Re: c++/1267: stepping does not work in shared libraries

[Daniel Jacobowitz <drow at mvista dot com>]

The following reply was made to PR c++/1267; it has been noted by GNATS.

From: Daniel Jacobowitz <drow@mvista.com>
To: ldm@engr.uconn.edu
Cc: gdb-gnats@sources.redhat.com
Subject: Re: c++/1267: stepping does not work in shared libraries
Date: Tue, 1 Jul 2003 12:16:15 -0500

 On Tue, Jul 01, 2003 at 04:59:01PM -0000, ldm@engr.uconn.edu wrote:
 > 
 > >Number:         1267
 > >Category:       c++
 > >Synopsis:       stepping does not work in shared libraries
 > >Confidential:   no
 > >Severity:       serious
 > >Priority:       medium
 > >Responsible:    unassigned
 > >State:          open
 > >Class:          change-request
 > >Submitter-Id:   net
 > >Arrival-Date:   Tue Jul 01 17:08:00 UTC 2003
 > >Closed-Date:
 > >Last-Modified:
 > >Originator:     ldm@engr.uconn.edu
 > >Release:        GNU gdb 5.3.90_2003-06-29-cvs-debian
 > >Organization:
 > >Environment:
 > aegir /home/ldm/willow/proposals/CAREER [131] -> uname -a
 > Linux aegir 2.4.20 #17 SMP Thu May 1 17:10:22 EDT 2003 i686 GNU/Linux
 > Debian/sid
 > >Description:
 > I cannot use the next command when my code is compiled with shared libraries. Usually, I place a breakpoint in main, run the program, then set a breakpoint somewhere in one of my shared library and continue. 
 > 
 > When the program stops at the breakpoint the only commands that work reliably are
 > 1) set breakpoint
 > 2) step inside (step)
 > 
 > The step over (next) command has the same effect as continue. This makes the debugger almost unusable at I have to create tons of temporary breakpoints.
 > 
 > This has been going on for a while now (probably several months). 
 
 What version of GCC are you using, and has it changed recently?  If
 it's the Debian/unstable system compiler, then it recently changed from
 3.2 to 3.3.
 
 > >How-To-Repeat:
 > A C++ program. A C++ shared library (no dynamic loading necessary). Follow the step above after that. 
 
 It works for me.  But I've seen another report very much like this
 recently, so there must be something in your testcase that isn't in
 mine.  Could you make a small testcase that shows the problem, so we
 know we're on the same page?
 
 -- 
 Daniel Jacobowitz                           Debian GNU/Linux Developer
 MontaVista Software                         Carnegie Mellon University