svn+ssh://pythondev@svn.python.org/python/trunk
........
r74972 | ronald.oussoren | 2009-09-20 20:54:16 +0200 (Sun, 20 Sep 2009) | 5 lines
Merge a newer version of libffi_osx, based on the
version of libffi in OSX 10.6.1.
This fixes issue6918
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74943 | georg.brandl | 2009-09-19 09:35:07 +0200 (Sa, 19 Sep 2009) | 1 line
#6944: the argument to PyArg_ParseTuple should be a tuple, otherwise a SystemError is set. Also clean up another usage of PyArg_ParseTuple.
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74921 | thomas.heller | 2009-09-18 22:05:44 +0200 (Fr, 18 Sep 2009) | 3 lines
Issue #4606: Passing 'None' if ctypes argtype is set to POINTER(...)
does now always result in NULL.
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74903 | georg.brandl | 2009-09-18 11:18:27 +0200 (Fr, 18 Sep 2009) | 1 line
#6938: "ident" is always a string, so use a format code which works.
........
........
r74901 | georg.brandl | 2009-09-18 11:14:52 +0200 (Fr, 18 Sep 2009) | 1 line
#6905: use better exception messages in inspect when the argument is of the wrong type.
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74873 | georg.brandl | 2009-09-17 13:48:31 +0200 (Do, 17 Sep 2009) | 1 line
#6844 followup: the warning when setting Exception.message was removed, do not test for it.
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74869 | georg.brandl | 2009-09-17 13:28:09 +0200 (Do, 17 Sep 2009) | 4 lines
Issue #6922: Fix an infinite loop when trying to decode an invalid
UTF-32 stream with a non-raising error handler like "replace" or "ignore".
........
........
r74853 | mark.dickinson | 2009-09-16 23:10:56 +0100 (Wed, 16 Sep 2009) | 5 lines
Issue #6713: Improve performance of str(n) and repr(n) for integers n
(up to 3.1 times faster in tests), by special-casing base 10 in
_PyLong_Format. (Backport of r74851 from py3k.)
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74845 | georg.brandl | 2009-09-16 22:30:09 +0200 (Mi, 16 Sep 2009) | 5 lines
#6844: do not emit DeprecationWarnings on access if Exception.message has been set by the user.
This works by always setting it in __dict__, except when it's implicitly set in __init__.
........
........
r74843 | mark.dickinson | 2009-09-16 21:26:31 +0100 (Wed, 16 Sep 2009) | 4 lines
Remove outdated include; this include was breaking OS X builds using
non-Apple gcc4.3 and gcc4.4 (because CoreFoundation/CoreFoundation.h
won't compile under non-Apple gcc).
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74841 | thomas.wouters | 2009-09-16 12:55:54 -0700 (Wed, 16 Sep 2009) | 23 lines
Fix issue #1590864, multiple threads and fork() can cause deadlocks, by
acquiring the import lock around fork() calls. This prevents other threads
from having that lock while the fork happens, and is the recommended way of
dealing with such issues. There are two other locks we care about, the GIL
and the Thread Local Storage lock. The GIL is obviously held when calling
Python functions like os.fork(), and the TLS lock is explicitly reallocated
instead, while also deleting now-orphaned TLS data.
This only fixes calls to os.fork(), not extension modules or embedding
programs calling C's fork() directly. Solving that requires a new set of API
functions, and possibly a rewrite of the Python/thread_*.c mess. Add a
warning explaining the problem to the documentation in the mean time.
This also changes behaviour a little on AIX. Before, AIX (but only AIX) was
getting the import lock reallocated, seemingly to avoid this very same
problem. This is not the right approach, because the import lock is a
re-entrant one, and reallocating would do the wrong thing when forking while
holding the import lock.
Will backport to 2.6, minus the tiny AIX behaviour change.
........
svn+ssh://pythondev@svn.python.org/python/trunk
........
r74798 | ronald.oussoren | 2009-09-15 20:33:33 +0200 (Tue, 15 Sep 2009) | 8 lines
MacOSX: detect the architectures supported by
Tk.framework and build _tkinter only for those
architectures.
This replaces the hardcoded solution that is no
longer valid now that 64-bit capable versions of
Tk are available on OSX.
........
........
r74750 | lars.gustaebel | 2009-09-12 12:28:15 +0200 (Sat, 12 Sep 2009) | 9 lines
Issue #6856: Add a filter keyword argument to TarFile.add().
The filter argument must be a function that takes a TarInfo
object argument, changes it and returns it again. If the
function returns None the TarInfo object will be excluded
from the archive.
The exclude argument is deprecated from now on, because it
does something similar but is not as flexible.
........
........
r74723 | mark.dickinson | 2009-09-08 21:20:19 +0100 (Tue, 08 Sep 2009) | 3 lines
Issue #6857: Fix Decimal formatting to be consistent with existing float
formatting: both are now right-aligned by default.
........
Ronald Oussoren
Ezio Melotti
Thomas Heller
Tarek Ziadé
Georg Brandl
Mark Dickinson
Thomas Wouters
Benjamin Peterson
Matthias Klose
Lars Gustäbel