Build fbgemm_gpu for TORCH_CUDA_ARCH_LIST=10.0 and CUDA 12.8 and 12.9 (#162544)

## Summary
- pytorch is not built for *a variants of SM architectures, due to non-portability. However, we need fbgemm_gpu kernels built for sm100a (see #162209)

## Changes
- **Setting USE_FBGEMM_GENAI for CUDA builds**: fbgemm_gpu builds for sm100a if using CUDA 12.8 or 12.9 ([source](2033a0a08f/.github/scripts/nova_dir.bash (L29-L32))), so I follow the same rule here.
- **Extra nvcc flags**: if USE_FBGEMM_GENAI and USE_CUDA are set, we add extra nvcc flags for sm100a

## Test plan

Test build:
```
echo $CUDA_HOME
/usr/local/cuda-12.9

export TORCH_CUDA_ARCH_LIST=10.0
python -m pip install --no-build-isolation -v -e .
```

Check build logs:
```
  CMake Warning at CMakeLists.txt:901 (message):
    Setting USE_FBGEMM_GENAI to ON, doing CUDA build for SM100a
```

Run unit tests:
- `pytest test/test_matmul_cuda.py  -k test_mxfp8_scaled_grouped_mm`

Pull Request resolved: https://github.com/pytorch/pytorch/pull/162544
Approved by: https://github.com/drisspg

CMakeLists.txt

@@ -902,9 +902,9 @@ IF(USE_FBGEMM_GENAI AND USE_ROCM AND NOT "gfx942" IN_LIST PYTORCH_ROCM_ARCH)
   set(USE_FBGEMM_GENAI off)
 endif()
 
-# Set USE_FBGEMM_GENAI to ON for CUDA build on SM100
-if(USE_CUDA AND "$ENV{TORCH_CUDA_ARCH_LIST}" MATCHES "10.0a")
-  message(WARNING "Setting USE_FBGEMM_GENAI to ON for CUDA build on SM100")
+# Set USE_FBGEMM_GENAI to ON for CUDA build on SM100.
+if(USE_CUDA AND "$ENV{TORCH_CUDA_ARCH_LIST}" MATCHES "10.0" AND CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 12.8)
+  message(STATUS "Setting USE_FBGEMM_GENAI to ON, doing CUDA build for SM100a")
   set(USE_FBGEMM_GENAI ON)
 endif()
 

aten/src/ATen/CMakeLists.txt

@@ -265,6 +265,14 @@ IF(USE_FBGEMM_GENAI)
       "${FBGEMM_GENAI_SRCS}/cutlass_extensions/**/*.cu")
     list(FILTER fbgemm_genai_native_cuda_cu INCLUDE REGEX ${FBGEMM_CUTLASS_KERNELS_REGEX})
 
+    # PyTorch is not built for 10.0a in CI, due to lack of portability,
+    # so we need to explicitly build these files for 10.0a.
+    foreach(cu_file ${fbgemm_genai_native_cuda_cu})
+      _BUILD_FOR_ADDITIONAL_ARCHS(
+        "${cu_file}"
+        "100a")
+    endforeach()
+
     file(GLOB_RECURSE fbgemm_genai_native_cuda_cpp
       "${FBGEMM_GENAI_SRCS}/common/*.cpp"
     )