Fixes the regex fuzzer so that when searching for an operator to
replace, it takes account of the fact that the string being searched may
be very small, avoiding an issue where unsigned integer underflow would
occur.
Bug: crbug.com/1359193
Change-Id: I653a20429dc20385a64f8d684c81d023702458e6
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102641
Kokoro: Kokoro <noreply+kokoro@google.com>
Auto-Submit: Alastair Donaldson <afdx@google.com>
Commit-Queue: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
• Add `cts validate` command used to check for expectation collisions.
Can be used as a presubmit check.
This is more tightly checked than the previous logic, as this works on just
the expectations, instead of results.
• Fix an issue where the test result reduction could introduce collisions with
'Skip' expectations.
To fix this, the update process first adds 'consumed' results for the skipped
tests, preventing test tree reduction for that part of the tree.
• Fix a bug in the generation of 'New failures' and 'New flakes' which produced
more expectations than was necessary.
The issue here was that the tree roots could contain overlaps, and roots could
be processed before sub-trees, resulting in inefficient expectations.
• Fix collisions in the expectations file, and update with results from
the most recent roll.
Change-Id: I7b64553408998fb4416458ce564fc49c8f6d4d07
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101860
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Shuffle the transform orders to ensure that these are embedded in a structure before running the Std140 transform.
Add more end-to-end tests for these.
As pointed out in tint:1673, arrays of matrices are not correctly decomposed by the Std140 transform.
This will be addressed by a later change.
Bug: tint:1673
Change-Id: I47c93e458ff48578922d576819792e8ed3a5723c
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102541
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Zhaoming Jiang <zhaoming.jiang@intel.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
We already compute the "first" and "last" basic block that
uses a value, so we could know when to hoist a value into
a var declaration. You have to do this sometimes to make
sure all uses are in scope of the declaration.
Until now we tracked Phis with an entirely different mechanism.
But there are cases which broke down. That's what happens
in crbug.com/tint/1649.
Additionally, GraphicsFuzz cases generarte similar weirdness.
Also, be more careful about ensuring that the assignments
generated to feed phis behave as if they occur in parallel.
Within a single batch of such assignments, generate and
use intermediate let-declarations for phis that that batch
will overwrite.
Also, unwrap-references when rectifying the signedness of
binary operators.
Skip tests that fail due to crbug.comt/tint/98:
test/tint/bug/tint/749.spvasm.*
Fixed: tint:1649
Change-Id: I7314c351b74a10bfa9a18011f3d80a520568011c
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101220
Auto-Submit: David Neto <dneto@google.com>
Commit-Queue: David Neto <dneto@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Metal configures the query set and query index for the beginning and end
of passes in pass descriptor when beginning a pass encoder, so we need
to record all timestamp writes information in BeginXxxPassCmd. For the
platfroms that support timestamp query, it must support timestamp write
at command boundary or stage boundary, if the stage boundary is
supported, use sampleBufferAttachments API for Metal implementation,
otherwise simulate timestamp write using sampleCountersInBuffer API
after begining a pass and before ending a pass.
Bug: dawn:1250
Change-Id: I462cb05a0102521cd2df4db3ac6f71863419b933
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/93940
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Hao Li <hao.x.li@intel.com>
This removes a serialization optimization where structs could be
memcpy'ed. memcpy is still used for deserialization of structs.
Structs may contain padding bytes which when copied can leak
uninitialized data across a trusted boundary. In Chrome, this
means previously-written or uninitialized bytes in the GPU process
could leak into the Renderer process.
In the future, we may be able to bring this back by introducing
a concept of safe-to-memcpy structs, or by forking the code so
that Renderer->GPU process uses the unsafe memcpy, and
GPU->Renderer uses to the safe member-by-member version.
An alternative might be to ensure that everything returned from
the WebGPU API has initialized padding, but this invariant is not
trivial to guarantee.
Fixed: chromium:1359098
Change-Id: I91151251d1ab999e0f5552e1efccc472e451cc10
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102461
Reviewed-by: Brandon Jones <bajones@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
This patch updates the validations on the inter-stage shader variables to
match the latest WebGPU SPEC (in chapter "validating-inter-stage-interfaces").
With this patch the below validation tests in WebGPU CTS will pass:
- render_pipeline,inter_stage:max_shader_variable_location:*
- render_pipeline,inter_stage:max_components_count,*
Fixed: dawn:1448
Test: dawn_unittests
Change-Id: I3e4d98f03ec18e5d1642a4d7ecd3eed1b7ae04d0
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102104
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
- Adds Prepend function to LinkedList to avoid directly using the
insert functions on the LinkNodes. (And tests for this as well.)
- Adds ApiObjectList class for tracking lists of objects for
destruction.
- Renames and virtualizes some tracking interfaces so that they can be
overriden for the TextureView/Texture cases.
- Removes explicit destroying of TextureViews from Device since
destroying Textures will destroy TextureViews now.
Fixed: dawn:1355
Change-Id: I3522383ea7724d6e41ac0c805793a6c34d9bec27
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101762
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Loko Kung <lokokung@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This reverts commit c1f5112462.
Reason for revert: underyling issue in crbug.com/dawn/1539 fixed
Original change's description:
> Remove code to set Vulkan backend debug labels
>
> This is hitting a memory corruption issue inside the NVIDIA driver.
> Speculatively delete the code in the Vulkan backend until further
> analysis finds the root cause.
>
> Bug: dawn:1539
> Change-Id: Ie7bf5bed31976da5f13325c81033e787c4d376b9
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102100
> Reviewed-by: Loko Kung <lokokung@google.com>
> Kokoro: Kokoro <noreply+kokoro@google.com>
> Commit-Queue: Austin Eng <enga@chromium.org>
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: dawn:1539
Change-Id: Ic1e4c4ad05abd6f2e244e4a5364bf54e6288cffb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102107
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Loko Kung <lokokung@google.com>
FencedDeleter does not need to be used at all since we were deleting
a yet-to-be-used VkShaderModule.
Also, set the VkShaderModule label before cache deduplication since to
avoid a race where the VkShaderModule is in use by another thread.
The other thread may also be setting the label, or using it in a
pipeline creation.
Bug: dawn:1539
Change-Id: I5e3d7ce214c4c089c9cc3272f373aa8233017965
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102105
Commit-Queue: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Loko Kung <lokokung@google.com>
Update with additional argument docs, and docs on uploading builds
to Swarming. I've found this particularly useful lately.
Bug: none
Change-Id: Iee04043b7260c505d4d63025c4038e3eb50db338
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102106
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
This patch replaces NULL with nullptr as nullptr is preferred as null
pointers in Chromium coding style.
Bug: None
Change-Id: Ie6ab9d606d791bad2c50001815062c22e9ec0d25
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102102
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
may_exonerate indicates that a test failed for a known issue that
we could exonerate. Merging of test results now removes results
with may_exonerate unless all of them were tagged as such. So, if
for example, a test fails for a known timeout issue, but has a
subsequent pass, the timeout will be ignored.
This serves to reduce the impact of known, hard-to-fix issues and
allow the CTS roller to make progress with less noise.
Change-Id: I5103a666496398a17b3aa6ccf3f267421e40ba97
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101804
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
Following https://github.com/gpuweb/cts/pull/1836, this will let Chrome's
CTS runner bypass the builtin CTS timeouts in favor of our own harness
timeouts.
Bug: chromium:1340602
Change-Id: I09f25de26f28f4595c6d277050f2a71969566fdb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/102200
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
Since GLSL ES does not support the offset= attribute, struct members
with explicit @align or @size attributes require adding explicit
padding members. This in turn requires rewriting any constructor
calls to initialize the new padding to zero, handled in the same
transform.
Note that this is currently overly-verbose, and will add padding where
GLSL doesn't technically need it (e.g., padding a vec3 out to 16 bytes).
Bug: tint:1415
Change-Id: Ia9ba513066a0e84f4c43247fcbbe02f5fadd6630
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101720
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Alastair Donaldson
Ben Clayton
Antonio Maiorano
Dawn Autoroller
Austin Eng
David Neto
Li Hao
Jiawei Shao
Loko Kung
Seto
Wang Qing
Shrek Shao
Yan,Shaobo
Stephen White