From tracker at bugs.pypy.org  Sat Sep  1 17:51:54 2012
From: tracker at bugs.pypy.org (David)
Date: Sat, 01 Sep 2012 15:51:54 +0000
Subject: [pypy-issue] [issue1249] Itertools.tee not working correctly with
	pypy?
In-Reply-To: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>
Message-ID: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>


New submission from David <davidhalter88 at gmail.com>:

Hi! I'm the developer of Jedi: https://github.com/davidhalter/jedi, which is an 
auto-completion library. I try to support Python 2.5-3.2. Out of curiosity, I 
tried to use PyPy, which worked incredibly well! But:
Some tests just won't work sometimes with PyPy.

# for i in $(seq 10); do pypy test/run.py goto 92; done

sometimes works, and sometimes doesn't. It seems like, itertools.tee sometimes 
returns old values. PyPy seems working perfectly fine, if I use iter(list()) 
arround all the itertools.tee objects. (I'm speaking about the iterator on 
jedi/evaluate.py:1338).

Is there something that I'm missing?

Tests of Python 2.5-3.2 are passing:
http://travis-ci.org/#!/davidhalter/jedi

Cheers!
David

----------
messages: 4691
nosy: David, pypy-issue
priority: bug
release: 1.8
status: chatting
title: Itertools.tee not working correctly with pypy?

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1249>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 18:06:37 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 01 Sep 2012 16:06:37 +0000
Subject: [pypy-issue] [issue1249] Itertools.tee not working correctly with
	pypy?
In-Reply-To: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>
Message-ID: <1346515597.2.0.420744160685.issue1249@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

In evaluate(), I see that you use weakrefs a lot, and you seem to do very
different things depending on whether weakrefs are still alive or already dead.
 I would need to look more deeply into the code, but it seems to me that it is
enough to cause the behavior under PyPy to not match the behavior under CPython.
 In PyPy, a weakref's target can still be alive even after the object is
supposedly dead, for some unspecified amount of time.

On CPython you get a partially more consistent behavior.  But it is wrong to
rely on it anyway: e.g. if you create circular references between objects A and
B, and there is a pointer from A to C and a weakref from D to C, then on CPython
too the weakref will appear to be alive for an unexpected longer time.

----------
nosy: +arigo

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1249>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 18:48:09 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 01 Sep 2012 16:48:09 +0000
Subject: [pypy-issue] [issue1249] Itertools.tee not working correctly with
	pypy?
In-Reply-To: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>
Message-ID: <1346518089.42.0.538089476423.issue1249@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Ok, found it.  You're right, it's really a bug in itertools.tee().  Test in
7307034ccd39, will be fixed soon.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1249>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 18:50:18 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 01 Sep 2012 16:50:18 +0000
Subject: [pypy-issue] [issue1249] Itertools.tee not working correctly with
	pypy?
In-Reply-To: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>
Message-ID: <1346518218.92.0.531189282444.issue1249@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Fix in b20a8c159d25.

----------
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1249>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 20:38:35 2012
From: tracker at bugs.pypy.org (Erik Allik)
Date: Sat, 01 Sep 2012 18:38:35 +0000
Subject: [pypy-issue] [issue1250] "Bus error: 10" with infinite @property
	rec. inside	nose.twistedtools.deferred
In-Reply-To: <1346524715.28.0.595140415331.issue1250@bugs.pypy.org>
Message-ID: <1346524715.28.0.595140415331.issue1250@bugs.pypy.org>


New submission from Erik Allik <eallik at gmail.com>:

To reproduce:


    from nose.twistedtools import deferred
     
     
    class Foo(object):  # both with new-style and old-style
        @property  # needed to reproduce the bug
        def prop(self):
            return self.prop
     
     
    @deferred(timeout=None)  # needed to reproduce the bug
    def test():
        Foo().prop

...and run with `nosetests whateverfilename.py`




$ pypy --version
Python 2.7.2 (341e1e3821ff, Jun 07 2012, 15:42:54)
[PyPy 1.9.0 with GCC 4.2.1]

----------
messages: 4695
nosy: eallik, pypy-issue
priority: bug
release: 1.9
status: unread
title: "Bus error: 10" with infinite @property rec. inside nose.twistedtools.deferred

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1250>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 20:39:54 2012
From: tracker at bugs.pypy.org (Erik Allik)
Date: Sat, 01 Sep 2012 18:39:54 +0000
Subject: [pypy-issue] [issue1250] "Bus error: 10" with infinite @property
	rec. inside	nose.twistedtools.deferred
In-Reply-To: <1346524715.28.0.595140415331.issue1250@bugs.pypy.org>
Message-ID: <1346524794.22.0.579890830529.issue1250@bugs.pypy.org>


Erik Allik <eallik at gmail.com> added the comment:

*might* be relevant: nose.twistedtools.deferred runs the test in a thread.

----------
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1250>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 21:32:32 2012
From: tracker at bugs.pypy.org (Erik Allik)
Date: Sat, 01 Sep 2012 19:32:32 +0000
Subject: [pypy-issue] [issue1250] "Bus error: 10" with infinite @property
	rec. inside	nose.twistedtools.deferred
In-Reply-To: <1346524715.28.0.595140415331.issue1250@bugs.pypy.org>
Message-ID: <1346527952.84.0.125014279936.issue1250@bugs.pypy.org>


Erik Allik <eallik at gmail.com> added the comment:

verified on pypy-c-jit-57061-5f0a7929c8ea-osx64 -- crashes

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1250>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  1 21:51:35 2012
From: tracker at bugs.pypy.org (Fijal)
Date: Sat, 01 Sep 2012 19:51:35 +0000
Subject: [pypy-issue] [issue1250] "Bus error: 10" with infinite @property
	rec. inside	nose.twistedtools.deferred
In-Reply-To: <1346524715.28.0.595140415331.issue1250@bugs.pypy.org>
Message-ID: <1346529095.59.0.24676565182.issue1250@bugs.pypy.org>


Fijal <fijall at gmail.com> added the comment:

Unreproduced on linux 64. might be something wrong with OS X stack depth 
detection code.

----------
nosy: +fijal

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1250>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  2 01:46:22 2012
From: tracker at bugs.pypy.org (David)
Date: Sat, 01 Sep 2012 23:46:22 +0000
Subject: [pypy-issue] [issue1249] Itertools.tee not working correctly with
	pypy?
In-Reply-To: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>
Message-ID: <1346543182.47.0.777954934249.issue1249@bugs.pypy.org>


David <davidhalter88 at gmail.com> added the comment:

oh, that was fast!
thank you very much.

that's the commit, if anyone is searching:
https://bitbucket.org/pypy/pypy/changeset/b20a8c159d25b682cfcfaa1462008b3f

----------
status: resolved -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1249>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  2 01:59:48 2012
From: tracker at bugs.pypy.org (David)
Date: Sat, 01 Sep 2012 23:59:48 +0000
Subject: [pypy-issue] [issue1249] Itertools.tee not working correctly with
	pypy?
In-Reply-To: <1346514714.49.0.707536841312.issue1249@bugs.pypy.org>
Message-ID: <1346543988.69.0.414023391819.issue1249@bugs.pypy.org>


David <davidhalter88 at gmail.com> added the comment:

sorry, messed up the status. great work anyways!

----------
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1249>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 12:02:32 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 10:02:32 +0000
Subject: [pypy-issue] [issue1233] str() types can't be used with ctypes's
	py_object()
In-Reply-To: <1344763493.64.0.519996205744.issue1233@bugs.pypy.org>
Message-ID: <1346666551.92.0.359711110112.issue1233@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Passing a "payload" argument for a callback is needed if you were writing all C
code, but not if you are writing the callback in Python.  Can't you use the
following scheme instead?

def make_callback(payload_string):
    def callback(args..):
        ... use payload_string ...
    return callback

Then you pass to C code the callback from "make_callback(string)".  The c_void_p
argument is not useful and can always be NULL.

----------
nosy: +arigo

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1233>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 12:07:40 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 10:07:40 +0000
Subject: [pypy-issue] [issue1233] str() types can't be used with ctypes's
	py_object()
In-Reply-To: <1344763493.64.0.519996205744.issue1233@bugs.pypy.org>
Message-ID: <1346666860.08.0.392402201462.issue1233@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

For example, the function add_timer() from timer.py can be rewritten much more
simply as follows (untested code):

def add_timer(interval, callback, param=None):
    if interval < 0:
        raise ValueError("interval must not be smaller than 0")

    def llcallback(x, y, null_ignored):
        return callback(x, y, param)
    timerid = dll.SDL_AddTimer(interval, llcallback, 0)
    if timerid == 0 or not bool(timerid):
        raise SDLError()
    return timerid

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1233>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 18:48:19 2012
From: tracker at bugs.pypy.org (Marcus von Appen)
Date: Mon, 03 Sep 2012 16:48:19 +0000
Subject: [pypy-issue] [issue1233] str() types can't be used with ctypes's
	py_object()
In-Reply-To: <1344763493.64.0.519996205744.issue1233@bugs.pypy.org>
Message-ID: <1346690899.35.0.567522576721.issue1233@bugs.pypy.org>


Marcus von Appen <mva at sysfault.org> added the comment:

I know that using a separate hook would be possible (as well as using an
object class wrapper for the callback). As written earlier, it would be
somewhat more comfortable, though, if one could just pass the payload
directly to the C code hook.

This also would allow 3rd party C code APIs to pick up the data and do
further processing with it without the need to create a separate wrapper,
which needs to be unpacked elsewhere beforehand.

For pure Python interaction, the callback wrapper is an easy solution,
for a more generic approach in PyPy, one currently would need to put
the argument into another capsule to pass it around, which makes code
supposed to be run in a CPython or PyPy environment more complex.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1233>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 19:38:08 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 17:38:08 +0000
Subject: [pypy-issue] [issue1233] str() types can't be used with ctypes's
	py_object()
In-Reply-To: <1344763493.64.0.519996205744.issue1233@bugs.pypy.org>
Message-ID: <1346693888.32.0.127264910786.issue1233@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

I suppose "py_object" can be supported with enough hacks.  You're free to try,
too: start with writing py_object.__init__(x) that stores 'x' in a global list
and returns the index.  It may look obscure, but that's exactly how we would
implement it on pypy.  More precisely, we would add some (pure Python) logic to
lib_pypy/_ctypes/*.py.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1233>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 19:40:33 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 17:40:33 +0000
Subject: [pypy-issue] [issue1248] please build with older GNU libc on
	Linux/amd64
In-Reply-To: <1346476259.72.0.54490436525.issue1248@bugs.pypy.org>
Message-ID: <1346694033.54.0.778370298832.issue1248@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

How?  If you have a precise step-by-step guide of how to build a Linux binary
using some older dependencies than the ones that happen to be installed on the
system, and if this guide is not too long and complicated, we could try. 
Otherwise, sorry, we do our best for Linux binaries but keep hitting these
issues all the time (we just fixed the one with openssl).

----------
nosy: +arigo
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1248>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 19:45:07 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 17:45:07 +0000
Subject: [pypy-issue] [issue1244] bogus test
In-Reply-To: <1345958585.27.0.0708688786784.issue1244@bugs.pypy.org>
Message-ID: <1346694307.43.0.795850284517.issue1244@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Fixed in 5b12e8df1272.

----------
nosy: +arigo
status: unread -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1244>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 19:52:15 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 17:52:15 +0000
Subject: [pypy-issue] [issue1242] lib_pypy/_ctypes/function.py
	accesses	_rawffi.get_last_error in linux
In-Reply-To: <1345899640.51.0.331465536862.issue1242@bugs.pypy.org>
Message-ID: <1346694735.59.0.853633617055.issue1242@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Fixed (2071a44dfde1), differently: by making sure _rawffi.{get,set}_last_error
exists.  On non-Windows it is a pair of no-op functions.

----------
nosy: +arigo
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1242>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 21:06:25 2012
From: tracker at bugs.pypy.org (Marcus von Appen)
Date: Mon, 03 Sep 2012 19:06:25 +0000
Subject: [pypy-issue] [issue1233] str() types can't be used with ctypes's
	py_object()
In-Reply-To: <1344763493.64.0.519996205744.issue1233@bugs.pypy.org>
Message-ID: <1346699185.55.0.226796787326.issue1233@bugs.pypy.org>


Marcus von Appen <mva at sysfault.org> added the comment:

Am I right with the assumption that the global list would emulate the pointer 
value, if we talk about PyObject*? So - without looking at the code of the PyPy 
implementation on py_object - it acts as a sort of proxy, which returns (or would 
return) the index of the real object for accessing?

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1233>
________________________________________

From tracker at bugs.pypy.org  Mon Sep  3 21:10:12 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 03 Sep 2012 19:10:12 +0000
Subject: [pypy-issue] [issue1233] str() types can't be used with ctypes's
	py_object()
In-Reply-To: <1344763493.64.0.519996205744.issue1233@bugs.pypy.org>
Message-ID: <1346699412.93.0.455252633594.issue1233@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Yes, exactly.  "py_object" is not implemented so far in PyPy, but that's how we
would need to implement it.  There is no direct solution because you can't take
the address of a PyPy object and pass it to C code --- to start with we have a
moving garbage collector.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1233>
________________________________________

From tracker at bugs.pypy.org  Tue Sep  4 22:07:45 2012
From: tracker at bugs.pypy.org (Andy Pilate)
Date: Tue, 04 Sep 2012 20:07:45 +0000
Subject: [pypy-issue] [issue1251] distutil but in pypy when installing
	Limnoria
In-Reply-To: <1346789265.58.0.511392643252.issue1251@bugs.pypy.org>
Message-ID: <1346789265.58.0.511392643252.issue1251@bugs.pypy.org>


New submission from Andy Pilate <cubox at cubox.me>:

There is a pypy but when installing Limnoria.
I don't know where pypy fail, but it's failing and it's not the fault of Limnoria.

How to reproduce : 
Pypy on ubuntu precise installed via apt-get
git clone http://github.com/ProgVal/Limnoria
cd Limnoria
sudo pypy setup.py install
cat src/version.py
cat /usr/local/lib/pypy2.7/dist-packages/supybot/version.py

----------
messages: 4711
nosy: cubox, pypy-issue
priority: bug
status: unread
title: distutil but in pypy when installing Limnoria

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1251>
________________________________________

From tracker at bugs.pypy.org  Tue Sep  4 22:23:33 2012
From: tracker at bugs.pypy.org (Andy Pilate)
Date: Tue, 04 Sep 2012 20:23:33 +0000
Subject: [pypy-issue] [issue1252] Limnoria can't run with pypy
In-Reply-To: <1346790213.2.0.651413002973.issue1252@bugs.pypy.org>
Message-ID: <1346790213.2.0.651413002973.issue1252@bugs.pypy.org>


New submission from Andy Pilate <cubox at cubox.me>:

Don't know if it's the fault of pypy or Limnoria, but it is working with cpython.
Log is at http://cubox.me/files/prog.txt
How to reproduce : 
git clone http://github.com/ProgVal/Limnoria
cd Limnoria
sudo pypy setup.py install
cp src/version.py /usr/local/lib/pypy2.7/dist-packages/supybot/version.py #See an 
other bug report by me for this issue.
mkdir /tmp/lol/42 && cd /tmp/lol/42
supybot-wizard # I will don't explain you, just enter random values.
pypy /usr/local/bin/supybot *.conf
Enjoy!

----------
messages: 4712
nosy: cubox, pypy-issue
priority: bug
status: unread
title: Limnoria can't run with pypy

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1252>
________________________________________

From tracker at bugs.pypy.org  Wed Sep  5 02:05:16 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Wed, 05 Sep 2012 00:05:16 +0000
Subject: [pypy-issue] [issue1252] Limnoria can't run with pypy
In-Reply-To: <1346790213.2.0.651413002973.issue1252@bugs.pypy.org>
Message-ID: <1346803516.92.0.778819287086.issue1252@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

See minimal test case below.  It seems that re.sub() returns the input string when no match was found; this optimization 
should not be done when a string subclass is given.

PyPy should be fixed, but this could be considered an implementation detail as well.
The class IrcString should be more robust and ensure that x.lowered is not a IrcString! otherwise the __hash__ method 
enters a nasty recursion.
   x.lowered = str(toLower(x))


import re
class MyString(str):
    pass
s1 = MyString('zz')
s2 = re.compile('aa').sub('bb', s1)
print s1 is s2
# True in Pypy which returns a str subclass;
# False for CPython which creates an exact str instance.

----------
nosy: +afa
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1252>
________________________________________

From tracker at bugs.pypy.org  Wed Sep  5 02:11:02 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Wed, 05 Sep 2012 00:11:02 +0000
Subject: [pypy-issue] [issue1251] distutil bug in pypy when installing
	Limnoria
In-Reply-To: <1346789265.58.0.511392643252.issue1251@bugs.pypy.org>
Message-ID: <1346803862.26.0.212733107184.issue1251@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

setup.py has this code::
   open(...version.py', 'w').write(...)
See in http://pypy.readthedocs.org/en/latest/cpython_differences.html#differences-related-to-
garbage-collection-strategies
the code should really close the file, or use a "with" statement to ensure that data is flushed:

with open(...version.py', 'w') as f:
   f.write(...)

----------
nosy: +afa
status: unread -> wontfix

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1251>
________________________________________

From tracker at bugs.pypy.org  Wed Sep  5 09:15:13 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Wed, 05 Sep 2012 07:15:13 +0000
Subject: [pypy-issue] [issue1252] Limnoria can't run with pypy
In-Reply-To: <1346790213.2.0.651413002973.issue1252@bugs.pypy.org>
Message-ID: <1346829313.27.0.193218349674.issue1252@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Fixed in 89ae68436535.

----------
nosy: +arigo
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1252>
________________________________________

From tracker at bugs.pypy.org  Wed Sep  5 10:44:04 2012
From: tracker at bugs.pypy.org (Andy Pilate)
Date: Wed, 05 Sep 2012 08:44:04 +0000
Subject: [pypy-issue] [issue1251] distutil bug in pypy when installing
	Limnoria
In-Reply-To: <1346789265.58.0.511392643252.issue1251@bugs.pypy.org>
Message-ID: <1346834644.8.0.361387720228.issue1251@bugs.pypy.org>


Andy Pilate <cubox at cubox.me> added the comment:

Oh, okey. Reading the code is not my job ;)
I will submit a merge request with the fix in Limnoria.
Thanks for your time :D

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1251>
________________________________________

From tracker at bugs.pypy.org  Fri Sep  7 11:44:03 2012
From: tracker at bugs.pypy.org (Pigmej)
Date: Fri, 07 Sep 2012 09:44:03 +0000
Subject: [pypy-issue] [issue1253] compile => RuntimeError: maximum recursion
	depth exceeded
In-Reply-To: <1347011043.22.0.852237039186.issue1253@bugs.pypy.org>
Message-ID: <1347011043.22.0.852237039186.issue1253@bugs.pypy.org>


New submission from Pigmej <jedrzej.nowak at codernity.com>:

Run attached code on CPython and on PyPy. It will work fine on CPython, but will crash on 
PyPy, because o "RuntimeError: maximum recursion depth exceeded".

At current trunk it crashes on 6400, so in case of change please verify the while loop.


// I know that code makes no sense at all.

----------
files: pypy_rec.py
messages: 4717
nosy: pigmej, pypy-issue
priority: bug
status: unread
title: compile => RuntimeError: maximum recursion depth exceeded

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1253>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  8 11:24:35 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 08 Sep 2012 09:24:35 +0000
Subject: [pypy-issue] [issue1244] bogus test
In-Reply-To: <1345958585.27.0.0708688786784.issue1244@bugs.pypy.org>
Message-ID: <1347096275.62.0.580297853048.issue1244@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Checked in a65ca3de676a.  Does it help?  (Please, can you tell us *in all cases*
at least which test files you're talking about?  There are two completely
different places with a similar few lines.)

----------
status: chatting -> testing

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1244>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  8 11:26:16 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 08 Sep 2012 09:26:16 +0000
Subject: [pypy-issue] [issue1254] meta-bug for post-stdlib-2.7.3 win32
	breakage
In-Reply-To: <1347093369.93.0.363725480893.issue1254@bugs.pypy.org>
Message-ID: <1347096376.89.0.0454012268586.issue1254@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Define it ourselves to what value?

----------
nosy: +arigo
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1254>
________________________________________

From tracker at bugs.pypy.org  Sat Sep  8 22:32:54 2012
From: tracker at bugs.pypy.org (mattip)
Date: Sat, 08 Sep 2012 20:32:54 +0000
Subject: [pypy-issue] [issue1254] meta-bug for post-stdlib-2.7.3 win32
	breakage
In-Reply-To: <1347093369.93.0.363725480893.issue1254@bugs.pypy.org>
Message-ID: <1347136374.86.0.344812115414.issue1254@bugs.pypy.org>


mattip <matti.picus at gmail.com> added the comment:

vs2008 defines _MAX_ENV in stdlib.h according to msdn
Fixed with changeset 2050ccf55674 which no longer crashes test
rlib/test/test_rwin32.py
here is the msdn link - 
http://msdn.microsoft.com/en-us/library/e8abs9d2(v=vs.80).aspx

----------
nosy: +mattip
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1254>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 06:19:09 2012
From: tracker at bugs.pypy.org (mattip)
Date: Sun, 09 Sep 2012 04:19:09 +0000
Subject: [pypy-issue] [issue1254] meta-bug for post-stdlib-2.7.3 win32
	breakage
In-Reply-To: <1347093369.93.0.363725480893.issue1254@bugs.pypy.org>
Message-ID: <1347164348.71.0.409710094176.issue1254@bugs.pypy.org>


mattip <matti.picus at gmail.com> added the comment:

new problem: self.random_context = None breaks translation on Windows
module/posix/interp_posix.py contains a State class that has a _freeze_ method 
that sets self.random_context to None. This breaks translation on the current 
trunk. It does not show up on module/posix/test/test_posix2 (which still has 4 
failing tests, but none seem to be related). How can I write a test for _freeze_?

----------
nosy: +afa
status: resolved -> in-progress

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1254>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 19:22:51 2012
From: tracker at bugs.pypy.org (dremon)
Date: Sun, 09 Sep 2012 17:22:51 +0000
Subject: [pypy-issue] [issue1255] signal.pause blocks socket event processing
In-Reply-To: <1347211371.82.0.592320296484.issue1255@bugs.pypy.org>
Message-ID: <1347211371.82.0.592320296484.issue1255@bugs.pypy.org>


New submission from dremon <dmitry at pankratov.net>:

When select.select is used for server socket in the separate thread and the main
thread pauses with signal.pause call no socket events are processed.
Attached is a sample script (works under CPython but breaks in PyPy).

----------
files: tcpserv.py
messages: 4725
nosy: dremon, pypy-issue
priority: bug
status: unread
title: signal.pause blocks socket event processing

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1255>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 20:19:22 2012
From: tracker at bugs.pypy.org (Julian Berman)
Date: Sun, 09 Sep 2012 18:19:22 +0000
Subject: [pypy-issue] [issue1256] Writing bytes to a StringIO expecting
	unicode exhibits a	confusing error message
In-Reply-To: <1347214762.49.0.627922484954.issue1256@bugs.pypy.org>
Message-ID: <1347214762.49.0.627922484954.issue1256@bugs.pypy.org>


New submission from Julian Berman <Julian+PyPy at GrayVines.com>:

This was fixed in CPython here, http://bugs.python.org/issue12434 , apparently 
without a test.

----------
files: errmsg.patch
messages: 4726
nosy: Julian, pypy-issue
priority: feature
status: unread
title: Writing bytes to a StringIO expecting unicode exhibits a confusing error message

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1256>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 20:35:24 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sun, 09 Sep 2012 18:35:24 +0000
Subject: [pypy-issue] [issue1256] Writing bytes to a StringIO expecting
	unicode exhibits a	confusing error message
In-Reply-To: <1347214762.49.0.627922484954.issue1256@bugs.pypy.org>
Message-ID: <1347215724.23.0.426306793526.issue1256@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Thanks!  Fixed (7d5b0a808fc9).  Indeed, the message was clearly bogus IMHO too.

----------
nosy: +arigo
status: unread -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1256>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 20:43:48 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sun, 09 Sep 2012 18:43:48 +0000
Subject: [pypy-issue] [issue1255] signal.pause blocks socket event processing
In-Reply-To: <1347211371.82.0.592320296484.issue1255@bugs.pypy.org>
Message-ID: <1347216228.9.0.380756157398.issue1255@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

thread.signal() does not release the GIL in PyPy!  That's bogus.  Testing the
fix now...

----------
nosy: +arigo
status: unread -> in-progress

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1255>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 20:49:09 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sun, 09 Sep 2012 18:49:09 +0000
Subject: [pypy-issue] [issue1255] signal.pause blocks socket event processing
In-Reply-To: <1347211371.82.0.592320296484.issue1255@bugs.pypy.org>
Message-ID: <1347216549.1.0.127445059998.issue1255@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

er, I mean signal.pause(), sorry.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1255>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 21:31:04 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sun, 09 Sep 2012 19:31:04 +0000
Subject: [pypy-issue] [issue1255] signal.pause blocks socket event processing
In-Reply-To: <1347211371.82.0.592320296484.issue1255@bugs.pypy.org>
Message-ID: <1347219064.96.0.754483337163.issue1255@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Fixed (5f6bd03d6b4c), thanks for the report!

----------
status: in-progress -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1255>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 21:59:22 2012
From: tracker at bugs.pypy.org (dremon)
Date: Sun, 09 Sep 2012 19:59:22 +0000
Subject: [pypy-issue] [issue1257] Memory leak when using web.py
In-Reply-To: <1347220762.49.0.575714347026.issue1257@bugs.pypy.org>
Message-ID: <1347220762.49.0.575714347026.issue1257@bugs.pypy.org>


New submission from dremon <dmitry at pankratov.net>:

web.py, used as a standalone HTTP server. I noticed a steady resident memory
growth of the pypy process, probably objects are not being gc'ed. With pyopenssl
adapter it grows faster.
Attached is a test script (requires web.py and pyopenssl).
Commands used for client connection:
while `/bin/true`; do curl http://localhost:8700; done
while `/bin/true`; do curl -k https://localhost:8700; done

----------
files: test_httpservice.py
messages: 4731
nosy: dremon, pypy-issue
priority: bug
status: unread
title: Memory leak when using web.py

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1257>
________________________________________

From tracker at bugs.pypy.org  Sun Sep  9 22:57:22 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sun, 09 Sep 2012 20:57:22 +0000
Subject: [pypy-issue] [issue1254] meta-bug for post-stdlib-2.7.3 win32
	breakage
In-Reply-To: <1347093369.93.0.363725480893.issue1254@bugs.pypy.org>
Message-ID: <1347224242.34.0.645362505877.issue1254@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

You can't write a simple test for _freeze_, precisely because it's only used for
translation.  Attempted fix in 1feab101071c (not tested).

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1254>
________________________________________

From tracker at bugs.pypy.org  Mon Sep 10 05:14:10 2012
From: tracker at bugs.pypy.org (mattip)
Date: Mon, 10 Sep 2012 03:14:10 +0000
Subject: [pypy-issue] [issue1254] meta-bug for post-stdlib-2.7.3 win32
	breakage
In-Reply-To: <1347093369.93.0.363725480893.issue1254@bugs.pypy.org>
Message-ID: <1347246850.92.0.406711939769.issue1254@bugs.pypy.org>


mattip <matti.picus at gmail.com> added the comment:

translation succeeds.
Reinstalling VS2008 with sp1
http://www.microsoft.com/visualstudio/en-us/products/2008-editions/express
yields _MAX_ENV in stdlib.h, see uploaded file

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1254>
________________________________________
-------------- next part --------------
/***
*stdlib.h - declarations/definitions for commonly used library functions
*
*       Copyright (c) Microsoft Corporation. All rights reserved.
*
*Purpose:
*       This include file contains the function declarations for commonly
*       used library functions which either don't fit somewhere else, or,
*       cannot be declared in the normal place for other reasons.
*       [ANSI]
*
*       [Public]
*
****/

#if     _MSC_VER > 1000
#pragma once
#endif

#ifndef _INC_STDLIB
#define _INC_STDLIB

#include <crtdefs.h>
#include <limits.h>

#ifdef  _MSC_VER
/*
 * Currently, all MS C compilers for Win32 platforms default to 8 byte
 * alignment.
 */
#pragma pack(push,_CRT_PACKING)
#endif  /* _MSC_VER */

#ifdef  __cplusplus
extern "C" {
#endif

/* Define NULL pointer value */
#ifndef NULL
#ifdef __cplusplus
#define NULL    0
#else
#define NULL    ((void *)0)
#endif
#endif

/* Definition of the argument values for the exit() function */

#define EXIT_SUCCESS    0
#define EXIT_FAILURE    1


#ifndef _ONEXIT_T_DEFINED

#if !defined (_M_CEE_PURE)
typedef int (__cdecl * _onexit_t)(void);
#else
typedef int (__clrcall * _onexit_t)(void);
typedef _onexit_t _onexit_m_t;
#endif

#if defined (_M_CEE_MIXED)
typedef int (__clrcall * _onexit_m_t)(void);
#endif

#if     !__STDC__
/* Non-ANSI name for compatibility */
#define onexit_t _onexit_t
#endif

#define _ONEXIT_T_DEFINED
#endif


/* Data structure definitions for div and ldiv runtimes. */

#ifndef _DIV_T_DEFINED

typedef struct _div_t {
        int quot;
        int rem;
} div_t;

typedef struct _ldiv_t {
        long quot;
        long rem;
} ldiv_t;

#define _DIV_T_DEFINED
#endif

/*
 * structs used to fool the compiler into not generating floating point
 * instructions when copying and pushing [long] double values
 */

#ifndef _CRT_DOUBLE_DEC

#ifndef _LDSUPPORT

#pragma pack(4)
typedef struct {
    unsigned char ld[10];
} _LDOUBLE;
#pragma pack()

#define _PTR_LD(x) ((unsigned char  *)(&(x)->ld))

#else

/* push and pop long, which is #defined as __int64 by a spec2k test */
#pragma push_macro("long")
#undef long
typedef long double _LDOUBLE;
#pragma pop_macro("long")

#define _PTR_LD(x) ((unsigned char  *)(x))

#endif

typedef struct {
        double x;
} _CRT_DOUBLE;

typedef struct {
    float f;
} _CRT_FLOAT;

/* push and pop long, which is #defined as __int64 by a spec2k test */
#pragma push_macro("long")
#undef long

typedef struct {
        /*
         * Assume there is a long double type
         */
        long double x;
} _LONGDOUBLE;

#pragma pop_macro("long")

#pragma pack(4)
typedef struct {
    unsigned char ld12[12];
} _LDBL12;
#pragma pack()

#define _CRT_DOUBLE_DEC
#endif

/* Maximum value that can be returned by the rand function. */

#define RAND_MAX 0x7fff

/*
 * Maximum number of bytes in multi-byte character in the current locale
 * (also defined in ctype.h).
 */
#ifndef MB_CUR_MAX
#ifdef _MT
#define MB_CUR_MAX ___mb_cur_max_func()
#else
#define MB_CUR_MAX __mb_cur_max
#endif
#if !defined(_M_CEE_PURE)
_CRTIMP extern int __mb_cur_max;
#else
_CRTIMP int* __cdecl __p___mb_cur_max();
#define __mb_cur_max (*__p___mb_cur_max())
#endif /* !defined(_M_CEE_PURE) */
_CRTIMP int __cdecl ___mb_cur_max_func(void);
_CRTIMP int __cdecl ___mb_cur_max_l_func(_locale_t);
#endif  /* MB_CUR_MAX */

/* Minimum and maximum macros */

#define __max(a,b)  (((a) > (b)) ? (a) : (b))
#define __min(a,b)  (((a) < (b)) ? (a) : (b))

/*
 * Sizes for buffers used by the _makepath() and _splitpath() functions.
 * note that the sizes include space for 0-terminator
 */
#define _MAX_PATH   260 /* max. length of full pathname */
#define _MAX_DRIVE  3   /* max. length of drive component */
#define _MAX_DIR    256 /* max. length of path component */
#define _MAX_FNAME  256 /* max. length of file name component */
#define _MAX_EXT    256 /* max. length of extension component */

/*
 * Argument values for _set_error_mode().
 */
#define _OUT_TO_DEFAULT 0
#define _OUT_TO_STDERR  1
#define _OUT_TO_MSGBOX  2
#define _REPORT_ERRMODE 3

/*
 * Argument values for _set_abort_behavior().
 */
#define _WRITE_ABORT_MSG    0x1
#define _CALL_REPORTFAULT   0x2

/*
 * Sizes for buffers used by the getenv/putenv family of functions.
 */
#define _MAX_ENV 32767   

#if !defined(_M_CEE_PURE)
/* a purecall handler procedure. Never returns normally */
typedef void (__cdecl *_purecall_handler)(void); 

/* establishes a purecall handler for the process */
_CRTIMP _purecall_handler __cdecl _set_purecall_handler(_In_opt_ _purecall_handler _Handler);
_CRTIMP _purecall_handler __cdecl _get_purecall_handler();
#endif

#if defined(__cplusplus)
extern "C++"
{
#if defined(_M_CEE_PURE)
    typedef void (__clrcall *_purecall_handler)(void);
    typedef _purecall_handler _purecall_handler_m;
    _MRTIMP _purecall_handler __cdecl _set_purecall_handler(_In_opt_ _purecall_handler _Handler);
#endif

#if defined(_M_CEE_MIXED)
    typedef void (__clrcall *_purecall_handler_m)(void);

    _MRTIMP _purecall_handler_m __cdecl _set_purecall_handler(_In_opt_ _purecall_handler_m _Handler);

    _MRTIMP _purecall_handler __cdecl _set_purecall_handler(_In_ int _Handler); /* To disambiguate NULL/0 */
#endif
}
#endif

#if !defined(_M_CEE_PURE)
/* a invalid_arg handler procedure. */
typedef void (__cdecl *_invalid_parameter_handler)(const wchar_t *, const wchar_t *, const wchar_t *, unsigned int, uintptr_t); 

/* establishes a invalid_arg handler for the process */
_CRTIMP _invalid_parameter_handler __cdecl _set_invalid_parameter_handler(_In_opt_ _invalid_parameter_handler _Handler);
_CRTIMP _invalid_parameter_handler __cdecl _get_invalid_parameter_handler(void);
#endif

#if defined(__cplusplus)
extern "C++"
{
#if defined(_M_CEE_PURE)
    typedef void (__clrcall *_invalid_parameter_handler)(const wchar_t *, const wchar_t *, const wchar_t *, unsigned int, uintptr_t);
    typedef _invalid_parameter_handler _invalid_parameter_handler_m;
    _MRTIMP _invalid_parameter_handler __cdecl _set_invalid_parameter_handler(_In_opt_ _invalid_parameter_handler _Handlerh);
#endif

#if defined(_M_CEE_MIXED)
    typedef void (__clrcall *_invalid_parameter_handler_m)(const wchar_t *, const wchar_t *, const wchar_t *, unsigned int, uintptr_t);
    _MRTIMP _invalid_parameter_handler_m __cdecl _set_invalid_parameter_handler(_In_opt_ _invalid_parameter_handler_m _Handlerh);

    _MRTIMP _invalid_parameter_handler __cdecl _set_invalid_parameter_handler(_In_ int _Handler); /* To disambiguate NULL/0, only when we have both */
#endif
}
#endif

/* External variable declarations */
#ifndef _CRT_ERRNO_DEFINED
#define _CRT_ERRNO_DEFINED
_CRTIMP extern int * __cdecl _errno(void);
#define errno   (*_errno())

errno_t __cdecl _set_errno(_In_ int _Value);
errno_t __cdecl _get_errno(_Out_ int * _Value);
#endif

_CRTIMP unsigned long * __cdecl __doserrno(void);
#define _doserrno   (*__doserrno())

errno_t __cdecl _set_doserrno(_In_ unsigned long _Value);
errno_t __cdecl _get_doserrno(_Out_ unsigned long * _Value);

/* you can't modify this, but it is non-const for backcompat */
_CRTIMP _CRT_INSECURE_DEPRECATE(strerror) char ** __cdecl __sys_errlist(void);
#define _sys_errlist (__sys_errlist())

_CRTIMP _CRT_INSECURE_DEPRECATE(strerror) int * __cdecl __sys_nerr(void);
#define _sys_nerr (*__sys_nerr())

#if     defined(_DLL) && defined(_M_IX86)

_CRTIMP int *          __cdecl __p___argc(void);
_CRTIMP char ***       __cdecl __p___argv(void);
_CRTIMP wchar_t ***    __cdecl __p___wargv(void);
_CRTIMP char ***       __cdecl __p__environ(void);
_CRTIMP wchar_t ***    __cdecl __p__wenviron(void);
_CRTIMP char **        __cdecl __p__pgmptr(void);
_CRTIMP wchar_t **     __cdecl __p__wpgmptr(void);


#endif  /* _M_IX86 && _DLL */

#if !defined(_M_CEE_PURE)
_CRTIMP extern int __argc;          /* count of cmd line args */
_CRTIMP extern char ** __argv;      /* pointer to table of cmd line args */
_CRTIMP extern wchar_t ** __wargv;  /* pointer to table of wide cmd line args */
#else
_CRTIMP int* __cdecl __p___argc();
_CRTIMP char*** __cdecl __p___argv();
_CRTIMP wchar_t*** __cdecl __p___wargv();
#define __argv (*__p___argv())
#define __argc (*__p___argc())
#define __wargv (*__p___wargv())
#endif

#if !defined(_M_CEE_PURE)
#ifdef  _POSIX_
extern char ** environ;             /* pointer to environment table */
#else
_CRTIMP extern char ** _environ;    /* pointer to environment table */
_CRTIMP extern wchar_t ** _wenviron;    /* pointer to wide environment table */
#endif  /* _POSIX_ */

_CRT_INSECURE_DEPRECATE_GLOBALS(_get_pgmptr) _CRTIMP extern char * _pgmptr;      /* points to the module (EXE) name */
_CRT_INSECURE_DEPRECATE_GLOBALS(_get_wpgmptr) _CRTIMP extern wchar_t * _wpgmptr;  /* points to the module (EXE) wide name */


#else

_CRTIMP char*** __cdecl __p__environ();
_CRTIMP wchar_t*** __cdecl __p__wenviron();
_CRT_INSECURE_DEPRECATE_GLOBALS(_get_pgmptr) _CRTIMP char** __cdecl __p__pgmptr();
_CRT_INSECURE_DEPRECATE_GLOBALS(_get_wpgmptr) _CRTIMP wchar_t** __cdecl __p__wpgmptr();

#define _environ   (*__p__environ())
#define _wenviron  (*__p__wenviron())
#define _pgmptr    (*__p__pgmptr())
#define _wpgmptr   (*__p__wpgmptr())

#endif /* !defined(_M_CEE_PURE) */

errno_t __cdecl _get_pgmptr(_Deref_out_z_ char ** _Value);
errno_t __cdecl _get_wpgmptr(_Deref_out_z_ wchar_t ** _Value);


#if !defined(_M_CEE_PURE)
_CRT_INSECURE_DEPRECATE_GLOBALS(_get_fmode) _CRTIMP extern int _fmode;          /* default file translation mode */
#else
_CRTIMP int* __cdecl __p__fmode();
#define _fmode (*__p__fmode())
#endif /* !defined(_M_CEE_PURE) */

_CRTIMP errno_t __cdecl _set_fmode(_In_ int _Mode);
_CRTIMP errno_t __cdecl _get_fmode(_Out_ int * _PMode);

/* _countof helper */
#if !defined(_countof)
#if !defined(__cplusplus)
#define _countof(_Array) (sizeof(_Array) / sizeof(_Array[0]))
#else
extern "C++"
{
template <typename _CountofType, size_t _SizeOfArray>
char (*__countof_helper(UNALIGNED _CountofType (&_Array)[_SizeOfArray]))[_SizeOfArray];
#define _countof(_Array) sizeof(*__countof_helper(_Array))
}
#endif
#endif

/* function prototypes */

#ifndef _CRT_TERMINATE_DEFINED
#define _CRT_TERMINATE_DEFINED
_CRTIMP __declspec(noreturn) void __cdecl exit(_In_ int _Code);
_CRTIMP __declspec(noreturn) void __cdecl _exit(_In_ int _Code);
_CRTIMP void __cdecl abort(void);
#endif

_CRTIMP unsigned int __cdecl _set_abort_behavior(_In_ unsigned int _Flags, _In_ unsigned int _Mask);

#ifndef _CRT_ABS_DEFINED
#define _CRT_ABS_DEFINED
        int     __cdecl abs(_In_ int _X);
        long    __cdecl labs(_In_ long _X);
#endif

        __int64    __cdecl _abs64(__int64);
#if _MSC_VER >= 1400 && defined(_M_CEE)
#pragma warning (suppress: 4985)
        _Check_return_ int    __clrcall _atexit_m_appdomain(_In_opt_ void (__clrcall * _Func)(void));
#if defined(_M_CEE_MIXED)
#pragma warning (suppress: 4985)
        _Check_return_ int    __clrcall _atexit_m(_In_opt_ void (__clrcall * _Func)(void));
#else
        _Check_return_ inline int __clrcall _atexit_m(_In_opt_ void (__clrcall *_Function)(void))
#pragma warning (suppress: 4985)
        {
            return _atexit_m_appdomain(_Function);
        }
#endif
#endif
#if defined(_M_CEE_PURE)
        /* In pure mode, atexit is the same as atexit_m_appdomain */
extern "C++"
{
inline  int    __clrcall atexit
(
    void (__clrcall *_Function)(void)
)
{
    return _atexit_m_appdomain(_Function);
}
}
#else
        int    __cdecl atexit(void (__cdecl *)(void));
#endif
#ifndef _CRT_ATOF_DEFINED
#define _CRT_ATOF_DEFINED
_Check_return_ _CRTIMP double  __cdecl atof(_In_z_ const char *_String);
_Check_return_ _CRTIMP double  __cdecl _atof_l(_In_z_ const char *_String, _In_opt_ _locale_t _Locale);
#endif
_Check_return_ _CRTIMP _CRT_JIT_INTRINSIC int    __cdecl atoi(_In_z_ const char *_Str);
_Check_return_ _CRTIMP int    __cdecl _atoi_l(_In_z_ const char *_Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP long   __cdecl atol(_In_z_ const char *_Str);
_Check_return_ _CRTIMP long   __cdecl _atol_l(_In_z_ const char *_Str, _In_opt_ _locale_t _Locale);
#ifndef _CRT_ALGO_DEFINED
#define _CRT_ALGO_DEFINED
#if __STDC_WANT_SECURE_LIB__
_Check_return_ _CRTIMP void * __cdecl bsearch_s(_In_ const void * _Key, _In_bytecount_x_(_NumOfElements * _SizeOfElements) const void * _Base, 
        _In_ rsize_t _NumOfElements, _In_ rsize_t _SizeOfElements,
        _In_ int (__cdecl * _PtFuncCompare)(void *, const void *, const void *), void * _Context);
#endif
_Check_return_ _CRTIMP void * __cdecl bsearch(_In_ const void * _Key, _In_bytecount_x_(_NumOfElements * _SizeOfElements) const void * _Base, 
        _In_ size_t _NumOfElements, _In_ size_t _SizeOfElements,
        _In_ int (__cdecl * _PtFuncCompare)(const void *, const void *));

#if __STDC_WANT_SECURE_LIB__
_CRTIMP void __cdecl qsort_s(_Inout_bytecap_x_(_NumOfElements* _SizeOfElements) void * _Base, 
        _In_ rsize_t _NumOfElements, _In_ rsize_t _SizeOfElements,
        _In_ int (__cdecl * _PtFuncCompare)(void *, const void *, const void *), void *_Context);
#endif
_CRTIMP void __cdecl qsort(_Inout_bytecap_x_(_NumOfElements * _SizeOfElements) void * _Base, 
	_In_ size_t _NumOfElements, _In_ size_t _SizeOfElements, 
        _In_ int (__cdecl * _PtFuncCompare)(const void *, const void *));
#endif
        _Check_return_ unsigned short __cdecl _byteswap_ushort(_In_ unsigned short _Short);
        _Check_return_ unsigned long  __cdecl _byteswap_ulong (_In_ unsigned long _Long);
        _Check_return_ unsigned __int64 __cdecl _byteswap_uint64(_In_ unsigned __int64 _Int64);
_Check_return_ _CRTIMP div_t  __cdecl div(_In_ int _Numerator, _In_ int _Denominator);
_Check_return_ _CRTIMP _CRT_INSECURE_DEPRECATE(_dupenv_s) char * __cdecl getenv(_In_z_ const char * _VarName);
#if __STDC_WANT_SECURE_LIB__
_Check_return_opt_ _CRTIMP errno_t __cdecl getenv_s(_Out_ size_t * _ReturnSize, _Out_z_cap_(_DstSize) char * _DstBuf, _In_ rsize_t _DstSize, _In_z_ const char * _VarName);
#endif
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, getenv_s, _Out_ size_t *, _ReturnSize, char, _Dest, _In_z_ const char *, _VarName)
#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)
#pragma push_macro("_dupenv_s")
#undef _dupenv_s
#endif

_Check_return_opt_ _CRTIMP errno_t __cdecl _dupenv_s(_Out_ _Deref_post_opt_z_cap_(*_PBufferSizeInBytes) char **_PBuffer, _Out_opt_ size_t * _PBufferSizeInBytes, _In_z_ const char * _VarName);

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)
#pragma pop_macro("_dupenv_s")
#endif

_Check_return_opt_ _CRTIMP errno_t __cdecl _itoa_s(_In_ int _Value, _Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ int _Radix);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _itoa_s, _In_ int, _Value, char, _Dest, _In_ int, _Radix)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_1_1(char *, __RETURN_POLICY_DST, _CRTIMP, _itoa, _In_ int, _Value, _Pre_notnull_ _Post_z_, char, _Dest, _In_ int, _Radix)
_Check_return_opt_ _CRTIMP errno_t __cdecl _i64toa_s(_In_ __int64 _Val, _Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ int _Radix);
_CRTIMP _CRT_INSECURE_DEPRECATE(_i64toa_s) char * __cdecl _i64toa(_In_ __int64 _Val, _Pre_notnull_ _Post_z_ char * _DstBuf, _In_ int _Radix);
_Check_return_opt_ _CRTIMP errno_t __cdecl _ui64toa_s(_In_ unsigned __int64 _Val, _Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ int _Radix);
_CRTIMP _CRT_INSECURE_DEPRECATE(_ui64toa_s) char * __cdecl _ui64toa(_In_ unsigned __int64 _Val, _Pre_notnull_ _Post_z_ char * _DstBuf, _In_ int _Radix);
_Check_return_ _CRTIMP __int64 __cdecl _atoi64(_In_z_ const char * _String);
_Check_return_ _CRTIMP __int64 __cdecl _atoi64_l(_In_z_ const char * _String, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP __int64 __cdecl _strtoi64(_In_z_ const char * _String, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_ int _Radix);
_Check_return_ _CRTIMP __int64 __cdecl _strtoi64_l(_In_z_ const char * _String, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_ int _Radix, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP unsigned __int64 __cdecl _strtoui64(_In_z_ const char * _String, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_ int _Radix);
_Check_return_ _CRTIMP unsigned __int64 __cdecl _strtoui64_l(_In_z_ const char * _String, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_ int  _Radix, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP ldiv_t __cdecl ldiv(_In_ long _Numerator, _In_ long _Denominator);
#ifdef __cplusplus
extern "C++"
{
    inline ldiv_t  div(_In_ long _A1, _In_ long _A2)
    {
        return ldiv(_A1, _A2);
    }
}
#endif
_Check_return_opt_ _CRTIMP errno_t __cdecl _ltoa_s(_In_ long _Val, _Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ int _Radix);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _ltoa_s, _In_ long, _Value, char, _Dest, _In_ int, _Radix)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_1_1(char *, __RETURN_POLICY_DST, _CRTIMP, _ltoa, _In_ long, _Value, _Pre_notnull_ _Post_z_, char, _Dest, _In_ int, _Radix)
_Check_return_ _CRTIMP int    __cdecl mblen(_In_opt_bytecount_(_MaxCount) _Pre_opt_z_ const char * _Ch, _In_ size_t _MaxCount);
_Check_return_ _CRTIMP int    __cdecl _mblen_l(_In_opt_bytecount_(_MaxCount) _Pre_opt_z_ const char * _Ch, _In_ size_t _MaxCount, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP size_t __cdecl _mbstrlen(_In_z_ const char * _Str);
_Check_return_ _CRTIMP size_t __cdecl _mbstrlen_l(_In_z_ const char *_Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP size_t __cdecl _mbstrnlen(_In_z_ const char *_Str, _In_ size_t _MaxCount);
_Check_return_ _CRTIMP size_t __cdecl _mbstrnlen_l(_In_z_ const char *_Str, _In_ size_t _MaxCount, _In_opt_ _locale_t _Locale);
_CRTIMP int    __cdecl mbtowc(_Pre_notnull_ _Post_z_ wchar_t * _DstCh, _In_opt_bytecount_(_SrcSizeInBytes) _Pre_opt_z_ const char * _SrcCh, _In_ size_t _SrcSizeInBytes);
_CRTIMP int    __cdecl _mbtowc_l(_Pre_notnull_ _Post_z_ wchar_t * _DstCh, _In_opt_bytecount_(_SrcSizeInBytes) _Pre_opt_z_ const char * _SrcCh, _In_ size_t _SrcSizeInBytes, _In_opt_ _locale_t _Locale);
_Check_return_opt_ _CRTIMP errno_t __cdecl mbstowcs_s(_Out_opt_ size_t * _PtNumOfCharConverted, _Out_opt_cap_post_count_(_SizeInWords, *_PtNumOfCharConverted) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_count_(_MaxCount) _Pre_z_ const char * _SrcBuf, _In_ size_t _MaxCount );
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_2(errno_t, mbstowcs_s, _Out_opt_ size_t *, _PtNumOfCharConverted, _Deref_post_z_ wchar_t, _Dest, _In_z_ const char *, _Source, _In_ size_t, _MaxCount)
__DEFINE_CPP_OVERLOAD_STANDARD_NFUNC_0_2_SIZE(_CRTIMP, mbstowcs, _Out_opt_z_cap_(_MaxCount), wchar_t, _Dest, _In_z_ const char *, _Source, _In_ size_t, _MaxCount)

_Check_return_opt_ _CRTIMP errno_t __cdecl _mbstowcs_s_l(_Out_opt_ size_t * _PtNumOfCharConverted, _Out_opt_cap_post_count_(_SizeInWords, *_PtNumOfCharConverted) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_count_(_MaxCount) _Pre_z_ const char * _SrcBuf, _In_ size_t _MaxCount, _In_opt_ _locale_t _Locale);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_3(errno_t, _mbstowcs_s_l, _Out_opt_ size_t *, _PtNumOfCharConverted, wchar_t, _Dest, _In_z_ const char *, _Source, _In_ size_t, _MaxCount, _In_opt_ _locale_t, _Locale)
__DEFINE_CPP_OVERLOAD_STANDARD_NFUNC_0_3_SIZE_EX(_CRTIMP, _mbstowcs_l, _mbstowcs_s_l, _Out_opt_z_cap_(_Size) wchar_t, _Out_z_cap_(_MaxCount), wchar_t, _Dest, _In_z_ const char *, _Source, _In_ size_t, _MaxCount, _In_opt_ _locale_t, _Locale)

_Check_return_ _CRTIMP int    __cdecl rand(void);
#if defined(_CRT_RAND_S)
_CRTIMP errno_t __cdecl rand_s ( _Out_ unsigned int *_RandomValue);
#endif

_Check_return_opt_ _CRTIMP int    __cdecl _set_error_mode(_In_ int _Mode);

_CRTIMP void   __cdecl srand(_In_ unsigned int _Seed);
_Check_return_ _CRTIMP double __cdecl strtod(_In_z_ const char * _Str, _Out_opt_ _Deref_post_z_ char ** _EndPtr);
_Check_return_ _CRTIMP double __cdecl _strtod_l(_In_z_ const char * _Str, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP long   __cdecl strtol(_In_z_ const char * _Str, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_ int _Radix );
_Check_return_ _CRTIMP long   __cdecl _strtol_l(_In_z_ const char *_Str, _Out_opt_ _Deref_post_z_ char **_EndPtr, _In_ int _Radix, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP unsigned long __cdecl strtoul(_In_z_ const char * _Str, _Out_opt_ _Deref_post_z_ char ** _EndPtr, _In_ int _Radix);
_Check_return_ _CRTIMP unsigned long __cdecl _strtoul_l(const char * _Str, _Out_opt_ _Deref_post_z_ char **_EndPtr, _In_ int _Radix, _In_opt_ _locale_t _Locale);
#ifndef _CRT_SYSTEM_DEFINED
#define _CRT_SYSTEM_DEFINED
_CRTIMP int __cdecl system(_In_opt_z_ const char * _Command);
#endif
_Check_return_opt_ _CRTIMP errno_t __cdecl _ultoa_s(_In_ unsigned long _Val, _Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ int _Radix);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _ultoa_s, _In_ unsigned long, _Value, char, _Dest, _In_ int, _Radix)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_1_1(char *, __RETURN_POLICY_DST, _CRTIMP, _ultoa, _In_ unsigned long, _Value, _Pre_notnull_ _Post_z_, char, _Dest, _In_ int, _Radix)
_CRTIMP _CRT_INSECURE_DEPRECATE(wctomb_s) int    __cdecl wctomb(_Out_opt_z_bytecap_c_(MB_LEN_MAX) char * _MbCh, _In_ wchar_t _WCh);
_CRTIMP _CRT_INSECURE_DEPRECATE(_wctomb_s_l) int    __cdecl _wctomb_l(_Pre_maybenull_ _Post_z_ char * _MbCh, _In_ wchar_t _WCh, _In_opt_ _locale_t _Locale);
#if __STDC_WANT_SECURE_LIB__
_Check_return_wat_ _CRTIMP errno_t __cdecl wctomb_s(_Out_opt_ int * _SizeConverted, _Out_opt_bytecap_post_bytecount_(_SizeInBytes, *_SizeConverted) char * _MbCh, _In_ rsize_t _SizeInBytes, _In_ wchar_t _WCh);
#endif
_Check_return_wat_ _CRTIMP errno_t __cdecl _wctomb_s_l(_Out_opt_ int * _SizeConverted, _Out_opt_z_bytecap_(_SizeInBytes) char * _MbCh, _In_ size_t _SizeInBytes, _In_ wchar_t _WCh, _In_opt_ _locale_t _Locale);
_Check_return_wat_ _CRTIMP errno_t __cdecl wcstombs_s(_Out_opt_ size_t * _PtNumOfCharConverted, _Out_opt_bytecap_post_bytecount_(_DstSizeInBytes, *_PtNumOfCharConverted) char * _Dst, _In_ size_t _DstSizeInBytes, _In_z_ const wchar_t * _Src, _In_ size_t _MaxCountInBytes);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_2(errno_t, wcstombs_s, _Out_opt_ size_t *, _PtNumOfCharConverted, _Out_opt_bytecap_(_Size) char, _Dest, _In_z_ const wchar_t *, _Source, _In_ size_t, _MaxCount)
__DEFINE_CPP_OVERLOAD_STANDARD_NFUNC_0_2_SIZE(_CRTIMP, wcstombs, _Out_opt_z_cap_(_MaxCount), char, _Dest, _In_z_ const wchar_t *, _Source, _In_ size_t, _MaxCount)
_Check_return_wat_ _CRTIMP errno_t __cdecl _wcstombs_s_l(_Out_opt_ size_t * _PtNumOfCharConverted, _Out_opt_bytecap_post_bytecount_(_DstSizeInBytes, *_PtNumOfCharConverted) char * _Dst, _In_ size_t _DstSizeInBytes, _In_z_ const wchar_t * _Src, _In_ size_t _MaxCountInBytes, _In_opt_ _locale_t _Locale);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_3(errno_t, _wcstombs_s_l, _Out_opt_ size_t *,_PtNumOfCharConverted, _Out_opt_cap_(_Size) char, _Dest, _In_z_ const wchar_t *, _Source, _In_ size_t, _MaxCount, _In_opt_ _locale_t, _Locale)
__DEFINE_CPP_OVERLOAD_STANDARD_NFUNC_0_3_SIZE_EX(_CRTIMP, _wcstombs_l, _wcstombs_s_l, _Out_opt_z_cap_(_Size) char, _Out_z_cap_(_MaxCount), char, _Dest, _In_z_ const wchar_t *, _Source, _In_ size_t, _MaxCount, _In_opt_ _locale_t, _Locale)

#if _MSC_VER >= 1400 && defined(__cplusplus) && defined(_M_CEE)
/*
 * Managed search routines. Note __cplusplus, this is because we only support
 * managed C++.
 */
extern "C++"
{
#if __STDC_WANT_SECURE_LIB__
_Check_return_ void * __clrcall bsearch_s(_In_ const void * _Key, _In_bytecount_x_(_NumOfElements*_SizeOfElements) const void * _Base, _In_ rsize_t _NumOfElements, _In_ rsize_t _SizeOfElements, 
        _In_ int (__clrcall * _PtFuncCompare)(void *, const void *, const void *), void * _Context);
#endif
_Check_return_ void * __clrcall bsearch  (_In_ const void * _Key, _In_bytecount_x_(_NumOfElements*_SizeOfElements) const void * _Base, _In_ size_t _NumOfElements, _In_ size_t _SizeOfElements,
        _In_ int (__clrcall * _PtFuncCompare)(const void *, const void *));

#if __STDC_WANT_SECURE_LIB__
void __clrcall qsort_s(_Inout_bytecap_x_(_NumOfElements*_SizeOfElements) _Post_bytecount_x_(_NumOfElements*_SizeOfElements) void * _Base, 
        _In_ rsize_t _NumOfElements, _In_ rsize_t _SizeOfElment, 
        _In_ int (__clrcall * _PtFuncCompare)(void *, const void *, const void *), void * _Context);
#endif
void __clrcall qsort(_Inout_bytecap_x_(_NumOfElements*_SizeOfElements) _Post_bytecount_x_(_NumOfElements*_SizeOfElements) void * _Base, 
        _In_ size_t _NumOfElements, _In_ size_t _SizeOfElements, 
        _In_ int (__clrcall * _PtFuncCompare)(const void *, const void *));

}
#endif

#ifndef _CRT_ALLOCATION_DEFINED
#define _CRT_ALLOCATION_DEFINED

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)

#pragma push_macro("calloc")
#pragma push_macro("free")
#pragma push_macro("malloc")
#pragma push_macro("realloc")
#pragma push_macro("_recalloc")
#pragma push_macro("_aligned_free")
#pragma push_macro("_aligned_malloc")
#pragma push_macro("_aligned_offset_malloc")
#pragma push_macro("_aligned_realloc")
#pragma push_macro("_aligned_recalloc")
#pragma push_macro("_aligned_offset_realloc")
#pragma push_macro("_aligned_offset_recalloc")
#pragma push_macro("_aligned_msize")

#undef calloc
#undef free
#undef malloc
#undef realloc
#undef _recalloc
#undef _aligned_free
#undef _aligned_malloc
#undef _aligned_offset_malloc
#undef _aligned_realloc
#undef _aligned_recalloc
#undef _aligned_offset_realloc
#undef _aligned_offset_recalloc
#undef _aligned_msize

#endif
_Check_return_ _Ret_opt_bytecap_x_(_NumOfElements* _SizeOfElements) _CRTIMP _CRT_JIT_INTRINSIC _CRTNOALIAS _CRTRESTRICT    void * __cdecl calloc(_In_ size_t _NumOfElements, _In_ size_t _SizeOfElements);
_CRTIMP                     _CRTNOALIAS                                                                             void   __cdecl free(_Inout_opt_ void * _Memory);
_Check_return_ _Ret_opt_bytecap_(_Size) _CRTIMP _CRT_JIT_INTRINSIC _CRTNOALIAS _CRTRESTRICT                              void * __cdecl malloc(_In_ size_t _Size);
_Check_return_ _Ret_opt_bytecap_(_NewSize) _CRTIMP _CRTNOALIAS _CRTRESTRICT                           void * __cdecl realloc(_In_opt_ void * _Memory, _In_ size_t _NewSize);
_Check_return_ _Ret_opt_bytecap_x_(_Count*_Size) _CRTIMP _CRTNOALIAS _CRTRESTRICT                       void * __cdecl _recalloc(_In_opt_ void * _Memory, _In_ size_t _Count, _In_ size_t _Size);
_CRTIMP                     _CRTNOALIAS                                                                             void   __cdecl _aligned_free(_Inout_opt_ void * _Memory);
_Check_return_ _Ret_opt_bytecap_(_Size) _CRTIMP _CRTNOALIAS _CRTRESTRICT                              void * __cdecl _aligned_malloc(_In_ size_t _Size, _In_ size_t _Alignment);
_Check_return_ _Ret_opt_bytecap_(_Size) _CRTIMP _CRTNOALIAS _CRTRESTRICT                              void * __cdecl _aligned_offset_malloc(_In_ size_t _Size, _In_ size_t _Alignment, _In_ size_t _Offset);
_Check_return_ _Ret_opt_bytecap_(_NewSize) _CRTIMP _CRTNOALIAS _CRTRESTRICT                              void * __cdecl _aligned_realloc(_In_opt_ void * _Memory, _In_ size_t _NewSize, _In_ size_t _Alignment);
_Check_return_ _Ret_opt_bytecap_x_(_Count*_Size) _CRTIMP _CRTNOALIAS _CRTRESTRICT                       void * __cdecl _aligned_recalloc(_In_opt_ void * _Memory, _In_ size_t _Count, _In_ size_t _Size, _In_ size_t _Alignment);
_Check_return_ _Ret_opt_bytecap_(_NewSize) _CRTIMP _CRTNOALIAS _CRTRESTRICT                              void * __cdecl _aligned_offset_realloc(_In_opt_ void * _Memory, _In_ size_t _NewSize, _In_ size_t _Alignment, _In_ size_t _Offset);
_Check_return_ _Ret_opt_bytecap_x_(_Count*_Size) _CRTIMP _CRTNOALIAS _CRTRESTRICT                       void * __cdecl _aligned_offset_recalloc(_In_opt_ void * _Memory, _In_ size_t _Count, _In_ size_t _Size, _In_ size_t _Alignment, _In_ size_t _Offset);
_Check_return_ _CRTIMP                                                  size_t __cdecl _aligned_msize(_In_ void * _Memory, _In_ size_t _Alignment, _In_ size_t _Offset);


#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)

#pragma pop_macro("_aligned_msize")
#pragma pop_macro("_aligned_offset_recalloc")
#pragma pop_macro("_aligned_offset_realloc")
#pragma pop_macro("_aligned_recalloc")
#pragma pop_macro("_aligned_realloc")
#pragma pop_macro("_aligned_offset_malloc")
#pragma pop_macro("_aligned_malloc")
#pragma pop_macro("_aligned_free")
#pragma pop_macro("_recalloc")
#pragma pop_macro("realloc")
#pragma pop_macro("malloc")
#pragma pop_macro("free")
#pragma pop_macro("calloc")

#endif

#endif /*_CRT_ALLOCATION_DEFINED */

#ifndef _WSTDLIB_DEFINED

/* wide function prototypes, also declared in wchar.h  */

_Check_return_wat_ _CRTIMP errno_t __cdecl _itow_s (_In_ int _Val, _Out_z_cap_(_SizeInWords) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_ int _Radix);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _itow_s, _In_ int, _Value, wchar_t, _Dest, _In_ int, _Radix)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_1_1(wchar_t *, __RETURN_POLICY_DST, _CRTIMP, _itow, _In_ int, _Value, _Pre_notnull_ _Post_z_, wchar_t, _Dest, _In_ int, _Radix)
_Check_return_wat_ _CRTIMP errno_t __cdecl _ltow_s (_In_ long _Val, _Out_z_cap_(_SizeInWords) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_ int _Radix);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _ltow_s, _In_ long, _Value, wchar_t, _Dest, _In_ int, _Radix)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_1_1(wchar_t *, __RETURN_POLICY_DST, _CRTIMP, _ltow, _In_ long, _Value, _Pre_notnull_ _Post_z_, wchar_t, _Dest, _In_ int, _Radix)
_Check_return_wat_ _CRTIMP errno_t __cdecl _ultow_s (_In_ unsigned long _Val, _Out_z_cap_(_SizeInWords) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_ int _Radix);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _ultow_s, _In_ unsigned long, _Value, wchar_t, _Dest, _In_ int, _Radix)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_1_1(wchar_t *, __RETURN_POLICY_DST, _CRTIMP, _ultow, _In_ unsigned long, _Value, _Pre_notnull_ _Post_z_, wchar_t, _Dest, _In_ int, _Radix)
_Check_return_ _CRTIMP double __cdecl wcstod(_In_z_ const wchar_t * _Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr);
_Check_return_ _CRTIMP double __cdecl _wcstod_l(_In_z_ const wchar_t *_Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP long   __cdecl wcstol(_In_z_ const wchar_t *_Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, int _Radix);
_Check_return_ _CRTIMP long   __cdecl _wcstol_l(_In_z_ const wchar_t *_Str, _Out_opt_ _Deref_post_z_ wchar_t **_EndPtr, int _Radix, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP unsigned long __cdecl wcstoul(_In_z_ const wchar_t *_Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, int _Radix);
_Check_return_ _CRTIMP unsigned long __cdecl _wcstoul_l(_In_z_ const wchar_t *_Str, _Out_opt_ _Deref_post_z_ wchar_t **_EndPtr, int _Radix, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP _CRT_INSECURE_DEPRECATE(_wdupenv_s) wchar_t * __cdecl _wgetenv(_In_z_ const wchar_t * _VarName);
_Check_return_wat_ _CRTIMP errno_t __cdecl _wgetenv_s(_Out_ size_t * _ReturnSize, _Out_z_cap_(_DstSizeInWords) wchar_t * _DstBuf, _In_ size_t _DstSizeInWords, _In_z_ const wchar_t * _VarName);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_1_1(errno_t, _wgetenv_s, _Out_ size_t *, _ReturnSize, wchar_t, _Dest, _In_z_ const wchar_t *, _VarName)

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)
#pragma push_macro("_wdupenv_s")
#undef _wdupenv_s
#endif

_Check_return_wat_ _CRTIMP errno_t __cdecl _wdupenv_s(_Out_ _Deref_post_opt_z_cap_(*_BufferSizeInWords) wchar_t **_Buffer, _Out_opt_ size_t *_BufferSizeInWords, _In_z_ const wchar_t *_VarName);

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)
#pragma pop_macro("_wdupenv_s")
#endif

#ifndef _CRT_WSYSTEM_DEFINED
#define _CRT_WSYSTEM_DEFINED
_CRTIMP int __cdecl _wsystem(_In_opt_z_ const wchar_t * _Command);
#endif
_Check_return_ _CRTIMP double __cdecl _wtof(_In_z_ const wchar_t *_Str);
_Check_return_ _CRTIMP double __cdecl _wtof_l(_In_z_ const wchar_t *_Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP int __cdecl _wtoi(_In_z_ const wchar_t *_Str);
_Check_return_ _CRTIMP int __cdecl _wtoi_l(_In_z_ const wchar_t *_Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP long __cdecl _wtol(_In_z_ const wchar_t *_Str);
_Check_return_ _CRTIMP long __cdecl _wtol_l(_In_z_ const wchar_t *_Str, _In_opt_ _locale_t _Locale);

_Check_return_wat_ _CRTIMP errno_t __cdecl _i64tow_s(_In_ __int64 _Val, _Out_z_cap_(_SizeInWords) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_ int _Radix);
_CRTIMP _CRT_INSECURE_DEPRECATE(_i64tow_s) wchar_t * __cdecl _i64tow(_In_ __int64 _Val, _Pre_notnull_ _Post_z_ wchar_t * _DstBuf, _In_ int _Radix);
_Check_return_wat_ _CRTIMP errno_t __cdecl _ui64tow_s(_In_ unsigned __int64 _Val, _Out_z_cap_(_SizeInWords) wchar_t * _DstBuf, _In_ size_t _SizeInWords, _In_ int _Radix);
_CRTIMP _CRT_INSECURE_DEPRECATE(_ui64tow_s) wchar_t * __cdecl _ui64tow(_In_ unsigned __int64 _Val, _Pre_notnull_ _Post_z_ wchar_t * _DstBuf, _In_ int _Radix);
_Check_return_ _CRTIMP __int64   __cdecl _wtoi64(_In_z_ const wchar_t *_Str);
_Check_return_ _CRTIMP __int64   __cdecl _wtoi64_l(_In_z_ const wchar_t *_Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP __int64   __cdecl _wcstoi64(_In_z_ const wchar_t * _Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, _In_ int _Radix);
_Check_return_ _CRTIMP __int64   __cdecl _wcstoi64_l(_In_z_ const wchar_t * _Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, _In_ int _Radix, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP unsigned __int64  __cdecl _wcstoui64(_In_z_ const wchar_t * _Str, _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, _In_ int _Radix);
_Check_return_ _CRTIMP unsigned __int64  __cdecl _wcstoui64_l(_In_z_ const wchar_t *_Str , _Out_opt_ _Deref_post_z_ wchar_t ** _EndPtr, _In_ int _Radix, _In_opt_ _locale_t _Locale);

#define _WSTDLIB_DEFINED
#endif


#ifndef _POSIX_

/* 
Buffer size required to be passed to _gcvt, fcvt and other fp conversion routines
*/
#define _CVTBUFSIZE (309+40) /* # of digits in max. dp value + slop */

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)

#pragma push_macro("_fullpath")
#undef _fullpath

#endif

_Check_return_ _CRTIMP char * __cdecl _fullpath(_Out_opt_z_cap_(_SizeInBytes) char * _FullPath, _In_z_ const char * _Path, _In_ size_t _SizeInBytes);

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)

#pragma pop_macro("_fullpath")

#endif

_Check_return_wat_ _CRTIMP errno_t __cdecl _ecvt_s(_Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ double _Val, _In_ int _NumOfDights, _Out_ int * _PtDec, _Out_ int * _PtSign);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_0_4(errno_t, _ecvt_s, char, _Dest, _In_ double, _Value, _In_ int, _NumOfDigits, _Out_ int *, _PtDec, _Out_ int *, _PtSign)
_Check_return_ _CRTIMP _CRT_INSECURE_DEPRECATE(_ecvt_s) char * __cdecl _ecvt(_In_ double _Val, _In_ int _NumOfDigits, _Out_ int * _PtDec, _Out_ int * _PtSign);
_Check_return_wat_ _CRTIMP errno_t __cdecl _fcvt_s(_Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ double _Val, _In_ int _NumOfDec, _Out_ int * _PtDec, _Out_ int * _PtSign);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_0_4(errno_t, _fcvt_s, char, _Dest, _In_ double, _Value, _In_ int, _NumOfDigits, _Out_ int *, _PtDec, _Out_ int *, _PtSign)
_Check_return_ _CRTIMP _CRT_INSECURE_DEPRECATE(_fcvt_s) char * __cdecl _fcvt(_In_ double _Val, _In_ int _NumOfDec, _Out_ int * _PtDec, _Out_ int * _PtSign);
_CRTIMP errno_t __cdecl _gcvt_s(_Out_z_cap_(_Size) char * _DstBuf, _In_ size_t _Size, _In_ double _Val, _In_ int _NumOfDigits);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_0_2(errno_t, _gcvt_s, char, _Dest, _In_ double, _Value, _In_ int, _NumOfDigits)
_CRTIMP _CRT_INSECURE_DEPRECATE(_gcvt_s) char * __cdecl _gcvt(_In_ double _Val, _In_ int _NumOfDigits, _Pre_notnull_ _Post_z_ char * _DstBuf);

_Check_return_ _CRTIMP int __cdecl _atodbl(_Out_ _CRT_DOUBLE * _Result, _In_z_ char * _Str);
_Check_return_ _CRTIMP int __cdecl _atoldbl(_Out_ _LDOUBLE * _Result, _In_z_ char * _Str);
_Check_return_ _CRTIMP int __cdecl _atoflt(_Out_ _CRT_FLOAT * _Result, _In_z_ char * _Str);
_Check_return_ _CRTIMP int __cdecl _atodbl_l(_Out_ _CRT_DOUBLE * _Result, _In_z_ char * _Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP int __cdecl _atoldbl_l(_Out_ _LDOUBLE * _Result, _In_z_ char * _Str, _In_opt_ _locale_t _Locale);
_Check_return_ _CRTIMP int __cdecl _atoflt_l(_Out_ _CRT_FLOAT * _Result, _In_z_ char * _Str, _In_opt_ _locale_t _Locale);
        _Check_return_ unsigned long __cdecl _lrotl(_In_ unsigned long _Val, _In_ int _Shift);
        _Check_return_ unsigned long __cdecl _lrotr(_In_ unsigned long _Val, _In_ int _Shift);
_Check_return_wat_ _CRTIMP_ALTERNATIVE errno_t   __cdecl _makepath_s(_Out_z_cap_(_SizeInWords) char * _PathResult, _In_ size_t _SizeInWords, _In_opt_z_ const char * _Drive, _In_opt_z_ const char * _Dir, _In_opt_z_ const char * _Filename,
        _In_opt_z_ const char * _Ext);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_0_4(errno_t, _makepath_s, char, _Path, _In_opt_z_ const char *, _Drive, _In_opt_z_ const char *, _Dir, _In_opt_z_ const char *, _Filename, _In_opt_z_ const char *, _Ext)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_0_4(void, __RETURN_POLICY_VOID, _CRTIMP, _makepath, _Pre_notnull_ _Post_z_, char, _Path, _In_opt_z_ const char *, _Drive, _In_opt_z_ const char *, _Dir, _In_opt_z_ const char *, _Filename, _In_opt_z_ const char *, _Ext)

#if _MSC_VER >= 1400 && defined(_M_CEE)
		_onexit_m_t    __clrcall _onexit_m_appdomain(_onexit_m_t _Function);
	#if defined(_M_CEE_MIXED)
		_onexit_m_t    __clrcall _onexit_m(_onexit_m_t _Function);
	#else
		inline _onexit_m_t    __clrcall _onexit_m(_onexit_t _Function)
		{
			return _onexit_m_appdomain(_Function);
		}
	#endif
        
#endif
#if defined(_M_CEE_PURE)
        /* In pure mode, _onexit is the same as _onexit_m_appdomain */
extern "C++"
{
inline  _onexit_t    __clrcall _onexit
(
    _onexit_t _Function
)
{
    return _onexit_m_appdomain(_Function);
}
}
#else
        _onexit_t __cdecl _onexit(_In_opt_ _onexit_t _Func);
#endif
        
#ifndef _CRT_PERROR_DEFINED
#define _CRT_PERROR_DEFINED
_CRTIMP void __cdecl perror(_In_opt_z_ const char * _ErrMsg);
#endif

#pragma warning (push)
#pragma warning (disable:6540) // the functions below have declspecs in their declarations in the windows headers, causing PREfast to fire 6540 here
_Check_return_ _CRTIMP int    __cdecl _putenv(_In_z_ const char * _EnvString);
_Check_return_wat_ _CRTIMP errno_t __cdecl _putenv_s(_In_z_ const char * _Name, _In_z_ const char * _Value);
        _Check_return_ unsigned int __cdecl _rotl(_In_ unsigned int _Val, _In_ int _Shift);
        _Check_return_ unsigned __int64 __cdecl _rotl64(_In_ unsigned __int64 _Val, _In_ int _Shift);
        _Check_return_ unsigned int __cdecl _rotr(_In_ unsigned int _Val, _In_ int _Shift);
        _Check_return_ unsigned __int64 __cdecl _rotr64(_In_ unsigned __int64 _Val, _In_ int _Shift);
#pragma warning (pop)

_CRTIMP errno_t __cdecl _searchenv_s(_In_z_ const char * _Filename, _In_z_ const char * _EnvVar, _Out_z_cap_(_SizeInBytes) char * _ResultPath, _In_ size_t _SizeInBytes);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_2_0(errno_t, _searchenv_s, _In_z_ const char *, _Filename, _In_z_ const char *, _EnvVar, char, _ResultPath)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_2_0(void, __RETURN_POLICY_VOID, _CRTIMP, _searchenv, _In_z_ const char *, _Filename, _In_z_ const char *, _EnvVar, _Pre_notnull_ _Post_z_, char, _ResultPath)

_CRT_INSECURE_DEPRECATE(_splitpath_s) _CRTIMP void   __cdecl _splitpath(_In_z_ const char * _FullPath, _Pre_maybenull_ _Post_z_ char * _Drive, _Pre_maybenull_ _Post_z_ char * _Dir, _Pre_maybenull_ _Post_z_ char * _Filename, _Pre_maybenull_ _Post_z_ char * _Ext);
_Check_return_wat_ _CRTIMP_ALTERNATIVE errno_t  __cdecl _splitpath_s(_In_z_ const char * _FullPath, 
		_Out_opt_z_cap_(_DriveSize) char * _Drive, _In_ size_t _DriveSize, 
		_Out_opt_z_cap_(_DirSize) char * _Dir, _In_ size_t _DirSize, 
		_Out_opt_z_cap_(_FilenameSize) char * _Filename, _In_ size_t _FilenameSize, 
		_Out_opt_z_cap_(_ExtSize) char * _Ext, _In_ size_t _ExtSize);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_SPLITPATH(errno_t, _splitpath_s,  char, _Dest)

_CRTIMP void   __cdecl _swab(_Inout_cap_(_SizeInBytes) _Post_count_(_SizeInBytes) char * _Buf1, _Inout_cap_(_SizeInBytes) _Post_count_(_SizeInBytes) char * _Buf2, int _SizeInBytes);

#ifndef _WSTDLIBP_DEFINED

/* wide function prototypes, also declared in wchar.h  */

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)
#pragma push_macro("_wfullpath")
#undef _wfullpath
#endif

_Check_return_ _CRTIMP wchar_t * __cdecl _wfullpath(_Out_opt_z_cap_(_SizeInWords) wchar_t * _FullPath, _In_z_ const wchar_t * _Path, _In_ size_t _SizeInWords);

#if defined(_DEBUG) && defined(_CRTDBG_MAP_ALLOC)
#pragma pop_macro("_wfullpath")
#endif

_Check_return_wat_ _CRTIMP_ALTERNATIVE errno_t __cdecl _wmakepath_s(_Out_z_cap_(_SIZE) wchar_t * _PathResult, _In_ size_t _SIZE, _In_opt_z_ const wchar_t * _Drive, _In_opt_z_ const wchar_t * _Dir, _In_opt_z_ const wchar_t * _Filename,
        _In_opt_z_ const wchar_t * _Ext);        
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_0_4(errno_t, _wmakepath_s, wchar_t, _ResultPath, _In_opt_z_ const wchar_t *, _Drive, _In_opt_z_ const wchar_t *, _Dir, _In_opt_z_ const wchar_t *, _Filename, _In_opt_z_ const wchar_t *, _Ext)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_0_4(void, __RETURN_POLICY_VOID, _CRTIMP, _wmakepath, _Pre_notnull_ _Post_z_, wchar_t, _ResultPath, _In_opt_z_ const wchar_t *, _Drive, _In_opt_z_ const wchar_t *, _Dir, _In_opt_z_ const wchar_t *, _Filename, _In_opt_z_ const wchar_t *, _Ext)
#ifndef _CRT_WPERROR_DEFINED
#define _CRT_WPERROR_DEFINED
_CRTIMP void __cdecl _wperror(_In_opt_z_ const wchar_t * _ErrMsg);
#endif 
_Check_return_ _CRTIMP int    __cdecl _wputenv(_In_z_ const wchar_t * _EnvString);
_Check_return_wat_ _CRTIMP errno_t __cdecl _wputenv_s(_In_z_ const wchar_t * _Name, _In_z_ const wchar_t * _Value);
_CRTIMP errno_t __cdecl _wsearchenv_s(_In_z_ const wchar_t * _Filename, _In_z_ const wchar_t * _EnvVar, _Out_z_cap_(_SizeInWords) wchar_t * _ResultPath, _In_ size_t _SizeInWords);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_2_0(errno_t, _wsearchenv_s, _In_z_ const wchar_t *, _Filename, _In_z_ const wchar_t *, _EnvVar, wchar_t, _ResultPath)
__DEFINE_CPP_OVERLOAD_STANDARD_FUNC_2_0(void, __RETURN_POLICY_VOID, _CRTIMP, _wsearchenv, _In_z_ const wchar_t *, _Filename, _In_z_ const wchar_t *, _EnvVar, _Pre_notnull_ _Post_z_, wchar_t, _ResultPath)
_CRT_INSECURE_DEPRECATE(_wsplitpath_s) _CRTIMP void   __cdecl _wsplitpath(_In_z_ const wchar_t * _FullPath, _Pre_maybenull_ _Post_z_ wchar_t * _Drive, _Pre_maybenull_ _Post_z_ wchar_t * _Dir, _Pre_maybenull_ _Post_z_ wchar_t * _Filename, _Pre_maybenull_ _Post_z_ wchar_t * _Ext);
_CRTIMP_ALTERNATIVE errno_t __cdecl _wsplitpath_s(_In_z_ const wchar_t * _FullPath, 
		_Out_opt_z_cap_(_DriveSize) wchar_t * _Drive, _In_ size_t _DriveSize, 
		_Out_opt_z_cap_(_DirSize) wchar_t * _Dir, _In_ size_t _DirSize, 
		_Out_opt_z_cap_(_FilenameSize) wchar_t * _Filename, _In_ size_t _FilenameSize, 
		_Out_opt_z_cap_(_ExtSize) wchar_t * _Ext, _In_ size_t _ExtSize);
__DEFINE_CPP_OVERLOAD_SECURE_FUNC_SPLITPATH(errno_t, _wsplitpath_s, wchar_t, _Path)

#define _WSTDLIBP_DEFINED
#endif

/* The Win32 API SetErrorMode, Beep and Sleep should be used instead. */
_CRT_OBSOLETE(SetErrorMode) _CRTIMP void __cdecl _seterrormode(_In_ int _Mode);
_CRT_OBSOLETE(Beep) _CRTIMP void __cdecl _beep(_In_ unsigned _Frequency, _In_ unsigned _Duration);
_CRT_OBSOLETE(Sleep) _CRTIMP void __cdecl _sleep(_In_ unsigned long _Duration);

#endif  /* _POSIX_ */

#if     !__STDC__

#ifndef _POSIX_

/* Non-ANSI names for compatibility */

#ifndef __cplusplus
#define max(a,b)    (((a) > (b)) ? (a) : (b))
#define min(a,b)    (((a) < (b)) ? (a) : (b))
#endif

#define sys_errlist _sys_errlist
#define sys_nerr    _sys_nerr
#define environ     _environ

#pragma warning(push)
#pragma warning(disable: 4141) /* Using deprecated twice */ 
_Check_return_ _CRT_NONSTDC_DEPRECATE(_ecvt) _CRT_INSECURE_DEPRECATE(_ecvt_s) _CRTIMP char * __cdecl ecvt(_In_ double _Val, _In_ int _NumOfDigits, _Out_ int * _PtDec, _Out_ int * _PtSign);
_Check_return_ _CRT_NONSTDC_DEPRECATE(_fcvt) _CRT_INSECURE_DEPRECATE(_fcvt_s) _CRTIMP char * __cdecl fcvt(_In_ double _Val, _In_ int _NumOfDec, _Out_ int * _PtDec, _Out_ int * _PtSign);
_CRT_NONSTDC_DEPRECATE(_gcvt) _CRT_INSECURE_DEPRECATE(_fcvt_s)		_CRTIMP char * __cdecl gcvt(_In_ double _Val, _In_ int _NumOfDigits, _Pre_notnull_ _Post_z_ char * _DstBuf);
_CRT_NONSTDC_DEPRECATE(_itoa) _CRT_INSECURE_DEPRECATE(_itoa_s)		_CRTIMP char * __cdecl itoa(_In_ int _Val, _Pre_notnull_ _Post_z_ char * _DstBuf, _In_ int _Radix);
_CRT_NONSTDC_DEPRECATE(_ltoa) _CRT_INSECURE_DEPRECATE(_ltoa_s)		_CRTIMP char * __cdecl ltoa(_In_ long _Val, _Pre_notnull_ _Post_z_ char * _DstBuf, _In_ int _Radix);
_Check_return_ _CRT_NONSTDC_DEPRECATE(_putenv) _CRTIMP int    __cdecl putenv(_In_z_ const char * _EnvString);
_CRT_NONSTDC_DEPRECATE(_swab)										_CRTIMP void   __cdecl swab(_Inout_z_bytecap_(_SizeInBytes) char * _Buf1,_Inout_z_bytecap_(_SizeInBytes) char * _Buf2, _In_ int _SizeInBytes);
_CRT_NONSTDC_DEPRECATE(_ultoa) _CRT_INSECURE_DEPRECATE(_ultoa_s)	_CRTIMP char * __cdecl ultoa(_In_ unsigned long _Val, _Pre_notnull_ _Post_z_ char * _Dstbuf, _In_ int _Radix);
#pragma warning(pop)
onexit_t __cdecl onexit(_In_opt_ onexit_t _Func);

#endif  /* _POSIX_ */

#endif  /* __STDC__ */

#ifdef  __cplusplus
}

#endif

#ifdef  _MSC_VER
#pragma pack(pop)
#endif  /* _MSC_VER */

#endif  /* _INC_STDLIB */

From tracker at bugs.pypy.org  Tue Sep 11 02:49:50 2012
From: tracker at bugs.pypy.org (Mark Roberts)
Date: Tue, 11 Sep 2012 00:49:50 +0000
Subject: [pypy-issue] [issue1199] psycopg2ct 2.4.4 segmentation fault on
	Ubuntu 12.04 64-bit
In-Reply-To: <1341163597.62.0.605992819875.issue1199@bugs.pypy.org>
Message-ID: <1347324590.3.0.334334347041.issue1199@bugs.pypy.org>


Mark Roberts <wizzat at gmail.com> added the comment:

I am able to reproduce this.  It appears to be fixed in the latest nightly build 
(pypy-c-jit-57259-1feab101071c-linux64)

----------
nosy: +wizzat
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1199>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 11:06:41 2012
From: tracker at bugs.pypy.org (Zooko)
Date: Tue, 11 Sep 2012 09:06:41 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>


New submission from Zooko <zooko at zooko.com>:

This script has never done this under CPython, but just now I got this failure:

MAIN zompu:~/playground/stringchain$ time
PYTHONPATH=/usr/local/site-packages/pyutil-1.9.3_2-py2.7.egg/:/usr/local/site-packages/statprof-0.1.2-py2.7.egg
pypy -OOu -c 'from stringchain.bench import bench; bench.slow_bench(profile=True)'
N is the total number of bytes produced and consumed; results are in nanoseconds
per byte (scientific notation)

impl:  StringChain
task:  accumulate_then_one_gulp
N:  100000000Traceback (most recent call last):
  File "app_main.py", line 51, in run_toplevel
  File "app_main.py", line 575, in run_it
  File "<string>", line 1, in <module>
  File "stringchain/bench/bench.py", line 469, in slow_bench
    benchutil.rep_bench(task, BSIZE, initfunc=initfunc,
UNITS_PER_SECOND=UNITS_PER_SECOND)#, profile=profile,
profresultsfileobj=profresultsfileobj)
  File
"/usr/local/stow/pypy-1.9/site-packages/pyutil-1.9.3_2-py2.7.egg/pyutil/benchutil.py",
line 142, in rep_bench
    initfunc(n)
  File "stringchain/bench/bench.py", line 337, in init
    self.l.append(randstr(nextchunksize))
  File
"/usr/local/stow/pypy-1.9/site-packages/pyutil-1.9.3_2-py2.7.egg/pyutil/randutil.py",
line 84, in insecurerandstr
    return os.urandom(n)
  File "/usr/local/stow/pypy-1.9/lib-python/2.7/os.py", line 756, in urandom
    bs += read(_urandomfd, n - len(bs))
OSError: [Errno 4] Interrupted system call


real    0m14.677s
user    0m1.703s
sys     0m12.936s
MAIN zompu:~/playground/stringchain$ time
PYTHONPATH=/usr/local/site-packages/pyutil-1.9.3_2-py2.7.egg/:/usr/local/site-packages/statprof-0.1.2-py2.7.egg
pypy -OOu -c 'from stringchain.bench import bench; bench.slow_bench(profile=True)'
N is the total number of bytes produced and consumed; results are in nanoseconds
per byte (scientific notation)

impl:  StringChain
task:  accumulate_then_one_gulp
N:  100000000Traceback (most recent call last):
  File "app_main.py", line 51, in run_toplevel
  File "app_main.py", line 575, in run_it
  File "<string>", line 1, in <module>
  File "stringchain/bench/bench.py", line 469, in slow_bench
    benchutil.rep_bench(task, BSIZE, initfunc=initfunc,
UNITS_PER_SECOND=UNITS_PER_SECOND)#, profile=profile,
profresultsfileobj=profresultsfileobj)
  File
"/usr/local/stow/pypy-1.9/site-packages/pyutil-1.9.3_2-py2.7.egg/pyutil/benchutil.py",
line 142, in rep_bench
    initfunc(n)
  File "stringchain/bench/bench.py", line 337, in init
    self.l.append(randstr(nextchunksize))
  File
"/usr/local/stow/pypy-1.9/site-packages/pyutil-1.9.3_2-py2.7.egg/pyutil/randutil.py",
line 84, in insecurerandstr
    return os.urandom(n)
  File "/usr/local/stow/pypy-1.9/lib-python/2.7/os.py", line 756, in urandom
    bs += read(_urandomfd, n - len(bs))
OSError: [Errno 4] Interrupted system call


real    0m28.911s
user    0m3.436s
sys     0m25.242s

It seems to take a different amount of time to fail each time I run it. I'm
using statprof -- http://pypi.python.org/pypi/statprof/ -- which sets a timer
signal -- which is probably part of the problem.

----------
messages: 4735
nosy: pypy-issue, zooko
priority: bug
status: unread
title: Interrupted system call from os.urandom

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 11:40:22 2012
From: tracker at bugs.pypy.org (David)
Date: Tue, 11 Sep 2012 09:40:22 +0000
Subject: [pypy-issue] [issue1259] module __file__ attribute shows wrong path
In-Reply-To: <1347356422.5.0.518215816362.issue1259@bugs.pypy.org>
Message-ID: <1347356422.5.0.518215816362.issue1259@bugs.pypy.org>


New submission from David <david.schneider at uni-duesseldorf.de>:

The __file__ attribute for some modules shows a path corresponding to the system 
where pypy was translated and not where it is running. 
This causes one of the module/math app-level tests that use this attribute to load 
the math_testcases.txt file to fail, if the test are run on a different machine as 
the one used for translation. 

Examples:
>>>> import math; math.__file__
'/Users/pypy/buildslave/pypy-c-jit-macosx-x86-64/build/pypy/module/math'
or 
>>>> import abc; abc.__file__
'/Users/pypy/buildslave/pypy-c-jit-macosx-x86-64/build/lib-python/2.7/abc.py'

In other cases, like ctypes, the behavior is as expected, e.g.:
>>>> import ctypes; ctypes.__file__
'/opt/pypy/lib-python/2.7/ctypes/__init__.pyc'

----------
messages: 4736
nosy: bivab, pypy-issue
priority: bug
status: unread
title: module __file__ attribute shows wrong path

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1259>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 12:02:54 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Tue, 11 Sep 2012 10:02:54 +0000
Subject: [pypy-issue] [issue1259] module __file__ attribute shows wrong path
In-Reply-To: <1347356422.5.0.518215816362.issue1259@bugs.pypy.org>
Message-ID: <1347357774.36.0.616660652573.issue1259@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

"math" should maybe not have a __file__ attribute at all (it doesn't in CPython).

"abc" has a broken __file__ attribute probably coming from a .pyc file?  But I
don't understand why there would be a pyc file in the tarball...

----------
nosy: +arigo
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1259>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 13:25:28 2012
From: tracker at bugs.pypy.org (David)
Date: Tue, 11 Sep 2012 11:25:28 +0000
Subject: [pypy-issue] [issue1259] module __file__ attribute shows wrong path
In-Reply-To: <1347356422.5.0.518215816362.issue1259@bugs.pypy.org>
Message-ID: <1347362728.55.0.209241283563.issue1259@bugs.pypy.org>


David <david.schneider at uni-duesseldorf.de> added the comment:

For "abc" this also happens when placing the pypy-c binary in a clean checkout and 
starting it there.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1259>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 18:18:30 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Tue, 11 Sep 2012 16:18:30 +0000
Subject: [pypy-issue] [issue1259] module __file__ attribute shows wrong path
In-Reply-To: <1347356422.5.0.518215816362.issue1259@bugs.pypy.org>
Message-ID: <1347380310.66.0.612421135206.issue1259@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

"abc.py" is imported before the ObjSpace is frozen; the whole module is probably 
a prebuilt constant.

----------
nosy: +afa

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1259>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 18:54:47 2012
From: tracker at bugs.pypy.org (mattip)
Date: Tue, 11 Sep 2012 16:54:47 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347382487.04.0.583840539628.issue1258@bugs.pypy.org>


mattip <matti.picus at gmail.com> added the comment:

usually one needs to reinstall all site-packages into the untarred or unzipped 
pypy, in general you cannot use the /usr/local/site-packages from cpython for 
pypy.

----------
nosy: +mattip
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 19:03:19 2012
From: tracker at bugs.pypy.org (Zooko)
Date: Tue, 11 Sep 2012 17:03:19 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347382999.97.0.426088690515.issue1258@bugs.pypy.org>


Zooko <zooko at zooko.com> added the comment:

All of the Python code was installed specifically for Pypy. There are no CPython
packages being imported.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 20:48:49 2012
From: tracker at bugs.pypy.org (mattip)
Date: Tue, 11 Sep 2012 18:48:49 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347389329.59.0.419430511088.issue1258@bugs.pypy.org>


mattip <matti.picus at gmail.com> added the comment:

help me reproduce the problem, here is what I have so far, starting from last 
night's 32 bit linux pypy

$ virtualenv -p pypy-c-jit-57259-1feab101071c-linux/bin/pypy 
pypy_for_stringchains
$ pypy_for_stringchains/bin/pip install statprof
$ pypy_for_stringchains/bin/pip install pyutil
$ pypy_for_stringchains/bin/pip install stringchain
$ time pypy_for_stringchains/bin/pypy -OOu -c 'from stringchain.bench import 
bench; bench.slow_bench(profile=True)'

output from the last command is

Traceback (most recent call last):
  File "app_main.py", line 51, in run_toplevel
  File "app_main.py", line 546, in run_it
  File "<string>", line 1, in <module>
TypeError: slow_bench() takes no arguments (1 given)

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 20:54:16 2012
From: tracker at bugs.pypy.org (Zooko)
Date: Tue, 11 Sep 2012 18:54:16 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347389656.02.0.341191819955.issue1258@bugs.pypy.org>


Zooko <zooko at zooko.com> added the comment:

Okay, I'll get back to you. If you want to try a stab at isolating what I
suspect is the problem, just use statprof and do something like this:

import statprof
statprof.start()
while True:
    os.urandom(32)

But I'll get back to you with instructions to reproduce the exact software I
have here, once I package it up and upload it to PyPI...

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 20:58:55 2012
From: tracker at bugs.pypy.org (mattip)
Date: Tue, 11 Sep 2012 18:58:55 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347389935.94.0.745913116364.issue1258@bugs.pypy.org>


mattip <matti.picus at gmail.com> added the comment:

indeed, the 4 line test crashes for me too. Much simpler :)

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 21:01:45 2012
From: tracker at bugs.pypy.org (Zooko)
Date: Tue, 11 Sep 2012 19:01:45 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347390105.87.0.592012512819.issue1258@bugs.pypy.org>


Zooko <zooko at zooko.com> added the comment:

Great! Thanks.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 11 23:00:46 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Tue, 11 Sep 2012 21:00:46 +0000
Subject: [pypy-issue] [issue1258] Interrupted system call from os.urandom
In-Reply-To: <1347354401.32.0.392307653813.issue1258@bugs.pypy.org>
Message-ID: <1347397246.27.0.613304377851.issue1258@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

The script fails consistently for me with CPython 2.7.2, and seems to pass with 2.7.3.
The difference is that os.urandom() has been rewritten: it's now a builtin function, but more 
importantly it correctly handles EINTR and retries.

PyPy has just been updated to version 2.7.3, and this version runs the script without failure 
(during at least 1 minute).

----------
nosy: +afa
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1258>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 12 01:38:48 2012
From: tracker at bugs.pypy.org (Julian Berman)
Date: Tue, 11 Sep 2012 23:38:48 +0000
Subject: [pypy-issue] [issue1260] sys.exc_info not being set before entering
	a context	manager's __exit__
In-Reply-To: <1347406728.03.0.619393304673.issue1260@bugs.pypy.org>
Message-ID: <1347406728.03.0.619393304673.issue1260@bugs.pypy.org>


New submission from Julian Berman <Julian+PyPy at GrayVines.com>:

The following code raises a KeyboardInterrupt on CPython but a TypeError on PyPy 
(1.9.0):

class Foo(object):
    def __enter__(self):
        return self

    def __exit__(self, exc_type, exc_value, traceback):
        if exc_type == KeyboardInterrupt:
            raise

with Foo():
    raise KeyboardInterrupt()

----------
messages: 4747
nosy: Julian, pypy-issue
priority: bug
status: unread
title: sys.exc_info not being set before entering a context manager's __exit__

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1260>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 12 07:58:01 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Wed, 12 Sep 2012 05:58:01 +0000
Subject: [pypy-issue] [issue1260] sys.exc_info not being set before entering
	a context	manager's __exit__
In-Reply-To: <1347406728.03.0.619393304673.issue1260@bugs.pypy.org>
Message-ID: <1347429481.19.0.886971277288.issue1260@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

Interesting corner case... PyPy behavior is:

Traceback (most recent call last):
  File "app_main.py", line 51, in run_toplevel
  File "/tmp/t.py", line 10, in <module>
    raise KeyboardInterrupt()
  File "/tmp/t.py", line 7, in __exit__
    raise
TypeError: raise: no active exception to re-raise

CPython2.7 acts as expected, but note that python2.6 is similar to pypy here:
Traceback (most recent call last):
  File "/tmp/t.py", line 10, in <module>
    raise KeyboardInterrupt()
  File "/tmp/t.py", line 7, in __exit__
    raise
TypeError: exceptions must be old-style classes or derived from BaseException, not NoneType

And CPython3.2 (or 3.3) does not display the correct traceback:
Traceback (most recent call last):
  File "/tmp/t.py", line 10, in <module>
    raise KeyboardInterrupt()
  File "/tmp/t.py", line 10, in <module>
    raise KeyboardInterrupt()
KeyboardInterrupt

----------
nosy: +afa
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1260>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 12 13:59:52 2012
From: tracker at bugs.pypy.org (Ariel Ben-Yehuda)
Date: Wed, 12 Sep 2012 11:59:52 +0000
Subject: [pypy-issue] [issue1261] Implement PEP 3138
In-Reply-To: <1347451192.44.0.431539503569.issue1261@bugs.pypy.org>
Message-ID: <1347451192.44.0.431539503569.issue1261@bugs.pypy.org>


New submission from Ariel Ben-Yehuda <arielb1 at mail.tau.ac.il>:

Also fixes a bug with surrogate handling

----------
files: pep3138-rc.patch
messages: 4749
nosy: arielby, pypy-issue
priority: feature
status: unread
title: Implement PEP 3138

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1261>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 12 14:01:22 2012
From: tracker at bugs.pypy.org (Ariel Ben-Yehuda)
Date: Wed, 12 Sep 2012 12:01:22 +0000
Subject: [pypy-issue] [issue1262] Implement PEP 3138
In-Reply-To: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>
Message-ID: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>


New submission from Ariel Ben-Yehuda <arielb1 at mail.tau.ac.il>:

Also fixes a bug with surrogate handling

don't forget to regenerate the unicode tables after applying the patch

----------
files: pep3138-rc.patch
messages: 4750
nosy: arielby, pypy-issue
priority: feature
status: unread
title: Implement PEP 3138

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1262>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 12 14:24:35 2012
From: tracker at bugs.pypy.org (Ariel Ben-Yehuda)
Date: Wed, 12 Sep 2012 12:24:35 +0000
Subject: [pypy-issue] [issue1262] Implement PEP 3138
In-Reply-To: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>
Message-ID: <1347452675.96.0.875000594555.issue1262@bugs.pypy.org>


Ariel Ben-Yehuda <arielb1 at mail.tau.ac.il> added the comment:

new patch

----------
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1262>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 12 14:47:12 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Wed, 12 Sep 2012 12:47:12 +0000
Subject: [pypy-issue] [issue1261] Implement PEP 3138
In-Reply-To: <1347451192.44.0.431539503569.issue1261@bugs.pypy.org>
Message-ID: <1347454032.88.0.191368533606.issue1261@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

Duplicate. see issue1262 instead.

----------
nosy: +afa
status: unread -> invalid

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1261>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 13 00:12:10 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Wed, 12 Sep 2012 22:12:10 +0000
Subject: [pypy-issue] [issue1260] sys.exc_info not being set before entering
	a context	manager's __exit__
In-Reply-To: <1347406728.03.0.619393304673.issue1260@bugs.pypy.org>
Message-ID: <1347487930.13.0.343991561335.issue1260@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

Fixed in 90debdd98abb. Thanks for the nice test case!

----------
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1260>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 13 19:52:23 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Thu, 13 Sep 2012 17:52:23 +0000
Subject: [pypy-issue] [issue1263] virtual-ize raw_mallocs in the JIT due to
	CFFI
In-Reply-To: <1347558743.79.0.46208223373.issue1263@bugs.pypy.org>
Message-ID: <1347558743.79.0.46208223373.issue1263@bugs.pypy.org>


New submission from Armin Rigo <armin.rigo at gmail.com>:

The JIT front-end (or maybe the optimizer) needs to optimize away the
raw_mallocs left around by CFFI calls.

----------
messages: 4755
nosy: arigo, pypy-issue
priority: performance bug
status: unread
title: virtual-ize raw_mallocs in the JIT due to CFFI

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1263>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 13 19:53:22 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Thu, 13 Sep 2012 17:53:22 +0000
Subject: [pypy-issue] [issue1263] Port CFFI to Windows
In-Reply-To: <1347558743.79.0.46208223373.issue1263@bugs.pypy.org>
Message-ID: <1347558802.82.0.839150861735.issue1263@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

_cffi_backend needs to be ported to Windows

----------
priority: performance bug -> bug
status: unread -> chatting
title: virtual-ize raw_mallocs in the JIT due to CFFI -> Port CFFI to Windows

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1263>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 15 09:30:33 2012
From: tracker at bugs.pypy.org (bdk)
Date: Sat, 15 Sep 2012 07:30:33 +0000
Subject: [pypy-issue] [issue1264] segmentation faults/fatal rpython errors
	with threading +	import
In-Reply-To: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>
Message-ID: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>


New submission from bdk <bdkearns at gmail.com>:

minimal test case attached, crashes on my system almost 100% repeatedly. get same 
failures replacing xattr with other .so modules.
with linux64 nightly Python 2.7.3 (f07eec99afa5, Sep 13 2012, 04:52:35)

----------
messages: 4757
nosy: bdk, pypy-issue
priority: bug
status: unread
title: segmentation faults/fatal rpython errors with threading + import

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1264>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 15 12:58:01 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Sat, 15 Sep 2012 10:58:01 +0000
Subject: [pypy-issue] [issue1264] segmentation faults/fatal rpython errors
	with threading +	import
In-Reply-To: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>
Message-ID: <1347706681.43.0.757069584349.issue1264@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

I have an awful thought: is it possible that cpyext releases the GIL when it 
calls C functions?
We should check make_generic_cpy_call() in module/cpyext/api.py.

----------
nosy: +afa
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1264>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 15 18:21:13 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 15 Sep 2012 16:21:13 +0000
Subject: [pypy-issue] [issue1264] segmentation faults/fatal rpython errors
	with threading +	import
In-Reply-To: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>
Message-ID: <1347726073.82.0.549148652707.issue1264@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Found the bug together with Amaury.  Will be fixed, but in the meantime, as a
workaround, you can start your program with "import cpyext", anywhere before you
create threads.

----------
nosy: +arigo

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1264>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 15 19:47:50 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Sat, 15 Sep 2012 17:47:50 +0000
Subject: [pypy-issue] [issue1264] segmentation faults/fatal rpython errors
	with threading +	import
In-Reply-To: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>
Message-ID: <1347731270.29.0.142511804844.issue1264@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

Fixed with c23cd52909fb. Thanks for the report!

----------
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1264>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 15 19:57:11 2012
From: tracker at bugs.pypy.org (bdk)
Date: Sat, 15 Sep 2012 17:57:11 +0000
Subject: [pypy-issue] [issue1264] segmentation faults/fatal rpython errors
	with threading +	import
In-Reply-To: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>
Message-ID: <1347731831.26.0.921588606278.issue1264@bugs.pypy.org>


bdk <bdkearns at gmail.com> added the comment:

this workaround prevents the crash in this test case of import, but is not a 
general workaround for the bug -- if i run the second test case that calls a 
function in the extension module it still crashes 5-10% of the runs.

does this fix work for the second test case as well? i'm running the second in a 
loop: for i in {1..100}; do pypy xxx.py; done

----------
status: resolved -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1264>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 15 21:16:27 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Sat, 15 Sep 2012 19:16:27 +0000
Subject: [pypy-issue] [issue1264] segmentation faults/fatal rpython errors
	with threading +	import
In-Reply-To: <1347694233.09.0.704847414829.issue1264@bugs.pypy.org>
Message-ID: <1347736587.32.0.623451191992.issue1264@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

You also found a race condition in PyEval_SaveThread...
Fixed in 9bf06ba21208.

----------
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1264>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 18 00:58:38 2012
From: tracker at bugs.pypy.org (bdk)
Date: Mon, 17 Sep 2012 22:58:38 +0000
Subject: [pypy-issue] [issue1265] crashes when testing rsocket.py under pypy
In-Reply-To: <1347922718.7.0.490936971262.issue1265@bugs.pypy.org>
Message-ID: <1347922718.7.0.490936971262.issue1265@bugs.pypy.org>


New submission from bdk <bdkearns at gmail.com>:

~/pypy$ ../pypy-c-jit-57366-073896dbe4e3-linux64/bin/pypy test_getaddrinfo.py 
*** glibc detected *** ../pypy-c-jit-57366-073896dbe4e3-linux64/bin/pypy: double 
free or corruption (out): 0x00007f06098a2780 ***
======= Backtrace: =========
/lib/x86_64-linux-gnu/libc.so.6(+0x7e626)[0x7f0607ba5626]
/lib/x86_64-linux-gnu/libc.so.6(freeaddrinfo+0x20)[0x7f0607bf4b50]
/usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call_unix64+0x4c)[0x7f0608769a14]
/usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call+0x1e5)[0x7f0608769435]
../pypy-c-jit-57366-073896dbe4e3-
linux64/bin/pypy(pypy_g_ccall_ffi_call__ffi_cifPtr_arrayPtr_arrayPtr_arr+0x6b)
[0x20f10db]

----------
files: test_getaddrinfo.py
messages: 4763
nosy: bdk, pypy-issue
priority: bug
status: unread
title: crashes when testing rsocket.py under pypy

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1265>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 18 04:30:47 2012
From: tracker at bugs.pypy.org (bdk)
Date: Tue, 18 Sep 2012 02:30:47 +0000
Subject: [pypy-issue] [issue1265] crashes when testing rsocket.py under pypy
In-Reply-To: <1347922718.7.0.490936971262.issue1265@bugs.pypy.org>
Message-ID: <1347935447.73.0.315517439089.issue1265@bugs.pypy.org>


bdk <bdkearns at gmail.com> added the comment:

potential fix here https://bitbucket.org/pypy/pypy/pull-request/87/ but not 
familiar enough with pypy internals to know if it's 100% correct. hoping someone 
can review.

----------
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1265>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 18 16:03:18 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Tue, 18 Sep 2012 14:03:18 +0000
Subject: [pypy-issue] [issue1265] crashes when testing rsocket.py under pypy
In-Reply-To: <1347922718.7.0.490936971262.issue1265@bugs.pypy.org>
Message-ID: <1347976998.11.0.0951090759011.issue1265@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Located the problem as "p.contents" not having a "_b_base_" pointing back to
"p".  Fixed in d3f714e26354.

----------
nosy: +arigo
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1265>
________________________________________

From tracker at bugs.pypy.org  Tue Sep 18 17:29:36 2012
From: tracker at bugs.pypy.org (Ralf Schmitt)
Date: Tue, 18 Sep 2012 15:29:36 +0000
Subject: [pypy-issue] [issue1248] please build with older GNU libc on
	Linux/amd64
In-Reply-To: <1346476259.72.0.54490436525.issue1248@bugs.pypy.org>
Message-ID: <1347982176.93.0.314320451071.issue1248@bugs.pypy.org>


Ralf Schmitt <ralf at systemexit.de> added the comment:

I had great success with using a centos 5 machine as base and installing gentoo
prefix on it: http://www.gentoo.org/proj/en/gentoo-alt/prefix/

I'm distributing my apps with every dynamic library besides what comes with glibc.


I've also used http://nixos.org/patchelf.html in order to adapt the library
search path.

the instructions are only short if you download a ready made image.

----------
nosy: +schmir

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1248>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 19 17:50:04 2012
From: tracker at bugs.pypy.org (Brian)
Date: Wed, 19 Sep 2012 15:50:04 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>


New submission from Brian <thefollowingcanbefound at gmail.com>:

The following code causes Fatal RPython error: IndexError.
d = range(1)
d.pop()
d.append(1)

>From what I can tell the exception is always thrown if a range is popped until empty and then append 
is called with any argument.  I have observed this behavior on both 1.8 and 1.9.  No exception is 
thrown by cPython 2.7.3.  Replacing the first line with "d = list(range(1))" does not prevent the 
exception, but "d = [x for x in range(1)]" does.

Here is the raw output running the attached file (same as code above).

RPython traceback:
  File "translator_goal_targetpypystandalone.c", line 2395, in entry_point
  File "interpreter_pyframe.c", line 984, in PyFrame_execute_frame
  File "jit_metainterp_warmspot.c", line 865, in ll_portal_runner__Unsigned_Bool_pypy_interpreter
  File "module_pypyjit_interp_jit.c", line 145, in portal_2
  File "interpreter_pyopcode.c", line 2594, in handle_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 7365, in dispatch_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 14837, in call_function__AccessDirect_None
  File "interpreter_pyframe.c", line 984, in PyFrame_execute_frame
  File "jit_metainterp_warmspot.c", line 865, in ll_portal_runner__Unsigned_Bool_pypy_interpreter
  File "module_pypyjit_interp_jit.c", line 145, in portal_2
  File "interpreter_pyopcode.c", line 2594, in handle_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 6978, in dispatch_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 14837, in call_function__AccessDirect_None
  File "interpreter_pyframe.c", line 984, in PyFrame_execute_frame
  File "jit_metainterp_warmspot.c", line 865, in ll_portal_runner__Unsigned_Bool_pypy_interpreter
  File "module_pypyjit_interp_jit.c", line 145, in portal_2
  File "interpreter_pyopcode.c", line 2594, in handle_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 8125, in dispatch_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 14837, in call_function__AccessDirect_None
  File "interpreter_pyframe.c", line 984, in PyFrame_execute_frame
  File "jit_metainterp_warmspot.c", line 865, in ll_portal_runner__Unsigned_Bool_pypy_interpreter
  File "module_pypyjit_interp_jit.c", line 145, in portal_2
  File "interpreter_pyopcode.c", line 2594, in handle_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 8018, in dispatch_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 20396, in EXEC_STMT__AccessDirect_None
  File "interpreter_pyframe.c", line 984, in PyFrame_execute_frame
  File "jit_metainterp_warmspot.c", line 865, in ll_portal_runner__Unsigned_Bool_pypy_interpreter
  File "module_pypyjit_interp_jit.c", line 145, in portal_2
  File "interpreter_pyopcode.c", line 2594, in handle_bytecode__AccessDirect_None
  File "interpreter_pyopcode.c", line 9978, in dispatch_bytecode__AccessDirect_None
  File "objspace_std_callmethod.c", line 3506, in CALL_METHOD__AccessDirect_star_1
  File "interpreter_gateway.c", line 1111, in BuiltinCode2_fastcall_2
  File "interpreter_gateway.c", line 3166, in BuiltinCode_handle_exception
Fatal RPython error: IndexError
Abort

----------
files: broken.py
messages: 4767
nosy: brianwgoldman, pypy-issue
priority: bug
release: 1.9
status: unread
title: Popping and appending to range causes Fatal RPython error

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 19 19:32:02 2012
From: tracker at bugs.pypy.org (Philip Jenvey)
Date: Wed, 19 Sep 2012 17:32:02 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348075922.68.0.465047798742.issue1266@bugs.pypy.org>


Philip Jenvey <pjenvey at underboss.org> added the comment:

fixed in efca15c9d78e, thanks

----------
status: unread -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 19 19:45:22 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Wed, 19 Sep 2012 17:45:22 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348076722.34.0.81278235541.issue1266@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

http://bpaste.net/show/46442/

In other words I don't see the point of append(), which is prone to overflow
bugs like this one.  Killing it completely looks the safest option to me.

----------
nosy: +arigo
status: resolved -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 19 19:54:35 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Wed, 19 Sep 2012 17:54:35 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348077275.69.0.946042308101.issue1266@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

More fun, overflowing the multiplication in getslice():
http://bpaste.net/show/46446/

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 20 01:56:59 2012
From: tracker at bugs.pypy.org (Philip Jenvey)
Date: Wed, 19 Sep 2012 23:56:59 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348099019.2.0.200210613534.issue1266@bugs.pypy.org>


Philip Jenvey <pjenvey at underboss.org> added the comment:

The attached patch fixes Armin's bugs and a couple other corner cases in 
reversing operations. I pretty much just removed all of the affected 
optimizations altogether.

getslice maybe seems a little more useful of an optimization than 
append/reverse/sort(True) are, but probably not that useful either. Armin, let 
me know if you'd prefer workarounds for any of these instead, otherwise I'll 
commit as is

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 20 19:45:38 2012
From: tracker at bugs.pypy.org (Ariel Ben-Yehuda)
Date: Thu, 20 Sep 2012 17:45:38 +0000
Subject: [pypy-issue] [issue1262] Implement PEP 3138
In-Reply-To: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>
Message-ID: <1348163138.47.0.782255711694.issue1262@bugs.pypy.org>


Ariel Ben-Yehuda <arielb1 at mail.tau.ac.il> added the comment:

updated patch - is this legal RPython?

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1262>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 20 19:49:44 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Thu, 20 Sep 2012 17:49:44 +0000
Subject: [pypy-issue] [issue1262] Implement PEP 3138
In-Reply-To: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>
Message-ID: <1348163384.41.0.209926973041.issue1262@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Just checking: it seems to me that PEP 3138 is about a change done *only* to
Python 3, not back-ported to Python 2.  Why are you patch not for the py3k
branch of PyPy then?  I may be wrong...

----------
nosy: +arigo

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1262>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 20 19:58:53 2012
From: tracker at bugs.pypy.org (Ariel Ben-Yehuda)
Date: Thu, 20 Sep 2012 17:58:53 +0000
Subject: [pypy-issue] [issue1262] Implement PEP 3138
In-Reply-To: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>
Message-ID: <1348163933.39.0.496798062207.issue1262@bugs.pypy.org>


Ariel Ben-Yehuda <arielb1 at mail.tau.ac.il> added the comment:

The patch is for py3k

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1262>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 20 20:00:41 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Thu, 20 Sep 2012 18:00:41 +0000
Subject: [pypy-issue] [issue1262] Implement PEP 3138
In-Reply-To: <1347451282.6.0.989781872121.issue1262@bugs.pypy.org>
Message-ID: <1348164041.0.0.283791430877.issue1262@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Indeed, sorry.  I though "hg diff" would write the name of the branch somewhere
in the patch, but it doesn't.  Only the revision it's based on can indirectly
point to the branch.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1262>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 20 21:20:03 2012
From: tracker at bugs.pypy.org (Fijal)
Date: Thu, 20 Sep 2012 19:20:03 +0000
Subject: [pypy-issue] [issue1199] psycopg2ct 2.4.4 segmentation fault on
	Ubuntu 12.04 64-bit
In-Reply-To: <1341163597.62.0.605992819875.issue1199@bugs.pypy.org>
Message-ID: <1348168803.32.0.420007734861.issue1199@bugs.pypy.org>


Fijal <fijall at gmail.com> added the comment:

Closing if fixed. PLease reopen if it still appears.

----------
nosy: +fijal
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1199>
________________________________________

From tracker at bugs.pypy.org  Fri Sep 21 10:05:41 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Fri, 21 Sep 2012 08:05:41 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348214741.47.0.747735469506.issue1266@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

It's fine to kill.  After all, for comparison, xrange() objects support very
little of the list interface; I would say that there is no point in doing
optimizations with our range() objects that go beyond what xrange() supports.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Fri Sep 21 19:51:04 2012
From: tracker at bugs.pypy.org (Philip Jenvey)
Date: Fri, 21 Sep 2012 17:51:04 +0000
Subject: [pypy-issue] [issue1266] Popping and appending to range causes
	Fatal RPython error
In-Reply-To: <1348069804.45.0.46260374891.issue1266@bugs.pypy.org>
Message-ID: <1348249864.8.0.867486205726.issue1266@bugs.pypy.org>


Philip Jenvey <pjenvey at underboss.org> added the comment:

committed the further related fixes in ad495c36a765

----------
status: chatting -> resolved

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1266>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 22 22:15:16 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 22 Sep 2012 20:15:16 +0000
Subject: [pypy-issue] [issue1067] Performance on Stack-Based VM machine
In-Reply-To: <1329857569.22.0.328879135116.issue1067@bugs.pypy.org>
Message-ID: <1348344915.94.0.128635205199.issue1067@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Finally found a simple idea (c7c74257157c): when we hit the limit
'max_unroll_loops', then instead of aborting, retry to compile by
disabling the 'unroll' optimization.

This alone fixes go2.py, which runs 235 times faster now than before,
and 8.7 times faster than CPython.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1067>
________________________________________

From tracker at bugs.pypy.org  Sun Sep 23 00:20:49 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 22 Sep 2012 22:20:49 +0000
Subject: [pypy-issue] [issue1253] compile => RuntimeError: maximum recursion
	depth exceeded
In-Reply-To: <1347011043.22.0.852237039186.issue1253@bugs.pypy.org>
Message-ID: <1348352449.51.0.461771779631.issue1253@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Testing, close if it works (417403f0dc1a)

----------
nosy: +arigo
status: unread -> testing

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1253>
________________________________________

From tracker at bugs.pypy.org  Mon Sep 24 12:52:09 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Mon, 24 Sep 2012 10:52:09 +0000
Subject: [pypy-issue] [issue1267] RPython MemoryError in the GC
In-Reply-To: <1348483929.66.0.926922716548.issue1267@bugs.pypy.org>
Message-ID: <1348483929.66.0.926922716548.issue1267@bugs.pypy.org>


New submission from Armin Rigo <armin.rigo at gmail.com>:

I think we don't have this as a bug report, so here it is:

We need to look at when the GC can raise an RPython-level MemoryError and fix
what occurs then.  We'd also need to do something about the situation of getting
out of memory, which right now means unrecoverable death.

http://mail.python.org/pipermail/pypy-dev/2012-August/010394.html

----------
messages: 4781
nosy: arigo, pypy-issue
priority: bug
status: unread
title: RPython MemoryError in the GC

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1267>
________________________________________

From tracker at bugs.pypy.org  Wed Sep 26 14:55:04 2012
From: tracker at bugs.pypy.org (Simon Lundberg)
Date: Wed, 26 Sep 2012 12:55:04 +0000
Subject: [pypy-issue] [issue1268] file.writelines() very slow on Windows
In-Reply-To: <1348664104.59.0.290334033173.issue1268@bugs.pypy.org>
Message-ID: <1348664104.59.0.290334033173.issue1268@bugs.pypy.org>


New submission from Simon Lundberg <simon.lundberg at yahoo.com>:

Using file.writelines() in 32-bit PyPy on Windows seems very slow (40x-60x 
slower than in CPython). Using file.write() with large chunks of data is much 
faster (though still not as fast as in CPython).

Attached is a small function that takes a file path and number of lines in the 
file as arguments. It creates a set of dummy data, and writes it out using 
file.write() and file.writelines(), and prints out the timings. I'm getting the 
following results:

CPython 64-bit:
f.write(): 0.098
f.writelines(): 0.161

CPython 32-bit:
f.write(): 0.127
f.writelines(): 0.205

PyPy 32-bit:
f.write(): 0.297
f.writelines(): 9.47

----------
files: writeTest.py
messages: 4782
nosy: SimonLundberg, pypy-issue
priority: performance bug
status: unread
title: file.writelines() very slow on Windows

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1268>
________________________________________
-------------- next part --------------
import time

def writeTest(path, numLines=1000000):
    t = [time.time()]
    data = '\n'.join(['this is a line of data' for i in range(numLines)])
    t.append(time.time())
    print 'Creating data:  ', t[-1] - t[-2]
    with file(path, 'w') as f: f.write(data)
    t.append(time.time())
    print 'f.write():      ', t[-1] - t[-2]

    data = ['this is a line of data\n' for i in range(numLines)]
    t.append(time.time())
    print 'Creating data:  ', t[-1] - t[-2]
    with file(path, 'w') as f: f.writelines(data)
    t.append(time.time())
    print 'f.writelines(): ', t[-1] - t[-2]

From tracker at bugs.pypy.org  Thu Sep 27 08:55:57 2012
From: tracker at bugs.pypy.org (Alexey Poryadin)
Date: Thu, 27 Sep 2012 06:55:57 +0000
Subject: [pypy-issue] [issue1269] weakref.WeakSet does not work correctly
In-Reply-To: <1348728957.11.0.250940191016.issue1269@bugs.pypy.org>
Message-ID: <1348728957.11.0.250940191016.issue1269@bugs.pypy.org>


New submission from Alexey Poryadin <alexey.poryadin at gmail.com>:

WeakSet works incorrectly, try sample below:

from weakref import WeakSet
class Obj(object): pass

o1 = Obj()
o2 = Obj()
ws = WeakSet([o1, o2])
del o2
assert len(tuple(ws))==1

----------
messages: 4783
nosy: alesh, pypy-issue
priority: bug
status: unread
title: weakref.WeakSet does not work correctly

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1269>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 27 09:41:51 2012
From: tracker at bugs.pypy.org (Amaury Forgeot d Arc)
Date: Thu, 27 Sep 2012 07:41:51 +0000
Subject: [pypy-issue] [issue1269] weakref.WeakSet does not work correctly
In-Reply-To: <1348728957.11.0.250940191016.issue1269@bugs.pypy.org>
Message-ID: <1348731711.59.0.604899306422.issue1269@bugs.pypy.org>


Amaury Forgeot d Arc <amauryfa at gmail.com> added the comment:

This is an assumed difference with CPython; PyPy does not use reference counting to manage 
memory.

See this paragraph about PyPy's garbage collector:
http://doc.pypy.org/en/latest/cpython_differences.html#differences-related-to-garbage-
collection-strategies
"""... It also means that weak references may stay alive for a bit longer than expected."""

You could add a call to gc.collect() to clear weak references, but of course this has a 
large performance impact.

----------
nosy: +afa
status: unread -> wontfix

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1269>
________________________________________

From tracker at bugs.pypy.org  Thu Sep 27 14:19:25 2012
From: tracker at bugs.pypy.org (Alexey Poryadin)
Date: Thu, 27 Sep 2012 12:19:25 +0000
Subject: [pypy-issue] [issue1269] weakref.WeakSet does not work correctly
In-Reply-To: <1348728957.11.0.250940191016.issue1269@bugs.pypy.org>
Message-ID: <1348748365.28.0.857349503561.issue1269@bugs.pypy.org>


Alexey Poryadin <alexey.poryadin at gmail.com> added the comment:

I understand, but saddened :) This difference would not allow to port of certain 
programs from CPython. Programs that used weakref usually expects what weak 
references becoming dead immediately.

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1269>
________________________________________

From tracker at bugs.pypy.org  Fri Sep 28 01:58:23 2012
From: tracker at bugs.pypy.org (Antony Lee)
Date: Thu, 27 Sep 2012 23:58:23 +0000
Subject: [pypy-issue] [issue1270] ctypes structure should allow unicode
	field names
In-Reply-To: <1348790303.3.0.849012689351.issue1270@bugs.pypy.org>
Message-ID: <1348790303.3.0.849012689351.issue1270@bugs.pypy.org>


New submission from Antony Lee <anntzer.lee at gmail.com>:

from ctypes import *
class C(Structure): _fields_ = [(u"field", c_int)]

fails with pypy (TypeError: cannot add non-string keys to dict of a type), but
works with cpython2.  The bug is in structure.py, line 71:

self.__dict__.update(fields)

which should be replaced by

for k, v in fields: setattr(self, k, v)

because setattr does (I think?) an implicit unicode->bytes conversion.

----------
messages: 4786
nosy: antony, pypy-issue
priority: bug
status: unread
title: ctypes structure should allow unicode field names

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1270>
________________________________________

From tracker at bugs.pypy.org  Fri Sep 28 20:46:32 2012
From: tracker at bugs.pypy.org (Ronny Pfannschmidt)
Date: Fri, 28 Sep 2012 18:46:32 +0000
Subject: [pypy-issue] [issue1271] nightly requires a libffi shared lib thats
	newer than	what debian has
In-Reply-To: <1348857992.41.0.377657301298.issue1271@bugs.pypy.org>
Message-ID: <1348857992.41.0.377657301298.issue1271@bugs.pypy.org>


New submission from Ronny Pfannschmidt <Ronny.Pfannschmidt at gmx.de>:

$ pypy-bin               
/home/ronny/Projects/pypy/pypy-bin/bin/pypy: error while loading shared
libraries: libffi.so.6: cannot open shared object file: No such file or directory

debian only has libffi5

----------
messages: 4787
nosy: pypy-issue, ronny
priority: bug
status: unread
title: nightly requires a libffi shared lib thats newer than what debian has

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1271>
________________________________________

From tracker at bugs.pypy.org  Fri Sep 28 20:49:34 2012
From: tracker at bugs.pypy.org (Ronny Pfannschmidt)
Date: Fri, 28 Sep 2012 18:49:34 +0000
Subject: [pypy-issue] [issue1271] nightly requires a libffi shared lib thats
	newer than	what debian has
In-Reply-To: <1348857992.41.0.377657301298.issue1271@bugs.pypy.org>
Message-ID: <1348858174.4.0.696849110208.issue1271@bugs.pypy.org>


Ronny Pfannschmidt <Ronny.Pfannschmidt at gmx.de> added the comment:

extra note, even when being evil and copying libffi.so.5 in that place it fails
wrt glibc version

LD_LIBRARY_PATH=lib pypy-bin
/home/ronny/Projects/pypy/pypy-bin/bin/pypy: /lib/x86_64-linux-gnu/libc.so.6:
version `GLIBC_2.14' not found (required by
/home/ronny/Projects/pypy/pypy-bin/bin/pypy)
/home/ronny/Projects/pypy/pypy-bin/bin/pypy: /lib/x86_64-linux-gnu/libc.so.6:
version `GLIBC_2.15' not found (required by
/home/ronny/Projects/pypy/pypy-bin/bin/pypy)

----------
status: unread -> chatting

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1271>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 29 11:22:42 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sat, 29 Sep 2012 09:22:42 +0000
Subject: [pypy-issue] [issue1271] nightly requires a libffi shared lib thats
	newer than	what debian has
In-Reply-To: <1348857992.41.0.377657301298.issue1271@bugs.pypy.org>
Message-ID: <1348910562.12.0.808834968036.issue1271@bugs.pypy.org>


Armin Rigo <armin.rigo at gmail.com> added the comment:

Looks like a duplicate of Issue1248.

----------
nosy: +arigo

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1271>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 29 12:00:04 2012
From: tracker at bugs.pypy.org (Fijal)
Date: Sat, 29 Sep 2012 10:00:04 +0000
Subject: [pypy-issue] [issue1248] please build with older GNU libc on
	Linux/amd64
In-Reply-To: <1346476259.72.0.54490436525.issue1248@bugs.pypy.org>
Message-ID: <1348912804.0.0.792513665009.issue1248@bugs.pypy.org>


Fijal <fijall at gmail.com> added the comment:

Just for some information

The Ubuntu we use has a standard libc (whatever comes with ubuntu) and a new gcc 
(4.7). Everything else is completely standard

----------
nosy: +fijal

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1248>
________________________________________

From tracker at bugs.pypy.org  Sat Sep 29 22:14:59 2012
From: tracker at bugs.pypy.org (Marcus Smith)
Date: Sat, 29 Sep 2012 20:14:59 +0000
Subject: [pypy-issue] [issue1272] user and home install schemes broke in pypy
In-Reply-To: <1348949699.1.0.101888055363.issue1272@bugs.pypy.org>
Message-ID: <1348949699.1.0.101888055363.issue1272@bugs.pypy.org>


New submission from Marcus Smith <qwcode at gmail.com>:

the user and home installation schemes are broke in pypy 1.9 due it's custom 
"pypy" installation scheme.

user scheme: http://docs.python.org/dev/distutils/install.html#alternate-
installation-the-user-scheme

home scheme: http://docs.python.org/dev/distutils/install.html#alternate-
installation-the-home-scheme

there's a wake of issues/workarounds in pip and virtualenv related to the "pypy" 
scheme:
-https://github.com/pypa/pip/issues/629
-https://github.com/pypa/pip/issues/624
-https://github.com/pypa/pip/issues/625
-https://github.com/pypa/virtualenv/issues/306
-https://github.com/pypa/virtualenv/issues/302

I created a diff that shows the difference for distutils.command.install between 
py2.7 and pypy1.9, and discuss options.

https://github.com/qwcode/pypy_scheme/commit/b7856156f302c192698cd311df814dfd3bf
7a8a0

----------
messages: 4791
nosy: pypy-issue, qwcode
priority: bug
release: 1.9
status: unread
title: user and home install schemes broke in pypy

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1272>
________________________________________

From tracker at bugs.pypy.org  Sun Sep 30 15:59:09 2012
From: tracker at bugs.pypy.org (Armin Rigo)
Date: Sun, 30 Sep 2012 13:59:09 +0000
Subject: [pypy-issue] [issue1273] test_multiprocessing is skipped
In-Reply-To: <1349013548.68.0.193914332397.issue1273@bugs.pypy.org>
Message-ID: <1349013548.68.0.193914332397.issue1273@bugs.pypy.org>


New submission from Armin Rigo <armin.rigo at gmail.com>:

test_multiprocessing.py is skipped, and if we enable it it hangs forever on
Linux.  Investigate.

----------
messages: 4792
nosy: arigo, pypy-issue
priority: bug
status: unread
title: test_multiprocessing is skipped

________________________________________
PyPy bug tracker <tracker at bugs.pypy.org>
<https://bugs.pypy.org/issue1273>
________________________________________