Charalampos Stratakis added the comment:
SSLContext.minimum_version is added here on the master branch:
https://github.com/python/cpython/commit/698dde16f60729d9e3f53c23a4ddb8e5ffe818bf
But I'd be also reluctant to partially backport a new feature to fix the test
Charalampos Stratakis added the comment:
Getting those failures on RHEL8 as well, which can be worked around by setting
the env OPENSSL_CONF=/non-existing-file
==
ERROR: test_protocol_sslv23 (test.test_ssl.ThreadedTests
Charalampos Stratakis added the comment:
Fedora 29 has openssl 1.1.1 which seems to be related.
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue35
Change by Charalampos Stratakis :
--
pull_requests: +11913
___
Python tracker
<https://bugs.python.org/issue33570>
___
___
Python-bugs-list mailing list
Unsub
Change by Charalampos Stratakis :
--
pull_requests: +11911
___
Python tracker
<https://bugs.python.org/issue29136>
___
___
Python-bugs-list mailing list
Unsub
Change by Charalampos Stratakis :
--
pull_requests: +11910
___
Python tracker
<https://bugs.python.org/issue32947>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
Hi Brett,
Those issues should be reported on Red Hat's bugzilla, if it's definite that
the kernel version (or a new cpython release downstream) is the one to blame
for it.
Also RHEL 7's python version while being 2.7.5, carries
Change by Charalampos Stratakis :
--
pull_requests: +11696, 11698
___
Python tracker
<https://bugs.python.org/issue1294959>
___
___
Python-bugs-list mailin
Change by Charalampos Stratakis :
--
pull_requests: +11696, 11697
___
Python tracker
<https://bugs.python.org/issue1294959>
___
___
Python-bugs-list mailin
Change by Charalampos Stratakis :
--
pull_requests: +11696
___
Python tracker
<https://bugs.python.org/issue1294959>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
Closing this as it's not really a bug in the code, and I don't think spending
too much time on python2 is worth it.
--
stage: -> resolved
status: open -> closed
___
Python tracker
<http
Charalampos Stratakis added the comment:
Possibly relevant:
https://fedoraproject.org/wiki/Changes/PPC64LE_Float128_Transition#Detailed_Description
But the work is not complete.
--
___
Python tracker
<https://bugs.python.org/issue35
Change by Charalampos Stratakis :
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue35752>
___
___
Python-bugs-list mailing list
Unsubscribe:
Charalampos Stratakis added the comment:
Indeed it's not a bug per se, more like code readability issue, if however it's
not deemed as an issue, it can be closed.
--
___
Python tracker
<https://bugs.python.o
New submission from Charalampos Stratakis :
Results from a recent static analysis scan for python2:
Error: USE_AFTER_FREE (CWE-825):
Python-2.7.15/Modules/_bsddb.c:6697: freed_arg: "free" frees "name".
Python-2.7.15/Modules/_bsddb.c:6715: pass_freed_arg: Passing freed p
Charalampos Stratakis added the comment:
Small correction. This regression has been on 3.7 for some time now (since it
was the master branch then), but then I requested to have the buggy commit
backported to 3.6 to fix the --with-lto flag there. Which unfortunately
introduced the issue to
Charalampos Stratakis added the comment:
And also 3.7.2rc1
--
___
Python tracker
<https://bugs.python.org/issue35257>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
This change fixes a regression introduced in 3.6.8rc1 with
https://bugs.python.org/issue31354
--
___
Python tracker
<https://bugs.python.org/issue35
Charalampos Stratakis added the comment:
Maybe a release blocker for 3.7.2 and 3.6.8?
--
nosy: +cstratak, ned.deily
___
Python tracker
<https://bugs.python.org/issue35
Charalampos Stratakis added the comment:
The PR is pending another round of review.
--
___
Python tracker
<https://bugs.python.org/issue35257>
___
___
Python-bug
Charalampos Stratakis added the comment:
Hi Ned,
I recently pushed a fix on the master and 3.7 for this exact issue:
https://bugs.python.org/issue35351 but it builds on top of
https://github.com/python/cpython/pull/10922 which is not yet in 3.6.
Thus 3.7 is fine for the rc cutoff.
However
Change by Charalampos Stratakis :
--
title: Add LDFLAGS_NODIST for the LDFLAGS not intended for propagation to C
extensions. -> Avoid leaking linker flags into distutils.
___
Python tracker
<https://bugs.python.org/issu
Charalampos Stratakis added the comment:
PR has been finalized.
--
nosy: +vstinner
___
Python tracker
<https://bugs.python.org/issue35257>
___
___
Python-bug
Change by Charalampos Stratakis :
--
pull_requests: +10161
___
Python tracker
<https://bugs.python.org/issue28015>
___
___
Python-bugs-list mailing list
Unsub
Change by Charalampos Stratakis :
--
pull_requests: +10161, 10162, 10163
___
Python tracker
<https://bugs.python.org/issue28015>
___
___
Python-bugs-list mailin
Change by Charalampos Stratakis :
--
pull_requests: +10161, 10162
___
Python tracker
<https://bugs.python.org/issue28015>
___
___
Python-bugs-list mailin
Charalampos Stratakis added the comment:
And here is the difference between compiling the extension with the current
tip, comparing to applying my current draft PR:
Master branch with the linker flags propagated:
running build
running build_ext
building 'demo' extension
crea
Charalampos Stratakis added the comment:
So to better illustrate the actual issue I'll be using an example from the
python documentation [0][1].
Get the demo.c and the setup.py. Compile cpython first with --with-lto and then
compile the demo.c with ./python3 setup.py build.
You will n
Change by Charalampos Stratakis :
Added file: https://bugs.python.org/file47977/setup.py
___
Python tracker
<https://bugs.python.org/issue35257>
___
___
Python-bug
Change by Charalampos Stratakis :
--
keywords: +patch
pull_requests: +10140
stage: -> patch review
___
Python tracker
<https://bugs.python.org/issu
Charalampos Stratakis added the comment:
Also on Fedora the same set of security policies can be set as RHEL8 by
utilizing 'update-crypto-policies --set NEXT'
--
___
Python tracker
<https://bugs.python.o
Change by Charalampos Stratakis :
--
keywords: +patch
pull_requests: +10067
stage: -> patch review
___
Python tracker
<https://bugs.python.org/issu
Charalampos Stratakis added the comment:
I believe I figured out the issue, at least on the master branch.
While checking the certificates used by asyncio tests within the test_asyncio
folder I noticed they were quite outdated when compared to the more recent
updated ones with the test
Charalampos Stratakis added the comment:
It seems I can reproduce it on Fedora as well by setting stronger crypto
defaults through 'update-crypto-policies --set FUTURE'.
Repo located here:
https://gitlab.com/redhat-crypto/fedora-crypto-policies/tree/master
The changes are many
Change by Charalampos Stratakis :
Added file: https://bugs.python.org/file47957/asyncio_failures_3.6.log
___
Python tracker
<https://bugs.python.org/issue35352>
___
___
Charalampos Stratakis added the comment:
Seems that 3.8 and 3.7 get the same failures, while 3.6 gets 14 instead of 8.
--
___
Python tracker
<https://bugs.python.org/issue35
New submission from Charalampos Stratakis :
When compiling python3 on RHEL8 and run the tests, test_asyncio fails with 8
failures (and more on older branches).
openssl version is 1.1.1
Attaching the failures for the various branches.
--
components: Tests, asyncio
files
Change by Charalampos Stratakis :
Added file: https://bugs.python.org/file47956/asyncio_failures_3.7.log
___
Python tracker
<https://bugs.python.org/issue35352>
___
___
Change by Charalampos Stratakis :
--
keywords: +patch
pull_requests: +10044
stage: -> patch review
___
Python tracker
<https://bugs.python.org/issu
New submission from Charalampos Stratakis :
Through configure.ac and this change [0] LTOFLAGS and more specifically -flto
is baked into BASECFLAGS when using link time optimizations, which will make
this flag propagate to distutils.
[0]
https://github.com/python/cpython/commit
Change by Charalampos Stratakis :
--
pull_requests: +9851
stage: commit review -> patch review
___
Python tracker
<https://bugs.python.org/issue28043>
___
_
Change by Charalampos Stratakis :
--
pull_requests: +9850
___
Python tracker
<https://bugs.python.org/issue33570>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
Correction: The second commit is referring to
https://github.com/python/cpython/pull/9908/
--
___
Python tracker
<https://bugs.python.org/issue35
New submission from Charalampos Stratakis :
Through acb8c5234302f8057b331abaafb2cc8697daf58f the CFLAGS_NODIST variable was
created, in order to place there compiler flags used by the interpreter, but
not intended to be propagated to C extensions.
I saw a similar issue when working on
Charalampos Stratakis added the comment:
I'd say yes. In general python may have been compiled with -flto, but it's
still a bit buggy from the compilers' side. It doesn't work well always or at
all depending on the toolchain, and even if python was compiled with -flto
Charalampos Stratakis added the comment:
This change exports -flto in cflags.
You can check it with python3-config --cflags after this commit.
Which means that every c extension which will use those cflags will also
utilize -flto which wasn't happening before, thus I'd say it'
Change by Charalampos Stratakis :
--
pull_requests: +9697
___
Python tracker
<https://bugs.python.org/issue31625>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
The workaround indeed works.
--
___
Python tracker
<https://bugs.python.org/issue28015>
___
___
Python-bugs-list mailin
Charalampos Stratakis added the comment:
The issue with the 3.6 branch is https://bugs.python.org/issue31625 which
hasn't been backported to 3.6. Would it make sense to backport it?
I have the backports ready locally but not sure if pushing
https://bugs.python.org/issue31625 to 3
Charalampos Stratakis added the comment:
Would it make sense to backport that to 3.6 as well? Currently this blocks
https://bugs.python.org/issue28015 from being backported to 3.6
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.
Change by Charalampos Stratakis :
--
pull_requests: +9572
___
Python tracker
<https://bugs.python.org/issue31354>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
Alright the static scanner warns in the case of a typo being made where the
variable is actually used, but since the variable is indeed not used, that
doesn't not make it a bug.
--
___
Python tr
Charalampos Stratakis added the comment:
Generic reference on the issue:
https://github.com/koalaman/shellcheck/wiki/SC2034
--
___
Python tracker
<https://bugs.python.org/issue34
Charalampos Stratakis added the comment:
Adding to this, the reason I initially caught up with that, was due to a
coverity scan.
More explicitly:
Error: SHELLCHECK_WARNING: [#def1]
/usr/lib64/python3.6/config-3.6dm-x86_64-linux-gnu/install-sh:63:1: warning:
transform_arg appears unused
Charalampos Stratakis added the comment:
Hello Shivank. I had a PR ready locally which I was about to push, so you
posted just at the right time :) Feel free to work on this issue.
--
___
Python tracker
<https://bugs.python.org/issue35
New submission from Charalampos Stratakis :
Analyzing some coverity scan results I stumbled upon this issue:
Python-3.6.5/Lib/xml/dom/minidom.py:1914: original:
"n._call_user_data_handler(operation, n, notation)" looks like the original
copy.
Python-3.6.5/Lib/xml/dom/minid
New submission from Charalampos Stratakis :
There is an Python test that checks the minimum TLS version -
test_min_max_version (test.test_ssl.ContextTests).
Fedora 29+ sets TLSv1 as explicit minimum version. Python's test suite assumes
that the minimum protocol version is set to a
Change by Charalampos Stratakis :
--
pull_requests: +9301
___
Python tracker
<https://bugs.python.org/issue34623>
___
___
Python-bugs-list mailing list
Unsub
Change by Charalampos Stratakis :
--
pull_requests: +9284
___
Python tracker
<https://bugs.python.org/issue34623>
___
___
Python-bugs-list mailing list
Unsub
Charalampos Stratakis added the comment:
Will this change be backported to 3.5 and 3.4? It applied cleanly on both
however on 3.4 there is a test failure:
==
ERROR: test_del_attribute (test.test_xml_etree_c.MiscTests
Change by Charalampos Stratakis :
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue34834>
___
___
Python-bugs-list mailing list
Unsubscribe:
Change by Charalampos Stratakis :
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue34836>
___
___
Python-bugs-list mailing list
Unsubscribe:
Change by Charalampos Stratakis :
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue34818>
___
___
Python-bugs-list mailing list
Unsubscribe:
Change by Charalampos Stratakis :
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue34814>
___
___
Python-bugs-list mailing list
Unsubscribe:
Charalampos Stratakis added the comment:
Tested so far only on a 3.7.0 build.
--
versions: -Python 3.8
___
Python tracker
<https://bugs.python.org/issue34
Change by Charalampos Stratakis :
--
keywords: +patch
pull_requests: +8990
stage: -> patch review
___
Python tracker
<https://bugs.python.org/issu
Charalampos Stratakis added the comment:
FAIL: test_NULL_ob_type (test.test_gdb.PrettyPrintTests)
Ensure that a PyObject* with NULL ob_type is handled gracefully
--
Traceback (most recent call last):
File "/builddir/
Change by Charalampos Stratakis :
--
versions: +Python 3.7, Python 3.8
___
Python tracker
<https://bugs.python.org/issue34811>
___
___
Python-bugs-list mailin
New submission from Charalampos Stratakis :
In Fedora we got a new build of gdb which makes python's test_gdb fail on
x86_64, i686 and aarch64(arm 64 bits) architectures.
gdb's commits between the passing and failing tests:
https://github.com/bminor/binutils-g
New submission from Charalampos Stratakis :
The install-sh file that python uses for autotools is horribly outdated. Last
update was 16 years ago and it's being copied from automake's source code.
Updating it to modern standards could potentially fix issues for systems that
use
Charalampos Stratakis added the comment:
As mentioned before, this is due to missing libffi-devel on a Centos 7 system.
Verified it as well on one of my systems.
Could you try again, but before just do a 'yum install libffi-devel'?
Maybe a warning could be added to the install o
Charalampos Stratakis added the comment:
Yes test_poplib and test_ftplib on fedora rawhide when run against openssl
1.1.1 pre8. Haven't tried the pr7, but assuming that the tests were fine before
here is the list of changes between pre7 and pre8:
https://github.com/openssl/openssl/co
Charalampos Stratakis added the comment:
On Fedora after the fix we had also to BuildRequire libnsl2-devel and
libtirpc-devel after the glibc change was implemented for the correct headers
to be included within the buildroot.
I am not familiar with suse's build system webui to find the
Charalampos Stratakis added the comment:
Matej is this about Python 2? Because the solution didn't actually work for
Python 2 and on Fedora we had to implement a workaround [0]. Unfortunately
there weren't enough free cycles so far to investigate further.
[0]
https://src.fedorap
Charalampos Stratakis added the comment:
I'd say there are use cases where gdb will be used with optimizations
especially in downstream distribution.
--
___
Python tracker
<https://bugs.python.org/is
Charalampos Stratakis added the comment:
The latest stable Fedora's have glibc >= 2.26
Maybe the buildbot needs to be updated?
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org
Charalampos Stratakis added the comment:
Ping. Could someone take a look? There is a PR ready.
--
___
Python tracker
<https://bugs.python.org/issue32962>
___
___
Charalampos Stratakis added the comment:
Is it possible/feasible to fix that for the 3.7 and 3.6 branches as well?
--
___
Python tracker
<https://bugs.python.org/issue33
Change by Charalampos Stratakis :
--
nosy: +Dormouse759
___
Python tracker
<https://bugs.python.org/issue32962>
___
___
Python-bugs-list mailing list
Unsubscribe:
Charalampos Stratakis added the comment:
This is an issue with the stack size.
It was encountered recently while building Python 3.6 under CentOS 6 [0] and
the way to fix it was to increase the maximum stack size using ulimit e.g. [1]
[0] https://bugzilla.redhat.com/show_bug.cgi?id=1572150
Charalampos Stratakis added the comment:
The bug exists indeed but I would like to know what is the actual use case here.
--
___
Python tracker
<https://bugs.python.org/issue11
Charalampos Stratakis added the comment:
I could make a PR out of the patch but I wonder if anyone is actually using
this functionality. Wearing my Fedora's python maintainer hat, we have a lot of
tooling around rpm that handles those things, so I don't think there is a valid
u
Charalampos Stratakis added the comment:
A PR has been filed, could someone take a look?
--
___
Python tracker
<https://bugs.python.org/issue29673>
___
___
Pytho
Change by Charalampos Stratakis :
--
nosy: +cstratak
___
Python tracker
<https://bugs.python.org/issue29673>
___
___
Python-bugs-list mailing list
Unsubscribe:
Charalampos Stratakis added the comment:
We are getting hit by that quite often on Fedora, with the transition to gcc 8
and it creates unnecessary noise at our build logs. Thanks for working on that.
When you sent your PR I can test it within our build system and verify if it
works
Charalampos Stratakis added the comment:
The bug is still present on the 3.6 branch.
--
___
Python tracker
<https://bugs.python.org/issue29804>
___
___
Python-bug
Charalampos Stratakis added the comment:
Tested the fix and ctypes is linked successfully with libdl when utilizing the
strict symbol check.
--
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.p
Charalampos Stratakis added the comment:
Reopenning the issue.
python2 is still failing with the fix applied:
gcc -pthread -fPIC -fno-strict-aliasing -O2 -g -pipe -Wall
-Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS
-fexceptions -fstack-protector-strong -grecord
Charalampos Stratakis added the comment:
Attaching the build log from python2. Near the end you can see the undefined
references errors.
--
Added file: https://bugs.python.org/file47426/py2buildlog
___
Python tracker
<https://bugs.python.
Charalampos Stratakis added the comment:
Just tested it for python3 on a rawhide system with the flag enabled and it
works, ctypes is compiled successfully.
python2 is not affected the same way though, various modules fail there with
different undefined references issues when setting the
Charalampos Stratakis added the comment:
@Stephane,
Currently the flag has been disabled on rawhide due to too many breakages but
the bug is still there if you add '-z defs' to the linker flags.
More info:
https://src.fedoraproject.org/rpms/redhat-rp
Charalampos Stratakis added the comment:
So what would be the best way to proceed from here?
Add a check or just link to dl unconditionally?
--
___
Python tracker
<https://bugs.python.org/issue32
Change by Charalampos Stratakis :
--
pull_requests: +5374
___
Python tracker
<https://bugs.python.org/issue9216>
___
___
Python-bugs-list mailing list
Unsub
Change by Charalampos Stratakis :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Charalampos Stratakis added the comment:
Unfortunately when trying to reproduce it by compiling from source and adding
the '-z defs' flag to the linker, undefined symbol issues appeared for all the
extension modules of the stdlib.
--
components: +Extensi
Charalampos Stratakis added the comment:
Relevant part from the build log:
running build_ext
building '_ctypes' extension
gcc -pthread -Wno-unused-result -Wsign-compare -DDYNAMIC_ANNOTATIONS_ENABLED=1
-DNDEBUG -O2 -g -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOU
Change by Charalampos Stratakis :
--
components: +Build
___
Python tracker
<https://bugs.python.org/issue32647>
___
___
Python-bugs-list mailing list
Unsubscribe:
New submission from Charalampos Stratakis :
Original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1537489
As soon as strict symbol checking was enabled on the development branch of
Fedora, our build of CPython started failing due to undefined symbols when
compiling ctypes.
Build
Change by Charalampos Stratakis :
--
pull_requests: +5130
stage: -> patch review
___
Python tracker
<https://bugs.python.org/issue32635>
___
___
Python-
Change by Charalampos Stratakis :
--
components: +Extension Modules
___
Python tracker
<https://bugs.python.org/issue32635>
___
___
Python-bugs-list mailin
101 - 200 of 293 matches
Mail list logo