On Mon, Jan 10, 2005 at 10:54:55AM -0500, Robert Baruch wrote: > I'm having a problem with a subreg(mem(X)) insn being generated as a > result of reload. Whatever created that subreg, and didn't resolve it, is the bug. r~