Author pitrou
Recipients Albert.Strasheim, aljungberg, asksol, bquinlan, brian.curtin, gdb, gkcn, haypo, hongqn, jnoller, neologix, pitrou, vlasovskikh
Date 2011-05-13.13:15:04
SpamBayes Score 0.00114854
Marked as misclassified No
Message-id <1305292501.3561.24.camel@localhost.localdomain>
In-reply-to <BANLkTikFRc7d=Cs0Q5-623e=iiTE_R9o3A@mail.gmail.com>
Content
> But Lib/multiprocessing/connection.py does:
> 
>     def _send_bytes(self, buf):
>         # For wire compatibility with 3.2 and lower
>         n = len(buf)
>         self._send(struct.pack("=i", len(buf)))
>         # The condition is necessary to avoid "broken pipe" errors
>         # when sending a 0-length buffer if the other end closed the pipe.
>         if n > 0:
>             self._send(buf)
> 
> This is definitely not atomic.

Indeed, it isn't, Pipe objects are not meant to be safe against multiple
access. Queue objects (in multiprocessing/queues.py) use locks so they
are safe.
History
Date User Action Args
2011-05-13 13:15:05pitrousetrecipients: + pitrou, bquinlan, haypo, jnoller, hongqn, brian.curtin, asksol, vlasovskikh, neologix, gdb, Albert.Strasheim, aljungberg, gkcn
2011-05-13 13:15:04pitroulinkissue9205 messages
2011-05-13 13:15:04pitroucreate