[Bug testsuite/4466] New: system_func (0,1,1) fails on ppc64.

[Bug testsuite/4466] New: system_func (0,1,1) fails on ppc64.

[srinivasa at in dot ibm dot com]

My environment is kernel-2.6.21-rc7, systemtap-20070505, elfutils-0.125

Systemtap.samples/system_func.stp probes sys_open and prints "sys_open". But
system_func.stp taking some time to print "sys_open" and  hence system_func.exp
which greps "sys_open" is failing.

==================================================================
Running
/home/systemtap/tmp/stap_testing_200705070516/src/testsuite/systemtap.samples/system_func.e
xp ...
cat: __xyzzy123ABC__: No such file or directory
root

Waititing for processes to exit
DONE
FAIL: system_func (0,1,1)
testcase
/home/systemtap/tmp/stap_testing_200705070516/src/testsuite/systemtap.samples/system_func.
exp completed in 6 seconds
================================================
I have added extra 10 secs timer to wait for "sys_open" message. This fixes the
problem.
=======================================
#! stap

# test the system() function

probe kernel.function("sys_open") {
  # very inefficient. Testing only. DO NOT DO THIS
  msg="echo sys_open"
  system(msg)
}

probe begin {
  # should succeed
  system("whoami")

  # should fail
  system("cat __xyzzy123ABC__")

}

probe timer.ms(10000)
{
exit()
}

probe end {
  # should succeed
  system("echo DONE")
}
============================

-- 
           Summary: system_func (0,1,1) fails on ppc64.
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: testsuite
        AssignedTo: systemtap at sources dot redhat dot com
        ReportedBy: srinivasa at in dot ibm dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=4466

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug testsuite/4466] system_func (0,1,1) fails on ppc64.

[fche at redhat dot com]

------- Additional Comments From fche at redhat dot com  2007-05-07 14:46 -------
This is an example of a common race between "begin" and other probes.
Or at least, there used to be a race.  Now, begin probes are strictly
done first, and kprobes etc. are only registered afterward.

What could work here even better is an added timer probe that is bound
to run while the kprobes are active, and thus trigger sys_open soon:

probe timer.ms(500) { system ("cat DOESNOTEXIST") }


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4466

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug testsuite/4466] system_func (0,1,1) fails on ppc64.

[srinivasa at in dot ibm dot com]

------- Additional Comments From srinivasa at in dot ibm dot com  2007-05-29 04:48 -------
Bug is getting reproduced on systemtap-20070526 build.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |


http://sourceware.org/bugzilla/show_bug.cgi?id=4466

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug testsuite/4466] system_func (0,1,1) fails on ppc64.

[hunt at redhat dot com]

------- Additional Comments From hunt at redhat dot com  2007-09-17 17:06 -------
Looks like there is still a race. Checked in some additional fixes.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED


http://sourceware.org/bugzilla/show_bug.cgi?id=4466

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.