re PR target/92048 (armeb regression after r276645)

2019-10-15  Richard Biener  <rguenther@suse.de>

	PR testsuite/92048
	* gcc.dg/vect/fast-math-vect-pr29925.c: Avoid unrolling of
	inner loop.

From-SVN: r276991

gcc/testsuite/ChangeLog

+2019-10-15  Richard Biener  <rguenther@suse.de>
+
+	PR testsuite/92048
+	* gcc.dg/vect/fast-math-vect-pr29925.c: Avoid unrolling of
+	inner loop.
+
 2019-10-15  Ilya Leoshkevich  <iii@linux.ibm.com>
 
 	* gcc.target/s390/load-thread-pointer-once.c: New test.

gcc/testsuite/gcc.dg/vect/fast-math-vect-pr29925.c

@@ -13,6 +13,7 @@ void interp_pitch(float *exc, float *interp, int pitch, int len)
    for (i=0;i<len;i++)
    {
       float tmp = 0;
+#pragma GCC unroll 0
       for (k=0;k<7;k++)
       {
          tmp += exc[i-pitch+k+maxj-6];