On Wed, 28 Mar 2001, Dan Drake wrote: > Line 158 of the newcpu.h file mentioned is nothing special; it's a > simple function that does one line of pointer arithmetic. Well, I've had a rough time with pointer arithmetic in the past, but I can't imagine that gcc would barf that badly; surely it traps invalid pointers? I'd submit the bug report, though. Have you ever seen this code compile with your own eyes anywhere? -- "To misattribute a quote is unforgivable." --Anonymous