Suppress error generation in create*PipelineAsync, clarify async steps by kainino0x · Pull Request #4752 · gpuweb/gpuweb
Captures errors from when `create*PipelineAsync` call into `create*Pipeline`. Also improves the definitions of all device-timeline-async operations: - GPUDevice.createComputePipelineAsync() - GPUDevice.createRenderPipelineAsync() - GPUShaderModule.getCompilationInfo() - GPUQueue.onSubmittedWorkDone() - GPUBuffer.mapAsync() (GPUDevice.popErrorScope() returns a promise, but its device-timeline steps are synchronous.) Improves some language in "Errors & Debugging" and around the device-timeline "device becomes lost" event hooking. Fixes 4714 Fixes 4742