Hello, looking at PR 6152 by Michael Hitch, I expect that it is a problem for _any_ machine using in-memory breakpoint instructions with seperate I/D-cache... - did anybody else notice the problem? - how does gdb insert breakpoints on our different architectures? -is