From: Kevin Buettner <kevinb@redhat.com>
To: "Kris Warkentin" <kewarken@qnx.com>, <gdb@sources.redhat.com>
Subject: Re: solib-search-path not honoured after program start
Date: Wed, 19 Feb 2003 19:44:00 -0000 [thread overview]
Message-ID: <1030219194420.ZM9267@localhost.localdomain> (raw)
In-Reply-To: "Kris Warkentin" <kewarken@qnx.com> "solib-search-path not honoured after program start" (Feb 19, 2:04pm)
On Feb 19, 2:04pm, Kris Warkentin wrote:
> Here's the problem:
>
> Run a program under gdb and break at main. If gdb can't find all the shared
> libs, it complains about it like so:
>
> Error while mapping shared library sections:
> libtestLib_g.so.1: No such file or directory.
>
> So, at main, if I 'info shared', I see something like this:
>
> >From To Syms Read Shared Object Library
> No libtestLib_g.so.1
> 0xb0312504 0xb0349b06 Yes /t/x86/lib/libc.so.2
>
> If I now go and set solib-search-path such that it can find
> libtestLib_g.so.1, and type 'shared', it still doesn't find it.
>
> If I restart the program, there is no problem. For whatever reason, after
> the process has started, gdb never tries to find the shlibs again.
>
> I spent some time tracing around but didn't see exactly where this might be
> fixable. Looks like solib_open does the searching but isn't called later
> on. The shared command calls solib_add which doesn't seem to do any
> searching on solib_search_path.
>
> Can/should this be fixed?
I think this so.
At first glance, the problem seems to be that update_solib_list() only
considers mapping newly found shared libraries and not ones that were
previously known, but which could not be found.
It might not be too hard to change things so that it attempts to map
previously unmappable libraries, but I wonder what should happen to
already mapped and/or loaded shared libraries when solib-search-path
is changed. If changing the search path (or the absolute prefix)
would cause different libraries to be found (than were found
previously), should the old ones be unmapped/discarded? (I suspect
the answer is yes.)
If so, maybe the right way to fix this problem is to have the
"set solib-search-path" and "set solib-absolute-prefix" commands
simply unload all known shared libraries and then invoke solib_add().
Kevin
next prev parent reply other threads:[~2003-02-19 19:44 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-02-19 19:04 Kris Warkentin
2003-02-19 19:44 ` Kevin Buettner [this message]
2003-02-19 19:58 ` Kris Warkentin
2003-02-19 21:11 ` Kevin Buettner
2003-02-19 21:31 ` Kris Warkentin
2003-02-20 18:27 ` Kevin Buettner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1030219194420.ZM9267@localhost.localdomain \
--to=kevinb@redhat.com \
--cc=gdb@sources.redhat.com \
--cc=kewarken@qnx.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox