New submission from Gregory P. Smith:
The problem: without the stack being 16-byte aligned, code generated by modern
compilers like recent gcc/g++ or clang assumed that the stack is 16 byte
aligned and uses SSE instructions in some circumstances that require this.
Without this fix, any
Gregory P. Smith added the comment:
http://bugs.python.org/issue17245 filed to track the stack alignment issue.
The only reason i set this as release blocker is to let a release manager
decide which of these two issues to proceed with for 2.7.4 and 3.2.4 (and
3.3.1
Gregory P. Smith added the comment:
yeah that looks like stdin was closed or was /dev/null.
--
resolution: -> invalid
status: open -> closed
___
Python tracker
<http://bugs.python.org/i
Gregory P. Smith added the comment:
File "/home/riley/pywikipedia/userinterfaces/terminal_interface_base.py", line
129, in input
text = getpass.getpass('')
you can't call getpass without stdin or a terminal and expect it to do
anything. your problem is in
New submission from Gregory P. Smith:
bytearray has many methods that return a *new* bytearray object rather than
applying the transformation to modify the bytearray in place. Given that one
use of bytearray's is to avoid making extra copies... There should be in-place
variants o
Gregory P. Smith added the comment:
Translate isn't a text operation. (That's the api I wanted). The others I
only lists for completeness because someone else would complain if I
hadn't. ;)
On Mar 1, 2013 12:57 PM, "Terry J. Reedy" wrote:
>
> Terry J. Reedy add
Gregory P. Smith added the comment:
Yes! 2to3 features are allowed in stable releases.
On Mar 2, 2013 5:17 AM, "Ezio Melotti" wrote:
>
> Ezio Melotti added the comment:
>
> Can this still go on 2.7/3.2/3.3?
>
> --
>
> ___
Gregory P. Smith added the comment:
looks sane and cleaner than the silly x32 hack in there, i'll take care of it.
FYI - your patch forgot to add the "unsigned char d_type;" struct member before
name. fixed and applied.
When i wrote that code I was probably assuming that st
Gregory P. Smith added the comment:
The bytes (and bytearray?) version of this should generate memoryview's instead
of new bytes objects to avoid a copy.
While not required, It'd be useful if the implementation of this pre-scanned
the data internally so that the length of the
Gregory P. Smith added the comment:
Indeed, a bytearray version would require the talked about but not implemented
due to complexity (in pep3118) support for locking a buffer from other
mutations. best concentrate on bytes then.
Do we have a memoryview equivalent for PyUnicode? If not, we
Gregory P. Smith added the comment:
The important reasons for this are memory use and cache thrashing, not just
CPU time.
--
___
Python tracker
<http://bugs.python.org/issue17
Gregory P. Smith added the comment:
The gain will be more noticeable the faster the Python implementation it is
running under is. It is going to avoid logN relloc's in just about all
implementations. That CPython is relatively slow is not a justification to
avoid adding the feature.
I
Changes by Gregory P. Smith :
--
versions: +Python 3.4 -Python 3.3
___
Python tracker
<http://bugs.python.org/issue13229>
___
___
Python-bugs-list mailin
Gregory P. Smith added the comment:
I'd like to take care of this at Python. At least for posix (someone else can
deal with the windows side if they want).
I just stumbled upon an extension module at work that someone wrote
specifically because os.listdir consumed way too much r
Gregory P. Smith added the comment:
right he has a separate issue open tracking the walkdir stuff in
issue13229. I saw it first before finding this issue which is exactly what
I wanted. :)
--
___
Python tracker
<http://bugs.python.org/issue11
Gregory P. Smith added the comment:
It'd perhaps have been better if things like memoryview were never exposed to
the user at all as a distinct type and became an internal implementation detail
behind PyBytes and PyUnicode objects (they could hold a reference to something
else or col
Changes by Gregory P. Smith :
--
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue3329>
___
___
Python-bugs-list mailing list
Unsubscribe:
Gregory P. Smith added the comment:
If this is done, it should probably be on by default on all --with-pydebug
buildbots. Otherwise I suspect nobody will _ever_ look at its output and we
should just remove the feature all together.
Being off in the main process on the build bots would still
Gregory P. Smith added the comment:
I'd personally say don't bother with this. Let people who _need_ this use
their own C extension modules to handle all secure data as we're not in a
position to make and test any guarantees about what happens to data anywhere
within a Python
Changes by Gregory P. Smith :
--
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue3982>
___
___
Python-bugs-list mailing list
Unsubscribe:
Changes by Gregory P. Smith :
--
nosy: gregory.p.smith
priority: normal
severity: normal
status: open
title: imaplib.IMAP4_stream subprocess is opened unbuffered but ignores short
reads
___
Python tracker
<http://bugs.python.org/issue17
New submission from Gregory P. Smith:
imaplib.IMAP4_stream subprocess is opened unbuffered but ignores short reads
when reading the message body. Depending on timing, message body size and
kernel pipe buffer size and phase of the moon and whether you're debugging the
thing or not... I
Gregory P. Smith added the comment:
The error does not happen when running the same code under 2.7, despite the
same default bufsize=0 subprocess behavior. This is likely due to differences
in the Python 2.x old style io library when os.fdopen(fd, 'rb', bufsize) is
used vs 3.x wh
Gregory P. Smith added the comment:
Yes imaplib can be fixed pretty easily and should use buffered IO regardless.
I'm pondering if the default behavior of subprocess needs fixing as
existing python 2.x code being ported to 3 doesn't expect this changed
behavior of the PIPE file ob
Changes by Gregory P. Smith :
--
assignee: -> gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue17192>
___
___
Python-bugs-list mailing list
Un
Gregory P. Smith added the comment:
I'm taking care of this with a libffi update per #17192 and #19245. Thanks for
the upstream fix into libffi 3.0.13 alex. :)
--
nosy: +gregory.p.smith
resolution: -> duplicate
status: open -> closed
___
Changes by Gregory P. Smith :
--
superseder: -> ctypes libffi needs to align the x86 stack to 16 bytes
___
Python tracker
<http://bugs.python.org/issu
Changes by Gregory P. Smith :
--
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue14803>
___
___
Python-bugs-list mailing list
Unsubscribe:
Changes by Gregory P. Smith :
--
nosy: +jcea -gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue17477>
___
___
Python-bugs-list mailing list
Unsub
Gregory P. Smith added the comment:
that patch looks good for imaplib.
i'll follow up on the subprocess side of things to see if the default
behavior should be changed to better match what happened in 2.7 (or if not:
to make sure the change in behavior is sufficiently documented and not
r
Gregory P. Smith added the comment:
fyi, I un-cc'd myself as I don't maintain this module anymore; jcea does. :)
I agree with the patch. Support for anything older than 4.3 is completely
pointless in 2.7 and supporting 5.x is likely required for future OS distros
wanting t
Gregory P. Smith added the comment:
2.7, 3.2, 3.3 and 3.4 have all be updated to ctypes v3.0.13.
fwiw, I tried a build of 3.3 on Sparc Solaris 10 using snakebite (nitrogen) and
ctypes builds and links but two tests fail: ctypes.test.test_bitfields.C_Test
test_ints and test_shorts failed
Changes by Gregory P. Smith :
--
nosy: -gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue17477>
___
___
Python-bugs-list mailing list
Unsubscribe:
Gregory P. Smith added the comment:
I've updated libffi to v3.0.13 which contains the fixes for this in 2.7, 3.2,
3.3 and 3.4.
if someone wants to apply a surgical patch to fix this problem, the one
mentioned above is good as well as:
https://github.com/atgreen/libffi/c
New submission from Gregory P. Smith:
The subprocess module in Python 3 uses io.open(file_descriptor, mode, bufsize)
to create its Popen stdout, stderr and stdin file objects.
In Python 2, it used the old os.fdopen which created an old-style python 2 file
object that simply wraps libc's
Gregory P. Smith added the comment:
fyi - sem_open is unrelated, nothing in subprocess needs that.
--
___
Python tracker
<http://bugs.python.org/issue12
Gregory P. Smith added the comment:
Some way for me to reproduce this is needed.
--
stage: -> test needed
___
Python tracker
<http://bugs.python.org/issu
Changes by Gregory P. Smith :
--
resolution: -> fixed
stage: needs patch -> commit review
status: open -> closed
___
Python tracker
<http://bugs.python.or
Gregory P. Smith added the comment:
fyi - subprocess uses /proc to get a list of open fd's in the child process
before exec when possible. search for FD_DIR in
http://hg.python.org/cpython/file/d674bbbed333/Modules/_posixsubprocess.c
--
___
P
Gregory P. Smith added the comment:
It looks like someone has already done the Python-ast cleanup and I don't
understand the point of the distutils test one or the libffi one (ffi has been
updated but it's trivial and would still apply).
--
nosy: +grego
Gregory P. Smith added the comment:
I'm not inclined to apply the patch to ffi unless someone can demonstrate that
it is an actual problem. given that nobody has applied it to *upstream* libffi
(since we just pulled in v3.0.13 earlier this week) I doubt it matters to
anyone.
everything
Gregory P. Smith added the comment:
The number of things we'll break by changing this errant behavior to be
_correct_ is way less than the number of things that are already broken due to
it.
If the bufsize=0 default is left in place the behavior differs between Windows
and POSIX plat
Gregory P. Smith added the comment:
(actually I'm not sure about the windows vs posix behavior difference, that may
not be true; I don't have a windows system handy to test that on)
--
___
Python tracker
<http://bugs.python.o
Gregory P. Smith added the comment:
great! fixing now. :)
--
___
Python tracker
<http://bugs.python.org/issue17488>
___
___
Python-bugs-list mailing list
Unsub
Changes by Gregory P. Smith :
--
resolution: -> fixed
stage: -> committed/rejected
status: open -> closed
___
Python tracker
<http://bugs.python.or
Gregory P. Smith added the comment:
the pythonrun issue had already been fixed. the others were still there (valid
bugs, but were unlikely to be causing problems given the codepaths). fixed.
I didn't write a Misc/NEWS entry on the commit because I didn't know how to
usefully desc
Gregory P. Smith added the comment:
Here's is an os.scandir(path='.') implementation that iterates reading the
directory on the fly instead of pre-building a list.
os.listdir's implementation should ultimately be replaced by this as:
def listdir(path=None):
if path is
Gregory P. Smith added the comment:
you'll see my code already has TODOs in there for that. windows API
documentation suggests that windows returns even more (stat-like) info when
traversing a directory. returning a namedtuple with the relevant info from the
platform at hand would be
Gregory P. Smith added the comment:
Your objection is noted but it is wrong.
A Python program today cannot process arbitrarily large directories within a
fixed amount of ram today due to os.listdir. This makes it unsuitable for file
system cleanup tasks that we have run into on production
Gregory P. Smith added the comment:
phew, yay. because i didn't see anything obviously wrong with the #16962
commit.
also, regarding checking for EINTR on some of the close() calls. It isn't a
big deal. This _posixsubprocess.c code that contains those checks is always
execut
Changes by Gregory P. Smith :
--
nosy: +twouters
___
Python tracker
<http://bugs.python.org/issue11406>
___
___
Python-bugs-list mailing list
Unsubscribe:
Gregory P. Smith added the comment:
Here's an updated patch that fixes the windows build based on twouters'
comments. It goes ahead and removes the old C implementation of listdir in
favor of the trivial Python wrapping of scandir.
--
Added file: http://bugs.python.org
Gregory P. Smith added the comment:
While i don't personally like things with iter in the name I won't object.
That way the scandir name would be left available for a future version of this
that yields namedtuples of directory entry details as Martin wa
Changes by Gregory P. Smith :
--
title: Python 2.7.4 Breaks ZipFile Extraction -> Python 2.7.4 breaks ZipFile
extraction of zip files with unicode member paths
versions: +Python 3.2, Python 3.3, Python 3.4
___
Python tracker
<http://bugs.pyth
Gregory P. Smith added the comment:
> > Note that the patch uses type punning through a union
>
> What is the standard and portable way to cast an array of bytes to size_t?
I'd expect just casting the pointer type before dereferencing:
unsigned char *p;
...
hash = (
Changes by Gregory P. Smith :
--
priority: high -> release blocker
___
Python tracker
<http://bugs.python.org/issue17656>
___
___
Python-bugs-list mai
Changes by Gregory P. Smith :
--
nosy: +benjamin.peterson, larry
priority: normal -> release blocker
___
Python tracker
<http://bugs.python.org/issu
Gregory P. Smith added the comment:
consensus here appears to be "bad idea... don't do this."
--
nosy: +gregory.p.smith
priority: high -> normal
resolution: -> wont fix
stage: -> committed/rejected
status: open -> closed
__
Gregory P. Smith added the comment:
okay, but i don't personally find any of these to be good ideas as "codecs"
given they don't have anything to do with translating between bytes<->unicode.
--
resolution: wont fix ->
stage: committed/reject
Changes by Gregory P. Smith :
--
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue17826>
___
___
Python-bugs-list mailing list
Unsubscribe:
Gregory P. Smith added the comment:
it _looks_ like the libffi updates i put in 3.2, 3.3 and default (3.4) are
messed up with some changes missing in 3.2 and more missing in 3.3 and 3.4.
(i'm comparing vs upstream 3.0.13 <-> 2.7 <-> 3.2 <-> 3.3 <-> 3.4).
i blame
Gregory P. Smith added the comment:
I've got this and all of the fixes necessary for 3.3, default and 2.7 in my
local repositories.
odd thing in the 3.2 patch: it is entirely unclear to me if the issue 10309 fix
needs to be kept. i included it as it was missing from 3.2 after pre
Changes by Gregory P. Smith :
--
resolution: -> fixed
status: open -> closed
___
Python tracker
<http://bugs.python.org/issue17192>
___
___
Python-bugs-
Gregory P. Smith added the comment:
Actually I'm thinking this duck may only have a beak. Instead of a bunch of
fields of None I'd prefer just not having that attribute defined on the
object. I consider the os specific "stat-like" info from reading a
directory to be so
Gregory P. Smith added the comment:
The recommendation for anyone using regular expressions on hostile input is to
(a) don't do that. (b) use a better regexp without this possible behavior and
(c) use something like re2 (there's a Python binding at
https://github.com/axiak/pyre2)
Gregory P. Smith added the comment:
Indeed, doing this _without a regexp_ is preferred. :)
--
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue17
New submission from Gregory P. Smith:
In Python 2.7.3 through 2.7.5 the telnetlib select.poll based implementation
assumes that select.error has an errno attribute when handling errors. it does
not. select.error is not an EnvironmentError derived exception.
http://hg.python.org/cpython/file
Gregory P. Smith added the comment:
As this is only on the select.poll code path, a workaround for code that isn't
going to hit select.select file descriptor limits is to set their
telnetlib.Telnet instance _has_poll attribute to False before using it.
my_telnet = telnetlib.T
New submission from Gregory P. Smith:
1) Checkout a 3.4 tree. Mine was at revision d244e1770f1b.
2) ./configure
3) make -j14 profile-opt
4) ./python -m test.regrtest -v test_distutils
FAIL: test_sysconfig_module (distutils.tests.test_sysconfig.SysconfigTestCase
Gregory P. Smith added the comment:
Should _sysconfigdata.py contain these flags for reference, implying that it is
this test or distutils.sysconfig implementation which is really wrong?
--
___
Python tracker
<http://bugs.python.org/issue22
Changes by Gregory P. Smith :
--
title: make profile-opt fails to update _sysconfigdata.py during the rebuild ->
make profile-opt includes -fprofile* flags in _sysconfigdata CFLAGS
___
Python tracker
<http://bugs.python.org/issu
Gregory P. Smith added the comment:
attaching a proposed workaround in the unittest.
--
keywords: +patch
Added file:
http://bugs.python.org/file37231/issue22904-test-workaround-gps01.diff
___
Python tracker
<http://bugs.python.org/issue22
New submission from Gregory P. Smith:
When running a parallel make -j12 test, test_pydoc fails reasonably often with:
test test_pydoc failed -- Traceback (most recent call last):
File "/.../Lib/test/test_pydoc.py", line 556, in test_synopsis_sourceless
synopsis = pydo
Changes by Gregory P. Smith :
--
assignee: gregory.p.smith ->
___
Python tracker
<http://bugs.python.org/issue22910>
___
___
Python-bugs-list mailing list
Un
Gregory P. Smith added the comment:
I suspect flakiness is due to parallel test execution. Is some other test
possibly executing at the same time removing __pycache__ directories or .pyc
files to recreate them (test_compileall?)? If the test were adjusted to point
to a .py file of its own
Gregory P. Smith added the comment:
fyi - tracking the new issue koobs reported in http://bugs.python.org/issue22910
--
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue20
New submission from Gregory P. Smith:
$ cat >test.py <
info = f.next()
File "/usr/lib/python2.7/tarfile.py", line 2319, in next
self.fileobj.seek(self.offset)
File "/usr/lib/python2.7/tarfile.py", line 555, in seek
raise StreamError("s
Gregory P. Smith added the comment:
removing it would also be fine, it's not like it gets used anymore. the fixes
were trivial. done.
--
resolution: -> fixed
stage: -> resolved
status: open -> closed
___
Python tracker
<http:
Gregory P. Smith added the comment:
thanks for the cleanup refactoring!
--
___
Python tracker
<http://bugs.python.org/issue23234>
___
___
Python-bugs-list mailin
Changes by Gregory P. Smith :
--
nosy: -gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue3566>
___
___
Python-bugs-list mailing list
Unsubscribe:
Gregory P. Smith added the comment:
I can apply this to subprocess32 but it is going to look much more like:
+#ifndef MS_WINDOWS /* WTF is anyone compiling on Windows? Shouldn't work! */
+# define HAVE_UNISTD_H 1
+#endif
+#ifdef HAVE_UNISTD_H
#include
+#endif
The real question is w
Changes by Gregory P. Smith :
--
assignee: -> gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue23223>
___
___
Python-bugs-list mailing list
Un
Changes by Gregory P. Smith :
--
assignee: gregory.p.smith ->
___
Python tracker
<http://bugs.python.org/issue1060>
___
___
Python-bugs-list mailing list
Un
Gregory P. Smith added the comment:
This bug prevents zipfile's writestr() from writing large data (longer than
UINT_MAX) to a 64-bit zip file.
The zlib.crc32 function which, as written, cannot accept input with a size
larger than an unsigned int.
https://hg.python.org/cpython
Changes by Gregory P. Smith :
--
assignee: -> gregory.p.smith
nosy: +gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue23342>
___
___
Python-
Gregory P. Smith added the comment:
A 1) Opting not to capture by default is good. Let people explicitly request
that.
A 2) "check" seems like a reasonable parameter name for the "should i raise if
rc != 0" bool. I don't have any other good bikeshed name suggestio
Gregory P. Smith added the comment:
Ethan: check_output combines them when stdout=subprocess.STDOUT is passed (
https://docs.python.org/3.5/library/subprocess.html#subprocess.STDOUT).
Never pass stdout=PIPE or stderr= PIPE to call() or check*() methods as
that will lead to a deadlock when a pipe
Gregory P. Smith added the comment:
Likely a dupe of http://bugs.python.org/issue22904 which is why i set the
buildbot up. :)
I like your patch here better than my non-fixing poor hack in the other issue.
I hadn't noticed CFLAGS_NODIST b
Gregory P. Smith added the comment:
see also issue23390
--
superseder: -> make profile-opt: test_distutils failure
___
Python tracker
<http://bugs.python.org/issu
Changes by Gregory P. Smith :
--
resolution: -> duplicate
status: open -> closed
___
Python tracker
<http://bugs.python.org/issue22904>
___
___
Python-bugs-
Changes by Gregory P. Smith :
--
assignee: -> gregory.p.smith
___
Python tracker
<http://bugs.python.org/issue23390>
___
___
Python-bugs-list mailing list
Un
Changes by Gregory P. Smith :
--
resolution: -> fixed
status: open -> closed
___
Python tracker
<http://bugs.python.org/issue23390>
___
___
Python-bugs-
Gregory P. Smith added the comment:
Also FWIW, I setup a buildbot earlier this year building in profile-opt mode.
--
___
Python tracker
<http://bugs.python.org/issue24
Gregory P. Smith added the comment:
The Makefile's default "make all" target is meant for developers and iterative
development.
make profile-opt is not.
If anything, this issue should document it somewhere if we even have docs on
how to make "release" builds of
New submission from Gregory P. Smith:
for bytes, \v (0x0b) is not considered a line break. for unicode, it is.
this traces back to the Objects/stringlib/ code where unicode defers to the
decision made by Objects/unicodeobject.c's ascii_linebreak table which contains
7 line breaks in
Gregory P. Smith added the comment:
hah, i should've searched the tracker first. looks like the other open issues
cover this.
--
resolution: -> duplicate
status: open -> closed
superseder: -> str.splitlines splitting on non-\r\n characters
versions: +Python 2.7, Pyth
Gregory P. Smith added the comment:
The obvious fix seems to be to not use splitlines but explicitly split on the
allowed characters for ASCII based protocols and formats that only want \r and
\n to be considered.
I don't think we can rightfully change the unicode splitlines beh
Gregory P. Smith added the comment:
If this isn't already mentioned in 2 to 3 porting notes it is worth
highlighting there. code which uses a str in python 2 and still uses a str in
python 3 is now splitting on many more characters.
That seems to be the source of bugs like issue
Gregory P. Smith added the comment:
i'm updating the title to be more accurate.
turning it on by default is likely not desirable as the makefile is primarily
used by developers who are iterating on changes.
but having it use a good workload (regrtest) and work with llvm and os x are
Gregory P. Smith added the comment:
My gut feeling is to say "don't do that" when it comes to passing subprocess
instances between processes (ie: pickling and unpickling them) and expecting
them to work as if nothing had changed... You already acknowledge that this is
a stran
3001 - 3100 of 3274 matches
Mail list logo
