This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: [DISCUSS]: cross-debugging shared libraries
- To: "Peter.Schauer" <Peter dot Schauer at regent dot e-technik dot tu-muenchen dot de>
- Subject: Re: [DISCUSS]: cross-debugging shared libraries
- From: Michael Snyder <msnyder at redhat dot com>
- Date: Thu, 26 Oct 2000 11:56:36 -0700
- CC: Alexandre Oliva <aoliva at redhat dot com>, kettenis at wins dot uva dot nl, gdb-patches at sourceware dot cygnus dot com, rfolden at cygnus dot com, ebachalo at cygnus dot com, jld at cygnus dot com
- Organization: Red Hat, Inc.
- References: <200010260906.LAA09732@reisser.regent.e-technik.tu-muenchen.de>
Peter.Schauer wrote:
>
> > > I abhor environment variables, in case of problems debugging becomes
> > > a nightmare, as you might not be able to reproduce the problem without
> > > the correct set of environment variables.
> >
> > Hmm... Good point. But this is no different from needing a certain
> > .gdbinit (or the corresponding gdb settings) to reproduce a problem.
>
> It is different.
>
> I could request e.g the output of a
> show solib-search-path
> command from the problem reporter and see the effective value immediately.
>
> show env
> shows the environment the inferior will see, not necessarily GDB's own
> environment. If someone does a set env in his .gdbinit file, you are lost.
> And yes, I really have seen such scenarios ...
What if I made GDB check the environment variable at start-up
(ONLY), and do a "set solib-search-path" internally at
that time? Then your "show solib-search-path" would work.