dawn/node: Fix crash when using device after destroy()

Do not assign null to the wgpu::Device, as this will result in a nullptr dereference if you attempt to use the device after destruction.
The wgpu device correctly handles the error state of being used after destroy().

Fixes crashes of CTS tests:
'webgpu:api,validation,state,device_lost,destroy,*'

Change-Id: Ibee1078436efadf25f53735fbaa47d5fc5f74898
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/97442
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>

src/dawn/node/binding/GPUDevice.cpp

@@ -172,7 +172,7 @@ GPUDevice::~GPUDevice() {
     // Without this, we'll get a 'Promise not resolved or rejected' fatal message as the
     // lost_promise_ is left hanging. We'll also not clean up any GPU objects before terminating the
     // process, which is not a good idea.
-    if (device_) {
+    if (!destroyed_) {
         destroy(env_);
     }
 }
@@ -204,7 +204,7 @@ void GPUDevice::destroy(Napi::Env env) {
             env_, interop::GPUDeviceLostReason::kDestroyed, "device was destroyed"));
     }
     device_.Destroy();
-    device_ = nullptr;
+    destroyed_ = true;
 }
 
 interop::Interface<interop::GPUBuffer> GPUDevice::createBuffer(

src/dawn/node/binding/GPUDevice.h

@@ -111,6 +111,8 @@ class GPUDevice final : public interop::GPUDevice {
     // This promise's JS object lives as long as the device because it is stored in .lost
     // of the wrapper JS object.
     interop::Promise<interop::Interface<interop::GPUDeviceLostInfo>> lost_promise_;
+
+    bool destroyed_ = false;
 };
 
 }  // namespace wgpu::binding