Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
FYI, cygwin-pyio-setmode.patch includes two extra parts for running on Cygwin.
First, fix import error for ctypes (unopened issue). Second, fix building
_ctypes module (#4032 - P
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
This issue seems to duplicate #28459.
#28459 has a patch (implement _setmode() by ctypes on _pyio). I'm not sure
_setmode() will be used in other places in the future, but the patch just works
at the moment.
In addition,
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
Serhiy, would you open a PR? Your patch looks good to me.
I've been looking at the warning on my machine (ubuntu 16.04 x86), it's a bit
annoying to me.
--
nosy: +masamoto
___
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
This issue has been out-of-date since Python 3.4 maintenance became security
status.
--
___
Python tracker <rep...@bugs.python.org>
<https://bugs.pytho
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
Probably, yes. ncurses on Cygwin has provided is_pad() [*]. In addition, the
old version of Cygwin will become to not define WINDOW_HAS_FLAGS by issue25720.
[*] newer version than patchlevel 20090906 was provided on
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
I opened PR 4164 to improve the is_pad configure check and previous PR was
closed.
--
___
Python tracker <rep...@bugs.python.org>
<https://bugs.python
Change by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +4133
___
Python tracker <rep...@bugs.python.org>
<https://bugs.python
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
I opened PR 4153 that is an alternative for PR 4136 (issue2445).
--
___
Python tracker <rep...@bugs.python.org>
<https://bugs.pytho
Change by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +4121
___
Python tracker <rep...@bugs.python.org>
<https://bugs.pytho
Change by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
keywords: +patch
pull_requests: +4067
stage: -> patch review
___
Python tracker <rep...@bugs.python.org>
<https://bugs.pyt
Masayuki Yamamoto <ma3yuki.8mam...@gmail.com> added the comment:
Or Py_tss_NEEDS_INIT is removed from the C API document to become a private
feature; in other words, the Python interpreter will recommend that the static
allocation for the TSS key shouldn't be used in the API client
Masayuki Yamamoto added the comment:
FYI, PEP 539 was accepted (see python-dev threads [1] [2]).
The only thing missing is a reference implementation, next I complete it.
[1] https://mail.python.org/pipermail/python-dev/2017-August/149091.html
[2] https://mail.python.org/pipermail/python-dev
Masayuki Yamamoto added the comment:
Since previous comment, I've studied the switch for show/hide implementation
detail. As the result, I have understood the Py_BUILD_CORE macro hasn't been
generally used for hiding implementation detail (and Py_LIMITED_API does the
part).
Therefore, I
Masayuki Yamamoto added the comment:
oh, I found a mistake.
- replace Py_LIMITED_API with Py_BUILD_CORE on Include/pythread.h Py_tss_t
definition
--
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
Masayuki Yamamoto added the comment:
Nick and Erik, thank you for the comments! I merged proposal into the PR.
Well, I'm interested in the hide implementation detail for TSS API (lately,
I've read the python-ideas thread "PEP: Hide implementation details in the C
API" which Victor
Masayuki Yamamoto added the comment:
Hi, I attempted Nick's proposal and removed unused codes from TLS
implementation (bpo-30279, bpo-30832). This change looks good to me (PR 1362).
As the result, I think ready to be more slim API because the own implementation
for TLS was removed. Therefore
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +2654
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +2651
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +2652
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
New submission from Masayuki Yamamoto:
I found a syntax error when compiling without threads.
In that place, the colon has been used instead of semicolon at end of statement.
build commands and error log:
$ uname -a
Linux masayuki-P35-DS3 4.4.0-83-generic #106-Ubuntu SMP Mon Jun 26 17:54:25 UTC
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +2604
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
New submission from Masayuki Yamamoto:
CPython has provided the own implementation for thread-local storage (TLS) on
Python/thread.c, it's used in the case which a platform has not supplied native
TLS. However, currently all supported platforms (NT and pthreads) have
provided native TLS
Masayuki Yamamoto added the comment:
Ping. I updated PR a bit: macOS is joined to new compile condition and remove
platform-specific condition.
--
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
Masayuki Yamamoto added the comment:
@rigordo Are you using mintty? If I remember rightly, mintty hasn't been set
256 colors after installation (at least in past release, I'm not sure
currently).
--
___
Python tracker <rep...@bugs.python.org>
Masayuki Yamamoto added the comment:
Hi, I finished various things and tackle the issue again, I opened PR 1689 at
last.
Changes from previous patch:
* If ncurses doesn't have both is_pad function and _flags field of WINDOW,
NCURSES_OPAQUE is defined as zero to make WINDOW to non-opaque type
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +1782
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
New submission from Masayuki Yamamoto:
Python/thread_foobar.h is the template code that is threading adaptation for
new platforms, and it hasn't been used on actual platforms. As a matter of
fact, Python/thread_pthread.h and Python/thread_nt.h give concrete examples of
adaptation, therefore
Masayuki Yamamoto added the comment:
Victor,
The tracemalloc module is not available on platforms that provides own TLS
implementation that uses PyMem_* functions. In the own implementation, it
occurs deadlock by recursion call in thread key search. PyOnceVar API handles
extendable array
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
--
pull_requests: +1472
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
Masayuki Yamamoto added the comment:
We should avoid to submit quickfix to official as much as we can. Especially
the case of unofficial (or inarticulate) support platform, a patch for specific
platform is probably rejected because the code won't be maintained.
Thread Specific Storage API
Masayuki Yamamoto added the comment:
The patch can split to several issues, moreover, some issues are able to pick
up as general not Cygwin specific. Perhaps, it will be an opportunity to fix
even if Cygwin specific, if we could explain the implicit issue on
implementation. See #4032, #25658
Masayuki Yamamoto added the comment:
Above said, I updated minor changes to the version 2 patch. Several codes have
kept the words "thread local" and "TLS" because they have pointed programming
method or other meanings, not CPython TLS API itself. (e.g. _decimal module)
-
Masayuki Yamamoto added the comment:
I commented at the Rietveld, since I think that patch needs a bit more
modification.
* rename PyThread_ReInitTLS
* update comments and messages that have explained CPython TLS API
--
___
Python tracker <
Masayuki Yamamoto added the comment:
After Erik posted PEP 539 draft, we've discussed features of new API on the
Python-ideas [*]. As the result of discussion, features of new API have been
changed below points. Thus, I updated patch based on the result.
1. API uses opaque type because
Masayuki Yamamoto added the comment:
#25720 is a similar issue that is originated by opaque type WINDOW.
--
nosy: +masamoto
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
Masayuki Yamamoto added the comment:
Building 2.7 is for testing purposes. It is used to judge whether it is a
problem (specifically for Cygwin-specific) originated in the old version.
_setmode is an important function for setting input and output of CPython, so
in Cygwin I feel annoying
Masayuki Yamamoto added the comment:
I wrote a patch based on msg281227's idea. I confirmed to pass related tests
(test_capi, test_threading and test_tracemalloc) on Cygwin x86 and Ubuntu 16.04
x86.
This patch adds to change CPython about:
1. Avoid compile error on Currently TLS API
Masayuki Yamamoto added the comment:
Elik, Ed,
I have overlooked tracemalloc module raises deadlock if apply the patch.
I found out a source comment on Modules/_tracemalloc.c:161
/* If your OS does not provide native thread local storage, you can implement
it manually using a lock
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
Added file: http://bugs.python.org/file45685/sys-executable-suffix-3.patch
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
Removed file: http://bugs.python.org/file45684/sys-executable-suffix-3.patch
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
Masayuki Yamamoto added the comment:
I updated a patch to check the made path because previous patch has added
suffix even to symbolic link. New patch works that If made path is invalid,
revert to original.
--
Added file: http://bugs.python.org/file45684/sys-executable-suffix-3.patch
New submission from Masayuki Yamamoto:
We're discussing a cause of this issue at #25720.
--
nosy: +masamoto
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
Changes by Masayuki Yamamoto <ma3yuki.8mam...@gmail.com>:
Added file: http://bugs.python.org/file45593/2.7-include-io.h.patch
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python
New submission from Masayuki Yamamoto:
Platform that appeared warning is Vista Cygwin x86. Interpreter execution
doesn't crash because _setmode function is supplied from cygwin1.dll that
always linked.
Warning reason is header io.h [*] doesn't include to source file. Therefore I
wrote two
Masayuki Yamamoto added the comment:
I wrote a patch to avoid compile error for platforms that pthread_key_t is not
integer.
This patch changes to turn off Py_HAVE_NATIVE_TLS if pthread_key_t is not
integer. Hence the platforms use TLS functions implemented by CPython self.
And, I would
Masayuki Yamamoto added the comment:
On, I got complex type in cloudABI. I see my patch doesn't solve it.
https://github.com/NuxiNL/cloudlibc/blob/master/src/include/sys/types.h#L94
https://github.com/NuxiNL/cloudlibc/blob/master/src/include/_/types.h#L209
Masayuki Yamamoto added the comment:
Umm, API seems a design that is passing into function by integer or pointer
because the users don't need type detail. I think the implementation of complex
type is not realistic. Actually, CouldABI and Cygwin are used pointer, and type
detail is hided
Masayuki Yamamoto added the comment:
Hi, I came from #28656.
I have read past discussions, And I've understood the pthread_key_t has
possible of either integer or pointer. So I think there is a simple solution
that replaces key type to intptr_t.
Thus I wrote a patch, but it maybe need
Masayuki Yamamoto added the comment:
I have missed a case of empty progpath. I updated the patch to avoid calling
add_exe_suffix() at that case.
--
Added file: http://bugs.python.org/file45327/sys-executable-suffix-2.patch
___
Python tracker <
Masayuki Yamamoto added the comment:
And updated patch for 2.7
--
Added file: http://bugs.python.org/file45262/2.7-unixccompiler-implib.patch
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/
Masayuki Yamamoto added the comment:
Move version to 3.7 and 2.7, and I updated two patches adding import library
type for file searching.
Current Cygwin is used UnixCCompiler, and the compiler has be able to build
extension module that doesn't need to library link at build time (e.g. array
Masayuki Yamamoto added the comment:
I agree to use setmode() at _pyio module for Cygwin. However, I have two
reasons that I disagree to the implementation of os.setmode().
First, FreeBSD has another setmode() that operate file permission bits [1].
Therefore the implementation of os.setmode
Masayuki Yamamoto added the comment:
Oh, that's enough to work, Martin.
I confirmed too.
--
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
New submission from Masayuki Yamamoto:
_socket module has failed to compile with --without-threads flag since
554fb699af8c, because Py_END_ALLOW_THREADS macro exists behind the done label (
Modules/socketmodule.c:666 ).
If --without-threads flag goes on, Py_END_ALLOW_THREADS macro replaces
Masayuki Yamamoto added the comment:
I agree to add the suffix to solve ambiguous path on Cygwin. And I think the
patch should apply to all platforms having executable suffix.
Therefore, I modified your patch to remove #if directive for Cygwin, and I
changed title to spread covered platforms
Masayuki Yamamoto added the comment:
This patch has impact to end user, thus I don't agree to apply as far as Cygwin
to avoid user surprise.
I think to need a consistency between other platforms having executable suffix
(e.g. MSYS2), and also the patch needs short documentation of changing
Masayuki Yamamoto added the comment:
Thanks, INADA.
I confirmed your solution, and has succeeded to build _futures extension. Your
patch has a little misspell. I fixed it.
--
Added file: http://bugs.python.org/file45046/gc_track-2.patch
___
Python
New submission from Masayuki Yamamoto:
I tried to build cpython on cygwin (vista x86).
Core interpretor has built to success, but '_futures' extension module has
failed (compile error on Modules/_futuresmodule.c:946 ). It has occured since
f8815001a390
part of build log:
$ uname -a
CYGWIN_NT
Masayuki Yamamoto added the comment:
I searched declaration for _PyGC_generation0, And it has be found at
Include/objimpl.h:259
The found declaration hasn't used PyAPI_DATA macro. Hence, I wrote a patch to
wrap declaration by the macro. I has succeeded to build _futures module using
the patch
Masayuki Yamamoto added the comment:
Thank you for confirming, Yen :)
In this case, It seems necessary that resolves headers. I think missing headers
issue maybe solve by #28190. I wrote a join test patch for #28190 and #25720.
Would you be able to resolve headers using this?
--
Added
Masayuki Yamamoto added the comment:
Added comment in review.
Yen, I tried to build without curses.h file (overwrite to ncurses.h). It was
failed on checking header. Hence, I confirmed curses headers on Cygwin and
Ubuntu.
I found out that curses.h includes the directive "#in
Masayuki Yamamoto added the comment:
I updated the patch that add configuration check for is_pad. the is_pad is
wrapped into py_is_pad at Modules/_cursesmodule.c:932 by either of three ways.
Case one -- is_pad is found: Define the macro that is simple wrapping.
Case two -- is_pad is not found
Masayuki Yamamoto added the comment:
I tried to build curses module on Cygwin (Vista x86) using #25720 patch. And it
has been succeeded.
When test_curses ran without skip condition, it was same result as msg278060
(#28190).
I found out build success reasons for cases of applying patch:
#25720
Masayuki Yamamoto added the comment:
test_curses has been skipped almost. the skip reason has been written
"cygwin's curses mostly just hangs" in test case class.
I removed the skip condition, and ran test. Tests was almost passed. The only,
unget_wch was failed by edge case '
Masayuki Yamamoto added the comment:
Now, Cygwin platform is able to build core interpreter on default branch. But
the curses module has been failed to build. Therefore I tried to build curses
module on Cygwin (Vista x86) using this patch. And it has been succeeded.
The patch effect at build
Masayuki Yamamoto added the comment:
I don't agree with #14438 patch to apply branch. this is quickfix for just
Cygwin. I think that curses library should be chosen using configure script.
Erik, How about patch for #14598 or #28190? I prefer #28190 because better
header is chosen using
Masayuki Yamamoto added the comment:
unpack_iterator type has not been registered into _struct module. And all users
get only unpack_iterator object from function returning iterator. The object
iterating doesn't need reference to type. Therefore, I think issue finish by
PyType_Ready fix
Masayuki Yamamoto added the comment:
I wrote a patch to add the unpackiter_type initialization into PyInit__struct
function. I has confirmed solve #28337 on ubuntu x86 16.04.
--
nosy: +masamoto
Added file: http://bugs.python.org/file44929/PyType_Ready-unpackiter_type.patch
Masayuki Yamamoto added the comment:
I made a mistake that is issue number. The correct issue number is #21124.
--
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
New submission from Masayuki Yamamoto:
After #21224 was applied into default branch, the _struct.unpack_iterator has
crashed by segfault.
reproduce segfault:
$ uname -a
Linux masayuki-P35-DS3 4.4.0-38-generic #57-Ubuntu SMP Tue Sep 6 15:41:41 UTC
2016 i686 i686 i686 GNU/Linux
$ hg update -C
Masayuki Yamamoto added the comment:
I don't know why Cygwin packaging script (cygport) doesn't copy direct
libpythonX.Ym.dll.a to /usr/lib. Hence I wrote the patch that doesn't conflict
with cygport script.
In current makefile rule for Cygwin, Install $(LDLIBRARY) into $(LIBPL).
Therefore
Masayuki Yamamoto added the comment:
I found -L option addition at Lib/distutils/command/build_ext.py:222. This
place is a process that finalize options for the build_ext command.
Comment in there: for extensions under Cygwin and AtheOS Python's library
directory must be appended
Masayuki Yamamoto added the comment:
I'm so sorry that I misspelled your name, Erik.
--
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/i
Masayuki Yamamoto added the comment:
Hello, Eric.
I tried to build the '_speedups' C extension module of simplejson (third party
library) using the python built on cygwin. As you pointed out, the build of
'_speedups' C extension module at the stage of the link failed.
build log:
Python3.5.2
Masayuki Yamamoto added the comment:
Oh, I made a mistake that is checking the si_band field without signal header.
Here is a modified patch adding signal header at checking the si_band field.
I passed tests compiling cpython 3.5.2+ on ubuntu 16.04.1 (x86) using this
patch.
--
Added
Masayuki Yamamoto added the comment:
Hello, I'm writer for past patch.
3.4-issue21085-struct_siginfo.patch removes si_band field from struct_siginfo
if the C siginfo_t doesn't have si_band field.
This modification raises a incompatibility between platforms having
struct_siginfo.
Now, I wrote
Masayuki Yamamoto added the comment:
I tried another two case changing variable declaration.
First case of just removing extern, Compiler similarly warns, and test passed.
Second case of removing _PyOS_ReadlineTState declaration, Compiler has not
warned, and test passed.
First:
$ hg diff
diff
Masayuki Yamamoto added the comment:
um, Compiler warned, but test passed. It seems a only warning.
build log:
$ ./configure --prefix=/opt/py34
$ make
...
building 'readline' extension
gcc -Wno-unused-result -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes
-Werror=declaration-after-statement
New submission from Masayuki Yamamoto:
Compiler warns case of define HAVE_DECLSPEC_DLL.
In Modules/readline.c:1065, _PyOS_ReadlineTState variable declaration is
different to Include/pythonrun.h:275.
--
components: Build, Extension Modules
hgrepos: 307
messages: 242013
nosy: masamoto
Masayuki Yamamoto added the comment:
Here is a patch modifying variable declaration to same as Include/pytonrun.h.
--
keywords: +patch
Added file:
http://bugs.python.org/file39205/3.4-issue24058-readline-_PyOS_ReadlineTState.patch
___
Python tracker
Masayuki Yamamoto added the comment:
I wrote the patch to replace $(PYTHON).
Changes:
Add TYPESLOTSGEN to configure.ac. TYPESLOTSGEN values is the built-in python.
If the built-in python is not found, TYPESLOTSGEN echoes message that is
skipping like ASDLGEN.
Generated the configure script
New submission from Masayuki Yamamoto:
Objects/typeslots.inc is broken, and make stop on cygwin-1.7.35-i686.
If built-in python is not found in make processing, Objects/typeslots.inc
generating will fail.
Because Objects/typeslots.inc is generated using $(PYTHON) in Makefile, but
$(PYTHON
Masayuki Yamamoto added the comment:
This issue resolved on default branch in #20597 .
In 3.4 branch latest, PATH_MAX seems unused already in Modules/main.c:12 and
Python/pythonrun.c:35.
I want to cherry-pick #20597 to 3.4 branch.
--
nosy: +masamoto
Masayuki Yamamoto added the comment:
CygwinCCompiler seems legacy code.
In v3.4 latest rivision, Distutils has made to work that is a patching to
UnixCCompiler.
So, I have a think that doesn't need to use CygwinCCompiler to build on Cygwin.
I have upload a patch for UnixCCompiler. The patch
Changes by Masayuki Yamamoto light2happy@gmail.com:
--
nosy: +masamoto
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue14438
___
___
Python-bugs
Changes by Masayuki Yamamoto light2happy@gmail.com:
--
nosy: +masamoto
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue13756
___
___
Python-bugs
Masayuki Yamamoto added the comment:
I rewrote the patch for Python 3.4.
The configure and pyconfig.h.in were generated by autoreconf.
If evaluate to the curses WINDOW doesn't have the _flags member,
The configure script defines related macros and checks the member again.
And removed the quick
Changes by Masayuki Yamamoto light2happy@gmail.com:
--
nosy: +masamoto
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue23338
___
___
Python-bugs
Masayuki Yamamoto added the comment:
New patch uses configure script to set the compile condition for
struct_siginfo.si_band.
The script was generated from configure.ac using autoreconf.
Could you make sure of having the si_band in another platform?
--
Added file: http
Masayuki Yamamoto added the comment:
similar issue #23338: PyErr_Format in ctypes uses invalid parameter
--
nosy: +masamoto
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue23663
Masayuki Yamamoto added the comment:
Victor,
In that case, May I edit configure script to generate the HAVE_* defines?
I'd like to add a statement to check the si_band to configure.ac.
And please generate patches not the git format. Otherwise, Rietveld is unable
to generated the review link
Masayuki Yamamoto added the comment:
I wrote improved patch to remove the 'si_band'. This patch modifies the
'n_in_sequence' to conform to the number of structure members.
And I tested manually for struct_siginfo. struct_siginfo objects builds well.
That's all for now.
$ python3.4.exe
Python
Changes by Masayuki Yamamoto light2happy@gmail.com:
--
nosy: -masamoto
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue21088
___
___
Python
Masayuki Yamamoto added the comment:
I have solved about compiling _struct module too.
I have no authority of this issues, So please close of issues.
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue21124
Changes by Masayuki Yamamoto light2happy@gmail.com:
--
nosy: -masamoto
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue21124
___
___
Python
Masayuki Yamamoto added the comment:
I wrote a patch file.
In other extention module source codes, global variable PyTypeObject has
initialized to using PyVarObject_HEAD_INIT(NULL, 0).
And so, as with other modules, I tried to edit and compiling _struct.c in
Cygwin 1.7.28.
The module compiling
New submission from Masayuki Yamamoto:
There is a test code that is RB characters display on screen. I expected
displaying R to right, and displaying B to bottom. It was run as expected
in Python 2.7.3 and 3.2.3 on Cygwin.
But they were displayed reverse in Python 3.4.0. And when addch
97 matches
Mail list logo
