[Bf-blender-cvs] [f1aeb2c] master: this is an attempted Fix: T38679 Cycles GPU Performance Regression

Martijn Berger noreply at git.blender.org
Mon Feb 17 20:13:26 CET 2014


Commit: f1aeb2ccf4deb4625f7ae59d673317d09b9580ca
Author: Martijn Berger
Date:   Mon Feb 17 20:11:45 2014 +0100
https://developer.blender.org/rBf1aeb2ccf4deb4625f7ae59d673317d09b9580ca

this is an attempted Fix: T38679
Cycles GPU Performance Regression

>From my testing this (what i should have done in the first place) reduces the regression a lot.
Lets hope it is enough or we have to go back to busy waiting.

===================================================================

M	intern/cycles/device/device_cuda.cpp

===================================================================

diff --git a/intern/cycles/device/device_cuda.cpp b/intern/cycles/device/device_cuda.cpp
index 3073f07..fa65aef 100644
--- a/intern/cycles/device/device_cuda.cpp
+++ b/intern/cycles/device/device_cuda.cpp
@@ -654,9 +654,6 @@ public:
 		cuda_assert(cuFuncSetBlockShape(cuPathTrace, xthreads, ythreads, 1))
 		cuda_assert(cuLaunchGridAsync(cuPathTrace, xblocks, yblocks, cuStream))
 
-		cuda_assert(cuEventRecord(tileDone, cuStream ))
-		cuda_assert(cuEventSynchronize(tileDone))
-
 		cuda_pop_context();
 	}
 
@@ -989,6 +986,8 @@ public:
 
 					task->update_progress(tile);
 				}
+				cuda_assert(cuEventRecord(tileDone, cuStream ))
+				cuda_assert(cuEventSynchronize(tileDone))
 
 				task->release_tile(tile);
 			}




More information about the Bf-blender-cvs mailing list