This is the mail archive of the
gdb-prs@sources.redhat.com
mailing list for the GDB project.
backtrace/1660: GDB 6.1 can't display backtrace from Solaris 8 core files
- From: rdiez_antispam1 at activenav dot com
- To: gdb-gnats at sources dot redhat dot com
- Date: 28 May 2004 09:40:08 -0000
- Subject: backtrace/1660: GDB 6.1 can't display backtrace from Solaris 8 core files
- Reply-to: rdiez_antispam1 at activenav dot com
>Number: 1660
>Category: backtrace
>Synopsis: GDB 6.1 can't display backtrace from Solaris 8 core files
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: unassigned
>State: open
>Class: change-request
>Submitter-Id: net
>Arrival-Date: Fri May 28 09:48:01 UTC 2004
>Closed-Date:
>Last-Modified:
>Originator: rdiez_antispam1@activenav.com
>Release: unknown-1.0
>Organization:
>Environment:
Crashed executable and GDB both built with GCC 3.4.0, binutils 2.14
>Description:
I've got a core file from an executable built with GCC 3.4.0 / binutils 2.14 on Solaris 8. With GDB 6.1, the call stack is wrong, note the warning at the bottom:
#0 0xff259764 in __sigprocmask () from /usr/lib/libthread.so.1
#1 0xff24e978 in _resetsig () from /usr/lib/libthread.so.1
#2 0xff24e118 in _sigon () from /usr/lib/libthread.so.1
#3 0xff251158 in _thrp_kill () from /usr/lib/libthread.so.1
#4 0xff04b908 in raise () from /usr/lib/libc.so.1
#5 0xff0358f4 in abort () from /usr/lib/libc.so.1
#6 0xff035b98 in _assert () from /usr/lib/libc.so.1
#7 0x00019120 in ?? ()
Previous frame identical to this frame (corrupt stack?)
I built GDB 6.1 myself with the same GCC and binutils. The configuration is "sparc-sun-solaris2.8".
The executable that asserted was a debug build, so there were no optimisations enabled.
However, with an older GDB 5.3 that came from www.sunfreeware.com, the call stack is slightly dodgy at the end, but still usable.
I've always had stack backtrace problems with GDB under Solaris 7 or 8. I'm not an expert, but I know now a fair bit, and I find the whole GCC/GDB/Solaris experience frustrating.
There doesn't appear to be anything on the GDB or GCC documentation to tell you what you need to do to get everything working.
My requests are:
1) Can anyone either fix the Solaris backtrace problems, or at least aknowledge there are some?
2) Can someone add to the GDB documentation some notes about what to do to get the Solaris backtrace to work properly? I suspect there will be some spooky combination of compiler/system/debugger/build settings.
>How-To-Repeat:
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted: