i386/70330: Re-Open 33262? - gdb does not handle pending signals
properly when single stepping during a remote debug session.
Ernie Smallis
esmallis at stbernard.com
Wed Aug 11 15:50:26 PDT 2004
>Number: 70330
>Category: i386
>Synopsis: Re-Open 33262? - gdb does not handle pending signals properly when single stepping during a remote debug session.
>Confidential: no
>Severity: critical
>Priority: high
>Responsible: freebsd-i386
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: sw-bug
>Submitter-Id: current-users
>Arrival-Date: Wed Aug 11 22:50:22 GMT 2004
>Closed-Date:
>Last-Modified:
>Originator: Ernie Smallis
>Release: 5.2.1, 5.2.CURRENT
>Organization:
St. Bernard Software (San Diego, CA)
>Environment:
uname -snrmv
FreeBSD gromit.rapid.stbernard.com 5.2.1-RELEASE FreeBSD 5.2.1-RELEASE #0: Mon Feb 23 20:45:55 GMT 2004 root at wv1u.btc.adaptec.com:/usr/obj/usr/src/sys/GENERIC i386
>Description:
I am seeing a similar problem in both 5.2.1 and 5.2 current.
I am receiving a SIGSEGV for gdb during a remote kernel debug session.
The messages are:
1. From infrun.c Program received signal SIGSEGV, Segmentation fault.
2. From remote.c Can't send signals to this remote system. SIGSEGV not sent.
#2 I can understand because the host gdb is bombing.
#1 is possibly bombing for a stack overflow?
When the host is signalled, process_event_stop_test( ) in infrun.c is called and I expect that either remote_wait( ) or remote_async_wait( ) are called as part of telling the world that gdb has crashed.
It also appears that handle_inferior_event( ) is where we pick this up. I haven't debugged any further and would like to know your opinion and get some direction.
>How-To-Repeat:
If you use the remote gdb to debug a target kernel and you continually step and/or set a breakpoint and continue in any way, you will receive the above messages.
Thanks for any info./help you can provide.
Ernie ;-)
Ernest Smallis
Sr. Software Engineer
RAPID Development
esmallis at stbernard.com
858-524-2137 (PDT)
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted:
More information about the freebsd-i386
mailing list