Open Bug 307572 Opened 19 years ago Updated 2 years ago

cvar2 test fails on HP-UX

Categories

(NSPR :: NSPR, defect)

4.6.1
HP
HP-UX
defect

Tracking

(Not tracked)

People

(Reporter: jason.m.reid, Unassigned)

Details

charm[svbld]:/share/builds/mccrel3/security/securitytip/builds/20050908.1/wozzeck_Solaris8/mozilla/nsprpub/HP-UXB.11.11_DBG.OBJ/pr/tests>
./cvar2

CondVar Test:

Simple test creates several local and global threads; half use a single,
shared condvar, and the other half have their own condvar.  The main
thread then loops notifying them to wakeup.

The timeout test is very similar except that the threads are not
notified.  They will all wakeup on a 1 second timeout.

The mixed test combines the simple test and the timeout test; every
third thread is notified, the other threads are expected to timeout
correctly.

Lastly, the combined test creates a thread for each of the above three
cases and they all run simultaneously.

This test is run with 5, 10, 15, and 20 threads of each type.


5 Thread tests
           Condvar simple test shared UU:  90.00 usec
           Condvar simple test shared UK:  70.00 usec
             Condvar simple test priv UU: 190.00 usec
             Condvar simple test priv UK: 190.00 usec
                 Condvar simple test All: 440.00 usec
                    Condvar timeout test: 910.00 usec

10 Thread tests
           Condvar simple test shared UU: 140.00 usec
           Condvar simple test shared UK: 140.00 usec
             Condvar simple test priv UU: 340.00 usec
             Condvar simple test priv UK: 310.00 usec
                 Condvar simple test All: 900.00 usec
                    Condvar timeout test: 1010.00 usec

15 Thread tests
           Condvar simple test shared UU: 220.00 usec
           Condvar simple test shared UK: 210.00 usec
             Condvar simple test priv UU: 470.00 usec
             Condvar simple test priv UK: 490.00 usec
                 Condvar simple test All: 1490.00 usec
                    Condvar timeout test: 1190.00 usec

20 Thread tests
           Condvar simple test shared UU: 280.00 usec
           Condvar simple test shared UK: 270.00 usec
             Condvar simple test priv UU: 660.00 usec
             Condvar simple test priv UK: 620.00 usec
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11
error creating thread
: PR_INSUFFICIENT_RESOURCES_ERROR(-5974), oserror = 11


This test fails on all cases. It loops until killed.

r> grep cvar2 *HP*summary
charm.1-HP-UXB.11.11_DBG.OBJ-.summary:cvar2./runtests.sh[261]: 20558 Terminated
charm.2-HP-UXB.11.11_OPT.OBJ-.summary:cvar2./runtests.sh[261]: 1806 Terminated
charm.3-HP-UXB.11.11_64_DBG.OBJ-.summary:cvar2./runtests.sh[261]: 2672 Terminated
charm.4-HP-UXB.11.11_64_OPT.OBJ-.summary:cvar2./runtests.sh[261]: 9473 Terminated

$ uname -a
HP-UX charm B.11.11 U 9000/800 1549706547 unlimited-user license
QA Contact: wtchang → nspr
Severity: normal → S3

The bug assignee is inactive on Bugzilla, so the assignee is being reset.

Assignee: wtc → nobody
You need to log in before you can comment on or make changes to this bug.