diff options
| -rw-r--r-- | gcc/ChangeLog | 7 | ||||
| -rw-r--r-- | gcc/testsuite/ChangeLog | 4 | ||||
| -rw-r--r-- | gcc/testsuite/gcc.dg/vect/costmodel/x86_64/costmodel-alignpeel.c | 9 | ||||
| -rw-r--r-- | gcc/tree-vect-data-refs.c | 14 |
4 files changed, 27 insertions, 7 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 8dee382..0c54109 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,5 +1,12 @@ 2017-05-03 Richard Biener <rguenther@suse.de> + * tree-vect-data-refs.c (vect_enhance_data_refs_alignment): + When all DRs have unknown misaligned do not always peel + when there is a store but apply the same costing model as if + there were only loads. + +2017-05-03 Richard Biener <rguenther@suse.de> + Revert PR tree-optimization/80492 * tree-ssa-alias.c (decl_refs_may_alias_p): Handle diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog index 3f6190bd..2d4d8e0 100644 --- a/gcc/testsuite/ChangeLog +++ b/gcc/testsuite/ChangeLog @@ -1,3 +1,7 @@ +2017-05-03 Richard Biener <rguenther@suse.de> + + * gcc.dg/vect/costmodel/x86_64/costmodel-alignpeel.c: New testcase. + 2017-05-03 Jakub Jelinek <jakub@redhat.com> PR tree-optimization/79472 diff --git a/gcc/testsuite/gcc.dg/vect/costmodel/x86_64/costmodel-alignpeel.c b/gcc/testsuite/gcc.dg/vect/costmodel/x86_64/costmodel-alignpeel.c new file mode 100644 index 0000000..9e20b4a --- /dev/null +++ b/gcc/testsuite/gcc.dg/vect/costmodel/x86_64/costmodel-alignpeel.c @@ -0,0 +1,9 @@ +/* { dg-do compile } */ + +void func(double * __restrict__ v1, double * v2, unsigned n) +{ + for (unsigned i = 0; i < n; ++i) + v1[i] = v2[i]; +} + +/* { dg-final { scan-tree-dump-not "Alignment of access forced using peeling" "vect" } } */ diff --git a/gcc/tree-vect-data-refs.c b/gcc/tree-vect-data-refs.c index aa504b6..c974268 100644 --- a/gcc/tree-vect-data-refs.c +++ b/gcc/tree-vect-data-refs.c @@ -1715,18 +1715,18 @@ vect_enhance_data_refs_alignment (loop_vec_info loop_vinfo) dr0 = first_store; } - /* In case there are only loads with different unknown misalignments, use - peeling only if it may help to align other accesses in the loop or + /* Use peeling only if it may help to align other accesses in the loop or if it may help improving load bandwith when we'd end up using unaligned loads. */ tree dr0_vt = STMT_VINFO_VECTYPE (vinfo_for_stmt (DR_STMT (dr0))); - if (!first_store - && !STMT_VINFO_SAME_ALIGN_REFS ( - vinfo_for_stmt (DR_STMT (dr0))).length () + if (STMT_VINFO_SAME_ALIGN_REFS + (vinfo_for_stmt (DR_STMT (dr0))).length () == 0 && (vect_supportable_dr_alignment (dr0, false) != dr_unaligned_supported - || (builtin_vectorization_cost (vector_load, dr0_vt, 0) - == builtin_vectorization_cost (unaligned_load, dr0_vt, -1)))) + || (DR_IS_READ (dr0) + && (builtin_vectorization_cost (vector_load, dr0_vt, 0) + == builtin_vectorization_cost (unaligned_load, + dr0_vt, -1))))) do_peeling = false; } |