This will help check that the Vulkan devices are enough for WebGPU in
a following CL.
In addition to additional limits this CL:
- Change maxColorAttachments 4 -> 8 to match WebGPU
- Renames kMinDynamicBufferOffsetAlignment to
kMinUniformBufferOffsetAlignment.
- Renames kMaxVertexBufferStride to kMaxVertexBufferArrayStride.
- Changes validation of buffer offsets to use the separate uniform and
storage limits (but no test is added because they are the same).
- Adds validation and a test for kMaxStorageBufferBindingSize.
- Augment the null::Device memory limit for that new test (it allocates
a buffer of 512MB).
- Fix the maxColorAttachment test to not use hardcoded values.
Bug: dawn:796
Change-Id: Ibe4219130a44355ae91c02aaa0a41cf5d9f9e234
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56081
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
- Add zero-sized implementation in each backend.
- Test creating zero-sized query set for each type.
Bug: dawn:954
Change-Id: Ib2b8a7d890dae6ee40f5a05aac570a4600dedab8
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56260
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jiawei Shao <jiawei.shao@intel.com>
Commit-Queue: Hao Li <hao.x.li@intel.com>
This change implements 3D texture sampling with an en2end test.
It turns out that the implementation has already been done. The
test can pass with minor changes.
Bug: dawn:547
Change-Id: I5dfe1a446de3287392e39cb4dd58143e115b02cd
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56000
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Yunchao He <yunchao.he@intel.com>
Add the SampleRateShading capability if the sampling type is `sample`.
Bug: tint:746
Change-Id: I20fb25913f5c0919b6d16a9d0e9fc8b1551ff7ea
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56247
Kokoro: Kokoro <noreply+kokoro@google.com>
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Transforms that handle entry point IO need to strip interpolate
attributes when creating undecorated structures, and not strip them
when recreating the IO parameters and return types.
Bug: tint:746
Change-Id: Ifa7d3b9fe3950154689417ad387026ece3df81bb
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56243
Kokoro: Kokoro <noreply+kokoro@google.com>
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Add E2E tests to cover all of the parameter combinations.
Mark the attribute as unimplemented in the other backends.
Bug: tint:746
Change-Id: I86881ff0b224fe93670db42473341ae185eeabdd
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56244
Kokoro: Kokoro <noreply+kokoro@google.com>
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
They are only valid on entry point parameters and return types, and
struct members. They must only be used on floating point scalar and
vector types. If the interpolation type is flat, the sampling type
must not be specified.
Bug: tint:746
Change-Id: Iab17816bc9947a74593a5937bdf513ac9ec664f1
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56241
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
We're migrating to run all tests with both Tint and SPIRV-Cross.
Remove explicitly instantiating only with Tint, and add
skips for tests that will not work if use_tint_generator is not
enabled.
Bug: dawn:571
Change-Id: I8322dec688e75b6efa8c4f18c6b9439a2e325631
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56261
Auto-Submit: Austin Eng <enga@chromium.org>
Commit-Queue: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
RefBase::Detatch() method in RefBase.h calls std::move,
but no appropriate header was included. Error example was:
RefBase.h:137:22: error: no member named 'move' in namespace 'std'
Change-Id: I60f84bffba06ba5fd3fea7fb9c4908cfe3989472
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56281
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
This is fixup for the
https://dawn-review.googlesource.com/c/dawn/+/54800
where the macre was not properly declared.
Clang doesn't claims about it, but GCC does.
The errors example for GCC 9.3.0:
/dawn/src/dawn_native/CopyTextureForBrowserHelper.cpp:40:5:
error: unterminated raw string
40 | R"(
| ^
/dawn/src/dawn_native/CopyTextureForBrowserHelper.cpp:46:6:
warning: missing terminating " character
46 | )"
| ^
Bug: chromium:1217153, chromium:819294
Change-Id: I47aa2dac37d9dfa7c02532caeb3341edd22fcd07
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56066
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
As all loops in the parser require `synchronized_` to be true, don't set this to true if we've hit the maximum number of error diagnostics.
This lets the parser bubble up as soon as the limit was reached, instead of only aborting once we finally reach the top level loop.
ClusterFuzz has uncovered a number of cases where, in a loop, it can produce an error, then resynchronize, then error again on the next loop iteration. This produces more than the max limit of errors, and can stall the tests long enough to time out.
No unit tests for this, as it requires a really contrived input to trigger, and to exhaustively catch all the places we now call maybe_set_synchronized() would result in a large colleciton of horrible tests. Instead, let ClusterFuzz do the testing.
Fixed: chromium:1224031
Fixed: chromium:1224032
Fixed: chromium:1224042
Fixed: chromium:1224049
Fixed: chromium:1224050
Fixed: chromium:1224130
Fixed: chromium:1224131
Fixed: chromium:1224132
Fixed: chromium:1224144
Fixed: chromium:1224191
Change-Id: I63f160e28d13b85ee92e1b921239d7fb8ac22a08
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56070
Auto-Submit: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: James Price <jrprice@google.com>
Add `out` parameters to expression and type generators.
Use the new helper classes in TextGenerator.
Cleans up bad formatting.
Prepares the writer generating 'pre' statements, required for atomics.
If-else statements are generated slightly differently. This is done so that 'pre' statements for the else conditions are scoped correctly. This is identical to the HLSL writer.
Bug tint:892
Change-Id: I4c6e96c90673ba30898b3682bf3198497d63a2d4
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56067
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: James Price <jrprice@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
logical negation operand must be 'bool' or 'vecN<bool>'
complement operand must be be 'i32', 'u32', 'vecN<i32>' or 'vecN<u32>'
Negation operand must be 'i32', 'f32', 'vecN<i32>' or 'vecN<f32>'
Bug: tint:916 chromium:1216597
Change-Id: Ic88a124a32d16b542560da3ca1c159968d4043a0
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55860
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Fix missing Release() on RefBase::operator=(const T&).
Always acquire the new reference before dropping the existing reference.
Consider:
* Object A holds the only reference to object B.
* Ref of A is assigned B.
If A were released before referencing B, then B would be a dead pointer.
Remove RefBase::kNullValue, just use Traits::kNullValue. Avoids an unnecessary constexpr copy.
Add even more tests.
Change-Id: I111079d56cbf8c09162aa6e493078bead9ae97fa
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55882
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
This makes SPIRV ingestion always go through a SPIRV->WGSL step first.
This will be used to make sure that path works correctly on existing
WebGPU SPIRV sites so we're confident the SPIRV->WGSL translation will
work when compiled to WASM.
Bug: dawn:960
Change-Id: I17efd8c64d9d60ff033ba89b8fff9295e66524ef
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56080
Reviewed-by: Austin Eng <enga@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Auto-Submit: Corentin Wallez <cwallez@chromium.org>
Functions that are not entry points cannot have any attributes on
their return types.
Validate the builtin store types for return types.
Bug: tint:851
Change-Id: I718356b3ab06db4b4502a53b81790e4de0ecfeac
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/56100
Auto-Submit: James Price <jrprice@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
We are now passing all of Dawn's E2E tests with the Tint generator.
Bug: dawn:571
Change-Id: Ie74a07bde805b8652ac23af2bda2fa11597d04ce
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55980
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
WGSL specifies that the size of a host-shareable structure is a
multiple of the largest alignment of any member in that
structure. While SPIR-V cross doesn't generate this trailing padding,
Tint does, and so we need to pad the host structure to match to avoid
validation errors.
Bug: dawn:571
Change-Id: Id5de89125a74c0469ae69d6e86b4c570bbdb31ab
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56120
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Fix a capitalization error in CMakeLists.txt
Link dxgi.lib only in UWP compilation in CMakeLists.txt
Bug: dawn:766
Change-Id: I011f44c8147998a9f4b3a55b25888ca1ad05a3e1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56180
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
This patch adds a missing check in QueueGL::WriteTexture() to return
a validation error when we try to call writeTexture() on the stencil
aspect on OpenGL backends as copying from a buffer into a stencil
texture is not supported on OpenGL.
BUG=dawn:439
Change-Id: I203d40aefe346aea663a492c5a065248b74f8ea3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56002
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
This patch fixes a wrong buffer type declaration in the shader used
in BufferZeroInitTest.BoundAsReadonlyStorageBuffer. In this test the
buffer should be bound as a read-only storage buffer, so it should be
declared as var<storage, read> instead of var<storage, read_write>.
BUG: none
Change-Id: Ib85bf199f64e205b944b7233aae31323cbec1819
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/56042
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
Remove `pre` and `out` parameters from most generator methods.
Use the `out_` string stream in TextGenerator, add helpers to TextGenerator to simplify line printing.
Remove the `pre` and `out` fields from TestHelper.
Cleans up the `pre` aspects of the HLSL writer, so the same concept can be used by the MSL writer.
Fixes indentation bugs in formatting.
Change-Id: Ia35daf632c7c7b84a6fbf3b9ae42baaeb3c97649
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55960
Auto-Submit: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
Tint has rolled, so this test can be re-enabled.
SPIRV-Cross produces bad results - likely the UBO layout for mat2x2 is
incorrectly padded - so this test is suppressed for !UseTintGenerator.
Fixed: dawn:946
Change-Id: If1d030314f51759e6a2a00dd8edc2597efb1eb4a
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55440
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Allows transforms to assert their dependencies have been run before they
are.
Also allows the backends to validate that their sanitizers have been run
before they're used.
Change-Id: I1e97afe06f9e7371283bade54bbb2e2c41f87a00
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55442
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: James Price <jrprice@google.com>
It may not be the root cause of CTS memory leak, but we found some
issues in D3D12 memory allocation for large buffers (> 4M).
The memory allocation is not freed correctly when we call buffer
destroy. When we create a large buffer, it always allocate committed
resources from the system memory and append it to LRU cache, and will
try to free the expired memory in LRU cache when the left memory size
is not enough for the current allocation. When we call buffer destroy,
we remove its heap from LRU cache, release its allocation, but will
add the allocation reference to a delete queue and wait to be freed
together in the next device.tick. At this time, the destroyed memory is
not really released to the system memory. If no tick is triggered during
the buffer allocation and deallocation, these memories will never be
reused again and there will be a memory leak. We need to free these
memories when we detect there is not enough memory, do not wait for
device.tick to do that.
Add a stress case to track the issue, and need to discuss how to fix it.
Bug: dawn:957
Change-Id: Id4bcc97496d372ad7c27c31675327ca12e133217
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55740
Commit-Queue: Hao Li <hao.x.li@intel.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Core WebGPU requires some Vulkan features to be implemented. Check for
them so that we don't return an adapter for a Vulkan device with
insufficient features.
Bug: dawn:796
Change-Id: I445760f184807b8f2d4ee248289b512c18038f33
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55920
Reviewed-by: Jiawei Shao <jiawei.shao@intel.com>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Auto-Submit: Corentin Wallez <cwallez@chromium.org>
This ensures ObjC objects do not leak. On non-Apple
platforms, the scope does nothing.
Fixed: dawn:546
Change-Id: Id43702e6536bf9cb37825a5449511e10cf0734f5
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55841
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Entry point function parameter can only have builtin and location
attributes (unless a disable validation attribute is present). Other
functions cannot have any decorations on their parameters.
Fix parameter creation in the CanonicalizeEntryPointIO transform for a
case where it was not stripping attributes that are not valid for
function parameters.
Change-Id: I000908aa2dc007c52d100c5f65bca051a49bec9a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55863
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Use a DisableValidationDecoration to allow these storage classes only
for variables generated by the SPIR-V sanitizer. This is also used in
the Inspector tests, since the Inspector is currently run *after* the
SPIR-V sanitizer. These tests will be removed when this is no longer
the case.
Also validate that builtin/location decorations are not used on
variables (unless they have input/output storage class).
Fix or delete all of the other tests that were wrongly using these
storage classes and attributes.
Bug: tint:697
Change-Id: I8be7fb16191f5e2bed9f7dfb700e51f3b97fd1fe
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55862
Auto-Submit: James Price <jrprice@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Describes what Tint system raised the diagnostic.
Use this information in the fuzzers to distinguish between expected and unexpected failure cases in the Transform fuzzer tests.
Fixed: chromium:1206407
Fixed: chromium:1207154
Change-Id: I3b807acafe384a2fc363d2a4165a29693450b3cf
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55254
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
HLSL usually implicitly casts a vector down to a scalar, but this breaks when passing the vector to RWByteAddressBuffer.Store (for DXC only).
Fixed: tint:827
Change-Id: I67d0bc6e9185de3d434a7aaeb575d83850111ec5
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55760
Auto-Submit: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
This test is invalid for multiple reasons (wrong types, wrong stages,
global variable instead of a parameter). The shader IO E2E tests give
better coverage for this anyway, so just delete the unit test.
Change-Id: I1627addf1e407e473e33043341d742958ceebca5
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/55861
Auto-Submit: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Set BUILD_EXAMPLE to OFF in UWP.
Disable Vulkan in UWP.
Compile GLFW related sources only on supported platform.
Remove linking to user32.lib in UWP.
Add linking to dxgi.lib only in debug configuration.
Use EmptyDebugLogger in UWP.
Bug: dawn:766
Change-Id: I3e6149058048d1b597267dfb9d63342b6a54e0af
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55260
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
This patch modifies the ValidateLinearTextureCopyOffset for T2B and B2T
commands to make sure that buffer offset is a multiple of 4 when the
texture is of a depth/stencil format. The unittests are also updated
to cover these cases.
Bug: dawn:721
Change-Id: I5cb9487e61e7f6ac7075157a092a1f3577844042
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/55720
Commit-Queue: Zhaoming Jiang <zhaoming.jiang@intel.com>
Reviewed-by: Austin Eng <enga@chromium.org>
