From pypy-dev-issue at codespeak.net  Fri Sep 10 00:08:35 2010
From: pypy-dev-issue at codespeak.net (Amaury Forgeot d Arc)
Date: Thu, 09 Sep 2010 22:08:35 +0000
Subject: [PyPy-issue] [issue554] main thread not calling
	threading._shutdown() when exiting
Message-ID: <1284070115.09.0.175390423373.issue554@>


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

Fixed with r76841.

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

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue554>
_______________________________________________________


From pypy-dev-issue at codespeak.net  Fri Sep 24 06:07:39 2010
From: pypy-dev-issue at codespeak.net (Zeev Tarantov)
Date: Fri, 24 Sep 2010 04:07:39 +0000
Subject: [PyPy-issue] [issue559] pypy jit doesn't build on 64-bit machine
Message-ID: <1285301259.45.0.674221545631.issue559@>


New submission from Zeev Tarantov <zeev.tarantov at gmail.com>:

pypy trunk revision 77317
command line python translate.py --opt=jit targetpypystandalone.py
cpython 2.6.5
gcc gcc version 4.5.2-pre rev. 164254
Gentoo Linux ~amd64 (64bit unstable)

looks like it coldn't compile any of the C code.

[translation:ERROR] Error:
[translation:ERROR]  Traceback (most recent call last):
[translation:ERROR]    File "translate.py", line 289, in main
[translation:ERROR]     drv.proceed(goals)
[translation:ERROR]    File "/home/wolf/pypy-trunk/pypy/translator/driver.py",
line 800, in proceed
[translation:ERROR]     return self._execute(goals, task_skip = self._maybe_skip())
[translation:ERROR]    File
"/home/wolf/pypy-trunk/pypy/translator/tool/taskengine.py", line 116, in _execute
[translation:ERROR]     res = self._do(goal, taskcallable, *args, **kwds)
[translation:ERROR]    File "/home/wolf/pypy-trunk/pypy/translator/driver.py",
line 293, in _do
[translation:ERROR]     res = func()
[translation:ERROR]    File "/home/wolf/pypy-trunk/pypy/translator/driver.py",
line 563, in task_compile_c
[translation:ERROR]     cbuilder.compile(**kwds)
[translation:ERROR]    File "/home/wolf/pypy-trunk/pypy/translator/c/genc.py",
line 517, in compile
[translation:ERROR]     extra_opts)
[translation:ERROR]    File
"/home/wolf/pypy-trunk/pypy/translator/platform/posix.py", line 183, in
execute_makefile
[translation:ERROR]     self._handle_error(returncode, stdout, stderr,
path.join('make'))
[translation:ERROR]    File
"/home/wolf/pypy-trunk/pypy/translator/platform/__init__.py", line 127, in
_handle_error
[translation:ERROR]     raise CompilationError(stdout, stderr)
[translation:ERROR]  CompilationError: <CompilationError err="make: Entering
directory `/tmp/usession-trunk-0/testing_1'\ngcc -O3 -pthread
-fomit-frame-pointer -Wall -Wno-unused -...es_6.s nonfuncnodes_30.s
nonfuncnodes_14.s implement_20.s\nmake: Leaving directory
`/tmp/usession-trunk-0/testing_1'\n">

build log compressed with bzip2 attached

----------
effort: ???
files: build.log.bz2
messages: 1834
nosy: pypy-issue, t_zeev
priority: bug
release: ???
status: unread
title: pypy jit doesn't build on 64-bit machine

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue559>
_______________________________________________________
-------------- next part --------------
A non-text attachment was scrubbed...
Name: build.log.bz2
Type: application/x-bzip2
Size: 16244 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/pypy-issue/attachments/20100924/824b9343/attachment.bin>

From pypy-dev-issue at codespeak.net  Fri Sep 24 10:05:32 2010
From: pypy-dev-issue at codespeak.net (Amaury Forgeot d Arc)
Date: Fri, 24 Sep 2010 08:05:32 +0000
Subject: [PyPy-issue] [issue559] pypy jit doesn't build on 64-bit machine
Message-ID: <1285315532.34.0.658046566806.issue559@>


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

The failure is in trackgcroot.py, probably an assembler instruction we did not 
encounter before.
Can you join the (compressed!) file /tmp/usession-trunk-0/testing_1/implement_51.s, so we can figure it out?

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

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue559>
_______________________________________________________


From pypy-dev-issue at codespeak.net  Fri Sep 24 18:10:16 2010
From: pypy-dev-issue at codespeak.net (Zeev Tarantov)
Date: Fri, 24 Sep 2010 16:10:16 +0000
Subject: [PyPy-issue] [issue559] pypy jit doesn't build on 64-bit machine
Message-ID: <1285344616.35.0.947856303595.issue559@>


Zeev Tarantov <zeev.tarantov at gmail.com> added the comment:

That file did not exist, but I looked at the build log and found the command
line that produces it, ran it and got you the file. Attached in bzip2.

That "[platform:ERROR]" next to the compilation command line looks alarming.
Actually, the build log has lots of exceptions and errors. I just don't know
which are expected.

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue559>
_______________________________________________________
-------------- next part --------------
A non-text attachment was scrubbed...
Name: implement_51.s.bz2
Type: application/x-bzip2
Size: 95739 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/pypy-issue/attachments/20100924/0a30ef62/attachment.bin>

From pypy-dev-issue at codespeak.net  Fri Sep 24 18:18:13 2010
From: pypy-dev-issue at codespeak.net (Amaury Forgeot d Arc)
Date: Fri, 24 Sep 2010 16:18:13 +0000
Subject: [PyPy-issue] [issue559] pypy jit doesn't build on 64-bit machine
Message-ID: <1285345093.02.0.965287294908.issue559@>


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

OK, I just submitted a fix (xchgq was not handled by the 64bit compiler)

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue559>
_______________________________________________________


From pypy-dev-issue at codespeak.net  Fri Sep 24 21:05:48 2010
From: pypy-dev-issue at codespeak.net (Zeev Tarantov)
Date: Fri, 24 Sep 2010 19:05:48 +0000
Subject: [PyPy-issue] [issue559] pypy jit doesn't build on 64-bit machine
Message-ID: <1285355148.29.0.273287883078.issue559@>


Zeev Tarantov <zeev.tarantov at gmail.com> added the comment:

Awesome response time!

[Timer] Timings:
[Timer] annotate                       --- 1488.9 s
[Timer] rtype_lltype                   --- 1029.4 s
[Timer] pyjitpl_lltype                 --- 1634.5 s
[Timer] backendopt_lltype              ---  759.5 s
[Timer] stackcheckinsertion_lltype     ---  114.7 s
[Timer] database_c                     --- 1212.6 s
[Timer] source_c                       --- 1614.4 s
[Timer] compile_c                      ---  808.3 s
[Timer] ===========================================
[Timer] Total:                         --- 8662.2 s

And the resulting pypy-c appears to work. You can close this ticket as "fixed".

It outputs some stuff to stderr like this:

Tracing:        116     0.513629
Backend:        114     0.082529
Running asm:            30985
Blackhole:              18402
TOTAL:                  7.233677
ops:                    183617
recorded ops:           87669
  calls:                2637
guards:                 48225
opt ops:                13079
opt guards:             2936
forcings:               711
abort: trace too long:  2
abort: compiling:       0
abort: vable escape:    0
nvirtuals:              22644
nvholes:                2547
nvreused:               13160
Is there documentation on how to interpret this information?

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue559>
_______________________________________________________


From pypy-dev-issue at codespeak.net  Fri Sep 24 22:05:46 2010
From: pypy-dev-issue at codespeak.net (Amaury Forgeot d Arc)
Date: Fri, 24 Sep 2010 20:05:46 +0000
Subject: [PyPy-issue] [issue559] pypy jit doesn't build on 64-bit machine
Message-ID: <1285358746.34.0.645973730425.issue559@>


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

I don't know the exact meaning of all these figures, except that the overhead 
caused by the tracing JIT was not too high: around 0.59s for a total of 7s.

I suggest you to connect to the #pypy channel and bug people there to improve the 
documentation!

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

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue559>
_______________________________________________________


From pypy-dev-issue at codespeak.net  Thu Sep 30 21:15:25 2010
From: pypy-dev-issue at codespeak.net (Alex Gaynor)
Date: Thu, 30 Sep 2010 19:15:25 +0000
Subject: [PyPy-issue] [issue560] Possible to segfault pypy using
	_pickle_support
Message-ID: <1285874125.67.0.671323190959.issue560@>


New submission from Alex Gaynor <alex.gaynor at gmail.com>:

The following segfaults pypy:

alex at alex-laptop:~/projects/pypy$ ./pypy-c 
Python 2.5.2 (77472, Sep 29 2010, 17:21:04)
[PyPy 1.3.0] on linux2
Type "help", "copyright", "credits" or "license" for more information.
And now for something completely different: ``we believe in pancakes, maybe''
>>>> import _pickle_support
>>>> _pickle_support.method_new(None, None)
Segmentation fault (core dumped)

----------
effort: ???
messages: 1840
nosy: agaynor, pypy-issue
priority: bug
release: ???
status: unread
title: Possible to segfault pypy using _pickle_support

_______________________________________________________
PyPy development tracker <pypy-dev-issue at codespeak.net>
<https://codespeak.net/issue/pypy-dev/issue560>
_______________________________________________________