[Kernel] Add expert_map support to Cutlass FP8 MOE

csrc/quantization/cutlass_w8a8/moe/moe_data.cu

@@ -46,14 +46,26 @@ __global__ void compute_expert_offsets(
 }
 
 __global__ void compute_arg_sorts(const int* __restrict__ topk_ids,
+                                  const int32_t* __restrict__ expert_offsets,
                                   int32_t* input_permutation,
                                   int32_t* output_permutation,
                                   int32_t* atomic_buffer, const int topk_length,
                                   const int topk) {
-  int expert_id = blockIdx.x;
+  int const blk_expert_id = blockIdx.x;
+  int const num_experts = gridDim.x;
+  int32_t const num_tokens = expert_offsets[num_experts];
 
   for (int i = threadIdx.x; i < topk_length; i += THREADS_PER_EXPERT) {
-    if (topk_ids[i] == expert_id) {
+    int const expert_id = topk_ids[i];
+    if (expert_id == -1 && blockIdx.x == 0) {
+      // output_permutation is used to re-order the moe outputs. It is
+      // used as c2 = c2[c_map], where c2 is a torch.tensor that is the
+      // output of the cutlass kernels and c_map is the output_permutation.
+      // c2 is initialized to zeros, therefore by setting the output_permutation
+      // to num_tokens, we are guaranteed to fill the moe outputs to zero
+      // for "invalid" topk_ids.
+      output_permutation[i] = num_tokens;
+    } else if (expert_id == blk_expert_id) {
       int start = atomicAdd(&atomic_buffer[expert_id], 1);
       input_permutation[start] = i / topk;
       output_permutation[i] = start;
@@ -83,6 +95,7 @@ void get_cutlass_moe_mm_data_caller(
       static_cast<int32_t*>(atomic_buffer.data_ptr()), num_experts);
   compute_arg_sorts<<<num_experts, num_threads, 0, stream>>>(
       static_cast<const int32_t*>(topk_ids.data_ptr()),
+      static_cast<const int32_t*>(expert_offsets.data_ptr()),
       static_cast<int32_t*>(input_permutation.data_ptr()),
       static_cast<int32_t*>(output_permutation.data_ptr()),
       static_cast<int32_t*>(atomic_buffer.data_ptr()), topk_ids.numel(),