- Uses an anonymous function to delete command pool/buffers.
- Shuffles the code around a bit so that the CommandPoolAndBuffer are
clearly next to the EncodingContext stuff to make it clear that we
may be able to consolidate them in the future.
Bug: chromium:1372772
Change-Id: I92a1d0333b7a85d439b5963a58db69ac685c03a4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/112181
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
- Remove Types variant, and replace with a type-erasing Value class
instead. This is not only better for compile times, but makes the code
much easier to understand.
- Value wraps an internal shared_ptr to a const detail::ValueBase,
allowing it to be used as a value-type (i.e. copyable), while behaving
polymorphically.
- Add static_asserts to Val, Vec, and Mat creation helpers to emit a
more useful error message when the wrong type is passed in.
Bug: tint:1581
Change-Id: Icd0d08522bedb3eab12c44efa0d1555ed6e96458
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/111700
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
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 CL updates the if and switch nodes to store the condition value in
a register. The EmitExpression is updated to return a Register and the
builder updated to emit the expressions for the if, break-if, while,
and switch expressions.
Bug: tint:1718
Change-Id: Ie710812c74e8b9423a4aa997db451d9cdf304feb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/110784
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
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>
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>
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>
Add a transform to truncate unused user interstage variables by
adding a new truncated shader io struct wrapper of the original
one, with a truncate function to do the assignments called
at the return statement.
This transform is meant to be run after CanonicalizeEntryPointIO,
and will only be run under hlsl/generator_impl.cc to workaround
the extra register limitation for interstage variables on D3D FXC.
Bug: dawn:1493
Change-Id: I69081189ad7d4b76f2371fcc079f67dced2e9944
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/104620
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Shrek Shao <shrekshao@google.com>
