summaryrefslogtreecommitdiff
path: root/src/tuning/kernels
diff options
context:
space:
mode:
authorCedric Nugteren <web@cedricnugteren.nl>2016-10-03 19:32:01 +0200
committerCedric Nugteren <web@cedricnugteren.nl>2016-10-03 19:32:01 +0200
commitc1c4bc5d209280e4ec9be5c0a26f7c94077a6b20 (patch)
tree8f5781b17e2b2d7b96bf8d9e593e1e471fc67a30 /src/tuning/kernels
parent243cef73db53b5d8ce6c55f95e18c4412539d210 (diff)
Re-organised GEMM direct kernel and added faster fall-back version for incomplete rectangles
Diffstat (limited to 'src/tuning/kernels')
-rw-r--r--src/tuning/kernels/xgemm_direct.cpp1
1 files changed, 1 insertions, 0 deletions
diff --git a/src/tuning/kernels/xgemm_direct.cpp b/src/tuning/kernels/xgemm_direct.cpp
index c6948ef5..204e0be4 100644
--- a/src/tuning/kernels/xgemm_direct.cpp
+++ b/src/tuning/kernels/xgemm_direct.cpp
@@ -35,6 +35,7 @@ class TuneXgemmDirect {
#include "../src/kernels/common.opencl"
#include "../src/kernels/level3/xgemm_direct_part1.opencl"
#include "../src/kernels/level3/xgemm_direct_part2.opencl"
+ #include "../src/kernels/level3/xgemm_direct_part3.opencl"
;
}