dawn:
* Zero-initialize the SpecializationDataEntry. For booleans, the first
byte may be initialized, while the 3 other bytes may contain garbage.
This can cause boolean specializations to be 'true' when 'false' was
specified.
dawn_node:
* Implement conversion of GPUProgrammableStage's constants.
* Fix Napi::Object -> unordered_map conversion.
* Fix member name in error messages generated by
'DictionaryMembersFromJS'
Bug: dawn:1041
Change-Id: I016ec4916fc73226dfe1bd8e7dc52a1e88e46a6b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67383
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
These expressions were not emitted as spec-ops, failing new CTS tests.
Change-Id: I56e8f56a22de2b8dac9a8bd7a2d694d8d81dca35
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67480
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Reviewed-by: James Price <jrprice@google.com>
Until all backend implementations are done.
Bug: dawn:1169
Change-Id: I16d184bec03bbc9a255ad33cf11aa615b1986389
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67422
Commit-Queue: Shrek Shao <shrekshao@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Use a transform to convert these to the vector form for the MSL and
SPIR-V backends.
MSL only has the scalar form from version 2.0 onwards.
Fixed: tint:1123
Change-Id: I384abd9872d9eae52a10a37cbd6aa96004692e9c
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67360
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This reverts commit d84d7f1ea3280e90e27242df50334133513ed8da.
Reason for revert: CTS tests for this are now enabled.
Original change's description:
> Disable DrawIndexedIndirect
>
> Not passing CTS yet
>
> Bug: dawn:809
> Change-Id: Ib7c240372a8fdbd45803230292c374e2957c8d15
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/65601
> Reviewed-by: Ben Clayton <bclayton@google.com>
> Reviewed-by: Kai Ninomiya <kainino@chromium.org>
> Commit-Queue: Austin Eng <enga@chromium.org>
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: dawn:809
Change-Id: I3018fec67c278343956a6da68041bc6ec24d55f4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67420
Reviewed-by: Ben Clayton <bclayton@google.com>
Auto-Submit: Austin Eng <enga@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
A call-statement can now be made on a function that returns a value.
Bug: tint:1256
Change-Id: I9a3cc4e330f7441e09dc408bdc1cfa515adac392
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67382
Commit-Queue: Ben Clayton <bclayton@google.com>
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
The rules have been relaxed about function values always needing to be
consumed, so a bunch of these phony-assignments can be replaced with the
RHS, but this will be done as a follow-up.
Bug: tint:1213
Change-Id: Ie7c4280f87b4ad7e5a429994b0b88ac22c2f3a9f
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67300
Commit-Queue: Ben Clayton <bclayton@google.com>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Use spec-spelling of 'constructible'.
Add missing test file to 'test/BUILD.gn'
See https://dawn-review.googlesource.com/c/tint/+/67064
Change-Id: Ie39773617fd0a363d63cc6449bf3905c9eb6786d
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67380
Auto-Submit: Ben Clayton <bclayton@google.com>
Commit-Queue: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: James Price <jrprice@google.com>
It's been removed from the spec:
https://github.com/gpuweb/gpuweb/pull/2127
Fixed: tint:1213
Change-Id: I163fe807765bb1ac0580b398f4897daea555216a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67067
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
The `Ignore()` intrinsic is about to be deprecated, so don't use it for testing.
Bug: tint:1213
Change-Id: Ib5d5966da6d566a9f02940970ebd79d520b5e0e8
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67065
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
This reverts commit c7e6bb0d8d8f3bed3ca27ff822631c1671a4e5b2.
Reason for revert: clusterfuzz identified issue https://bugs.chromium.org/p/chromium/issues/detail?id=1262112
Original change's description:
> Improve validation errors for encoders
>
> Improves the validation messages in ComputePassEncoder.cpp,
> ProgrammablePassEncoder.cpp, RenderBundleEncoder.cpp, and
> EncodingContext.cpp/h to give them more contextual information.
>
> Bug: dawn:563
> Change-Id: I87c46c4bfda1375809fae93239029ea4e3b9c0a2
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67000
> Commit-Queue: Brandon Jones <bajones@chromium.org>
> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: dawn:563
Change-Id: I259ccde1735c4201ff2736562cfe4689e9a22f62
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67321
Reviewed-by: Brandon Jones <bajones@chromium.org>
Commit-Queue: Brandon Jones <bajones@chromium.org>
Produce a warning if the attribute is missing for integral vertex
outputs or fragment inputs. This will become an error in the future,
as per the WGSL spec.
Add the attribute to E2E tests.
Bug: tint:1224
Change-Id: Ia1f353f36cb7db516cf9e8b4877423dec3b3e711
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67160
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Same as Chrome's args, this allows us to set the
DeviceDescriptor::forceEnabledToggles and forceDisabledToggles when
creating the GPUDevice.
Example: node cmdline.ts ... --gpu-provider-flag=enable-dawn-features=dump_shaders ...
Bug: dawn:1163
Change-Id: Ib5db71355f72e5d08f8fe87313c5e3d63ee236c3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66963
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Only used on Windows builds for now, this flag adds the input path to
the DLL search paths. The main purpose is to ensure we load the right
version of D3D dlls, such as d3dcompiler_47.dll from the Chrome output
dir, rather than the default one in the Systems directory.
Bug: dawn:1163
Change-Id: I8e696dd877ec715e1e54d8589af8275e62c90937
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66962
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Do not emit block-decorated structs.
Combine them with the variable into a uniform or buffer interface block.
Bug: tint:1223
Change-Id: I16263ea449ecab705319dc440275b7d169021bc9
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67243
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
As an alternative to setting the "DAWNNODE_BACKEND" env var.
Bug: dawn:1163
Change-Id: I33ace14e7603b97caf7286a109a31160853f62e0
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66961
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Covered by GlslGeneratorImplTest_Function.Emit_Function_WithArrayReturn
Bug: tint:1258
Change-Id: I181641ebe170ec359d9833e9022e6e70747fde9a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67240
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
This will be needed for emitting UBO and SSBO interface
blocks.
Bug: tint:1223
Change-Id: I1cdc75b67a4fe612dcf8094a7d73b8bf1f85f40a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67242
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
This change replaces the ".gpu" export with ".create()" function that
accepts an array of flags. These will be used by cmdline.ts to set flags
such as what dawn backend to use. We currenly environment variables, but
this will be more flexible.
Bug: dawn:1163
Change-Id: If2fb35811cac45e16121fbd828f997ef3d795f36
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66960
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Prevent reusing offsets from a previous direct draw.
Update test to verify that values are updated correctly
for each draw. Add tests for indirect draw offsets.
Bug: dawn:548
Change-Id: Ice8325a8a41b8a4375767156dbaba3ee3d714f3b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67121
Reviewed-by: Austin Eng <enga@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
Improves the validation messages in ComputePassEncoder.cpp,
ProgrammablePassEncoder.cpp, RenderBundleEncoder.cpp, and
EncodingContext.cpp/h to give them more contextual information.
Bug: dawn:563
Change-Id: I87c46c4bfda1375809fae93239029ea4e3b9c0a2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/67000
Commit-Queue: Brandon Jones <bajones@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Metal 1.x does not support swizzling on packed_vec types.
Use array-index for single element selection (permitted on LHS and RHS of assignment)
Cast the packed_vec to a vec for multiple element swizzles (not permitted as the LHS of an assignment).
Fixed: tint:1249
Change-Id: I70cbb0c22a935b06b3905d24484bdc2edfb95fc2
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/67060
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Ben Clayton <bclayton@google.com>
Reviewed-by: James Price <jrprice@google.com>
It is always on now when using tint::CommonFuzzer, and runs before &
after the transform step.
This CL also adds missing API coverage to the Inspector fuzzing code.
Errors found with the Inspector are now reported as fuzzer failures
and should generate bug reports.
BUG=tint:1250,tint:1251,tint:1250
Change-Id: I1c1bcbddf81a35620f89c5b7a648c44e6a1f2952
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/66980
Auto-Submit: Ryan Harrison <rharrison@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Alastair Donaldson <afdx@google.com>
This patch implements [[num_workgroups]] on the API side for
Dispatch() calls by setting num_workgroups.xyz as root constants.
This patch also adds a temporary validation that on D3D12 backend
using a compute pipeline with [[num_workgroups]] in a
DispatchIndirect call is not supported.
BUG=dawn:839
TEST=dawn_end2end_tests
Change-Id: Iaee2ffd162e9420e4e80944fbb222f10a4600c6a
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66580
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
This commit also unifies the initialization process for Adapters.
InitializeImpl() initializes the actual backend adapter.
InitializeSupportedFeaturesImpl() checks base WebGPU features and
discovers additional supported features.
InitializeSupportedLimitsImpl() checks base WebGPU limits and
queries the adapter's maximum supported limits.
Some of these limits from the backend are still overriden in the
frontend because they are limited by internal Dawn constants.
Bug: dawn:685
Change-Id: I43efb0b678dd45f8f89cd62d13104dd00b197da1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/64980
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Brandon Jones <bajones@chromium.org>
- Moving the call into DeleteThis should make it so that derived classes don't need to explicitly implement a destructor that calls DestroyApiObject.
Bug: dawn:628
Change-Id: I145f42e7e4c144cc0d2d7c7f609744399d514fe1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66840
Commit-Queue: Loko Kung <lokokung@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
This allows tests to be streamed to N node processes, without incurring the cost of re-scanning the test lists for each case.
It also means that there is an increased chance of state leakage. `--isolate` is a new flag that uses the old behavior of running each test in a separate process.
Depends on: https://github.com/gpuweb/cts/pull/789
Change-Id: Ifc92d1cc07a9de3b2751bed0971f3424d0c247ae
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66920
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
If we use render bundle, the compability validation between
writeDepth/Stencil in DepthStencilState in render pipeline and
depth/stencilReadOnly in DepthStencilAttachment in render pass
will become two steps:
1. validation between render pipeline and render bundle during
RenderBundleEncoder's SetPipeline().
2. validation between render bundle and render pass during
RenderPassEncoder's ExecuteBundles().
So, render bundle is like a bridge for this compability validation
between pipeline and pass.
The first step has been done in previous patch. The patch does
the second step.
Bug: dawn:485
Change-Id: I1226494e901c07bdb9f565bce7b9073d420f2fe2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66842
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Yunchao He <yunchao.he@intel.com>
Spinning up new devices for each test can take a long time.
Specifying --j 0 will run a single instance of node, with the given query to run.
Change-Id: I27c161bb76f5deaaa505ab5ae361ea6a0942a130
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66880
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
https://github.com/gpuweb/cts/pull/788 significantly improves the performance of incremental builds, but it can still take around 2 seconds to do that incremental build.
Have the run-cts tool scan the CTS source tree for the most recent modification, and only rebuild if the timestamp is more recent than the last build time.
Significantly reduces startup overhead of the tool.
Change-Id: I7e9227127c02e6d4dc8875a577178956bf5e2481
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66822
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
These should always be resolved or rejected.
The Fatal() call, when a promise is not resolved or rejected, is currently disabled due to https://github.com/gpuweb/cts/issues/784.
Bug: dawn:1123
Change-Id: Ie0e8ac187ad70be0fea41cd66956d0bfd9c53212
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66821
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
The AsyncRunner will enqueue a call to `Device::Tick()` when the runner count moves from 0 async tasks to 1.
It has been observed that some 'async' tasks are actually synchronious, which results in multiple tick callbacks being enqueued before the first has a chance to run.
Fix this by using another boolean to track whether the function has been queued.
Bug: dawn:1127
Change-Id: I7dd81d33d601bf1d3cefb5c4dad6c237883e51ee
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/66820
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
