This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: --gc-section leftovers workaround.
On Wednesday 28 October 2009 20:13:43, Paul Pluzhnikov wrote:
> Since we've just sorted them, (FDE+1)->initial_location == 0 implies that
> FDE->initial_location == 0 as well, doesn't it?
Correct, but it may be FDE+1 that needs discarding, not FDE.
We could even have FDE,FDE+1,FDE+2,FDE+3, where
FDE,FDE+1,FDE+2 all have initial_location == 0, and FDE+3 not.
We want to compare each of FDE,FDE+1,FDE+2 with FDE+3 for
overlap and discard those that do overlap,
not FDE with FDE+1 ; FDE+1 with FDE+2 ; FDE+2 with FDE+3.
--
Pedro Alves