This CL implements f16 in uniform and storage address space, allowing
using f16 types in uniform and storage buffers on all backends. Tint
uint tests and Dawn E2E tests are added to validate the f16 types work
as expected.
Bug: tint:1473, tint:1502
Change-Id: I15e3de1033d3727f2ea33f4657f682c5f13c2153
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/106320
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Zhaoming Jiang <zhaoming.jiang@intel.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL change the DXC version checking logic to get both DXC compiler
and validator version, which are not necessarily identical, and require
both version being 1.6 or higher to enable the use_dxc toggle.
This CL also modify the src/dawn/tests/BUILD.gn and add a copy target as
data_deps for "dawn_test" template, which copy DXC binaries from Windows
10 SDK 20348 to out directory, to ensure that windows trybots running
dawn_end2end_tests.exe (e.g. win-dawn-rel) can access a DXC of version
1.6 and can run end-to-end tests with DXC.
Bug: tint:1719
Change-Id: I39b48f3dffdf121d3749af7aa4b3d0bed1c22ea8
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110340
Commit-Queue: Zhaoming Jiang <zhaoming.jiang@intel.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Repeatedly creating and destroying WebGPU devices in a loop causes
large amounts of memory to pile up in the GPU process. Much of this
memory comes from the D3D12 command queue.
Releasing the command queue early in DestroyImpl before the destructor
runs goes a long way towards relieving the memory pressure.
Bug: chromium:1377789
Change-Id: I3ff9a5f6cb3ea3136e41079343532cbe732b6cc4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/111280
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Rafael Cintron <rafael.cintron@microsoft.com>
This patch updates the validations about CreateBuffer() with dawn_wire
to match the latest WebGPU SPEC.
According to the SPEC, the validations in CreateBuffer() should be
executed in the below order:
1. If mappedAtCreation == true, return nullptr and a RangeError will be
generated in Chromium.
2. Validate BufferDescriptor and check if there is OOM at device timeline
3. Check if there is OOM at content timeline
Bug: dawn:1586
Change-Id: I97ff5f82a42208442ddf6e46e66381c3b3680450
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/109040
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
dawn.node now uses tint::Initialize() and tint::Shutdown(), and so
needs to link against libtint.
Fixed: tint:1765
Change-Id: I03e575b4709c43a6052ed3ca635376251c3323c0
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/111080
Auto-Submit: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Pads serialized wire command buffers to 8 bytes so that we don't have
misaligned write/reads which can cause SIGILL depending on platform and
compilation mode, i.e. -c dbg in google3 builds.
- Adds helpers for aligning sizeof calls.
- Adds constant for wire padding (8u).
- Modifies BufferConsumer to allocate according to padding. This
guarantees that when we [de]serialize stuff, the padding should be
equal on both sides.
- Modifies extra byte serialization code (adding CommandExtension
struct). This makes it clearer that each extension needs to be
padded independently. Otherwise, before in wire/client/Buffer.cpp,
since the read/write handle sizes were being passed as a sum, but
read out separately from the BufferConsumer, we corrupt our pointers.
- Adds some simple unit tests.
Bug: dawn:1334
Change-Id: Id80e7c01a34b9f01c3f02b3e6c04c3bb3ad0eff9
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110501
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Loko Kung <lokokung@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Adds tint_public_config to Dawn's internal config so Tint headers
can be included.
Fixed: dawn:1594
Change-Id: I4068fd95b6eae3138fbcc04f29f054c7cffdcf12
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/111260
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Shrek Shao <shrekshao@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
Kokoro: Austin Eng <enga@chromium.org>
Sync up with current WebGPU spec to allow FS input being a
subset of VS output instead of requiring a strict match.
This patch involves changing the validation and adding tests,
together with using the TruncateInterstageVariables for hlsl
generator to workaround the extra limit for D3D12 backend.
Bug: dawn:1493
Change-Id: I2d4ba7f43dbe57f17ecd5c5d659f4ca93bb682a3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/109460
Commit-Queue: Shrek Shao <shrekshao@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Auto-Submit: Shrek Shao <shrekshao@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Calling UnmapInternal would set the state to Unmapped, allowing the
buffer to be mapped again even though it is destroyed.
Bug: chromium:1388920
Change-Id: Ibb4da332bafd44a0d4900c8ea5bfbd674bbc35e0
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/111121
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Loko Kung <lokokung@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
- Messed up the conditon in previous change so it required both the
enable flag to be on and the disable to be off, but then we can't
deprecate the enable flag in Chromium.
Bug: dawn:549
Change-Id: I1c730939104aafaef48182238fa32ed7fa6a1e16
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110983
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Loko Kung <lokokung@google.com>
- The disable toggle will supercede the enable one once Chromium side
deprecates usage of the enable one.
Bug: dawn:549
Change-Id: I5c5bd60161917fe2654cfce55a6f29e8a7e79962
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110728
Commit-Queue: Loko Kung <lokokung@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
This is a catch-all that handles all cases where an operation can result
in non-finite values, such as from calls to std::cosh and std::sinh.
Bug: tint:1581
Bug: tint:1747
Change-Id: Ibb55466fea01b263c98d598459c788fd22cf5bb7
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110726
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This will be required to use the new data-caching functionality provided by:
https://github.com/gpuweb/cts/pull/1985
Also propagate `--verbose` down, if passed to the frontend.
Change-Id: Ic436e8ed754296cec859c45bd4db703634c31ab1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/109764
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
- The test was failing on the google3 TAP bots.
- Also adjusts the size so that it is the aligned min given the limit,
not just the aligned limit since that can be larger than the limit.
Bug: dawn:1217
Change-Id: I02a68d14d61099dc5c1a2450cbaadc2320f5b8a9
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110760
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Loko Kung <lokokung@google.com>
This patch adds a workaround on Intel Gen9.5 and Gen11 GPUs to always
allocate 2D textures with CopyDst as committed resources instead of
placed resources to mitigate a driver bug about CreatePlacedResource().
Bug: chromium:1237175
Test: dawn_end2end_tests
Change-Id: I64ab9c083c8835fb2971660eed51252fecac416c
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/100641
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
- Issue found in LinkedListTests when running against ASAN since the
LinkedList was being destroyed after the Nodes, thereby triggering
a RemoveFromList on the root node, but the other nodes were never
removed from the list and are dangling pointers.
Change-Id: I136abbc5d73c35142990c9fe4669e5fc6d5ef644
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110500
Reviewed-by: Austin Eng <enga@chromium.org>
Auto-Submit: Loko Kung <lokokung@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Since overrides can be used to size workgroup arrays and also as
initializers to module-scope variables, we cannot just consider
overrides that are directly referenced in the shader functions.
This change makes the Resolver track references to overrides whilst
resolving array types and module-scope variable declarations, so that
they are included in the set of overrides reported by the Inspector in
these scenarios.
Fixed: tint:1762
Change-Id: If7501abf3ddcb87a87134ddd578aa4904d204de6
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110460
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: James Price <jrprice@google.com>
Auto-Submit: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This reverts commit f3666c45f3362c7faf7cbdd9328ac76d21327a2a.
Reason for revert: Did not realize these assertions could knowingly be reached after this fix for crbug.com/1340654
Original change's description:
> Add assertions to ServerInlineMemoryTransferService to catch corrupted states while fuzzing with DawnWireServerFuzzer and tests.
>
> Bug: chromium:1340654,chromium:1374495,chromium:1376477
> Change-Id: Icfb008a1cd6dbd8af32f3aedc90ef29e29a0465b
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/106041
> Commit-Queue: Brendon Tiszka <tiszka@chromium.org>
> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
> Reviewed-by: Austin Eng <enga@chromium.org>
> Kokoro: Kokoro <noreply+kokoro@google.com>
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: chromium:1340654,chromium:1374495
Change-Id: I5bff56e32266721ec418a545ad72c1065c73d526
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/109461
Commit-Queue: Brendon Tiszka <tiszka@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
In cases that a debug message does happen, we shouldn't use
unreachable as it will make the compiler use a false assumption.
Instead, only ASSERT(false) which crashes in debug, and does
nothing in release builds.
Bug: chromium:1375131
Change-Id: I7733151c241ee875ac40969ce22f037351141e89
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110600
Auto-Submit: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
This patch adds a test to dawn_end2end_tests to reproduce a driver
issue about creating textures with CreatePlacedResource() on Intel
D3D12 drivers.
Bug: chromium:1237175
Test: dawn_end2end_tests
Change-Id: I26fe6c9b827d8a05cfe2336405e43c549e52ea50
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/100567
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Use 'Mock' instead of non-inclusive language 'Dummy' in toggle
MetalUseDummyBlitEncoderForWriteTimestamp to pass PRESUBMIT.py check.
Bug: dawn:1250
Change-Id: I47dff4adec08d4d076b5ae5b4e0a521b235d5868
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110060
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>
The DXBC asm listing attempts to be helpful when display literals by
making them human readable,but when it displays values as floats, they
will only have a precision of 6.
For example, 0x09909909 ends up getting displayed as 0.000000 because as
a float, this value is 3.48106525683e-33. FXC has an option to output
literals as hex values, so let's use that.
Change-Id: I2b3017bd834eac89248fe01cae85ba1bb4033e59
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/109582
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
This CL moves reserved words from a deprecation to an error.
Bug: tint:1463
Change-Id: I5c66baa15dc748215877c8152171c690495bc0c2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/108861
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL moves module-scope let from a deprecation to an error.
Change-Id: Iffecbb667cf79515234b6510ce7c5bbbb6e673bc
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/108862
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Add more coverage for timestamp query on D3D12 backend to make sure
timestamps are converted correctly:
- All timestamp queries inside and outside passes
- The 'disable_timestamp_query_conversion' toggle disabled and enabled
Bug: dawn:1250
Change-Id: Ibdc6b35faed7cc1e1a8b60df4a5032914b411bc1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/108022
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Hao Li <hao.x.li@intel.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Change-Id: I2f9ba7f98acdf9441d80d3a1169b36bee44a2e0e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/108545
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
WebGPU specification was updated that GPUBuffer.unmap()
shouldn't validate buffer state.
https://github.com/gpuweb/gpuweb/pull/3368
This commit reflects it to the Dawn implementation.
Bug: dawn:1528
Change-Id: Ie66f68214bd6896a4d674ed00addc3ffb539c235
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/106264
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Takahiro <hogehoge@gachapin.jp>
Have Initialize() bind the Program printer - which is helpful for debugging.
Call these from dawn/node.
This allows dawn/node to print programs when things go wrong.
Change-Id: I32d8805381d2939e82dc6ea383b9860fbb5fb69e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/107684
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Only tag to submit when the total size is larger than the threshold,
so that we can make as few submits as possible meanwhile avoiding OOM.
Bug: chromium:1258986
Change-Id: I7190e1bb942bfaffc5cd424ce4743173735b25e3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/106418
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Jie A Chen <jie.a.chen@intel.com>
Current ExternalTexture visible related info only contains width and height.
But not all the visible rect started at (0, 0). This CL add visibleOrigin to contain origin info and record (width, height) in
visibleSize
Bug: chromium:1361363
Change-Id: I3d8931e490c97740f152653383f07d0a2d984dd3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/108024
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Shaobo Yan <shaobo.yan@intel.com>
Timestamp period is also needed by the timestamp-query-inside-passes
feature, which is enabled separately from the timestamp-query feature.
Bug: dawn:1193
Change-Id: I8a1f87f8d7931261b87608306820daefc4c3dc55
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/107880
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Hao Li <hao.x.li@intel.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
`-o` will emit the files to the given root output directory
`--verbose` will print what's going on, to help with debugging
Omitting these flags will behave as before.
Also consolidate the utils package into fileutils. These were two packages with near identical functionality.
Change-Id: I855dd4b57807fb9239a52e7f357842d4ba2517ee
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/107687
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Without this other Tint transforms may end up calling user code instead
of builtins (for example for the min() used in robustness).
This commit does the following changes:
- Changes ShaderModuleVk to return a CompiledSpirv object instead of
just a Spirv Blob so that a remappedEntryPoint can be stored in the
cache alongside the SPIR-V.
- Inlines the logic and simplifies TransformedConcurrentShaderModuleCache
slightly (by introducing a struct instead of std::pair, and adding a
conversion method to ModuleAndSpirv).
- Adds the Renamer transform to ShaderModuleVk and adapt the code to
use the remappedEntryPoint where needed (pipeline creation and
post-compilation reflection).
- Adds a test where the min() used by the robustness transform is
overriden to return a constant 0.
- Moves the Renamer transform to be just after the SingleEntryPoint
transform in D3D12 and Metal as well so as to make the test pass.
Fixed: dawn:1583
Bug: dawn:1585 dawn:1587
Change-Id: Ia9de38d391a7901ed04b097f4a8d439759f7556e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/107020
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
This adds 'HasScheduledCommands()', with which Dawn no longer needs the
future serial to tick and track the async tasks.
Bug: dawn:1413
Change-Id: Ide9ba69b796a46fa8bb70b002f4e2aeb1622bffd
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/98720
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Jie A Chen <jie.a.chen@intel.com>
The assert failure is caused by the validation orders. Current
validation order cannot ensure "destination" is valid when it is
passed to "ValidateTextureToTextureCopyCommonRestrictions".
This CL seperate "ValidateCopyForBrowserCommonRestrictions" to
"ValidateCopyForBrowserDestination" and "ValidateCopyForBrowserOptions".
Correcting the order and adding more comments.
Bug: chromium:1379001
Change-Id: I9bdbd773659827d0056cd7c37e78ac02ce22451c
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/107560
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
