[issue28275] LZMADecompressor.decompress Use After Free

Serhiy Storchaka added the comment: Thanks John. Could you please add a test based on your reproducer? -- nosy: +serhiy.storchaka stage: -> test needed ___ Python tracker

[issue28207] Use pkg-config to find dependencies

Zachary Ware added the comment: I don't have any philosophical opposition to using pkg-config, but it would be nice to avoid making configure/Makefile/setup.py any more complex than they already are. If you can somehow simplify those files (without breaking anything) by using pkg-config, I'd

[issue28275] LZMADecompressor.decompress Use After Free

Changes by Alex Gaynor : -- nosy: +nadeem.vawda ___ Python tracker ___ ___

[issue28275] LZMADecompressor.decompress Use After Free

Changes by Alex Gaynor : -- keywords: +security_issue ___ Python tracker ___ ___

[issue28275] LZMADecompressor.decompress Use After Free

Changes by John Leitch : Added file: http://bugs.python.org/file44828/Py35_LZMADecompressor.py ___ Python tracker ___

[issue28275] LZMADecompressor.decompress Use After Free

New submission from John Leitch: Python 3.5.2 suffers from a use after free vulnerability caused by the behavior of the LZMADecompressor.decompress method. The problem exists due to a dangling pointer created by an incomplete error path in the _lzma!decompress function. static PyObject *

[issue24201] _winreg PyHKEY Type Confusion

Zachary Ware added the comment: I agree with Eryk that this is not a winreg bug. If I'm understanding this correctly, you would need to pass unsanitized remote input into a function that's going to affect your registry. That strikes me as an incredibly ridiculous thing to do; anyone doing

Re: Stompy

> > My error: >> Traceback (most recent call last): >> File "temp.py", line 8, in >> my_stomp.connect(AMQ_USERNAME, AMQ_PASSWORD) >> File "/lclapps/oppen/thirdparty/stompy/stomp.py", line 48, in connect >> self.frame.connect(self.sock, username=username, password=password, >>

Re: Is 'learning python 5th edition' a good book to beginner?

On Sun, Sep 25, 2016 at 3:50 PM, Lawrence D’Oliveiro wrote: > Judging from the table of contents > , it spends too much time > (i.e. > 0) talking about Python 2. I own this book. The 5th edition covers both Python 3.3

Dynamically updating Graph in python

0 down vote favorite I have a M x N 2D array: ith row represents that value of N points at time i. I want to visualize the points [1 row of the array] in the form of a graph where the values get updated after a small interval. Thus the graph shows 1 row at a time, then update the

[issue28207] SQLite headers are not searched in custom locations

Santiago Castro added the comment: Okay, I checked out again and bz2 and openssl were in standard paths in fact, my bad. But I think python should also take into account pkg-config. I left a Dockerfile with how I think it should work with pyenv and Linuxbrew:

[issue28260] mock._Any and mock._Call implement __eq__ but not __hash__

Changes by Berker Peksag : -- resolution: -> not a bug stage: -> resolved status: open -> closed ___ Python tracker ___

[issue27141] Fix collections.UserList shallow copy

Serhiy Storchaka added the comment: UserList.copy() doesn't copy instance attributes, but copy.copy() should copy them. -- ___ Python tracker ___

[issue11957] re.sub confusion between count and flags args

Changes by Serhiy Storchaka : Added file: http://bugs.python.org/file44826/re_deprecate_positional_count.patch ___ Python tracker

[issue11957] re.sub confusion between count and flags args

Serhiy Storchaka added the comment: Here are two alternative patches. The first patch checks if count or maxsplit arguments are re.RegexFlag and raise TypeError if it is true. This makes misusing flags fail fast. The second patch deprecates passing count and maxsplit arguments as positional

Re: Is 'learning python 5th edition' a good book to beginner?

Judging from the table of contents , it spends too much time (i.e. > 0) talking about Python 2. -- https://mail.python.org/mailman/listinfo/python-list

Is 'learning python 5th edition' a good book to beginner?

I want to find a python book like C++ primer which provides me details to understand the language. -- https://mail.python.org/mailman/listinfo/python-list

[issue27998] Remove support of bytes paths in os.scandir()

Guido van Rossum added the comment: Nice! -- ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue28274] asyncio does not call exception handler if task stored

R. David Murray added the comment: In the first case, once the loop has run the task it no longer has a reference to it, and it gets GCed. The __del__ method of the task calls your exception handler. In the second case, you have a reference to it, so __del__ does not get called. If you

[issue28260] mock._Any and mock._Call implement __eq__ but not __hash__

Roy Williams added the comment: OK, let's close this issue and I'll ping the TIP thread/this issue again. On Sun, Sep 25, 2016, 11:16 AM Ned Batchelder wrote: > > Ned Batchelder added the comment: > > Roy, the code on GitHub isn't a literal copy of the Python 3 code,

[issue28274] asyncio does not call exception handler if task stored

New submission from Billy Foster: I found a very strange bug in asyncio where whether exception handlers are called depends on whether a task is stored. To illustrate, the following code works as expected, printing out that it made it to the exception handler: import asyncio async def run():

[issue28226] compileall does not support pathlib

New submission from Brett Cannon: Is there a fix for compileall missing from the patch? All I see are a new test. -- assignee: -> ethan.furman ___ Python tracker

[issue23407] os.walk always follows Windows junctions

Craig Holmquist added the comment: Updated patch with changes to Win32JunctionTests. -- Added file: http://bugs.python.org/file44824/issue23407-2.patch ___ Python tracker

[issue28231] zipfile does not support pathlib

Brett Cannon added the comment: the patch LGTM, but Serhiy has a point that maybe we should add tests for dealing with other paths such as those contained within the zipfile. -- ___ Python tracker

[issue28225] bz2 does not support pathlib

Brett Cannon added the comment: Docs might need updating, but otherwise LGTM. -- assignee: -> ethan.furman ___ Python tracker ___

[issue23407] os.walk always follows Windows junctions

Craig Holmquist added the comment: Actually, it looks like there is already a way to create junctions and a test for them in test_os. However, it includes this line: # Junctions are not recognized as links. self.assertFalse(os.path.islink(self.junction)) That suggests the old

Re: Stompy

On 2016-09-25 18:46, Jason Friedman wrote: My goal is to send messages to an AMQ server using Python 3.3. I found Stompy and performed 2to3-3.3 before building. I am open to non-Stompy solutions. My code: from stompy.stomp import Stomp my_stomp = Stomp(AMQ_HOST, AMQ_PORT)

[issue28260] mock._Any and mock._Call implement __eq__ but not __hash__

Ned Batchelder added the comment: Roy, the code on GitHub isn't a literal copy of the Python 3 code, since it uses six, "from __future__ import", conditional imports of builtins, and so on. This can be fixed in GitHub. -- ___ Python tracker

[issue28260] mock._Any and mock._Call implement __eq__ but not __hash__

Antti Haapala added the comment: Well, as far as I understand, this is a bug in the backport only. The Python 2.6+ backport does not behave identically to Python 3 code in Python 2, unless `__hash__ = None` is added. unittest.mock in Python 3 is not affected in any way. --

Stompy

My goal is to send messages to an AMQ server using Python 3.3. I found Stompy and performed 2to3-3.3 before building. I am open to non-Stompy solutions. My code: from stompy.stomp import Stomp my_stomp = Stomp(AMQ_HOST, AMQ_PORT) my_stomp.connect(AMQ_USERNAME, AMQ_PASSWORD) My error:

[issue11957] re.sub confusion between count and flags args

Roundup Robot added the comment: New changeset 216e8b809e4e by Serhiy Storchaka in branch '3.5': Issue #11957: Restored re tests for passing count and maxsplit as positional https://hg.python.org/cpython/rev/216e8b809e4e New changeset b39b09290718 by Serhiy Storchaka in branch '3.6': Issue

[issue23407] os.walk always follows Windows junctions

Craig Holmquist added the comment: The attached patch changes _Py_attribute_data_to_stat to set S_IFLNK for both symlinks and junctions, and changes win_readlink to return the target path for junctions (IO_REPARSE_TAG_MOUNT_POINT) as well as symlinks. I'm not sure what to do as far as adding

[issue27141] Fix collections.UserList shallow copy

Raymond Hettinger added the comment: Yes. I will look at this shortly. -- ___ Python tracker ___ ___

[issue28273] Make os.waitpid() option parameter optional.

Changes by Marcos Dione : -- components: +Library (Lib) type: -> enhancement versions: +Python 3.7 ___ Python tracker ___

[issue28273] Make os.waitpid() option parameter optional.

New submission from Marcos Dione: According to the documentation, the second parameter for os.waitpid(), called options, must be most of the time 0. So why not make it the default value? The attached patch implements the proposed change, which is API compatible with current usages of the

[issue28260] mock._Any and mock._Call implement __eq__ but not __hash__

Roy Williams added the comment: Hey Antti My understanding is the version available on PyPi is now a strict backport of what's in Python 3, so they suggested filling the bug in bugs.python.org so it can be backported.

[issue17206] Py_XDECREF() expands its argument multiple times

Changes by SilentGhost : -- resolution: -> fixed status: open -> closed type: -> behavior ___ Python tracker ___

[issue27665] Make create_server able to listen on several ports

Alexander Bayandin added the comment: Yury, in 3.6 is added support for multiple hosts for create_server but not for multiple ports which I suggest to add. -- resolution: out of date -> status: closed -> open versions: -Python 3.5 ___ Python

[issue27611] test_tix cannot import _default_root after test_idle

Changes by Serhiy Storchaka : -- status: open -> closed ___ Python tracker ___

[issue19795] Formatting of True/False/None in docs

Serhiy Storchaka added the comment: Only the first patch was committed. The patch for None is not reviewed yet. Here is updated patch. -- resolution: fixed -> stage: resolved -> patch review status: closed -> open title: Formatting of True/False in docs -> Formatting of

[issue20947] -Wstrict-overflow findings

Serhiy Storchaka added the comment: Use following command for reproducing: make -s CFLAGS=-Wstrict-overflow -- ___ Python tracker ___

[issue20947] -Wstrict-overflow findings

Serhiy Storchaka added the comment: I can reproduce warnings in Modules/_posixsubprocess.c: gcc -pthread -fPIC -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -std=c99 -Wextra -Wno-unused-result -Wno-unused-parameter -Wno-missing-field-initializers

[issue27923] PEP 467 -- Minor API improvements for binary sequences

Nick Coghlan added the comment: Just what Ethan noted above, and the fact that Ethan should also list himself as a co-author now :) That said, one major alternative it would be good to consider further between now and 3.7 is Serhiy's "seqtools" idea:

[issue27611] test_tix cannot import _default_root after test_idle

Roundup Robot added the comment: New changeset 94a26aa1b1e0 by Serhiy Storchaka in branch '2.7': Issue #27611: Fixed support of default root window in the Tix module. https://hg.python.org/cpython/rev/94a26aa1b1e0 New changeset 7b458bcdab75 by Serhiy Storchaka in branch '3.5': Issue #27611:

[issue18170] define built-in exceptions in Python code

Changes by Christian Heimes : -- nosy: +christian.heimes versions: +Python 3.7 -Python 3.4 ___ Python tracker ___

[issue1175984] Make subprocess.Popen support file-like objects (win)

Changes by Christian Heimes : -- dependencies: -subprocess: more general (non-buffering) communication resolution: -> out of date stage: test needed -> resolved status: open -> closed ___ Python tracker

[issue20947] -Wstrict-overflow findings

Christian Heimes added the comment: I can no longer reproduce the warnings with $ CFLAGS=-Wstrict-overflow ./configure -C --with-pydebug --silent $ make --silent -- nosy: +christian.heimes stage: -> resolved status: open -> closed ___ Python

[issue19795] Formatting of True/False in docs

Changes by Christian Heimes : -- resolution: -> fixed stage: patch review -> resolved status: open -> closed ___ Python tracker ___

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Vitaly Murashev added the comment: > And why not open MS_WIN64 to any Windows compiler It would be very good idea Patches suggested here are just the drafts which just work. Actually I don't believe they will be accepted, so just dropped here for history --

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Steve Dower added the comment: FWIW I'm not bothered by this patch. Maybe we can one day tidy this all up but merging this now is a good stop-gap. (I probably won't get a chance in the next week.) Other compilers can pass architecture flags on the command line instead of inferring them at

[issue28270] [MinGW] Can't compile Modules/posixmodule.c by MinGW - several macro are missed

Vitaly Murashev added the comment: We (crystax.net) are building Python for Windows completely by cmake and MinGW using PC/pyconfig.h without any configure scripts -- ___ Python tracker

[issue28272] a redundant check in maybe_small_long

Changes by Oren Milman : -- keywords: +patch Added file: http://bugs.python.org/file44819/issue28272_ver1.diff ___ Python tracker ___

[issue28272] a redundant check in maybe_small_long

Changes by Oren Milman : Added file: http://bugs.python.org/file44818/CPythonTestOutput.txt ___ Python tracker ___

[issue28272] a redundant check in maybe_small_long

Changes by Oren Milman : Added file: http://bugs.python.org/file44817/patchedCPythonTestOutput_ver1.txt ___ Python tracker ___

[issue28272] a redundant check in maybe_small_long

Changes by Oren Milman : Removed file: http://bugs.python.org/file44815/CPythonTestOutput.txt ___ Python tracker ___

[issue28272] a redundant check in maybe_small_long

Changes by Oren Milman : Removed file: http://bugs.python.org/file44816/patchedCPythonTestOutput_ver1.txt ___ Python tracker ___

[issue28272] a redundant check in maybe_small_long

Changes by Oren Milman : Added file: http://bugs.python.org/file44816/patchedCPythonTestOutput_ver1.txt ___ Python tracker ___

[issue28272] a redundant check in maybe_small_long

New submission from Oren Milman: current state In Objects/longobject.c, the function maybe_small_long first checks whether v (the received PyLongObject pointer) is not NULL. However, currently in every call to maybe_small_long, it is already guaranteed that v is not

[issue28271] [MinGW] Can't compile _ctypes/callproc.c - SEH not supported by MinGW

Changes by Vitaly Murashev : Added file: http://bugs.python.org/file44814/callproc.c.2.7.mingw.patch ___ Python tracker ___

[issue28271] [MinGW] Can't compile _ctypes/callproc.c - SEH not supported by MinGW

New submission from Vitaly Murashev: Structured exception handling not supported by MinGW, and as a result file 'Modules/_ctypes/callproc.c' is not compilable by MinGW without patching As I know the patch was initially introduced in Google's repo, and fixed file 'callproc.c' now can be found

[issue28270] [MinGW] Can't compile Modules/posixmodule.c by MinGW - several macro are missed

Martin Panter added the comment: This competes with the patch at Issue 17598. How are you building Python? I presume you are not using the configure script. It would be good to come up with a patch that addresses for both cases. -- nosy: +martin.panter

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Martin Panter added the comment: It looks like Issue 4709 may also be a duplicate. Your added definition is distant from the comment explaining it. And why not open MS_WIN64 to any Windows compiler, rather than limiting it to just MSC and MINGW? I presume you are not building by running the

[issue28270] [MinGW] Can't compile Modules/posixmodule.c by MinGW - several macro are missed

New submission from Vitaly Murashev: 'posixmodule.c' is written pretty well, but some important MinGW realated macro are missed. And as a result an attempt to complile Modules/posixmodule.c by MinGW fails. So suggested patch (for 3.5.2 and 2.7.12) just turns on missed MinGW related macro

[issue28270] [MinGW] Can't compile Modules/posixmodule.c by MinGW - several macro are missed

Changes by Vitaly Murashev : Added file: http://bugs.python.org/file44812/posixmodule.c.2.7.mingw.patch ___ Python tracker ___

[issue10504] Trivial mingw compile fixes

Martin Panter added the comment: The patch seems to share some changes to Modules/posixmodule.c with parts of Issue 17598’s patch (and one common part has already been applied). Issue 17591 already fixed to lowercase. Just now Issue 28269 has been opened about strcasecmp(). --

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Changes by Vitaly Murashev : -- type: -> crash ___ Python tracker ___ ___

[issue28269] [MinGW] Can't compile Python/dynload_win.c due to static strcasecmp

Changes by Vitaly Murashev : -- type: -> compile error ___ Python tracker ___ ___

[issue28260] mock._Any and mock._Call implement __eq__ but not __hash__

Antti Haapala added the comment: So you mean that unittest.mock should explicitly set `__hash__ = None`; but this is already what the unittest.mock does, in Python 3; there is no need to set `__hash__ = None`; it is the backport that needs to do this in order to stay compatible. There is no

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Vitaly Murashev added the comment: Patch suggested here is actually the most trivial as it could be. And at the same time we (crystax.net) can prove that after this patch Python being compiled by MInGW for 64-bit Windows actually works well. There are other minor fixes but this one is the

Re: How to convert 'ö' to 'oe' or 'o' (or other similar things) in a string?

On Sun, 25 Sep 2016 09:08 am, Thomas 'PointedEars' Lahn wrote: > Christian Gollwitzer wrote: > >> Am 17.09.16 um 23:19 schrieb Thomas 'PointedEars' Lahn: >>> Peng Yu wrote: Hi, I want to convert strings in which the characters with accents should be converted to the ones without

[issue27998] Remove support of bytes paths in os.scandir()

Serhiy Storchaka added the comment: I don't know what is the benefit of removing support of bytes paths in os.scandir(), but supporting bytes paths allows to remove ugly workarounds from os and glob modules. Lib/glob.py | 16 !! Lib/os.py | 67

Re: How to convert 'ö' to 'oe' or 'o' (or other similar things) in a string?

Am 25.09.16 um 01:08 schrieb Thomas 'PointedEars' Lahn: Christian Gollwitzer wrote: Am 17.09.16 um 23:19 schrieb Thomas 'PointedEars' Lahn: Peng Yu wrote: Hi, I want to convert strings in which the characters with accents should be converted to the ones without accents. […] […] ./main.py

[issue28269] [MinGW] Can't compile Python/dynload_win.c due to static strcasecmp

Changes by Vitaly Murashev : Added file: http://bugs.python.org/file44809/dynload_win.c.2.7.mingw.patch ___ Python tracker ___

[issue28269] [MinGW] Can't compile Python/dynload_win.c due to static strcasecmp

New submission from Vitaly Murashev: Attempt to complile Python/dynload_win.c by MinGW fails due to static reimplementation of strcasecmp function in this file: --- /* Case insensitive string compare, to avoid any dependencies on particular C RTL implementations */ static int strcasecmp

[issue26148] String literals are not interned if in a tuple

Serhiy Storchaka added the comment: The patch is provided in issue27942. -- resolution: -> duplicate stage: -> resolved status: open -> closed superseder: -> Default value identity regression ___ Python tracker

[issue28190] Detect curses headers correctly for cross-compiling

Changes by Serhiy Storchaka : -- nosy: +serhiy.storchaka ___ Python tracker ___

[issue28268] bz2.open does not use __fspath__ (PEP 519)

Serhiy Storchaka added the comment: This is a duplicate of issue27511 and issue28225. -- nosy: +serhiy.storchaka resolution: -> duplicate stage: -> resolved status: open -> closed superseder: -> Add PathLike objects support to BZ2File ___ Python

[issue27942] Default value identity regression

Serhiy Storchaka added the comment: Proposed patch interns string constants recursively in tuples and frozensets. This fixes issue26148 and issue25981. -- keywords: +patch stage: -> patch review versions: +Python 3.7 Added file:

[issue27995] Upgrade Python 3.4 to OpenSSL 1.0.2h on Windows

Christian Heimes added the comment: 1.0.2i is the latest version of the 1.0.2 series, #27995 -- resolution: -> out of date stage: -> resolved status: open -> closed superseder: -> Upgrade Python 3.4 to OpenSSL 1.0.2h on Windows ___ Python tracker

[issue28268] bz2.open does not use __fspath__ (PEP 519)

New submission from chrysn: The bz2.open function accepts a file name; after PEP 519 (file system path protocol) it "is expected that most APIs in Python's standard library that currently accept a file system path will be updated appropriately to accept path objects". BZ2File explicitly

[issue28248] Upgrade installers to OpenSSL 1.0.2i

Changes by Christian Heimes : -- superseder: Upgrade Python 3.4 to OpenSSL 1.0.2h on Windows -> ___ Python tracker ___

[issue28248] Upgrade installers to OpenSSL 1.0.2i

Changes by Christian Heimes : -- superseder: -> Upgrade Python 3.4 to OpenSSL 1.0.2h on Windows versions: +Python 3.4 ___ Python tracker

[issue15206] uuid module falls back to unsuitable RNG

Christian Heimes added the comment: Past me was a bit too eager... Only UUID4 are suppose to be random and unpredictable. uuid.UUID4 uses os.urandom() to as RNG. UUID1, UUID3 and UUID5 are more concerned with reducing collisions. -- status: open -> pending type: security -> behavior

Re: How to convert 'ö' to 'oe' or 'o' (or other similar things) in a string?

Den 2016-09-25 skrev wxjmfa...@gmail.com : > > As an European guy, I recommend to use the characters > set used in the "official" font families used in Germany: > BundesSerif / BundesSans. > > See >

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Martin Panter added the comment: This probably duplicates Issue 17590, although the extra context here is nice :) What do you think of my patch there, which also removes them from PC/pyconfig.h? I think it is better to define things in one place if possible. -- nosy: +martin.panter

[issue23505] Urlparse insufficient validation leads to open redirect

Antti Haapala added the comment: *I mean the problem exists in `urlparse`, not only in `urlunparse` -- ___ Python tracker ___

[issue28203] complex() gives wrong error when the second argument has an invalid type

Soumya Sharma added the comment: Thanks for the merge! -- ___ Python tracker ___ ___ Python-bugs-list mailing

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

Changes by Vitaly Murashev : Added file: http://bugs.python.org/file44806/pyconfig.h.2.7.mingw.patch ___ Python tracker ___

[issue28267] [MinGW] Crash at start when compiled by MinGW for 64-bit Windows using PC/pyconfig.h

New submission from Vitaly Murashev: Hi,here the issue: We (crystax.net) use custom builds of cpython, which for windows are compiled by MinGW with pyconfig.h taken from PC/pyconfig.h And for 32-bit Windows everything works well, but for 64-bit Windows - doesn't. The root cause of this issue is

[issue28190] Detect curses headers correctly for cross-compiling

Changes by Chi Hsuan Yen : -- title: Cross-build _curses failed if host ncurses headers and target ncurses headers have different layouts -> Detect curses headers correctly for cross-compiling ___ Python tracker

[issue28266] setup.py uses build Python's configuration when cross-compiling

Rouslan Korneychuk added the comment: > The problem is: setting PYTHON_FOR_BUILD manually breaks cross-compiling. > Please remove it, add python3.6 to $PATH and cross-compiling should be OK. I see. > For the problem of libffi: since issue27976 using bundled libffi is > deprecated. Please

[issue23505] Urlparse insufficient validation leads to open redirect

Antti Haapala added the comment: The problem is in `urlunparse`. If you check the RFC 2396, it has the following regular expression: ^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))? where group 3 is the //netloc, and 4 is netloc substring, of an url http://netloc/foobar - and

[issue28266] setup.py uses build Python's configuration when cross-compiling

Chi Hsuan Yen added the comment: The problem is: setting PYTHON_FOR_BUILD manually breaks cross-compiling. Please remove it, add python3.6 to $PATH and cross-compiling should be OK. For the problem of libffi: since issue27976 using bundled libffi is deprecated. Please build libffi outside

[issue28266] setup.py uses build Python's configuration when cross-compiling

Rouslan Korneychuk added the comment: I just tried compiling 3.6 unmodified for Android and it failed when trying to build libffi. The error message included these two lines: configure: error: cannot run C compiled programs. If you meant to cross compile, use `--host'. so 3.6b1 is still

[issue28207] SQLite headers are not searched in custom locations

Chi Hsuan Yen added the comment: Hmmm, currently only _ctypes uses pkg-config to detect libffi's header path, as it's not easy to determine without pkg-config. Is there a magic that enables openssl and bz2 outside standard paths :) Anyway, using pkg-config is not a bad idea. I'd like to hear

[issue28266] setup.py uses build Python's configuration when cross-compiling

Chi Hsuan Yen added the comment: This is fixed two weeks ago. See issue28046. The relevant fix is at http://bugs.python.org/issue28046#msg275845. Xavier pushed it to default before the 3.6 branch is created. I guess it should already be in 3.6b1?? Current hg tip works for me: $ find