#pypy on 2018-07-03 — irc logs at freenode.irclog.whitequark.org

2018-02-26 15:52 cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://botbot.me/freenode/pypy/ ) | use cffi for calling C | the secret reason for us trying to get PyPy users: to test the JIT well enough that we're somewhat confident about it

00:15 <__pv> there's a --source flag or something like it in the pypy translation pass that outputs a set of .c sources and a Makefile

00:16 adamholmberg has joined #pypy

00:21 adamholmberg has quit [Ping timeout: 256 seconds]

00:22 asmeurer_ has joined #pypy

00:38 inhahe_ has quit []

00:41 inhahe has joined #pypy

00:53 asmeurer_ has quit [Quit: asmeurer_]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/build-pypy-c-jit-linux-armel/builds/2350 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/build-pypy-c-jit-linux-armhf-raspbian/builds/2111 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/3856 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/5562 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/6837 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/3992 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/5940 [py3.5]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/4707 [py3.5]

01:07 dddddd has quit [Remote host closed the connection]

01:16 asmeurer has joined #pypy

02:03 inhahe has quit []

02:04 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/4707 [py3.5]

02:05 <bbot2> Success: http://buildbot.pypy.org/builders/build-pypy-c-jit-linux-armel/builds/2350 [py3.5]

02:11 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/5562 [py3.5]

02:13 <bbot2> Success: http://buildbot.pypy.org/builders/build-pypy-c-jit-linux-armhf-raspbian/builds/2111 [py3.5]

02:13 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-armhf-raspbian/builds/1802 [py3.5]

02:16 gavinc has quit [Quit: Konversation terminated!]

02:16 DIRT has quit [Quit: Leaving]

02:18 adamholmberg has joined #pypy

02:19 jsza has quit [Quit: Connection closed for inactivity]

02:20 jcea has quit [Quit: jcea]

02:22 adamholmberg has quit [Ping timeout: 240 seconds]

02:23 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/6837 [py3.5]

02:43 jsza has joined #pypy

03:19 adamholmberg has joined #pypy

03:21 DIRT has joined #pypy

03:23 adamholmberg has quit [Ping timeout: 260 seconds]

04:15 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/5940 [py3.5]

04:31 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/3856 [py3.5]

04:33 wleslie has joined #pypy

04:41 wleslie has quit [Quit: ~~~ Crash in JIT!]

05:16 cjwelborn has quit [Remote host closed the connection]

05:20 adamholmberg has joined #pypy

05:25 adamholmberg has quit [Ping timeout: 260 seconds]

05:26 cjwelborn has joined #pypy

06:09 maxxam has joined #pypy

06:10 <maxxam> ok, so I’m trying to troubleshoot a PyPy3.5 segfault

06:11 <maxxam> is there a list of known-unstable imports somewhere?

06:11 <maxxam> (cfbolz previously mentioned multiprocessing isn’t great)

06:11 <maxxam> or should I just paste a list of my imports here and you guys tell me if it contains any land mines...

06:24 <njs> maxxam: I'm pretty sure they'd like a bug report regardless. The ideal is if you can provide a reproducer, and minimize the reproducer (e.g. start chopping out pieces of the reproducer until the segfault goes away, then put that bit back, repeat until you can't remove anything without removing the segfault).

06:24 tayfun26 has joined #pypy

06:24 <maxxam> njs: this is a bug that crops up in production only, after 48h+ of continuous running

06:26 <maxxam> njs: when I discussed this with cfbolz, the only realistic troubleshooting option involved compiling pypy with debug symbols… which is a pain for a variety of reasons even beyond the obvious

06:27 <maxxam> njs: so I’m first trying fixes like „remove the detritus imports“ which have incidentally led to significant improvements in performance

06:27 <maxxam> (as a result of rewriting sections of code that were just awful)

06:41 <maxxam> „dowsing for segfaults“

06:43 <maxxam> question… would PyPy ever speculatively look into code paths that could never be executed? e.g if I have a return statement before `import pdb; pdb.set_trace()`

07:10 realitix has joined #pypy

07:22 adamholmberg has joined #pypy

07:27 adamholmberg has quit [Ping timeout: 260 seconds]

07:28 asmeurer has quit [Quit: asmeurer]

07:35 <cfbolz> maxxam: no

07:36 <maxxam> ok, thanks

07:38 <maxxam> cfbolz: do you mind if I list off some packages and you tell me if any have known issues?

07:48 <LarstiQ> maxxam: you can at least list and see if there is a response

07:49 <LarstiQ> here if not much, or pastebin/something else otherwise

07:49 <maxxam> LarstiQ: ok

07:49 <maxxam> time

07:49 <maxxam> datetime

07:49 <maxxam> atexit

07:49 <maxxam> threading

07:49 <maxxam> sys

07:49 <maxxam> os

07:49 <maxxam> traceback

07:49 <maxxam> logging

07:49 <maxxam> collections

07:49 <maxxam> concurrent.futures

07:50 <maxxam> random

07:50 <maxxam> statistics

07:50 <maxxam> json

07:50 <maxxam> __future__.absolute_import

07:50 <maxxam> decimal

07:51 <maxxam> base64

07:51 <maxxam> uuid

07:51 <maxxam> hashlib

07:51 <maxxam> hmac

07:51 <maxxam> urllib.parse

07:51 <maxxam> math

07:51 <maxxam> functools

07:51 <maxxam> queue

07:51 <maxxam> itertools

07:51 <maxxam> operator

07:52 <maxxam> websocket

07:52 <maxxam> ssl

07:52 <maxxam> future

07:52 <maxxam> requests

07:52 <maxxam> sortedcollections

07:52 <maxxam> sortedcontainers

07:52 <maxxam> ujson

07:52 <maxxam> websocket_client

07:53 <maxxam> …at least those are the primary suspects

08:10 <bbot2> Retry: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/3992 [py3.5]

08:13 <cfbolz> maxxam: uh, please use a pastebin next time

08:13 <maxxam> cfbolz: right, sorry

08:13 <cfbolz> ujson is slower than just the stdlib json module on PyPy

08:14 <fijal> also possibly explodes

08:14 <cfbolz> fijal: does it?

08:14 <maxxam> cfbolz: huh, wow, didn’t expect that

08:15 <fijal> cfbolz: I remember some C hacks

08:15 <fijal> maaaaybe cpyext grew enough hacks in the meantime

08:15 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/3993 [py3.5]

08:15 <maxxam> fijal: well ujson is being pretty well stress tested here, that could well be it

08:15 <fijal> maxxam: I really don't know. swap it with json and see if it keeps crashing

08:16 <maxxam> fijal: will do.

08:16 <fijal> seriously, who knows

08:17 dddddd has joined #pypy

08:17 <cfbolz> maxxam: nothing else sticks out much

08:18 <fijal> I have no idea what is "websocket" - is it stdlib or something else?

08:20 <maxxam> fijal: it’s kind of a mystery to me as well. let me look

08:20 antocuni has joined #pypy

08:22 TheAdversary has quit [Ping timeout: 268 seconds]

08:23 adamholmberg has joined #pypy

08:23 Hasimir has quit [Ping timeout: 240 seconds]

08:23 <maxxam> fijal: ok, looks like it’s an alias for websocket-client

08:23 <fijal> which is?

08:23 <maxxam> https://github.com/websocket-client/websocket-client

08:24 oberstet has joined #pypy

08:24 <maxxam> ah. and it might be using numpy or wsaccel

08:24 <fijal> seems harmless enough

08:25 <maxxam> * is using numpy

08:27 stevenja_ has joined #pypy

08:28 <maxxam> well, the ujson-ectomy is complete

08:28 adamholmberg has quit [Ping timeout: 245 seconds]

08:28 <maxxam> I’ll let this run for a while and see if that fixes it

08:30 <maxxam> can confirm that there may be a slight performance benefit to switching to json from ujson…

08:32 <maxxam> go pypy, I did not expect that

08:32 stevenja_ has quit [Ping timeout: 264 seconds]

09:04 <cfbolz> maxxam: pypy's json is not completely written in pure python, but has an rpython part

09:12 DIRT has quit [Ping timeout: 260 seconds]

09:33 lritter has joined #pypy

09:35 Ganwell has quit [Ping timeout: 245 seconds]

09:38 Ganwell has joined #pypy

09:59 _whitelogger has joined #pypy

10:00 solarjoe4 has joined #pypy

10:01 solarjoe_ has joined #pypy

10:01 solarjoe_ has quit [Remote host closed the connection]

10:01 solarjoe4 has quit [Client Quit]

10:02 solarjoe4 has joined #pypy

10:12 void_sec3301 has quit [Ping timeout: 256 seconds]

10:16 marvin has quit [Ping timeout: 256 seconds]

10:17 ivh has quit [Ping timeout: 260 seconds]

10:17 antocuni has quit [Ping timeout: 240 seconds]

10:17 Hodgestar has quit [Ping timeout: 276 seconds]

10:25 adamholmberg has joined #pypy

10:29 adamholmberg has quit [Ping timeout: 256 seconds]

10:49 Hasimir has joined #pypy

10:57 Hodgestar has joined #pypy

11:13 oberstet has quit [Ping timeout: 240 seconds]

11:26 adamholmberg has joined #pypy

11:30 adamholmberg has quit [Ping timeout: 240 seconds]

11:32 marky1991 has joined #pypy

11:36 oberstet has joined #pypy

11:40 ivh has joined #pypy

11:42 Hasimir has quit [Read error: Connection reset by peer]

11:43 marvin has joined #pypy

11:45 Hasimir has joined #pypy

11:55 ivh has left #pypy [#pypy]

11:57 TheAdversary has joined #pypy

12:27 stevenja_ has joined #pypy

12:31 stevenja_ has quit [Ping timeout: 265 seconds]

12:50 oberstet has quit [Ping timeout: 255 seconds]

12:51 adamholmberg has joined #pypy

13:02 oberstet has joined #pypy

13:54 stevenja_ has joined #pypy

13:56 DIRT has joined #pypy

13:57 stevenja_ has quit [Remote host closed the connection]

13:57 stevenja_ has joined #pypy

13:58 stevenja_ has quit [Remote host closed the connection]

13:58 stevenja_ has joined #pypy

14:01 stevenja_ has quit [Remote host closed the connection]

14:03 stevenja_ has joined #pypy

14:26 solarjoe4 has quit [Quit: Leaving]

14:28 mefistof1les has joined #pypy

14:35 Gonsor has joined #pypy

14:38 marky1991 has quit [Remote host closed the connection]

14:39 marky1991 has joined #pypy

14:40 marky1991 has quit [Remote host closed the connection]

14:40 marky1991 has joined #pypy

14:47 antocuni has joined #pypy

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

15:00 stevenja_ has joined #pypy

15:04 stevenja_ has quit [Ping timeout: 245 seconds]

15:08 stevenja_ has joined #pypy

15:12 oberstet has quit [Ping timeout: 268 seconds]

15:14 DIRT has quit [Ping timeout: 264 seconds]

15:19 stevenja_ has quit [Remote host closed the connection]

15:20 stevenja_ has joined #pypy

15:25 stevenja_ has quit [Ping timeout: 276 seconds]

15:30 tayfun26 has quit [Quit: tayfun26]

15:38 stevenja_ has joined #pypy

15:40 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/3993 [py3.5]

16:17 stevenja_ has quit [Remote host closed the connection]

16:17 stevenja_ has joined #pypy

16:17 stevenja_ has quit [Remote host closed the connection]

16:18 stevenja_ has joined #pypy

16:20 stevenj__ has joined #pypy

16:20 stevenja_ has quit [Read error: Connection reset by peer]

16:22 ronan has quit [Ping timeout: 256 seconds]

16:24 stevenj__ has quit [Ping timeout: 240 seconds]

16:40 ronan has joined #pypy

16:59 kipras is now known as kipras`away

17:01 asmeurer_ has joined #pypy

17:07 realitix has quit [Ping timeout: 240 seconds]

17:12 antocuni has quit [Ping timeout: 240 seconds]

17:15 realitix has joined #pypy

17:26 realitix has quit [Ping timeout: 256 seconds]

18:13 ronan has quit [Remote host closed the connection]

18:16 ronan has joined #pypy

18:24 asmeurer_ has quit [Quit: asmeurer_]

18:25 asmeurer_ has joined #pypy

18:26 marky1991 has quit [Ping timeout: 245 seconds]

18:44 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-armhf-raspbian/builds/1802 [py3.5]

18:53 marky1991 has joined #pypy

18:53 marky1991 has quit [Remote host closed the connection]

18:54 marky1991 has joined #pypy

19:03 stevenja_ has joined #pypy

19:08 stevenja_ has quit [Ping timeout: 245 seconds]

19:11 stevenja_ has joined #pypy

19:13 asmeurer_ has quit [Quit: asmeurer_]

19:18 Gonsor_ has joined #pypy

19:21 Gonsor has quit [Ping timeout: 245 seconds]

19:39 asmeurer__ has joined #pypy

19:45 jamesaxl has joined #pypy

19:46 <kenaan> arigo default 60d37209763d /rpython/rlib/rvmprof/dummy.py: Fix rvmprof/dummy: stop_sampling() is supposed to return an integer, not None

19:54 Taggnostr has quit [Remote host closed the connection]

19:56 Taggnostr has joined #pypy

19:58 DIRT has joined #pypy

20:00 mattip has joined #pypy

20:01 DIRT has quit [Max SendQ exceeded]

20:01 DIRT has joined #pypy

20:40 mattip_ has quit [Ping timeout: 240 seconds]

20:50 mattip_ has joined #pypy

20:52 raynold has joined #pypy

20:59 stevenja_ has quit [Remote host closed the connection]

20:59 stevenja_ has joined #pypy

21:04 stevenja_ has quit [Ping timeout: 276 seconds]