[Bug testsuite/93216] [10 regression] gcc.dg/optimize-bswaphi-1.c fails starting with r280034

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93216

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2020-01-10
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
   Target Milestone|---                         |10.0
     Ever confirmed|0                           |1

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Huh.  I thought those testcases are basically endian independent (somehow),
but obviously they are not.  So we find a HImode bswap instead of a HImode load
for

unsigned short
get_unaligned_16 (unsigned char *p)
{
  return p[0] | (p[1] << 8);
}

looks like the trick in the testcase is to have both BE and LE variants
so the counts will match up.  Will fix.