Issue 5663 in pharo: Snapshot from background blocks image

Status: New
Owner: ----

New issue 5663 by [hidden email]: Snapshot from background blocks  
image
http://code.google.com/p/pharo/issues/detail?id=5663

Hi guys,

This code reproduces the blockade in fresh 1.4 OneClick (with or without
Transcript window open):

        [
          (Delay forSeconds: 5) wait.
           SmalltalkImage current saveSession.
        ] forkAt: Processor userBackgroundPriority

Note that fork on default priority does not block while on user
background priority it blocks. It does not always block immediately
(sometimes yes sometimes not) but it after image restart it is always
blocked.

But if you periodically write to Transcript from process on user
background priority, it does not block, even if you heavily browse
around, write to transcript, do manual snapshot:

        [ [true] whileTrue:
            [(Delay forMilliseconds: 500) wait.
            Transcript show: '.tst.']
        ] forkAt: Processor userBackgroundPriority

So, it seems not only thread-safe Transcript but something with snapshot
(from a background priority, why?) is in play here.

SmalltalkImage>>snapshot:andQuit: , shouldn’t run it on highest priority
to block all other processes during the preparation for snapshot?


This is Pharo 1.4 OneClick on openSuse 12.1 (x86_64), kernel  
3.1.9-1.4-desktop

Original discussion:  
http://forum.world.st/Unresponsive-1-4-image-tp4574693p4576377.html


Best regards
Janko


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Comment #1 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

I also can reproduce the same bug on Pharo 1.4 OneClick  on mac osx lion  
10.7.3, Cog vm


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Updates:
        Status: Workneeded
        Labels: Type-Bug Milestone-1.4 Milestone-2.0

Comment #2 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

(No comment was entered for this change.)


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Comment #3 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

seems to be related to Issue 5868


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Comment #4 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

Issue 5868 has been merged into this issue.


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Comment #6 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

We could block the current thread using semaphores and always spawn the  
image saving in a new thread with the appropriate priority.


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Updates:
        Cc: [hidden email]

Comment #7 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

(No comment was entered for this change.)


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Comment #8 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

hmm.. this one requires some thought (since debugging won't help at all).
What if you fork at higher priority? Does it blocks or not?

I guess, what happens is that when you saving image in background priority  
thread,
once you restart it, during startup sequence right after starting up Delay  
process (which is highest priority), the scheduler quickly switching to UI  
process, because it has higher priority, leaving the rest startup  
procedures unfinished (like Sensor startup etc), at so UI process doesn't  
receives any events.

But then, by all means, it should activate the background process  
eventually, to continue running it (and finish startup), unless it keeps  
running endlessly in some higher priority process. If i right, in addition  
to blockade, you should observe 100% CPU load as well.





_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Comment #9 on issue 5663 by [hidden email]: Snapshot from background  
blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

Following Marcus' request to review the importance of issues for 2.0,
this seems fairly advanced and unusual for most users to do.  Apparently no  
action since September, vote -> 3.0.


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker

Updates:
        Labels: -Milestone-1.4 -Milestone-2.0 Milestone-3.0

Comment #10 on issue 5663 by [hidden email]: Snapshot from  
background blocks image
http://code.google.com/p/pharo/issues/detail?id=5663

(No comment was entered for this change.)


_______________________________________________
Pharo-bugtracker mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-bugtracker