sw: Remove all use of custom compiler intrinsics

Author: colluca

sw/tests/src/openmp_for_static_schedule.c

@@ -25,26 +25,22 @@ unsigned __attribute__((noinline)) static_schedule(void) {
 #pragma omp parallel firstprivate(data_a, data_x, data_y)
     {
         int nthreads = omp_get_num_threads();
-        // DM, rep, bound, stride, data
-        __builtin_ssr_setup_1d_r(
-            0, 0, AXPY_N / nthreads - 1, sizeof(double),
-            &data_x[AXPY_N / nthreads * omp_get_thread_num()]);
-        __builtin_ssr_setup_1d_r(
-            1, 0, AXPY_N / nthreads - 1, sizeof(double),
-            &data_y[AXPY_N / nthreads * omp_get_thread_num()]);
-        __builtin_ssr_setup_1d_w(
-            2, 0, AXPY_N / nthreads - 1, sizeof(double),
-            &data_y[AXPY_N / nthreads * omp_get_thread_num()]);
-        __builtin_ssr_enable();
+        snrt_ssr_loop_1d(SNRT_SSR_DM0, AXPY_N / nthreads, sizeof(double));
+        snrt_ssr_read(SNRT_SSR_DM0, SNRT_SSR_1D, &data_x[AXPY_N / nthreads * omp_get_thread_num()]);
+        snrt_ssr_loop_1d(SNRT_SSR_DM1, AXPY_N / nthreads, sizeof(double));
+        snrt_ssr_read(SNRT_SSR_DM1, SNRT_SSR_1D, &data_y[AXPY_N / nthreads * omp_get_thread_num()]);
+        snrt_ssr_loop_1d(SNRT_SSR_DM2, AXPY_N / nthreads, sizeof(double));
+        snrt_ssr_write(SNRT_SSR_DM2, SNRT_SSR_1D, &data_y[AXPY_N / nthreads * omp_get_thread_num()]);
+        snrt_ssr_enable();
 #pragma omp for schedule(static)
         for (unsigned i = 0; i < AXPY_N; i++) {
-            // data_y[i] = data_a * data_x[i] + data_y[i];
-            // data_y[i] = data_a * __builtin_ssr_pop(0) +
-            __builtin_ssr_pop(1);
-            __builtin_ssr_push(
-                2, data_a * __builtin_ssr_pop(0) + __builtin_ssr_pop(1));
+            asm volatile(
+                "fmadd.d ft2, %[a], ft0, ft1\n"
+                :
+                : [ a ] "f"(data_a)
+                : "ft0", "ft1", "ft2", "memory");
         }
-        __builtin_ssr_disable();
+        snrt_ssr_disable();
     }
 
     // check data