Cycles: add HIP device support for AMD GPUs

NOTE: this feature is not ready for user testing, and not yet enabled in daily
builds. It is being merged now for easier collaboration on development.

HIP is a heterogenous compute interface allowing C++ code to be executed on
GPUs similar to CUDA. It is intended to bring back AMD GPU rendering support
on Windows and Linux.

https://github.com/ROCm-Developer-Tools/HIP.

As of the time of writing, it should compile and run on Linux with existing
HIP compilers and driver runtimes. Publicly available compilers and drivers
for Windows will come later.

See task T91571 for more details on the current status and work remaining
to be done.

Credits:

Sayak Biswas (AMD)
Arya Rafii (AMD)
Brian Savery (AMD)

Differential Revision: https://developer.blender.org/D12578

1 files changed, 68 insertions, 0 deletions

diff --git a/intern/cycles/device/hip/queue.h b/intern/cycles/device/hip/queue.h
new file mode 100644
index 00000000000..04c8a5982ce
--- /dev/null
+++ b/intern/cycles/device/hip/queue.h
@@ -0,0 +1,68 @@
+/*
+ * Copyright 2011-2021 Blender Foundation
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#pragma once
+
+#ifdef WITH_HIP
+
+#  include "device/device_kernel.h"
+#  include "device/device_memory.h"
+#  include "device/device_queue.h"
+
+#  include "device/hip/util.h"
+
+CCL_NAMESPACE_BEGIN
+
+class HIPDevice;
+class device_memory;
+
+/* Base class for HIP queues. */
+class HIPDeviceQueue : public DeviceQueue {
+ public:
+  HIPDeviceQueue(HIPDevice *device);
+  ~HIPDeviceQueue();
+
+  virtual int num_concurrent_states(const size_t state_size) const override;
+  virtual int num_concurrent_busy_states() const override;
+
+  virtual void init_execution() override;
+
+  virtual bool kernel_available(DeviceKernel kernel) const override;
+
+  virtual bool enqueue(DeviceKernel kernel, const int work_size, void *args[]) override;
+
+  virtual bool synchronize() override;
+
+  virtual void zero_to_device(device_memory &mem) override;
+  virtual void copy_to_device(device_memory &mem) override;
+  virtual void copy_from_device(device_memory &mem) override;
+
+  virtual hipStream_t stream()
+  {
+    return hip_stream_;
+  }
+
+  // TODO : (Arya) Enable this after stabilizing the dev branch
+  virtual unique_ptr<DeviceGraphicsInterop> graphics_interop_create() override;
+
+ protected:
+  HIPDevice *hip_device_;
+  hipStream_t hip_stream_;
+};
+
+CCL_NAMESPACE_END
+
+#endif /* WITH_HIP */