pass unittest for tensor core gemm.

[haruhi55]

There are still several cases where the tensor core GEMM does not pass the unit tests. The implementation contains some bugs.

TiledCUDA/tests/cpp/cell/test_gemm.cu

Lines 357 to 359 in 471fd6f

 // run_test<32, 32, 32, tl::RowMajor<2, 1>, 32>(); 
 // run_test<64, 64, 32, tl::RowMajor<2, 2>, 32>(); 
 // run_test<64, 32, 128, tl::RowMajor<2, 2>, 32>();