This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: IA32 bootstrap error
- From: Daniel Berlin <dan at cgsoftware dot com>
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: <robertlipe at usa dot net>, <gcc at gcc dot gnu dot org>
- Date: Sun, 30 Dec 2001 18:47:13 -0500 (EST)
- Subject: Re: IA32 bootstrap error
On Sun, 30 Dec 2001, Richard Kenner wrote:
> I can see that they're somewhat co-dependent as they're each the
> primary consumer/producer of this sort of stuff, but that's not the
> promise that ELF and Dwarf were trying to make. The goal was to be
> interoperable across the object file level, wasn't it?
>
> I strongly agree.
So do I.
And if it was a case of some co-dependency, i'd be with you.
However, the real problem is GDB's problems with perfectly standard
dwarf2.
Nothing more, nothing less.
It lacks support for various things that are in the standard, in this
case, .debug_str support.
Jakub was kind enough add it to gdb at the same time he added support for
emitting them in gcc.
There are in fact, other things in dwarf2 we can't emit right now because
gdb doesn't support them. Location lists come to mind. As do inter-CU
references. There are plenty of other caveats. C++ debugging, is
completely dependent on the fact that we emit a DW_AT_MIPS_linkage_name
attribute that is non-standard. GDB falls down a lot without it (though
it's being fixed).
What about *that* kind of co-dependency?
Every other dwarf2 reader i know of, and utilities that use things like
libdwarf (SGI's now-free dwarf2 reading library), work perfectly with
these features.
It's only gdb that doesn't support it.
If you want to say we shouldn't be co-dependent, then the patch that
makes the cvs gdb requirement is nothing to complain about.
--Dan