Issue 4655 in pharo: Additional Socket woes

Status: FixProposed
Owner: [hidden email]
Labels: Milestone-1.3 Milestone-1.4

New issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

If allocating throwaway Sockets faster than finalization is triggered, one  
can easily run into the maxSemaphoreTable size.

For this reason, it's a good idea to do an explicit GC trying to free some  
slots, before deciding to actually increase the size and potentially lose  
signals.

See Issue 4505 for previous alterations connected to this.


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

Comment #1 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

SLICE-Issue-4655-Try-GC-to-finalize-externalObjects-before-increasing-maxSemaphore-size-HenrikSperreJohansen.1  
in  
Inbox.




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

Comment #2 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

SLICE-Issue-4655-1-3ExternalObjectFixesForCog-HenrikSperreJohansen.1

Changes from issue 4504 and this one for 1.3, from the latest version I  
could find on website.

They really are quite essential if you plan on using sockets or other  
object using externally registered objects on a Cog vm... We ship that as  
default in 1.3, right?


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

Updates:
        Labels: Importance-High

Comment #3 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

(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 #4 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Plah, SLICE's gone with recent SqueakSource outage...
Reupping new versions shortly, which additionally separate the locks on  
externalObject table for additions and removals.
That way, we don't block the finalizer when trying to free old slots as  
part of addition.


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

Comment #5 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Slices back up.

Now with more fixes, comments and other stuff than ever!


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

Comment #6 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Oh yes, forgot to mention in the commits, also included values in the  
vmParameterAt: comment for those (for now at least) specific to Cog. Mostly  
just a copy paste from the VMMaker version of the primitive.


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

Comment #7 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Thanks henrik.



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

Comment #8 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Hmm... 77 conflicts when merging in 1.4... (due to the move of methods to  
VirtualMachine)


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

Comment #9 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

That's impossible unless merging is critically flawed, there were like 15  
changed methods from the baseline.


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

Comment #10 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Are you sure you are not trying to merge the slice for 1.3? (there are 2  
separate ones)


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

Comment #11 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

Uploaded v3 of slice for 1.4, with the 4 conflicts resolved.


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

Comment #12 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

in 14098

TODO: 1.3


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

Updates:
        Status: Closed

Comment #13 on issue 4655 by [hidden email]: Additional Socket woes
http://code.google.com/p/pharo/issues/detail?id=4655

in 13299


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