state *colorop_state,
u64 val;
colorop_state->colorop = colorop;
+ colorop_state->bypass = true;
if (colorop->curve_1d_type_property) {
drm_object_property_get_default_value(&colorop->base,
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index dd6943117461..7ecf4e3dca4f 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -72,6 +72,14 @@ struct drm_colorop_state {
/* colorop properties */
+ /**
+ * @bypass:
+ *
+ * True if colorop shall be bypassed. False if colorop is
+ * enabled.
+ */
+ bool bypass;
+
/**
* @curve_1d_type:
*
@@ -150,6 +158,18 @@ struct drm_colorop {
*/
struct drm_property *type_property;
+ /**
+ * @bypass_property:
+ *
+ * Boolean property to control enablement of the color
+ * operation. Setting bypass to "true" shall always be supported
+ * in order to allow compositors to quickly fall back to
+ * alternate methods of color processing. This is important
+ * since setting color operations can fail due to unique
+ * HW constraints.
+ */
+ struct drm_property *bypass_property;
+
/**
* @curve_1d_type_property:
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 10/44] drm/colorop: Add NEXT property
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (8 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 09/44] drm/colorop: Add BYPASS property Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 11/44] drm/colorop: Add atomic state print for drm_colorop Harry Wentland
` (35 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- move next comment here from Add 3x4 CTM patch (Sebastian)
- Add kernel doc
v4:
- Allow setting of NEXT property to NULL (Chaitanya Kumar Borah)
v3:
- Add next pointer to colorop to be used by drivers
and in DRM core
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
drivers/gpu/drm/drm_colorop.c | 30 ++++++++++++++++++++++++++++++
include/drm/drm_colorop.h | 20 ++++++++++++++++++++
2 files changed, 50 insertions(+)
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 054e7f01151b..685dd18126aa 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -88,6 +88,7 @@ static 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++;
@@ -120,6 +121,16 @@ static 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)
+ return -ENOMEM;
+ colorop->next_property = prop;
+ drm_object_attach_property(&colorop->base,
+ colorop->next_property,
+ 0);
+
return ret;
}
@@ -293,3 +304,22 @@ 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 7ecf4e3dca4f..5255cf44f4b1 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -150,6 +150,14 @@ struct drm_colorop {
*/
enum drm_colorop_type type;
+ /**
+ * @next:
+ *
+ * Read-only
+ * Pointer to next drm_colorop in pipeline
+ */
+ struct drm_colorop *next;
+
/**
* @type_property:
*
@@ -177,10 +185,20 @@ 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_colorop_find - look up a Colorop object from its ID
* @dev: DRM device
@@ -252,4 +270,6 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type);
*/
const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
+void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
+
#endif /* __DRM_COLOROP_H__ */
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 11/44] drm/colorop: Add atomic state print for drm_colorop
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (9 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 10/44] drm/colorop: Add NEXT property Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 12/44] drm/plane: Add COLOR PIPELINE property Harry Wentland
` (34 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 ++++++++++++++++++++++++-
1 file changed, 24 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index b09db9835e1f..096460d38545 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -780,6 +780,19 @@ 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)
{
@@ -801,7 +814,8 @@ 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);
}
@@ -1848,6 +1862,7 @@ 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;
@@ -1857,6 +1872,14 @@ static void __drm_state_dump(struct drm_device *dev, struct drm_printer *p,
if (!drm_drv_uses_atomic_modeset(dev))
return;
+ list_for_each_entry(colorop, &config->colorop_list, head) {
+ if (take_locks)
+ drm_modeset_lock(&colorop->plane->mutex, NULL);
+ drm_atomic_colorop_print_state(p, colorop->state);
+ if (take_locks)
+ drm_modeset_unlock(&colorop->plane->mutex);
+ }
+
list_for_each_entry(plane, &config->plane_list, head) {
if (take_locks)
drm_modeset_lock(&plane->mutex, NULL);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 12/44] drm/plane: Add COLOR PIPELINE property
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (10 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 11/44] drm/colorop: Add atomic state print for drm_colorop Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 13/44] drm/colorop: Add NEXT to colorop state print Harry Wentland
` (33 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 096460d38545..d333433319f6 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -1466,6 +1466,52 @@ 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 419813be21b0..d8d5be358b3e 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 a28b22fdd7a4..e3a85a27fa7a 100644
--- a/drivers/gpu/drm/drm_plane.c
+++ b/drivers/gpu/drm/drm_plane.c
@@ -1793,3 +1793,55 @@ int drm_plane_add_size_hints_property(struct drm_plane *plane,
return 0;
}
EXPORT_SYMBOL(drm_plane_add_size_hints_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);
diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index d2b44121b7fb..d8232fbb002b 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -852,6 +852,9 @@ drm_atomic_add_affected_connectors(struct drm_atomic_state *state,
int __must_check
drm_atomic_add_affected_planes(struct drm_atomic_state *state,
struct drm_crtc *crtc);
+int __must_check
+drm_atomic_add_affected_colorops(struct drm_atomic_state *state,
+ struct drm_plane *plane);
int __must_check drm_atomic_check_only(struct drm_atomic_state *state);
int __must_check drm_atomic_commit(struct drm_atomic_state *state);
diff --git a/include/drm/drm_atomic_uapi.h b/include/drm/drm_atomic_uapi.h
index 70a115d523cd..436315523326 100644
--- a/include/drm/drm_atomic_uapi.h
+++ b/include/drm/drm_atomic_uapi.h
@@ -50,6 +50,8 @@ drm_atomic_set_crtc_for_plane(struct drm_plane_state *plane_state,
struct drm_crtc *crtc);
void drm_atomic_set_fb_for_plane(struct drm_plane_state *plane_state,
struct drm_framebuffer *fb);
+void drm_atomic_set_colorop_for_plane(struct drm_plane_state *plane_state,
+ struct drm_colorop *colorop);
int __must_check
drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
struct drm_crtc *crtc);
diff --git a/include/drm/drm_plane.h b/include/drm/drm_plane.h
index 2a9b902623c0..8a769c6a8251 100644
--- a/include/drm/drm_plane.h
+++ b/include/drm/drm_plane.h
@@ -774,6 +774,14 @@ struct drm_plane {
*/
struct drm_property *color_range_property;
+ /**
+ * @color_pipeline_property:
+ *
+ * Optional "COLOR_PIPELINE" enum property for specifying
+ * a color pipeline to use on the plane.
+ */
+ struct drm_property *color_pipeline_property;
+
/**
* @scaling_filter_property: property to apply a particular filter while
* scaling.
@@ -997,4 +1005,7 @@ int drm_plane_add_size_hints_property(struct drm_plane *plane,
const struct drm_plane_size_hint *hints,
int num_hints);
+int drm_plane_create_color_pipeline_property(struct drm_plane *plane,
+ struct drm_prop_enum_list *pipelines,
+ int num_pipelines);
#endif
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 13/44] drm/colorop: Add NEXT to colorop state print
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (11 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 12/44] drm/plane: Add COLOR PIPELINE property Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 14/44] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
` (32 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland
v5:
- Drop unused header definitions
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 +
1 file changed, 1 insertion(+)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index d333433319f6..9919a6f1092f 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -791,6 +791,7 @@ 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,
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 14/44] drm/vkms: Add enumerated 1D curve colorop
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (12 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 13/44] drm/colorop: Add NEXT to colorop state print Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 15/44] drm/vkms: Add kunit tests for linear and sRGB LUTs Harry Wentland
` (31 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Squash with "Pull apply_colorop out of pre_blend_color_transform"
(Sebastian)
- Fix warnings
- Fix include
- Drop TODOs
v4:
- Drop _tf_ from color_pipeline init function
- Pass supported TFs into colorop init
- Create bypass pipeline in DRM helper (Pekka)
v2:
- Add commit description
- Fix sRGB EOTF LUT definition
- Add linear and sRGB inverse EOTF LUTs
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 | 72 +++
drivers/gpu/drm/vkms/vkms_composer.c | 51 ++
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, 946 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..f61dfde47156
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -0,0 +1,72 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#include <linux/slab.h>
+#include <drm/drm_colorop.h>
+#include <drm/drm_print.h>
+#include <drm/drm_property.h>
+#include <drm/drm_plane.h>
+
+#include "vkms_drv.h"
+
+#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);
+
+static 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 3ecda70c2b55..bc116d16e378 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 +164,53 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
}
}
+static void apply_colorop(struct pixel_argb_u16 *pixel, struct drm_colorop *colorop)
+{
+ struct drm_colorop_state *colorop_state = colorop->state;
+
+ 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;
+
+ while (colorop) {
+ struct drm_colorop_state *colorop_state;
+
+ if (!colorop)
+ return;
+
+ 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);
+
+ colorop = colorop->next;
+ }
+}
+
/**
* blend - blend the pixels from all planes and compute crc
* @wb: The writeback frame buffer metadata
@@ -200,6 +248,9 @@ 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 5e46ea5b96dc..278cf3533f58 100644
--- a/drivers/gpu/drm/vkms/vkms_drv.h
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -169,4 +169,8 @@ 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;
}
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 15/44] drm/vkms: Add kunit tests for linear and sRGB LUTs
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (13 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 14/44] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 16/44] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Harry Wentland
` (30 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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),
{}
};
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 16/44] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (14 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 15/44] drm/vkms: Add kunit tests for linear and sRGB LUTs Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 17/44] drm/colorop: Add 3x4 CTM type Harry Wentland
` (29 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Fix kernel docs
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 d8d5be358b3e..5dd834e3979b 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);
@@ -1186,6 +1192,12 @@ 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;
@@ -1198,7 +1210,6 @@ 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 51f39912866f..6b5021b68e35 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..a2fc85ecdea1 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 the COLOR_PIPELINE
+ * property on a &drm_plane, as well as drm_colorop properties.
+ *
+ * Drivers will ignore these drm_plane 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;
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 17/44] drm/colorop: Add 3x4 CTM type
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (15 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 16/44] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 18/44] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
` (28 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Add function signature for init (Sebastian)
- Fix kernel-doc
v4:
- Create helper function for creating 3x4 CTM colorop
- Fix CTM indexes in comment (Pekka)
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 | 39 +++++++++++++++++++++++++++++++
include/drm/drm_colorop.h | 21 +++++++++++++++++
include/uapi/drm/amdgpu_drm.h | 9 -------
include/uapi/drm/drm_mode.h | 33 ++++++++++++++++++++++----
6 files changed, 131 insertions(+), 14 deletions(-)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 9919a6f1092f..c2fcde96e218 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -790,7 +790,19 @@ 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 5dd834e3979b..4a9cb1f7dbf7 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 685dd18126aa..73fd4b7a4ce3 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -63,6 +63,7 @@
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[] = {
@@ -201,6 +202,43 @@ 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)
{
@@ -287,6 +325,7 @@ void drm_colorop_reset(struct drm_colorop *colorop)
static const char * const colorop_type_name[] = {
[DRM_COLOROP_1D_CURVE] = "1D Curve",
+ [DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
};
const char *drm_get_colorop_type_name(enum drm_colorop_type type)
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 5255cf44f4b1..1566d88fd018 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -87,6 +87,14 @@ 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;
};
@@ -185,6 +193,17 @@ struct drm_colorop {
*/
struct drm_property *curve_1d_type_property;
+ /**
+ * @data_property:
+ *
+ * blob property for any TYPE that requires a blob of data,
+ * such as 1DLUT, CTM, 3DLUT, etc.
+ *
+ * The way this blob is interpreted depends on the TYPE of
+ * this
+ */
+ struct drm_property *data_property;
+
/**
* @next_property:
*
@@ -220,6 +239,8 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
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);
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index efe5de6ce208..e44362e74fa1 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -1283,15 +1283,6 @@ struct drm_amdgpu_info_gpuvm_fault {
#define AMDGPU_FAMILY_GC_11_5_0 150 /* GC 11.5.0 */
#define AMDGPU_FAMILY_GC_12_0_0 152 /* GC 12.0.0 */
-/* FIXME wrong namespace! */
-struct drm_color_ctm_3x4 {
- /*
- * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude
- * (not two's complement!) format.
- */
- __u64 matrix[12];
-};
-
#if defined(__cplusplus)
}
#endif
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index 3dd89ae6515a..da2be67609c8 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -847,6 +847,20 @@ struct drm_color_ctm {
__u64 matrix[9];
};
+struct drm_color_ctm_3x4 {
+ /*
+ * Conversion matrix with 3x4 dimensions in S31.32 sign-magnitude
+ * (not two's complement!) format.
+ *
+ * out matrix in
+ * |R| |0 1 2 3 | | R |
+ * |G| = |4 5 6 7 | x | G |
+ * |B| |8 9 10 11| | B |
+ * |1.0|
+ */
+ __u64 matrix[12];
+};
+
struct drm_color_lut {
/*
* Values are mapped linearly to 0.0 - 1.0 range, with 0x0 == 0.0 and
@@ -861,16 +875,27 @@ struct drm_color_lut {
/**
* enum drm_colorop_type - Type of color operation
*
- * @DRM_COLOROP_1D_CURVE: A 1D curve that is being applied to all color channels.
- * The curve is specified via tha CURVE_1D_TYPE colorop
- * property.
*
* drm_colorops can be of many different types. Each type behaves differently
* and defines a different set of properties. This enum defines all types and
* gives a high-level description.
*/
enum drm_colorop_type {
- DRM_COLOROP_1D_CURVE
+ /**
+ * @DRM_COLOROP_1D_CURVE:
+ *
+ * A 1D curve that is being applied to all color channels. The
+ * curve is specified via tha CURVE_1D_TYPE colorop property.
+ */
+ DRM_COLOROP_1D_CURVE,
+
+ /**
+ * @DRM_COLOROP_CTM_3X4:
+ *
+ * A 3x4 matrix. Its values are specified via the
+ * &drm_color_ctm_3x4 struct provided via the DATA property.
+ */
+ DRM_COLOROP_CTM_3X4,
};
/**
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 18/44] drm/vkms: Use s32 for internal color pipeline precision
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (16 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 17/44] drm/colorop: Add 3x4 CTM type Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
` (27 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 | 55 ++++++++++++++++++++++------
drivers/gpu/drm/vkms/vkms_drv.h | 4 ++
2 files changed, 47 insertions(+), 12 deletions(-)
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index bc116d16e378..6e939d3a6d5c 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)
{
struct drm_colorop_state *colorop_state = colorop->state;
@@ -190,24 +190,55 @@ 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;
- 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;
+
+ 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 278cf3533f58..b78bc2611746 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;
};
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (17 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 18/44] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
` (26 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 +++++++++++++++++++++++
2 files changed, 58 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
index f61dfde47156..adcb08153a09 100644
--- a/drivers/gpu/drm/vkms/vkms_colorop.c
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -37,7 +37,37 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
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 6e939d3a6d5c..bd1df06ced85 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)
{
struct drm_colorop_state *colorop_state = colorop->state;
@@ -184,6 +208,9 @@ 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);
}
}
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (18 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-21 1:07 ` Jeff Johnson
` (2 more replies)
2024-08-19 20:56 ` [PATCH v5 21/44] drm/vkms: Add tests for CTM handling Harry Wentland
` (25 subsequent siblings)
45 siblings, 3 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 56dab563abd7..bd69df0eee33 100644
--- a/drivers/gpu/drm/tests/Makefile
+++ b/drivers/gpu/drm/tests/Makefile
@@ -20,6 +20,7 @@ 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
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 21/44] drm/vkms: Add tests for CTM handling
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (19 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 22/44] drm/colorop: pass plane_color_pipeline client cap to atomic check Harry Wentland
` (24 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Make apply_3x4_matrix static
v4:
- Comment on origin of bt709_enc matrix (Pekka)
- Use full opaque alpha (Pekka)
- Add additional check for Y < 0xffff (Pekka)
- Remove unused code (Pekka)
- 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 ++++++++++++++++++
1 file changed, 251 insertions(+)
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),
{}
};
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 22/44] drm/colorop: pass plane_color_pipeline client cap to atomic check
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (20 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 21/44] drm/vkms: Add tests for CTM handling Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 23/44] drm/colorop: define a new macro for_each_new_colorop_in_state Harry Wentland
` (23 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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.
v5:
- Fix kernel docs
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
drivers/gpu/drm/drm_atomic_uapi.c | 1 +
include/drm/drm_atomic.h | 18 ++++++++++++++++++
2 files changed, 19 insertions(+)
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 4a9cb1f7dbf7..39805f822d44 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -1578,6 +1578,7 @@ 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 d8232fbb002b..c908ad75b95a 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -414,6 +414,24 @@ struct drm_atomic_state {
*/
bool duplicated : 1;
+ /**
+ * @plane_color_pipeline:
+ *
+ * Indicates whether this atomic state originated with a client that
+ * set the DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE.
+ *
+ * Drivers and helper functions should use this to ignore legacy
+ * properties that are incompatible with the drm_plane COLOR_PIPELINE
+ * behavior, such as:
+ *
+ * - COLOR_RANGE
+ * - COLOR_ENCODING
+ *
+ * or any other driver-specific properties that might affect pixel
+ * values.
+ */
+ bool plane_color_pipeline : 1;
+
/**
* @colorops:
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 23/44] drm/colorop: define a new macro for_each_new_colorop_in_state
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (21 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 22/44] drm/colorop: pass plane_color_pipeline client cap to atomic check Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 24/44] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Harry Wentland
` (22 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 c908ad75b95a..ed767cb7a55b 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -1035,6 +1035,26 @@ 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
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 24/44] drm/amd/display: Ignore deprecated props when plane_color_pipeline set
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (22 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 23/44] drm/colorop: define a new macro for_each_new_colorop_in_state Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 25/44] drm/amd/display: Add bypass COLOR PIPELINE Harry Wentland
` (21 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 7e7929f24ae4..3257bd22bf6b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -5319,6 +5319,10 @@ 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;
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 25/44] drm/amd/display: Add bypass COLOR PIPELINE
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (23 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 24/44] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 26/44] drm/amd/display: Skip color pipeline initialization for cursor plane Harry Wentland
` (20 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 a83bd0331c3b..d260db42d407 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
@@ -1746,6 +1746,20 @@ 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,
@@ -1849,7 +1863,12 @@ 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);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 26/44] drm/amd/display: Skip color pipeline initialization for cursor plane
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (24 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 25/44] drm/amd/display: Add bypass COLOR PIPELINE Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 27/44] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
` (19 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 d260db42d407..8fc0f73fce33 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
@@ -1754,6 +1754,9 @@ 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);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 27/44] drm/amd/display: Add support for sRGB EOTF in DEGAM block
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (25 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 26/44] drm/amd/display: Skip color pipeline initialization for cursor plane Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
` (18 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 ebabfe3a512f..adb1e4b3522d 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 8fc0f73fce33..8f950ca1f5ba 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"
@@ -1753,10 +1754,19 @@ 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);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (26 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 27/44] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-21 4:52 ` kernel test robot
2024-08-19 20:56 ` [PATCH v5 29/44] drm/amd/display: Add support for sRGB EOTF in BLND block Harry Wentland
` (17 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 adb1e4b3522d..58a996090509 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);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 29/44] drm/amd/display: Add support for sRGB EOTF in BLND block
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (27 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 30/44] drm/colorop: Add PQ 125 EOTF and its inverse Harry Wentland
` (16 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 58a996090509..4009895c3b6d 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);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 30/44] drm/colorop: Add PQ 125 EOTF and its inverse
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (28 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 29/44] drm/amd/display: Add support for sRGB EOTF in BLND block Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 31/44] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Harry Wentland
` (15 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 | 21 ++++++++++++++++++++-
2 files changed, 22 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 73fd4b7a4ce3..f9dcc522c94d 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -69,6 +69,8 @@ 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 1566d88fd018..d953d16c32f7 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -35,7 +35,6 @@
* enum drm_colorop_curve_1d_type - type of 1D curve
*
* Describes a 1D curve to be applied by the DRM_COLOROP_1D_CURVE colorop.
- *
*/
enum drm_colorop_curve_1d_type {
/**
@@ -55,6 +54,26 @@ enum drm_colorop_curve_1d_type {
*/
DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+ /**
+ * @DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
+ *
+ * The PQ transfer function, scaled by 125.0f, so that 10,000
+ * nits correspond to 125.0f.
+ *
+ * Transfer characteristics of the PQ function as defined by
+ * SMPTE ST 2084 (2014) for 10-, 12-, 14-, and 16-bit systems
+ * and Rec. ITU-R BT.2100-2 perceptual quantization (PQ) system.
+ * Equivalent to H.273 TransferCharacteristics code point 16.
+ */
+ DRM_COLOROP_1D_CURVE_PQ_125_EOTF,
+
+ /**
+ * @DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF:
+ *
+ * The inverse of DRM_COLOROP_1D_CURVE_PQ_125_EOTF.
+ */
+ DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF,
+
/**
* @DRM_COLOROP_1D_CURVE_COUNT:
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 31/44] drm/amd/display: Enable support for PQ 125 EOTF and Inverse
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (29 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 30/44] drm/colorop: Add PQ 125 EOTF and its inverse Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 32/44] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Harry Wentland
` (14 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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 4009895c3b6d..a1299ca83bac 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);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 32/44] drm/colorop: add BT2020/BT709 OETF and Inverse OETF
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (30 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 31/44] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Harry Wentland
@ 2024-08-19 20:56 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 33/44] drm/amd/display: Add support for BT.709 and BT.2020 TFs Harry Wentland
` (13 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:56 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).
v5:
- Add kernel docs
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
drivers/gpu/drm/drm_colorop.c | 2 ++
include/drm/drm_colorop.h | 19 +++++++++++++++++++
2 files changed, 21 insertions(+)
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index f9dcc522c94d..032ca65299e5 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -69,6 +69,8 @@ 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 d953d16c32f7..d8d0e4b571ca 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -54,6 +54,25 @@ enum drm_colorop_curve_1d_type {
*/
DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF,
+ /**
+ * @DRM_COLOROP_1D_CURVE_BT2020_INV_OETF:
+ *
+ * The inverse of &DRM_COLOROP_1D_CURVE_BT2020_OETF
+ */
+ DRM_COLOROP_1D_CURVE_BT2020_INV_OETF,
+
+ /**
+ * @DRM_COLOROP_1D_CURVE_BT2020_OETF:
+ *
+ * The BT.2020/BT.709 transfer function. The BT.709 and BT.2020
+ * transfer functions are the same, the only difference is that
+ * BT.2020 is defined with more precision for 10 and 12-bit
+ * encodings.
+ *
+ *
+ */
+ DRM_COLOROP_1D_CURVE_BT2020_OETF,
+
/**
* @DRM_COLOROP_1D_CURVE_PQ_125_EOTF:
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 33/44] drm/amd/display: Add support for BT.709 and BT.2020 TFs
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (31 preceding siblings ...)
2024-08-19 20:56 ` [PATCH v5 32/44] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 34/44] drm/colorop: Add 1D Curve Custom LUT type Harry Wentland
` (12 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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 a1299ca83bac..8f1721bdc243 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)
{
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 34/44] drm/colorop: Add 1D Curve Custom LUT type
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (32 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 33/44] drm/amd/display: Add support for BT.709 and BT.2020 TFs Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 35/44] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Harry Wentland
` (11 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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.
v5:
- Add kernel doc
- Define SIZE in similar manner to GAMMA_SIZE on drm_crtc (Melissa)
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 | 47 +++++++++++++++++++++++++++++--
include/drm/drm_colorop.h | 16 +++++++++++
include/uapi/drm/drm_mode.h | 9 ++++++
5 files changed, 79 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index c2fcde96e218..10af940de8de 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -796,6 +796,10 @@ 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 39805f822d44..5bbce7360262 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 032ca65299e5..e760d3301d48 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -63,6 +63,7 @@
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"}
};
@@ -218,12 +219,53 @@ 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;
}
+/**
+ * drm_colorop_curve_1d_lut_init - Initialize a DRM_COLOROP_1D_LUT
+ *
+ * @dev: DRM device
+ * @colorop: The drm_colorop object to initialize
+ * @plane: The associated drm_plane
+ * @lut_size: LUT size supported by driver
+ * @return zero on success, -E value on failure
+ */
+int drm_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_IMMUTABLE, "SIZE",
+ 0, UINT_MAX);
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->size_property = prop;
+ drm_object_attach_property(&colorop->base, colorop->size_property, lut_size);
+
+ /* 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)
{
@@ -329,6 +371,7 @@ void drm_colorop_reset(struct drm_colorop *colorop)
static const char * const colorop_type_name[] = {
[DRM_COLOROP_1D_CURVE] = "1D Curve",
+ [DRM_COLOROP_1D_LUT] = "1D Curve Custom LUT",
[DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
};
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index d8d0e4b571ca..6e5da057ef44 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -125,6 +125,13 @@ 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:
*
@@ -231,6 +238,13 @@ 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_property:
*
@@ -277,6 +291,8 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
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 da2be67609c8..83bf461558b5 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -889,6 +889,15 @@ enum drm_colorop_type {
*/
DRM_COLOROP_1D_CURVE,
+ /**
+ * @DRM_COLOROP_1D_LUT:
+ *
+ * A simple 1D LUT of uniformly spaced &drm_color_lut entries,
+ * packed into a blob via the DATA property. The driver's expected
+ * LUT size is advertised via the SIZE property.
+ */
+ DRM_COLOROP_1D_LUT,
+
/**
* @DRM_COLOROP_CTM_3X4:
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 35/44] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (33 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 34/44] drm/colorop: Add 1D Curve Custom LUT type Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop Harry Wentland
` (10 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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 8f1721bdc243..c66f249ae70b 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
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (34 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 35/44] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-21 3:19 ` kernel test robot
2024-08-21 9:22 ` kernel test robot
2024-08-19 20:57 ` [PATCH v5 37/44] drm/colorop: Add mutliplier type Harry Wentland
` (9 subsequent siblings)
45 siblings, 2 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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 c66f249ae70b..ea9bd287cddd 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) {
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 37/44] drm/colorop: Add mutliplier type
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (35 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 38/44] drm/amd/display: add multiplier colorop Harry Wentland
` (8 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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.
v5:
- Fix atomic state print
- Add kernel doc
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 | 37 +++++++++++++++++++++++++++++--
include/drm/drm_colorop.h | 16 +++++++++++++
include/uapi/drm/drm_mode.h | 9 ++++++++
5 files changed, 67 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 10af940de8de..2ab797e9d946 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -803,6 +803,9 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
case DRM_COLOROP_CTM_3X4:
drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
break;
+ case DRM_COLOROP_MULTIPLIER:
+ drm_printf(p, "\tmultiplier=%llu\n", state->multiplier);
+ break;
default:
break;
}
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 5bbce7360262..142824de33ca 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 e760d3301d48..9634effab73a 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -64,7 +64,8 @@
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[] = {
@@ -285,6 +286,37 @@ int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop
}
EXPORT_SYMBOL(drm_colorop_ctm_3x4_init);
+/**
+ * drm_colorop_mult_init - Initialize a DRM_COLOROP_MULTIPLIER
+ *
+ * @dev: DRM device
+ * @colorop: The drm_colorop object to initialize
+ * @plane: The associated drm_plane
+ * @return zero on success, -E value on failure
+ */
+int drm_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)
{
@@ -372,7 +404,8 @@ void drm_colorop_reset(struct drm_colorop *colorop)
static const char * const colorop_type_name[] = {
[DRM_COLOROP_1D_CURVE] = "1D Curve",
[DRM_COLOROP_1D_LUT] = "1D Curve Custom LUT",
- [DRM_COLOROP_CTM_3X4] = "3x4 Matrix"
+ [DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
+ [DRM_COLOROP_MULTIPLIER] = "Multiplier",
};
const char *drm_get_colorop_type_name(enum drm_colorop_type type)
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 6e5da057ef44..09e96b9a8eff 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -125,6 +125,13 @@ 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:
*
@@ -238,6 +245,13 @@ struct drm_colorop {
*/
struct drm_property *curve_1d_type_property;
+ /**
+ * @multiplier_property:
+ *
+ * Multiplier property for plane gain
+ */
+ struct drm_property *multiplier_property;
+
/**
* @size_property:
*
@@ -295,6 +309,8 @@ 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 83bf461558b5..8e9886200f4d 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -905,6 +905,15 @@ enum drm_colorop_type {
* &drm_color_ctm_3x4 struct provided via the DATA property.
*/
DRM_COLOROP_CTM_3X4,
+
+ /**
+ * @DRM_COLOROP_MULTIPLIER:
+ *
+ * A simple multiplier, applied to all color values. The
+ * multiplier is specified as a S31.32 via the MULTIPLIER
+ * property.
+ */
+ DRM_COLOROP_MULTIPLIER,
};
/**
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 38/44] drm/amd/display: add multiplier colorop
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (36 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 37/44] drm/colorop: Add mutliplier type Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 39/44] drm/amd/display: Swap matrix and multiplier Harry Wentland
` (7 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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 ea9bd287cddd..a14dc67dbbdf 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) {
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 39/44] drm/amd/display: Swap matrix and multiplier
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (37 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 38/44] drm/amd/display: add multiplier colorop Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 40/44] drm/colorop: Define LUT_1D interpolation Harry Wentland
` (6 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung
From: Alex Hung <alex.hung@amd.com>
Swap the order of matrix and multiplier as designed in hardware.
Signed-off-by: Alex Hung <alex.hung@amd.com>
---
.../amd/display/amdgpu_dm/amdgpu_dm_color.c | 22 +++++++-------
.../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 30 +++++++++----------
2 files changed, 26 insertions(+), 26 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 a14dc67dbbdf..2f0b06fb5f81 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
@@ -1479,17 +1479,6 @@ 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;
-
/* Multiplier */
colorop = colorop->next;
if (!colorop) {
@@ -1501,6 +1490,17 @@ 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 b739d6cb3e6b..8a86849e1529 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
@@ -69,21 +69,6 @@ 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;
-
/* Multiplier */
op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
if (!op) {
@@ -99,6 +84,21 @@ 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) {
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 40/44] drm/colorop: Define LUT_1D interpolation
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (38 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 39/44] drm/amd/display: Swap matrix and multiplier Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 41/44] drm/colorop: allow non-bypass colorops Harry Wentland
` (5 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland
We want to make sure userspace is aware of the 1D LUT
interpolation. While linear interpolation is common it
might not be supported on all HW. Give driver implementers
a way to specify their interpolation.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
.../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 6 ++--
drivers/gpu/drm/drm_atomic.c | 1 +
drivers/gpu/drm/drm_colorop.c | 36 ++++++++++++++++++-
include/drm/drm_colorop.h | 19 +++++++++-
include/uapi/drm/drm_mode.h | 13 +++++++
5 files changed, 71 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 8a86849e1529..c85e344eaccb 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
@@ -121,7 +121,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES);
+ ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES,
+ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
if (ret)
return ret;
@@ -151,7 +152,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES);
+ ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES,
+ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
if (ret)
return ret;
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 2ab797e9d946..dbb8776ae433 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -798,6 +798,7 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
break;
case DRM_COLOROP_1D_LUT:
drm_printf(p, "\tsize=%d\n", state->size);
+ drm_printf(p, "\tinterpolation=%s\n", drm_get_colorop_lut1d_interpolation_name(colorop->lut1d_interpolation));
drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
break;
case DRM_COLOROP_CTM_3X4:
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 9634effab73a..47a9877099c3 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -77,6 +77,9 @@ static const char * const colorop_curve_1d_type_names[] = {
[DRM_COLOROP_1D_CURVE_PQ_125_INV_EOTF] = "PQ 125 Inverse EOTF",
};
+static const struct drm_prop_enum_list drm_colorop_lut1d_interpolation_list[] = {
+ { DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR, "Linear" },
+};
/* Init Helpers */
@@ -233,10 +236,12 @@ static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_color
* @colorop: The drm_colorop object to initialize
* @plane: The associated drm_plane
* @lut_size: LUT size supported by driver
+ * @lut1d_interpolation: 1D LUT interpolation type
* @return zero on success, -E value on failure
*/
int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane, uint32_t lut_size)
+ struct drm_plane *plane, uint32_t lut_size,
+ enum drm_colorop_lut1d_interpolation_type lut1d_interpolation)
{
struct drm_property *prop;
int ret;
@@ -255,6 +260,17 @@ int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *co
colorop->size_property = prop;
drm_object_attach_property(&colorop->base, colorop->size_property, lut_size);
+ /* Interpolation */
+ prop = drm_property_create_enum(dev, DRM_MODE_PROP_IMMUTABLE, "LUT1D_INTERPOLATION",
+ drm_colorop_lut1d_interpolation_list,
+ ARRAY_SIZE(drm_colorop_lut1d_interpolation_list));
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->lut1d_interpolation_property = prop;
+ drm_object_attach_property(&colorop->base, prop, lut1d_interpolation);
+ colorop->lut1d_interpolation = lut1d_interpolation;
+
/* data */
ret = drm_colorop_create_data_prop(dev, colorop);
if (ret)
@@ -407,6 +423,9 @@ static const char * const colorop_type_name[] = {
[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
[DRM_COLOROP_MULTIPLIER] = "Multiplier",
};
+static const char * const colorop_lut1d_interpolation_name[] = {
+ [DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR] = "Linear",
+};
const char *drm_get_colorop_type_name(enum drm_colorop_type type)
{
@@ -424,6 +443,21 @@ const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type ty
return colorop_curve_1d_type_names[type];
}
+/**
+ * drm_get_colorop_lut1d_interpolation_name: return a string for interpolation type
+ * @type: interpolation type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_interpolation_type type)
+{
+ if (WARN_ON(type >= ARRAY_SIZE(colorop_lut1d_interpolation_name)))
+ return "unknown";
+
+ return colorop_lut1d_interpolation_name[type];
+}
+
/**
* drm_colorop_set_next_property - sets the next pointer
* @colorop: drm colorop
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 09e96b9a8eff..1304b19c96b5 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -238,6 +238,21 @@ struct drm_colorop {
*/
struct drm_property *bypass_property;
+ /**
+ * @lut1d_interpolation:
+ *
+ * Read-only
+ * Interpolation for DRM_COLOROP_1D_LUT
+ */
+ enum drm_colorop_lut1d_interpolation_type lut1d_interpolation;
+
+ /**
+ * @lut1d_interpolation_property:
+ *
+ * Read-only property for DRM_COLOROP_1D_LUT interpolation
+ */
+ struct drm_property *lut1d_interpolation_property;
+
/**
* @curve_1d_type_property:
*
@@ -306,7 +321,8 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
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);
+ struct drm_plane *plane, uint32_t lut_size,
+ enum drm_colorop_lut1d_interpolation_type lut1d_interpolation);
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,
@@ -360,6 +376,7 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type);
* const pointer and hence is threadsafe.
*/
const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
+const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_interpolation_type type);
void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index 8e9886200f4d..b64f9515bd0a 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -916,6 +916,19 @@ enum drm_colorop_type {
DRM_COLOROP_MULTIPLIER,
};
+/**
+ * enum drm_colorop_lut1d_interpolation_type - type of interpolation for 1D LUTs
+ */
+enum drm_colorop_lut1d_interpolation_type {
+ /**
+ * @DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR:
+ *
+ * Linear interpolation. Values between points of the LUT will be
+ * linearly interpolated.
+ */
+ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR,
+};
+
/**
* struct drm_plane_size_hint - Plane size hints
*
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 41/44] drm/colorop: allow non-bypass colorops
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (39 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 40/44] drm/colorop: Define LUT_1D interpolation Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:57 ` [PATCH v5 42/44] drm/colorop: Add 3D LUT supports to color pipeline Harry Wentland
` (4 subsequent siblings)
45 siblings, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland
Not all HW will be able to do bypass on all color
operations. Introduce an 'allow_bypass' boolean for
all colorop init functions and only create the BYPASS
property when it's true.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
.../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 22 +++++---
drivers/gpu/drm/drm_atomic.c | 3 +-
drivers/gpu/drm/drm_colorop.c | 51 ++++++++++++-------
drivers/gpu/drm/vkms/vkms_colorop.c | 8 +--
include/drm/drm_colorop.h | 10 ++--
5 files changed, 60 insertions(+), 34 deletions(-)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index c85e344eaccb..05fdec7c28ad 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
@@ -60,7 +60,9 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_degam_tfs);
+ ret = drm_colorop_curve_1d_init(dev, op, plane,
+ amdgpu_dm_supported_degam_tfs,
+ true);
if (ret)
return ret;
@@ -76,7 +78,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_mult_init(dev, op, plane);
+ ret = drm_colorop_mult_init(dev, op, plane, true);
if (ret)
return ret;
@@ -91,7 +93,7 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_ctm_3x4_init(dev, op, plane);
+ ret = drm_colorop_ctm_3x4_init(dev, op, plane, true);
if (ret)
return ret;
@@ -106,7 +108,9 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_shaper_tfs);
+ ret = drm_colorop_curve_1d_init(dev, op, plane,
+ amdgpu_dm_supported_shaper_tfs,
+ true);
if (ret)
return ret;
@@ -122,7 +126,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
}
ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES,
- DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
+ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR,
+ true);
if (ret)
return ret;
@@ -137,7 +142,9 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_init(dev, op, plane, amdgpu_dm_supported_blnd_tfs);
+ ret = drm_colorop_curve_1d_init(dev, op, plane,
+ amdgpu_dm_supported_blnd_tfs,
+ true);
if (ret)
return ret;
@@ -153,7 +160,8 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
}
ret = drm_colorop_curve_1d_lut_init(dev, op, plane, MAX_COLOR_LUT_ENTRIES,
- DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR);
+ DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR,
+ true);
if (ret)
return ret;
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index dbb8776ae433..02ff576aa7a9 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -789,7 +789,8 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
drm_printf(p, "colorop[%u]:\n", colorop->base.id);
drm_printf(p, "\ttype=%s\n", drm_get_colorop_type_name(colorop->type));
- drm_printf(p, "\tbypass=%u\n", state->bypass);
+ if (colorop->bypass_property)
+ drm_printf(p, "\tbypass=%u\n", state->bypass);
switch (colorop->type) {
case DRM_COLOROP_1D_CURVE:
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 47a9877099c3..2b7526c9608e 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -84,7 +84,8 @@ static const struct drm_prop_enum_list drm_colorop_lut1d_interpolation_list[] =
/* Init Helpers */
static int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane, enum drm_colorop_type type)
+ struct drm_plane *plane, enum drm_colorop_type type,
+ bool allow_bypass)
{
struct drm_mode_config *config = &dev->mode_config;
struct drm_property *prop;
@@ -120,16 +121,18 @@ static int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
colorop->type_property,
colorop->type);
- /* bypass */
- prop = drm_property_create_bool(dev, DRM_MODE_PROP_ATOMIC,
- "BYPASS");
- if (!prop)
- return -ENOMEM;
+ if (allow_bypass) {
+ /* bypass */
+ prop = drm_property_create_bool(dev, DRM_MODE_PROP_ATOMIC,
+ "BYPASS");
+ if (!prop)
+ return -ENOMEM;
- colorop->bypass_property = prop;
- drm_object_attach_property(&colorop->base,
- colorop->bypass_property,
- 1);
+ colorop->bypass_property = prop;
+ drm_object_attach_property(&colorop->base,
+ colorop->bypass_property,
+ 1);
+ }
/* next */
prop = drm_property_create_object(dev, DRM_MODE_PROP_IMMUTABLE | DRM_MODE_PROP_ATOMIC,
@@ -153,10 +156,13 @@ static int drm_colorop_init(struct drm_device *dev, struct drm_colorop *colorop,
* @supported_tfs: A bitfield of supported drm_colorop_curve_1d_init enum values,
* created using BIT(curve_type) and combined with the OR '|'
* operator.
+ * @allow_bypass: true if BYPASS property should be created, false if bypass of
+ * this colorop is not possible
* @return zero on success, -E value on failure
*/
int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane, u64 supported_tfs)
+ struct drm_plane *plane, u64 supported_tfs,
+ bool allow_bypass)
{
struct drm_prop_enum_list enum_list[DRM_COLOROP_1D_CURVE_COUNT];
int i, len;
@@ -177,7 +183,8 @@ int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *coloro
return -EINVAL;
}
- ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_CURVE);
+ ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_CURVE,
+ allow_bypass);
if (ret)
return ret;
@@ -237,16 +244,20 @@ static int drm_colorop_create_data_prop(struct drm_device *dev, struct drm_color
* @plane: The associated drm_plane
* @lut_size: LUT size supported by driver
* @lut1d_interpolation: 1D LUT interpolation type
+ * @allow_bypass: true if BYPASS property should be created, false if bypass of
+ * this colorop is not possible
* @return zero on success, -E value on failure
*/
int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
struct drm_plane *plane, uint32_t lut_size,
- enum drm_colorop_lut1d_interpolation_type lut1d_interpolation)
+ enum drm_colorop_lut1d_interpolation_type lut1d_interpolation,
+ bool allow_bypass)
{
struct drm_property *prop;
int ret;
- ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_LUT);
+ ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_1D_LUT,
+ allow_bypass);
if (ret)
return ret;
@@ -284,11 +295,12 @@ int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *co
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)
+ struct drm_plane *plane, bool allow_bypass)
{
int ret;
- ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_CTM_3X4);
+ ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_CTM_3X4,
+ allow_bypass);
if (ret)
return ret;
@@ -308,15 +320,18 @@ EXPORT_SYMBOL(drm_colorop_ctm_3x4_init);
* @dev: DRM device
* @colorop: The drm_colorop object to initialize
* @plane: The associated drm_plane
+ * @allow_bypass: true if BYPASS property should be created, false if bypass of
+ * this colorop is not possible
* @return zero on success, -E value on failure
*/
int drm_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane)
+ struct drm_plane *plane, bool allow_bypass)
{
struct drm_property *prop;
int ret;
- ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_MULTIPLIER);
+ ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_MULTIPLIER,
+ allow_bypass);
if (ret)
return ret;
diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
index adcb08153a09..0a87cb14b8e6 100644
--- a/drivers/gpu/drm/vkms/vkms_colorop.c
+++ b/drivers/gpu/drm/vkms/vkms_colorop.c
@@ -28,7 +28,7 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs);
+ ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs, true);
if (ret)
return ret;
@@ -44,7 +44,7 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_ctm_3x4_init(dev, op, plane);
+ ret = drm_colorop_ctm_3x4_init(dev, op, plane, true);
if (ret)
return ret;
@@ -59,7 +59,7 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_ctm_3x4_init(dev, op, plane);
+ ret = drm_colorop_ctm_3x4_init(dev, op, plane, true);
if (ret)
return ret;
@@ -74,7 +74,7 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
return -ENOMEM;
}
- ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs);
+ ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs, true);
if (ret)
return ret;
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 1304b19c96b5..044294b498e4 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -319,14 +319,16 @@ static inline struct drm_colorop *drm_colorop_find(struct drm_device *dev,
}
int drm_colorop_curve_1d_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane, u64 supported_tfs);
+ struct drm_plane *plane, u64 supported_tfs,
+ bool allow_bypass);
int drm_colorop_curve_1d_lut_init(struct drm_device *dev, struct drm_colorop *colorop,
struct drm_plane *plane, uint32_t lut_size,
- enum drm_colorop_lut1d_interpolation_type lut1d_interpolation);
+ enum drm_colorop_lut1d_interpolation_type lut1d_interpolation,
+ bool allow_bypass);
int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane);
+ struct drm_plane *plane, bool allow_bypass);
int drm_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
- struct drm_plane *plane);
+ struct drm_plane *plane, bool allow_bypass);
struct drm_colorop_state *
drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 42/44] drm/colorop: Add 3D LUT supports to color pipeline
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (40 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 41/44] drm/colorop: allow non-bypass colorops Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 43/44] drm/amd/display: add 3D LUT colorop Harry Wentland
` (3 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Alex Hung
From: Alex Hung <alex.hung@amd.com>
It is to be used to enable HDR by allowing userpace to create and pass
3D LUTs to kernel and hardware.
1. new drm_colorop_type: DRM_COLOROP_3D_LUT.
2. 3D LUT modes define hardware capabilities to userspace applications.
3. mode index points to current 3D LUT mode in lut_3d_modes.
Signed-off-by: Alex Hung <alex.hung@amd.com>
---
drivers/gpu/drm/drm_atomic.c | 21 ++++++++
drivers/gpu/drm/drm_atomic_uapi.c | 17 +++++++
drivers/gpu/drm/drm_colorop.c | 69 +++++++++++++++++++++++++
include/drm/drm_colorop.h | 42 ++++++++++++++++
include/uapi/drm/drm_mode.h | 84 +++++++++++++++++++++++++++++++
5 files changed, 233 insertions(+)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 02ff576aa7a9..1d080c367f4b 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -786,6 +786,9 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
const struct drm_colorop_state *state)
{
struct drm_colorop *colorop = state->colorop;
+ struct drm_property_blob *modes = state->lut_3d_modes;
+ struct drm_mode_3dlut_mode *mode_3dlut;
+ int i;
drm_printf(p, "colorop[%u]:\n", colorop->base.id);
drm_printf(p, "\ttype=%s\n", drm_get_colorop_type_name(colorop->type));
@@ -808,6 +811,24 @@ static void drm_atomic_colorop_print_state(struct drm_printer *p,
case DRM_COLOROP_MULTIPLIER:
drm_printf(p, "\tmultiplier=%llu\n", state->multiplier);
break;
+ case DRM_COLOROP_3D_LUT:
+ mode_3dlut = (struct drm_mode_3dlut_mode *) modes->data;
+
+ drm_printf(p, "\tlut_3d_modes blob id=%d\n", modes ? modes->base.id : 0);
+ for (i = 0; i < modes->length / sizeof(struct drm_mode_3dlut_mode); i++) {
+ drm_printf(p, "\t lut_size=%d\n", mode_3dlut[i].lut_size);
+ drm_printf(p, "\t lut_strides=%d %d %d\n", mode_3dlut[i].lut_stride[0],
+ mode_3dlut[i].lut_stride[1],
+ mode_3dlut[i].lut_stride[2]);
+ drm_printf(p, "\t interpolation=%s\n",
+ drm_get_colorop_lut3d_interpolation_name(mode_3dlut[i].interpolation));
+ drm_printf(p, "\t color_depth=%d\n", mode_3dlut[i].color_depth);
+ drm_printf(p, "\t color_format=%X\n", mode_3dlut[i].color_format);
+ drm_printf(p, "\t traversal_order=%X\n", mode_3dlut[i].traversal_order);
+ }
+ drm_printf(p, "\tlut_3d_mode_index=%d\n", state->lut_3d_mode_index);
+ drm_printf(p, "\tdata blob id=%d\n", state->data ? state->data->base.id : 0);
+ break;
default:
break;
}
diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
index 142824de33ca..50212eeba272 100644
--- a/drivers/gpu/drm/drm_atomic_uapi.c
+++ b/drivers/gpu/drm/drm_atomic_uapi.c
@@ -695,9 +695,11 @@ static int drm_atomic_color_set_data_property(struct drm_colorop *colorop,
struct drm_colorop_state *state,
struct drm_property *property, uint64_t val)
{
+ struct drm_mode_3dlut_mode *modes;
ssize_t elem_size = -1;
ssize_t size = -1;
bool replaced = false;
+ uint32_t index;
switch (colorop->type) {
case DRM_COLOROP_1D_LUT:
@@ -706,6 +708,15 @@ static int drm_atomic_color_set_data_property(struct drm_colorop *colorop,
case DRM_COLOROP_CTM_3X4:
size = sizeof(struct drm_color_ctm_3x4);
break;
+ case DRM_COLOROP_3D_LUT:
+ index = state->lut_3d_mode_index;
+ if (index >= (state->lut_3d_modes->length / sizeof(struct drm_mode_3dlut_mode)))
+ return -EINVAL;
+
+ modes = (struct drm_mode_3dlut_mode *) state->lut_3d_modes->data;
+ size = modes[index].lut_stride[0] * modes[index].lut_stride[1] * modes[index].lut_stride[2] *
+ sizeof(struct drm_color_lut);
+ break;
default:
/* should never get here */
return -EINVAL;
@@ -729,6 +740,8 @@ static int drm_atomic_colorop_set_property(struct drm_colorop *colorop,
state->curve_1d_type = val;
} else if (property == colorop->multiplier_property) {
state->multiplier = val;
+ } else if (property == colorop->lut_3d_mode_index_property) {
+ state->lut_3d_mode_index = val;
} else if (property == colorop->data_property) {
return drm_atomic_color_set_data_property(colorop,
state, property, val);
@@ -756,6 +769,10 @@ drm_atomic_colorop_get_property(struct drm_colorop *colorop,
*val = state->curve_1d_type;
} else if (property == colorop->multiplier_property) {
*val = state->multiplier;
+ } else if (property == colorop->lut_3d_modes_property) {
+ *val = (state->lut_3d_modes) ? state->lut_3d_modes->base.id : 0;
+ } else if (property == colorop->lut_3d_mode_index_property) {
+ *val = state->lut_3d_mode_index;
} 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 2b7526c9608e..86f7f084a73f 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -66,6 +66,7 @@ static const struct drm_prop_enum_list drm_colorop_type_enum_list[] = {
{ DRM_COLOROP_1D_LUT, "1D Curve Custom LUT" },
{ DRM_COLOROP_CTM_3X4, "3x4 Matrix"},
{ DRM_COLOROP_MULTIPLIER, "Multiplier"},
+ { DRM_COLOROP_3D_LUT, "3D LUT"},
};
static const char * const colorop_curve_1d_type_names[] = {
@@ -348,6 +349,53 @@ int drm_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
}
EXPORT_SYMBOL(drm_colorop_mult_init);
+int drm_colorop_3dlut_init(struct drm_device *dev, struct drm_colorop *colorop,
+ struct drm_plane *plane, struct drm_mode_3dlut_mode *mode_3dlut,
+ size_t num, bool allow_bypass)
+{
+ struct drm_property_blob *blob;
+ struct drm_property *prop;
+ int ret;
+
+ ret = drm_colorop_init(dev, colorop, plane, DRM_COLOROP_3D_LUT, allow_bypass);
+ if (ret)
+ return ret;
+
+ /* lut_3d_modes */
+ prop = drm_property_create(dev, DRM_MODE_PROP_BLOB | DRM_MODE_PROP_IMMUTABLE, "3DLUT_MODES", 0);
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->lut_3d_modes_property = prop;
+
+
+ blob = drm_property_create_blob(colorop->dev, num * sizeof(struct drm_mode_3dlut_mode),
+ mode_3dlut);
+ if (IS_ERR(blob))
+ return PTR_ERR(blob);
+
+ drm_object_attach_property(&colorop->base, colorop->lut_3d_modes_property, blob ? blob->base.id : 0);
+ drm_colorop_reset(colorop);
+
+ drm_property_replace_blob(&colorop->state->lut_3d_modes, blob);
+
+ /* lut_3d_modes index */
+ prop = drm_property_create_range(dev, DRM_MODE_PROP_ATOMIC, "3DLUT_MODE_INDEX", 0, num - 1);
+ if (!prop)
+ return -ENOMEM;
+
+ colorop->lut_3d_mode_index_property = prop;
+ drm_object_attach_property(&colorop->base, colorop->lut_3d_mode_index_property, 0);
+
+ /* data */
+ ret = drm_colorop_create_data_prop(dev, colorop);
+ if (ret)
+ return ret;
+
+ return 0;
+}
+EXPORT_SYMBOL(drm_colorop_3dlut_init);
+
static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop,
struct drm_colorop_state *state)
{
@@ -437,7 +485,13 @@ static const char * const colorop_type_name[] = {
[DRM_COLOROP_1D_LUT] = "1D Curve Custom LUT",
[DRM_COLOROP_CTM_3X4] = "3x4 Matrix",
[DRM_COLOROP_MULTIPLIER] = "Multiplier",
+ [DRM_COLOROP_3D_LUT] = "3D LUT",
};
+
+static const char * const colorop_lu3d_interpolation_name[] = {
+ [DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL] = "Tetrahedral",
+};
+
static const char * const colorop_lut1d_interpolation_name[] = {
[DRM_COLOROP_LUT1D_INTERPOLATION_LINEAR] = "Linear",
};
@@ -473,6 +527,21 @@ const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_inte
return colorop_lut1d_interpolation_name[type];
}
+/**
+ * drm_get_colorop_lut3d_interpolation_name - return a string for interpolation type
+ * @type: interpolation type to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
+const char *drm_get_colorop_lut3d_interpolation_name(enum drm_colorop_lut3d_interpolation_type type)
+{
+ if (WARN_ON(type >= ARRAY_SIZE(colorop_lu3d_interpolation_name)))
+ return "unknown";
+
+ return colorop_lu3d_interpolation_name[type];
+}
+
/**
* drm_colorop_set_next_property - sets the next pointer
* @colorop: drm colorop
diff --git a/include/drm/drm_colorop.h b/include/drm/drm_colorop.h
index 044294b498e4..285ab8cc12f0 100644
--- a/include/drm/drm_colorop.h
+++ b/include/drm/drm_colorop.h
@@ -139,6 +139,29 @@ struct drm_colorop_state {
*/
uint32_t size;
+ /**
+ * @lut_3d_modes:
+ *
+ * Mode blob for displaying a list of supported 3dlut modes.
+ *
+ * To setup a 3D LUT, lut_3d_modes, lut_3d_modes and data are expected
+ * to be used in the following sequence:
+ *
+ * 1. device driver sets a list of supported drm_mode_3dlut_mode in "lut_3d_modes".
+ * 2. userspace reads "lut_3d_modes" to determines an appropriate mode.
+ * 3. userspace sets "lut_3d_mode_index" pointing the selected mode.
+ * 4. userspace passes a 3D LUT via "data"
+ * 5. usersapce commits to device driver
+ */
+ struct drm_property_blob *lut_3d_modes;
+
+ /**
+ * @lut_3d_mode_index:
+ *
+ * A zero-based index pointing to current lut_3d_mode.
+ */
+ uint16_t lut_3d_mode_index;
+
/**
* @data:
*
@@ -274,6 +297,21 @@ struct drm_colorop {
*/
struct drm_property *size_property;
+ /**
+ * @lut_3d_modes_property:
+ *
+ * 3DLUT mode property used to convert the framebuffer's colors
+ * to non-linear gamma.
+ */
+ struct drm_property *lut_3d_modes_property;
+
+ /**
+ * @lut_3d_mode_index_property:
+ *
+ * 3DLUT mode index property for choosing 3D LUT mode.
+ */
+ struct drm_property *lut_3d_mode_index_property;
+
/**
* @data_property:
*
@@ -329,6 +367,9 @@ int drm_colorop_ctm_3x4_init(struct drm_device *dev, struct drm_colorop *colorop
struct drm_plane *plane, bool allow_bypass);
int drm_colorop_mult_init(struct drm_device *dev, struct drm_colorop *colorop,
struct drm_plane *plane, bool allow_bypass);
+int drm_colorop_3dlut_init(struct drm_device *dev, struct drm_colorop *colorop,
+ struct drm_plane *plane, struct drm_mode_3dlut_mode *mode_3dlut,
+ size_t num, bool allow_bypass);
struct drm_colorop_state *
drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colorop);
@@ -379,6 +420,7 @@ const char *drm_get_colorop_type_name(enum drm_colorop_type type);
*/
const char *drm_get_colorop_curve_1d_type_name(enum drm_colorop_curve_1d_type type);
const char *drm_get_colorop_lut1d_interpolation_name(enum drm_colorop_lut1d_interpolation_type type);
+const char *drm_get_colorop_lut3d_interpolation_name(enum drm_colorop_lut3d_interpolation_type type);
void drm_colorop_set_next_property(struct drm_colorop *colorop, struct drm_colorop *next);
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index b64f9515bd0a..ab807045003f 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -914,6 +914,90 @@ enum drm_colorop_type {
* property.
*/
DRM_COLOROP_MULTIPLIER,
+ /**
+ * @DRM_COLOROP_3D_LUT:
+ *
+ * A 3D LUT of &drm_color_lut entries,
+ * packed into a blob via the DATA property. The driver's expected
+ * LUT size is advertised via the SIZE property.
+ */
+ DRM_COLOROP_3D_LUT,
+};
+
+/**
+ * enum drm_colorop_lut3d_interpolation_type - type of 3DLUT interpolation
+ *
+ */
+enum drm_colorop_lut3d_interpolation_type {
+ /**
+ * @DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL:
+ *
+ * Tetrahedral 3DLUT interpolation
+ */
+ DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
+};
+
+/**
+ * enum drm_colorop_lut3d_traversal_order - traversal order of the 3D LUT
+ *
+ * This enum describes the order of traversal of 3DLUT elements.
+ */
+enum drm_colorop_lut3d_traversal_order {
+ /**
+ * @DRM_COLOROP_LUT3D_TRAVERSAL_RGB:
+ *
+ * the LUT elements are traversed like so:
+ * for R in range 0..n
+ * for G in range 0..n
+ * for B in range 0..n
+ * color = lut3d[R][G][B]
+ */
+ DRM_COLOROP_LUT3D_TRAVERSAL_RGB,
+ /**
+ * @DRM_COLOROP_LUT3D_TRAVERSAL_BGR:
+ *
+ * the LUT elements are traversed like so:
+ * for R in range 0..n
+ * for G in range 0..n
+ * for B in range 0..n
+ * color = lut3d[B][G][R]
+ */
+ DRM_COLOROP_LUT3D_TRAVERSAL_BGR,
+};
+
+/**
+ * struct drm_mode_3dlut_mode - 3D LUT mode
+ *
+ * The mode describes the supported and selected format of a 3DLUT.
+ */
+struct drm_mode_3dlut_mode {
+ /**
+ * @lut_size: 3D LUT size - can be 9, 17 or 33
+ */
+ __u16 lut_size;
+ /**
+ * @lut_stride: dimensions of 3D LUT. Must be larger than lut_size
+ */
+ __u16 lut_stride[3];
+ /**
+ * @interpolation: interpolation algorithm for 3D LUT. See drm_colorop_lut3d_interpolation_type
+ */
+ __u16 interpolation;
+ /**
+ * @color_depth: color depth - can be 8, 10 or 12
+ */
+ __u16 color_depth;
+ /**
+ * @color_format: color format specified by fourcc values
+ * ex. DRM_FORMAT_XRGB16161616 - color in order of RGB, each is 16bit.
+ */
+ __u32 color_format;
+ /**
+ * @traversal_order:
+ *
+ * Traversal order when parsing/writing the 3D LUT. See enum drm_colorop_lut3d_traversal_order
+ */
+ __u16 traversal_order;
};
/**
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 43/44] drm/amd/display: add 3D LUT colorop
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (41 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 42/44] drm/colorop: Add 3D LUT supports to color pipeline Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-19 20:57 ` [PATCH v5 44/44] drm/amd/display: Add AMD color pipeline doc Harry Wentland
` (2 subsequent siblings)
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 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 3D LUT.
The color pipeline now consists of the following colorops:
1. 1D curve colorop
2. Multiplier
3. 3x4 CTM
4. 1D curve colorop
5. 1D LUT
6. 3D LUT
7. 1D curve colorop
8. 1D LUT
Signed-off-by: Alex Hung <alex.hung@amd.com>
---
.../amd/display/amdgpu_dm/amdgpu_dm_color.c | 113 +++++++++++++++++-
.../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 26 ++++
2 files changed, 136 insertions(+), 3 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 2f0b06fb5f81..cc3d68d1386f 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
@@ -1283,7 +1283,8 @@ __set_dm_plane_colorop_multiplier(struct drm_plane_state *plane_state,
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 *colorop,
+ bool *enabled)
{
struct drm_colorop *old_colorop;
struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
@@ -1311,6 +1312,7 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
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);
+ *enabled = true;
}
/* 1D LUT - SHAPER LUT */
@@ -1338,8 +1340,101 @@ __set_dm_plane_colorop_shaper(struct drm_plane_state *plane_state,
shaper_size = shaper_lut != NULL ? shaper_size : 0;
/* Custom LUT size must be the same as supported size */
- if (shaper_size == colorop_state->size)
+ if (shaper_size == colorop_state->size) {
__set_output_tf(tf, shaper_lut, shaper_size, false);
+ *enabled = true;
+ }
+ }
+
+ return 0;
+}
+
+/* __set_colorop_3dlut - set DRM 3D LUT to DC stream
+ * @drm_lut3d: user 3D LUT
+ * @drm_lut3d_size: size of 3D LUT
+ * @drm_mode_3dlut: drm_mode selected by userspace
+ * @lut3d: DC 3D LUT
+ *
+ * Map user 3D LUT data to DC 3D LUT and all necessary bits to program it
+ * on DCN accordingly.
+ */
+static void __set_colorop_3dlut(const struct drm_color_lut *drm_lut3d,
+ uint32_t drm_lut3d_size,
+ struct drm_mode_3dlut_mode *drm_mode_3dlut,
+ struct dc_3dlut *lut)
+{
+ if (!drm_lut3d_size)
+ return;
+
+ lut->state.bits.initialized = 0;
+
+ /* Only supports 17x17x17 3D LUT (12-bit) now */
+ if (drm_mode_3dlut->color_depth == 12)
+ lut->lut_3d.use_12bits = true;
+ else
+ return;
+
+ if (drm_mode_3dlut->lut_size == 17)
+ lut->lut_3d.use_tetrahedral_9 = false;
+ else
+ return;
+
+ lut->state.bits.initialized = 1;
+ __drm_3dlut_to_dc_3dlut(drm_lut3d, drm_lut3d_size, &lut->lut_3d,
+ lut->lut_3d.use_tetrahedral_9,
+ drm_mode_3dlut->color_depth);
+
+}
+
+static int
+__set_dm_plane_colorop_3dlut(struct drm_plane_state *plane_state,
+ struct dc_plane_state *dc_plane_state,
+ struct drm_colorop *colorop,
+ bool shaper_enabled)
+{
+ struct drm_colorop *old_colorop;
+ struct drm_colorop_state *colorop_state = NULL, *new_colorop_state;
+ struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
+ struct drm_atomic_state *state = plane_state->state;
+ const struct amdgpu_device *adev = drm_to_adev(colorop->dev);
+ const struct drm_device *dev = colorop->dev;
+ struct drm_mode_3dlut_mode *mode;
+ const struct drm_color_lut *lut3d;
+ uint32_t lut3d_size, index;
+ int i = 0;
+
+ /* 3D LUT */
+ old_colorop = colorop;
+ for_each_new_colorop_in_state(state, colorop, new_colorop_state, i) {
+ if (new_colorop_state->colorop == old_colorop &&
+ new_colorop_state->colorop->type == DRM_COLOROP_3D_LUT) {
+ colorop_state = new_colorop_state;
+ break;
+ }
+ }
+
+ if (colorop_state && !colorop_state->bypass && colorop->type == DRM_COLOROP_3D_LUT) {
+ if (!adev->dm.dc->caps.color.dpp.hw_3d_lut) {
+ drm_dbg(dev, "3D LUT is not supported by hardware\n");
+ return 0;
+ }
+
+ drm_dbg(dev, "3D LUT colorop with ID: %d\n", colorop->base.id);
+ mode = (struct drm_mode_3dlut_mode *) colorop_state->lut_3d_modes->data;
+ index = colorop_state->lut_3d_mode_index;
+ lut3d = __extract_blob_lut(colorop_state->data, &lut3d_size);
+ lut3d_size = lut3d != NULL ? lut3d_size : 0;
+ __set_colorop_3dlut(lut3d, lut3d_size, &mode[index], &dc_plane_state->lut3d_func);
+
+ /* 3D LUT requires shaper. If shaper colorop is bypassed, enable shaper curve
+ * with TRANSFER_FUNCTION_LINEAR
+ */
+ if (!shaper_enabled) {
+ tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+ tf->tf = TRANSFER_FUNCTION_LINEAR;
+ tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
+ __set_output_tf(tf, NULL, 0, false);
+ }
}
return 0;
@@ -1468,6 +1563,7 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
{
struct drm_colorop *colorop = plane_state->color_pipeline;
struct drm_device *dev = plane_state->plane->dev;
+ bool shaper_enabled = false;
int ret;
/* 1D Curve - DEGAM TF */
@@ -1508,7 +1604,7 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
return -EINVAL;
}
- ret = __set_dm_plane_colorop_shaper(plane_state, dc_plane_state, colorop);
+ ret = __set_dm_plane_colorop_shaper(plane_state, dc_plane_state, colorop, &shaper_enabled);
if (ret)
return ret;
@@ -1517,6 +1613,17 @@ amdgpu_dm_plane_set_colorop_properties(struct drm_plane_state *plane_state,
if (!colorop)
return -EINVAL;
+ /* 3D LUT */
+ colorop = colorop->next;
+ if (!colorop) {
+ drm_dbg(dev, "no 3D LUT colorop found\n");
+ return -EINVAL;
+ }
+
+ ret = __set_dm_plane_colorop_3dlut(plane_state, dc_plane_state, colorop, shaper_enabled);
+ if (ret)
+ return ret;
+
/* 1D Curve & LUT - BLND TF & LUT */
colorop = colorop->next;
if (!colorop) {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
index 05fdec7c28ad..a81f81240341 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
@@ -47,6 +47,17 @@ const u64 amdgpu_dm_supported_blnd_tfs =
BIT(DRM_COLOROP_1D_CURVE_PQ_125_EOTF) |
BIT(DRM_COLOROP_1D_CURVE_BT2020_INV_OETF);
+struct drm_mode_3dlut_mode lut3d_modes[] = {
+ {
+ .lut_size = 17,
+ .lut_stride = {17, 17, 17},
+ .interpolation = DRM_COLOROP_LUT3D_INTERPOLATION_TETRAHEDRAL,
+ .color_depth = 12,
+ .color_format = DRM_FORMAT_XRGB16161616,
+ .traversal_order = DRM_COLOROP_LUT3D_TRAVERSAL_RGB,
+ },
+};
+
int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_prop_enum_list *list)
{
struct drm_colorop *op, *prev_op;
@@ -135,6 +146,21 @@ int amdgpu_dm_initialize_default_pipeline(struct drm_plane *plane, struct drm_pr
prev_op = op;
+ /* 3D LUT */
+ op = kzalloc(sizeof(struct drm_colorop), GFP_KERNEL);
+ if (!op) {
+ DRM_ERROR("KMS: Failed to allocate colorop\n");
+ return -ENOMEM;
+ }
+
+ ret = drm_colorop_3dlut_init(dev, op, plane, lut3d_modes, ARRAY_SIZE(lut3d_modes), true);
+ 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) {
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* [PATCH v5 44/44] drm/amd/display: Add AMD color pipeline doc
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (42 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 43/44] drm/amd/display: add 3D LUT colorop Harry Wentland
@ 2024-08-19 20:57 ` Harry Wentland
2024-08-27 17:49 ` [PATCH v5 00/44] Color Pipeline API w/ VKMS Louis Chauvet
2024-08-29 14:55 ` Xaver Hugl
45 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-19 20:57 UTC (permalink / raw)
To: dri-devel, amd-gfx; +Cc: wayland-devel, Harry Wentland
A short description about the AMD color pipeline.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
.../amd/display/amdgpu_dm/amdgpu_dm_color.c | 122 +++++++++++++++---
1 file changed, 102 insertions(+), 20 deletions(-)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index cc3d68d1386f..d48dcaee9e2e 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -33,6 +33,32 @@
/**
* DOC: overview
*
+ * We have three types of color management in the AMD display driver.
+ * 1. the legacy &drm_crtc DEGAMMA, CTM, and GAMMA properties
+ * 2. AMD driver private color management on &drm_plane and &drm_crtc
+ * 3. AMD plane color pipeline
+ *
+ * The CRTC properties are the original color management. When they were
+ * implemented per-plane color management was not a thing yet. Because
+ * of that we could get away with plumbing the DEGAMMA and CTM
+ * properties to pre-blending HW functions. This is incompatible with
+ * per-plane color management, such as via the AMD private properties or
+ * the new drm_plane color pipeline. The only compatible CRTC property
+ * with per-plane color management is the GAMMA property as it is
+ * applied post-blending.
+ *
+ * The AMD driver private color management properties are only exposed
+ * when the kernel is built explicitly with -DAMD_PRIVATE_COLOR. They
+ * are temporary building blocks on the path to full-fledged &drm_plane
+ * and &drm_crtc color pipelines and lay the driver's groundwork for the
+ * color pipelines.
+ *
+ * The AMD plane color pipeline describes AMD's &drm_colorops via the
+ * &drm_plane's COLOR_PIPELINE property.
+ *
+ * drm_crtc Properties
+ * -------------------
+ *
* The DC interface to HW gives us the following color management blocks
* per pipe (surface):
*
@@ -43,33 +69,89 @@
* - Surface regamma LUT (normalized)
* - Output CSC (normalized)
*
- * But these aren't a direct mapping to DRM color properties. The current DRM
- * interface exposes CRTC degamma, CRTC CTM and CRTC regamma while our hardware
- * is essentially giving:
+ * But these aren't a direct mapping to DRM color properties. The
+ * current DRM interface exposes CRTC degamma, CRTC CTM and CRTC regamma
+ * while our hardware is essentially giving:
*
* Plane CTM -> Plane degamma -> Plane CTM -> Plane regamma -> Plane CTM
*
- * The input gamma LUT block isn't really applicable here since it operates
- * on the actual input data itself rather than the HW fp representation. The
- * input and output CSC blocks are technically available to use as part of
- * the DC interface but are typically used internally by DC for conversions
- * between color spaces. These could be blended together with user
- * adjustments in the future but for now these should remain untouched.
+ * The input gamma LUT block isn't really applicable here since it
+ * operates on the actual input data itself rather than the HW fp
+ * representation. The input and output CSC blocks are technically
+ * available to use as part of the DC interface but are typically used
+ * internally by DC for conversions between color spaces. These could be
+ * blended together with user adjustments in the future but for now
+ * these should remain untouched.
*
- * The pipe blending also happens after these blocks so we don't actually
- * support any CRTC props with correct blending with multiple planes - but we
- * can still support CRTC color management properties in DM in most single
- * plane cases correctly with clever management of the DC interface in DM.
+ * The pipe blending also happens after these blocks so we don't
+ * actually support any CRTC props with correct blending with multiple
+ * planes - but we can still support CRTC color management properties in
+ * DM in most single plane cases correctly with clever management of the
+ * DC interface in DM.
*
- * As per DRM documentation, blocks should be in hardware bypass when their
- * respective property is set to NULL. A linear DGM/RGM LUT should also
- * considered as putting the respective block into bypass mode.
+ * As per DRM documentation, blocks should be in hardware bypass when
+ * their respective property is set to NULL. A linear DGM/RGM LUT should
+ * also considered as putting the respective block into bypass mode.
*
- * This means that the following
- * configuration is assumed to be the default:
+ * This means that the following configuration is assumed to be the
+ * default:
+ *
+ * Plane DGM Bypass -> Plane CTM Bypass -> Plane RGM Bypass -> ... CRTC
+ * DGM Bypass -> CRTC CTM Bypass -> CRTC RGM Bypass
+ *
+ * AMD Private Color Management on drm_plane
+ * -----------------------------------------
+ *
+ * The AMD private color management properties on a &drm_plane are:
+ *
+ * - AMD_PLANE_DEGAMMA_LUT
+ * - AMD_PLANE_DEGAMMA_LUT_SIZE
+ * - AMD_PLANE_DEGAMMA_TF
+ * - AMD_PLANE_HDR_MULT
+ * - AMD_PLANE_CTM
+ * - AMD_PLANE_SHAPER_LUT
+ * - AMD_PLANE_SHAPER_LUT_SIZE
+ * - AMD_PLANE_SHAPER_TF
+ * - AMD_PLANE_LUT3D
+ * - AMD_PLANE_LUT3D_SIZE
+ * - AMD_PLANE_BLEND_LUT
+ * - AMD_PLANE_BLEND_LUT_SIZE
+ * - AMD_PLANE_BLEND_TF
+ *
+ * The AMD private color management property on a &drm_crtc is:
+ *
+ * - AMD_CRTC_REGAMMA_TF
+ *
+ * Use of these properties is discouraged.
+ *
+ * AMD plane color pipeline
+ * ------------------------
+ *
+ * The AMD &drm_plane color pipeline is advertised for DCN generations
+ * 3.0 and newer. It exposes these elements in this order:
+ *
+ * 1. 1D curve colorop
+ * 2. Multiplier
+ * 3. 3x4 CTM
+ * 4. 1D curve colorop
+ * 5. 1D LUT
+ * 6. 3D LUT
+ * 7. 1D curve colorop
+ * 8. 1D LUT
+ *
+ * The multiplier (#2) is a simple multiplier that is applied to all
+ * channels.
+ *
+ * The 3x4 CTM (#3) is a simple 3x4 matrix.
+ *
+ * #1, and #7 are non-linear to linear curves. #4 is a linear to
+ * non-linear curve. They support sRGB, PQ, and BT.709/BT.2020 EOTFs or
+ * their inverse.
+ *
+ * The 1D LUTs (#5 and #8) are plain 4096 entry LUTs.
+ *
+ * The 3DLUT (#6) is a tetrahedrally interpolated 17 cube LUT.
*
- * Plane DGM Bypass -> Plane CTM Bypass -> Plane RGM Bypass -> ...
- * CRTC DGM Bypass -> CRTC CTM Bypass -> CRTC RGM Bypass
*/
#define MAX_DRM_LUT_VALUE 0xFFFF
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* Re: [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling
2024-08-19 20:56 ` [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
@ 2024-08-21 1:05 ` Jeff Johnson
2024-08-27 17:49 ` Louis Chauvet
1 sibling, 0 replies; 69+ messages in thread
From: Jeff Johnson @ 2024-08-21 1:05 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx; +Cc: wayland-devel, Arthur Grillo
On 8/19/24 13:56, 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
>
> v5:
> - Bring back static for lerp_u16 and get_lut_index (Arthur)
>
> v4:
> - Test the critical points of the lerp function (Pekka)
>
> v3:
> - Use include way of testing static functions (Arthur)
note version information is usually after the "---" so that it not part
of the final git history
>
> 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 | 4 +
> 4 files changed, 176 insertions(+)
> 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
...
> +kunit_test_suite(vkms_color_test_suite);
> +
> +MODULE_LICENSE("GPL");
> \ No newline at end of file
Since commit 1fffe7a34c89 ("script: modpost: emit a warning when the
description is missing"), a module without a MODULE_DESCRIPTION() will
result in a warning when built with make W=1. Recently, multiple
developers have been eradicating these warnings treewide, and very few
are left, so please don't introduce a new one :)
Please add the missing MODULE_DESCRIPTION()
/jeff
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h
2024-08-19 20:56 ` [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
@ 2024-08-21 1:07 ` Jeff Johnson
2024-08-21 4:00 ` kernel test robot
2024-08-22 14:55 ` kernel test robot
2 siblings, 0 replies; 69+ messages in thread
From: Jeff Johnson @ 2024-08-21 1:07 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx; +Cc: wayland-devel
On 8/19/24 13:56, Harry Wentland 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 56dab563abd7..bd69df0eee33 100644
> --- a/drivers/gpu/drm/tests/Makefile
> +++ b/drivers/gpu/drm/tests/Makefile
> @@ -20,6 +20,7 @@ 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 @@
...
> +MODULE_AUTHOR("AMD");
> +MODULE_LICENSE("GPL and additional rights");
> \ No newline at end of file
Please add the missing MODULE_DESCRIPTION()
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop
2024-08-19 20:57 ` [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop Harry Wentland
@ 2024-08-21 3:19 ` kernel test robot
2024-08-21 9:22 ` kernel test robot
1 sibling, 0 replies; 69+ messages in thread
From: kernel test robot @ 2024-08-21 3:19 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx
Cc: oe-kbuild-all, wayland-devel, Alex Hung, Harry Wentland
Hi Harry,
kernel test robot noticed the following build warnings:
[auto build test WARNING on drm/drm-next]
[also build test WARNING on drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-misc/drm-misc-next drm-tip/drm-tip linus/master v6.11-rc4 next-20240820]
[cannot apply to drm-exynos/exynos-drm-next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Harry-Wentland/drm-Add-helper-for-conversion-from-signed-magnitude/20240820-050138
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link: https://lore.kernel.org/r/20240819205714.316380-37-harry.wentland%40amd.com
patch subject: [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop
config: arm-randconfig-002-20240821 (https://download.01.org/0day-ci/archive/20240821/202408211016.TkunD7H3-lkp@intel.com/config)
compiler: arm-linux-gnueabi-gcc (GCC) 14.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240821/202408211016.TkunD7H3-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408211016.TkunD7H3-lkp@intel.com/
All warnings (new ones prefixed by >>):
In file included from include/linux/device.h:15,
from include/drm/drm_print.h:31,
from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_ring.h:29,
from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_ctx.h:29,
from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:43,
from drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:25:
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c: In function '__set_dm_plane_colorop_3x4_matrix':
>> drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:41:22: warning: format '%ld' expects argument of type 'long int', but argument 3 has type 'size_t' {aka 'unsigned int'} [-Wformat=]
41 | #define dev_fmt(fmt) "amdgpu: " fmt
| ^~~~~~~~~~
include/linux/dev_printk.h:110:30: note: in definition of macro 'dev_printk_index_wrap'
110 | _p_func(dev, fmt, ##__VA_ARGS__); \
| ^~~
include/linux/dev_printk.h:156:61: note: in expansion of macro 'dev_fmt'
156 | dev_printk_index_wrap(_dev_warn, KERN_WARNING, dev, dev_fmt(fmt), ##__VA_ARGS__)
| ^~~~~~~
include/drm/drm_print.h:470:9: note: in expansion of macro 'dev_warn'
470 | dev_##level##type((drm) ? (drm)->dev : NULL, "[drm] " fmt, ##__VA_ARGS__)
| ^~~~
include/drm/drm_print.h:480:9: note: in expansion of macro '__drm_printk'
480 | __drm_printk((drm), warn,, fmt, ##__VA_ARGS__)
| ^~~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1247:25: note: in expansion of macro 'drm_warn'
1247 | drm_warn(dev, "blob->length (%ld) isn't equal to drm_color_ctm_3x4 (%ld)\n",
| ^~~~~~~~
>> drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:41:22: warning: format '%ld' expects argument of type 'long int', but argument 4 has type 'unsigned int' [-Wformat=]
41 | #define dev_fmt(fmt) "amdgpu: " fmt
| ^~~~~~~~~~
include/linux/dev_printk.h:110:30: note: in definition of macro 'dev_printk_index_wrap'
110 | _p_func(dev, fmt, ##__VA_ARGS__); \
| ^~~
include/linux/dev_printk.h:156:61: note: in expansion of macro 'dev_fmt'
156 | dev_printk_index_wrap(_dev_warn, KERN_WARNING, dev, dev_fmt(fmt), ##__VA_ARGS__)
| ^~~~~~~
include/drm/drm_print.h:470:9: note: in expansion of macro 'dev_warn'
470 | dev_##level##type((drm) ? (drm)->dev : NULL, "[drm] " fmt, ##__VA_ARGS__)
| ^~~~
include/drm/drm_print.h:480:9: note: in expansion of macro '__drm_printk'
480 | __drm_printk((drm), warn,, fmt, ##__VA_ARGS__)
| ^~~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1247:25: note: in expansion of macro 'drm_warn'
1247 | drm_warn(dev, "blob->length (%ld) isn't equal to drm_color_ctm_3x4 (%ld)\n",
| ^~~~~~~~
vim +41 drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h
539489fc91ea77 Aurabindo Pillai 2020-04-08 40
539489fc91ea77 Aurabindo Pillai 2020-04-08 @41 #define dev_fmt(fmt) "amdgpu: " fmt
539489fc91ea77 Aurabindo Pillai 2020-04-08 42
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h
2024-08-19 20:56 ` [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
2024-08-21 1:07 ` Jeff Johnson
@ 2024-08-21 4:00 ` kernel test robot
2024-08-22 14:55 ` kernel test robot
2 siblings, 0 replies; 69+ messages in thread
From: kernel test robot @ 2024-08-21 4:00 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx
Cc: oe-kbuild-all, wayland-devel, Harry Wentland
Hi Harry,
kernel test robot noticed the following build errors:
[auto build test ERROR on drm/drm-next]
[also build test ERROR on drm-exynos/exynos-drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-misc/drm-misc-next drm-tip/drm-tip linus/master v6.11-rc4 next-20240820]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Harry-Wentland/drm-Add-helper-for-conversion-from-signed-magnitude/20240820-050138
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link: https://lore.kernel.org/r/20240819205714.316380-21-harry.wentland%40amd.com
patch subject: [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h
config: parisc-randconfig-001-20240821 (https://download.01.org/0day-ci/archive/20240821/202408211157.n8ubZpFl-lkp@intel.com/config)
compiler: hppa-linux-gcc (GCC) 14.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240821/202408211157.n8ubZpFl-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408211157.n8ubZpFl-lkp@intel.com/
All errors (new ones prefixed by >>):
In file included from drivers/gpu/drm/tests/drm_fixp_test.c:6:
drivers/gpu/drm/tests/drm_fixp_test.c: In function 'drm_test_sm2fixp':
>> drivers/gpu/drm/tests/drm_fixp_test.c:11:66: error: integer overflow in expression of type 'long long int' results in '9223372036854775807' [-Werror=overflow]
11 | KUNIT_EXPECT_EQ(test, 0x7fffffffffffffffll, ((1LL << 63) - 1));
| ^
include/kunit/test.h:725:40: note: in definition of macro 'KUNIT_BASE_BINARY_ASSERTION'
725 | const typeof(right) __right = (right); \
| ^~~~~
include/kunit/test.h:920:9: note: in expansion of macro 'KUNIT_BINARY_INT_ASSERTION'
920 | KUNIT_BINARY_INT_ASSERTION(test, \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~
include/kunit/test.h:917:9: note: in expansion of macro 'KUNIT_EXPECT_EQ_MSG'
917 | KUNIT_EXPECT_EQ_MSG(test, left, right, NULL)
| ^~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/tests/drm_fixp_test.c:11:9: note: in expansion of macro 'KUNIT_EXPECT_EQ'
11 | KUNIT_EXPECT_EQ(test, 0x7fffffffffffffffll, ((1LL << 63) - 1));
| ^~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
vim +11 drivers/gpu/drm/tests/drm_fixp_test.c
8
9 static void drm_test_sm2fixp(struct kunit *test)
10 {
> 11 KUNIT_EXPECT_EQ(test, 0x7fffffffffffffffll, ((1LL << 63) - 1));
12
13 /* 1 */
14 KUNIT_EXPECT_EQ(test, drm_int2fixp(1), drm_sm2fixp(1ull << DRM_FIXED_POINT));
15
16 /* -1 */
17 KUNIT_EXPECT_EQ(test, drm_int2fixp(-1), drm_sm2fixp((1ull << 63) | (1ull << DRM_FIXED_POINT)));
18
19 /* 0.5 */
20 KUNIT_EXPECT_EQ(test, drm_fixp_from_fraction(1, 2), drm_sm2fixp(1ull << (DRM_FIXED_POINT - 1)));
21
22 /* -0.5 */
23 KUNIT_EXPECT_EQ(test, drm_fixp_from_fraction(-1, 2), drm_sm2fixp((1ull << 63) | (1ull << (DRM_FIXED_POINT - 1))));
24
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
2024-08-19 20:56 ` [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
@ 2024-08-21 4:52 ` kernel test robot
0 siblings, 0 replies; 69+ messages in thread
From: kernel test robot @ 2024-08-21 4:52 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx
Cc: llvm, oe-kbuild-all, wayland-devel, Alex Hung, Harry Wentland
Hi Harry,
kernel test robot noticed the following build warnings:
[auto build test WARNING on drm/drm-next]
[also build test WARNING on drm-exynos/exynos-drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-misc/drm-misc-next drm-tip/drm-tip linus/master v6.11-rc4 next-20240820]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Harry-Wentland/drm-Add-helper-for-conversion-from-signed-magnitude/20240820-050138
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link: https://lore.kernel.org/r/20240819205714.316380-29-harry.wentland%40amd.com
patch subject: [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
config: x86_64-randconfig-003-20240821 (https://download.01.org/0day-ci/archive/20240821/202408211241.xAvubshv-lkp@intel.com/config)
compiler: clang version 18.1.5 (https://github.com/llvm/llvm-project 617a15a9eac96088ae5e9134248d8236e34b91b1)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240821/202408211241.xAvubshv-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408211241.xAvubshv-lkp@intel.com/
All warnings (new ones prefixed by >>):
>> drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1232:42: warning: variable 'shaper_size' is uninitialized when used here [-Wuninitialized]
1232 | return __set_output_tf(tf, shaper_lut, shaper_size, false);
| ^~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1214:22: note: initialize the variable 'shaper_size' to silence this warning
1214 | uint32_t shaper_size;
| ^
| = 0
1 warning generated.
vim +/shaper_size +1232 drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c
1205
1206 static int
1207 __set_colorop_in_shaper_1d_curve(struct dc_plane_state *dc_plane_state,
1208 struct drm_colorop_state *colorop_state)
1209 {
1210 struct dc_transfer_func *tf = &dc_plane_state->in_shaper_func;
1211 struct drm_colorop *colorop = colorop_state->colorop;
1212 struct drm_device *drm = colorop->dev;
1213 const struct drm_color_lut *shaper_lut;
1214 uint32_t shaper_size;
1215
1216 if (colorop->type != DRM_COLOROP_1D_CURVE &&
1217 colorop_state->curve_1d_type != DRM_COLOROP_1D_CURVE_SRGB_INV_EOTF)
1218 return -EINVAL;
1219
1220 if (colorop_state->bypass) {
1221 tf->type = TF_TYPE_BYPASS;
1222 tf->tf = TRANSFER_FUNCTION_LINEAR;
1223 return 0;
1224 }
1225
1226 drm_dbg(drm, "Shaper colorop with ID: %d\n", colorop->base.id);
1227
1228 if (colorop->type == DRM_COLOROP_1D_CURVE) {
1229 tf->type = TF_TYPE_DISTRIBUTED_POINTS;
1230 tf->tf = amdgpu_colorop_tf_to_dc_tf(colorop_state->curve_1d_type);
1231 tf->sdr_ref_white_level = SDR_WHITE_LEVEL_INIT_VALUE;
> 1232 return __set_output_tf(tf, shaper_lut, shaper_size, false);
1233 }
1234
1235 return -EINVAL;
1236 }
1237
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop
2024-08-19 20:57 ` [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop Harry Wentland
2024-08-21 3:19 ` kernel test robot
@ 2024-08-21 9:22 ` kernel test robot
1 sibling, 0 replies; 69+ messages in thread
From: kernel test robot @ 2024-08-21 9:22 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx
Cc: oe-kbuild-all, wayland-devel, Alex Hung, Harry Wentland
Hi Harry,
kernel test robot noticed the following build errors:
[auto build test ERROR on drm/drm-next]
[also build test ERROR on drm-exynos/exynos-drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-misc/drm-misc-next drm-tip/drm-tip linus/master v6.11-rc4 next-20240821]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Harry-Wentland/drm-Add-helper-for-conversion-from-signed-magnitude/20240820-050138
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link: https://lore.kernel.org/r/20240819205714.316380-37-harry.wentland%40amd.com
patch subject: [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop
config: parisc-randconfig-001-20240821 (https://download.01.org/0day-ci/archive/20240821/202408211502.eV7fXuRD-lkp@intel.com/config)
compiler: hppa-linux-gcc (GCC) 14.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240821/202408211502.eV7fXuRD-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408211502.eV7fXuRD-lkp@intel.com/
All errors (new ones prefixed by >>):
In file included from include/linux/device.h:15,
from include/drm/drm_print.h:31,
from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_ring.h:29,
from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu_ctx.h:29,
from drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:43,
from drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:25:
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c: In function '__set_dm_plane_colorop_3x4_matrix':
>> drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:41:22: error: format '%ld' expects argument of type 'long int', but argument 3 has type 'size_t' {aka 'unsigned int'} [-Werror=format=]
41 | #define dev_fmt(fmt) "amdgpu: " fmt
| ^~~~~~~~~~
include/linux/dev_printk.h:110:30: note: in definition of macro 'dev_printk_index_wrap'
110 | _p_func(dev, fmt, ##__VA_ARGS__); \
| ^~~
include/linux/dev_printk.h:156:61: note: in expansion of macro 'dev_fmt'
156 | dev_printk_index_wrap(_dev_warn, KERN_WARNING, dev, dev_fmt(fmt), ##__VA_ARGS__)
| ^~~~~~~
include/drm/drm_print.h:470:9: note: in expansion of macro 'dev_warn'
470 | dev_##level##type((drm) ? (drm)->dev : NULL, "[drm] " fmt, ##__VA_ARGS__)
| ^~~~
include/drm/drm_print.h:480:9: note: in expansion of macro '__drm_printk'
480 | __drm_printk((drm), warn,, fmt, ##__VA_ARGS__)
| ^~~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1247:25: note: in expansion of macro 'drm_warn'
1247 | drm_warn(dev, "blob->length (%ld) isn't equal to drm_color_ctm_3x4 (%ld)\n",
| ^~~~~~~~
>> drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h:41:22: error: format '%ld' expects argument of type 'long int', but argument 4 has type 'unsigned int' [-Werror=format=]
41 | #define dev_fmt(fmt) "amdgpu: " fmt
| ^~~~~~~~~~
include/linux/dev_printk.h:110:30: note: in definition of macro 'dev_printk_index_wrap'
110 | _p_func(dev, fmt, ##__VA_ARGS__); \
| ^~~
include/linux/dev_printk.h:156:61: note: in expansion of macro 'dev_fmt'
156 | dev_printk_index_wrap(_dev_warn, KERN_WARNING, dev, dev_fmt(fmt), ##__VA_ARGS__)
| ^~~~~~~
include/drm/drm_print.h:470:9: note: in expansion of macro 'dev_warn'
470 | dev_##level##type((drm) ? (drm)->dev : NULL, "[drm] " fmt, ##__VA_ARGS__)
| ^~~~
include/drm/drm_print.h:480:9: note: in expansion of macro '__drm_printk'
480 | __drm_printk((drm), warn,, fmt, ##__VA_ARGS__)
| ^~~~~~~~~~~~
drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_color.c:1247:25: note: in expansion of macro 'drm_warn'
1247 | drm_warn(dev, "blob->length (%ld) isn't equal to drm_color_ctm_3x4 (%ld)\n",
| ^~~~~~~~
cc1: all warnings being treated as errors
vim +41 drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgpu.h
539489fc91ea773 Aurabindo Pillai 2020-04-08 40
539489fc91ea773 Aurabindo Pillai 2020-04-08 @41 #define dev_fmt(fmt) "amdgpu: " fmt
539489fc91ea773 Aurabindo Pillai 2020-04-08 42
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h
2024-08-19 20:56 ` [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
2024-08-21 1:07 ` Jeff Johnson
2024-08-21 4:00 ` kernel test robot
@ 2024-08-22 14:55 ` kernel test robot
2 siblings, 0 replies; 69+ messages in thread
From: kernel test robot @ 2024-08-22 14:55 UTC (permalink / raw)
To: Harry Wentland, dri-devel, amd-gfx
Cc: oe-kbuild-all, wayland-devel, Harry Wentland
Hi Harry,
kernel test robot noticed the following build errors:
[auto build test ERROR on drm/drm-next]
[also build test ERROR on drm-exynos/exynos-drm-next drm-intel/for-linux-next drm-intel/for-linux-next-fixes drm-misc/drm-misc-next drm-tip/drm-tip linus/master v6.11-rc4 next-20240822]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Harry-Wentland/drm-Add-helper-for-conversion-from-signed-magnitude/20240820-050138
base: git://anongit.freedesktop.org/drm/drm drm-next
patch link: https://lore.kernel.org/r/20240819205714.316380-21-harry.wentland%40amd.com
patch subject: [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h
config: um-randconfig-r123-20240821 (https://download.01.org/0day-ci/archive/20240822/202408222209.vpdzQng3-lkp@intel.com/config)
compiler: clang version 20.0.0git (https://github.com/llvm/llvm-project 26670e7fa4f032a019d23d56c6a02926e854e8af)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240822/202408222209.vpdzQng3-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408222209.vpdzQng3-lkp@intel.com/
All errors (new ones prefixed by >>):
>> drivers/gpu/drm/tests/drm_fixp_test.c:11:59: error: overflow in expression; result is 9'223'372'036'854'775'807 with type 'long long' [-Werror,-Winteger-overflow]
11 | KUNIT_EXPECT_EQ(test, 0x7fffffffffffffffll, ((1LL << 63) - 1));
| ~~~~~~~~~~~~^~~
include/kunit/test.h:917:34: note: expanded from macro 'KUNIT_EXPECT_EQ'
917 | KUNIT_EXPECT_EQ_MSG(test, left, right, NULL)
| ^~~~~
include/kunit/test.h:922:18: note: expanded from macro 'KUNIT_EXPECT_EQ_MSG'
922 | left, ==, right, \
| ^~~~~
include/kunit/test.h:758:19: note: expanded from macro 'KUNIT_BINARY_INT_ASSERTION'
758 | left, op, right, \
| ^~~~~
include/kunit/test.h:725:33: note: expanded from macro 'KUNIT_BASE_BINARY_ASSERTION'
725 | const typeof(right) __right = (right); \
| ^~~~~
1 error generated.
vim +11 drivers/gpu/drm/tests/drm_fixp_test.c
8
9 static void drm_test_sm2fixp(struct kunit *test)
10 {
> 11 KUNIT_EXPECT_EQ(test, 0x7fffffffffffffffll, ((1LL << 63) - 1));
12
13 /* 1 */
14 KUNIT_EXPECT_EQ(test, drm_int2fixp(1), drm_sm2fixp(1ull << DRM_FIXED_POINT));
15
16 /* -1 */
17 KUNIT_EXPECT_EQ(test, drm_int2fixp(-1), drm_sm2fixp((1ull << 63) | (1ull << DRM_FIXED_POINT)));
18
19 /* 0.5 */
20 KUNIT_EXPECT_EQ(test, drm_fixp_from_fraction(1, 2), drm_sm2fixp(1ull << (DRM_FIXED_POINT - 1)));
21
22 /* -0.5 */
23 KUNIT_EXPECT_EQ(test, drm_fixp_from_fraction(-1, 2), drm_sm2fixp((1ull << 63) | (1ull << (DRM_FIXED_POINT - 1))));
24
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 00/44] Color Pipeline API w/ VKMS
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (43 preceding siblings ...)
2024-08-19 20:57 ` [PATCH v5 44/44] drm/amd/display: Add AMD color pipeline doc Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
2024-10-03 20:09 ` Harry Wentland
2024-08-29 14:55 ` Xaver Hugl
45 siblings, 1 reply; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 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
Le 19/08/24 - 16:56, Harry Wentland a écrit :
> This is an RFC set for a color pipeline API, along with implementations
> in VKMS and amdgpu. It is tested with a set of IGT tests that can be
> found at [1]. The IGT tests run a pixel-by-pixel comparison with an
> allowable delta variation as the goal for these transformations is
> perceptual correctness, not complete pixel accuracy.
>
> v5 of this patchset fleshed out documentation for colorops and the
> various defines that are being introduced.
>
> VKMS supports two named transfer function colorops and two matrix
> colorops.
>
> Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
>
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D Curve EOTF
> 8. 1D LUT
>
> The supported curves for the 1D Curve type are:
> - sRGB EOTF and its inverse
> - PQ EOTF, scaled to [0.0, 125.0] and its inverse
> - BT.2020/BT.709 OETF and its inverse
>
> Note that the 1st and 5th colorops take the EOTF or Inverse
> OETF while the 3rd colorop takes the Inverse EOTF or OETF.
>
> The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
> exists for other drivers to describe their own 3D LUT capability.
>
> This mirrors the color pipeline used by gamescope and presented by
> Melissa Wen, with the exception of the DEGAM LUT, which is not currently
> used. See [1]
> https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
>
> At this point we're hoping to see gamescope and kwin implementations
> take shape. The existing pipeline should be enough to satisfy the
> gamescope use-cases on the drm_plane.
>
> In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
> support to the color pipeline. I have sketched these out already but
> don't have it all hooked up yet. This should not hinder adoption of this
> API for gaming use-cases.
>
> We'll also want to advertise IN_FORMATS on a color pipeline as some
> color pipelines won't be able to work for all IN_FORMATS on a plane.
> Again, I have a sketch but no full implementation yet. This is not
> currently required by the AMD color pipeline and could be added after
> the merge of this set.
>
> VKMS patches could still be improved in a few ways, though the
> payoff might be limited and I would rather focus on other work
> at the moment. The most obvious thing to improve would be to
> eliminate the hard-coded LUTs for identity, and sRGB, and replace
> them with fixed-point math instead.
Hi!
I reviewed your VKMS patches and added a few comments in your series. This
series looks very good.
Thanks for this work,
Louis Chauvet
> There are plenty of things that I would like to see, but they could
> be added after the merge of this patchset:
> - COLOR_ENCODING and COLOR_RANGE
> - IN_FORMATS for a color pipeline
> - Is it possible to support HW which can't bypass entire pipeline?
> - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
> - read-only scaling colorop which defines scaling taps and position
> - named matrices, for things like converting YUV to RGB
> - Add custom LUT colorops to VKMS
>
> IGT tests can be found at [1] or on the igt-dev mailing list.
>
> A kernel branch can be found at [2].
>
> I've also rebased Uma and Chaitanya's patches for the Intel color
> pipeline on top of this to show how I envision them to mesh with
> my changes. The relevant branches can be found at [3] for the kernel
> and [4] for IGT. There were some rebase conflicts in i915 and I'm
> not entirely sure I've resolved all of them correctly, but the branch
> compiles and shows my thoughts for the new DRM concepts to support
> Intel's pipeline.
>
> [1] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-color-pipeline-v5
> [2] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v5
> [3] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-intel-color-pipeline-v5
> [4] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-intel-color-pipeline-v5
>
>
> v5:
> - amdgpu 3D LUT
> - Don't require BYPASS
> - update RFC docs and add to TOC tree
> - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
> - add amdgpu color pipeline doc
> - define SIZE property similar to drm_crtc's GAMMA_SIZE
> - various minor fixes and cleanups
>
> v4:
> - Add amdgpu color pipeline (WIP)
> - Don't block setting of deprecated properties, instead pass client cap
> to atomic check so drivers can ignore these props
> - Drop IOCTL definitions (Pekka)
> - Use enum property for colorop TYPE (Pekka)
> - A few cleanups to the docs (Pekka)
> - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
> - Add missing function declarations (Chaitanya Kumar Borah)
> - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
> - Add helper for creation of pipeline drm_plane property (Pekka)
> - Always create Bypass pipeline (Pekka)
> - A bunch of changes to VKMS kunit tests (Pekka)
> - Fix index in CTM doc (Pekka)
>
> v3:
> - Abandon IOCTLs and discover colorops as clients iterate the pipeline
> - Remove need for libdrm
> - Add color_pipeline client cap and make mutually exclusive with
> COLOR_RANGE and COLOR_ENCODING properties
> - add CTM colorop to VKMS
> - Use include way for kunit testing static functions (Arthur)
> - Make TYPE a range property
> - Move enum drm_colorop_type to uapi header
> - and a bunch of smaller bits that are highlighted in the relevant commit
> description
>
> v2:
> - Rebased on drm-misc-next
> - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
> - Incorporate feedback in color_pipeline.rst doc
> - Add support for sRGB inverse EOTF
> - Add 2nd enumerated TF colorop to VKMS
> - Fix LUTs and some issues with applying LUTs in VKMS
>
> Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
> Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
> Cc: Simon Ser <contact@emersion.fr>
> Cc: Harry Wentland <harry.wentland@amd.com>
> Cc: Melissa Wen <mwen@igalia.com>
> Cc: Jonas Ådahl <jadahl@redhat.com>
> Cc: Sebastian Wick <sebastian.wick@redhat.com>
> Cc: Shashank Sharma <shashank.sharma@amd.com>
> Cc: Alexander Goins <agoins@nvidia.com>
> Cc: Joshua Ashton <joshua@froggi.es>
> Cc: Michel Dänzer <mdaenzer@redhat.com>
> Cc: Aleix Pol <aleixpol@kde.org>
> Cc: Xaver Hugl <xaver.hugl@gmail.com>
> Cc: Victoria Brekenfeld <victoria@system76.com>
> Cc: Sima <daniel@ffwll.ch>
> Cc: Uma Shankar <uma.shankar@intel.com>
> Cc: Naseer Ahmed <quic_naseer@quicinc.com>
> Cc: Christopher Braga <quic_cbraga@quicinc.com>
> Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> Cc: Hector Martin <marcan@marcan.st>
> Cc: Liviu Dudau <Liviu.Dudau@arm.com>
> Cc: Sasha McIntosh <sashamcintosh@google.com>
> Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
>
> Alex Hung (13):
> drm/colorop: define a new macro for_each_new_colorop_in_state
> drm/amd/display: Skip color pipeline initialization for cursor plane
> drm/amd/display: Add support for sRGB EOTF in DEGAM block
> drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
> drm/amd/display: Add support for sRGB EOTF in BLND block
> drm/colorop: Add 1D Curve Custom LUT type
> drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
> drm/amd/display: add 3x4 matrix colorop
> drm/colorop: Add mutliplier type
> drm/amd/display: add multiplier colorop
> drm/amd/display: Swap matrix and multiplier
> drm/colorop: Add 3D LUT supports to color pipeline
> drm/amd/display: add 3D LUT colorop
>
> Harry Wentland (31):
> drm: Add helper for conversion from signed-magnitude
> drm/vkms: Round fixp2int conversion in lerp_u16
> drm/vkms: Add kunit tests for VKMS LUT handling
> drm/doc/rfc: Describe why prescriptive color pipeline is needed
> drm/colorop: Introduce new drm_colorop mode object
> drm/colorop: Add TYPE property
> drm/colorop: Add 1D Curve subtype
> Documentation/gpu: document drm_colorop
> 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: Use s32 for internal color pipeline precision
> drm/vkms: add 3x4 matrix in color pipeline
> drm/tests: Add a few tests around drm_fixed.h
> drm/vkms: Add tests for CTM handling
> drm/colorop: pass plane_color_pipeline client cap to atomic check
> drm/amd/display: Ignore deprecated props when plane_color_pipeline set
> drm/amd/display: Add bypass COLOR PIPELINE
> drm/colorop: Add PQ 125 EOTF and its inverse
> drm/amd/display: Enable support for PQ 125 EOTF and Inverse
> drm/colorop: add BT2020/BT709 OETF and Inverse OETF
> drm/amd/display: Add support for BT.709 and BT.2020 TFs
> drm/colorop: Define LUT_1D interpolation
> drm/colorop: allow non-bypass colorops
> drm/amd/display: Add AMD color pipeline doc
>
> Documentation/gpu/drm-kms.rst | 15 +
> Documentation/gpu/rfc/color_pipeline.rst | 376 ++++++++
> Documentation/gpu/rfc/index.rst | 3 +
> drivers/gpu/drm/Makefile | 1 +
> .../gpu/drm/amd/display/amdgpu_dm/Makefile | 3 +-
> .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +
> .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 581 ++++++++++++-
> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 196 +++++
> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 36 +
> .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 32 +
> drivers/gpu/drm/drm_atomic.c | 184 +++-
> drivers/gpu/drm/drm_atomic_helper.c | 12 +
> drivers/gpu/drm/drm_atomic_state_helper.c | 5 +
> drivers/gpu/drm/drm_atomic_uapi.c | 175 ++++
> drivers/gpu/drm/drm_colorop.c | 562 ++++++++++++
> 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 | 5 +
> 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 | 102 +++
> drivers/gpu/drm/vkms/vkms_composer.c | 115 ++-
> 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 | 130 +++
> include/drm/drm_atomic_uapi.h | 3 +
> include/drm/drm_colorop.h | 427 ++++++++++
> include/drm/drm_file.h | 7 +
> include/drm/drm_fixed.h | 18 +
> include/drm/drm_mode_config.h | 18 +
> include/drm/drm_plane.h | 19 +
> include/uapi/drm/amdgpu_drm.h | 9 -
> include/uapi/drm/drm.h | 16 +
> include/uapi/drm/drm_mode.h | 156 ++++
> 40 files changed, 4595 insertions(+), 34 deletions(-)
> create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> create mode 100644 drivers/gpu/drm/drm_colorop.c
> create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
> create mode 100644 drivers/gpu/drm/vkms/tests/.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
>
> --
> 2.46.0
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 05/44] drm/colorop: Introduce new drm_colorop mode object
2024-08-19 20:56 ` [PATCH v5 05/44] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
0 siblings, 0 replies; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel
Le 19/08/24 - 16:56, Harry Wentland a écrit :
[...]
> +#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 */
Can you add a comment like this, so it is clear that not all the fields
are valid and only a few of them will be used.
/*
* Color properties
*
* The following fields are not always valid, their usage depends
* on the colorop type. See their associated comment for more
* information.
*/
> +
> + /** @state: backpointer to global drm_atomic_state */
> + struct drm_atomic_state *state;
> +};
> +
> +/**
> + * struct drm_colorop - DRM color operation control structure
> + *
> + * A colorop represents one color operation. They can be chained via
> + * the 'next' pointer to build a color pipeline.
> + */
> +struct drm_colorop {
> + /** @dev: parent DRM device */
> + struct drm_device *dev;
> +
> + /**
> + * @head:
> + *
> + * List of all colorops on @dev, linked from &drm_mode_config.colorop_list.
> + * Invariant over the lifetime of @dev and therefore does not need
> + * locking.
> + */
> + struct list_head head;
> +
> + /**
> + * @index: Position inside the mode_config.list, can be used as an array
> + * index. It is invariant over the lifetime of the colorop.
> + */
> + unsigned index;
> +
> + /** @base: base mode object */
> + struct drm_mode_object base;
> +
> + /**
> + * @plane:
> + *
> + * The plane on which the colorop sits. A drm_colorop is always unique
> + * to a plane.
> + */
> + struct drm_plane *plane;
> +
> + /**
> + * @state:
> + *
> + * Current atomic state for this colorop.
> + *
> + * This is protected by @mutex. Note that nonblocking atomic commits
> + * access the current colorop state without taking locks.
> + */
> + struct drm_colorop_state *state;
> +
> + /* colorop properties */
Maybe the same kind of comment here?
> + /** @properties: property tracking for this colorop */
> + struct drm_object_properties properties;
> +
> +};
> +
[...]
Louis Chauvet
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 14/44] drm/vkms: Add enumerated 1D curve colorop
2024-08-19 20:56 ` [PATCH v5 14/44] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
0 siblings, 0 replies; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Alex Hung
Le 19/08/24 - 16:56, Harry Wentland a écrit :
> +static 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;
> + }
There is no need to cleanup previously created colorop here?
> + ret = drm_colorop_curve_1d_init(dev, op, plane, supported_tfs);
> + if (ret)
> + return ret;
ditto?
> + 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++;
The usage of len is not very clear, can you maybe remove it? Or do you
plan to use it for instanciating multiple pipelines?
struct drm_prop_enum_list pipeline;
ret = vkms_initialize_color_pipeline(plane, &pipeline);
drm_plane_create_color_pipeline_property(plane, &pipeline, 1);
> + /* Create COLOR_PIPELINE property and attach */
> + drm_plane_create_color_pipeline_property(plane, pipelines, len);
This function can fail, can you also test the result here?
> +
> + return 0;
> +}
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index 3ecda70c2b55..bc116d16e378 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 +164,53 @@ static void apply_lut(const struct vkms_crtc_state *crtc_state, struct line_buff
> }
> }
>
> +static void apply_colorop(struct pixel_argb_u16 *pixel, struct drm_colorop *colorop)
> +{
> + struct drm_colorop_state *colorop_state = colorop->state;
> +
> + 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);
Maybe add a ONCE/RATELIMITED here, this function is called for every
pixel.
> + 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;
> +
> + while (colorop) {
> + struct drm_colorop_state *colorop_state;
> +
> + if (!colorop)
> + return;
> +
> + colorop_state = colorop->state;
> +
> + if (!colorop_state)
> + return;
> +
> + for (size_t x = 0; x < output_buffer->n_pixels; x++)
> + if (!colorop_state->bypass)
Maybe we can swap the for and the if? I don't know the performance impact
is huge, but it feel more logical to check bypass before iterating.
> + apply_colorop(&output_buffer->pixels[x], colorop);
> +
> + colorop = colorop->next;
> + }
> +}
> +
> /**
> * blend - blend the pixels from all planes and compute crc
> * @wb: The writeback frame buffer metadata
> @@ -200,6 +248,9 @@ 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_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 },
For this LUT and the other, can you add a comment to explain how the
values were generated/found?
[...]
Thanks,
Louis Chauvet
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 02/44] drm/vkms: Round fixp2int conversion in lerp_u16
2024-08-19 20:56 ` [PATCH v5 02/44] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
0 siblings, 0 replies; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel
Le 19/08/24 - 16:56, Harry Wentland a écrit :
> 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 e7441b227b3c..3d6785d081f2 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)
> --
> 2.46.0
>
Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 18/44] drm/vkms: Use s32 for internal color pipeline precision
2024-08-19 20:56 ` [PATCH v5 18/44] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
0 siblings, 0 replies; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel
Le 19/08/24 - 16:56, Harry Wentland a écrit :
> 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.
Maybe you can do this inversion on your first commit so it will reduce the
code change here?
> 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 | 55 ++++++++++++++++++++++------
> drivers/gpu/drm/vkms/vkms_drv.h | 4 ++
> 2 files changed, 47 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index bc116d16e378..6e939d3a6d5c 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)
> {
> struct drm_colorop_state *colorop_state = colorop->state;
>
> @@ -190,24 +190,55 @@ 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;
>
> - 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;
I think this is useless, the while should be sufficient.
> +
> + 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);
clamp can't be used here? And can't we store the result directly in
output_buffer?
> +
> + /* 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 278cf3533f58..b78bc2611746 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;
> };
> --
> 2.46.0
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline
2024-08-19 20:56 ` [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
2024-09-06 17:08 ` Harry Wentland
0 siblings, 1 reply; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel
Le 19/08/24 - 16:56, Harry Wentland a écrit :
> 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.
Is there a reason to use signed-magnitude and not 2s-complement here? I
did not read the whole amd driver, but it seems that the matrix is always
converted to fixed point notation (amdgpu_dm_fixpt_from_s3132 in
amdgpu_dm_color.c). It may reduce the complexity here and in the amd
driver too.
>
> 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 +++++++++++++++++++++++
> 2 files changed, 58 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
> index f61dfde47156..adcb08153a09 100644
> --- a/drivers/gpu/drm/vkms/vkms_colorop.c
> +++ b/drivers/gpu/drm/vkms/vkms_colorop.c
> @@ -37,7 +37,37 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
>
> 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;
> + }
Same as before, don't we leak memory/properties here?
> + 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 6e939d3a6d5c..bd1df06ced85 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);
There is no need to round here, like done in [1]?
I don't know if the performance improvment is huge, bug maybe you can
precompute drm_int2fixp(pixel->r/g/b)?
[1]: https://lore.kernel.org/all/20240802-yuv-v9-12-08a706669e16@bootlin.com/
> +}
> +
> static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
> {
> struct drm_colorop_state *colorop_state = colorop->state;
> @@ -184,6 +208,9 @@ 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);
> }
>
> }
> --
> 2.46.0
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 41/44] drm/colorop: allow non-bypass colorops
2024-08-19 20:57 ` [PATCH v5 41/44] drm/colorop: allow non-bypass colorops Harry Wentland
@ 2024-08-27 17:49 ` Louis Chauvet
0 siblings, 0 replies; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel
Le 19/08/24 - 16:57, Harry Wentland a écrit :
> Not all HW will be able to do bypass on all color
> operations. Introduce an 'allow_bypass' boolean for
> all colorop init functions and only create the BYPASS
> property when it's true.
You did not change the documentation of struct drm_colorop_state, so can
we expect state->bypass to be always valid (ie. false when bypass is not
possible)?
I don't think so, because drm_atomic_helper_colorop_duplicate_state
/ drm_colorop_reset are setting the bypass to true. Maybe you can add
something like this?
/**
* @bypass:
*
* When the property BYPASS exists on this colorop, this stores
* the requested bypass state: true if colorop shall be bypassed,
* false if colorop is enabled.
*/
[...]
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling
2024-08-19 20:56 ` [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
2024-08-21 1:05 ` Jeff Johnson
@ 2024-08-27 17:49 ` Louis Chauvet
2024-08-28 13:36 ` Harry Wentland
1 sibling, 1 reply; 69+ messages in thread
From: Louis Chauvet @ 2024-08-27 17:49 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel, Arthur Grillo
Le 19/08/24 - 16:56, Harry Wentland a écrit :
[...]
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index 3d6785d081f2..3ecda70c2b55 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -435,3 +435,7 @@ int vkms_set_crc_source(struct drm_crtc *crtc, const char *src_name)
diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
index 3d6785d081f2..3ecda70c2b55 100644
--- a/drivers/gpu/drm/vkms/vkms_composer.c
+++ b/drivers/gpu/drm/vkms/vkms_composer.c
@@ -435,3 +435,7 @@ 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>
> return ret;
> }
> +
> +#ifdef CONFIG_DRM_VKMS_KUNIT_TESTS
> +#include "tests/vkms_color_tests.c"
> +#endif
This is very strange to include a .c in this file, is it something done a
lot in the kernel? I can find only one occurence of this pattern in the
kernel [1], the other tests are in their own modules.
In addition it crate many warning during compilations:
warning: symbol 'test_*' was not declared. Should it be static?
As other tests will be introduced (yuv [2], config [3]), it is maybe
interesting to introduce a new module as [2] is doing?
[1]: https://elixir.bootlin.com/linux/v6.11-rc5/source/fs/ext4/mballoc.c#L7047
[2]: https://lore.kernel.org/all/20240809-yuv-v10-14-1a7c764166f7@bootlin.com/
[3]: https://lore.kernel.org/all/20240814-google-remove-crtc-index-from-parameter-v1-15-6e179abf9fd4@bootlin.com/
^ permalink raw reply related [flat|nested] 69+ messages in thread
* Re: [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling
2024-08-27 17:49 ` Louis Chauvet
@ 2024-08-28 13:36 ` Harry Wentland
0 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-08-28 13:36 UTC (permalink / raw)
To: Louis Chauvet; +Cc: dri-devel, amd-gfx, wayland-devel, Arthur Grillo
On 2024-08-27 13:49, Louis Chauvet wrote:
> Le 19/08/24 - 16:56, Harry Wentland a écrit :
>
> [...]
>
>> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
>> index 3d6785d081f2..3ecda70c2b55 100644
>> --- a/drivers/gpu/drm/vkms/vkms_composer.c
>> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
>> @@ -435,3 +435,7 @@ int vkms_set_crc_source(struct drm_crtc *crtc, const char *src_name)
> diff --git a/drivers/gpu/drm/vkms/vkms_composer.c b/drivers/gpu/drm/vkms/vkms_composer.c
> index 3d6785d081f2..3ecda70c2b55 100644
> --- a/drivers/gpu/drm/vkms/vkms_composer.c
> +++ b/drivers/gpu/drm/vkms/vkms_composer.c
> @@ -435,3 +435,7 @@ 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>
>> return ret;
>> }
>> +
>> +#ifdef CONFIG_DRM_VKMS_KUNIT_TESTS
>> +#include "tests/vkms_color_tests.c"
>> +#endif
>
> This is very strange to include a .c in this file, is it something done a
> lot in the kernel? I can find only one occurence of this pattern in the
> kernel [1], the other tests are in their own modules.
>
> In addition it crate many warning during compilations:
> warning: symbol 'test_*' was not declared. Should it be static?
>
> As other tests will be introduced (yuv [2], config [3]), it is maybe
> interesting to introduce a new module as [2] is doing?
The VISIBLE_IF_KUNIT et al. is much nicer than including a .c file.
Thanks for pointing me to them. Will change this.
Harry
>
> [1]: https://elixir.bootlin.com/linux/v6.11-rc5/source/fs/ext4/mballoc.c#L7047
> [2]: https://lore.kernel.org/all/20240809-yuv-v10-14-1a7c764166f7@bootlin.com/
> [3]: https://lore.kernel.org/all/20240814-google-remove-crtc-index-from-parameter-v1-15-6e179abf9fd4@bootlin.com/
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 00/44] Color Pipeline API w/ VKMS
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
` (44 preceding siblings ...)
2024-08-27 17:49 ` [PATCH v5 00/44] Color Pipeline API w/ VKMS Louis Chauvet
@ 2024-08-29 14:55 ` Xaver Hugl
2024-09-09 18:40 ` Harry Wentland
2024-09-10 19:24 ` [PATCH] drm/colorop: get DATA blob ref at duplicate_state Harry Wentland
45 siblings, 2 replies; 69+ messages in thread
From: Xaver Hugl @ 2024-08-29 14:55 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, Victoria Brekenfeld, Sima,
Uma Shankar, Naseer Ahmed, Christopher Braga, Abhinav Kumar,
Arthur Grillo, Hector Martin, Liviu Dudau, Sasha McIntosh,
Chaitanya Kumar Borah
Hi,
I have a WIP implementation of this for KWin at
https://invent.kde.org/plasma/kwin/-/commits/work/zamundaaa/drm-colorop.
It maps KWin's color pipeline to the drm one to get (primary plane
only, for now) direct scanout of HDR content on SDR displays while
doing tone mapping (or SDR on HDR, but that's been possible before).
It currently uses the two 1D custom luts and the 3D lut for this; once
I add support for named 1D curves, it should also make use of that and
the matrix. I haven't tested the Intel version yet, but if it can
match the pipeline, it should work the same there.
In testing your amd-color-pipeline-v5 branch I get two crashes in the
kernel though: https://invent.kde.org/-/snippets/3217
The null pointer dereference only happens sometimes, seemingly
randomly, but if I make KWin generate new blobs each frame, I can 100%
reliably reproduce the list_add corruption problem by just opening
glxgears in fullscreen while the color profile of the display is set
to "built in" (which makes it use the 3D LUT for gamut mapping).
Here's a branch that generates the color pipeline each frame and
triggers this: https://invent.kde.org/plasma/kwin/-/commits/work/zamundaaa/drm-colorop-list-add-corruption
Am Mo., 19. Aug. 2024 um 22:57 Uhr schrieb Harry Wentland
<harry.wentland@amd.com>:
>
> This is an RFC set for a color pipeline API, along with implementations
> in VKMS and amdgpu. It is tested with a set of IGT tests that can be
> found at [1]. The IGT tests run a pixel-by-pixel comparison with an
> allowable delta variation as the goal for these transformations is
> perceptual correctness, not complete pixel accuracy.
>
> v5 of this patchset fleshed out documentation for colorops and the
> various defines that are being introduced.
>
> VKMS supports two named transfer function colorops and two matrix
> colorops.
>
> Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
>
> 1. 1D Curve EOTF
> 2. 3x4 CTM
> 3. Multiplier
> 4. 1D Curve Inverse EOTF
> 5. 1D LUT
> 6. 3D LUT
> 7. 1D Curve EOTF
> 8. 1D LUT
>
> The supported curves for the 1D Curve type are:
> - sRGB EOTF and its inverse
> - PQ EOTF, scaled to [0.0, 125.0] and its inverse
> - BT.2020/BT.709 OETF and its inverse
>
> Note that the 1st and 5th colorops take the EOTF or Inverse
> OETF while the 3rd colorop takes the Inverse EOTF or OETF.
>
> The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
> exists for other drivers to describe their own 3D LUT capability.
>
> This mirrors the color pipeline used by gamescope and presented by
> Melissa Wen, with the exception of the DEGAM LUT, which is not currently
> used. See [1]
> https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
>
> At this point we're hoping to see gamescope and kwin implementations
> take shape. The existing pipeline should be enough to satisfy the
> gamescope use-cases on the drm_plane.
>
> In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
> support to the color pipeline. I have sketched these out already but
> don't have it all hooked up yet. This should not hinder adoption of this
> API for gaming use-cases.
>
> We'll also want to advertise IN_FORMATS on a color pipeline as some
> color pipelines won't be able to work for all IN_FORMATS on a plane.
> Again, I have a sketch but no full implementation yet. This is not
> currently required by the AMD color pipeline and could be added after
> the merge of this set.
>
> VKMS patches could still be improved in a few ways, though the
> payoff might be limited and I would rather focus on other work
> at the moment. The most obvious thing to improve would be to
> eliminate the hard-coded LUTs for identity, and sRGB, and replace
> them with fixed-point math instead.
>
> There are plenty of things that I would like to see, but they could
> be added after the merge of this patchset:
> - COLOR_ENCODING and COLOR_RANGE
> - IN_FORMATS for a color pipeline
> - Is it possible to support HW which can't bypass entire pipeline?
> - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
> - read-only scaling colorop which defines scaling taps and position
> - named matrices, for things like converting YUV to RGB
> - Add custom LUT colorops to VKMS
>
> IGT tests can be found at [1] or on the igt-dev mailing list.
>
> A kernel branch can be found at [2].
>
> I've also rebased Uma and Chaitanya's patches for the Intel color
> pipeline on top of this to show how I envision them to mesh with
> my changes. The relevant branches can be found at [3] for the kernel
> and [4] for IGT. There were some rebase conflicts in i915 and I'm
> not entirely sure I've resolved all of them correctly, but the branch
> compiles and shows my thoughts for the new DRM concepts to support
> Intel's pipeline.
>
> [1] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-color-pipeline-v5
> [2] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v5
> [3] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-intel-color-pipeline-v5
> [4] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-intel-color-pipeline-v5
>
>
> v5:
> - amdgpu 3D LUT
> - Don't require BYPASS
> - update RFC docs and add to TOC tree
> - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
> - add amdgpu color pipeline doc
> - define SIZE property similar to drm_crtc's GAMMA_SIZE
> - various minor fixes and cleanups
>
> v4:
> - Add amdgpu color pipeline (WIP)
> - Don't block setting of deprecated properties, instead pass client cap
> to atomic check so drivers can ignore these props
> - Drop IOCTL definitions (Pekka)
> - Use enum property for colorop TYPE (Pekka)
> - A few cleanups to the docs (Pekka)
> - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
> - Add missing function declarations (Chaitanya Kumar Borah)
> - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
> - Add helper for creation of pipeline drm_plane property (Pekka)
> - Always create Bypass pipeline (Pekka)
> - A bunch of changes to VKMS kunit tests (Pekka)
> - Fix index in CTM doc (Pekka)
>
> v3:
> - Abandon IOCTLs and discover colorops as clients iterate the pipeline
> - Remove need for libdrm
> - Add color_pipeline client cap and make mutually exclusive with
> COLOR_RANGE and COLOR_ENCODING properties
> - add CTM colorop to VKMS
> - Use include way for kunit testing static functions (Arthur)
> - Make TYPE a range property
> - Move enum drm_colorop_type to uapi header
> - and a bunch of smaller bits that are highlighted in the relevant commit
> description
>
> v2:
> - Rebased on drm-misc-next
> - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
> - Incorporate feedback in color_pipeline.rst doc
> - Add support for sRGB inverse EOTF
> - Add 2nd enumerated TF colorop to VKMS
> - Fix LUTs and some issues with applying LUTs in VKMS
>
> Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
> Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
> Cc: Simon Ser <contact@emersion.fr>
> Cc: Harry Wentland <harry.wentland@amd.com>
> Cc: Melissa Wen <mwen@igalia.com>
> Cc: Jonas Ådahl <jadahl@redhat.com>
> Cc: Sebastian Wick <sebastian.wick@redhat.com>
> Cc: Shashank Sharma <shashank.sharma@amd.com>
> Cc: Alexander Goins <agoins@nvidia.com>
> Cc: Joshua Ashton <joshua@froggi.es>
> Cc: Michel Dänzer <mdaenzer@redhat.com>
> Cc: Aleix Pol <aleixpol@kde.org>
> Cc: Xaver Hugl <xaver.hugl@gmail.com>
> Cc: Victoria Brekenfeld <victoria@system76.com>
> Cc: Sima <daniel@ffwll.ch>
> Cc: Uma Shankar <uma.shankar@intel.com>
> Cc: Naseer Ahmed <quic_naseer@quicinc.com>
> Cc: Christopher Braga <quic_cbraga@quicinc.com>
> Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> Cc: Hector Martin <marcan@marcan.st>
> Cc: Liviu Dudau <Liviu.Dudau@arm.com>
> Cc: Sasha McIntosh <sashamcintosh@google.com>
> Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
>
> Alex Hung (13):
> drm/colorop: define a new macro for_each_new_colorop_in_state
> drm/amd/display: Skip color pipeline initialization for cursor plane
> drm/amd/display: Add support for sRGB EOTF in DEGAM block
> drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
> drm/amd/display: Add support for sRGB EOTF in BLND block
> drm/colorop: Add 1D Curve Custom LUT type
> drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
> drm/amd/display: add 3x4 matrix colorop
> drm/colorop: Add mutliplier type
> drm/amd/display: add multiplier colorop
> drm/amd/display: Swap matrix and multiplier
> drm/colorop: Add 3D LUT supports to color pipeline
> drm/amd/display: add 3D LUT colorop
>
> Harry Wentland (31):
> drm: Add helper for conversion from signed-magnitude
> drm/vkms: Round fixp2int conversion in lerp_u16
> drm/vkms: Add kunit tests for VKMS LUT handling
> drm/doc/rfc: Describe why prescriptive color pipeline is needed
> drm/colorop: Introduce new drm_colorop mode object
> drm/colorop: Add TYPE property
> drm/colorop: Add 1D Curve subtype
> Documentation/gpu: document drm_colorop
> 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: Use s32 for internal color pipeline precision
> drm/vkms: add 3x4 matrix in color pipeline
> drm/tests: Add a few tests around drm_fixed.h
> drm/vkms: Add tests for CTM handling
> drm/colorop: pass plane_color_pipeline client cap to atomic check
> drm/amd/display: Ignore deprecated props when plane_color_pipeline set
> drm/amd/display: Add bypass COLOR PIPELINE
> drm/colorop: Add PQ 125 EOTF and its inverse
> drm/amd/display: Enable support for PQ 125 EOTF and Inverse
> drm/colorop: add BT2020/BT709 OETF and Inverse OETF
> drm/amd/display: Add support for BT.709 and BT.2020 TFs
> drm/colorop: Define LUT_1D interpolation
> drm/colorop: allow non-bypass colorops
> drm/amd/display: Add AMD color pipeline doc
>
> Documentation/gpu/drm-kms.rst | 15 +
> Documentation/gpu/rfc/color_pipeline.rst | 376 ++++++++
> Documentation/gpu/rfc/index.rst | 3 +
> drivers/gpu/drm/Makefile | 1 +
> .../gpu/drm/amd/display/amdgpu_dm/Makefile | 3 +-
> .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +
> .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 581 ++++++++++++-
> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 196 +++++
> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 36 +
> .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 32 +
> drivers/gpu/drm/drm_atomic.c | 184 +++-
> drivers/gpu/drm/drm_atomic_helper.c | 12 +
> drivers/gpu/drm/drm_atomic_state_helper.c | 5 +
> drivers/gpu/drm/drm_atomic_uapi.c | 175 ++++
> drivers/gpu/drm/drm_colorop.c | 562 ++++++++++++
> 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 | 5 +
> 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 | 102 +++
> drivers/gpu/drm/vkms/vkms_composer.c | 115 ++-
> 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 | 130 +++
> include/drm/drm_atomic_uapi.h | 3 +
> include/drm/drm_colorop.h | 427 ++++++++++
> include/drm/drm_file.h | 7 +
> include/drm/drm_fixed.h | 18 +
> include/drm/drm_mode_config.h | 18 +
> include/drm/drm_plane.h | 19 +
> include/uapi/drm/amdgpu_drm.h | 9 -
> include/uapi/drm/drm.h | 16 +
> include/uapi/drm/drm_mode.h | 156 ++++
> 40 files changed, 4595 insertions(+), 34 deletions(-)
> create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> create mode 100644 drivers/gpu/drm/drm_colorop.c
> create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
> create mode 100644 drivers/gpu/drm/vkms/tests/.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
>
> --
> 2.46.0
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline
2024-08-27 17:49 ` Louis Chauvet
@ 2024-09-06 17:08 ` Harry Wentland
0 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-09-06 17:08 UTC (permalink / raw)
To: Louis Chauvet; +Cc: dri-devel, amd-gfx, wayland-devel
On 2024-08-27 13:49, Louis Chauvet wrote:
> Le 19/08/24 - 16:56, Harry Wentland a écrit :
>> 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.
>
> Is there a reason to use signed-magnitude and not 2s-complement here? I
> did not read the whole amd driver, but it seems that the matrix is always
> converted to fixed point notation (amdgpu_dm_fixpt_from_s3132 in
> amdgpu_dm_color.c). It may reduce the complexity here and in the amd
> driver too.
>
It's so as to keep the 3x4 matrix the same as the 3x3 one
(drm_color_ctm_3x4 and drm_color_ctm) and the original
3x3 one was defined to use S31.32 sign-magnitude. I'd
prefer 2s-complement myself but that ship has sailed.
>>
>> 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 +++++++++++++++++++++++
>> 2 files changed, 58 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/vkms/vkms_colorop.c b/drivers/gpu/drm/vkms/vkms_colorop.c
>> index f61dfde47156..adcb08153a09 100644
>> --- a/drivers/gpu/drm/vkms/vkms_colorop.c
>> +++ b/drivers/gpu/drm/vkms/vkms_colorop.c
>> @@ -37,7 +37,37 @@ static int vkms_initialize_color_pipeline(struct drm_plane *plane, struct drm_pr
>>
>> 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;
>> + }
>
> Same as before, don't we leak memory/properties here?
>
Thanks. Fix for this will be in v6 (for both vkms and amdgpu).
>> + 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 6e939d3a6d5c..bd1df06ced85 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);
>
> There is no need to round here, like done in [1]?
>
Good call. This makes the CTM test in IGT pass with a [0, 0] bracket,
i.e., it means the CTM is applied with zero error. :)
> I don't know if the performance improvment is huge, bug maybe you can
> precompute drm_int2fixp(pixel->r/g/b)?
>
Good call. Doesn't hurt to precompute.
Harry
> [1]: https://lore.kernel.org/all/20240802-yuv-v9-12-08a706669e16@bootlin.com/
>
>> +}
>> +
>> static void apply_colorop(struct pixel_argb_s32 *pixel, struct drm_colorop *colorop)
>> {
>> struct drm_colorop_state *colorop_state = colorop->state;
>> @@ -184,6 +208,9 @@ 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);
>> }
>>
>> }
>> --
>> 2.46.0
>>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 00/44] Color Pipeline API w/ VKMS
2024-08-29 14:55 ` Xaver Hugl
@ 2024-09-09 18:40 ` Harry Wentland
2024-09-10 20:37 ` Alex Goins
2024-09-10 19:24 ` [PATCH] drm/colorop: get DATA blob ref at duplicate_state Harry Wentland
1 sibling, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-09-09 18:40 UTC (permalink / raw)
To: Xaver Hugl
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, Victoria Brekenfeld, Sima,
Uma Shankar, Naseer Ahmed, Christopher Braga, Abhinav Kumar,
Arthur Grillo, Hector Martin, Liviu Dudau, Sasha McIntosh,
Chaitanya Kumar Borah
On 2024-08-29 10:55, Xaver Hugl wrote:
> Hi,
>
> I have a WIP implementation of this for KWin at
> https://invent.kde.org/plasma/kwin/-/commits/work/zamundaaa/drm-colorop.
> It maps KWin's color pipeline to the drm one to get (primary plane
> only, for now) direct scanout of HDR content on SDR displays while
> doing tone mapping (or SDR on HDR, but that's been possible before).
> It currently uses the two 1D custom luts and the 3D lut for this; once
> I add support for named 1D curves, it should also make use of that and
> the matrix. I haven't tested the Intel version yet, but if it can
> match the pipeline, it should work the same there.
>
> In testing your amd-color-pipeline-v5 branch I get two crashes in the
> kernel though: https://invent.kde.org/-/snippets/3217
> The null pointer dereference only happens sometimes, seemingly
> randomly, but if I make KWin generate new blobs each frame, I can 100%
> reliably reproduce the list_add corruption problem by just opening
> glxgears in fullscreen while the color profile of the display is set
> to "built in" (which makes it use the 3D LUT for gamut mapping).
> Here's a branch that generates the color pipeline each frame and
> triggers this: https://invent.kde.org/plasma/kwin/-/commits/work/zamundaaa/drm-colorop-list-add-corruption
>
Thanks for the test branch. Got a repro and will investigate.
Harry
>
>
> Am Mo., 19. Aug. 2024 um 22:57 Uhr schrieb Harry Wentland
> <harry.wentland@amd.com>:
>>
>> This is an RFC set for a color pipeline API, along with implementations
>> in VKMS and amdgpu. It is tested with a set of IGT tests that can be
>> found at [1]. The IGT tests run a pixel-by-pixel comparison with an
>> allowable delta variation as the goal for these transformations is
>> perceptual correctness, not complete pixel accuracy.
>>
>> v5 of this patchset fleshed out documentation for colorops and the
>> various defines that are being introduced.
>>
>> VKMS supports two named transfer function colorops and two matrix
>> colorops.
>>
>> Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
>>
>> 1. 1D Curve EOTF
>> 2. 3x4 CTM
>> 3. Multiplier
>> 4. 1D Curve Inverse EOTF
>> 5. 1D LUT
>> 6. 3D LUT
>> 7. 1D Curve EOTF
>> 8. 1D LUT
>>
>> The supported curves for the 1D Curve type are:
>> - sRGB EOTF and its inverse
>> - PQ EOTF, scaled to [0.0, 125.0] and its inverse
>> - BT.2020/BT.709 OETF and its inverse
>>
>> Note that the 1st and 5th colorops take the EOTF or Inverse
>> OETF while the 3rd colorop takes the Inverse EOTF or OETF.
>>
>> The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
>> exists for other drivers to describe their own 3D LUT capability.
>>
>> This mirrors the color pipeline used by gamescope and presented by
>> Melissa Wen, with the exception of the DEGAM LUT, which is not currently
>> used. See [1]
>> https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
>>
>> At this point we're hoping to see gamescope and kwin implementations
>> take shape. The existing pipeline should be enough to satisfy the
>> gamescope use-cases on the drm_plane.
>>
>> In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
>> support to the color pipeline. I have sketched these out already but
>> don't have it all hooked up yet. This should not hinder adoption of this
>> API for gaming use-cases.
>>
>> We'll also want to advertise IN_FORMATS on a color pipeline as some
>> color pipelines won't be able to work for all IN_FORMATS on a plane.
>> Again, I have a sketch but no full implementation yet. This is not
>> currently required by the AMD color pipeline and could be added after
>> the merge of this set.
>>
>> VKMS patches could still be improved in a few ways, though the
>> payoff might be limited and I would rather focus on other work
>> at the moment. The most obvious thing to improve would be to
>> eliminate the hard-coded LUTs for identity, and sRGB, and replace
>> them with fixed-point math instead.
>>
>> There are plenty of things that I would like to see, but they could
>> be added after the merge of this patchset:
>> - COLOR_ENCODING and COLOR_RANGE
>> - IN_FORMATS for a color pipeline
>> - Is it possible to support HW which can't bypass entire pipeline?
>> - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>> - read-only scaling colorop which defines scaling taps and position
>> - named matrices, for things like converting YUV to RGB
>> - Add custom LUT colorops to VKMS
>>
>> IGT tests can be found at [1] or on the igt-dev mailing list.
>>
>> A kernel branch can be found at [2].
>>
>> I've also rebased Uma and Chaitanya's patches for the Intel color
>> pipeline on top of this to show how I envision them to mesh with
>> my changes. The relevant branches can be found at [3] for the kernel
>> and [4] for IGT. There were some rebase conflicts in i915 and I'm
>> not entirely sure I've resolved all of them correctly, but the branch
>> compiles and shows my thoughts for the new DRM concepts to support
>> Intel's pipeline.
>>
>> [1] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-color-pipeline-v5
>> [2] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v5
>> [3] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-intel-color-pipeline-v5
>> [4] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-intel-color-pipeline-v5
>>
>>
>> v5:
>> - amdgpu 3D LUT
>> - Don't require BYPASS
>> - update RFC docs and add to TOC tree
>> - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
>> - add amdgpu color pipeline doc
>> - define SIZE property similar to drm_crtc's GAMMA_SIZE
>> - various minor fixes and cleanups
>>
>> v4:
>> - Add amdgpu color pipeline (WIP)
>> - Don't block setting of deprecated properties, instead pass client cap
>> to atomic check so drivers can ignore these props
>> - Drop IOCTL definitions (Pekka)
>> - Use enum property for colorop TYPE (Pekka)
>> - A few cleanups to the docs (Pekka)
>> - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
>> - Add missing function declarations (Chaitanya Kumar Borah)
>> - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
>> - Add helper for creation of pipeline drm_plane property (Pekka)
>> - Always create Bypass pipeline (Pekka)
>> - A bunch of changes to VKMS kunit tests (Pekka)
>> - Fix index in CTM doc (Pekka)
>>
>> v3:
>> - Abandon IOCTLs and discover colorops as clients iterate the pipeline
>> - Remove need for libdrm
>> - Add color_pipeline client cap and make mutually exclusive with
>> COLOR_RANGE and COLOR_ENCODING properties
>> - add CTM colorop to VKMS
>> - Use include way for kunit testing static functions (Arthur)
>> - Make TYPE a range property
>> - Move enum drm_colorop_type to uapi header
>> - and a bunch of smaller bits that are highlighted in the relevant commit
>> description
>>
>> v2:
>> - Rebased on drm-misc-next
>> - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
>> - Incorporate feedback in color_pipeline.rst doc
>> - Add support for sRGB inverse EOTF
>> - Add 2nd enumerated TF colorop to VKMS
>> - Fix LUTs and some issues with applying LUTs in VKMS
>>
>> Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
>> Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
>> Cc: Simon Ser <contact@emersion.fr>
>> Cc: Harry Wentland <harry.wentland@amd.com>
>> Cc: Melissa Wen <mwen@igalia.com>
>> Cc: Jonas Ådahl <jadahl@redhat.com>
>> Cc: Sebastian Wick <sebastian.wick@redhat.com>
>> Cc: Shashank Sharma <shashank.sharma@amd.com>
>> Cc: Alexander Goins <agoins@nvidia.com>
>> Cc: Joshua Ashton <joshua@froggi.es>
>> Cc: Michel Dänzer <mdaenzer@redhat.com>
>> Cc: Aleix Pol <aleixpol@kde.org>
>> Cc: Xaver Hugl <xaver.hugl@gmail.com>
>> Cc: Victoria Brekenfeld <victoria@system76.com>
>> Cc: Sima <daniel@ffwll.ch>
>> Cc: Uma Shankar <uma.shankar@intel.com>
>> Cc: Naseer Ahmed <quic_naseer@quicinc.com>
>> Cc: Christopher Braga <quic_cbraga@quicinc.com>
>> Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
>> Cc: Arthur Grillo <arthurgrillo@riseup.net>
>> Cc: Hector Martin <marcan@marcan.st>
>> Cc: Liviu Dudau <Liviu.Dudau@arm.com>
>> Cc: Sasha McIntosh <sashamcintosh@google.com>
>> Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
>>
>> Alex Hung (13):
>> drm/colorop: define a new macro for_each_new_colorop_in_state
>> drm/amd/display: Skip color pipeline initialization for cursor plane
>> drm/amd/display: Add support for sRGB EOTF in DEGAM block
>> drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
>> drm/amd/display: Add support for sRGB EOTF in BLND block
>> drm/colorop: Add 1D Curve Custom LUT type
>> drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
>> drm/amd/display: add 3x4 matrix colorop
>> drm/colorop: Add mutliplier type
>> drm/amd/display: add multiplier colorop
>> drm/amd/display: Swap matrix and multiplier
>> drm/colorop: Add 3D LUT supports to color pipeline
>> drm/amd/display: add 3D LUT colorop
>>
>> Harry Wentland (31):
>> drm: Add helper for conversion from signed-magnitude
>> drm/vkms: Round fixp2int conversion in lerp_u16
>> drm/vkms: Add kunit tests for VKMS LUT handling
>> drm/doc/rfc: Describe why prescriptive color pipeline is needed
>> drm/colorop: Introduce new drm_colorop mode object
>> drm/colorop: Add TYPE property
>> drm/colorop: Add 1D Curve subtype
>> Documentation/gpu: document drm_colorop
>> 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: Use s32 for internal color pipeline precision
>> drm/vkms: add 3x4 matrix in color pipeline
>> drm/tests: Add a few tests around drm_fixed.h
>> drm/vkms: Add tests for CTM handling
>> drm/colorop: pass plane_color_pipeline client cap to atomic check
>> drm/amd/display: Ignore deprecated props when plane_color_pipeline set
>> drm/amd/display: Add bypass COLOR PIPELINE
>> drm/colorop: Add PQ 125 EOTF and its inverse
>> drm/amd/display: Enable support for PQ 125 EOTF and Inverse
>> drm/colorop: add BT2020/BT709 OETF and Inverse OETF
>> drm/amd/display: Add support for BT.709 and BT.2020 TFs
>> drm/colorop: Define LUT_1D interpolation
>> drm/colorop: allow non-bypass colorops
>> drm/amd/display: Add AMD color pipeline doc
>>
>> Documentation/gpu/drm-kms.rst | 15 +
>> Documentation/gpu/rfc/color_pipeline.rst | 376 ++++++++
>> Documentation/gpu/rfc/index.rst | 3 +
>> drivers/gpu/drm/Makefile | 1 +
>> .../gpu/drm/amd/display/amdgpu_dm/Makefile | 3 +-
>> .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +
>> .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 581 ++++++++++++-
>> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 196 +++++
>> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 36 +
>> .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 32 +
>> drivers/gpu/drm/drm_atomic.c | 184 +++-
>> drivers/gpu/drm/drm_atomic_helper.c | 12 +
>> drivers/gpu/drm/drm_atomic_state_helper.c | 5 +
>> drivers/gpu/drm/drm_atomic_uapi.c | 175 ++++
>> drivers/gpu/drm/drm_colorop.c | 562 ++++++++++++
>> 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 | 5 +
>> 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 | 102 +++
>> drivers/gpu/drm/vkms/vkms_composer.c | 115 ++-
>> 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 | 130 +++
>> include/drm/drm_atomic_uapi.h | 3 +
>> include/drm/drm_colorop.h | 427 ++++++++++
>> include/drm/drm_file.h | 7 +
>> include/drm/drm_fixed.h | 18 +
>> include/drm/drm_mode_config.h | 18 +
>> include/drm/drm_plane.h | 19 +
>> include/uapi/drm/amdgpu_drm.h | 9 -
>> include/uapi/drm/drm.h | 16 +
>> include/uapi/drm/drm_mode.h | 156 ++++
>> 40 files changed, 4595 insertions(+), 34 deletions(-)
>> create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
>> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
>> create mode 100644 drivers/gpu/drm/drm_colorop.c
>> create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
>> create mode 100644 drivers/gpu/drm/vkms/tests/.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
>>
>> --
>> 2.46.0
>>
^ permalink raw reply [flat|nested] 69+ messages in thread
* [PATCH] drm/colorop: get DATA blob ref at duplicate_state
2024-08-29 14:55 ` Xaver Hugl
2024-09-09 18:40 ` Harry Wentland
@ 2024-09-10 19:24 ` Harry Wentland
2024-09-13 0:17 ` Xaver Hugl
1 sibling, 1 reply; 69+ messages in thread
From: Harry Wentland @ 2024-09-10 19:24 UTC (permalink / raw)
To: xaver.hugl; +Cc: dri-devel, amd-gfx, wayland-devel, Harry Wentland
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
---
It was a stupid mistake on my part. The duplicate_state
function needs to take a reference to the blob.
This should fix it. Please give it a try if you can.
I'll roll it into the patch that introduces the bug in
my v6.
Harry
drivers/gpu/drm/drm_colorop.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
index 86f7f084a73f..fd1cd934df48 100644
--- a/drivers/gpu/drm/drm_colorop.c
+++ b/drivers/gpu/drm/drm_colorop.c
@@ -401,6 +401,9 @@ static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colo
{
memcpy(state, colorop->state, sizeof(*state));
+ if (state->data)
+ drm_property_blob_get(state->data);
+
state->bypass = true;
}
--
2.46.0
^ permalink raw reply related [flat|nested] 69+ messages in thread
* Re: [PATCH v5 00/44] Color Pipeline API w/ VKMS
2024-09-09 18:40 ` Harry Wentland
@ 2024-09-10 20:37 ` Alex Goins
2024-10-03 20:00 ` Harry Wentland
0 siblings, 1 reply; 69+ messages in thread
From: Alex Goins @ 2024-09-10 20:37 UTC (permalink / raw)
To: Harry Wentland
Cc: Xaver Hugl, dri-devel, amd-gfx, wayland-devel, Ville Syrjala,
Pekka Paalanen, Simon Ser, Melissa Wen, Jonas Ådahl,
Sebastian Wick, Shashank Sharma, Joshua Ashton,
Michel Dänzer, Aleix Pol, Victoria Brekenfeld, Sima,
Uma Shankar, Naseer Ahmed, Christopher Braga, Abhinav Kumar,
Arthur Grillo, Hector Martin, Liviu Dudau, Sasha McIntosh,
Chaitanya Kumar Borah
[-- Attachment #1: Type: text/plain, Size: 16363 bytes --]
Hi Harry,
Thanks for this. I just want to remind about a few things that would be required
for NVIDIA hardware, as discussed at the Display Next Hackfest -- fully
understand that they aren't currently included in this series because they
aren't required on AMD hardware. Allowing color ops to be non-bypassable is a
big improvement in this direction.
In each of these cases, it should be possible to represent NVIDIA's color
pipeline using the existing color ops, but it would rely on rejecting invalid
configurations in atomic check, with no way to communicate the limitations to
clients.
- Tag IN_FORMATS for a color pipeline (as you have mentioned).
+ Required to communicate that the first NVIDIA pre-blending LUT is
incompatible with FP16 inputs.
- Introduce a single-channel 1DLUT color op, or add the ability to flag
individual channels of an ordinary 1DLUT as read-only.
+ Required to handle NVIDIA's 1D tone-mapping LUT designed to operate only
on the intensity component of ICtCp.
- Allow marking color ops as read-only and non-bypassable (halfway there)
and/or allow marking a color op as requiring a modeset to update.
+ Required to handle specific LUTs that can't have their entries changed
while in use without visible glitches/tearing, yet must be enabled to
support tone-mapping and pre-blend scaling.
+ In lieu of the first option, it may be suitable to just use the existing
1D curve colorop with only a single curve supported, marked as
non-bypassable. Allowing modesets as in the second option would allow for
the possibility of selecting different curves, however.
Thanks,
Alex Goins
NVIDIA
On Mon, 9 Sep 2024, Harry Wentland wrote:
> External email: Use caution opening links or attachments
>
>
> On 2024-08-29 10:55, Xaver Hugl wrote:
> > Hi,
> >
> > I have a WIP implementation of this for KWin at
> > https://invent.kde.org/plasma/kwin/-/commits/work/zamundaaa/drm-colorop.
> > It maps KWin's color pipeline to the drm one to get (primary plane
> > only, for now) direct scanout of HDR content on SDR displays while
> > doing tone mapping (or SDR on HDR, but that's been possible before).
> > It currently uses the two 1D custom luts and the 3D lut for this; once
> > I add support for named 1D curves, it should also make use of that and
> > the matrix. I haven't tested the Intel version yet, but if it can
> > match the pipeline, it should work the same there.
> >
> > In testing your amd-color-pipeline-v5 branch I get two crashes in the
> > kernel though: https://invent.kde.org/-/snippets/3217
> > The null pointer dereference only happens sometimes, seemingly
> > randomly, but if I make KWin generate new blobs each frame, I can 100%
> > reliably reproduce the list_add corruption problem by just opening
> > glxgears in fullscreen while the color profile of the display is set
> > to "built in" (which makes it use the 3D LUT for gamut mapping).
> > Here's a branch that generates the color pipeline each frame and
> > triggers this: https://invent.kde.org/plasma/kwin/-/commits/work/zamundaaa/drm-colorop-list-add-corruption
> >
>
> Thanks for the test branch. Got a repro and will investigate.
>
> Harry
>
> >
> >
> > Am Mo., 19. Aug. 2024 um 22:57 Uhr schrieb Harry Wentland
> > <harry.wentland@amd.com>:
> >>
> >> This is an RFC set for a color pipeline API, along with implementations
> >> in VKMS and amdgpu. It is tested with a set of IGT tests that can be
> >> found at [1]. The IGT tests run a pixel-by-pixel comparison with an
> >> allowable delta variation as the goal for these transformations is
> >> perceptual correctness, not complete pixel accuracy.
> >>
> >> v5 of this patchset fleshed out documentation for colorops and the
> >> various defines that are being introduced.
> >>
> >> VKMS supports two named transfer function colorops and two matrix
> >> colorops.
> >>
> >> Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
> >>
> >> 1. 1D Curve EOTF
> >> 2. 3x4 CTM
> >> 3. Multiplier
> >> 4. 1D Curve Inverse EOTF
> >> 5. 1D LUT
> >> 6. 3D LUT
> >> 7. 1D Curve EOTF
> >> 8. 1D LUT
> >>
> >> The supported curves for the 1D Curve type are:
> >> - sRGB EOTF and its inverse
> >> - PQ EOTF, scaled to [0.0, 125.0] and its inverse
> >> - BT.2020/BT.709 OETF and its inverse
> >>
> >> Note that the 1st and 5th colorops take the EOTF or Inverse
> >> OETF while the 3rd colorop takes the Inverse EOTF or OETF.
> >>
> >> The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
> >> exists for other drivers to describe their own 3D LUT capability.
> >>
> >> This mirrors the color pipeline used by gamescope and presented by
> >> Melissa Wen, with the exception of the DEGAM LUT, which is not currently
> >> used. See [1]
> >> https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
> >>
> >> At this point we're hoping to see gamescope and kwin implementations
> >> take shape. The existing pipeline should be enough to satisfy the
> >> gamescope use-cases on the drm_plane.
> >>
> >> In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
> >> support to the color pipeline. I have sketched these out already but
> >> don't have it all hooked up yet. This should not hinder adoption of this
> >> API for gaming use-cases.
> >>
> >> We'll also want to advertise IN_FORMATS on a color pipeline as some
> >> color pipelines won't be able to work for all IN_FORMATS on a plane.
> >> Again, I have a sketch but no full implementation yet. This is not
> >> currently required by the AMD color pipeline and could be added after
> >> the merge of this set.
> >>
> >> VKMS patches could still be improved in a few ways, though the
> >> payoff might be limited and I would rather focus on other work
> >> at the moment. The most obvious thing to improve would be to
> >> eliminate the hard-coded LUTs for identity, and sRGB, and replace
> >> them with fixed-point math instead.
> >>
> >> There are plenty of things that I would like to see, but they could
> >> be added after the merge of this patchset:
> >> - COLOR_ENCODING and COLOR_RANGE
> >> - IN_FORMATS for a color pipeline
> >> - Is it possible to support HW which can't bypass entire pipeline?
> >> - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
> >> - read-only scaling colorop which defines scaling taps and position
> >> - named matrices, for things like converting YUV to RGB
> >> - Add custom LUT colorops to VKMS
> >>
> >> IGT tests can be found at [1] or on the igt-dev mailing list.
> >>
> >> A kernel branch can be found at [2].
> >>
> >> I've also rebased Uma and Chaitanya's patches for the Intel color
> >> pipeline on top of this to show how I envision them to mesh with
> >> my changes. The relevant branches can be found at [3] for the kernel
> >> and [4] for IGT. There were some rebase conflicts in i915 and I'm
> >> not entirely sure I've resolved all of them correctly, but the branch
> >> compiles and shows my thoughts for the new DRM concepts to support
> >> Intel's pipeline.
> >>
> >> [1] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-color-pipeline-v5
> >> [2] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v5
> >> [3] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-intel-color-pipeline-v5
> >> [4] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-intel-color-pipeline-v5
> >>
> >>
> >> v5:
> >> - amdgpu 3D LUT
> >> - Don't require BYPASS
> >> - update RFC docs and add to TOC tree
> >> - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
> >> - add amdgpu color pipeline doc
> >> - define SIZE property similar to drm_crtc's GAMMA_SIZE
> >> - various minor fixes and cleanups
> >>
> >> v4:
> >> - Add amdgpu color pipeline (WIP)
> >> - Don't block setting of deprecated properties, instead pass client cap
> >> to atomic check so drivers can ignore these props
> >> - Drop IOCTL definitions (Pekka)
> >> - Use enum property for colorop TYPE (Pekka)
> >> - A few cleanups to the docs (Pekka)
> >> - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
> >> - Add missing function declarations (Chaitanya Kumar Borah)
> >> - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
> >> - Add helper for creation of pipeline drm_plane property (Pekka)
> >> - Always create Bypass pipeline (Pekka)
> >> - A bunch of changes to VKMS kunit tests (Pekka)
> >> - Fix index in CTM doc (Pekka)
> >>
> >> v3:
> >> - Abandon IOCTLs and discover colorops as clients iterate the pipeline
> >> - Remove need for libdrm
> >> - Add color_pipeline client cap and make mutually exclusive with
> >> COLOR_RANGE and COLOR_ENCODING properties
> >> - add CTM colorop to VKMS
> >> - Use include way for kunit testing static functions (Arthur)
> >> - Make TYPE a range property
> >> - Move enum drm_colorop_type to uapi header
> >> - and a bunch of smaller bits that are highlighted in the relevant commit
> >> description
> >>
> >> v2:
> >> - Rebased on drm-misc-next
> >> - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
> >> - Incorporate feedback in color_pipeline.rst doc
> >> - Add support for sRGB inverse EOTF
> >> - Add 2nd enumerated TF colorop to VKMS
> >> - Fix LUTs and some issues with applying LUTs in VKMS
> >>
> >> Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
> >> Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
> >> Cc: Simon Ser <contact@emersion.fr>
> >> Cc: Harry Wentland <harry.wentland@amd.com>
> >> Cc: Melissa Wen <mwen@igalia.com>
> >> Cc: Jonas Ådahl <jadahl@redhat.com>
> >> Cc: Sebastian Wick <sebastian.wick@redhat.com>
> >> Cc: Shashank Sharma <shashank.sharma@amd.com>
> >> Cc: Alexander Goins <agoins@nvidia.com>
> >> Cc: Joshua Ashton <joshua@froggi.es>
> >> Cc: Michel Dänzer <mdaenzer@redhat.com>
> >> Cc: Aleix Pol <aleixpol@kde.org>
> >> Cc: Xaver Hugl <xaver.hugl@gmail.com>
> >> Cc: Victoria Brekenfeld <victoria@system76.com>
> >> Cc: Sima <daniel@ffwll.ch>
> >> Cc: Uma Shankar <uma.shankar@intel.com>
> >> Cc: Naseer Ahmed <quic_naseer@quicinc.com>
> >> Cc: Christopher Braga <quic_cbraga@quicinc.com>
> >> Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
> >> Cc: Arthur Grillo <arthurgrillo@riseup.net>
> >> Cc: Hector Martin <marcan@marcan.st>
> >> Cc: Liviu Dudau <Liviu.Dudau@arm.com>
> >> Cc: Sasha McIntosh <sashamcintosh@google.com>
> >> Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
> >>
> >> Alex Hung (13):
> >> drm/colorop: define a new macro for_each_new_colorop_in_state
> >> drm/amd/display: Skip color pipeline initialization for cursor plane
> >> drm/amd/display: Add support for sRGB EOTF in DEGAM block
> >> drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
> >> drm/amd/display: Add support for sRGB EOTF in BLND block
> >> drm/colorop: Add 1D Curve Custom LUT type
> >> drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
> >> drm/amd/display: add 3x4 matrix colorop
> >> drm/colorop: Add mutliplier type
> >> drm/amd/display: add multiplier colorop
> >> drm/amd/display: Swap matrix and multiplier
> >> drm/colorop: Add 3D LUT supports to color pipeline
> >> drm/amd/display: add 3D LUT colorop
> >>
> >> Harry Wentland (31):
> >> drm: Add helper for conversion from signed-magnitude
> >> drm/vkms: Round fixp2int conversion in lerp_u16
> >> drm/vkms: Add kunit tests for VKMS LUT handling
> >> drm/doc/rfc: Describe why prescriptive color pipeline is needed
> >> drm/colorop: Introduce new drm_colorop mode object
> >> drm/colorop: Add TYPE property
> >> drm/colorop: Add 1D Curve subtype
> >> Documentation/gpu: document drm_colorop
> >> 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: Use s32 for internal color pipeline precision
> >> drm/vkms: add 3x4 matrix in color pipeline
> >> drm/tests: Add a few tests around drm_fixed.h
> >> drm/vkms: Add tests for CTM handling
> >> drm/colorop: pass plane_color_pipeline client cap to atomic check
> >> drm/amd/display: Ignore deprecated props when plane_color_pipeline set
> >> drm/amd/display: Add bypass COLOR PIPELINE
> >> drm/colorop: Add PQ 125 EOTF and its inverse
> >> drm/amd/display: Enable support for PQ 125 EOTF and Inverse
> >> drm/colorop: add BT2020/BT709 OETF and Inverse OETF
> >> drm/amd/display: Add support for BT.709 and BT.2020 TFs
> >> drm/colorop: Define LUT_1D interpolation
> >> drm/colorop: allow non-bypass colorops
> >> drm/amd/display: Add AMD color pipeline doc
> >>
> >> Documentation/gpu/drm-kms.rst | 15 +
> >> Documentation/gpu/rfc/color_pipeline.rst | 376 ++++++++
> >> Documentation/gpu/rfc/index.rst | 3 +
> >> drivers/gpu/drm/Makefile | 1 +
> >> .../gpu/drm/amd/display/amdgpu_dm/Makefile | 3 +-
> >> .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +
> >> .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 581 ++++++++++++-
> >> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 196 +++++
> >> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 36 +
> >> .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 32 +
> >> drivers/gpu/drm/drm_atomic.c | 184 +++-
> >> drivers/gpu/drm/drm_atomic_helper.c | 12 +
> >> drivers/gpu/drm/drm_atomic_state_helper.c | 5 +
> >> drivers/gpu/drm/drm_atomic_uapi.c | 175 ++++
> >> drivers/gpu/drm/drm_colorop.c | 562 ++++++++++++
> >> 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 | 5 +
> >> 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 | 102 +++
> >> drivers/gpu/drm/vkms/vkms_composer.c | 115 ++-
> >> 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 | 130 +++
> >> include/drm/drm_atomic_uapi.h | 3 +
> >> include/drm/drm_colorop.h | 427 ++++++++++
> >> include/drm/drm_file.h | 7 +
> >> include/drm/drm_fixed.h | 18 +
> >> include/drm/drm_mode_config.h | 18 +
> >> include/drm/drm_plane.h | 19 +
> >> include/uapi/drm/amdgpu_drm.h | 9 -
> >> include/uapi/drm/drm.h | 16 +
> >> include/uapi/drm/drm_mode.h | 156 ++++
> >> 40 files changed, 4595 insertions(+), 34 deletions(-)
> >> create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
> >> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
> >> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
> >> create mode 100644 drivers/gpu/drm/drm_colorop.c
> >> create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
> >> create mode 100644 drivers/gpu/drm/vkms/tests/.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
> >>
> >> --
> >> 2.46.0
> >>
>
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH] drm/colorop: get DATA blob ref at duplicate_state
2024-09-10 19:24 ` [PATCH] drm/colorop: get DATA blob ref at duplicate_state Harry Wentland
@ 2024-09-13 0:17 ` Xaver Hugl
0 siblings, 0 replies; 69+ messages in thread
From: Xaver Hugl @ 2024-09-13 0:17 UTC (permalink / raw)
To: Harry Wentland; +Cc: dri-devel, amd-gfx, wayland-devel
Hi Harry,
I applied this patch and the two issues I mentioned before are gone.
I noticed a new problem though: Changes in the COLOR_PIPELINE value
aren't always applied immediately. For testing I played an HDR video
on an SDR screen with the work/zamundaaa/drm-colorop KWin branch, and
made it fullscreen. When I do that, the video is reinterpreted as sRGB
until the pipeline actually gets used at some later point in time, and
when opening another app to interrupt direct scanout while the
pipeline is being applied, the previous pipeline keeps being applied
until it switches to bypass. It often applies the proper value when
the cursor plane gets enabled or disabled, but that doesn't seem to be
reliable and it sometimes happens without cursor changes.
I checked with drm_info that KWin has actually set COLOR_PIPELINE to
the pipeline or bypass, so I'm confident that this issue is on the
kernel side.
When the pipeline is actually applied, it works correctly (outside of
some banding, but that's expected with the current state of the KWin
branch).
Am Di., 10. Sept. 2024 um 21:24 Uhr schrieb Harry Wentland
<harry.wentland@amd.com>:
>
> Signed-off-by: Harry Wentland <harry.wentland@amd.com>
> ---
>
> It was a stupid mistake on my part. The duplicate_state
> function needs to take a reference to the blob.
>
> This should fix it. Please give it a try if you can.
> I'll roll it into the patch that introduces the bug in
> my v6.
>
> Harry
>
> drivers/gpu/drm/drm_colorop.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/drivers/gpu/drm/drm_colorop.c b/drivers/gpu/drm/drm_colorop.c
> index 86f7f084a73f..fd1cd934df48 100644
> --- a/drivers/gpu/drm/drm_colorop.c
> +++ b/drivers/gpu/drm/drm_colorop.c
> @@ -401,6 +401,9 @@ static void __drm_atomic_helper_colorop_duplicate_state(struct drm_colorop *colo
> {
> memcpy(state, colorop->state, sizeof(*state));
>
> + if (state->data)
> + drm_property_blob_get(state->data);
> +
> state->bypass = true;
> }
>
> --
> 2.46.0
>
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 00/44] Color Pipeline API w/ VKMS
2024-09-10 20:37 ` Alex Goins
@ 2024-10-03 20:00 ` Harry Wentland
0 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-10-03 20:00 UTC (permalink / raw)
To: Alex Goins
Cc: Xaver Hugl, dri-devel, amd-gfx, wayland-devel, Ville Syrjala,
Pekka Paalanen, Simon Ser, Melissa Wen, Jonas Ådahl,
Sebastian Wick, Shashank Sharma, Joshua Ashton,
Michel Dänzer, Aleix Pol, Victoria Brekenfeld, Sima,
Uma Shankar, Naseer Ahmed, Christopher Braga, Abhinav Kumar,
Arthur Grillo, Hector Martin, Liviu Dudau, Sasha McIntosh,
Chaitanya Kumar Borah
On 2024-09-10 16:37, Alex Goins wrote:
> Hi Harry,
>
> Thanks for this. I just want to remind about a few things that would be required
> for NVIDIA hardware, as discussed at the Display Next Hackfest -- fully
> understand that they aren't currently included in this series because they
> aren't required on AMD hardware. Allowing color ops to be non-bypassable is a
> big improvement in this direction.
>
Thanks for your feedback.
> In each of these cases, it should be possible to represent NVIDIA's color
> pipeline using the existing color ops, but it would rely on rejecting invalid
> configurations in atomic check, with no way to communicate the limitations to
> clients.
>
> - Tag IN_FORMATS for a color pipeline (as you have mentioned).
> + Required to communicate that the first NVIDIA pre-blending LUT is
> incompatible with FP16 inputs.
>
I have a sketch in a branch [1] but the work was a bit larger than
I anticipated so I parked it for now.
> - Introduce a single-channel 1DLUT color op, or add the ability to flag
> individual channels of an ordinary 1DLUT as read-only.
> + Required to handle NVIDIA's 1D tone-mapping LUT designed to operate only
> on the intensity component of ICtCp.
>
I'm reluctant to add a new single-channel colorop type without having the
ability to test it. Though I guess VKMS could serve as a stand-in but that
would mean an implementation for it in VKMS.
> - Allow marking color ops as read-only and non-bypassable (halfway there)
> and/or allow marking a color op as requiring a modeset to update.
> + Required to handle specific LUTs that can't have their entries changed
> while in use without visible glitches/tearing, yet must be enabled to
> support tone-mapping and pre-blend scaling.
> + In lieu of the first option, it may be suitable to just use the existing
> 1D curve colorop with only a single curve supported, marked as
> non-bypassable. Allowing modesets as in the second option would allow for
> the possibility of selecting different curves, however.
>
Flags should be easy to add to a colorop because a colorop is just a DRM
object, so we can extend it with new properties. I envision this to be a
FLAGS bitmask property, so the implementation can probably look similar
to the rotation property [2]. And the flags can be REQUIRES_MODESET and
SINGLE_CHANNEL, or something like that.
[1] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v6-in-formats
[2] https://drmdb.emersion.fr/properties/3435973836/rotation
Harry
> Thanks,
> Alex Goins
> NVIDIA
^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [PATCH v5 00/44] Color Pipeline API w/ VKMS
2024-08-27 17:49 ` [PATCH v5 00/44] Color Pipeline API w/ VKMS Louis Chauvet
@ 2024-10-03 20:09 ` Harry Wentland
0 siblings, 0 replies; 69+ messages in thread
From: Harry Wentland @ 2024-10-03 20:09 UTC (permalink / raw)
To: Louis Chauvet
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 2024-08-27 13:49, Louis Chauvet wrote:
> Le 19/08/24 - 16:56, Harry Wentland a écrit :
>> This is an RFC set for a color pipeline API, along with implementations
>> in VKMS and amdgpu. It is tested with a set of IGT tests that can be
>> found at [1]. The IGT tests run a pixel-by-pixel comparison with an
>> allowable delta variation as the goal for these transformations is
>> perceptual correctness, not complete pixel accuracy.
>>
>> v5 of this patchset fleshed out documentation for colorops and the
>> various defines that are being introduced.
>>
>> VKMS supports two named transfer function colorops and two matrix
>> colorops.
>>
>> Amdgpu advertises the following pipeline for GPUs with DCN 3 or newer:
>>
>> 1. 1D Curve EOTF
>> 2. 3x4 CTM
>> 3. Multiplier
>> 4. 1D Curve Inverse EOTF
>> 5. 1D LUT
>> 6. 3D LUT
>> 7. 1D Curve EOTF
>> 8. 1D LUT
>>
>> The supported curves for the 1D Curve type are:
>> - sRGB EOTF and its inverse
>> - PQ EOTF, scaled to [0.0, 125.0] and its inverse
>> - BT.2020/BT.709 OETF and its inverse
>>
>> Note that the 1st and 5th colorops take the EOTF or Inverse
>> OETF while the 3rd colorop takes the Inverse EOTF or OETF.
>>
>> The 3D LUT is a 17^3 tetrahedrally interpolated LUT but the mechanism
>> exists for other drivers to describe their own 3D LUT capability.
>>
>> This mirrors the color pipeline used by gamescope and presented by
>> Melissa Wen, with the exception of the DEGAM LUT, which is not currently
>> used. See [1]
>> https://indico.freedesktop.org/event/4/contributions/186/attachments/138/218/xdc2023-TheRainbowTreasureMap-MelissaWen.pdf
>>
>> At this point we're hoping to see gamescope and kwin implementations
>> take shape. The existing pipeline should be enough to satisfy the
>> gamescope use-cases on the drm_plane.
>>
>> In order to support YUV we'll need to add COLOR_ENCODING and COLOR_RANGE
>> support to the color pipeline. I have sketched these out already but
>> don't have it all hooked up yet. This should not hinder adoption of this
>> API for gaming use-cases.
>>
>> We'll also want to advertise IN_FORMATS on a color pipeline as some
>> color pipelines won't be able to work for all IN_FORMATS on a plane.
>> Again, I have a sketch but no full implementation yet. This is not
>> currently required by the AMD color pipeline and could be added after
>> the merge of this set.
>>
>> VKMS patches could still be improved in a few ways, though the
>> payoff might be limited and I would rather focus on other work
>> at the moment. The most obvious thing to improve would be to
>> eliminate the hard-coded LUTs for identity, and sRGB, and replace
>> them with fixed-point math instead.
>
> Hi!
>
> I reviewed your VKMS patches and added a few comments in your series. This
> series looks very good.
>
Thanks so much for your review. v6 just went out.
Harry
> Thanks for this work,
> Louis Chauvet
>
>> There are plenty of things that I would like to see, but they could
>> be added after the merge of this patchset:
>> - COLOR_ENCODING and COLOR_RANGE
>> - IN_FORMATS for a color pipeline
>> - Is it possible to support HW which can't bypass entire pipeline?
>> - Can we do a LOAD / COMMIT model for LUTs (and other properties)?
>> - read-only scaling colorop which defines scaling taps and position
>> - named matrices, for things like converting YUV to RGB
>> - Add custom LUT colorops to VKMS
>>
>> IGT tests can be found at [1] or on the igt-dev mailing list.
>>
>> A kernel branch can be found at [2].
>>
>> I've also rebased Uma and Chaitanya's patches for the Intel color
>> pipeline on top of this to show how I envision them to mesh with
>> my changes. The relevant branches can be found at [3] for the kernel
>> and [4] for IGT. There were some rebase conflicts in i915 and I'm
>> not entirely sure I've resolved all of them correctly, but the branch
>> compiles and shows my thoughts for the new DRM concepts to support
>> Intel's pipeline.
>>
>> [1] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-color-pipeline-v5
>> [2] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-color-pipeline-v5
>> [3] https://gitlab.freedesktop.org/hwentland/linux/-/tree/amd-intel-color-pipeline-v5
>> [4] https://gitlab.freedesktop.org/hwentland/igt-gpu-tools/-/tree/amd-intel-color-pipeline-v5
>>
>>
>> v5:
>> - amdgpu 3D LUT
>> - Don't require BYPASS
>> - update RFC docs and add to TOC tree
>> - add drm_colorop and COLOR_PIPELINE kernel docs (non-RFC)
>> - add amdgpu color pipeline doc
>> - define SIZE property similar to drm_crtc's GAMMA_SIZE
>> - various minor fixes and cleanups
>>
>> v4:
>> - Add amdgpu color pipeline (WIP)
>> - Don't block setting of deprecated properties, instead pass client cap
>> to atomic check so drivers can ignore these props
>> - Drop IOCTL definitions (Pekka)
>> - Use enum property for colorop TYPE (Pekka)
>> - A few cleanups to the docs (Pekka)
>> - Rework the TYPE enum to name relation to avoid code duplication (Pekka)
>> - Add missing function declarations (Chaitanya Kumar Borah)
>> - Allow setting of NEXT property to NULL in _set_ function (Chaitanya Kumar Borah)
>> - Add helper for creation of pipeline drm_plane property (Pekka)
>> - Always create Bypass pipeline (Pekka)
>> - A bunch of changes to VKMS kunit tests (Pekka)
>> - Fix index in CTM doc (Pekka)
>>
>> v3:
>> - Abandon IOCTLs and discover colorops as clients iterate the pipeline
>> - Remove need for libdrm
>> - Add color_pipeline client cap and make mutually exclusive with
>> COLOR_RANGE and COLOR_ENCODING properties
>> - add CTM colorop to VKMS
>> - Use include way for kunit testing static functions (Arthur)
>> - Make TYPE a range property
>> - Move enum drm_colorop_type to uapi header
>> - and a bunch of smaller bits that are highlighted in the relevant commit
>> description
>>
>> v2:
>> - Rebased on drm-misc-next
>> - Introduce a VKMS Kunit so we can test LUT functionality in vkms_composer
>> - Incorporate feedback in color_pipeline.rst doc
>> - Add support for sRGB inverse EOTF
>> - Add 2nd enumerated TF colorop to VKMS
>> - Fix LUTs and some issues with applying LUTs in VKMS
>>
>> Cc: Ville Syrjala <ville.syrjala@linux.intel.com>
>> Cc: Pekka Paalanen <pekka.paalanen@collabora.com>
>> Cc: Simon Ser <contact@emersion.fr>
>> Cc: Harry Wentland <harry.wentland@amd.com>
>> Cc: Melissa Wen <mwen@igalia.com>
>> Cc: Jonas Ådahl <jadahl@redhat.com>
>> Cc: Sebastian Wick <sebastian.wick@redhat.com>
>> Cc: Shashank Sharma <shashank.sharma@amd.com>
>> Cc: Alexander Goins <agoins@nvidia.com>
>> Cc: Joshua Ashton <joshua@froggi.es>
>> Cc: Michel Dänzer <mdaenzer@redhat.com>
>> Cc: Aleix Pol <aleixpol@kde.org>
>> Cc: Xaver Hugl <xaver.hugl@gmail.com>
>> Cc: Victoria Brekenfeld <victoria@system76.com>
>> Cc: Sima <daniel@ffwll.ch>
>> Cc: Uma Shankar <uma.shankar@intel.com>
>> Cc: Naseer Ahmed <quic_naseer@quicinc.com>
>> Cc: Christopher Braga <quic_cbraga@quicinc.com>
>> Cc: Abhinav Kumar <quic_abhinavk@quicinc.com>
>> Cc: Arthur Grillo <arthurgrillo@riseup.net>
>> Cc: Hector Martin <marcan@marcan.st>
>> Cc: Liviu Dudau <Liviu.Dudau@arm.com>
>> Cc: Sasha McIntosh <sashamcintosh@google.com>
>> Cc: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
>>
>> Alex Hung (13):
>> drm/colorop: define a new macro for_each_new_colorop_in_state
>> drm/amd/display: Skip color pipeline initialization for cursor plane
>> drm/amd/display: Add support for sRGB EOTF in DEGAM block
>> drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block
>> drm/amd/display: Add support for sRGB EOTF in BLND block
>> drm/colorop: Add 1D Curve Custom LUT type
>> drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT
>> drm/amd/display: add 3x4 matrix colorop
>> drm/colorop: Add mutliplier type
>> drm/amd/display: add multiplier colorop
>> drm/amd/display: Swap matrix and multiplier
>> drm/colorop: Add 3D LUT supports to color pipeline
>> drm/amd/display: add 3D LUT colorop
>>
>> Harry Wentland (31):
>> drm: Add helper for conversion from signed-magnitude
>> drm/vkms: Round fixp2int conversion in lerp_u16
>> drm/vkms: Add kunit tests for VKMS LUT handling
>> drm/doc/rfc: Describe why prescriptive color pipeline is needed
>> drm/colorop: Introduce new drm_colorop mode object
>> drm/colorop: Add TYPE property
>> drm/colorop: Add 1D Curve subtype
>> Documentation/gpu: document drm_colorop
>> 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: Use s32 for internal color pipeline precision
>> drm/vkms: add 3x4 matrix in color pipeline
>> drm/tests: Add a few tests around drm_fixed.h
>> drm/vkms: Add tests for CTM handling
>> drm/colorop: pass plane_color_pipeline client cap to atomic check
>> drm/amd/display: Ignore deprecated props when plane_color_pipeline set
>> drm/amd/display: Add bypass COLOR PIPELINE
>> drm/colorop: Add PQ 125 EOTF and its inverse
>> drm/amd/display: Enable support for PQ 125 EOTF and Inverse
>> drm/colorop: add BT2020/BT709 OETF and Inverse OETF
>> drm/amd/display: Add support for BT.709 and BT.2020 TFs
>> drm/colorop: Define LUT_1D interpolation
>> drm/colorop: allow non-bypass colorops
>> drm/amd/display: Add AMD color pipeline doc
>>
>> Documentation/gpu/drm-kms.rst | 15 +
>> Documentation/gpu/rfc/color_pipeline.rst | 376 ++++++++
>> Documentation/gpu/rfc/index.rst | 3 +
>> drivers/gpu/drm/Makefile | 1 +
>> .../gpu/drm/amd/display/amdgpu_dm/Makefile | 3 +-
>> .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +
>> .../amd/display/amdgpu_dm/amdgpu_dm_color.c | 581 ++++++++++++-
>> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.c | 196 +++++
>> .../amd/display/amdgpu_dm/amdgpu_dm_colorop.h | 36 +
>> .../amd/display/amdgpu_dm/amdgpu_dm_plane.c | 32 +
>> drivers/gpu/drm/drm_atomic.c | 184 +++-
>> drivers/gpu/drm/drm_atomic_helper.c | 12 +
>> drivers/gpu/drm/drm_atomic_state_helper.c | 5 +
>> drivers/gpu/drm/drm_atomic_uapi.c | 175 ++++
>> drivers/gpu/drm/drm_colorop.c | 562 ++++++++++++
>> 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 | 5 +
>> 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 | 102 +++
>> drivers/gpu/drm/vkms/vkms_composer.c | 115 ++-
>> 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 | 130 +++
>> include/drm/drm_atomic_uapi.h | 3 +
>> include/drm/drm_colorop.h | 427 ++++++++++
>> include/drm/drm_file.h | 7 +
>> include/drm/drm_fixed.h | 18 +
>> include/drm/drm_mode_config.h | 18 +
>> include/drm/drm_plane.h | 19 +
>> include/uapi/drm/amdgpu_drm.h | 9 -
>> include/uapi/drm/drm.h | 16 +
>> include/uapi/drm/drm_mode.h | 156 ++++
>> 40 files changed, 4595 insertions(+), 34 deletions(-)
>> create mode 100644 Documentation/gpu/rfc/color_pipeline.rst
>> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.c
>> create mode 100644 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_colorop.h
>> create mode 100644 drivers/gpu/drm/drm_colorop.c
>> create mode 100644 drivers/gpu/drm/tests/drm_fixp_test.c
>> create mode 100644 drivers/gpu/drm/vkms/tests/.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
>>
>> --
>> 2.46.0
>>
^ permalink raw reply [flat|nested] 69+ messages in thread
end of thread, other threads:[~2024-10-03 20:09 UTC | newest]
Thread overview: 69+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-08-19 20:56 [PATCH v5 00/44] Color Pipeline API w/ VKMS Harry Wentland
2024-08-19 20:56 ` [PATCH v5 01/44] drm: Add helper for conversion from signed-magnitude Harry Wentland
2024-08-19 20:56 ` [PATCH v5 02/44] drm/vkms: Round fixp2int conversion in lerp_u16 Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 03/44] drm/vkms: Add kunit tests for VKMS LUT handling Harry Wentland
2024-08-21 1:05 ` Jeff Johnson
2024-08-27 17:49 ` Louis Chauvet
2024-08-28 13:36 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 04/44] drm/doc/rfc: Describe why prescriptive color pipeline is needed Harry Wentland
2024-08-19 20:56 ` [PATCH v5 05/44] drm/colorop: Introduce new drm_colorop mode object Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 06/44] drm/colorop: Add TYPE property Harry Wentland
2024-08-19 20:56 ` [PATCH v5 07/44] drm/colorop: Add 1D Curve subtype Harry Wentland
2024-08-19 20:56 ` [PATCH v5 08/44] Documentation/gpu: document drm_colorop Harry Wentland
2024-08-19 20:56 ` [PATCH v5 09/44] drm/colorop: Add BYPASS property Harry Wentland
2024-08-19 20:56 ` [PATCH v5 10/44] drm/colorop: Add NEXT property Harry Wentland
2024-08-19 20:56 ` [PATCH v5 11/44] drm/colorop: Add atomic state print for drm_colorop Harry Wentland
2024-08-19 20:56 ` [PATCH v5 12/44] drm/plane: Add COLOR PIPELINE property Harry Wentland
2024-08-19 20:56 ` [PATCH v5 13/44] drm/colorop: Add NEXT to colorop state print Harry Wentland
2024-08-19 20:56 ` [PATCH v5 14/44] drm/vkms: Add enumerated 1D curve colorop Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 15/44] drm/vkms: Add kunit tests for linear and sRGB LUTs Harry Wentland
2024-08-19 20:56 ` [PATCH v5 16/44] drm/colorop: Introduce DRM_CLIENT_CAP_PLANE_COLOR_PIPELINE Harry Wentland
2024-08-19 20:56 ` [PATCH v5 17/44] drm/colorop: Add 3x4 CTM type Harry Wentland
2024-08-19 20:56 ` [PATCH v5 18/44] drm/vkms: Use s32 for internal color pipeline precision Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:56 ` [PATCH v5 19/44] drm/vkms: add 3x4 matrix in color pipeline Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-09-06 17:08 ` Harry Wentland
2024-08-19 20:56 ` [PATCH v5 20/44] drm/tests: Add a few tests around drm_fixed.h Harry Wentland
2024-08-21 1:07 ` Jeff Johnson
2024-08-21 4:00 ` kernel test robot
2024-08-22 14:55 ` kernel test robot
2024-08-19 20:56 ` [PATCH v5 21/44] drm/vkms: Add tests for CTM handling Harry Wentland
2024-08-19 20:56 ` [PATCH v5 22/44] drm/colorop: pass plane_color_pipeline client cap to atomic check Harry Wentland
2024-08-19 20:56 ` [PATCH v5 23/44] drm/colorop: define a new macro for_each_new_colorop_in_state Harry Wentland
2024-08-19 20:56 ` [PATCH v5 24/44] drm/amd/display: Ignore deprecated props when plane_color_pipeline set Harry Wentland
2024-08-19 20:56 ` [PATCH v5 25/44] drm/amd/display: Add bypass COLOR PIPELINE Harry Wentland
2024-08-19 20:56 ` [PATCH v5 26/44] drm/amd/display: Skip color pipeline initialization for cursor plane Harry Wentland
2024-08-19 20:56 ` [PATCH v5 27/44] drm/amd/display: Add support for sRGB EOTF in DEGAM block Harry Wentland
2024-08-19 20:56 ` [PATCH v5 28/44] drm/amd/display: Add support for sRGB Inverse EOTF in SHAPER block Harry Wentland
2024-08-21 4:52 ` kernel test robot
2024-08-19 20:56 ` [PATCH v5 29/44] drm/amd/display: Add support for sRGB EOTF in BLND block Harry Wentland
2024-08-19 20:56 ` [PATCH v5 30/44] drm/colorop: Add PQ 125 EOTF and its inverse Harry Wentland
2024-08-19 20:56 ` [PATCH v5 31/44] drm/amd/display: Enable support for PQ 125 EOTF and Inverse Harry Wentland
2024-08-19 20:56 ` [PATCH v5 32/44] drm/colorop: add BT2020/BT709 OETF and Inverse OETF Harry Wentland
2024-08-19 20:57 ` [PATCH v5 33/44] drm/amd/display: Add support for BT.709 and BT.2020 TFs Harry Wentland
2024-08-19 20:57 ` [PATCH v5 34/44] drm/colorop: Add 1D Curve Custom LUT type Harry Wentland
2024-08-19 20:57 ` [PATCH v5 35/44] drm/amd/display: add shaper and blend colorops for 1D Curve Custom LUT Harry Wentland
2024-08-19 20:57 ` [PATCH v5 36/44] drm/amd/display: add 3x4 matrix colorop Harry Wentland
2024-08-21 3:19 ` kernel test robot
2024-08-21 9:22 ` kernel test robot
2024-08-19 20:57 ` [PATCH v5 37/44] drm/colorop: Add mutliplier type Harry Wentland
2024-08-19 20:57 ` [PATCH v5 38/44] drm/amd/display: add multiplier colorop Harry Wentland
2024-08-19 20:57 ` [PATCH v5 39/44] drm/amd/display: Swap matrix and multiplier Harry Wentland
2024-08-19 20:57 ` [PATCH v5 40/44] drm/colorop: Define LUT_1D interpolation Harry Wentland
2024-08-19 20:57 ` [PATCH v5 41/44] drm/colorop: allow non-bypass colorops Harry Wentland
2024-08-27 17:49 ` Louis Chauvet
2024-08-19 20:57 ` [PATCH v5 42/44] drm/colorop: Add 3D LUT supports to color pipeline Harry Wentland
2024-08-19 20:57 ` [PATCH v5 43/44] drm/amd/display: add 3D LUT colorop Harry Wentland
2024-08-19 20:57 ` [PATCH v5 44/44] drm/amd/display: Add AMD color pipeline doc Harry Wentland
2024-08-27 17:49 ` [PATCH v5 00/44] Color Pipeline API w/ VKMS Louis Chauvet
2024-10-03 20:09 ` Harry Wentland
2024-08-29 14:55 ` Xaver Hugl
2024-09-09 18:40 ` Harry Wentland
2024-09-10 20:37 ` Alex Goins
2024-10-03 20:00 ` Harry Wentland
2024-09-10 19:24 ` [PATCH] drm/colorop: get DATA blob ref at duplicate_state Harry Wentland
2024-09-13 0:17 ` Xaver Hugl
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox