[Stackless] Stackless crash

Jeff Senn senn at maya.com
Fri Jun 30 20:37:25 CEST 2006



Richard Tew wrote:
> On 6/30/06, Jeff Senn <senn at maya.com> wrote:
>> ... seems a little "wrong" for the channel to hold a reference
>> to itself (but I really unsure how to the do the opposite: have the 
>> tasklet
>> hold the reference to the channel) and other things sometimes *seem* 
>> wrong
>> about stackless too :-) :-)
>>
>> The patch is pretty easy (included below), but I'm not *certain* it 
>> doesn't
>> change some basic refcnt assumption somewhere...
> 
> Well, I don't know much about ref counting, which is why I was going
> to try and avoid it.  But this is the only solution I can think of which
> should work.  Well, except for forcing receive() to do hard switching
> instead of soft switching.

Yuck! Please don't! :-) :-)

Aside: my switch back to stackless (after using my own greenlet-based
implementation while waiting for a modern stackless) seems like
it improved my performance generally.
That is: I think that the soft-switching is more efficient for most things
I'm doing than all that memory-copying of stacks.  Yay!

> I think I will wait for Christian's opinion on this before I apply
> the patch given the limited case in which the problem occurs.

Yep. Sounds right.

-Jas

_______________________________________________
Stackless mailing list
Stackless at stackless.com
http://www.stackless.com/mailman/listinfo/stackless



More information about the Stackless mailing list