This is the mail archive of the gdb-prs@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

threads/1265: GNU/Linux LWP layer gets confused when thread-db says a thread has exited


>Number:         1265
>Category:       threads
>Synopsis:       GNU/Linux LWP layer gets confused when thread-db says a thread has exited
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Jun 29 19:18:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator:     drow@mvista.com
>Release:        5.3, 6.0-cvs
>Organization:
>Environment:
i386-pc-linux-gnu
>Description:
Right now lin-lwp assumes that for every attached LWP, there
is a (or maybe at least one) thread.  But the LWP doesn't
exit right atomicly with thread-db's death report.  In fact,
it runs some additional code, including a call to kill ().
So if you set a breakpoint on kill and run past thread
exit, you'll sometimes see:

Program received signal SIGTRAP, Trace/breakpoint trap.
[Switching to Thread 65541 (zombie)]
0x00000000 in ?? ()
>How-To-Repeat:
print-threads.exp in the testsuite demonstrates this.
>Fix:
Maybe by reporting LWPs without associated threads as "fake
threads", creating a new thread ID for them?
>Release-Note:
>Audit-Trail:
>Unformatted:


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]