#pypy on 2019-08-08 — irc logs at freenode.irclog.whitequark.org

2018-11-25 07:18 arigato changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end ) | use cffi for calling C | mac OS and Fedora are not Windows

00:07 xcm is now known as Guest2408

00:07 Guest2408 has quit [Read error: Connection reset by peer]

00:07 dansan_ has joined #pypy

00:08 xcm has joined #pypy

00:10 dansan has quit [Ping timeout: 272 seconds]

00:11 speeder39_ has quit [Quit: Connection closed for inactivity]

00:17 dansan has joined #pypy

00:17 dansan_ has quit [Ping timeout: 246 seconds]

00:26 jcea has joined #pypy

00:30 jcea has quit [Remote host closed the connection]

00:45 jcea has joined #pypy

00:47 jcea has quit [Remote host closed the connection]

00:52 travis-ci has joined #pypy

00:52 <travis-ci> timpalpant/vmprof-python@flush-buffers: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122364550)

00:52 travis-ci has left #pypy [#pypy]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4747 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/6572 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/5350 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/4522 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/6359 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/7607 [py3.6]

01:02 IRC-Source_74 has joined #pypy

01:03 IRC-Source_74 has quit [Remote host closed the connection]

01:04 jcea has joined #pypy

01:18 dansan has quit [Remote host closed the connection]

01:18 dansan has joined #pypy

01:24 travis-ci has joined #pypy

01:24 travis-ci has left #pypy [#pypy]

01:24 <travis-ci> timpalpant/vmprof-python@master: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122364623)

01:45 travis-ci has joined #pypy

01:45 <travis-ci> timpalpant/vmprof-python@flush-buffers: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122365663)

01:45 travis-ci has left #pypy [#pypy]

01:53 travis-ci has joined #pypy

01:53 <travis-ci> timpalpant/vmprof-python@master: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122364670)

01:53 travis-ci has left #pypy [#pypy]

02:05 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/5350 [py3.6]

02:20 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/6359 [py3.6]

02:30 travis-ci has joined #pypy

02:30 <travis-ci> timpalpant/vmprof-python@master: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122365784)

02:30 travis-ci has left #pypy [#pypy]

02:35 jcea has quit [Quit: jcea]

02:36 travis-ci has joined #pypy

02:36 <travis-ci> timpalpant/vmprof-python@flush-buffers: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122368017)

02:36 travis-ci has left #pypy [#pypy]

02:37 dustinm has quit [Quit: Leaving]

02:37 <mattip> fijal: in the aarch64 instance, there is a venv/py directory. Is this the python to use for the buildbot builds?

02:39 dustinm has joined #pypy

02:44 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4746 [default]

02:56 dustinm- has joined #pypy

02:56 dustinm has quit [Ping timeout: 246 seconds]

03:06 lritter has quit [Ping timeout: 246 seconds]

03:07 lritter has joined #pypy

03:17 _whitelogger has joined #pypy

03:25 <kenaan> mattip buildbot 6a9387c1c1ad /bot2/pypybuildbot/: add aarch64 slave

03:45 <bbot2> Exception: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/6572 [py3.6]

03:45 <bbot2> Exception: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/7607 [py3.6]

03:45 <bbot2> Exception: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/1115 [default]

03:46 <bbot2> Exception: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/4522 [py3.6]

03:46 <bbot2> Exception: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4747 [py3.6]

03:46 bbot2 has quit [Quit: buildmaster reconfigured: bot disconnecting]

03:47 bbot2 has joined #pypy

03:48 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-aarch64/builds/0 [mattip: force build]

03:56 dddddd has quit [Remote host closed the connection]

04:16 speeder39_ has joined #pypy

04:37 lritter has quit [Quit: Leaving]

05:00 <bbot2> Started: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/2681

05:28 asmeurer__ has joined #pypy

05:46 asmeurer__ has quit [Quit: asmeurer__]

05:50 <runciter> can i debug an issue in rlib, e.g. in rsocket.py with pyinteractive.py?

05:53 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-aarch64/builds/0 [mattip: force build]

06:23 <phlebas> runciter:

06:23 <phlebas> runciter: yes

06:23 <phlebas> some issues, at least

06:24 <runciter> phlebas, hi! so it complains about `_socket` not existing - how would i get that?

06:24 <phlebas> that should be part of your python

06:24 <phlebas> what are you running pyinteractive.py with?

06:24 <runciter> oh, duh, i see

06:25 <runciter> phlebas, a python 2; unfortunately i want to muck with rsocket's implementation of _socket

06:25 <phlebas> using python2 is correct

06:32 <runciter> phlebas, context: https://bitbucket.org/pypy/pypy/issues/3049/systemerror-in-rsocket__setblocking-travis#comment-53329535

06:47 <arigato> runciter: ah, great that we have the minimal reproducer

06:48 <arigato> runciter: the next step is to try to write it as a test in rpython/rlib/test/test_rsocket.py

06:49 <arigato> note that the obscure error message we get on pypy means "an exception was raised in RPython, but it should be caught and turned into a proper Python-level exception by pypy somewhere"

06:50 <arigato> ...which actually means that the problem is likely a missing "except" statement in pypy/module/_socket/

06:50 <arigato> so the test should be written in pypy/module/_socket/test/, sorry

06:52 <arigato> runciter: to answer your question directly, you need: "./pyinteractive.py --withmod-_socket"

06:55 <arigato> interp_socket.py is just missing "try: ... except SocketError as e: raise converted_error(space, e)"

06:55 <arigato> around the call to self.sock.settimeout()

06:55 <arigato> we should actually review the whole interp_socket.py. For example I see a call to self.sock.connect_ex() that is not protected either

06:55 <arigato> self.sock.getsockopt()

06:56 <arigato> self.sock.gettimeout()

06:56 <arigato> self.sock.setblocking()

07:01 <arigato> I have no idea who raises the EBADF exception when we try to same example in pypy2 with socket.socket() instead of socketpair()

07:03 <arigato> OK, it's some obscure wrapper done by the pure-python socket.py from the stdlib

07:03 <arigato> I have no clue why socket.socket() returns this wrapper but not socket.socketpair()

07:03 <arigato> but that's a question for CPython

07:08 <arigato> a final note: CPython 2 never raises anything in setblocking() or settimeout() calls, for some reason

07:16 speeder39_ has quit [Quit: Connection closed for inactivity]

07:16 <bbot2> Success: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/2681

07:19 <kenaan> arigo py3.6 5a152ed4712a /pypy/interpreter/pyparser/test/unittest_samples.py: hg merge default

07:24 <arigato> runciter: I think I fixed it by adding the "try: except SocketError" around the few missing places in pypy/module/_socket/interp_socket

07:25 <arigato> thanks for trying to fix it too, and sorry about that---as usual, there are many levels in pypy and the real problem is finding the correct one

07:25 <arigato> once it's found the fix is trivial (at least in cases like this one)

07:37 <kenaan> arigo default 8819acbf633a /pypy/module/_socket/: Issue #3049 There were a missing "except SocketError" converting interp-level exceptions to app-level ones---or, a...

07:37 <kenaan> arigo py3.6 d84524cf112f /pypy/module/_socket/: hg merge default

09:26 _whitelogger has joined #pypy

09:27 <ionelmc> arigato: around?

09:52 <kenaan> mattip buildbot a36c3b0d16d1 /bot2/pypybuildbot/: add category for aarch64

09:53 bbot2 has quit [Quit: buildmaster reconfigured: bot disconnecting]

09:53 bbot2 has joined #pypy

09:56 <mattip> the buildbot summary page http://buildbot.pypy.org/summary?branch=%3Ctrunk%3E now sorts by some priority, and shows aarch64 category

10:05 xcm has quit [Remote host closed the connection]

10:07 xcm has joined #pypy

10:14 <mattip> ronan: running "python pytest.py -A" on py3.6 complains about a missing -D option, what is that option and how to use it?

10:18 <mattip> ahh, I need both -A and -D

10:18 <ronan> mattip: no you can't have both

10:18 <ronan> mattip: what are you doing exactly?

10:18 <mattip> how do I now run tests using cpython3?

10:19 <ronan> for the -A tests, you run them with python 2

10:19 <mattip> this complains

10:19 <mattip> python2 pytest.py -A pypy/objspace/std/test/test_stringformat.py

10:19 <mattip> this succeeds

10:19 <mattip> python2 pytest.py -A -D pypy/objspace/std/test/test_stringformat.py

10:20 <mattip> this obviously comlains too

10:20 <mattip> python3 pytest.py -D pypy/objspace/std/test/test_stringformat.py

10:21 <ronan> -D is for running the new apptest_xxx.py files

10:21 <ronan> "python2 pytest.py -A pypy/objspace/std/test/test_stringformat.py" works for me

10:22 <mattip> on py3.6 ?

10:22 <ronan> yes

10:23 <mattip> ok, must have been some stale files, after cleaning pyc it works, thanks

10:53 <kenaan> mattip stdlib-2.7.16 5e792e528dc7 /: merge default into branch

10:53 <kenaan> mattip py3.6 e5ea7d6f832f /pypy/objspace/std/: test, fix formatting with a memoryview

10:54 marky1991 has joined #pypy

11:02 <mattip> ronan: do you have an opinion on releasing PyPy8 with 3.6.1 or 3.6.9?

11:02 <mattip> It seems there are quite a few test failures.

11:02 <mattip> should we wait with the release or release with the older stdlib?

11:05 dddddd has joined #pypy

11:59 Masklinn has joined #pypy

12:15 rindolf has joined #pypy

12:16 <rindolf> Hi all! Turns out my numpy problem affected cpython3 as well

12:20 xcm has quit [Remote host closed the connection]

12:22 xcm has joined #pypy

12:42 marky1991 has quit [Read error: Connection reset by peer]

12:43 marky1991 has joined #pypy

12:57 <ronan> mattip: I'm not sure, most of the new failures seem to be new tests for preexisting issues

12:57 <mattip> ok, makes sense to wait then

12:58 <mattip> rindolf: hi. Not really on topic here, but why do you need to use a poorly supported distro of linux?

12:59 <ronan> mattip: well, the one thing against using 3.6.9 is that it kinda forces updating _ssl

12:59 <ronan> mattip: BTW, why do you want to call it pypy 8?

13:00 <mattip> changes in the C-API mean c-extensions must be recompiled

13:01 <ronan> but the so extension changes with the minor number as well

13:01 <mattip> updating _ssl is tedious but probably is more secure

13:02 <mattip> well, the purists would claim that backward incompatible changes in the API force a major version increase

13:02 <mattip> where if the API is only extended, we could do a minor version increase

13:03 <ronan> OK, that makes sense

13:03 <ronan> what changed?

13:06 xcm has quit [Remote host closed the connection]

13:08 xcm has joined #pypy

13:09 Masklinn has quit []

13:09 <mattip> cdatetime and PyDateTime_CAPI gained two new methods, that changed the size of the struct

13:10 <mattip> I guess it could be construed as backward compatible since the new methods were added to the end, but ...

13:21 <ronan> mattip: right. TBH, I have no clue whether there were other backwards-incompatible changes to pypy3 or not

13:29 <rindolf> mattip: i'm a mageia contributor

13:29 <rindolf> mattip: and this is the dev branch

13:29 dansan_ is now known as dansan

13:39 jcea has joined #pypy

13:42 <kenaan> rlamy __debug__-optimize 7d0654888c47 /: Close branch __debug__-optimize

13:43 <kenaan> rlamy py3.6 4a80ca49fc90 /: Merged in __debug__-optimize (pull request #662) Fix handling of __debug__ and sys.flags.optimize

13:46 <arigato> ionelmc: now yes

13:50 <mattip> ronan: any ideas about the persistent import errors on own-linux-x86-64 tests?

13:50 <mattip> http://buildbot.pypy.org/summary?builder=own-linux-x86-64&branch=py3.6

13:51 <ronan> mattip: nope

13:55 xcm has quit [Remote host closed the connection]

13:59 xcm has joined #pypy

14:16 <kenaan> rlamy py3.6 8c882c594b3f /pypy/doc/whatsnew-pypy3-head.rst: document branch

14:32 xcm has quit [Remote host closed the connection]

14:34 xcm has joined #pypy

14:35 travis-ci has joined #pypy

14:35 travis-ci has left #pypy [#pypy]

14:35 <travis-ci> timpalpant/vmprof-python@flush-buffers: The build passed. (https://travis-ci.com/timpalpant/vmprof-python/builds/122450031)

14:48 travis-ci has joined #pypy

14:48 <travis-ci> timpalpant/vmprof-python@master: The build failed. (https://travis-ci.com/timpalpant/vmprof-python/builds/122450451)

14:48 travis-ci has left #pypy [#pypy]

14:52 Rhy0lite has joined #pypy

14:59 xcm has quit [Remote host closed the connection]

15:00 <kenaan> rlamy py3.6 8b8aaf492f23 /pypy/module/_io/: Fix bpo-25862

15:00 xcm has joined #pypy

15:06 rubdos has quit [Quit: WeeChat 2.4]

15:07 marky1991 has quit [Ping timeout: 245 seconds]

15:07 jcea has quit [Quit: jcea]

15:14 marky1991 has joined #pypy

15:16 <kenaan> arigo sandbox-2 8655e95a8718 /rpython/translator/: When requested, dump compilation information, which the controller process can use to check for bad versions or m...

15:18 travis-ci has joined #pypy

15:18 <travis-ci> timpalpant/vmprof-python@broadcast-sigprof: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122458868)

15:18 travis-ci has left #pypy [#pypy]

15:28 <kenaan> rlamy default 5a0b5a0945e0 /pypy/objspace/std/test/: Move app-level tests to apptest_longobject.py

15:47 rubdos has joined #pypy

15:47 <kenaan> rlamy py3.6 d6fbbd74e9d4 /pypy/objspace/std/test/: Move app-level tests to apptest_longobject.py

15:49 travis-ci has joined #pypy

15:49 travis-ci has left #pypy [#pypy]

15:49 <travis-ci> timpalpant/vmprof-python@broadcast-sigprof: The build failed. (https://travis-ci.com/timpalpant/vmprof-python/builds/122459972)

15:50 BPL has joined #pypy

15:58 travis-ci has joined #pypy

15:58 <travis-ci> timpalpant/vmprof-python@master: The build has errored. (https://travis-ci.com/timpalpant/vmprof-python/builds/122460107)

15:58 travis-ci has left #pypy [#pypy]

16:12 marky1991 has quit [Ping timeout: 246 seconds]

16:35 travis-ci has joined #pypy

16:35 travis-ci has left #pypy [#pypy]

16:35 <travis-ci> timpalpant/vmprof-python@master: The build is still failing. (https://travis-ci.com/timpalpant/vmprof-python/builds/122463975)

17:03 <kenaan> mattip py3.6 c85847cc3955 /pypy/module/atexit/: test, fix raising SystemExit in atexit

17:18 marky1991 has joined #pypy

17:31 marky1991 has quit [Remote host closed the connection]

17:32 marky1991 has joined #pypy

17:32 asmeurer has joined #pypy

17:32 marky1991 has quit [Remote host closed the connection]

17:32 marky1991 has joined #pypy

17:36 asmeurer has quit [Client Quit]

17:48 jcea has joined #pypy

18:09 <mattip> the module.imp.test.test_import.py failures on own also fail when run with -A, so I think the tests are wrong

19:00 asmeurer has joined #pypy

19:03 Rhy0lite has quit [Quit: Leaving]

19:04 asmeurer has quit [Ping timeout: 248 seconds]

19:33 <kenaan> mattip py3.6 29d097325970 /pypy/module/__pypy__/test/test_signal.py: try to debug persistent buildbot failure that does not replicate locally

20:08 jcea has quit [Remote host closed the connection]

20:08 jcea has joined #pypy

20:18 dansan has quit [Excess Flood]

20:19 dansan has joined #pypy

20:20 dansan has quit [Read error: Connection reset by peer]

20:28 xcm has quit [Remote host closed the connection]

20:29 xcm has joined #pypy

20:44 dansan has joined #pypy

20:46 dansan_ has joined #pypy

20:51 dansan has quit [Ping timeout: 276 seconds]

21:23 xcm has quit [Killed (hitchcock.freenode.net (Nickname regained by services))]

21:25 xcm has joined #pypy