Cycles: Add support for P2P memory distribution (e.g. via NVLink)

This change modifies the multi-device implementation to support memory distribution
across devices, to reduce the overall memory footprint of large scenes and allow scenes to
fit entirely into combined GPU memory that previously had to fall back to host memory.

Reviewed By: brecht

Differential Revision: https://developer.blender.org/D7426

1 files changed, 2 insertions, 1 deletions

diff --git a/intern/cycles/blender/blender_python.cpp b/intern/cycles/blender/blender_python.cpp
index 79c16856462..0be19dbffd1 100644
--- a/intern/cycles/blender/blender_python.cpp
+++ b/intern/cycles/blender/blender_python.cpp
@@ -416,10 +416,11 @@ static PyObject *available_devices_func(PyObject * /*self*/, PyObject *args)
   for (size_t i = 0; i < devices.size(); i++) {
     DeviceInfo &device = devices[i];
     string type_name = Device::string_from_type(device.type);
-    PyObject *device_tuple = PyTuple_New(3);
+    PyObject *device_tuple = PyTuple_New(4);
     PyTuple_SET_ITEM(device_tuple, 0, pyunicode_from_string(device.description.c_str()));
     PyTuple_SET_ITEM(device_tuple, 1, pyunicode_from_string(type_name.c_str()));
     PyTuple_SET_ITEM(device_tuple, 2, pyunicode_from_string(device.id.c_str()));
+    PyTuple_SET_ITEM(device_tuple, 3, PyBool_FromLong(device.has_peer_memory));
     PyTuple_SET_ITEM(ret, i, device_tuple);
   }