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
lritter has quit [Quit: Leaving]
amaury has quit [Ping timeout: 256 seconds]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
asmeurer has quit [Quit: asmeurer]
inhahe has quit [Ping timeout: 264 seconds]
adamholmberg has joined #pypy
marr has quit [Ping timeout: 256 seconds]
pf_moore has quit [Quit: Connection closed for inactivity]
orion has joined #pypy
orion is now known as Guest83326
Guest83326 has quit [Remote host closed the connection]
asmeurer has joined #pypy
adamholmberg has quit [Ping timeout: 240 seconds]
asmeurer has quit [Quit: asmeurer]
asmeurer has joined #pypy
orion has joined #pypy
orion is now known as Guest1170
asmeurer has quit [Quit: asmeurer]
asmeurer has joined #pypy
Guest1170 has quit [Remote host closed the connection]
jcea has quit [Quit: jcea]
marky1991 has quit [Read error: Connection reset by peer]
orion has joined #pypy
jamesaxl has joined #pypy
orion is now known as Guest5333
forgottenone has joined #pypy
forgottenone has quit [Client Quit]
tbodt has joined #pypy
forgottenone has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Guest5333 has quit [Remote host closed the connection]
speeder39 has joined #pypy
__main__ has quit [Read error: Connection reset by peer]
_main_ has joined #pypy
_main_ is now known as __main__
dddddd has quit [Remote host closed the connection]
kirma has quit [Ping timeout: 240 seconds]
tayfun26 has joined #pypy
redj_ has joined #pypy
redj has quit [Ping timeout: 276 seconds]
redj_ has quit [Ping timeout: 268 seconds]
redj has joined #pypy
mcyprian has joined #pypy
<kenaan> arigo default c4654fc5b220 /rpython/rlib/: Extend FinalizerQueue to be able to handle low-level GCREF pointers directly, instead of high-level instances of so...
amaury has joined #pypy
Manis_ has quit [Quit: WeeChat 2.0.1]
inhahe__ has quit [Ping timeout: 260 seconds]
illume has joined #pypy
inhahe__ has joined #pypy
Remi_M has joined #pypy
Remi_M has quit [Client Quit]
jdb_ has joined #pypy
Remi_M has joined #pypy
kirma has joined #pypy
Remi_M1 has joined #pypy
Remi_M has quit [Ping timeout: 256 seconds]
mcyprian has left #pypy [#pypy]
solarjoe4 has joined #pypy
speeder39 has quit [Quit: Connection closed for inactivity]
amaury has quit [Ping timeout: 256 seconds]
ronan has quit [Ping timeout: 256 seconds]
orion has joined #pypy
orion is now known as Guest98283
ronan has joined #pypy
Guest98283 has quit [Ping timeout: 255 seconds]
<kenaan> arigo default 2804dc8a8645 /rpython/rlib/rgc.py: FinalizerQueue works with Boehm too
jdb_ has quit [Ping timeout: 256 seconds]
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
idnar has quit [Ping timeout: 256 seconds]
dpn` has quit [Ping timeout: 256 seconds]
idnar has joined #pypy
dpn` has joined #pypy
illume has joined #pypy
<phlebas> what's the default encoding in pypy3 for cpyext (Py_UNICODE) strings? the typedef is wchar_t, but will it still write the selected sytem encoding, or will it write something dependent on the size of wchar_t?
jdb_ has joined #pypy
<kenaan> arigo reverse-debugger-updated 78d7729c2379 /pypy/module/cpyext/api.py: Fix type
raynold has quit [Quit: Connection closed for inactivity]
asmeurer has quit [Quit: asmeurer]
hawkowl has joined #pypy
<hawkowl> hello I have a question... bug... bug question
<hawkowl> I am tracking down the cause of a Twisted bug that causes init in thread 1 -> weakref in thread 1 -> send to thread 2 -> del in thread 1 -> thread 2 completes -> weakref still resolves
<hawkowl> I seem to have tracked it down to this
<hawkowl> it says it's refererred to by <frame object at 0x00007f06d005cda0>
<hawkowl> if I see what refers to that, it is [<thread._local object at 0x00007f9c5ea289f8>, <thread._local object at 0x00007f9c580601b0>, <frame object at 0x00007f9c5805cbc0>, <thread._local object at 0x00007f9c5f99c890>, (<frame object at 0x00007f9c5805cbc0>,)]
<hawkowl> so thread._local is holding onto the reference because it was in scope
jdb_ has quit [Ping timeout: 256 seconds]
energizer has quit [Ping timeout: 256 seconds]
jdb_ has joined #pypy
antocuni has joined #pypy
<hawkowl> okay, hm. it's being held onto by a Twisted ThreadedContextTracker
<hawkowl> which is calling local()
iko_ has joined #pypy
<fijal> hawkowl: weakref can still resolve after del
<hawkowl> what
<hawkowl> uh
<fijal> hawkowl: of course
<hawkowl> so how do i make it super weak
<hawkowl> weakest ref
<fijal> you can't
<fijal> even on cpython (really)
<hawkowl> hm
<fijal> weakref is only guaranteed to *eventually* die
<hawkowl> ok so is this test just busted by design then?
<hawkowl> it does die, just after the test lifetime
<fijal> yeah, you can do gc.collect() at the end of the test
<fijal> and check after that
<fijal> for tests this is acceptable
<hawkowl> yeah gc.collect() doesn't get rid of it
<hawkowl> the thread local hangs around after the thread dies
<fijal> run it 2 or 3 times?
<fijal> ah
<fijal> well if it's kept alive by frame, you need to delete it there
oberstet has joined #pypy
jdb_ has quit [Ping timeout: 246 seconds]
<hawkowl> I don't know what is... making the frame
<hawkowl> _local is
<hawkowl> is that the execution context?
<cfbolz> fijal: I think it's entirely possible that our thread local implementation hangs on to stuff a bit too much
<hawkowl> i tried pulling apart the obj but nothing from python side seems to say it
<fijal> cfbolz: completely
<hawkowl> `twisted.test.test_threadpool.ThreadPoolTests.test_threadCreationArgumentsCallInThreadWithCallback` is the test
<hawkowl> it's the only one failing on pypy3
<hawkowl> apart from a flaky one
<hawkowl> twisted.protocols.test.test_tls.TLSMemoryBIOTests.test_hugeWrite_TLSv1_* is timing out for some reason. hm.
<hawkowl> not locally tho
jdb_ has joined #pypy
antocuni has quit [Ping timeout: 246 seconds]
pf_moore has joined #pypy
ariava has joined #pypy
orion has joined #pypy
orion is now known as Guest8295
Guest8295 has quit [Ping timeout: 276 seconds]
<kenaan> jdb cffi/cffi 9318184d748c /doc/source/overview.rst: Show the recommended use case: API/out-of-line first, in its own section.
marr has joined #pypy
solarjoe4 has quit [Quit: Leaving]
antocuni has joined #pypy
marky1991 has joined #pypy
<kenaan> mjacob buildbot d86c1810081c /master/templates/layout.html: Add link to py3.6 summary.
solarjoe4 has joined #pypy
<mjacob> ^^^^^^ please restart
tayfun26 has quit [Quit: tayfun26]
stevie has joined #pypy
jcea has joined #pypy
jcea has quit [Remote host closed the connection]
stevie has left #pypy ["Auf Wiedersehen"]
jcea has joined #pypy
<kenaan> mjacob py3.6 9f4edb1898f2 /pypy/interpreter/test/test_syntax.py: Remove EOL whitespace.
jcea has quit [Quit: jcea]
jcea has joined #pypy
adamholmberg has joined #pypy
orion has joined #pypy
orion is now known as Guest99505
<xorAxAx> our NP complete problem solution heuristic inserted ~1200 stack checks in my current translation
amaury has joined #pypy
Guest99505 has quit [Ping timeout: 245 seconds]
inad922 has joined #pypy
<illume> Does anyone know how to pypy-on-appveyor? They don't install pypy by default. https://www.appveyor.com/docs/build-environment/#python
<inad922> Hi
<inad922> The stm development for pypy is really dropped?
<inad922> I just read this article
<illume> I guess for now I'll download pypy win32 release with curl on appveyor.
<Rotonen> inad922: the more accurate way is to say the ways explored did not yield within budget - that stuff is hard
<arigato> cfbolz: so I *think* that what occurs in guard-compatible:
<arigato> AbstractAttribute._type_lookup_pure() is marked
<cfbolz> ok?
<arigato> with elidable_compatible
<arigato> I'm still looking but we suspect that it returns different objects for the same arguments (different but meaning the same thing for most purposes)
<cfbolz> arigato: ok, that would be an actual bug leading to miscompilation, right?
<inad922> Rotonen: Interesting. Thanks!
<kenaan> Alexander Schremmer reverse-debugger-updated 1f98c4723883 /rpython/translator/revdb/src-revdb/revdb.c: Make it compile, breaks identityhash of prebuilt objects.
lritter has joined #pypy
inad922 has quit [Ping timeout: 240 seconds]
amaury has quit [Ping timeout: 260 seconds]
<cfbolz> antocuni: you around?
inad922 has joined #pypy
<mjacob> cfbolz: he's not here right now, no clue when he comes back
<kenaan> mjacob py3.6 1938b1ea9975 /pypy/interpreter/: Raise SyntaxError when 'return' with value is used inside an async generator.
<antocuni> cfbolz: hi
<cfbolz> antocuni: random question, how do I run the pdbpp tests?
<cfbolz> and are they supposed to pass atm?
<antocuni> yes, I think there is even a travis build IIRC
<cfbolz> ah, nice
<cfbolz> so if I open a pull request travis will tell me :-)
<antocuni> yes :)
<antocuni> anyway, I think that "py.test" or "tox" should just work
<antocuni> (the last famous words)
<cfbolz> antocuni: I get lots of failures, of course
<antocuni> maybe missing dependencies?
<cfbolz> shouldn't tox install those?
adamholmberg has quit [Remote host closed the connection]
<antocuni> yes
adamholmberg has joined #pypy
<cfbolz> probably all the same problem
<cfbolz> something about hidden frames
<antocuni> so it's the "5 frames hidden" which causes the problem
<antocuni> this is on my machine, they pass with no problem http://paste.openstack.org/show/706057/
<cfbolz> antocuni: maybe it's a change in pytest?
<antocuni> yeah, could be
<cfbolz> :-(
<cfbolz> antocuni: I am trying to add a heuristic to distinguish between "list 140" and "list(iterator)"
<cfbolz> simply by checking whether the next character after "list" is a "(" or not
adamholmberg has quit [Ping timeout: 265 seconds]
<antocuni> very nice, I desired this feature for a long time but I always been too lazy to implement it :)
<cfbolz> antocuni: :-)
<cfbolz> me too
<cfbolz> the other thing I did was a small improvement to sticky_by_default
<cfbolz> when this is on, and you run pytest --pdb, you don't see the error message, because the screen is cleared
<cfbolz> antocuni: so I propose changing it to not clear the screen the first time
marky1991 has quit [Ping timeout: 240 seconds]
adamholmberg has joined #pypy
amaury has joined #pypy
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
raynold has joined #pypy
<kenaan> rlamy py3tests 488e2edbef85 /pypy/tool/pytest/: Inject a fake pytest module for untranslated new-style applevel tests
jdb_ has quit [Ping timeout: 256 seconds]
adamholmberg has quit [Remote host closed the connection]
adamholmberg has joined #pypy
marky1991 has joined #pypy
<kenaan> antocuni cpyext-fast-typecheck 3ee549601971 /: (anto, matti): start a branch in which to speed up things like PySlice_Check and similar
<kenaan> antocuni cpyext-fast-typecheck c05e38122aa5 /pypy/module/cpyext/test/test_sliceobject.py: (anto, matti): add a passing test
<kenaan> antocuni cpyext-fast-typecheck 185b284722b0 /pypy/module/cpyext/: (anto, matti) introduce a new way to implement Py*_Check efficiently: we add a new field to typeob...
<kenaan> antocuni cpyext-fast-typecheck 4744cec39496 /pypy/module/cpyext/: implement PySlice_Check as a fast macro instead of going through the slow roundtrip
<kenaan> antocuni cpyext-fast-typecheck 403e8744b785 /pypy/module/cpyext/sliceobject.py: (anto, matti): fix the tests broken by the previous commit: we are no longer able to call PySlice_...
orion has joined #pypy
orion is now known as Guest74398
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
<kenaan> mattip cpyext-fast-typecheck 833681add062 /pypy/module/cpyext/: anto, matti) implement PyBool_Check as a fast macro
mcyprian has joined #pypy
<kenaan> Raemi guard-compatible c13441bb1f3b /rpython/jit/metainterp/: (arigo, remi) fix test
huonw has quit [Quit: bye]
illume has joined #pypy
solarjoe4 has quit [Quit: Leaving]
ronan has quit [Quit: Ex-Chat]
ronan__ has joined #pypy
inad922 has quit [Ping timeout: 252 seconds]
jcea has quit [Remote host closed the connection]
jcea has joined #pypy
inad922 has joined #pypy
energizer has joined #pypy
Guest74398 has quit [Remote host closed the connection]
inad923 has joined #pypy
inad922 has quit [Ping timeout: 256 seconds]
mcyprian has left #pypy [#pypy]
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
jcea has quit [Remote host closed the connection]
jcea has joined #pypy
<kenaan> Raemi guard-compatible 931f5eaed82f /rpython/jit/metainterp/: (arigo, remi) test and fix a case of too many bridges see test for explanation
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
<kenaan> mjacob py3.6 02930c62826c /lib-python/3/test/test_asyncgen.py: Add gc_collect() to make it crash more reliably. Yay!
<tos9> arigato: (re: your answer to my errors) -- that error is while trying to build pypy
jacob22_ has quit [Quit: Konversation terminated!]
jacob22__ has joined #pypy
marky1991 has quit [Ping timeout: 256 seconds]
ronan__ is now known as ronan
dddddd has joined #pypy
jcea has quit [Ping timeout: 276 seconds]
marky1991 has joined #pypy
asmeurer has joined #pypy
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
asmeurer has quit [Quit: asmeurer]
inad923 has quit [Ping timeout: 240 seconds]
jcea has joined #pypy
raynold has quit [Quit: Connection closed for inactivity]
oberstet has quit [Ping timeout: 245 seconds]
inad923 has joined #pypy
raynold has joined #pypy
<kenaan> arigo guard-compatible ec6d67e6b1cb /rpython/jit/metainterp/optimizeopt/optimizer.py: (remi, arigo) Fix what looks like a bug by staring at the diff with default
inad923 has quit [Ping timeout: 260 seconds]
illume has joined #pypy
marky1991 has quit [Ping timeout: 248 seconds]
antocuni has quit [Ping timeout: 265 seconds]
illume has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
kolko has joined #pypy
kolko has quit [Ping timeout: 256 seconds]
marky1991 has joined #pypy
forgottenone has quit [Read error: Connection reset by peer]
forgottenone has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
krono has joined #pypy
adamholmberg has quit []
<nanonyme> I wonder why https://pypy.org/compat.html isn't mentioning "with contextlib.closing(gen()) as iterable: for e in iterable" as an alternative for making sure generators are closed...
<nanonyme> (and also why generators are simply not context managers always period)
forgottenone has quit [Ping timeout: 256 seconds]
asmeurer has joined #pypy
amaury has quit [Quit: Konversation terminated!]
amaury_ has joined #pypy
amaury_ has quit [Ping timeout: 246 seconds]
nimaje has quit [Remote host closed the connection]
nimaje has joined #pypy
amaury_ has joined #pypy
drolando_ has joined #pypy
drolando has quit [Read error: Connection reset by peer]
drolando has joined #pypy
drolando_ has quit [Read error: Connection reset by peer]
marky1991 has quit [Quit: Saliendo]
marky1991 has joined #pypy
marky1991 has quit [Client Quit]
marky1991 has joined #pypy
drolando has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
marky1991 has quit [Read error: Connection reset by peer]
drolando has joined #pypy
marky1991 has joined #pypy
asmeurer has quit [Quit: asmeurer]
asmeurer has joined #pypy
antocuni has joined #pypy
illume has joined #pypy
illume has quit [Client Quit]
drolando has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
asmeurer has quit [Quit: asmeurer]
drolando has joined #pypy
asmeurer has joined #pypy
marky1991 has quit [Read error: Connection reset by peer]
ariava has quit [Ping timeout: 260 seconds]
jdb_ has joined #pypy
<kenaan> rlamy py3tests 46d7ec4d97a8 /_pytest/assertion/rewrite.py: Revert changes to _pytest/ made by this branch
Remi_M1 has quit [Ping timeout: 256 seconds]
jdb_ has quit [Remote host closed the connection]
antocuni has quit [Ping timeout: 256 seconds]
lritter has quit [Ping timeout: 276 seconds]
<njs> cfbolz: fwiw cpython's threading.local goes to extreme measures to avoid pinning thing in memory
<njs> nanonyme: generators aren't context managers because one of the original design goals for generators was supporting @contextlib.contextmanager, and it was considered too confusing if you accidentally left off the decorator and still got a context manager out
<njs> nanonyme: (note I am not endorsing this; just that's what ncoghlan told me)
jamesaxl has quit [Quit: WeeChat 2.0.1]
jiffe has quit [Quit: WeeChat 2.0.1]
jiffe has joined #pypy
tbodt has quit [Read error: Connection reset by peer]
tbodt has joined #pypy
<kenaan> mattip py3.5 3da473ecc0a8 /lib-python/3/distutils/msvc9compiler.py: more debugging of finding cl.exe, 226287429316 did not print anything
<mattip> it seems to consistently appear after commit 57349ce40416
orion has joined #pypy
orion is now known as Guest92476
jredondo has joined #pypy
<jredondo> Hi there. It seems that my pypy3 installation (ubuntu precompiled 5.10.1) is not able to load modules that are *.so. Any suggestion? Thanks in advance.
<dash> jredondo: what are the symptoms?