Extended subgroups - use 128bit masks (#1215)

* Extended subgroups - use 128bit masks

* Refactoring to avoid kernels code duplication

* unification kernel names as test_ prefix +subgroups function name
* use string literals that improve readability
* use kernel templates that limit code duplication
* WorkGroupParams allows define default kernel - kernel template for multiple functions
* WorkGroupParams allows define  kernel for specific one subgroup function

Co-authored-by: Stuart Brady <stuart.brady@arm.com>
This commit is contained in:
Grzegorz Wawiorko
2021-10-01 12:28:37 +02:00
committed by GitHub
parent 903f1bf65d
commit 92844bead1
12 changed files with 592 additions and 1054 deletions

View File

@@ -18,15 +18,7 @@
#include "subhelpers.h"
extern const char* bcast_source;
extern const char* redadd_source;
extern const char* redmax_source;
extern const char* redmin_source;
extern const char* scinadd_source;
extern const char* scinmax_source;
extern const char* scinmin_source;
extern const char* scexadd_source;
extern const char* scexmax_source;
extern const char* scexmin_source;
extern std::string sub_group_reduction_scan_source;
extern std::string sub_group_generic_source;
#endif