This fixes a few failures on ARM and AArch64 due to a recent change in...

This fixes a few failures on ARM and AArch64 due to a recent change in
alignment peeling by switching the vector cost model off.

Tested on AArch64, ARM and x64 - committed as obvious.

    testsuite/
	* gcc.dg/vect/vect-44.c: Add -fno-vect-cost-model.
	* gcc.dg/vect/vect-50.c: Likewise.

From-SVN: r247737

diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index df971bd475076411dc79926c1dd4938ca99b1672..14f201b805795675632359fcde14323ade49e900 100644 (file)
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2017-05-08  Wilco Dijkstra  <wdijkstr@arm.com>
+
+       * testsuite/gcc.dg/vect/vect-44.c: Add -fno-vect-cost-model.
+       * gcc/testsuite/gcc.dg/vect/vect-50.c: Likewise.
+
 2017-05-07  Jeff Law  <law@redhat.com>
 
        Revert:

diff --git a/gcc/testsuite/gcc.dg/vect/vect-44.c b/gcc/testsuite/gcc.dg/vect/vect-44.c
index 186f9cfc9e26d6eb53514dec0fac176d696ec578..fbc593572429422c8e527c5e5559c515efd38aa6 100644 (file)
--- a/gcc/testsuite/gcc.dg/vect/vect-44.c
+++ b/gcc/testsuite/gcc.dg/vect/vect-44.c
@@ -1,4 +1,5 @@
 /* { dg-require-effective-target vect_float } */
+/* { dg-options "-fno-vect-cost-model" } */
 
 #include <stdarg.h>
 #include "tree-vect.h"

diff --git a/gcc/testsuite/gcc.dg/vect/vect-50.c b/gcc/testsuite/gcc.dg/vect/vect-50.c
index 78bfd8d3920445fe51c7393a82870ea85f62bb55..0d5febc165ee3bf3b7b595237168d9d4b9604d4b 100644 (file)
--- a/gcc/testsuite/gcc.dg/vect/vect-50.c
+++ b/gcc/testsuite/gcc.dg/vect/vect-50.c
@@ -1,4 +1,5 @@
 /* { dg-require-effective-target vect_float } */
+/* { dg-options "-fno-vect-cost-model" } */
 
 #include <stdarg.h>
 #include "tree-vect.h"