This is the mail archive of the
gdb-patches@sourceware.cygnus.com
mailing list for the GDB project.
Re: path for gdb/dwarf2read.c, support 16-bit targets in dwarf-2
- To: Andrew Cagney <ac131313 at cygnus dot com>
- Subject: Re: path for gdb/dwarf2read.c, support 16-bit targets in dwarf-2
- From: Jim Blandy <jimb at zwingli dot cygnus dot com>
- Date: 03 Apr 2000 12:42:10 -0500
- CC: gdb-patches at sourceware dot cygnus dot com
- References: <38D4DCB0.88313CB2@worldnet.fr> <38D5B6E0.50FF6A5E@cygnus.com><38D68C56.856CB00C@worldnet.fr> <38D74A9E.A85ED8EC@cygnus.com><38D7E6BC.79543EBA@worldnet.fr> <npd7og40xc.fsf@zwingli.cygnus.com><38E06721.6D3A08CD@cygnus.com> <npsnx91szl.fsf@zwingli.cygnus.com><38E28D76.6932DD70@cygnus.com>
> > But which is the correct address --- 0x0000000080000200, or
> > 0xffffffff80000200?
>
> That is target dependant but well defined.
>
> The MIPS ISA/ABI is carefully defined so MIPS2 on MIPS3 et.al. ensures
> that the upper 32 bits is signextended from the low 32 bits.
Okay --- so if the user actually has the address of that object in a
register, they're going to see 0xffffffff80000200, right? So doesn't
your change cause GDB to print incorrect addresses?
Based on my experience with the PA 2.0 port, 64-bit addresses are a
genuine pain to handle. It would be nice to find a comfortable way to
deal with this. But I don't think having GDB print addresses that are
related, but wrong, is a good idea.