- Note that by default these are already errors, not warnings.
Change-Id: If35284041963cd387839fe102efe4dca73e594dc
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128064
Reviewed-by: Shrek Shao <shrekshao@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
BufferBase uses a staging buffer to upload data for non-mappable
buffer created with mappedAtCreation. However BufferBase doesn't
unmap staging buffer before copy data from it. But d3d11 debug
layer complain this illegal usage. It causes test failures.
Bug: dawn:1772
Change-Id: Id1c386ac7c45f41487f9cc7ef4e431eab87ba1c2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128480
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Peng Huang <penghuang@chromium.org>
- Note that by default these are already errors, not warnings.
Change-Id: Iab9ecf3cfd54c0219777c2d1587a7752c3173595
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128102
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Loko Kung <lokokung@google.com>
Reviewed-by: Shrek Shao <shrekshao@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
- Note that by default these are already errors, not warnings.
Change-Id: I46afa6e54e7915ba54aa6990cd641288609108c1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128101
Reviewed-by: Shrek Shao <shrekshao@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
- Note that by default these are already errors, not warnings.
Change-Id: I4eadbb6f71b2d2a39efe5755dbb6b8c329f1674f
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128063
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
Reviewed-by: Shrek Shao <shrekshao@google.com>
- Note that by default these are already errors, not warnings.
Change-Id: I33202a0cca8167c4c79e4aeee680b6b7cf1ff830
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128062
Reviewed-by: Shrek Shao <shrekshao@google.com>
Commit-Queue: Loko Kung <lokokung@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This reverts commit 8cc6205bf7cfba2816052a0cc5d81debd8d52028.
Reason for revert: Graphite actually reuses the bind groups between
draw calls using different pipelines and this change prevents it
from happening.
Original change's description:
> Disable frontend cache when implicit device sync is on.
>
> Normal behavior of ApiObjectBase's APIRelease() which only locks the
> device when last ref dropped is not thread safe if the object is cached
> as raw pointers by the device. Example of cached objects: bind group
> layout, pipeline, sampler, shader module.
>
> The following scenario could happen:
> - thread A:
> - shaderModuleA.APIRealease()
> - shaderModuleA.refCount.Decrement() == true (ref count has reached zero)
> - going to call shaderModuleA.LockAndDeleteThis().
> - thread B:
> - device.CreateShaderModule().
> - lock()
> - device.GetOrCreateShaderModule()
> - shaderModuleA is in the cache, so return it.
> - unlock()
> - thread A:
> - starting to call shaderModuleA.LockAndDeleteThis()
> - lock()
> - erase shaderModuleA from the cache.
> - delete shaderModuleA.
> - unlock()
>
> This CL disables caching when ImplicitDeviceSynchronization is turned on
> until we find a better solution.
>
> Bug: dawn:1769
> Change-Id: Ideb2a717ece0a40e18bd1c2bef00817262bd25da
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127900
> Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
> Reviewed-by: Austin Eng <enga@chromium.org>
> Kokoro: Kokoro <noreply+kokoro@google.com>
TBR=cwallez@chromium.org,enga@chromium.org,noreply+kokoro@google.com,dawn-scoped@luci-project-accounts.iam.gserviceaccount.com,lehoangquyen@chromium.org
Change-Id: Ib13bba8005402d06963865fae919388a91e718f0
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: dawn:1769
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128440
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Normal behavior of ApiObjectBase's APIRelease() which only locks the
device when last ref dropped is not thread safe if the object is cached
as raw pointers by the device. Example of cached objects: bind group
layout, pipeline, sampler, shader module.
The following scenario could happen:
- thread A:
- shaderModuleA.APIRealease()
- shaderModuleA.refCount.Decrement() == true (ref count has reached zero)
- going to call shaderModuleA.LockAndDeleteThis().
- thread B:
- device.CreateShaderModule().
- lock()
- device.GetOrCreateShaderModule()
- shaderModuleA is in the cache, so return it.
- unlock()
- thread A:
- starting to call shaderModuleA.LockAndDeleteThis()
- lock()
- erase shaderModuleA from the cache.
- delete shaderModuleA.
- unlock()
This CL disables caching when ImplicitDeviceSynchronization is turned on
until we find a better solution.
Bug: dawn:1769
Change-Id: Ideb2a717ece0a40e18bd1c2bef00817262bd25da
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127900
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
This method will return supported usage flags that can be used to create
a swap chain.
Bug: dawn:1760
Change-Id: I7699c2c4ef7142c6bd06e72239d6e4f9112f15a3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127440
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
This CL consolidates the unicode code into utils. These are utility
libraries, so the extra namespace doesn't add much.
Change-Id: Id0de612b6be036392a3cb018bfe66733f2f1ebcb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127403
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Add alphaToCoverage validation regards to targets[0] has alpha channel.
This change reflecting WebGPU V1 spec update is aimed to ship together
with WebGPU in Chromium.
Bug: dawn:1759
Change-Id: I0aef60cf8c4dc828e05d6027644ffed35b33f652
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128061
Commit-Queue: Kai Ninomiya <kainino@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Auto-Submit: Shrek Shao <shrekshao@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Loko Kung <lokokung@google.com>
Don't skip the TruncateInterstageVariables transform when
user defined interstage attribute input for fragment stage
is empty. Because builtin inputs could also cause register
mismatch for D3D12 HLSL compiler.
Add a boolean flag to Tint hlsl generator option to indicate
whether to run TruncateInterstageVariables or not.
This defaults to false in Tint, while Dawn always set
this to true for vertex stage.
Bug: dawn:1733
Change-Id: Ie4c3648b226513bf15f0e03ae4ce7f3cc09fdef4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127206
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Shrek Shao <shrekshao@google.com>
The initial test tests bind group layout creation of
different sizes, cache / no-cache hit, with and without
multiple threads.
Change-Id: Ic9ed6c6f1c298d35cd1358c7ff492027c83649a7
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127346
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This CL enable build d3d11 by default, if dawn is not build with
chromium. d3d11 backend is not full implemented yet, so no tests
will run against d3d11 backend.
Bug: dawn:1705
Change-Id: Id689ab5168511af0f75f0b3537f246713de5ca45
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127260
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Peng Huang <penghuang@chromium.org>
This reverts commit 3fcf96dd8c262c69029a47ce6ad85314fcd3d69b.
Reason for revert: want to enable end2end test piece by piece on bots for d3d11
Original change's description:
> d3d11: add d3d11 backend in end2end tests
>
> Right now, many tests are not passed becasue unimplemented
> features in d3d11 backend. HoweverD3D11 backend is disabled on
> bots by default, so this CL will not break out bots.
>
> Bug: dawn:1705
> Change-Id: I57321b86a404bc245b71c467479fdee0464dee9b
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126260
> Commit-Queue: Peng Huang <penghuang@chromium.org>
> Reviewed-by: Austin Eng <enga@chromium.org>
> Kokoro: Kokoro <noreply+kokoro@google.com>
Bug: dawn:1705
Change-Id: I95a1cc9a0962b01a6b31ea32b6129f109f4b3e42
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127240
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Peng Huang <penghuang@chromium.org>
This option will allow tests to be run with
ImplicitiDeviceSynchronization feature.
Bug: dawn:1662
Change-Id: Ic001b2fa175f63e8d77eeb3b23d4d2cf52bb224e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126580
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
1. Auto-generate mutex locking code in DawnNative's ProcTable's
functions. Using a mutex owned by the related Device.
- Unless the function/class has "no autolock" attribute (new) in
dawn.json. In which cases, no locking code will be auto-generated.
- Currently Instance, Adapter, Surface, Encoder classes and
Reference/Release method have "no autolock".
2. Added Feature::ImplicitDeviceSynchronization to enable Device's
mutex.
- If this feature is disabled, lock/unlock Device's mutex is no-op.
Auto-generated locking code will have no effect. This is the default.
- This approach is used instead of generating two sets of ProcTable
because it's cleaner and the internal code doesn't need to care about
whether there is a mutex being locked or not. Furthermore, if there
were two sets of ProcTable, and user used dawnProcSetProcs() to set
global ProcTable, it would affect all other modules using different
Devices. Even though those modules don't need thread safety.
Bug: dawn:1662
Change-Id: I75f0d28959f333318e4159916b259131581f79f5
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/119940
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
With the wire, callbacks reference device-related data which is
freed when the device is unregistered from the wire. However,
the device that the wire refers to can live longer than this
and have its callbacks invoked at a later time, leading to a
use-after-free. Always flush the callbacks so that when the
wire unregisters a device and clears its callbacks, they are
called immediately and no dangling pointers remain.
Bug: chromium:1430221
Change-Id: Ib753b58cee5e7cb4d90ebd703958a0ddd5873573
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126481
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Loko Kung <lokokung@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Right now, many tests are not passed becasue unimplemented
features in d3d11 backend. HoweverD3D11 backend is disabled on
bots by default, so this CL will not break out bots.
Bug: dawn:1705
Change-Id: I57321b86a404bc245b71c467479fdee0464dee9b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126260
Commit-Queue: Peng Huang <penghuang@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
InstanceBase has two seperate ref counts: external & internal.
When we return it to the external user, we need to increase external
ref count as well.
Bug: chromium:1429835
Change-Id: I4b585d04149180418bd620e6e4d86522e033b64e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126080
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
CopyExternalTextureForBrowser() uses external texture visible rect
as source size in previous.
But video frame natural size is the only one developer could get
from HTMLVideoElement and it means the size browser present video
on screen.
This CL add "natural size" in imageCopyExternalTexture and uses this
size as CopyExternalTextureForBrowser() source size.
Bug:dawn:1694
Change-Id: I2a3bfa8e689df11d1d13320d40ad02c7090425e5
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/123380
Commit-Queue: Shaobo Yan <shaobo.yan@intel.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
When InstanceBase::ProcessEvents() iterates through list of devices, one
device might be being destructed on another thread. Even if we try to
increase ref count of that device inside the ProcessEvents(), the device
might be in the middle of destructor call on another thread, increasing
the ref count is invalid in this case.
This CL attempts to fix this issue by removing the device's pointer
from InstanceBase earlier: when DeviceBase::WillDropLastExternalRef()
is called. After this point, any callback registered to this device will
be forwarded to InstanceBase's callback queue instead.
Bug: dawn:752
Change-Id: I8ae86575e34f753e52a76f5fc774bbb5366a1b85
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/124281
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Reviewed-by: Loko Kung <lokokung@google.com>
This was fixed a while ago in
https://dawn-review.googlesource.com/c/dawn/+/48320
which filled the resolve buffer with 0 and only
resolved the available queries.
Bug: dawn:973
Change-Id: Icf6f83eeca65507ba3e6dda9e9f910d684465622
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/125140
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Kai Ninomiya <kainino@chromium.org>
Auto-Submit: Austin Eng <enga@chromium.org>
This CL enable using f16 override, and also fix related tests in Dawn
and Tint.
Bug: tint:1473, tint:1502
Change-Id: I8336770e8a73e5023c1aba224b7b5f21692fbaa6
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/124544
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Zhaoming Jiang <zhaoming.jiang@intel.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Fixes a bug where Dawn incorrectly did not re-apply state
when transitioning between compute and render passes. If
a compute and render pipeline share the same pipeline layout,
all of the resources for the graphics pipeline need to be rebound
since the graphics state in D3D12 is disjoint from the compute
state.
Fixed: dawn:1689
Change-Id: I7d25a1c7954039c4130e67b682ebc05324353e9a
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/124540
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Loko Kung <lokokung@google.com>
This CL make adapter base holds it supported features set as private
instead of protected and provide a method to set features enabled. This
CL also rename SetSupportedFeatures in null adapter to
SetSupportedFeaturesForTesting.
This is a pre-CL for implementing UseDXC as instance toggle, which may
require further refactor and adapter features logic to handle the toggles.
Bug: dawn:1495
Change-Id: I0a07e5653b43f18278cb4a2fe90985cc90b66068
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/124421
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Zhaoming Jiang <zhaoming.jiang@intel.com>
Reviewed-by: Austin Eng <enga@chromium.org>
In most backends, MultiplanarExternalTexture must run before the
BindingRemapper. The exception is the Vulkan SPIR-V backend where
it has to run before Multiplanar. This CL moves both of the transforms
to backend transforms and sets up the ordering as needed in the
generators.
Bug: tint:1855 chromium:1421379
Change-Id: I8223bce40babe901fe08bb1e8f033243837d7b18
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/123522
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Since https://dawn-review.googlesource.com/c/dawn/+/120940, callbacks
will be deferred to be executed in next device.APITick() instead of
immediately.
However, if the device is already destroyed (last ref dropped),
user/wire_server has no chance to call device.APITick() anymore, leading
to the callbacks waiting in queue forever.
This is also possibly the cause of memory leaks in cluserfuzz tests.
This CL attempt to fix it by implementing Instance::ProcessEvents():
In this method, every created device will invoke APITick() even if it is
already lost/externally released.
bug: chromium:1422507
bug: dawn:752
Change-Id: Iec69ad3b547a7e88c6e1a2225b13ad060a501a4f
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/123420
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
The Vulkan spec mandates the use of a fragment shader unless
rasterizer discard is enabled. For now, enable the placeholder
fragment shader in all cases, with a bug logged to optimize it
for the rasterizer discard case.
Bug: dawn:1696 dawn:1698
Change-Id: I9e85e6308a9952fc505382488c618897bd9abc7b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/123503
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
Bug: dawn:1696
Change-Id: Ie14e3d124342228edef17a122f380e2bdfaf1e57
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/123502
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
Auto-Submit: Stephen White <senorblanco@chromium.org>
Reviewed-by: dan sinclair <dsinclair@google.com>
- Adds an e2e test to test writing to outputs when at the limit.
Bug: dawn:1665
Change-Id: I2b2b9c2d700be0e454dc945ed8e3e1fe6b191974
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/122801
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Loko Kung <lokokung@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This replicates a subset of the flow control CTS tests. The purpose of
this is to provide relatively easy-to-repro cases to demonstrate flow
control bugs on Intel Graphics UHD 630 GPUs.
Bug: tint:1868
Change-Id: I34d692230b44d8a0a917dc773cc748bbf288d55a
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/122940
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
This CL moves the `using` statement for `BindingPoint` out to its own
file in `src/tint/writer/binding_point.h`. Usages of
`tint::transform::BindingPoint` have been updated to
`tint::writer::BindingPoint`.
Bug: tint:1855
Change-Id: Ife7d92807e0a65b93f21158d73a60d5068fc5da9
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/123000
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
- Suppresses maxStorageTexturesPerShaderStage CTS test because it needs
to be modified to adhere to the new limit as well.
Bug: dawn:1665
Change-Id: I66c62bd94b613059633888210ec7e7b42dc3a1dc
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/122461
Commit-Queue: Loko Kung <lokokung@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
This CL add Instance and Adapter toggle stage, and promote
DisallowUnsafeAPIs as an instance toggle, and can be required using
DawnTogglesDescriptor chained in instance descriptor when creating
instance. The instance's toggles state will get inherited to adapters
and devices it create. Related tests are implemented and updated.
Toggles inheritance can be overriden if not forced, so requiring
DisallowUnsafeAPIs when creating device is still available and working
like before.
Note that currently we don't have toggle of adapter stage, and can not
require toggles when creating adapter, until follow up CLs implement it.
Currently the toggles state of a adapter is simply inherited from
instance.
Bug: dawn:1495
Change-Id: I6bf7aa0f950a99451afcc2cab5322c924b7d9520
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/122021
Commit-Queue: Zhaoming Jiang <zhaoming.jiang@intel.com>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
There are some scenarios where buffers can be used in pending
commands that are not retained by a command buffer.
They must be retained in the set of mappable buffers for eager
transition to prevent a use-after-free violation.
Fixed: chromium:1421170
Change-Id: I452d80b2513a7726a003d44e2a7850292d798bb1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/122580
Auto-Submit: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Loko Kung <lokokung@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Peng Huang <penghuang@chromium.org>
Currently in the middle of some functions, we execute callbacks
immediately such as inside Buffer::APIMapAsync(), Device::HandleError()
or Queue::Submit().
Firstly, this has risks. The functions might be in a middle of modifying
internal states. By triggering callbacks, users might call API
functions again which could further modify the internal states
unexpectedly or access the states in an inconsistent way.
Secondly, upcoming thread safe API which locks the public functions with
a mutex might encounter deadlock. Because callbacks might cause
re-entrances which would unexpectedly lock the public function again.
This CL attempts to limit number of functions that are allowed to
trigger callbacks. Other functions that want to trigger callbacks will
instead enqueue a request to execute callbacks in the next
Device::APITick() call.
Currently the functions that will be allowed to trigger callbacks are:
- Device::WillDropLastExternalRef()
- Device::APITick()
- Device::APISetLoggingCallback()
- Device::APISetUncapturedErrorCallback()
- Device::APISetDeviceLostCallback()
Bug: dawn:1672
Change-Id: Iabca00f1b6f8f69eb5e966ffaa43dda5ae20fa8b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/120940
Commit-Queue: Quyen Le <lehoangquyen@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
