This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [rfc/NEWS] GDB works with GCC -feliminate-dwarf2-dups.


"Eli Zaretskii" <eliz@gnu.org> wrote:
> I don't understand what you are saying ...

It's like this.  When gdb and gcc don't work together, I have to decide
whether I think it's a bug in gdb (so I file a gdb PR) or a bug in gcc
(so I file a gcc PR).  Some contemplation of this relationship leads
me to say: gdb and gcc are truly separate entities.  It's not that
"gdb X works with gcc Y", it's more exactly "gdb X and gcc Y both
support standards Z1, Z2, and Z3".

This is especially true because gdb works with other compilers besides gcc,
and as far as I know, gcc works with other debuggers besides gdb.

It's a subtle point, and even after you see it, you might not agree
with it.

eli> GDB can now debug programs compiled with the -feliminate-dwarf2-dups
eli> option to GCC 3.3.4 or later, as well as programs compiled with
eli> proprietary compilers that produce similar debug information (a more
eli> compact representation of DWARF-2 debug information using the
eli> DW_FORM_ref_addr references).

Well, after fooling with this a little bit, I like my version better.
I'm comfortable claiming that GDB supports DW_FORM_ref_addr, so that if
a compiler produces this information, GDB can support it.  I'm not
comfortable claiming that GDB can support "programs compiled with
proprietary compilers that produce similar debug information".

But you're the doco guy, so if you like it that way, go for it.

Michael


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]