eSi-RISC has vector permute functionality, but no unaligned loads. We
see execution failures on gcc.dg/vect/slp-perm-12.c because loop
versioning is used to make the tptr aligned for the first loop
iteration, and then with a step of originally 11, 22 after
vectorization, and a vector alignment of 8 bytes, the second iteration
causes an AlignmentError exception.
The attached patch to tree-vect-data-refs.c suppresses attempts to align
data accesses where the
step alignment times the vectorization factor is insufficient to sustain
the alignment during the loop.
Bootstrapped and regression tested on x86_64-pc-linux-gnu .
I have also attached a matching testsuite patch to not expect SLP
vectorization for slp-perm-12 when
no unaligned loads are available, although in terms of testing, I can
only say that it works for us.
2018-12-15 Joern Rennecke <joern.renne...@riscy-ip.com>
* tree-vect-data-refs.c (vect_enhance_data_refs_alignment): Don't do
versioning for data accesses with misaligned step.
Index: tree-vect-data-refs.c
===================================================================
--- tree-vect-data-refs.c (revision 267262)
+++ tree-vect-data-refs.c (working copy)
@@ -2160,6 +2160,20 @@ vect_enhance_data_refs_alignment (loop_v
break;
}
+ /* Forcing alignment in the first iteration is no good if
+ we don't keep it across iterations. For now, just disable
+ versioning in this case.
+ ?? We could actually unroll the loop to archive the required
+ overall step alignemnt, and forcing the alignment could be
+ done by doing some iterations of the non-vectorized loop. */
+ if (maybe_lt (LOOP_VINFO_VECT_FACTOR (loop_vinfo)
+ * DR_STEP_ALIGNMENT (dr),
+ TYPE_ALIGN_UNIT (vectype)))
+ {
+ do_versioning = false;
+ break;
+ }
+
/* The rightmost bits of an aligned address must be zeros.
Construct the mask needed for this test. For example,
GET_MODE_SIZE for the vector mode V4SI is 16 bytes so the
2018-12-15 Joern Rennecke <joern.renne...@riscy-ip.com>
* testsuite/gcc.dg/vect/slp-perm-12.c (dg-final): Don't expect SLP
vectorization for ! vect_no_align.
Index: testsuite/gcc.dg/vect/slp-perm-12.c
===================================================================
--- testsuite/gcc.dg/vect/slp-perm-12.c (revision 5616)
+++ testsuite/gcc.dg/vect/slp-perm-12.c (revision 5617)
@@ -49,4 +49,4 @@ int main()
return 0;
}
-/* { dg-final { scan-tree-dump-times "vectorizing stmts using SLP" 1 "vect" {
target vect_perm } } } */
+/* { dg-final { scan-tree-dump-times "vectorizing stmts using SLP" 1 "vect" {
target { vect_perm && {! vect_no_align } } } } } */
