[ESIMD] Added a test for slm_load/slm_store APIs

SYCL/ESIMD/api/slm_gather_scatter.cpp

@@ -0,0 +1,133 @@
+// REQUIRES: gpu
+// UNSUPPORTED: cuda
+// RUN: %clangxx -fsycl %s -o %t.out
+// RUN: %GPU_RUN_PLACEHOLDER %t.out
+//
+// The test checks functionality of the slm_gather/slm_scatter ESIMD APIs.
+
+// TODO: Enable the test for 1 and 2 byte types when the implementation is fixed
+#define SKIP_ONE_AND_TWO_BYTE_BLOCKS 1
+
+#include "../esimd_test_utils.hpp"
+
+#include <CL/sycl.hpp>
+#include <CL/sycl/INTEL/esimd.hpp>
+#include <iostream>
+
+using namespace cl::sycl;
+
+template <typename T, unsigned VL, unsigned STRIDE> struct Kernel {
+  T *buf;
+  Kernel(T *buf) : buf(buf) {}
+
+  void operator()(id<1> i) const SYCL_ESIMD_KERNEL {
+    using namespace sycl::ext::intel::experimental::esimd;
+
+    // In this test, we have a single workitem. No barriers required.
+    slm_init(VL * STRIDE *
+             sizeof(typename sycl::ext::intel::experimental::esimd::detail::
+                        dword_type<T>::type));
+
+    simd<T, VL> valsIn;
+    valsIn.copy_from(buf);
+
+    simd<uint32_t, VL> offsets(0, STRIDE * sizeof(T));
+    slm_scatter<T, VL>(valsIn, offsets);
+
+    simd_mask<VL> pred = 1;
+    pred[VL - 1] = 0; // mask out the last lane
+    simd<T, VL> valsOut = slm_gather<T, VL>(offsets, pred);
+    valsOut *= -1;
+
+    valsOut.copy_to(buf);
+  }
+};
+
+template <typename T, unsigned VL, unsigned STRIDE> bool test(queue q) {
+  using namespace sycl::ext::intel::experimental::esimd;
+  constexpr size_t size = VL;
+
+  std::cout << "Testing T=" << typeid(T).name() << " VL=" << VL
+            << " STRIDE=" << STRIDE << "...\n";
+
+  auto dev = q.get_device();
+  auto ctxt = q.get_context();
+  T *A = static_cast<T *>(malloc_shared(size * sizeof(T), dev, ctxt));
+  T *gold = new T[size];
+
+  for (int i = 0; i < size; ++i) {
+    A[i] = (T)-i;
+    gold[i] = (T)i;
+  }
+
+  // Account for masked out last lane (with pred argument to slm_gather).
+  gold[(VL - 1)] = (T)0;
+
+  try {
+    range<1> glob_range{1};
+
+    auto e = q.submit([&](handler &cgh) {
+      Kernel<T, VL, STRIDE> kernel(A);
+      cgh.parallel_for(glob_range, kernel);
+    });
+    e.wait();
+  } catch (sycl::exception const &e) {
+    std::cerr << "SYCL exception caught: " << e.what() << '\n';
+    free(A, ctxt);
+    delete[] gold;
+    return static_cast<bool>(e.code());
+  }
+
+  int err_cnt = 0;
+  for (unsigned i = 0; i < size; ++i) {
+    if (A[i] != gold[i]) {
+      if (++err_cnt < VL) {
+        std::cout << "failed at index " << i << ": " << A[i]
+                  << " != " << gold[i] << " (gold)\n";
+      }
+    }
+  }
+
+  if (err_cnt > 0) {
+    std::cout << "  pass rate: "
+              << ((float)(size - err_cnt) / (float)size) * 100.0f << "% ("
+              << (size - err_cnt) << "/" << size << ")\n";
+  }
+
+  free(A, ctxt);
+  delete[] gold;
+
+  std::cout << (err_cnt > 0 ? "  FAILED\n" : "  Passed\n");
+  return err_cnt > 0 ? false : true;
+}
+
+template <typename T, unsigned VL> bool test(queue q) {
+  bool passed = true;
+  passed &= test<T, VL, 1>(q);
+  passed &= test<T, VL, 2>(q);
+  passed &= test<T, VL, 3>(q);
+  passed &= test<T, VL, 4>(q);
+  return passed;
+}
+
+int main(void) {
+  queue q(esimd_test::ESIMDSelector{}, esimd_test::createExceptionHandler());
+
+  auto dev = q.get_device();
+  std::cout << "Running on " << dev.get_info<info::device::name>() << "\n";
+
+  bool passed = true;
+
+#ifndef SKIP_ONE_AND_TWO_BYTE_BLOCKS
+  passed &= test<char, 16>(q);
+  passed &= test<char, 32>(q);
+  passed &= test<short, 16>(q);
+  passed &= test<short, 32>(q);
+#endif
+  passed &= test<int, 16>(q);
+  passed &= test<int, 32>(q);
+  passed &= test<float, 16>(q);
+  passed &= test<float, 32>(q);
+
+  return passed ? 0 : 1;
+}