This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap 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]

[Bug dyninst/15306] New: stapdyn IRPC on terminated process, child SEGV


http://sourceware.org/bugzilla/show_bug.cgi?id=15306

             Bug #: 15306
           Summary: stapdyn IRPC on terminated process, child SEGV
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: dyninst
        AssignedTo: systemtap@sourceware.org
        ReportedBy: jistone@redhat.com
                CC: adrian.m.negreanu@intel.com
    Classification: Unclassified


Probing process(".../libc.so.6").function("*") leads to dyninst complaining
about "IRPC on terminated process" and stapdyn reports the child SEGVed.

The IRPC failure is while stapdyn is attempting the oneTimeCode to run
stp_dyninst_shm_connect in the child process.

The SEGV is from enter_dyninst_uprobe doing atomic_read(session_state()), which
gdb presents as atomic_read (v=0x0).

So it seems that this probe is firing in early glibc setup code, before that
code for shm_connect has even had a chance to run.  The session_state is in shm
since it's shared across all processes, so the pointer will be invalid this
early.  We'll need to guard against this condition in the probe prologue.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


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