procyberian/gcc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Fix bogus SLP nvector compute in check_load_store_for_partial_vectors

Browse source 
We have a new overload for vect_get_num_copies that handles both
SLP and non-SLP.  Use it and avoid the division by group_size
for SLP when not using load-store lanes.

	* tree-vect-stmts.cc (check_load_store_for_partial_vectors):
	Use the new vect_get_num_copies overload.  Only divide by
	group_size for SLP for load-store lanes.
This commit is contained in:
Richard Biener 2024-09-24 10:42:01 +02:00 • committed by Richard Biener
parent 9c04112fdc
commit 3db9e99165
1 changed files with 3 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
gcc/tree-vect-stmts.cc
View file

@ -1507,19 +1507,15 @@ check_load_store_for_partial_vectors (loop_vec_info loop_vinfo, tree vectype,
if (memory_access_type == VMAT_INVARIANT)
return;
unsigned int nvectors;
if (slp_node)
/* ??? Incorrect for multi-lane lanes. */
nvectors = SLP_TREE_NUMBER_OF_VEC_STMTS (slp_node) / group_size;
else
nvectors = vect_get_num_copies (loop_vinfo, vectype);
unsigned int nvectors = vect_get_num_copies (loop_vinfo, slp_node, vectype);
vec_loop_masks *masks = &LOOP_VINFO_MASKS (loop_vinfo);
vec_loop_lens *lens = &LOOP_VINFO_LENS (loop_vinfo);
machine_mode vecmode = TYPE_MODE (vectype);
bool is_load = (vls_type == VLS_LOAD);
if (memory_access_type == VMAT_LOAD_STORE_LANES)
{
if (slp_node)
nvectors /= group_size;
internal_fn ifn
= (is_load ? vect_load_lanes_supported (vectype, group_size, true)
: vect_store_lanes_supported (vectype, group_size, true));