1  /* { dg-do compile } */
       2  /* { dg-options "-O3 -fno-tree-fre -fno-tree-pre -fdump-tree-optimized --param sra-max-scalarization-size-Ospeed=32" } */
       3  /* System Z needs hardware vector support for this to work (the optimization
       4     gets too complex without it.
       5     { dg-additional-options "-march=z13" { target s390x-*-* } } */
       6  /* Use generic tuning on x86 for the same reasons as why alpha, powerpc etc. are
       7     xfailed below.
       8     { dg-additional-options "-mtune=generic" { target i?86-*-* x86_64-*-* } } */
       9  
      10  int
      11  foo ()
      12  {
      13    const int a[8] = { 0, 1, 2, 3, 4, 5, 6, 7 };
      14    int i, sum;
      15  
      16    sum = 0;
      17    for (i = 0; i < sizeof (a) / sizeof (*a); i++)
      18      sum += a[i];
      19  
      20    return sum;
      21  }
      22  
      23  /* After late unrolling the above loop completely DOM should be
      24     able to optimize this to return 28.  */
      25  
      26  /* On alpha, the vectorizer generates writes of two vector elements at once,
      27     but the loop reads only one element at a time, and DOM cannot resolve these.
      28     The same happens on powerpc depending on the SIMD support available.  */
      29  
      30  /* { dg-final { scan-tree-dump "return 28;" "optimized" { xfail { { alpha*-*-* hppa*64*-*-* nvptx*-*-* mmix-knuth-mmixware } || { { { lp64 && { powerpc*-*-* sparc*-*-* riscv*-*-* } } || aarch64_sve } || { arm*-*-* && { ! arm_neon } } } } } } } */