https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126981
Bug ID: 126981
Summary: [17 Regression] Wrong matching of saturating
truncation
Product: gcc
Version: 17.0
Status: UNCONFIRMED
Keywords: wrong-code
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: ktkachov at gcc dot gnu.org
Target Milestone: ---
Target: aarch64
Since we enabled the sattrunc optabs for aarch64 we are hitting a wrong code
matching from match.pd:
/* { dg-require-effective-target vect_int } */
#define N 64
/* Clip to [0, 65535]. At INT_MIN the negation is its own inverse, so the
shift yields -1 and the result is 65535 rather than 0. */
__attribute__ ((noipa)) void
clip (unsigned short *__restrict out, const int *__restrict in, int n)
{
for (int i = 0; i < n; ++i)
{
int x = in[i];
out[i] = ((unsigned int) x > 65535u
? (int) (-(unsigned int) x) >> 31
: x);
}
}
int
main (void)
{
int in[N];
unsigned short out[N];
for (int i = 0; i < N; ++i)
in[i] = (i & 3) == 0 ? (-__INT_MAX__ - 1) : i * 12345 - 30000;
clip (out, in, N);
#pragma GCC novector
for (int i = 0; i < N; ++i)
{
int x = in[i];
unsigned short ref = ((unsigned int) x > 65535u
? (int) (-(unsigned int) x) >> 31
: x);
if (out[i] != ref)
__builtin_abort ();
}
return 0;
}
at -O3 this aborts but not at -O2 on aarch64
