Fix =~ $str_overloaded (5.10 regression)

Fix =~ $str_overloaded (5.10 regression)

[ DAPM: I just cherry-picked the tests from this commit, since my own
  changes have already fixed this bug. FC's two commits:
      15d9c083b08647e489d279a1059b4f14a3df187b
      3e1022372a8200bc4c7354e0f588c7f71584a888
  were unrolled at the start of this branch since they clashed with my own
  stuff; this commit is re-adding the bits of those commits that are
  still needed: i.e. just the tests.
]

In 5.8.x, this code:

  use overload '""'=>sub { warn "stringify"; --$| ? "gonzo" : chr 256 };
  my $obj = bless\do{my $x};
  warn "$obj";
  print "match\n" if chr(256) =~ $obj;

prints

  stringify at - line 1.
  gonzo at - line 3.
  stringify at - line 1.
  match

which is to be expected.

In 5.10+, the stringification happens one extra time, causing a failed match:

  stringify at - line 1.
  gonzo at - line 3.
  stringify at - line 1.
  stringify at - line 1.

This logic in pp_regcomp is faulty:

	    if (DO_UTF8(tmpstr)) {
		assert (SvUTF8(tmpstr));
	    } else if (SvUTF8(tmpstr)) {
		... copy under ‘use bytes’...
	    }
	    else if (SvAMAGIC(tmpstr)) {
		/* make a copy to avoid extra stringifies */
		tmpstr = newSVpvn_flags(t, len, SVs_TEMP | SvUTF8(tmpstr));
	    }

The SvAMAGIC check never happens when the UTF8 flag is on.