From 19f51596e8be9fe87d583616466581ab5740c19d Mon Sep 17 00:00:00 2001
From: chenwei <15601910741@163.com>
Date: Tue, 29 Oct 2024 23:56:59 +0800
Subject: [PATCH] feat: support kFactor 8 used in mma tensor op tile iterator
 (#1512)

---
 .../cutlass/gemm/warp/mma_tensor_op_tile_iterator.h  | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h b/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h
index 46690bf1..e6e6d70f 100644
--- a/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h
+++ b/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h
@@ -2315,7 +2315,17 @@ class MmaTensorOpMultiplicandTileIterator<
     int access_contiguous_idx = -1;
     int access_strided_idx = -1;
 
-    if (Layout::kFactor == 4) {
+    if (Layout::kFactor == 8) {
+      int factor_in_partition =
+          (Layout::PartitionShape::kContiguous * Layout::kFactor /
+           Layout::TileShape::kContiguous);
+
+      if (Policy::LdsmShape::kStrided == Policy::LdsmShape::kCount) {
+        partition_contiguous_idx = lane_in_quad_pair / factor_in_partition;
+        access_contiguous_idx = ((lane_in_quad) ^ (lane_id / Layout::kFactor));
+        access_strided_idx = lane_id / Layout::kFactor;
+      }
+    } else if (Layout::kFactor == 4) {
       // Super Integer matrix multiply Interleaved-32
 
       int factor_in_partition =