diff options
-rw-r--r-- | CHANGELOG | 1 | ||||
-rw-r--r-- | CMakeLists.txt | 5 | ||||
-rw-r--r-- | README.md | 1 |
3 files changed, 4 insertions, 3 deletions
@@ -8,6 +8,7 @@ Development version (next release) - Fixed a performance issue (caused by fp16 support) by optimizing alpha/beta parameter passing to kernels - Added an option (-warm_up) to do a warm-up run before timing in the performance clients - Improved performance significantly of rotated GEMV computations +- Various minor fixes and enhancements - Added tuned parameters for various devices (see README) Version 0.8.0 diff --git a/CMakeLists.txt b/CMakeLists.txt index 231b8e95..7393c6e7 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -230,8 +230,6 @@ endif() # the CLTune library (not included as part of the source). if(TUNERS) - # Includes CLTune - # Visual Studio requires the sources of non-exported objects/libraries set(TUNERS_COMMON ) if(MSVC) @@ -301,7 +299,8 @@ if(CLIENTS) else() # Creates the common performance-tests objects (requires CMake 2.8.8) add_library(test_performance_common OBJECT test/performance/client.cpp) - # Adds clblast's interface include pathes because we can't link to clblast here + + # Adds CLBlast's interface include paths because we can't link to CLBlast here target_include_directories(test_performance_common PRIVATE $<TARGET_PROPERTY:clblast,INTERFACE_INCLUDE_DIRECTORIES> ${clblast_SOURCE_DIR}) @@ -286,6 +286,7 @@ The contributing authors (code, pull requests, testing) so far are: * [Marco Hutter](https://github.com/gpus) * [Hugh Perkins](https://github.com/hughperkins) * [Gian-Carlo Pascutto](https://github.com/gcp) +* [Ivan Shapovalov](https://github.com/intelfx) Tuning and testing on a variety of OpenCL devices was made possible by: |