Merged revisions 46490-46494,46496,46498,46500,46506,46521,46538,46558,46563-46567,46570-46571,46583,46593,46595-46598,46604,46606,46609-46753 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk
........
r46610 | martin.v.loewis | 2006-06-03 09:42:26 +0200 (Sat, 03 Jun 2006) | 2 lines
Updated version (win32-icons2.zip) from #1490384.
........
r46612 | andrew.kuchling | 2006-06-03 20:09:41 +0200 (Sat, 03 Jun 2006) | 1 line
[Bug #1472084] Fix description of do_tag
........
r46614 | andrew.kuchling | 2006-06-03 20:33:35 +0200 (Sat, 03 Jun 2006) | 1 line
[Bug #1475554] Strengthen text to say 'must' instead of 'should'
........
r46616 | andrew.kuchling | 2006-06-03 20:41:28 +0200 (Sat, 03 Jun 2006) | 1 line
[Bug #1441864] Clarify description of 'data' argument
........
r46617 | andrew.kuchling | 2006-06-03 20:43:24 +0200 (Sat, 03 Jun 2006) | 1 line
Minor rewording
........
r46619 | andrew.kuchling | 2006-06-03 21:02:35 +0200 (Sat, 03 Jun 2006) | 9 lines
[Bug #1497414] _self is a reserved word in the WATCOM 10.6 C compiler.
Fix by renaming the variable.
In a different module, Neal fixed it by renaming _self to self. There's
already a variable named 'self' here, so I used selfptr.
(I'm committing this on a Mac without Tk, but it's a simple search-and-replace.
<crosses fingers>, so I'll watch the buildbots and see what happens.)
........
r46621 | fredrik.lundh | 2006-06-03 23:56:05 +0200 (Sat, 03 Jun 2006) | 5 lines
"_self" is a said to be a reserved word in Watcom C 10.6. I'm
not sure that's really standard compliant behaviour, but I guess
we have to fix that anyway...
........
r46622 | andrew.kuchling | 2006-06-04 00:44:42 +0200 (Sun, 04 Jun 2006) | 1 line
Update readme
........
r46623 | andrew.kuchling | 2006-06-04 00:59:23 +0200 (Sun, 04 Jun 2006) | 1 line
Drop 0 parameter
........
r46624 | andrew.kuchling | 2006-06-04 00:59:59 +0200 (Sun, 04 Jun 2006) | 1 line
Some code tidying; use curses.wrapper
........
r46625 | andrew.kuchling | 2006-06-04 01:02:15 +0200 (Sun, 04 Jun 2006) | 1 line
Use True; value returned from main is unused
........
r46626 | andrew.kuchling | 2006-06-04 01:07:21 +0200 (Sun, 04 Jun 2006) | 1 line
Use true division, and the True value
........
r46627 | andrew.kuchling | 2006-06-04 01:09:58 +0200 (Sun, 04 Jun 2006) | 1 line
Docstring fix; use True
........
r46628 | andrew.kuchling | 2006-06-04 01:15:56 +0200 (Sun, 04 Jun 2006) | 1 line
Put code in a main() function; loosen up the spacing to match current code style
........
r46629 | andrew.kuchling | 2006-06-04 01:39:07 +0200 (Sun, 04 Jun 2006) | 1 line
Use functions; modernize code
........
r46630 | andrew.kuchling | 2006-06-04 01:43:22 +0200 (Sun, 04 Jun 2006) | 1 line
This demo requires Medusa (not just asyncore); remove it
........
r46631 | andrew.kuchling | 2006-06-04 01:46:36 +0200 (Sun, 04 Jun 2006) | 2 lines
Remove xmlrpc demo -- it duplicates the SimpleXMLRPCServer module.
........
r46632 | andrew.kuchling | 2006-06-04 01:47:22 +0200 (Sun, 04 Jun 2006) | 1 line
Remove xmlrpc/ directory
........
r46633 | andrew.kuchling | 2006-06-04 01:51:21 +0200 (Sun, 04 Jun 2006) | 1 line
Remove dangling reference
........
r46634 | andrew.kuchling | 2006-06-04 01:59:36 +0200 (Sun, 04 Jun 2006) | 1 line
Add more whitespace; use a better socket name
........
r46635 | tim.peters | 2006-06-04 03:22:53 +0200 (Sun, 04 Jun 2006) | 2 lines
Whitespace normalization.
........
r46637 | tim.peters | 2006-06-04 05:26:02 +0200 (Sun, 04 Jun 2006) | 16 lines
In a PYMALLOC_DEBUG build obmalloc adds extra debugging info
to each allocated block. This was using 4 bytes for each such
piece of info regardless of platform. This didn't really matter
before (proof: no bug reports, and the debug-build obmalloc would
have assert-failed if it was ever asked for a chunk of memory
>= 2**32 bytes), since container indices were plain ints. But after
the Py_ssize_t changes, it's at least theoretically possible to
allocate a list or string whose guts exceed 2**32 bytes, and the
PYMALLOC_DEBUG routines would fail then (having only 4 bytes
to record the originally requested size).
Now we use sizeof(size_t) bytes for each of a PYMALLOC_DEBUG
build's extra debugging fields. This won't make any difference
on 32-bit boxes, but will add 16 bytes to each allocation in
a debug build on a 64-bit box.
........
r46638 | tim.peters | 2006-06-04 05:38:04 +0200 (Sun, 04 Jun 2006) | 4 lines
_PyObject_DebugMalloc(): The return value should add
2*sizeof(size_t) now, not 8. This probably accounts for
current disasters on the 64-bit buildbot slaves.
........
r46639 | neal.norwitz | 2006-06-04 08:19:31 +0200 (Sun, 04 Jun 2006) | 1 line
SF #1499797, Fix for memory leak in WindowsError_str
........
r46640 | andrew.macintyre | 2006-06-04 14:31:09 +0200 (Sun, 04 Jun 2006) | 2 lines
Patch #1454481: Make thread stack size runtime tunable.
........
r46641 | andrew.macintyre | 2006-06-04 14:59:59 +0200 (Sun, 04 Jun 2006) | 2 lines
clean up function declarations to conform to PEP-7 style.
........
r46642 | martin.blais | 2006-06-04 15:49:49 +0200 (Sun, 04 Jun 2006) | 15 lines
Fixes in struct and socket from merge reviews.
- Following Guido's comments, renamed
* pack_to -> pack_into
* recv_buf -> recv_into
* recvfrom_buf -> recvfrom_into
- Made fixes to _struct.c according to Neal Norwitz comments on the checkins
list.
- Converted some ints into the appropriate -- I hope -- ssize_t and size_t.
........
r46643 | ronald.oussoren | 2006-06-04 16:05:28 +0200 (Sun, 04 Jun 2006) | 3 lines
"Import" LDFLAGS in Mac/OSX/Makefile.in to ensure pythonw gets build with
the right compiler flags.
........
r46644 | ronald.oussoren | 2006-06-04 16:24:59 +0200 (Sun, 04 Jun 2006) | 2 lines
Drop Mac wrappers for the WASTE library.
........
r46645 | tim.peters | 2006-06-04 17:49:07 +0200 (Sun, 04 Jun 2006) | 3 lines
s_methods[]: Stop compiler warnings by casting
s_unpack_from to PyCFunction.
........
r46646 | george.yoshida | 2006-06-04 19:04:12 +0200 (Sun, 04 Jun 2006) | 2 lines
Remove a redundant word
........
r46647 | george.yoshida | 2006-06-04 19:17:25 +0200 (Sun, 04 Jun 2006) | 2 lines
Markup fix
........
r46648 | martin.v.loewis | 2006-06-04 21:36:28 +0200 (Sun, 04 Jun 2006) | 2 lines
Patch #1359618: Speed-up charmap encoder.
........
r46649 | georg.brandl | 2006-06-04 23:46:16 +0200 (Sun, 04 Jun 2006) | 3 lines
Repair refleaks in unicodeobject.
........
r46650 | georg.brandl | 2006-06-04 23:56:52 +0200 (Sun, 04 Jun 2006) | 4 lines
Patch #1346214: correctly optimize away "if 0"-style stmts
(thanks to Neal for review)
........
r46651 | georg.brandl | 2006-06-05 00:15:37 +0200 (Mon, 05 Jun 2006) | 2 lines
Bug #1500293: fix memory leaks in _subprocess module.
........
r46654 | tim.peters | 2006-06-05 01:43:53 +0200 (Mon, 05 Jun 2006) | 2 lines
Whitespace normalization.
........
r46655 | tim.peters | 2006-06-05 01:52:47 +0200 (Mon, 05 Jun 2006) | 16 lines
Revert revisions:
46640 Patch #1454481: Make thread stack size runtime tunable.
46647 Markup fix
The first is causing many buildbots to fail test runs, and there
are multiple causes with seemingly no immediate prospects for
repairing them. See python-dev discussion.
Note that a branch can (and should) be created for resolving these
problems, like
svn copy svn+ssh://svn.python.org/python/trunk -r46640 svn+ssh://svn.python.org/python/branches/NEW_BRANCH
followed by merging rev 46647 to the new branch.
........
r46656 | andrew.kuchling | 2006-06-05 02:08:09 +0200 (Mon, 05 Jun 2006) | 1 line
Mention second encoding speedup
........
r46657 | gregory.p.smith | 2006-06-05 02:31:01 +0200 (Mon, 05 Jun 2006) | 7 lines
bugfix: when log_archive was called with the DB_ARCH_REMOVE flag present
in BerkeleyDB >= 4.2 it tried to construct a list out of an uninitialized
char **log_list.
feature: export the DB_ARCH_REMOVE flag by name in the module on BerkeleyDB >= 4.2.
........
r46658 | gregory.p.smith | 2006-06-05 02:33:35 +0200 (Mon, 05 Jun 2006) | 5 lines
fix a bug in the previous commit. don't leak empty list on error return and
fix the additional rare (out of memory only) bug that it was supposed to fix
of not freeing log_list when the python allocator failed.
........
r46660 | tim.peters | 2006-06-05 02:55:26 +0200 (Mon, 05 Jun 2006) | 9 lines
"Flat is better than nested."
Move the long-winded, multiply-nested -R support out
of runtest() and into some module-level helper functions.
This makes runtest() and the -R code easier to follow.
That in turn allowed seeing some opportunities for code
simplification, and made it obvious that reglog.txt
never got closed.
........
r46661 | hyeshik.chang | 2006-06-05 02:59:54 +0200 (Mon, 05 Jun 2006) | 3 lines
Fix a potentially invalid memory access of CJKCodecs' shift-jis
decoder. (found by Neal Norwitz)
........
r46663 | gregory.p.smith | 2006-06-05 03:39:52 +0200 (Mon, 05 Jun 2006) | 3 lines
* support DBEnv.log_stat() method on BerkeleyDB >= 4.0 [patch #1494885]
........
r46664 | tim.peters | 2006-06-05 03:43:03 +0200 (Mon, 05 Jun 2006) | 3 lines
Remove doctest.testmod's deprecated (in 2.4) `isprivate`
argument. A lot of hair went into supporting that!
........
r46665 | tim.peters | 2006-06-05 03:47:24 +0200 (Mon, 05 Jun 2006) | 2 lines
Whitespace normalization.
........
r46666 | tim.peters | 2006-06-05 03:48:21 +0200 (Mon, 05 Jun 2006) | 2 lines
Make doctest news more accurate.
........
r46667 | gregory.p.smith | 2006-06-05 03:56:15 +0200 (Mon, 05 Jun 2006) | 3 lines
* support DBEnv.lsn_reset() method on BerkeleyDB >= 4.4 [patch #1494902]
........
r46668 | gregory.p.smith | 2006-06-05 04:02:25 +0200 (Mon, 05 Jun 2006) | 3 lines
mention the just committed bsddb changes
........
r46671 | gregory.p.smith | 2006-06-05 19:38:04 +0200 (Mon, 05 Jun 2006) | 3 lines
* add support for DBSequence objects [patch #1466734]
........
r46672 | gregory.p.smith | 2006-06-05 20:20:07 +0200 (Mon, 05 Jun 2006) | 3 lines
forgot to add this file in previous commit
........
r46673 | tim.peters | 2006-06-05 20:36:12 +0200 (Mon, 05 Jun 2006) | 2 lines
Whitespace normalization.
........
r46674 | tim.peters | 2006-06-05 20:36:54 +0200 (Mon, 05 Jun 2006) | 2 lines
Add missing svn:eol-style property to text files.
........
r46675 | gregory.p.smith | 2006-06-05 20:48:21 +0200 (Mon, 05 Jun 2006) | 4 lines
* fix DBCursor.pget() bug with keyword argument names when no data= is
supplied [SF pybsddb bug #1477863]
........
r46676 | andrew.kuchling | 2006-06-05 21:05:32 +0200 (Mon, 05 Jun 2006) | 1 line
Remove use of Trove name, which isn't very helpful to users
........
r46677 | andrew.kuchling | 2006-06-05 21:08:25 +0200 (Mon, 05 Jun 2006) | 1 line
[Bug #1470026] Include link to list of classifiers
........
r46679 | tim.peters | 2006-06-05 22:48:49 +0200 (Mon, 05 Jun 2006) | 10 lines
Access _struct attributes directly instead of mucking with getattr.
string_reverse(): Simplify.
assertRaises(): Raise TestFailed on failure.
test_unpack_from(), test_pack_into(), test_pack_into_fn(): never
use `assert` to test for an expected result (it doesn't test anything
when Python is run with -O).
........
r46680 | tim.peters | 2006-06-05 22:49:27 +0200 (Mon, 05 Jun 2006) | 2 lines
Add missing svn:eol-style property to text files.
........
r46681 | gregory.p.smith | 2006-06-06 01:38:06 +0200 (Tue, 06 Jun 2006) | 3 lines
add depends = ['md5.h'] to the _md5 module extension for correctness sake.
........
r46682 | brett.cannon | 2006-06-06 01:51:55 +0200 (Tue, 06 Jun 2006) | 4 lines
Add 3 more bytes to a buffer to cover constants in string and null byte on top of 10 possible digits for an int.
Closes bug #1501223.
........
r46684 | gregory.p.smith | 2006-06-06 01:59:37 +0200 (Tue, 06 Jun 2006) | 5 lines
- bsddb: the __len__ method of a DB object has been fixed to return correct
results. It could previously incorrectly return 0 in some cases.
Fixes SF bug 1493322 (pybsddb bug 1184012).
........
r46686 | tim.peters | 2006-06-06 02:25:07 +0200 (Tue, 06 Jun 2006) | 7 lines
_PySys_Init(): It's rarely a good idea to size a buffer to the
exact maximum size someone guesses is needed. In this case, if
we're really worried about extreme integers, then "cp%d" can
actually need 14 bytes (2 for "cp" + 1 for \0 at the end +
11 for -(2**31-1)). So reserve 128 bytes instead -- nothing is
actually saved by making a stack-local buffer tiny.
........
r46687 | neal.norwitz | 2006-06-06 09:22:08 +0200 (Tue, 06 Jun 2006) | 1 line
Remove unused variable (and stop compiler warning)
........
r46688 | neal.norwitz | 2006-06-06 09:23:01 +0200 (Tue, 06 Jun 2006) | 1 line
Fix a bunch of parameter strings
........
r46689 | thomas.heller | 2006-06-06 13:34:33 +0200 (Tue, 06 Jun 2006) | 6 lines
Convert CFieldObject tp_members to tp_getset, since there is no
structmember typecode for Py_ssize_t fields. This should fix some of
the errors on the PPC64 debian machine (64-bit, big endian).
Assigning to readonly fields now raises AttributeError instead of
TypeError, so the testcase has to be changed as well.
........
r46690 | thomas.heller | 2006-06-06 13:54:32 +0200 (Tue, 06 Jun 2006) | 1 line
Damn - the sentinel was missing. And fix another silly mistake.
........
r46691 | martin.blais | 2006-06-06 14:46:55 +0200 (Tue, 06 Jun 2006) | 13 lines
Normalized a few cases of whitespace in function declarations.
Found them using::
find . -name '*.py' | while read i ; do grep 'def[^(]*( ' $i /dev/null ; done
find . -name '*.py' | while read i ; do grep ' ):' $i /dev/null ; done
(I was doing this all over my own code anyway, because I'd been using spaces in
all defs, so I thought I'd make a run on the Python code as well. If you need
to do such fixes in your own code, you can use xx-rename or parenregu.el within
emacs.)
........
r46693 | thomas.heller | 2006-06-06 17:34:18 +0200 (Tue, 06 Jun 2006) | 1 line
Specify argtypes for all test functions. Maybe that helps on strange ;-) architectures
........
r46694 | tim.peters | 2006-06-06 17:50:17 +0200 (Tue, 06 Jun 2006) | 5 lines
BSequence_set_range(): Rev 46688 ("Fix a bunch of
parameter strings") changed this function's signature
seemingly by mistake, which is causing buildbots to fail
test_bsddb3. Restored the pre-46688 signature.
........
r46695 | tim.peters | 2006-06-06 17:52:35 +0200 (Tue, 06 Jun 2006) | 4 lines
On python-dev Thomas Heller said these were committed
by mistake in rev 46693, so reverting this part of
rev 46693.
........
r46696 | andrew.kuchling | 2006-06-06 19:10:41 +0200 (Tue, 06 Jun 2006) | 1 line
Fix comment typo
........
r46697 | brett.cannon | 2006-06-06 20:08:16 +0200 (Tue, 06 Jun 2006) | 2 lines
Fix coding style guide bug.
........
r46698 | thomas.heller | 2006-06-06 20:50:46 +0200 (Tue, 06 Jun 2006) | 2 lines
Add a hack so that foreign functions returning float now do work on 64-bit
big endian platforms.
........
r46699 | thomas.heller | 2006-06-06 21:25:13 +0200 (Tue, 06 Jun 2006) | 3 lines
Use the same big-endian hack as in _ctypes/callproc.c for callback functions.
This fixes the callback function tests that return float.
........
r46700 | ronald.oussoren | 2006-06-06 21:50:24 +0200 (Tue, 06 Jun 2006) | 5 lines
* Ensure that "make altinstall" works when the tree was configured
with --enable-framework
* Also for --enable-framework: allow users to use --prefix to specify
the location of the compatibility symlinks (such as /usr/local/bin/python)
........
r46701 | ronald.oussoren | 2006-06-06 21:56:00 +0200 (Tue, 06 Jun 2006) | 3 lines
A quick hack to ensure the right key-bindings for IDLE on osx: install patched
configuration files during a framework install.
........
r46702 | tim.peters | 2006-06-07 03:04:59 +0200 (Wed, 07 Jun 2006) | 4 lines
dash_R_cleanup(): Clear filecmp._cache. This accounts for
different results across -R runs (at least on Windows) of
test_filecmp.
........
r46705 | tim.peters | 2006-06-07 08:57:51 +0200 (Wed, 07 Jun 2006) | 17 lines
SF patch 1501987: Remove randomness from test_exceptions,
from ?iga Seilnacht (sorry about the name, but Firefox
on my box can't display the first character of the name --
the SF "Unix name" is zseil).
This appears to cure the oddball intermittent leaks across
runs when running test_exceptions under -R. I'm not sure
why, but I'm too sleepy to care ;-)
The thrust of the SF patch was to remove randomness in the
pickle protocol used. I changed the patch to use
range(pickle.HIGHEST_PROTOCOL + 1), to try both pickle and
cPickle, and randomly mucked with other test lines to put
statements on their own lines.
Not a bugfix candidate (this is fiddling new-in-2.5 code).
........
r46706 | andrew.kuchling | 2006-06-07 15:55:33 +0200 (Wed, 07 Jun 2006) | 1 line
Add an SQLite introduction, taken from the 'What's New' text
........
r46708 | andrew.kuchling | 2006-06-07 19:02:52 +0200 (Wed, 07 Jun 2006) | 1 line
Mention other placeholders
........
r46709 | andrew.kuchling | 2006-06-07 19:03:46 +0200 (Wed, 07 Jun 2006) | 1 line
Add an item; also, escape %
........
r46710 | andrew.kuchling | 2006-06-07 19:04:01 +0200 (Wed, 07 Jun 2006) | 1 line
Mention other placeholders
........
r46716 | ronald.oussoren | 2006-06-07 20:57:44 +0200 (Wed, 07 Jun 2006) | 2 lines
Move Mac/OSX/Tools one level up
........
r46717 | ronald.oussoren | 2006-06-07 20:58:01 +0200 (Wed, 07 Jun 2006) | 2 lines
Move Mac/OSX/PythonLauncher one level up
........
r46718 | ronald.oussoren | 2006-06-07 20:58:42 +0200 (Wed, 07 Jun 2006) | 2 lines
mv Mac/OSX/BuildScript one level up
........
r46719 | ronald.oussoren | 2006-06-07 21:02:03 +0200 (Wed, 07 Jun 2006) | 2 lines
Move Mac/OSX/* one level up
........
r46720 | ronald.oussoren | 2006-06-07 21:06:01 +0200 (Wed, 07 Jun 2006) | 2 lines
And the last bit: move IDLE one level up and adjust makefiles
........
r46723 | ronald.oussoren | 2006-06-07 21:38:53 +0200 (Wed, 07 Jun 2006) | 4 lines
- Patch the correct version of python in the Info.plists at build time, instead
of relying on a maintainer to update them before releases.
- Remove the now empty Mac/OSX directory
........
r46727 | ronald.oussoren | 2006-06-07 22:18:44 +0200 (Wed, 07 Jun 2006) | 7 lines
* If BuildApplet.py is used as an applet it starts with a version of
sys.exutable that isn't usuable on an #!-line. That results in generated
applets that don't actually work. Work around this problem by resetting
sys.executable.
* argvemulator.py didn't work on intel macs. This patch fixes this
(bug #1491468)
........
r46728 | tim.peters | 2006-06-07 22:40:06 +0200 (Wed, 07 Jun 2006) | 2 lines
Whitespace normalization.
........
r46729 | tim.peters | 2006-06-07 22:40:54 +0200 (Wed, 07 Jun 2006) | 2 lines
Add missing svn:eol-style property to text files.
........
r46730 | thomas.heller | 2006-06-07 22:43:06 +0200 (Wed, 07 Jun 2006) | 7 lines
Fix for foreign functions returning small structures on 64-bit big
endian machines. Should fix the remaininf failure in the PPC64
Debian buildbot.
Thanks to Matthias Klose for providing access to a machine to debug
and test this.
........
r46731 | brett.cannon | 2006-06-07 23:48:17 +0200 (Wed, 07 Jun 2006) | 2 lines
Clarify documentation for bf_getcharbuffer.
........
r46735 | neal.norwitz | 2006-06-08 07:12:45 +0200 (Thu, 08 Jun 2006) | 1 line
Fix a refleak in recvfrom_into
........
r46736 | gregory.p.smith | 2006-06-08 07:17:08 +0200 (Thu, 08 Jun 2006) | 9 lines
- bsddb: the bsddb.dbtables Modify method now raises the proper error and
aborts the db transaction safely when a modifier callback fails.
Fixes SF python patch/bug #1408584.
Also cleans up the bsddb.dbtables docstrings since thats the only
documentation that exists for that unadvertised module. (people
really should really just use sqlite3)
........
r46737 | gregory.p.smith | 2006-06-08 07:38:11 +0200 (Thu, 08 Jun 2006) | 4 lines
* Turn the deadlock situation described in SF bug #775414 into a
DBDeadLockError exception.
* add the test case for my previous dbtables commit.
........
r46738 | gregory.p.smith | 2006-06-08 07:39:54 +0200 (Thu, 08 Jun 2006) | 2 lines
pasted set_lk_detect line in wrong spot in previous commit. fixed. passes tests this time.
........
r46739 | armin.rigo | 2006-06-08 12:56:24 +0200 (Thu, 08 Jun 2006) | 6 lines
(arre, arigo) SF bug #1350060
Give a consistent behavior for comparison and hashing of method objects
(both user- and built-in methods). Now compares the 'self' recursively.
The hash was already asking for the hash of 'self'.
........
r46740 | andrew.kuchling | 2006-06-08 13:56:44 +0200 (Thu, 08 Jun 2006) | 1 line
Typo fix
........
r46741 | georg.brandl | 2006-06-08 14:45:01 +0200 (Thu, 08 Jun 2006) | 2 lines
Bug #1502750: Fix getargs "i" format to use LONG_MIN and LONG_MAX for bounds checking.
........
r46743 | georg.brandl | 2006-06-08 14:54:13 +0200 (Thu, 08 Jun 2006) | 2 lines
Bug #1502728: Correctly link against librt library on HP-UX.
........
r46745 | georg.brandl | 2006-06-08 14:55:47 +0200 (Thu, 08 Jun 2006) | 3 lines
Add news for recent bugfix.
........
r46746 | georg.brandl | 2006-06-08 15:31:07 +0200 (Thu, 08 Jun 2006) | 4 lines
Argh. "integer" is a very confusing word ;)
Actually, checking for INT_MAX and INT_MIN is correct since
the format code explicitly handles a C "int".
........
r46748 | nick.coghlan | 2006-06-08 15:54:49 +0200 (Thu, 08 Jun 2006) | 1 line
Add functools.update_wrapper() and functools.wraps() as described in PEP 356
........
r46751 | georg.brandl | 2006-06-08 16:50:21 +0200 (Thu, 08 Jun 2006) | 4 lines
Bug #1502805: don't alias file.__exit__ to file.close since the
latter can return something that's true.
........
r46752 | georg.brandl | 2006-06-08 16:50:53 +0200 (Thu, 08 Jun 2006) | 3 lines
Convert test_file to unittest.
........
number of tests, all because of the codecs/_multibytecodecs issue described
here (it's not a Py3K issue, just something Py3K discovers):
http://mail.python.org/pipermail/python-dev/2006-April/064051.html
Hye-Shik Chang promised to look for a fix, so no need to fix it here. The
tests that are expected to break are:
test_codecencodings_cn
test_codecencodings_hk
test_codecencodings_jp
test_codecencodings_kr
test_codecencodings_tw
test_codecs
test_multibytecodec
This merge fixes an actual test failure (test_weakref) in this branch,
though, so I believe merging is the right thing to do anyway.
(branch-creation time) up to 43067. 43068 and 43069 contain a little
swapping action between re.py and sre.py, and this mightily confuses svn
merge, so later changes are going in separately.
This merge should break no additional tests.
The last-merged revision is going in a 'last_merge' property on '.' (the
branch directory.) Arbitrarily chosen, really; if there's a BCP for this, I
couldn't find it, but we can easily change it afterwards ;)
managed by C, because it's possible for the block to be smaller than the
new requested size, and at the end of allocated VM. Trying to copy over
nbytes bytes to a Python small-object block can segfault then, and there's
no portable way to avoid this (we would have to know how many bytes
starting at p are addressable, and std C has no means to determine that).
Bugfix candidate. Should be backported to 2.4, but I'm out of time.
consistency checks, enabled only in a debug (Py_DEBUG) build. Note that
this never gets called automatically unless PYMALLOC_DEBUG is #define'd
too, and the envar PYTHONMALLOCSTATS exists.
Added code to call this when PYMALLOC_DEBUG is enabled, and envar
PYTHONMALLOCSTATS is set, whenever a new arena is obtained and once
late in the Python shutdown process.
_PyObject_DebugMalloc: explicitly cast PyObject_Malloc's result to the
target pointer type.
_PyObject_DebugDumpStats: change decl of arena_alignment from unsigned
int to unsigned long.
This is for the 2.3 release only (it's new code).
most of the work. In particular, if the underlying realloc is able to
grow the memory block in place, great (this routine used to do a fresh
malloc + memcpy every time a block grew). BTW, I'm not so keen here on
avoiding possible quadratic-time realloc patterns as I am on making
the debug pymalloc more invisible (the more it uses memory "just like"
the underlying allocator, the better the chance that a suspected memory
corruption bug won't vanish when the debug malloc is turned on).
The bug report pointed out a bogosity in the comment block explaining
thread safety for arena management. Repaired that comment, repaired a
couple others while I was at it, and added an assert.
_PyMalloc_DebugRealloc: If this needed to get more memory, but couldn't,
it erroneously freed the original memory. Repaired that.
This is for 2.3 only (unless we decide to backport the new pymalloc).
runtime multiplications and divisions, via the scheme developed with
Vladimir Marangozov on Python-Dev. The pool_header struct loses its
capacity member, but gains nextoffset and maxnextoffset members; this
still leaves it at 32 bytes on a 32-bit box (it has to be padded to a
multiple of 8 bytes).
what these do given a 0 size argument. This is so that when pymalloc
is enabled, we don't need to wrap pymalloc calls in goofy little
routines special-casing 0. Note that it's virtually impossible to meet
the doc's promise that malloc(0) will never return NULL; this makes a
best effort, but not an insane effort. The code does promise that
realloc(not-NULL, 0) will never return NULL (malloc(0) is much harder).
_PyMalloc_Realloc: Changed to take over all requests for 0 bytes, and
rearranged to be a little quicker in expected cases.
All over the place: when resorting to the platform allocator, call
free/malloc/realloc directly, without indirecting thru macros. This
should avoid needing a nightmarish pile of #ifdef-ery if PYMALLOC_DEBUG
is changed so that pymalloc takes over all Py(Mem, Object} memory
operations (which would add useful debugging info to PyMem_xyz
allocations too).
possible pool states. I think it's much clearer now.
Added a new long overdue block-management overview comment block.
I believe the comments are in good shape now.
Added two comments about possible small optimizations (one getting rid
of runtime multiplications at the cost of a new pool_header member; the
other getting rid of runtime divisions and the pool_header capacity
member, at the cost of a static const vector of 32 uints).
This displays stats about the # of arenas, pools, blocks and bytes, to
stderr, both used and reserved but unused.
CAUTION: Because PYMALLOC_DEBUG is on, the debug malloc routine adds
16 bytes to each request. This makes each block appear two size classes
higher than it would be if PYMALLOC_DEBUG weren't on.
So far, playing with this confirms the obvious: there's a lot of activity
in the "small dict" size class, but nothing in the core makes any use of
the 8-byte or 16-byte classes.
the code so that the most frequent cases come first. Added comments.
Found a hidden assumption that a pool contains room for at least two
blocks, and added an assert to catch a violation if it ever happens in
a place where that matters. Gave the normal "I allocated this block"
case a longer basic block to work with before it has to do its first
branch (via breaking apart an embedded assignment in an "if", and
hoisting common code out of both branches).
address obtained from system malloc/realloc without holding the GIL.
When the vector of arena base addresses has to grow, the old vector is
deliberately leaked. This makes "stale" x-thread references safe.
arenas and narenas are also declared volatile, and changed in an order
that prevents a thread from picking up a value of narenas too large
for the value of arenas it sees.
Added more asserts.
Fixed an old inaccurate comment.
Added a comment explaining why it's safe to call pymalloc free/realloc
with an address obtained from system malloc/realloc even when arenas is
still NULL (this is obscure, since the ADDRESS_IN_RANGE macro
appears <wink> to index into arenas).
this. But added an overflow check just in case there is.
Got rid of the ushort macro. It wasn't used anymore (it was only used
in the no-longer-exists off_t macro), and there's no plausible use for it.
waste the first pool if malloc happens to return a pool-aligned address.
This means the number of pools per arena can now vary by 1. Unfortunately,
the code counted up from 0 to a presumed constant number of pools. So
changed the increasing "watermark" counter to a decreasing "nfreepools"
counter instead, and fiddled various stuff accordingly. This also allowed
getting rid of two more macros.
Also changed the code to align the first address to a pool boundary
instead of a page boundary. These are two parallel sets of macro #defines
that happen to be identical now, but the page macros are in theory more
restrictive (bigger), and there's simply no reason I can see that it
wasn't aligning to the less restrictive pool size all along (the code
only relies on pool alignment).
Hmm. The "page size" macros aren't used for anything *except* defining
the pool size macros, and the comments claim the latter isn't necessary.
So this has the feel of a layer of indirection that doesn't serve a
purpose; should probably get rid of the page macros now.
are called without the GIL. It's incredibly unlikely to fail, but I can't
make this bulletproof without either adding a lock for exclusion, or
giving up on growing the arena base-address vector (it would be safe if
this were a static array).
+ A new scheme for determining whether an address belongs to a pymalloc
arena. This should be 100% reliable. The poolp->pooladdr and
poolp->magic members are gone. A new poolp->arenaindex member takes
their place. Note that the pool header overhead doesn't actually
shrink, though, since the header is padded to a multiple of 8 bytes.
+ _PyMalloc_Free and _PyMalloc_Realloc should now be safe to call for
any legit address, whether obtained from a _PyMalloc function or from
the system malloc/realloc. It should even be safe to call
_PyMalloc_Free when *not* holding the GIL, provided that the passed-in
address was obtained from system malloc/realloc. Since this is
accomplished without any locks, you better believe the code is subtle.
I hope it's sufficiently commented.
+ The above implies we don't need the new PyMalloc_{New, NewVar, Del}
API anymore, and could switch back to PyObject_XXX without breaking
existing code mixing PyObject_XXX with PyMem_{Del, DEL, Free, FREE}.
Nothing is done here about that yet, and I'd like to see this new
code exercised more first.
+ The small object threshhold is boosted to 256 (the max). We should
play with that some more, but the old 64 was way too small for 2.3.
+ Getting a new arena is now done via new function new_arena().
+ Removed some unused macros, and squashed out some macros that were
used only once to define other macros.
+ Arenas are no longer linked together. A new vector of arena base
addresses had to be created anyway to make address classification
bulletproof.
+ A lot of the patch size is an illusion: given the way address
classification works now, it was more convenient to switch the
sense of the prime "if" tests in the realloc and free functions,
so the "if" and "else" blocks got swapped.
+ Assorted minor code, comment and whitespace cleanup.
Back to the Windows installer <wink>.
Assorted: bump the serial number via a trivial new bumpserialno()
function. The point is to give a single place to set a breakpoint when
waiting for a specific serial number.
When WITH_PYMALLOC is defined, define PYMALLOC_DEBUG to enable the debug
allocator. This can be done independent of build type (release or debug).
A debug build automatically defines PYMALLOC_DEBUG when pymalloc is
enabled. It's a detected error to define PYMALLOC_DEBUG when pymalloc
isn't enabled.
Two debugging entry points defined only under PYMALLOC_DEBUG:
+ _PyMalloc_DebugCheckAddress(const void *p) can be used (e.g., from gdb)
to sanity-check a memory block obtained from pymalloc. It sprays
info to stderr (see next) and dies via Py_FatalError if the block is
detectably damaged.
+ _PyMalloc_DebugDumpAddress(const void *p) can be used to spray info
about a debug memory block to stderr.
A tiny start at implementing "API family" checks isn't good for
anything yet.
_PyMalloc_DebugRealloc() has been optimized to do little when the new
size is <= old size. However, if the new size is larger, it really
can't call the underlying realloc() routine without either violating its
contract, or knowing something non-trivial about how the underlying
realloc() works. A memcpy is always done in this case.
This was a disaster for (and only) one of the std tests: test_bufio
creates single text file lines up to a million characters long. On
Windows, fileobject.c's get_line() uses the horridly funky
getline_via_fgets(), which keeps growing and growing a string object
hoping to find a newline. It grew the string object 1000 bytes each
time, so for a million-character string it took approximately forever
(I gave up after a few minutes).
So, also:
fileobject.c, getline_via_fgets(): When a single line is outrageously
long, grow the string object at a mildly exponential rate, instead of
just 1000 bytes at a time.
That's enough so that a debug-build test_bufio finishes in about 5 seconds
on my Win98SE box. I'm curious to try this on Win2K, because it has very
different memory behavior than Win9X, and test_bufio always took a factor
of 10 longer to complete on Win2K. It *could* be that the endless
reallocs were simply killing it on Win2K even in the release build.
Also move all _PyMalloc_XXX entry points into obmalloc.c.
The Windows build works fine.
The Unix build is changed here (Makefile.pre.in), but not tested.
No other platform's build process has been fiddled.
May or may not be related to bug 407680 (obmalloc.c - looks like it's
corrupted). This repairs the illegal vrbl names, but leaves a pile of
illegal macro names (_THIS_xxx, _SYSTEM_xxx, _SET_HOOKS, _FETCH_HOOKS).
Thomas Wouters
Neal Norwitz
Martin v. Löwis
Tim Peters
Walter Dörwald
Guido van Rossum
Jeremy Hylton
Neil Schemenauer