This is the mail archive of the
gdb-prs@sources.redhat.com
mailing list for the GDB project.
gdb/258: SIGSEGV error only in gdb, not outside; File pointer garbled by gdb.
- From: jstoffels at xs4all dot nl
- To: gdb-gnats at sources dot redhat dot com
- Date: 14 Dec 2001 12:30:51 -0000
- Subject: gdb/258: SIGSEGV error only in gdb, not outside; File pointer garbled by gdb.
- Reply-to: jstoffels at xs4all dot nl
>Number: 258
>Category: gdb
>Synopsis: SIGSEGV error only in gdb, not outside; File pointer garbled by gdb.
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: unassigned
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Fri Dec 14 04:38:01 PST 2001
>Closed-Date:
>Last-Modified:
>Originator: jstoffels@xs4all.nl
>Release: cygwin gnu gdb 5.0 20010428-3
>Organization:
>Environment:
Windows 98, Pentium 3, cygwin
>Description:
I read a datastructure from an ASCII file from a file pointer to the input file, then copy it to stdout. Outside gdb this works fine. I pass the FILE * down through several functions.
When I include an object file I do not refer to, an error occurs later in gdb; By the way, the object file needs -lcurses and the loader warns about resolving COLS, LINES and stdscr to some other symbols.
When I run the program with the unused object file from the command line, no problem occurs.
When I run it inside gdb I get a SIGSEGV, and my FILE * value has changed through the stack. High up it is allright, at the crash site the FILE * has become nonsense, clearly causing the SIGSEGV on usage of the FILE *.
Has anyone a suggestion how to track the problem further?
>How-To-Repeat:
I just hope someone can give me a hint for the hunt at least?
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted: