| // Copyright 2023 The Dawn & Tint Authors |
| // |
| // Redistribution and use in source and binary forms, with or without |
| // modification, are permitted provided that the following conditions are met: |
| // |
| // 1. Redistributions of source code must retain the above copyright notice, this |
| // list of conditions and the following disclaimer. |
| // |
| // 2. Redistributions in binary form must reproduce the above copyright notice, |
| // this list of conditions and the following disclaimer in the documentation |
| // and/or other materials provided with the distribution. |
| // |
| // 3. Neither the name of the copyright holder nor the names of its |
| // contributors may be used to endorse or promote products derived from |
| // this software without specific prior written permission. |
| // |
| // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" |
| // AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE |
| // DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE |
| // FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL |
| // DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR |
| // SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER |
| // CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, |
| // OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| |
| #include <atomic> |
| #include <cstdint> |
| #include <utility> |
| #include <vector> |
| |
| #include "dawn/common/FutureUtils.h" |
| #include "dawn/tests/DawnTest.h" |
| #include "dawn/webgpu.h" |
| #include "gmock/gmock.h" |
| |
| namespace dawn { |
| namespace { |
| |
| using testing::AnyOf; |
| using testing::Eq; |
| |
| wgpu::Device CreateExtraDevice(wgpu::Instance instance) { |
| // IMPORTANT: DawnTest overrides RequestAdapter and RequestDevice and mixes |
| // up the two instances. We use these to bypass the override. |
| auto* requestAdapter = reinterpret_cast<WGPUProcInstanceRequestAdapter>( |
| wgpuGetProcAddress(nullptr, "wgpuInstanceRequestAdapter")); |
| auto* requestDevice = reinterpret_cast<WGPUProcAdapterRequestDevice>( |
| wgpuGetProcAddress(nullptr, "wgpuAdapterRequestDevice")); |
| |
| wgpu::Adapter adapter2; |
| requestAdapter( |
| instance.Get(), nullptr, |
| [](WGPURequestAdapterStatus status, WGPUAdapter adapter, const char*, void* userdata) { |
| ASSERT_EQ(status, WGPURequestAdapterStatus_Success); |
| *reinterpret_cast<wgpu::Adapter*>(userdata) = wgpu::Adapter::Acquire(adapter); |
| }, |
| &adapter2); |
| DAWN_ASSERT(adapter2); |
| |
| wgpu::Device device2; |
| requestDevice( |
| adapter2.Get(), nullptr, |
| [](WGPURequestDeviceStatus status, WGPUDevice device, const char*, void* userdata) { |
| ASSERT_EQ(status, WGPURequestDeviceStatus_Success); |
| *reinterpret_cast<wgpu::Device*>(userdata) = wgpu::Device::Acquire(device); |
| }, |
| &device2); |
| DAWN_ASSERT(device2); |
| |
| return device2; |
| } |
| |
| std::pair<wgpu::Instance, wgpu::Device> CreateExtraInstance(wgpu::InstanceDescriptor* desc) { |
| wgpu::Instance instance2 = wgpu::CreateInstance(desc); |
| |
| wgpu::Device device2 = CreateExtraDevice(instance2); |
| DAWN_ASSERT(device2); |
| |
| return std::pair(std::move(instance2), std::move(device2)); |
| } |
| |
| // EventCompletionTests |
| |
| enum class WaitType { |
| TimedWaitAny, |
| SpinWaitAny, |
| SpinProcessEvents, |
| }; |
| |
| enum class WaitTypeAndCallbackMode { |
| TimedWaitAny_WaitAnyOnly, |
| TimedWaitAny_AllowSpontaneous, |
| SpinWaitAny_WaitAnyOnly, |
| SpinWaitAny_AllowSpontaneous, |
| SpinProcessEvents_AllowProcessEvents, |
| SpinProcessEvents_AllowSpontaneous, |
| }; |
| |
| std::ostream& operator<<(std::ostream& o, WaitTypeAndCallbackMode waitMode) { |
| switch (waitMode) { |
| case WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly: |
| return o << "TimedWaitAny_WaitAnyOnly"; |
| case WaitTypeAndCallbackMode::SpinWaitAny_WaitAnyOnly: |
| return o << "SpinWaitAny_WaitAnyOnly"; |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowProcessEvents: |
| return o << "SpinProcessEvents_AllowProcessEvents"; |
| case WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous: |
| return o << "TimedWaitAny_AllowSpontaneous"; |
| case WaitTypeAndCallbackMode::SpinWaitAny_AllowSpontaneous: |
| return o << "SpinWaitAny_AllowSpontaneous"; |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowSpontaneous: |
| return o << "SpinProcessEvents_AllowSpontaneous"; |
| } |
| } |
| |
| DAWN_TEST_PARAM_STRUCT(EventCompletionTestParams, WaitTypeAndCallbackMode); |
| |
| class EventCompletionTests : public DawnTestWithParams<EventCompletionTestParams> { |
| protected: |
| wgpu::Instance testInstance; |
| wgpu::Device testDevice; |
| wgpu::Queue testQueue; |
| std::vector<wgpu::FutureWaitInfo> mFutures; |
| std::atomic<uint64_t> mCallbacksCompletedCount = 0; |
| uint64_t mCallbacksIssuedCount = 0; |
| uint64_t mCallbacksWaitedCount = 0; |
| |
| void SetUp() override { |
| DawnTestWithParams::SetUp(); |
| WaitTypeAndCallbackMode mode = GetParam().mWaitTypeAndCallbackMode; |
| if (UsesWire()) { |
| DAWN_TEST_UNSUPPORTED_IF(mode == WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly || |
| mode == |
| WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous); |
| } |
| testInstance = GetInstance(); |
| testDevice = device; |
| testQueue = queue; |
| // Make sure these aren't used accidentally (unfortunately can't do the same for instance): |
| device = nullptr; |
| queue = nullptr; |
| } |
| |
| void UseSecondInstance() { |
| wgpu::InstanceDescriptor desc; |
| desc.features.timedWaitAnyEnable = !UsesWire(); |
| std::tie(testInstance, testDevice) = CreateExtraInstance(&desc); |
| testQueue = testDevice.GetQueue(); |
| } |
| |
| void LoseTestDevice() { |
| EXPECT_CALL(mDeviceLostCallback, |
| Call(testing::_, WGPUDeviceLostReason_Unknown, testing::_, testing::_)) |
| .Times(1); |
| testDevice.ForceLoss(wgpu::DeviceLostReason::Unknown, "Device lost for testing"); |
| testInstance.ProcessEvents(); |
| } |
| |
| void TrivialSubmit() { |
| wgpu::CommandBuffer cb = testDevice.CreateCommandEncoder().Finish(); |
| testQueue.Submit(1, &cb); |
| } |
| |
| wgpu::CallbackMode GetCallbackMode() { |
| switch (GetParam().mWaitTypeAndCallbackMode) { |
| case WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::SpinWaitAny_WaitAnyOnly: |
| return wgpu::CallbackMode::WaitAnyOnly; |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowProcessEvents: |
| return wgpu::CallbackMode::AllowProcessEvents; |
| case WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous: |
| case WaitTypeAndCallbackMode::SpinWaitAny_AllowSpontaneous: |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowSpontaneous: |
| return wgpu::CallbackMode::AllowSpontaneous; |
| } |
| } |
| |
| bool IsSpontaneous() { return GetCallbackMode() == wgpu::CallbackMode::AllowSpontaneous; } |
| |
| void TrackForTest(wgpu::Future future) { |
| mCallbacksIssuedCount++; |
| |
| switch (GetParam().mWaitTypeAndCallbackMode) { |
| case WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous: |
| case WaitTypeAndCallbackMode::SpinWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::SpinWaitAny_AllowSpontaneous: |
| mFutures.push_back(wgpu::FutureWaitInfo{future, false}); |
| break; |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowProcessEvents: |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowSpontaneous: |
| break; |
| } |
| } |
| |
| wgpu::Future OnSubmittedWorkDone(WGPUQueueWorkDoneStatus expectedStatus) { |
| struct Userdata { |
| EventCompletionTests* self; |
| WGPUQueueWorkDoneStatus expectedStatus; |
| }; |
| Userdata* userdata = new Userdata{this, expectedStatus}; |
| |
| return testQueue.OnSubmittedWorkDone({ |
| nullptr, |
| GetCallbackMode(), |
| [](WGPUQueueWorkDoneStatus status, void* userdata) { |
| Userdata* u = reinterpret_cast<Userdata*>(userdata); |
| u->self->mCallbacksCompletedCount++; |
| ASSERT_EQ(status, u->expectedStatus); |
| delete u; |
| }, |
| userdata, |
| }); |
| } |
| |
| void TestWaitAll(bool loopOnlyOnce = false) { |
| switch (GetParam().mWaitTypeAndCallbackMode) { |
| case WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous: |
| return TestWaitImpl(WaitType::TimedWaitAny, loopOnlyOnce); |
| case WaitTypeAndCallbackMode::SpinWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::SpinWaitAny_AllowSpontaneous: |
| return TestWaitImpl(WaitType::SpinWaitAny, loopOnlyOnce); |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowProcessEvents: |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowSpontaneous: |
| return TestWaitImpl(WaitType::SpinProcessEvents, loopOnlyOnce); |
| } |
| } |
| |
| void TestWaitIncorrectly() { |
| switch (GetParam().mWaitTypeAndCallbackMode) { |
| case WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous: |
| case WaitTypeAndCallbackMode::SpinWaitAny_WaitAnyOnly: |
| case WaitTypeAndCallbackMode::SpinWaitAny_AllowSpontaneous: |
| return TestWaitImpl(WaitType::SpinProcessEvents); |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowProcessEvents: |
| case WaitTypeAndCallbackMode::SpinProcessEvents_AllowSpontaneous: |
| return TestWaitImpl(WaitType::SpinWaitAny); |
| } |
| } |
| |
| private: |
| void TestWaitImpl(WaitType waitType, bool loopOnlyOnce = false) { |
| uint64_t oldCompletedCount = mCallbacksCompletedCount; |
| |
| const auto start = std::chrono::high_resolution_clock::now(); |
| auto testTimeExceeded = [=]() -> bool { |
| return std::chrono::high_resolution_clock::now() - start > std::chrono::seconds(5); |
| }; |
| |
| switch (waitType) { |
| case WaitType::TimedWaitAny: { |
| bool emptyWait = mFutures.size() == 0; |
| // Loop at least once so we can test it with 0 futures. |
| do { |
| ASSERT_FALSE(testTimeExceeded()); |
| DAWN_ASSERT(!UsesWire()); |
| wgpu::WaitStatus status; |
| |
| uint64_t oldCompletionCount = mCallbacksCompletedCount; |
| // Any futures should succeed within a few milliseconds at most. |
| status = testInstance.WaitAny(mFutures.size(), mFutures.data(), UINT64_MAX); |
| ASSERT_EQ(status, wgpu::WaitStatus::Success); |
| bool mayHaveCompletedEarly = IsSpontaneous(); |
| if (!mayHaveCompletedEarly && !emptyWait) { |
| ASSERT_GT(mCallbacksCompletedCount, oldCompletionCount); |
| } |
| |
| // Verify this succeeds instantly because some futures completed already. |
| status = testInstance.WaitAny(mFutures.size(), mFutures.data(), 0); |
| ASSERT_EQ(status, wgpu::WaitStatus::Success); |
| |
| RemoveCompletedFutures(); |
| if (loopOnlyOnce) { |
| break; |
| } |
| } while (mFutures.size() > 0); |
| } break; |
| case WaitType::SpinWaitAny: { |
| bool emptyWait = mFutures.size() == 0; |
| // Loop at least once so we can test it with 0 futures. |
| do { |
| ASSERT_FALSE(testTimeExceeded()); |
| |
| uint64_t oldCompletionCount = mCallbacksCompletedCount; |
| FlushWire(); |
| auto status = testInstance.WaitAny(mFutures.size(), mFutures.data(), 0); |
| if (status == wgpu::WaitStatus::TimedOut) { |
| continue; |
| } |
| ASSERT_TRUE(status == wgpu::WaitStatus::Success); |
| bool mayHaveCompletedEarly = IsSpontaneous(); |
| if (!mayHaveCompletedEarly && !emptyWait) { |
| ASSERT_GT(mCallbacksCompletedCount, oldCompletionCount); |
| } |
| |
| RemoveCompletedFutures(); |
| if (loopOnlyOnce) { |
| break; |
| } |
| } while (mFutures.size() > 0); |
| } break; |
| case WaitType::SpinProcessEvents: { |
| do { |
| ASSERT_FALSE(testTimeExceeded()); |
| |
| FlushWire(); |
| testInstance.ProcessEvents(); |
| |
| if (loopOnlyOnce) { |
| break; |
| } |
| } while (mCallbacksCompletedCount < mCallbacksIssuedCount); |
| } break; |
| } |
| |
| if (!IsSpontaneous()) { |
| ASSERT_EQ(mCallbacksCompletedCount - oldCompletedCount, |
| mCallbacksIssuedCount - mCallbacksWaitedCount); |
| } |
| ASSERT_EQ(mCallbacksCompletedCount, mCallbacksIssuedCount); |
| mCallbacksWaitedCount = mCallbacksCompletedCount; |
| } |
| |
| void RemoveCompletedFutures() { |
| size_t oldSize = mFutures.size(); |
| if (oldSize > 0) { |
| mFutures.erase( |
| std::remove_if(mFutures.begin(), mFutures.end(), |
| [](const wgpu::FutureWaitInfo& info) { return info.completed; }), |
| mFutures.end()); |
| ASSERT_LT(mFutures.size(), oldSize); |
| } |
| } |
| }; |
| |
| // Wait when no events have been requested. |
| TEST_P(EventCompletionTests, NoEvents) { |
| TestWaitAll(); |
| } |
| |
| // WorkDone event after submitting some trivial work. |
| TEST_P(EventCompletionTests, WorkDoneSimple) { |
| TrivialSubmit(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| } |
| |
| // WorkDone event before device loss, wait afterward. |
| TEST_P(EventCompletionTests, WorkDoneAcrossDeviceLoss) { |
| TrivialSubmit(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| } |
| |
| // WorkDone event after device loss. |
| TEST_P(EventCompletionTests, WorkDoneAfterDeviceLoss) { |
| TrivialSubmit(); |
| LoseTestDevice(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| } |
| |
| // WorkDone event twice after submitting some trivial work. |
| TEST_P(EventCompletionTests, WorkDoneTwice) { |
| TrivialSubmit(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| } |
| |
| // WorkDone event without ever having submitted any work. |
| TEST_P(EventCompletionTests, WorkDoneNoWork) { |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| } |
| |
| // WorkDone event after all work has completed already. |
| TEST_P(EventCompletionTests, WorkDoneAlreadyCompleted) { |
| TrivialSubmit(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| TrackForTest(OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success)); |
| TestWaitAll(); |
| } |
| |
| // WorkDone events waited in reverse order. |
| TEST_P(EventCompletionTests, WorkDoneOutOfOrder) { |
| // With ProcessEvents or Spontaneous we can't control the order of completion. |
| DAWN_TEST_UNSUPPORTED_IF(GetCallbackMode() != wgpu::CallbackMode::WaitAnyOnly); |
| |
| TrivialSubmit(); |
| wgpu::Future f1 = OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success); |
| TrivialSubmit(); |
| wgpu::Future f2 = OnSubmittedWorkDone(WGPUQueueWorkDoneStatus_Success); |
| |
| // When using WaitAny, normally callback ordering guarantees would guarantee f1 completes before |
| // f2. But if we wait on f2 first, then f2 is allowed to complete first because f1 still hasn't |
| // had an opportunity to complete. |
| TrackForTest(f2); |
| TestWaitAll(); |
| TrackForTest(f1); |
| TestWaitAll(/*loopOnlyOnce=*/true); |
| } |
| |
| constexpr WGPUQueueWorkDoneStatus kStatusUninitialized = |
| static_cast<WGPUQueueWorkDoneStatus>(INT32_MAX); |
| |
| TEST_P(EventCompletionTests, WorkDoneDropInstanceBeforeEvent) { |
| // TODO(crbug.com/dawn/1987): Wire does not implement instance destruction correctly yet. |
| DAWN_TEST_UNSUPPORTED_IF(UsesWire()); |
| |
| UseSecondInstance(); |
| testInstance = nullptr; // Drop the last external ref to the instance. |
| |
| WGPUQueueWorkDoneStatus status = kStatusUninitialized; |
| testQueue.OnSubmittedWorkDone({nullptr, GetCallbackMode(), |
| [](WGPUQueueWorkDoneStatus status, void* userdata) { |
| *reinterpret_cast<WGPUQueueWorkDoneStatus*>(userdata) = |
| status; |
| }, |
| &status}); |
| |
| if (IsSpontaneous()) { |
| // TODO(crbug.com/dawn/2059): Once Spontaneous is implemented, this should no longer expect |
| // the callback to be cleaned up immediately (and should expect it to happen on a future |
| // Tick). |
| ASSERT_THAT(status, AnyOf(Eq(WGPUQueueWorkDoneStatus_Success), |
| Eq(WGPUQueueWorkDoneStatus_InstanceDropped))); |
| } else { |
| ASSERT_EQ(status, WGPUQueueWorkDoneStatus_InstanceDropped); |
| } |
| } |
| |
| TEST_P(EventCompletionTests, WorkDoneDropInstanceAfterEvent) { |
| // TODO(crbug.com/dawn/1987): Wire does not implement instance destruction correctly yet. |
| DAWN_TEST_UNSUPPORTED_IF(UsesWire()); |
| |
| UseSecondInstance(); |
| |
| WGPUQueueWorkDoneStatus status = kStatusUninitialized; |
| testQueue.OnSubmittedWorkDone({nullptr, GetCallbackMode(), |
| [](WGPUQueueWorkDoneStatus status, void* userdata) { |
| *reinterpret_cast<WGPUQueueWorkDoneStatus*>(userdata) = |
| status; |
| }, |
| &status}); |
| |
| if (IsSpontaneous()) { |
| testInstance = nullptr; // Drop the last external ref to the instance. |
| |
| // TODO(crbug.com/dawn/2059): Once Spontaneous is implemented, this should no longer expect |
| // the callback to be cleaned up immediately (and should expect it to happen on a future |
| // Tick). |
| ASSERT_THAT(status, AnyOf(Eq(WGPUQueueWorkDoneStatus_Success), |
| Eq(WGPUQueueWorkDoneStatus_InstanceDropped))); |
| } else { |
| ASSERT_EQ(status, kStatusUninitialized); |
| testInstance = nullptr; // Drop the last external ref to the instance. |
| ASSERT_EQ(status, WGPUQueueWorkDoneStatus_InstanceDropped); |
| } |
| } |
| |
| // TODO(crbug.com/dawn/1987): |
| // - Test any reentrancy guarantees (for ProcessEvents or WaitAny inside a callback), |
| // to make sure things don't blow up and we don't attempt to hold locks recursively. |
| // - Other tests? |
| |
| DAWN_INSTANTIATE_TEST_P(EventCompletionTests, |
| {D3D11Backend(), D3D11Backend({"d3d11_use_unmonitored_fence"}), |
| D3D12Backend(), MetalBackend(), VulkanBackend(), OpenGLBackend(), |
| OpenGLESBackend()}, |
| { |
| WaitTypeAndCallbackMode::TimedWaitAny_WaitAnyOnly, |
| WaitTypeAndCallbackMode::TimedWaitAny_AllowSpontaneous, |
| WaitTypeAndCallbackMode::SpinWaitAny_WaitAnyOnly, |
| WaitTypeAndCallbackMode::SpinWaitAny_AllowSpontaneous, |
| WaitTypeAndCallbackMode::SpinProcessEvents_AllowProcessEvents, |
| WaitTypeAndCallbackMode::SpinProcessEvents_AllowSpontaneous, |
| |
| // TODO(crbug.com/dawn/2059): The cases with the Spontaneous flag |
| // enabled were added before we implemented all of the spontaneous |
| // completions. They might accidentally be overly strict. |
| |
| // TODO(crbug.com/dawn/2059): Make guarantees that Spontaneous callbacks |
| // get called (as long as you're hitting "checkpoints"), and add the |
| // corresponding tests, for example: |
| // - SpinProcessEvents_Spontaneous, |
| // - SpinSubmit_Spontaneous, |
| // - SpinCheckpoint_Spontaneous (if wgpuDeviceCheckpoint is added). |
| // - (Note we don't want to guarantee Tick will process events - we |
| // could even test that it doesn't, if we make that true.) |
| }); |
| |
| // WaitAnyTests |
| |
| class WaitAnyTests : public DawnTest {}; |
| |
| TEST_P(WaitAnyTests, UnsupportedTimeout) { |
| wgpu::Instance instance2; |
| wgpu::Device device2; |
| |
| if (UsesWire()) { |
| // The wire (currently) never supports timedWaitAnyEnable, so we can run this test on the |
| // default instance/device. |
| instance2 = GetInstance(); |
| device2 = device; |
| } else { |
| // When not using the wire, DawnTest will unconditionally set timedWaitAnyEnable since it's |
| // useful for other tests. For this test, we need it to be false to test validation. |
| wgpu::InstanceDescriptor desc; |
| desc.features.timedWaitAnyEnable = false; |
| std::tie(instance2, device2) = CreateExtraInstance(&desc); |
| } |
| |
| // UnsupportedTimeout is still validated if no futures are passed. |
| for (uint64_t timeout : {uint64_t(1), uint64_t(0), UINT64_MAX}) { |
| ASSERT_EQ(instance2.WaitAny(0, nullptr, timeout), |
| timeout > 0 ? wgpu::WaitStatus::UnsupportedTimeout : wgpu::WaitStatus::Success); |
| } |
| |
| for (uint64_t timeout : {uint64_t(1), uint64_t(0), UINT64_MAX}) { |
| wgpu::FutureWaitInfo info{device2.GetQueue().OnSubmittedWorkDone( |
| {nullptr, wgpu::CallbackMode::WaitAnyOnly, [](WGPUQueueWorkDoneStatus, void*) {}, |
| nullptr})}; |
| wgpu::WaitStatus status = instance2.WaitAny(1, &info, timeout); |
| if (timeout == 0) { |
| ASSERT_TRUE(status == wgpu::WaitStatus::Success || |
| status == wgpu::WaitStatus::TimedOut); |
| } else { |
| ASSERT_EQ(status, wgpu::WaitStatus::UnsupportedTimeout); |
| } |
| } |
| } |
| |
| TEST_P(WaitAnyTests, UnsupportedCount) { |
| wgpu::Instance instance2; |
| wgpu::Device device2; |
| wgpu::Queue queue2; |
| |
| if (UsesWire()) { |
| // The wire (currently) never supports timedWaitAnyEnable, so we can run this test on the |
| // default instance/device. |
| instance2 = GetInstance(); |
| device2 = device; |
| queue2 = queue; |
| } else { |
| wgpu::InstanceDescriptor desc; |
| desc.features.timedWaitAnyEnable = true; |
| std::tie(instance2, device2) = CreateExtraInstance(&desc); |
| queue2 = device2.GetQueue(); |
| } |
| |
| for (uint64_t timeout : {uint64_t(0), uint64_t(1)}) { |
| // We don't support values higher than the default (64), and if you ask for lower than 64 |
| // you still get 64. DawnTest doesn't request anything (so requests 0) so gets 64. |
| for (size_t count : {kTimedWaitAnyMaxCountDefault, kTimedWaitAnyMaxCountDefault + 1}) { |
| std::vector<wgpu::FutureWaitInfo> infos; |
| for (size_t i = 0; i < count; ++i) { |
| infos.push_back( |
| {queue2.OnSubmittedWorkDone({nullptr, wgpu::CallbackMode::WaitAnyOnly, |
| [](WGPUQueueWorkDoneStatus, void*) {}, nullptr})}); |
| } |
| wgpu::WaitStatus status = instance2.WaitAny(infos.size(), infos.data(), timeout); |
| if (timeout == 0) { |
| ASSERT_TRUE(status == wgpu::WaitStatus::Success || |
| status == wgpu::WaitStatus::TimedOut); |
| } else if (UsesWire()) { |
| // Wire doesn't support timeouts at all. |
| ASSERT_EQ(status, wgpu::WaitStatus::UnsupportedTimeout); |
| } else if (count <= 64) { |
| ASSERT_EQ(status, wgpu::WaitStatus::Success); |
| } else { |
| ASSERT_EQ(status, wgpu::WaitStatus::UnsupportedCount); |
| } |
| } |
| } |
| } |
| |
| TEST_P(WaitAnyTests, UnsupportedMixedSources) { |
| wgpu::Instance instance2; |
| wgpu::Device device2; |
| wgpu::Queue queue2; |
| wgpu::Device device3; |
| wgpu::Queue queue3; |
| |
| if (UsesWire()) { |
| // The wire (currently) never supports timedWaitAnyEnable, so we can run this test on the |
| // default instance/device. |
| instance2 = GetInstance(); |
| device2 = device; |
| queue2 = queue; |
| device3 = CreateDevice(); |
| queue3 = device3.GetQueue(); |
| } else { |
| wgpu::InstanceDescriptor desc; |
| desc.features.timedWaitAnyEnable = true; |
| std::tie(instance2, device2) = CreateExtraInstance(&desc); |
| queue2 = device2.GetQueue(); |
| device3 = CreateExtraDevice(instance2); |
| queue3 = device3.GetQueue(); |
| } |
| |
| for (uint64_t timeout : {uint64_t(0), uint64_t(1)}) { |
| std::vector<wgpu::FutureWaitInfo> infos{{ |
| {queue2.OnSubmittedWorkDone({nullptr, wgpu::CallbackMode::WaitAnyOnly, |
| [](WGPUQueueWorkDoneStatus, void*) {}, nullptr})}, |
| {queue3.OnSubmittedWorkDone({nullptr, wgpu::CallbackMode::WaitAnyOnly, |
| [](WGPUQueueWorkDoneStatus, void*) {}, nullptr})}, |
| }}; |
| wgpu::WaitStatus status = instance2.WaitAny(infos.size(), infos.data(), timeout); |
| if (timeout == 0) { |
| ASSERT_TRUE(status == wgpu::WaitStatus::Success || |
| status == wgpu::WaitStatus::TimedOut); |
| } else if (UsesWire()) { |
| // Wire doesn't support timeouts at all. |
| ASSERT_EQ(status, wgpu::WaitStatus::UnsupportedTimeout); |
| } else { |
| ASSERT_EQ(status, wgpu::WaitStatus::UnsupportedMixedSources); |
| } |
| } |
| } |
| |
| DAWN_INSTANTIATE_TEST(WaitAnyTests, |
| D3D11Backend(), |
| D3D11Backend({"d3d11_use_unmonitored_fence"}), |
| D3D12Backend(), |
| MetalBackend(), |
| VulkanBackend(), |
| OpenGLBackend(), |
| OpenGLESBackend()); |
| |
| class FutureTests : public DawnTest {}; |
| |
| // Regression test for crbug.com/dawn/2460 where when we have mixed source futures in a process |
| // events call we were crashing. |
| TEST_P(FutureTests, MixedSourcePolling) { |
| // OnSubmittedWorkDone is implemented via a queue serial. |
| device.GetQueue().OnSubmittedWorkDone({nullptr, wgpu::CallbackMode::AllowProcessEvents, |
| [](WGPUQueueWorkDoneStatus, void*) {}, nullptr}); |
| |
| // PopErrorScope is implemented via a signal. |
| device.PushErrorScope(wgpu::ErrorFilter::Validation); |
| device.PopErrorScope(wgpu::CallbackMode::AllowProcessEvents, |
| [](wgpu::PopErrorScopeStatus, wgpu::ErrorType, const char*) {}); |
| |
| instance.ProcessEvents(); |
| } |
| |
| DAWN_INSTANTIATE_TEST(FutureTests, |
| D3D11Backend(), |
| D3D11Backend({"d3d11_use_unmonitored_fence"}), |
| D3D12Backend(), |
| MetalBackend(), |
| VulkanBackend(), |
| OpenGLBackend(), |
| OpenGLESBackend()); |
| |
| } // anonymous namespace |
| } // namespace dawn |