summaryrefslogtreecommitdiff
path: root/src/clblast.cc
diff options
context:
space:
mode:
authorCedric Nugteren <web@cedricnugteren.nl>2016-02-20 12:40:01 +0100
committerCedric Nugteren <web@cedricnugteren.nl>2016-02-20 12:40:01 +0100
commit8854a731276b3f32c9e381a228733de7c6d95760 (patch)
treefa4084578015ff03473120e9028114d79bec1b02 /src/clblast.cc
parentfadd76207fed5aeb87de7caf744397b008c6d784 (diff)
Added XGER routine, kernel, and tuner
Diffstat (limited to 'src/clblast.cc')
-rw-r--r--src/clblast.cc27
1 files changed, 19 insertions, 8 deletions
diff --git a/src/clblast.cc b/src/clblast.cc
index 77999aaf..aed3f141 100644
--- a/src/clblast.cc
+++ b/src/clblast.cc
@@ -38,6 +38,7 @@
#include "internal/routines/level2/xtrmv.h"
#include "internal/routines/level2/xtbmv.h"
#include "internal/routines/level2/xtpmv.h"
+#include "internal/routines/level2/xger.h"
// BLAS level-3 includes
#include "internal/routines/level3/xgemm.h"
@@ -835,14 +836,24 @@ template StatusCode Tpsv<double2>(const Layout, const Triangle, const Transpose,
// General rank-1 matrix update: SGER/DGER
template <typename T>
-StatusCode Ger(const Layout,
- const size_t, const size_t,
- const T,
- const cl_mem, const size_t, const size_t,
- const cl_mem, const size_t, const size_t,
- cl_mem, const size_t, const size_t,
- cl_command_queue*, cl_event*) {
- return StatusCode::kNotImplemented;
+StatusCode Ger(const Layout layout,
+ const size_t m, const size_t n,
+ const T alpha,
+ const cl_mem x_buffer, const size_t x_offset, const size_t x_inc,
+ const cl_mem y_buffer, const size_t y_offset, const size_t y_inc,
+ cl_mem a_buffer, const size_t a_offset, const size_t a_ld,
+ cl_command_queue* queue, cl_event* event) {
+ auto queue_cpp = Queue(*queue);
+ auto event_cpp = Event(*event);
+ auto routine = Xger<T>(queue_cpp, event_cpp);
+ auto status = routine.SetUp();
+ if (status != StatusCode::kSuccess) { return status; }
+ return routine.DoGer(layout,
+ m, n,
+ alpha,
+ Buffer<T>(x_buffer), x_offset, x_inc,
+ Buffer<T>(y_buffer), y_offset, y_inc,
+ Buffer<T>(a_buffer), a_offset, a_ld);
}
template StatusCode Ger<float>(const Layout,
const size_t, const size_t,