https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68559
Bug ID: 68559
Summary: Excessive peeling for gaps
Product: gcc
Version: 6.0
Status: UNCONFIRMED
Keywords: missed-optimization
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: rguenth at gcc dot gnu.org
Blocks: 53947
Target Milestone: ---
With N == 20 gcc.dg/vect/pr45752.c only contains the peeling-for-gap copy
because of the following premature check which doesn't consider the
group being vectorized using SLP (or three-element interleaving).
SLP will never require gap peeling for exact_log2 (groupsize) == -1
and three-element interleaving neither.
The LOOP_VINFO_PEELING_FOR_GAPS analysis should be postponed (apart from
the single-element interleaving case).
static bool
vect_analyze_group_access_1 (struct data_reference *dr)
{
...
/* If there is a gap in the end of the group or the group size cannot
be made a multiple of the vector element count then we access excess
elements in the last iteration and thus need to peel that off. */
if (loop_vinfo
&& (groupsize - last_accessed_element > 0
|| exact_log2 (groupsize) == -1))
{
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION, vect_location,
"Data access with gaps requires scalar "
"epilogue loop\n");
if (loop->inner)
{
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION, vect_location,
"Peeling for outer loop is not supported\n");
return false;
}
LOOP_VINFO_PEELING_FOR_GAPS (loop_vinfo) = true;
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53947
[Bug 53947] [meta-bug] vectorizer missed-optimizations
