Comment 27 for bug 685352

Building the following test case with current mainline on i386:

unsigned short test (unsigned char val) __attribute__ ((noinline));

unsigned short
test (unsigned char val)
{
  return val * 255;
}

int
main(int argc, char**argv)
{
  printf ("test(val=40) = %x\n", test(0x40));
  return 0;
}

We get the following (correct) output with -O0:
test(val=40) = 3fc0

and the following incorrect output with -O2:
test(val=40) = ffc0

The problem appears to be related to this piece of code in expand_expr_real2, case WIDEN_MULT_EXPR:

                  expand_operands (treeop0, treeop1, NULL_RTX, &op0, &op1,
                                   EXPAND_NORMAL);
                  temp = expand_widening_mult (mode, op0, op1, target,
                                               unsignedp, this_optab);

expand_operands will expand the constant 255 into QImode and return a (const_int -1) for op1. Passing this constant into expand_widening_mult then apparently generates a simple negation operation in HImode instead (via expand_const_mult) ...

It seems this code came in here:
http://gcc.gnu.org/ml/gcc-patches/2010-04/msg01327.html
Any suggestions how this ought to be handled?