Re: gdb/489: gdb changes g++ static initialization values

[Daniel Jacobowitz <drow at mvista dot com>]

The following reply was made to PR gdb/489; it has been noted by GNATS.

From: Daniel Jacobowitz <drow@mvista.com>
To: pbd@op.net
Cc: gdb-gnats@sources.redhat.com
Subject: Re: gdb/489: gdb changes g++ static initialization values
Date: Mon, 15 Apr 2002 23:21:50 -0400

 On Tue, Apr 16, 2002 at 02:30:16AM -0000, pbd@op.net wrote:
 > 
 > >Number:         489
 > >Category:       gdb
 > >Synopsis:       gdb changes g++ static initialization values
 > >Confidential:   no
 > >Severity:       critical
 > >Priority:       medium
 > >Responsible:    unassigned
 > >State:          open
 > >Class:          sw-bug
 > >Submitter-Id:   net
 > >Arrival-Date:   Mon Apr 15 19:38:01 PDT 2002
 > >Closed-Date:
 > >Last-Modified:
 > >Originator:     Paul Davis
 > >Release:        5.1.0.1
 > >Organization:
 > >Environment:
 > linux (kernel 2.4.19pre6, glibc 2.2.5, gcc/g++ 2.95.3)
 > >Description:
 > running a large progam under gdb changes the value
 > of some floating point values used during static
 > member initialization. a call to isnan() returns a
 > different value depending on whether the app is run
 > on its own or as a child of gdb.
 > >How-To-Repeat:
 > i cannot supply a small example of the bug. the
 > entire source tree is available via CVS, buts its
 > many 10's of thousands of lines. let me know if you would
 > like more details or help in tracking this down.
 
 Does this application use threads, perhaps?
 
 Especially if so, please try a recent CVS snapshot
 (http://sources.redhat.com/gdb/).  A floating-point bug was fixed
 recently.
 
 -- 
 Daniel Jacobowitz                           Carnegie Mellon University
 MontaVista Software                         Debian GNU/Linux Developer