public inbox for frysk-bugzilla@sourceware.org
help / color / mirror / Atom feed
* [Bug general/4174] New: Frysk blows up with heavily threading programs.
@ 2007-03-08 21:59 mcvet at redhat dot com
  2007-03-08 22:00 ` [Bug general/4174] " mcvet at redhat dot com
                   ` (5 more replies)
  0 siblings, 6 replies; 7+ messages in thread
From: mcvet at redhat dot com @ 2007-03-08 21:59 UTC (permalink / raw)
  To: frysk-bugzilla

Probably more of a stress-test edgecase; but when trying to open frysk on the 
following program from the GUI, it appears to hang for a few seconds, and then 
fail with the following:

java.lang.RuntimeException: destroyed
   at frysk.proc.LinuxPtraceTaskState$16.handleAddObservation(FryskGui)
   at frysk.proc.Task.handleAddObservation(FryskGui)
   at frysk.proc.TaskObservation.handleAdd(FryskGui)
   at frysk.proc.LinuxPtraceProcState$Attaching.allAttached(FryskGui)
   at 
frysk.proc.LinuxPtraceProcState$Attaching$ToOtherTasks.handleTaskAttachCompleted(FryskGui)
   at frysk.proc.Proc$4.execute(FryskGui)
   at frysk.event.EventLoop.runEventLoop(FryskGui)
   at frysk.event.EventLoop.run(FryskGui)
   at frysk.gui.Gui$4.run(FryskGui)
   at java.lang.Thread.run(libgcj.so.7rh)

The test program endlessly creates and destroys threads.

-- 
           Summary: Frysk blows up with heavily threading programs.
           Product: frysk
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: general
        AssignedTo: frysk-bugzilla at sourceware dot org
        ReportedBy: mcvet at redhat dot com
                CC: cagney at redhat dot com,mark at klomp dot org,mcvet at
                    redhat dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug general/4174] Frysk blows up with heavily threading programs.
  2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
@ 2007-03-08 22:00 ` mcvet at redhat dot com
  2007-03-28  8:55 ` mark at klomp dot org
                   ` (4 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: mcvet at redhat dot com @ 2007-03-08 22:00 UTC (permalink / raw)
  To: frysk-bugzilla


------- Additional Comments From mcvet at redhat dot com  2007-03-08 22:00 -------
Created an attachment (id=1604)
 --> (http://sourceware.org/bugzilla/attachment.cgi?id=1604&action=view)
Thread test

gcc -O0 -g -o thread_test -lpthread thread_test.c should do it.

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug general/4174] Frysk blows up with heavily threading programs.
  2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
  2007-03-08 22:00 ` [Bug general/4174] " mcvet at redhat dot com
@ 2007-03-28  8:55 ` mark at klomp dot org
  2007-03-28 12:09 ` mark at klomp dot org
                   ` (3 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: mark at klomp dot org @ 2007-03-28  8:55 UTC (permalink / raw)
  To: frysk-bugzilla


------- Additional Comments From mark at klomp dot org  2007-03-28 09:55 -------
The problem here is that frysk.rt.RunState throws a hissy fit when it sees any
addObservation failing:

  public void addFailed (Object o, Throwable w)
  {
    w.printStackTrace();
    stateProc.requestAbandonAndRunEvent(new RequestStopEvent(Manager.eventLoop)\
);
    System.exit(1);
  }

The TaskState really cannot help that the request for an observer addition was
done just at the moment the Task was about to be destroyed.

Is there anything the TaskState can do to help RunState recover in this case?

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug general/4174] Frysk blows up with heavily threading programs.
  2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
  2007-03-08 22:00 ` [Bug general/4174] " mcvet at redhat dot com
  2007-03-28  8:55 ` mark at klomp dot org
@ 2007-03-28 12:09 ` mark at klomp dot org
  2007-03-29 17:02 ` swagiaal at redhat dot com
                   ` (2 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: mark at klomp dot org @ 2007-03-28 12:09 UTC (permalink / raw)
  To: frysk-bugzilla


------- Additional Comments From mark at klomp dot org  2007-03-28 13:09 -------
As a workaround for now I just treat Tasks falling through addFailed() as if
they did an updateExecuted() so they disappear from the list of Tasks being
waited on.

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug general/4174] Frysk blows up with heavily threading programs.
  2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
                   ` (2 preceding siblings ...)
  2007-03-28 12:09 ` mark at klomp dot org
@ 2007-03-29 17:02 ` swagiaal at redhat dot com
  2007-03-29 19:54 ` pmuldoon at redhat dot com
  2007-03-29 19:59 ` mcvet at redhat dot com
  5 siblings, 0 replies; 7+ messages in thread
From: swagiaal at redhat dot com @ 2007-03-29 17:02 UTC (permalink / raw)
  To: frysk-bugzilla


------- Additional Comments From swagiaal at redhat dot com  2007-03-29 18:02 -------
isnt System.exit(1) kind of harsh.
It sounds like TaskState had a good reason for the add to fail, so RunState
should handle it and move on.

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug general/4174] Frysk blows up with heavily threading programs.
  2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
                   ` (3 preceding siblings ...)
  2007-03-29 17:02 ` swagiaal at redhat dot com
@ 2007-03-29 19:54 ` pmuldoon at redhat dot com
  2007-03-29 19:59 ` mcvet at redhat dot com
  5 siblings, 0 replies; 7+ messages in thread
From: pmuldoon at redhat dot com @ 2007-03-29 19:54 UTC (permalink / raw)
  To: frysk-bugzilla


------- Additional Comments From pmuldoon at redhat dot com  2007-03-29 20:54 -------
Can I suggest using never using System.exit in any case from the UI? Suggest using

Gui.quitFrysk()

which is a static method. It shuts down the GTK eventloop with mainQuit.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pmuldoon at redhat dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug general/4174] Frysk blows up with heavily threading programs.
  2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
                   ` (4 preceding siblings ...)
  2007-03-29 19:54 ` pmuldoon at redhat dot com
@ 2007-03-29 19:59 ` mcvet at redhat dot com
  5 siblings, 0 replies; 7+ messages in thread
From: mcvet at redhat dot com @ 2007-03-29 19:59 UTC (permalink / raw)
  To: frysk-bugzilla


------- Additional Comments From mcvet at redhat dot com  2007-03-29 20:59 -------
Well this is in RT so potentially it could be just fhpd running, but the 
System.exit was just an oversight - there are certainly better ways of handling 
that problem than just quitting. I'll look into a better solution.

-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=4174

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


^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2007-03-29 19:59 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-03-08 21:59 [Bug general/4174] New: Frysk blows up with heavily threading programs mcvet at redhat dot com
2007-03-08 22:00 ` [Bug general/4174] " mcvet at redhat dot com
2007-03-28  8:55 ` mark at klomp dot org
2007-03-28 12:09 ` mark at klomp dot org
2007-03-29 17:02 ` swagiaal at redhat dot com
2007-03-29 19:54 ` pmuldoon at redhat dot com
2007-03-29 19:59 ` mcvet at redhat dot com

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).