adding opencl convolution benchmark

This commit is contained in:
Blaise Tine
2023-11-14 22:31:30 -08:00
parent 4e7a536918
commit 61e3442ef8
16 changed files with 490 additions and 170 deletions

View File

@@ -12,10 +12,10 @@ inline uint32_t log2_fast(uint32_t x) {
}
void kernel_body(uint32_t task_id, kernel_arg_t* __UNIFORM__ arg) {
auto size = arg->size;
auto A = reinterpret_cast<TYPE*>(arg->A_addr);
auto A = reinterpret_cast<TYPE*>(arg->A_addr);
auto B = reinterpret_cast<TYPE*>(arg->B_addr);
auto C = reinterpret_cast<TYPE*>(arg->C_addr);
auto size = arg->size;
uint32_t row, col;
if (is_log2(size)) {