The attached patch to python 2.51 fixes the problem for me. I have submitted it to bugs.python.org.
On 21/12/2007, Richard Oudkerk <[EMAIL PROTECTED]> wrote:
> I could not reproduce the bug with the first test file I wrote, but a
> modified version reproduces the bug reliably for me with python2.4-dbg
> from ubuntu edgy x86:
>
> import time, os, thread
>
> def f():
> print "I am thread %d in process %d" % (thread.get_ident(),os.getpid())
> time.sleep(1)
>
> ident = thread.start_new_thread(f, ())
> print "started thread %d in process %d" % (ident, os.getpid())
> time.sleep(0.5) # give thread a chance to start
>
> pid = os.fork()
> if pid == 0:
> ident = thread.start_new_thread(f, ())
> print "started thread %d in process %d" % (ident, os.getpid())
> time.sleep(0.5) # give thread a chance to start
> os._exit(0)
>
> os.waitpid(pid, 0)
>
> The output is
>
> started thread -1210782816 in process 18687
> I am thread -1210782816 in process 18687
> started thread -1210782816 in process 18689
> Fatal Python error: Invalid thread state for this thread
> [5817 refs]
>
> I think the problem has to do with the fact that the child process's
> subthread has the same id as the parent process's subthread, so the
> thread local storage implementation gets confused.
>
> I will file a bug at bugs.python.org.
>
> Cheers,
>
> Richard.
>
>
> On 21/12/2007, Sandro Tosi <[EMAIL PROTECTED]> wrote:
> > Package: python-all-dbg
> > Version: 2.4.4-6
> > Severity: normal
> >
> > Hello,
> > while packaging processing (a python module to use processes like
> > threads) I've faced a strange error: running tests with python2.{4,5} no
> > problem, but using -dbg version an error comes out "Fatal Python error:
> > Invalid thread state for this thread".
> >
> > Richard (upstream author of processing) kindly provided me a small test
> > python program to spot the problem (dbg.py attached to this report):
> >
> > [EMAIL PROTECTED]:~$ python2.4 ~/tmp/dbg.py
> > parent sleeping ...
> > child sleeping ...
> > ... child waking
> > child exiting
> > ... parent waking
> > parent exiting
> > [EMAIL PROTECTED]:~$ python2.5 ~/tmp/dbg.py
> > parent sleeping ...
> > child sleeping ...
> > ... parent waking
> > ... child waking
> > child exiting
> > parent exiting
> > [EMAIL PROTECTED]:~$ python2.4-dbg ~/tmp/dbg.py
> > parent sleeping ...
> > Fatal Python error: Invalid thread state for this thread
> > ... parent waking
> > parent exiting
> > [7951 refs]
> > [EMAIL PROTECTED]:~$ python2.5-dbg ~/tmp/dbg.py
> > parent sleeping ...
> > Fatal Python error: Invalid thread state for this thread
> > ... parent waking
> > parent exiting
> > [9809 refs]
> >
> > The behaviour of python2.{4,5} is the intended one, while with -dbg the
> > error appears.
> >
> > It could be due to my arch (amd64), an issue on packaging or something
> > deeper in python code.
> >
> > Thanks,
> > Sandro
> >
> > -- System Information:
> > Debian Release: lenny/sid
> > APT prefers unstable
> > APT policy: (500, 'unstable'), (500, 'testing')
> > Architecture: amd64 (x86_64)
> >
> > Kernel: Linux 2.6.22-2-amd64 (SMP w/4 CPU cores)
> > Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
> > Shell: /bin/sh linked to /bin/bash
> >
> > Versions of packages python-all-dbg depends on:
> > ii python-all 2.4.4-6 Package depending on all
> > supported
> > ii python-dbg 2.4.4-6 Debug Build of the Python
> > Interpre
> > ii python2.4-dbg 2.4.4-6 Debug Build of the Python
> > Interpre
> > ii python2.5-dbg 2.5.1-5 Debug Build of the Python
> > Interpre
> >
> > python-all-dbg recommends no packages.
> >
> > -- no debconf information
> >
> >
>
fork-thread-patch
Description: Binary data
