[Stackless] Differences between tasklet pickling in Stackless Python and Pypy
Christian Tismer
tismer at stackless.com
Fri Jul 6 17:31:45 CEST 2012
Hi Alain,
On 7/6/12 5:06 PM, Alain Poirier wrote:
> Hi Christian,
>
> Pushing my tasklets and pickling tests on Pypy further, I see a lots
> of stuffs are pickled by Pypy when a single tasklet is pickled. For
> example the main tasklet is pickled too. In this case, I'm not really
> sure what happens on unpickling, if this can't lead to problems.
> What do you think?
>
> Here is a script displaying the serialized tasklets and showing the
> problem. On Stackless Python, only one tasklet is pickled :
>
> ...
> -----
> Serialized tasklets set([<stackless.tasklet object at 0x10047ea60>])
>
>
> On Pypy, the main tasklet is pickled too :
>
> ...
> -----
> Serialized tasklets set([<tasklet[, 1]>, <tasklet[main, 0]>])
Yes, this is related to the same wrong design principle.
The main tasklet should _not_ be included in the pickle in the first
place. Instead, it is crucial that a fresh one gets created, instead.
I think it makes no sense to investigate any further on this code --
it is simply known to be not perfect, to say the least.
I will try to provide a fix/re-write.
Cheers - Chris
--
Christian Tismer :^) <mailto:tismer at stackless.com>
tismerysoft GmbH : Have a break! Take a ride on Python's
Karl-Liebknecht-Str. 121 : *Starship* http://starship.python.net/
14482 Potsdam : PGP key -> http://pgp.uni-mainz.de
work +49 173 24 18 776 mobile +49 173 24 18 776 fax n.a.
PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04
whom do you want to sponsor today? http://www.stackless.com/
More information about the Stackless
mailing list