antocuni changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://botbot.me/freenode/pypy/ ) | use cffi for calling C | "PyPy: the Gradual Reduction of Magic (tm)"
<mattip>
but py3.5 can't use virtualenv, even on HEAD it fails
zmt00 has quit [Quit: Leaving]
<mattip>
on the 12.04 chroot linux 32 build slave
mattip has left #pypy ["bye"]
<fijal>
tos9: possible pong
<fijal>
but only 50%
marky1991 has quit [Read error: Connection reset by peer]
asmeurer_ has joined #pypy
asmeurer_ has quit [Client Quit]
marvin_ has quit [Remote host closed the connection]
marvin has joined #pypy
marvin is now known as Guest39186
marr has joined #pypy
slackyy has quit [Ping timeout: 268 seconds]
inhahe_ has quit [Ping timeout: 255 seconds]
inhahe_ has joined #pypy
<kenaan_>
arigo default eb297be15f48 /pypy/module/thread/test/test_import_lock.py: Fix this fragile test (it broke because of 167b802baf3b, unsure why)
antocuni has joined #pypy
the_drow has quit [Ping timeout: 260 seconds]
gclawes has quit [Ping timeout: 246 seconds]
the_drow has joined #pypy
marr has quit [Ping timeout: 248 seconds]
the_drow has quit [Ping timeout: 248 seconds]
asmeurer__ has joined #pypy
antocuni has quit [Ping timeout: 268 seconds]
asmeurer__ has quit [Ping timeout: 268 seconds]
the_drow has joined #pypy
the_drow has quit [Ping timeout: 240 seconds]
the_drow has joined #pypy
asmeurer has joined #pypy
asmeurer has quit [Ping timeout: 250 seconds]
the_drow has quit [Ping timeout: 250 seconds]
the_drow has joined #pypy
asmeurer has joined #pypy
asmeurer has quit [Ping timeout: 268 seconds]
the_drow has quit [Ping timeout: 268 seconds]
the_drow has joined #pypy
the_drow has quit [Ping timeout: 248 seconds]
asmeurer has joined #pypy
the_drow has joined #pypy
asmeurer has quit [Ping timeout: 248 seconds]
<carljm_>
fijal: sure, what were you wondering about?
carljm_ has quit [Quit: leaving]
carljm has joined #pypy
<carljm>
fijal: I didn't write the post or do much of the related work, I was just signal-boosting it on Twitter :-) but if you have questions I can probably put you in touch with the right person (and you probably already met them...)
<tos9>
fijal: :) hullo, how goes
asmeurer has joined #pypy
asmeurer has quit [Ping timeout: 248 seconds]
zmt00 has joined #pypy
the_drow has quit [Ping timeout: 248 seconds]
inhahe has joined #pypy
inhahe has quit [Client Quit]
the_drow has joined #pypy
<fijal>
tos9: hey hey
<fijal>
getting snowed out
<fijal>
carljm: yeah that was what I was wondering - I kinda arrived at similar conclusions about 10 years ago
<fijal>
and armin about 15 :)
asmeurer_ has joined #pypy
asmeurer_ has quit [Ping timeout: 258 seconds]
<arigato>
can you post the url for the post you're talking about?
<arigato>
can't find it here
<tos9>
fijal: did pypy adopt `perf`? and if so, do you have a place that you're actually putting the output?
raynold has quit [Quit: Connection closed for inactivity]
<fijal>
tos9: no, but we use similar strategies to perf for vmprof I would think?
<tos9>
fijal: will have a look (all I have is a SAAS problem I assume -- I want to have some benchmarks, and I want a service that will track the results of it over time :)
<fijal>
yeah I know :)
<fijal>
but our current answer is "there is not enough business to warrant massive effort, but we can do it as a contract"
* tos9
nods
<tos9>
fijal: when jsonschema makes me a millionaire you already know I'm giving you half :(
<tos9>
ok back to trying to get travis to not kill the container after an hour, or trying to get the benchmarks to not take an hour :)
<fijal>
tos9: well, I'm a bit torn, but ALSO I don't have that much time these days
<tos9>
fijal: Oh yeah I definitely do not expect to write anything :) -- just was figuring you'd have tackled it already
<tos9>
though I guess you have buildbot yourself, so it's less of a problem there
<tos9>
I also kind of expected the perf docs to suggest something
<tos9>
but maybe I'll file a ticket about that
<LarstiQ>
can the benchmark be split and run on, say, different branches?
<tos9>
LarstiQ: yeah so in theory I guess I can sample which ones I run or something -- I don't know yet what the rules are in travis, whether the 1 hour thing is per build or per sub-whatever...
<tos9>
but even that doesn't really solve "where is it that I store benchmark results and track them over time"
<LarstiQ>
right
asmeurer has joined #pypy
asmeurer has quit [Client Quit]
asmeurer__ has joined #pypy
mattip has joined #pypy
slackyy has joined #pypy
asmeurer__ has quit [Quit: asmeurer__]
<fijal>
mattip: hi
<mattip>
hi
<mattip>
fijal: did you get a chance to try the -A tests with a nightly macos build?
<fijal>
mattip: yes, seems to pass
the_drow has quit [Ping timeout: 240 seconds]
<fijal>
that was on memory accounting branch, I'll translate default
<mattip>
thanks, I'll reach out to the build slave owner to see if they can figure out what is going on
<mattip>
probably a corrupt file/directory somehow
<fijal>
mattip: antocuni broke translation I think?
<mattip>
I issued a pull request to vmprof-python to fix gcc warnings, could you review / see that it doesn't break clang macos ?
<fijal>
no worries
<fijal>
vmprof-python?
<fijal>
yeah
<fijal>
but after dinner
<mattip>
it seems strange to me that there is rpython-specific stuff there, but I don't want to change too much
<ronan>
mattip: to fix the virtualenv failure, the buildslave needs a newer virtualenv
<mattip>
could we do that as part of the build or does it need to be done by hand on the slave?
the_drow has quit [Ping timeout: 268 seconds]
<fijal>
mattip: test_posix2.py passes
<fijal>
./pypy-c ./pypy/test_all.py -A pypy/module/posix/test/test_posix2.py
<fijal>
97 passed, 3 skipped in 5.84 seconds
<ronan>
mattip: it has to be done on the slave - unless we change the build to do everything inside a virtualenv
<fijal>
it has 4 failures on cpython tho
asmeurer has joined #pypy
the_drow has joined #pypy
adamholmberg has joined #pypy
traverseda has quit [Ping timeout: 240 seconds]
asmeurer has quit [Quit: asmeurer]
traverseda has joined #pypy
adamholmberg has quit [Remote host closed the connection]
adamholmberg has joined #pypy
adamholmberg has quit [Ping timeout: 246 seconds]
antocuni has joined #pypy
antocuni has quit [Ping timeout: 248 seconds]
the_drow has quit [Ping timeout: 248 seconds]
<mattip>
arigato: it seems the linux32 buildbot has a very old virtualenv, could you update it?
the_drow has joined #pypy
<fijal>
mattip: what's the story with virtualenv not working on 3.5?
<mattip>
my bad, it seems numpy recipe to run tests also uses an old virtualenv
<mattip>
just like the linux32 build slave
asmeurer__ has joined #pypy
asmeurer__ has quit [Quit: asmeurer__]
the_drow has quit [Ping timeout: 240 seconds]
mattip has left #pypy ["bye"]
the_drow has joined #pypy
jamesaxl has quit [Quit: WeeChat 1.9.1]
asmeurer_ has joined #pypy
marr has joined #pypy
Guest39186 has quit [Ping timeout: 248 seconds]
raynold has joined #pypy
marvin has joined #pypy
marvin is now known as Guest95003
antocuni has joined #pypy
<kenaan_>
arigo default d00a16ef468f /rpython/annotator/annrpython.py: Follow-up for cb9634421fa2: revert the very general change and instead improve the logic at the point where it matt...
<arigato>
mattip (logs): maybe we need another approach then---e.g. including or downloading a recent virtualenv from the tests?
<kenaan_>
arigo default 60c4fa1b0539 /rpython/: Tweak: 16% speed increase of PyPy annotation, by avoiding the situation where a block is rescheduled many many time...