Catch2 Test Fixes (#2961)

* test fixes

* address PR comment

* PR comment fixing Nvidia pass

* add additional kernel launch checks

* pr comments

catch/unit/memory/hipMemcpyPeerAsync.cc

@@ -143,6 +143,7 @@ TEST_CASE("Unit_hipMemcpyPeerAsync_Basic") {
       hipLaunchKernelGGL(HipTest::vectorADD, dim3(1), dim3(1),
           0, 0, static_cast<const int*>(A_d),
           static_cast<const int*>(B_d), C_d, numElements*sizeof(int));
+      HIP_CHECK(hipGetLastError());
       HIP_CHECK(hipMemcpy(C_h, C_d, numElements*sizeof(int),
                hipMemcpyDeviceToHost));
       HipTest::checkVectorADD<int>(A_h, B_h, C_h, numElements);
@@ -166,6 +167,7 @@ TEST_CASE("Unit_hipMemcpyPeerAsync_Basic") {
       hipLaunchKernelGGL(HipTest::vectorADD, dim3(1), dim3(1),
           0, 0, static_cast<const int*>(X_d),
           static_cast<const int*>(Y_d), Z_d, numElements*sizeof(int));
+      HIP_CHECK(hipGetLastError());
       HIP_CHECK(hipMemcpy(C_h, Z_d, numElements*sizeof(int),
                hipMemcpyDeviceToHost));
       HipTest::checkVectorADD<int>(A_h, B_h, C_h, numElements);
@@ -227,6 +229,7 @@ TEST_CASE("Unit_hipMemcpyPeerAsync_StreamOnDiffDevice") {
       hipLaunchKernelGGL(HipTest::vectorADD, dim3(1), dim3(1),
           0, 0, static_cast<const int*>(A_d),
           static_cast<const int*>(B_d), C_d, numElements*sizeof(int));
+      HIP_CHECK(hipGetLastError());
       HIP_CHECK(hipMemcpy(C_h, C_d, numElements*sizeof(int),
                hipMemcpyDeviceToHost));
       HipTest::checkVectorADD<int>(A_h, B_h, C_h, numElements);
@@ -240,6 +243,7 @@ TEST_CASE("Unit_hipMemcpyPeerAsync_StreamOnDiffDevice") {
       hipLaunchKernelGGL(HipTest::vectorADD, dim3(1), dim3(1),
           0, 0, static_cast<const int*>(X_d),
           static_cast<const int*>(Y_d), Z_d, numElements*sizeof(int));
+      HIP_CHECK(hipGetLastError());
       HIP_CHECK(hipMemcpy(C_h, Z_d, numElements*sizeof(int),
                hipMemcpyDeviceToHost));