diff mbox series

tree-optimization/116609 - SLP live lane vectorization with partial vectors

Message ID 20240905085058.3CDC73864839@sourceware.org
State New
Headers show
Series tree-optimization/116609 - SLP live lane vectorization with partial vectors | expand

Commit Message

Richard Biener Sept. 5, 2024, 8:50 a.m. UTC 
The following implements the simple case of single-lane SLP when
using partial vectors which can use the VEC_EXTRACT_LAST code
generation without changes.  I'll keep the PR open for further
enhancements.

This avoids FAILs of gcc.target/aarch64/sve/live_1.c when using
single-lane SLP for non-grouped stores.

Bootstrap and regtest on x86_64-unknown-linux-gnu in progress.

	PR tree-optimization/116609
	* tree-vect-loop.cc (vectorizable_live_operation_1): Support
	partial vectors for single-lane SLP.
---
 gcc/tree-vect-loop.cc | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)
diff mbox series

Patch

diff --git a/gcc/tree-vect-loop.cc b/gcc/tree-vect-loop.cc
index 242d5e2d916..62cf9205059 100644
--- a/gcc/tree-vect-loop.cc
+++ b/gcc/tree-vect-loop.cc
@@ -10961,7 +10961,8 @@  vectorizable_live_operation_1 (loop_vec_info loop_vinfo,
 
 	 where VEC_LHS is the vectorized live-out result and MASK is
 	 the loop mask for the final iteration.  */
-      gcc_assert (ncopies == 1 && !slp_node);
+      gcc_assert (ncopies == 1
+		  && (!slp_node || SLP_TREE_LANES (slp_node) == 1));
       gimple_seq tem = NULL;
       gimple_stmt_iterator gsi = gsi_last (tem);
       tree len = vect_get_loop_len (loop_vinfo, &gsi,
@@ -10995,7 +10996,7 @@  vectorizable_live_operation_1 (loop_vec_info loop_vinfo,
 
 	 where VEC_LHS is the vectorized live-out result and MASK is
 	 the loop mask for the final iteration.  */
-      gcc_assert (!slp_node);
+      gcc_assert (!slp_node || SLP_TREE_LANES (slp_node) == 1);
       tree scalar_type = TREE_TYPE (STMT_VINFO_VECTYPE (stmt_info));
       gimple_seq tem = NULL;
       gimple_stmt_iterator gsi = gsi_last (tem);
@@ -11147,7 +11148,7 @@  vectorizable_live_operation (vec_info *vinfo, stmt_vec_info stmt_info,
       /* No transformation required.  */
       if (loop_vinfo && LOOP_VINFO_CAN_USE_PARTIAL_VECTORS_P (loop_vinfo))
 	{
-	  if (slp_node)
+	  if (slp_node && SLP_TREE_LANES (slp_node) != 1)
 	    {
 	      if (dump_enabled_p ())
 		dump_printf_loc (MSG_MISSED_OPTIMIZATION, vect_location,
@@ -11166,7 +11167,8 @@  vectorizable_live_operation (vec_info *vinfo, stmt_vec_info stmt_info,
 	    }
 	  else
 	    {
-	      gcc_assert (ncopies == 1 && !slp_node);
+	      gcc_assert (ncopies == 1
+			  && (!slp_node || SLP_TREE_LANES (slp_node) == 1));
 	      if (direct_internal_fn_supported_p (IFN_EXTRACT_LAST, vectype,
 						  OPTIMIZE_FOR_SPEED))
 		vect_record_loop_mask (loop_vinfo,
@@ -11213,8 +11215,9 @@  vectorizable_live_operation (vec_info *vinfo, stmt_vec_info stmt_info,
   if (slp_node)
     {
       gcc_assert (!loop_vinfo
-		  || (!LOOP_VINFO_FULLY_MASKED_P (loop_vinfo)
-		      && !LOOP_VINFO_FULLY_WITH_LENGTH_P (loop_vinfo)));
+		  || ((!LOOP_VINFO_FULLY_MASKED_P (loop_vinfo)
+		       && !LOOP_VINFO_FULLY_WITH_LENGTH_P (loop_vinfo))
+		      || SLP_TREE_LANES (slp_node) == 1));
 
       /* Get the correct slp vectorized stmt.  */
       vec_lhs = SLP_TREE_VEC_DEFS (slp_node)[vec_entry];