On Thu, 14 Apr 2011 02:31:40 +0100 Federico Schwindt <fg...@lodoss.net> wrote: > Hi,
Hello, thanks for your response.
> I just checked my logs and it was hanging as well, apologies for the initial
> confusion.
> Makefile.inc looks good but I wouldn't link 2.7 to the tree yet (the
> Makefile bit).
> A few comments:
>
> * add the fix for CVE-2011-1521. This should be applied to the other
> versions; I've mailed some people about it as I cannot provide diffs atm.
I added 3 patches, please verify it's ok:
- patch-Lib_test_test_urllib2_py
- patch-Lib_urllib2_py
- patch-Lib_test_test_urllib_py
> * use --with-system-expat and remove the extra chunk in patch-setup_py
done.
> * start with library 0.0
done.
> * I'd prefer if test_sys.py and test_signal.py are updated to use
> @unittest.skipIf rather than commenting out the test or doing it inside the
> test. I'd like us to submit some of the diffs upstream and following their
> style would make things easier.
I'm ok, however it's we have in 2.6, so I think we could stay with this for now
(but need to be modify later).
> * some other tests in test_signal.py need 'openbsd4' added (I wouldn't
> worry with startswith('openbsd') and maybe even 'openbsd3') .
> * in patch-setup_py:
>
> ++ if platform.startswith('openbsd4'):
>
> you probably want "in 'openbsd4'" or starswith('openbsd'). I prefer to
> enumerate rather than using startswith, ymmv.
done, thanks.
> Some of the failing/hanging tests (in test_io.py, test_signal.py and
> test_socket.py) are very likely due to pthreads and must be investigated.
> The others should be relatively trivial to fix.
> In any case probably committing it would make things easier so if you fix
> some of the things I mentioned above I'm OK. We can fix the remaining things
> in the tree.
Yes, thanks a lot.
The diff has not changed, tarball has with your remarks.
Remi.
Makefile_inc_python_port_mk.diff
Description: Binary data
python-2.7.tar.gz
Description: Binary data
