[SYCL][Bindless][UR] Add support for sRGB image formats by juanchuletas · Pull Request #22257 · intel/llvm

juanchuletas · 2026-06-09T01:41:09Z

Summary

This PR adds native sRGB decode support for bindless images through the full DPC++ stack: SYCL API, Unified Runtime, and the Level Zero adapter.

Description

The hardware sampler on Intel Arc GPUs supports native sRGB decode via ze_srgb_ext_desc_t, confirmed by direct Level Zero testing against driver 26.18.38308.1. The driver correctly applies the IEC 61966-2-1 piecewise formula on fetch, returning linear values to the kernel. This capability was not exposed anywhere in the DPC++ or UR stack.

The SYCL bindless image API had no way to request sRGB color space decoding. The image_descriptor only accepted num_channels and channel_type, with the channel order always derived internally. This made it impossible to distinguish between a linear RGBA texture and an sRGB-encoded RGBA texture, forcing applications to manually decode sRGB values on the CPU before upload at a significant memory cost.

Proposed Changes

`sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp`

Added a new image_color_space enum (linear / srgb) and a color_space field on image_descriptor, gated behind __INTEL_PREVIEW_BREAKING_CHANGES since it changes the descriptor's constructor signatures. Under that macro, all three range-based constructors (1D/2D/3D) gain a color_space parameter defaulting to image_color_space::linear, preserving existing call sites when the flag is off. Added a validation check in verify() (also macro-gated) requiring num_channels == 4 and channel_type == unorm_int8 whenever color_space == srgb, matching the only configuration the Level Zero driver accepts for sRGB images.

`sycl/source/detail/bindless_images.cpp`

Modified populate_ur_structs so that, under __INTEL_PREVIEW_BREAKING_CHANGES, urFormat.channelOrder is set to UR_IMAGE_CHANNEL_ORDER_SRGBA when desc.color_space == image_color_space::srgb, and falls back to the default channel order derived from num_channels otherwise. Without the preview flag, behavior is unchanged.

Proposed Usage

sycl::ext::oneapi::experimental::image_descriptor desc(
    sycl::range<2>(width, height),
    4,
    sycl::image_channel_type::unorm_int8,
    sycl::ext::oneapi::experimental::image_color_space::srgb
);

sycl::ext::oneapi::experimental::image_mem imgMem(desc, q);

Note: image_color_space and the color_space constructor parameter are only available when building with __INTEL_PREVIEW_BREAKING_CHANGES defined.

juanchuletas · 2026-06-09T07:20:37Z

Update,

I tested the changes using:

srgb_vs_rgb.cpp

and I got:

Device: Intel(R) Arc(TM) Graphics

Input raw pixel value:            186 (0xba)
Input normalized (unorm_int8):    0.729412
Expected linear :  0.491021

RGBA  result (no decode):  R=0.729412 G=0.729412 B=0.729412 A=1
sRGBA result (hardware decoded): R=0.490745 G=0.490745 B=0.490745 A=1

Difference in R channel:   0.238666

PASS: RGBA returns raw unorm, sRGBA returns decoded linear

kswiecicki

L0 adapter side LGTM, 2 nitpicks.

0x12CC

Thanks, @juanchuletas. Please add relevant tests.

0x12CC · 2026-06-09T18:03:44Z

These changes need corresponding changes in sycl/doc/extensions/experimental/sycl_ext_oneapi_bindless_images.asciidoc to ensure the implementation and specification remain consistent.

juanchuletas · 2026-06-10T06:30:08Z

@0x12CC @kswiecicki

I was thinking about the current implementation and some scenarios came out. The current implementation exposes channel_order as an std::optional<image_channel_order> on image_descriptor.

Nothing stops the user from doing the following:

sycl::ext::oneapi::experimental::image_descriptor desc(
    sycl::range<2>(width, height),
    4,
    sycl::image_channel_type::unorm_int8
);
desc.channel_order = sycl::image_channel_order::rgx;

We can prohibit that kind of usage from the verify() method

if (channel_order.has_value()) {
  if (channel_order.value() != image_channel_order::ext_oneapi_srgba) {
    throw sycl::exception(
        sycl::errc::invalid,
        "channel_order only supports ext_oneapi_srgba");
  }
  if (num_channels != 4) {
    throw sycl::exception(
        sycl::errc::invalid,
        "ext_oneapi_srgba channel order requires num_channels == 4");
  }
  if (channel_type != image_channel_type::unorm_int8) {
    throw sycl::exception(
        sycl::errc::invalid,
        "ext_oneapi_srgba channel order requires unorm_int8 channel type");
  }
}

but it feels weird to expose desc.channel_order and only allowing srgb. Other approach is to Make channel_order fully general: validate channel count consistency in verify() for any value set.

dyniols

This PR is ABI breaking because of changes in image_descriptor struct layout.

I wonder if we could re-work image_descriptor back to support channel_order since it was replaced with num_channels due to lack of support in CUDA.
Here is PR when this change was introduced: #13745
I believe that Level Zero, OpenCL backend should support channel order.

Please add tests exercising the new sRGB path with sampled, unsampled and using pitched memory allocation, through pitched_alloc_device.

juanchuletas · 2026-06-11T04:11:24Z

This PR is ABI breaking because of changes in image_descriptor struct layout.

I wonder if we could re-work image_descriptor back to support channel_order since it was replaced with num_channels due to lack of support in CUDA. Here is PR when this change was introduced: #13745 I believe that Level Zero, OpenCL backend should support channel order.

Please add tests exercising the new sRGB path with sampled, unsampled and using pitched memory allocation, through pitched_alloc_device.

Given that restoring channel_order as a primary field would also be an ABI break, and that CUDA does not support channel order natively, what is the preferred approach?

replacement of num_channels with channel_order
stay focused on sRGB hardware decode only

I am open to work on any of the approaches

dyniols · 2026-06-11T10:07:17Z

Given that restoring channel_order as a primary field would also be an ABI break.

I don't think the ABI break is avoidable here. Naturally, a change like this should land with the next major release. In the meantime, we can leverage __INTEL_PREVIEW_BREAKING_CHANGES macro to merge ABI-breaking change without delay and then remove it when possible.

what is the preferred approach?

replacement of num_channels with channel_order

stay focused on sRGB hardware decode only

I am open to work on any of the approaches

I lean towards replacement of num_channels with channel_order and using with image_channel_order enum class.
However I would ask @0x12CC on his perspective on such a change.

juanchuletas · 2026-06-15T01:12:21Z

I've been thinking about another approach. What do you think of this @dyniols , @0x12CC , @kswiecicki ?

Instead of reaching for image_channel_order at all. Why don't treat the sRGB as what it actually is: a property of how the sampler interprets the data:

enum class image_color_space : uint32_t {
    linear = 0,
    srgb   = 1,
};

struct image_descriptor {
    image_type        type          = image_type::standard;
    uint32_t          num_channels  = 0;
    image_channel_type channel_type;
    image_color_space color_space   = image_color_space::linear;  // the new field
   //Other fields
};

and usage like:

image_descriptor desc;
desc.num_channels = 4;
desc.channel_type  = image_channel_type::unorm_int8;
desc.color_space   = image_color_space::srgb;

Actually cuda does something similar, to enable the srgb

dyniols · 2026-06-19T14:40:30Z

Instead of reaching for image_channel_order at all. Why don't treat the sRGB as what it actually is: a property of how the sampler interprets the data:

enum class image_color_space : uint32_t {
    linear = 0,
    srgb   = 1,
};

struct image_descriptor {
    image_type        type          = image_type::standard;
    uint32_t          num_channels  = 0;
    image_channel_type channel_type;
    image_color_space color_space   = image_color_space::linear;  // the new field
   //Other fields
};

and usage like:

image_descriptor desc;
desc.num_channels = 4;
desc.channel_type  = image_channel_type::unorm_int8;
desc.color_space   = image_color_space::srgb;

Actually cuda does something similar, to enable the srgb

Sorry for the late reply. I think this direction makes sense, and using a new image_color_space field is probably a cleaner way to describe how the color data stored in an image should be interpreted.

juanchuletas · 2026-06-20T20:04:12Z

Instead of reaching for image_channel_order at all. Why don't treat the sRGB as what it actually is: a property of how the sampler interprets the data:
enum class image_color_space : uint32_t {
    linear = 0,
    srgb   = 1,
};

struct image_descriptor {
    image_type        type          = image_type::standard;
    uint32_t          num_channels  = 0;
    image_channel_type channel_type;
    image_color_space color_space   = image_color_space::linear;  // the new field
   //Other fields
};
and usage like:
image_descriptor desc;
desc.num_channels = 4;
desc.channel_type  = image_channel_type::unorm_int8;
desc.color_space   = image_color_space::srgb;
Actually cuda does something similar, to enable the srgb
Sorry for the late reply. I think this direction makes sense, and using a new image_color_space field is probably a cleaner way to describe how the color data stored in an image should be interpreted.

Great! Working on it!

juanchuletas · 2026-06-23T01:29:22Z

Hi! @dyniols, @kswiecicki, @0x12CC

On my last commit I moved the check to pupulate_ur_structs because adding the sRGB check inside verify() makes no sense as verify() is called inside the constructor and the current usage is like:

 // sRGBA image: hardware applies IEC 61966-2-1 decode on fetch
    syclexp::image_descriptor srgbaDesc(
        sycl::range<2>(1, 1),
        4,
        sycl::image_channel_type::unorm_int8
    );
    srgbaDesc.color_space = sycl::ext::oneapi::experimental::image_color_space::srgb;

sRGB color space is set after the object construction. I thought in adding the image_color_space field to the constructor, but that will lead into a very big change.

Totally apart, working on that I noticed that the user can set an image descriptor like this:

image_descriptor desc(range<2>{w, h}, 4, image_channel_type::unorm_int8);
desc.num_levels = 5;  // bypasses verify()
desc.array_size = 10; // bypasses verify()

And verify() never runs again.

What should we do here:

-Should we make an effort and add color_space to the constructor?
-Should we mark those fields as private ?
-Should we verify in other place?

dyniols · 2026-06-30T16:05:21Z

Hi @juanchuletas

I think adding image_color_space field to constructor seems the most reasonable but of course it will be API and ABI breaking change. So it will need to be marked as such.
Your thoughts @0x12CC?

Regarding verify() it runs in more places than just the constructor. It's also called at the start of the functions that are using the image descriptor (alloc_image_mem, create_image, map_external_image_memory, and the copy functions in handler.cpp) all call verify() right before passing it to unified runtime.

…er __INTEL_PREVIEW_BREAKING_CHANGES

juanchuletas · 2026-07-03T09:03:53Z

@dyniols

Hi! I made the changes, if that looks correct for you I can start adding the proper tests.

dyniols · 2026-07-06T12:25:58Z

if that looks correct for you I can start adding the proper tests.

Sure, I’ll review the current implementation changes.

There will be another requirement for this PR to update sycl/doc/extensions/experimental/sycl_ext_oneapi_bindless_images.asciidoc so the specification and implementation stay consistent. However, you can wait to do that until the implementation is accepted.

Copilot

Pull request overview

Note

Copilot couldn't run its full agentic review because no GitHub Actions runner was available. Make sure your repository has a runner available to run Copilot's review, or add a copilot-setup-steps.yml file specifying one with the runs-on attribute. See the docs for more details.

Adds end-to-end support for native sRGB decode for bindless images, plumbing a new sRGB selection from SYCL down through Unified Runtime to Level Zero via ze_srgb_ext_desc_t.

Changes:

Extends SYCL bindless image descriptors with a preview-only image_color_space and validation for sRGB constraints.
Propagates sRGB selection into UR image format (UR_IMAGE_CHANNEL_ORDER_SRGBA) under the preview macro.
Updates Level Zero adapter to map SRGBA layouts/swizzles and append ze_srgb_ext_desc_t into the pNext chain for bindless creation/allocation.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 6 comments.

File	Description
`unified-runtime/source/adapters/level_zero/image_common.cpp`	Adds `UR_IMAGE_CHANNEL_ORDER_SRGBA` mapping and chains `ze_srgb_ext_desc_t` into Level Zero image creation/allocation.
`sycl/source/detail/bindless_images.cpp`	Propagates SYCL descriptor sRGB intent into UR image format under preview macro.
`sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp`	Introduces preview-only `image_color_space` and adds validation for sRGB-required format constraints.

Comments suppressed due to low confidence (1)

unified-runtime/source/adapters/level_zero/image_common.cpp:623

UR_IMAGE_CHANNEL_ORDER_SRGBA is currently treated like generic 4-channel formats and will map to 16/32-bit layouts depending on channelType. However, the PR description indicates Level Zero sRGB decode support is limited (e.g., unorm_int8 / 8-bit). The UR adapter should enforce this at the mapping layer (return UR_RESULT_ERROR_UNSUPPORTED_IMAGE_FORMAT when SRGBA is used with non-8-bit and/or non-UNORM formats) so non-SYCL callers can’t produce a format that will be rejected or misinterpreted by the driver.

  case UR_IMAGE_CHANNEL_ORDER_ABGR:
  case UR_IMAGE_CHANNEL_ORDER_SRGBA: {
    switch (ZeImageFormatTypeSize) {
    case 8:
      ZeImageFormatLayout = ZE_IMAGE_FORMAT_LAYOUT_8_8_8_8;
      break;
    case 16:
      ZeImageFormatLayout = ZE_IMAGE_FORMAT_LAYOUT_16_16_16_16;
      break;
    case 32:
      ZeImageFormatLayout = ZE_IMAGE_FORMAT_LAYOUT_32_32_32_32;
      break;

Copilot

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 5 comments.

iclsrc

This PR adds image_color_space::srgb support to bindless images, gating the new color_space field and constructors behind __INTEL_PREVIEW_BREAKING_CHANGES and wiring UR_IMAGE_CHANNEL_ORDER_SRGBA through to the Level Zero ze_srgb_ext_desc_t extension. The core UR and SYCL runtime plumbing looks correct. Three issues are noted below: one correctness bug in get_mip_level_desc, one test-coverage gap (only R channel validated), and a suggestion for a device-independent unit test.

… sRGB constraints

juanchuletas · 2026-07-15T19:07:10Z

@iclsrc I saw several tests are failing, is there any else I can do ?

dyniols · 2026-07-16T09:59:44Z

@juanchuletas I will be gone on vacations and I will be back after 26th July FYI. You might need toping other bindless image reviewers.

dyniols · 2026-07-16T10:00:45Z

@iclsrc I saw several tests are failing, is there any else I can do ?

Can you merge the latest intel/llvm into your PR branch? I will review but I would prefer to re-run CI with newest intel/llvm changes.

…pport

juanchuletas · 2026-07-16T20:02:40Z

@iclsrc I saw several tests are failing, is there any else I can do ?

Can you merge the latest intel/llvm into your PR branch? I will review but I would prefer to re-run CI with newest intel/llvm changes.

Merge done! :)

juanchuletas · 2026-07-23T06:01:47Z

Hi @iclsrc , is anything else I can do here?

dyniols · 2026-07-27T13:11:40Z

Hi @iclsrc , is anything else I can do here?

Oh, this is a Claude bot, not a real person. I will take a look. I see Pre Commit is still failing.

Copilot

Pull request overview

Copilot reviewed 8 out of 8 changed files in this pull request and generated 3 comments.

Comments suppressed due to low confidence (4)

sycl/test-e2e/bindless_images/srgb/fetch_unsampled_srgb.cpp:20

Missing standard headers: this test uses uint8_t and std::vector but does not include <cstdint> / <vector>, which can break compilation depending on transitive includes.

#include <cmath>
#include <iostream>
#include <sycl/detail/core.hpp>
#include <sycl/ext/oneapi/bindless_images.hpp>

sycl/test-e2e/bindless_images/srgb/fetch_unsampled_srgb.cpp:96

Bindless image coordinates are (x, y) for a 2D descriptor created as {width, height}; the kernel currently passes (y, x) to fetch_image, which becomes out-of-bounds if width != height.

              acc[sycl::id<2>(dim0, dim1)] = syclexp::fetch_image<sycl::float4>(
                  linearImg, sycl::int2(dim0, dim1));

sycl/test-e2e/bindless_images/srgb/fetch_sampled_srgb.cpp:111

For a 2D descriptor created as {width, height}, normalized sample coordinates are (x/width, y/height). The test currently passes (y/height, x/width), which is inconsistent with the bindless image coordinate conventions (see coordinate_order_2D.cpp).

                           float fdim0 = (float(dim0) + 0.5f) / float(height);
                           float fdim1 = (float(dim1) + 0.5f) / float(width);
                           acc[sycl::id<2>(dim0, dim1)] =
                               syclexp::sample_image<sycl::float4>(
                                   linearImg, sycl::float2(fdim0, fdim1));

sycl/test-e2e/bindless_images/srgb/fetch_pitched_srgb.cpp:121

For a 2D descriptor created as {width, height}, normalized sample coordinates are (x/width, y/height). The test currently passes (y/height, x/width), which is inconsistent with the bindless image coordinate conventions (see coordinate_order_2D.cpp).

                           float fdim0 = (float(dim0) + 0.5f) / float(height);
                           float fdim1 = (float(dim1) + 0.5f) / float(width);
                           acc[sycl::id<2>(dim0, dim1)] =
                               syclexp::sample_image<sycl::float4>(
                                   linearImg, sycl::float2(fdim0, fdim1));

…ge fetch tests

juanchuletas · 2026-07-30T23:33:18Z

@dyniols

Changes suggested by copilot: Done

dyniols · 2026-08-07T12:43:47Z

@juanchuletas Can you resolve conflicts? Thanks and please ping me I will restart CI.

Copilot

Pull request overview

Copilot reviewed 8 out of 8 changed files in this pull request and generated 1 comment.

Suppressed comments (3)

unified-runtime/source/adapters/level_zero/image_common.cpp:623

Adding SRGBA to this shared translator also makes non-bindless urMemImageCreate accept the format. Both memory.cpp:1514-1523 and v2/memory.cpp:549-557 call ur2zeImageDesc and then zeImageCreate without chaining ze_srgb_ext_desc_t, so those APIs now return a linear RGBA image for an SRGBA request instead of decoding sRGB (or rejecting it). Keep SRGBA acceptance bindless-only, or attach the sRGB descriptor at every image-creation caller.

  case UR_IMAGE_CHANNEL_ORDER_SRGBA: {

sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp:56

This adds a public enum, descriptor member, constructor parameters, validation rules, and Level Zero-only behavior, but sycl_ext_oneapi_bindless_images.asciidoc still specifies the old image_descriptor API, while sycl_ext_oneapi_srgb.asciidoc still states that sRGB is OpenCL-only. Update the extension specifications so users can determine the API contract, preview gating, supported formats, and backend limitations.

/// image color space enum
enum class image_color_space : uint32_t {
  linear = 0,
  srgb = 1,
};

sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp:56

The PR description says image_color_space itself is available only with __INTEL_PREVIEW_BREAKING_CHANGES, but the enum is currently exposed in non-preview builds. Gate this declaration with the same macro as the field and constructors so the documented preview boundary matches the API.

This issue also appears on line 52 of the same file.

/// image color space enum
enum class image_color_space : uint32_t {
  linear = 0,
  srgb = 1,
};

…sampled images

juanchuletas · 2026-08-07T19:24:20Z

@dyniols Hi Daniel, I fixed the issue that copilot found.

dyniols · 2026-08-09T12:51:52Z

@juanchuletas Can you merge the latest intel/llvm into your PR branch and resolve conflicts in unified-runtime/source/adapters/level_zero/image_common.cpp?

…pport

juanchuletas · 2026-08-09T20:49:22Z

@juanchuletas Can you merge the latest intel/llvm into your PR branch and resolve conflicts in unified-runtime/source/adapters/level_zero/image_common.cpp?

Done! :)

dyniols

Okay, changes look good for me except few comments.
Let's wait for CI and see if all tests are passing.

dyniols · 2026-07-27T10:08:48Z

+                                syclexp::image_color_space::linear));
+}
+
+#endif // __INTEL_PREVIEW_BREAKING_CHANGES


Missing trailing newline.

dyniols · 2026-07-27T10:08:55Z

+
+  std::cout << "Test passed!" << std::endl;
+  return 0;
+}


Missing trailing newline.

dyniols · 2026-07-27T10:09:03Z

+
+  std::cout << "Test passed!" << std::endl;
+  return 0;
+}


Missing trailing newline.

dyniols · 2026-07-27T10:09:10Z

+
+  std::cout << "Test passed!" << std::endl;
+  return 0;
+}


Missing trailing newline.

dyniols · 2026-08-07T13:11:20Z

+// REQUIRES: aspect-ext_oneapi_bindless_images
+// REQUIRES: preview-breaking-changes-supported
+
+// UNSUPPORTED: cuda, hip


It's supported only in Level Zero. I would use REQUIRE: level_zero instead.

On second thought I think we should add new aspect that will help to check for support of SRGB channel order in bindless image extension. However it can be done as follow up in other PR.

dyniols · 2026-08-07T13:25:14Z

+            });
+      });
+
+      q.wait_and_throw();


This wait can be omitted since SYCL buffer's are emplaced within braces and destructor should handle synchronization and copying data back from device to host.

dyniols · 2026-08-07T13:27:34Z

+            });
+      });
+
+      q.wait_and_throw();


dyniols · 2026-08-07T13:27:45Z

+            });
+      });
+
+      q.wait_and_throw();


juanchuletas · 2026-08-10T18:53:58Z

@dyniols

I saw several tests failing on CI, so I checked and found that the test files hadn't been updated after I added image_color_space as the 4th argument to the image_descriptor constructor. They were still passing image_type in that slot. So, I edited the tests to pass image_color_space::linear before image_type, matching the constructor.

I also fixed the comments you left: replaced UNSUPPORTED: cuda, hip with REQUIRES: level_zero in the sRGB tests, since sRGB decode is only supported on Level Zero for now and I added the trailing newline.

…ructor Update image_descriptor calls to pass image_color_space before image_type, and require level_zero for sRGB tests instead of excluding cuda and hip.

juanchuletas · 2026-08-10T19:52:49Z

@dyniols

Some tests are still failing :( but I think the reason is teh failing ones have no the "preview-mode", other failing test is something about AMDGPU_TARGETS was not set, and system GPU detection was unsuccsesful

[SYCL][Bindless][UR] Add support for sRGB image formats

2900c7a

juanchuletas requested review from a team as code owners June 9, 2026 01:41

juanchuletas requested a review from 0x12CC June 9, 2026 01:41

[SYCL][Bindless][UR] applying clang format

15fc9ae

juanchuletas force-pushed the sycl-bindless-srgb-support branch from 8e6a729 to 15fc9ae Compare June 9, 2026 07:00

kswiecicki approved these changes Jun 9, 2026

View reviewed changes

Comment thread unified-runtime/source/adapters/level_zero/image_common.cpp

Comment thread unified-runtime/source/adapters/level_zero/image_common.cpp Outdated

0x12CC requested changes Jun 9, 2026

View reviewed changes

[SYCL][Bindless][UR] Address review feedback

26ae9ea

juanchuletas force-pushed the sycl-bindless-srgb-support branch from 7d07faf to 26ae9ea Compare June 10, 2026 03:46

dyniols reviewed Jun 10, 2026

View reviewed changes

Comment thread sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp Outdated

juanchuletas added 3 commits June 21, 2026 18:43

[SYCL][Bindless][UR] adding the image_color_space struct

69d0137

[SYCL][Bindless][UR] clang format

3d89061

[SYCL][Bindless][UR] moving srgb check to pupulate_ur_structs

b284a0d

juanchuletas force-pushed the sycl-bindless-srgb-support branch from d16f9d1 to b284a0d Compare June 23, 2026 01:09

[SYCL][Bindless][UR] Add image_color_space field and sRGB support und…

d978d66

…er __INTEL_PREVIEW_BREAKING_CHANGES

dyniols requested a review from Copilot July 6, 2026 12:19

Copilot AI reviewed Jul 6, 2026

View reviewed changes

Copilot started reviewing on behalf of dyniols July 6, 2026 14:19 View session

Copilot AI reviewed Jul 14, 2026

View reviewed changes

iclsrc reviewed Jul 14, 2026

View reviewed changes

Comment thread sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp

Comment thread sycl/test-e2e/bindless_images/srgb/fetch_unsampled_srgb.cpp

Comment thread sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp

iclsrc removed the claude-review label Jul 14, 2026

[SYCL][Bindless][UR] Solve comments and added unit tests for verify()…

30607fe

… sRGB constraints

Merge remote-tracking branch 'origin/sycl' into sycl-bindless-srgb-su…

4a5c7aa

…pport

dyniols requested a review from Copilot July 27, 2026 13:11

Copilot started reviewing on behalf of dyniols July 27, 2026 13:12 View session

Copilot AI reviewed Jul 27, 2026

View reviewed changes

Comment thread sycl/test-e2e/bindless_images/srgb/fetch_unsampled_srgb.cpp Outdated

Comment thread sycl/test-e2e/bindless_images/srgb/fetch_sampled_srgb.cpp Outdated

Comment thread sycl/test-e2e/bindless_images/srgb/fetch_pitched_srgb.cpp Outdated

[SYCL][Bindless][UR] Fix swapped x/y coordinates in sRGB bindless ima…

f3451a0

…ge fetch tests

johguenther mentioned this pull request Aug 5, 2026

Adding SYCL bindless image support RenderKit/ospray#612

Open

15 tasks

dyniols requested a balanced review from Copilot August 7, 2026 12:44

Copilot started reviewing on behalf of dyniols August 7, 2026 12:44 View session

Copilot AI reviewed Aug 7, 2026

View reviewed changes

Comment thread unified-runtime/source/adapters/level_zero/common/image_common.cpp

[SYCL][Bindless][UR] Chain sRGB descriptor before pitched desc for un…

e702dfb

…sampled images

Merge remote-tracking branch 'origin/sycl' into sycl-bindless-srgb-su…

88906eb

…pport

dyniols reviewed Aug 10, 2026

View reviewed changes

[SYCL][Bindless][UR] Fix e2e tests for updated image_descriptor const…

60859d5

…ructor Update image_descriptor calls to pass image_color_space before image_type, and require level_zero for sRGB tests instead of excluding cuda and hip.

Uh oh!

Conversation

juanchuletas commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Description

Proposed Changes

sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp

sycl/source/detail/bindless_images.cpp

Proposed Usage

Uh oh!

juanchuletas commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Update,

Uh oh!

kswiecicki left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

0x12CC left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

juanchuletas commented Jun 10, 2026

Uh oh!

dyniols left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

juanchuletas commented Jun 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dyniols commented Jun 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

juanchuletas commented Jun 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dyniols commented Jun 19, 2026

Uh oh!

juanchuletas commented Jun 20, 2026

Uh oh!

juanchuletas commented Jun 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dyniols commented Jun 30, 2026

Uh oh!

juanchuletas commented Jul 3, 2026

Uh oh!

dyniols commented Jul 6, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

iclsrc left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

juanchuletas commented Jul 15, 2026

juanchuletas commented Jun 9, 2026 •

edited

Loading

`sycl/include/sycl/ext/oneapi/bindless_images_descriptor.hpp`

`sycl/source/detail/bindless_images.cpp`

juanchuletas commented Jun 9, 2026 •

edited

Loading

juanchuletas commented Jun 11, 2026 •

edited

Loading

dyniols commented Jun 11, 2026 •

edited

Loading

juanchuletas commented Jun 15, 2026 •

edited

Loading

juanchuletas commented Jun 23, 2026 •

edited

Loading

dyniols commented Jul 16, 2026 •

edited

Loading

dyniols commented Jul 27, 2026 •

edited

Loading