Christian Heimes added the comment:
Brett removed a lot of stuff in 01cf4fb9c1aa567484c2ffb1b11f9b3fe9949b05 to
make the file more readable. I removed unnecessary -D, -I, and -L to make the
file even more readable. You can pass custom flags to ./configure.
Setup should not be edited by hand
Christian Heimes added the comment:
Typo, the last line should read "@HAVE_SCPROXY_FALSE@_scproxy"
--
___
Python tracker
<https://bugs.python.o
Christian Heimes added the comment:
Brett, we can use AM_CONDITIONAL() to conditionally enable/disable a feature
and AC_CONFIG_FILES() to create a Modules/Setup from a template:
Example:
The conditional
AM_CONDITIONAL([HAVE_SCPROXY], [test "$ac_sys_system" = "
Christian Heimes added the comment:
New changeset b5ee79494b2e0d484b7cf59f6746010e22567702 by Christian Heimes in
branch 'main':
bpo-45548: makesetup improvements (GH-29225)
https://github.com/python/cpython/commit/b5ee79494b2e0d484b7cf59f6746010e22567702
Change by Christian Heimes :
--
pull_requests: +27489
pull_request: https://github.com/python/cpython/pull/29225
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
New changeset 2b8677a3cd855eb3a579894c64588eab0e006269 by Christian Heimes in
branch 'main':
bpo-45548: FreeBSD doesn't like auto vars in makesetup (GH-29216)
https://github.com/python/cpython/commit/2b8677a3cd855eb3a579894c64588eab0e006269
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Change by Christian Heimes :
--
pull_requests: +27478
pull_request: https://github.com/python/cpython/pull/29216
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
New changeset ece916e92ce76cfdbfd208605a6fc827db0fbd52 by Christian Heimes in
branch 'main':
bpo-45548: Add missing extensions to Modules/Setup (GH-29199)
https://github.com/python/cpython/commit/ece916e92ce76cfdbfd208605a6fc827db0fbd52
Christian Heimes added the comment:
New changeset 81669d1b7bd668d1bd0fab92836e8b5e7976e362 by Christian Heimes in
branch 'main':
bpo-45595: Make extensions depend on header files (GH-29198)
https://github.com/python/cpython/commit/81669d1b7bd668d1bd0fab92836e8b5e7976e362
Christian Heimes added the comment:
New changeset 77e3f224d6ae6d38e5fc899fb5eaadf2b7c255a1 by Christian Heimes in
branch 'main':
bpo-45548: Remove checks for finite and gamma (GH-29206)
https://github.com/python/cpython/commit/77e3f224d6ae6d38e5fc899fb5eaadf2b7c255a1
Change by Christian Heimes :
--
pull_requests: +27474
pull_request: https://github.com/python/cpython/pull/29206
___
Python tracker
<https://bugs.python.org/issue45
Change by Christian Heimes :
--
pull_requests: +27468
pull_request: https://github.com/python/cpython/pull/29199
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
If we remove grep_headers_for(), then we can also close bpo-42325
--
___
Python tracker
<https://bugs.python.org/issue45
New submission from Christian Heimes :
The setup.py helper function grep_headers_for() is used by ctypes build step to
search header files for functions. The function kinda works on most platforms,
but only by accident. On my system it opens all header files that match
/usr/include/*.h
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27467
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29198
___
Python tracker
<https://bugs.python.org/issu
New submission from Christian Heimes :
Any change to a Python header file triggers a rebuild of Python core objects.
The extension modules are not automatically rebuild in several cases. This is
slightly annoying for core development because it forces me to do ``make
clean`` too often
Christian Heimes added the comment:
The trick would move the math function back into the core. Mark moved the math
functions out of the core on purpose, see bpo-7518.
--
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
Distutils is deprecated and should no longer be used. Please use twine to
upload software to PyPI,
https://blog.ganssle.io/articles/2021/10/setup-py-deprecated.html
--
___
Python tracker
<ht
Christian Heimes added the comment:
PR GH-29179 or GH-29181 address the issue with _math.o
PR GH-29164 adds the missing modules and also introduces pkg-config lookups for
dependencies.
Brett and I discussed that we have to introduce conditionals to Modules/Setup
to make use
Christian Heimes added the comment:
> Given that some essential third party modules are not going down this path,
> it is unlikely that general users would ever see any benefit.
I disagree with your statement. It does not reflect my experience, too.
Heap types are a prerequisite fo
Change by Christian Heimes :
--
pull_requests: +27459
pull_request: https://github.com/python/cpython/pull/29188
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
I ran into another problem. In shared mode, Modules/Setup places the shared
modules in ./Modules/ instead of the build directory ./build/lib.something/.
The ./Modules/ directory is not on sys.path. getpath.c only adds the build dir
to sys.path. It takes
Change by Christian Heimes :
--
pull_requests: +27453
pull_request: https://github.com/python/cpython/pull/29179
___
Python tracker
<https://bugs.python.org/issue45
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
We no longer accept patches. Contributors have to create a PR on GitHub, so we
can record contributions and verify the contributor license agreement.
--
___
Python tracker
<https://bugs.python.org/issue5
Christian Heimes added the comment:
New changeset f6e8b80d20159596cf641305bad3a833bedd2f4f by Christian Heimes in
branch 'main':
bpo-45571: use PY_CFLAGS_NODIST for shared Modules/Setup (GH-29161)
https://github.com/python/cpython/commit/f6e8b80d20159596cf641305bad3a833bedd2f4f
Christian Heimes added the comment:
I added several missing module definitions in
https://github.com/python/cpython/pull/29164
math/cmath conflict when _math.c is included twice.
--
___
Python tracker
<https://bugs.python.org/issue45
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
$ find build/lib.linux-x86_64-3.11/ -name '*.so' | sort
build/lib.linux-x86_64-3.11/_ctypes.cpython-311-x86_64-linux-gnu.so
build/lib.linux-x86_64-3.11/_ctypes_test.cpython-311-x86_64-linux-gnu.so
build/lib.linux-x86_64-3.11/_dbm.cpython-311-x86_64-linux
Christian Heimes added the comment:
I added some whitespace to Modules/Setup. All lines starting with ``#([a-z_])``
compile cleanly on Fedora 34 with all dependencies available. dbm, tkinter, and
sqlite are missing or untested.
$ sed -E -i 's/^#([a-z_])/\1/g' Modules/Setup
$ make
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27438
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29164
___
Python tracker
<https://bugs.python.org/issu
New submission from Christian Heimes :
pkg-config [1] is a standard tool on Linux and other platforms to detect
presence of dependencies as well as to figure out which compiler and linker
flags they require. Development packages provide a .pc file, e.g. ncurses
provides a ncursesw.pc file
Christian Heimes added the comment:
Could you or somebody else please create a PR with patch and a test case?
--
versions: +Python 3.11 -Python 2.7, Python 3.3, Python 3.4, Python 3.5, Python
3.6, Python 3.7, Python 3.8
___
Python tracker
<ht
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
New changeset 03e9f5dc751b8c441a85f428abc3f432ffe46345 by Christian Heimes in
branch 'main':
bpo-43974: Move Py_BUILD_CORE_MODULE into module code (GH-29157)
https://github.com/python/cpython/commit/03e9f5dc751b8c441a85f428abc3f432ffe46345
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27436
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29161
___
Python tracker
<https://bugs.python.org/issu
New submission from Christian Heimes :
I think that makesetup uses the wrong CFLAGS variable for shared modules. The
script uses PY_CFLAGS:
no) cc="$cc \$(CCSHARED) \$(PY_CFLAGS) \$(PY_CPPFLAGS)";;
while setup.py uses PY_CFLAGS_NODIST:
set_compiler_fla
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27434
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29159
___
Python tracker
<https://bugs.python.org/issu
New submission from Christian Heimes :
setup.py and Modules/Setup define a couple of macros for pyexpat and
_elementtree. The macros are no longer needed or should be moved to our
expat_config.h copy.
* HAVE_EXPAT_CONFIG_H is not used by our code and not used by system-wide expat
header
Change by Christian Heimes :
--
pull_requests: +27432
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29157
___
Python tracker
<https://bugs.python.org/issu
Christian Heimes added the comment:
The proposal is related to Brett's ticket bpo-45548. I no longer think that we
should define Py_BUILD_CORE_MODULE unconditionally. Instead I propose to move
the defines into each C module. This avoids duplication of macros in setup.py
and Modules/Setup
Christian Heimes added the comment:
Thanks for filing this feature request!
The DNS lookup part is out of scope for the ssl module. I don't want to get
into the DNS business. At $WORK I work on BIND, DNSSEC and DNS over TLS. Secure
DNS (and DNS in general) is already complicated when you
Christian Heimes added the comment:
The problem was fixed a couple of years ago. Python uses the reentrant
functions since commit 23e65b25557f957af840cf8fe68e80659ce28629 / GH-7081.
--
nosy: +christian.heimes
resolution: -> fixed
stage: -> resolved
status: open -> closed
s
Change by Christian Heimes :
--
type: -> enhancement
versions: -Python 3.6, Python 3.7, Python 3.8
___
Python tracker
<https://bugs.python.org/issu
Christian Heimes added the comment:
Andrew, could you please rebase your PR and get it submitted?
--
assignee: christian.heimes -> asvetlov
versions: +Python 3.11 -Python 3.10
___
Python tracker
<https://bugs.python.org/issu
Christian Heimes added the comment:
Fine with me! Thanks for your feedback.
--
resolution: -> rejected
stage: needs patch -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
Thankfully RDRAND is no longer a concern for us. All OpenSSL versions in
official Python.org installers and supported by recent Python versions have a
strong and fork-safe RNG.
--
resolution: -> fixed
stage: -> resolved
status: open -&g
Christian Heimes added the comment:
As Antoine said, you can use a partial function, closure, or an instance method
as a callable.
Besides nobody else has requested an additional argument in the past eight
years. I see this as a strong indication that an argument is not required
Christian Heimes added the comment:
We use os.urandom() because it is backed by a cryptographicly secure random
number generator. The random module uses a non-secure RNG. While RFC 4122 does
not mandate a CSRPNG, application often rely on unpredictable UUIDs.
Correctness and security
Christian Heimes added the comment:
I started
https://github.com/python/cpython/compare/main...tiran:configure_pkgconfig?expand=1
to integrate pkg-config with configure. The patchset also contains examples
how to pass flags from configure into Makefile and Modules/Setup.
--
nosy
Christian Heimes added the comment:
I need more information to assist you. Could you please provide:
- the steps how you compiled and install your custom OpenSSL installation
- the output of "ls /usr/lib64/openssl"
- your config.log file (please upload)
- the output of
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
OpenSSL 1.1.1 comes with sha512_256 and sha512_224. I propose that we use the
OpenSSL implementation and don't add the variants to our _sha512 module.
>>> hashlib.new("sha512_2
Change by Christian Heimes :
--
versions: +Python 3.11 -Python 3.4
___
Python tracker
<https://bugs.python.org/issue17924>
___
___
Python-bugs-list mailin
Christian Heimes added the comment:
Yes, the fix works well enough.
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
Yury, I don't remember how I found the issue.
It's still an issue in 3.9 and newer.
--
versions: +Python 3.10, Python 3.11, Python 3.9 -Python 3.6, Python 3.7
___
Python tracker
<https://bugs.python.
Christian Heimes added the comment:
Tests have been added as part of GH-24755.
--
resolution: -> fixed
stage: commit review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
I can still reproduce the issue with Python 3.9 and newer.
>>> import socket
>>> socket.socket(socket.AF_INET, socket.SOCK_STREAM).proto
0
--
type: -> behavior
versions: +Python 3.10, Python 3.11, Python 3.9 -Python 2.7, Py
Christian Heimes added the comment:
Victor, do we want to support "with open("somefile") as f: os.chmod(f, 0o644)"?
The feature request has been languishing for 8 years.
--
status: open -> pending
versions: +Python 3.11 -Python 3.4
___
Change by Christian Heimes :
--
versions: +Python 3.10, Python 3.11 -Python 3.8
___
Python tracker
<https://bugs.python.org/issue37505>
___
___
Python-bugs-list m
Christian Heimes added the comment:
There haven't been any activity on this feature request for eight years. I'm no
longer interested to implement my feature request. Closing...
--
resolution: -> rejected
stage: patch review -> resolved
status: open -&g
Christian Heimes added the comment:
2.7 and 3.5 are no longer supported
--
nosy: +christian.heimes
resolution: -> out of date
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
There is no agreement where to put the function. In the past seven years nobody
else has requested a pure Python implementation, too. I'm closing my feature
request as rejected.
--
resolution: -> rejected
stage: needs patch -> resolved
Christian Heimes added the comment:
My feature request has been around for 8 years without any progress. I don't
even recall why I needed the feature in the first place.
--
resolution: -> rejected
stage: patch review -> resolved
status: open -&g
Christian Heimes added the comment:
The issue has been languishing for 8 years. I'm closing it as "fixed".
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bug
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
New changeset 81520fe677d15cc7f9af5140bc5f9eca8409ad90 by Christian Heimes in
branch 'main':
bpo-45536: Check OpenSSL APIs in configure (GH-29088)
https://github.com/python/cpython/commit/81520fe677d15cc7f9af5140bc5f9eca8409ad90
Change by Christian Heimes :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Christian Heimes added the comment:
One of your optimization flags breaks the AX_C_FLOAT_WORDS_BIGENDIAN macro. The
optimized object file does not have seesnoon in its data section.
$ objdump -x conftest.o conftest_unoptimized.o
conftest.o: file format elf64-x86-64
conftest.o
Christian Heimes added the comment:
Your conftest.o does not contain the magic string "seesnoon". Either ICC uses a
different serialization format or your CFLAGS optimize the value away.
Please run
icc -c -o conftest_unoptimized.o conftest.c
and upload "conftes
Christian Heimes added the comment:
Please upload and attach the .o file to this bug. I need to see the content of
the file.
--
___
Python tracker
<https://bugs.python.org/issue45
Change by Christian Heimes :
--
type: -> enhancement
versions: +Python 3.10, Python 3.11
___
Python tracker
<https://bugs.python.org/issue14527>
___
___
Py
Christian Heimes added the comment:
The issue is fixed in Python 3 by commit
f72006f4429975a9d221e046e43dabd4f41eda23
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bug
Change by Christian Heimes :
--
versions: +Python 3.10, Python 3.11, Python 3.9 -Python 2.6, Python 2.7, Python
3.1, Python 3.2
___
Python tracker
<https://bugs.python.org/issue1760
Christian Heimes added the comment:
I'm closing this as duplicate.
--
resolution: -> duplicate
stage: -> resolved
status: open -> closed
superseder: -> XML vulnerabilities in Python
___
Python tracker
<https://bugs.python
Change by Christian Heimes :
--
versions: +Python 3.10, Python 3.11, Python 3.9 -Python 2.6
___
Python tracker
<https://bugs.python.org/issue7938>
___
___
Pytho
Christian Heimes added the comment:
Python 2 is no longer supported. Python 3's _PyLong_Add() function doesn't rely
on overflow.
--
nosy: +christian.heimes
resolution: -> out of date
stage: needs patch -> resolved
status: open -&g
Christian Heimes added the comment:
This is a very old bug report. Python has changed a lot in the past decade.
Please reopen the bug if you still can reproduce the issue with Python 3.9 or
newer.
--
nosy: +christian.heimes
resolution: -> out of date
stage: -> resolved
Christian Heimes added the comment:
Are Python 3.9 to 3.11 still affected by the issue?
The webbrowser module now uses subprocess.Popen() to start browsers. The
subprocess module calls wait() and performs additional cleanup to prevent
zombies.
--
nosy: +christian.heimes
status
Christian Heimes added the comment:
Does the issue still affect supported Python and Linux versions? Majority of
distros are using pipewire or pulseaudio these days.
--
nosy: +christian.heimes
status: open -> pending
versions: +Python 3.10, Python 3.11, Python 3.9 -Python
Christian Heimes added the comment:
Example output:
$ ./configure
checking whether compiling and linking against OpenSSL works... yes
checking for --with-openssl-rpath...
checking whether OpenSSL provides required APIs... yes
$ ./configure --with-openssl=/home/heimes/dev/python/multissl
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27354
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29088
___
Python tracker
<https://bugs.python.org/issu
New submission from Christian Heimes :
In thread [1] Robin Becker requested to check for working OpenSSL in configure
script. With reasonable effort it is possible to probe for basic APIs such as
minimum SSL and EVP interface.
[1]
https://mail.python.org/archives/list/python-...@python.org
Christian Heimes added the comment:
CPython internals can still use allocation on the stack. Only stable ABI
extensions have to use allocation on the heap.
--
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
A consumer will use the APIs:
---
Py_buffer *view;
int ndim;
const char *format;
const Py_ssize_t *shape, *strides, *suboffsets;
void *buf;
view = PyBuffer_New();
PyObject_GetBuffer(obj, view, flags);
ndim = PyBuffer_GetLayout(, , , );
buf
Christian Heimes added the comment:
All memory is owned by the exporter object. The exporter (aka producer) is the
Python type that implements Py_bf_getbuffer and Py_bf_releasebuffer. In
majority of cases the exporter doesn't have to set shape, strides, and
suboffsets. They are used
Change by Christian Heimes :
--
pull_requests: +27352
pull_request: https://github.com/python/cpython/pull/29086
___
Python tracker
<https://bugs.python.org/issue45
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27327
pull_request: https://github.com/python/cpython/pull/29056
___
Python tracker
<https://bugs.python.org/issue45
New submission from Christian Heimes :
Freelists are an optimization trick to avoid allocation and deallocation of
commonly used structures. Currently Python has freelists for frame, tuple,
float, list, dict, async generators, and context objects. Small ints are also
cached
Christian Heimes added the comment:
CC Antoine for his expertise of the buffer protocol
Opaque Py_Buffer and PyObject structs will require a different approach and
prevent some optimizations. The consumer will have to malloc() a Py_buffer
struct on the heap. In non-trivial cases
Change by Christian Heimes :
--
keywords: +patch
pull_requests: +27306
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/29035
___
Python tracker
<https://bugs.python.org/issu
Christian Heimes added the comment:
Please build conftest.o with compiler flags
icc -c -O3 -fPIC -fp-model strict -fp-model source -axCORE-AVX512,CORE-AVX2
-xAVX -ipo -prec-div -prec-sqrt
and attach the file.
--
___
Python tracker
<ht
Christian Heimes added the comment:
Python uses the AX_C_FLOAT_WORDS_BIGENDIAN macro from autoconf-archive to
detect byte order of floats. You can find the code at
https://www.gnu.org/software/autoconf-archive/ax_c_float_words_bigendian.html .
Please attach your config.log file
Christian Heimes added the comment:
We generally prefer to have only way way to accomplish a goal. If PEP 648 gets
accepted then your proposal won't be accepted. In case the PEP is rejected,
then your proposal may still be rejected or may require a new PEP. Preloading
has some complicated
Christian Heimes added the comment:
There is also https://www.python.org/dev/peps/pep-0648/
--
nosy: +christian.heimes
___
Python tracker
<https://bugs.python.org/issue45
Christian Heimes added the comment:
I'm glad you figured out the problem and fixed your system.
Python uses autoconf (configure) to detect the presence of features. It's a
standard approach. Since you had a stropts.h in your include directory,
configure declared HAVE_STROPTS_H. The system
Christian Heimes added the comment:
You are correct. To be more precise, it's caused by GH-16856. The commit
introduced PyAttributeErrorObject struct. Since AttributeError and OSError now
have an incompatible memory layout, it is no longer possible to create a
subclass that has
Christian Heimes added the comment:
You are running in a typical scope issue. The local and global scope of a
lambda work differently than you expect. You can work around the issue by
making fun a local variable:
for name, fun in namefun:
name2fun[name] = lambda x, fun=fun: fun(x
Christian Heimes added the comment:
This is not a problem that we can fix. Third party packages like numpy and
tensorflow have not build and publish builds for Python 3.10 for all platforms
yet.
--
nosy: +christian.heimes
resolution: -> third party
status: open ->
701 - 800 of 6455 matches
Mail list logo
