Remove the assumption about client_id from gpu memory buffer allocation on the browser side.

content/common/gpu/client/gpu_memory_buffer_impl_ozone.cc

Index: content/common/gpu/client/gpu_memory_buffer_impl_ozone.cc
diff --git a/content/common/gpu/client/gpu_memory_buffer_impl_ozone.cc b/content/common/gpu/client/gpu_memory_buffer_impl_ozone.cc
index 6f2ae5e36f4a324da5813ab76b12369a92047698..685dc7cc38d74c7f6c6050f7cb768f2261a0e440 100644
--- a/content/common/gpu/client/gpu_memory_buffer_impl_ozone.cc
+++ b/content/common/gpu/client/gpu_memory_buffer_impl_ozone.cc
@@ -13,11 +13,12 @@ namespace content {
 void GpuMemoryBufferImpl::Create(const gfx::Size& size,
                                  unsigned internalformat,
                                  unsigned usage,
+                                 int gpu_client_id,
                                  const CreationCallback& callback) {
   if (GpuMemoryBufferImplOzoneNativeBuffer::IsConfigurationSupported(
           internalformat, usage)) {
     GpuMemoryBufferImplOzoneNativeBuffer::Create(
-        size, internalformat, usage, callback);
+        size, internalformat, usage, gpu_client_id, callback);
     return;
   }