Add wgpuDeviceGetLimits. Split Required/Supported limit structs

Bug: dawn:685
Change-Id: Ibb5dd0479f5e887d4b2ca864c014ebaafb674dba
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/64443
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>

src/dawn_wire/client/Device.cpp

@@ -196,6 +196,12 @@ namespace dawn_wire { namespace client {
         return Buffer::CreateError(this);
     }
 
+    bool Device::GetLimits(WGPUSupportedLimits* limits) {
+        // Not implemented in the wire.
+        UNREACHABLE();
+        return false;
+    }
+
     WGPUQueue Device::GetQueue() {
         // The queue is lazily created because if a Device is created by
         // Reserve/Inject, we cannot send the GetQueue message until

src/dawn_wire/client/Device.h

@@ -64,6 +64,7 @@ namespace dawn_wire { namespace client {
                                                  WGPUCreatePipelineAsyncStatus status,
                                                  const char* message);
 
+        bool GetLimits(WGPUSupportedLimits* limits);
         WGPUQueue GetQueue();
 
         void CancelCallbacksForDisconnect() override;