[Executorch][SDPA] Remove slice creation (#9911)

Not needed because the only reason we were slicing was to get sliced seqlen of k and v that is used for causal attention.

Differential Revision: [D71370595](https://our.internmc.facebook.com/intern/diff/D71370595/)

ghstack-source-id: 276012275
Pull Request resolved: #9888

Author: pytorchbot

extension/llm/custom_ops/op_sdpa.cpp

@@ -273,7 +273,6 @@ Tensor& flash_attention_kernel_out(
   Format [n_layers, batch size, max_seq_len, num heads, head dim]
   ....
   @param[in] start_pos: sequence position
-  @param[in] seq_len: Seq length. e.g. seq_len dim of q_projected.
 */
 Tensor& custom_sdpa_out(
     RuntimeContext& ctx,
@@ -306,63 +305,7 @@ Tensor& custom_sdpa_out(
   const int64_t seq_len = q.size(1);
   auto q_seq_len = q.size(1);
 
-  // Refactor the following into create_view util perhaps using
-  // TensorPtr
-  std::array<::executorch::aten::DimOrderType, sdpa::impl::kKVDim>
-      sliced_key_dim_order{0, 1, 2, 3};
-  std::array<::executorch::aten::SizesType, sdpa::impl::kKVDim>
-      sliced_key_sizes;
-  sliced_key_sizes[0] = k.size(0);
-  sliced_key_sizes[1] = start_pos + seq_len; // key_cache.size(2);
-  sliced_key_sizes[2] = k.size(2);
-  sliced_key_sizes[3] = k.size(3);
-  std::array<::executorch::aten::StridesType, sdpa::impl::kKVDim>
-      sliced_key_strides;
-  dim_order_to_stride_nocheck(
-      sliced_key_sizes.data(),
-      sliced_key_dim_order.data(),
-      sdpa::impl::kKVDim,
-      sliced_key_strides.data());
-  // since the cache is sliced, the batch stride needs to stay the same.
-  sliced_key_strides[0] = k.strides()[0];
-  void* key_cache_data = k.mutable_data_ptr();
-  TensorImpl k_impl = TensorImpl(
-      k.scalar_type(),
-      sdpa::impl::kKVDim,
-      sliced_key_sizes.data(),
-      key_cache_data,
-      sliced_key_dim_order.data(),
-      sliced_key_strides.data(),
-      TensorShapeDynamism::STATIC);
-  Tensor sliced_key_cache(&k_impl);
-
-  std::array<::executorch::aten::DimOrderType, sdpa::impl::kKVDim>
-      sliced_value_dim_order{0, 1, 2, 3};
-  std::array<::executorch::aten::SizesType, sdpa::impl::kKVDim>
-      sliced_value_sizes;
-  sliced_value_sizes[0] = v.size(0);
-  sliced_value_sizes[1] = start_pos + seq_len; // value_cache.size(2);
-  sliced_value_sizes[2] = v.size(2);
-  sliced_value_sizes[3] = v.size(3);
-  std::array<::executorch::aten::StridesType, sdpa::impl::kKVDim>
-      sliced_value_strides;
-  dim_order_to_stride_nocheck(
-      sliced_value_sizes.data(),
-      sliced_value_dim_order.data(),
-      sdpa::impl::kKVDim,
-      sliced_value_strides.data());
-  // since the cache is sliced, the batch stride needs to stay the same.
-  sliced_value_strides[0] = v.strides()[0];
-  void* value_cache_data = v.mutable_data_ptr();
-  TensorImpl value_impl = TensorImpl(
-      v.scalar_type(),
-      sdpa::impl::kKVDim,
-      sliced_value_sizes.data(),
-      value_cache_data,
-      sliced_value_dim_order.data(),
-      sliced_value_strides.data(),
-      TensorShapeDynamism::STATIC);
-  Tensor sliced_value_cache(&value_impl);
+  const int64_t num_keys_for_causal_attention = start_pos + seq_len;
 
   ET_KERNEL_CHECK(
       ctx,
@@ -380,38 +323,41 @@ Tensor& custom_sdpa_out(
       sdpa::impl::cpu_flash_attention<CTYPE, 256, 512>(
           output,
           q,
-          sliced_key_cache,
-          sliced_value_cache,
+          k,
+          v,
           dropout_p,
           is_causal,
           attn_mask,
           scale,
           true, /* is_seq_at_dim_1 */
-          start_pos);
+          start_pos,
+          num_keys_for_causal_attention);
     } else if (q_seq_len >= 192) {
       sdpa::impl::cpu_flash_attention<CTYPE, 64, 512>(
           output,
           q,
-          sliced_key_cache,
-          sliced_value_cache,
+          k,
+          v,
           dropout_p,
           is_causal,
           attn_mask,
           scale,
           true, /* is_seq_at_dim_1 */
-          start_pos);
+          start_pos,
+          num_keys_for_causal_attention);
     } else {
       sdpa::impl::cpu_flash_attention<CTYPE, 32, 512>(
           output,
           q,
-          sliced_key_cache,
-          sliced_value_cache,
+          k,
+          v,
           dropout_p,
           is_causal,
           attn_mask,
           scale,
           true, /* is_seq_at_dim_1 */
-          start_pos);
+          start_pos,
+          num_keys_for_causal_attention);
     }
   });
   return output;

extension/llm/custom_ops/op_sdpa_impl.h

@@ -212,7 +212,8 @@ void cpu_flash_attention(
     const optional<Tensor>& attn_mask,
     const optional<double>& scale,
     bool is_seq_at_dim_1 = false,
-    const int64_t start_pos = 0) {
+    const int64_t start_pos = 0,
+    const int64_t num_keys_for_causal_attention = -1) {
   (void)dropout_p;
   // Query (Batch x Num_heads  x Q_seq_len  x Dim_per_head)
   // Key   (Batch x Num_heads  x KV_seq_len x Dim_per_head)
@@ -258,6 +259,13 @@ void cpu_flash_attention(
     kvSize = value.size(1);
   }
 
+  if (num_keys_for_causal_attention > 0) {
+    ET_CHECK_MSG(
+        num_keys_for_causal_attention <= kvSize,
+        "num_keys_for_causal_attention must be <= kvSize");
+    kvSize = num_keys_for_causal_attention;
+  }
+
   ET_CHECK_MSG(
       num_heads_kv <= num_head,
       "FlashAttention does not support num kv heads > num query heads.Got num query heads=%" PRId64