Subject: [firstname.lastname@example.org: Re: gdb 5.0 in NetBSD?]
To: None <email@example.com>
From: Frank van der Linden <firstname.lastname@example.org>
Date: 05/31/2000 10:55:18
I'm forwarding this one to Andrew Cagney and tech-toolchain, hopefully
it can be analyzed/fixed.
----- Forwarded message from Brett Lymn <email@example.com> -----
From: firstname.lastname@example.org (Brett Lymn)
Subject: Re: gdb 5.0 in NetBSD?
Your doubts are correct, at least on i386. GDB barfs in gdb/solib.c
in, to me, a non-obvious way -they have these defines at the top of
#define SOLIB_EXTRACT_ADDRESS(member) \
extract_address (&member, sizeof (member))
#define LM_ADDR(so) (SOLIB_EXTRACT_ADDRESS ((so) -> lm.lm_addr))
and this line of code:
p->addr += LM_ADDR (so);
gets translated to this:
p->addr += (extract_address (& ( so ) -> lm. som_addr , sizeof ( ( so ) -> lm. som_addr )) ) ;
Note the "som_addr" - what puzzles me is how that turned up.