kipras`away is now known as kipras
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
kipras is now known as kipras`away
marr has quit [Ping timeout: 255 seconds]
yuyichao has quit [Ping timeout: 240 seconds]
yuyichao has joined #pypy
tilgovi has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
ArneBab has joined #pypy
ArneBab_ has quit [Ping timeout: 252 seconds]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
lritter_ has joined #pypy
lritter has quit [Read error: Connection reset by peer]
marky1991 has quit [Ping timeout: 240 seconds]
lritter_ has quit [Remote host closed the connection]
tbodt has joined #pypy
<njs> haypo: for ad hoc measurements ('perf timeit') what I generally do is just have a loop that just keeps printing measurements until I hit control-C :-). It's not a rigorous comparative methodology, but for quickly learning something about performance it's pretty effective.
oberstet2 has joined #pypy
oberstet has quit [Ping timeout: 240 seconds]
tilgovi has quit [Remote host closed the connection]
mattip has joined #pypy
<mattip> another IRC logger just added #pypy https://irclog.whitequark.org/pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
tbodt has quit [Client Quit]
capuk has quit [Remote host closed the connection]
jacob22 has quit [Quit: Konversation terminated!]
kipras`away is now known as kipras
kipras is now known as kipras`away
jamadden has quit [Quit: Leaving.]
vkirilichev has joined #pypy
_whitelogger has joined #pypy
vkirilichev has quit [Remote host closed the connection]
vkirilichev has joined #pypy
vkirilichev has quit [Ping timeout: 255 seconds]
sven__ has joined #pypy
sven__ has quit [Remote host closed the connection]
sven__ has joined #pypy
jamesaxl has joined #pypy
benbange1 has joined #pypy
tmarkovich has quit [Ping timeout: 268 seconds]
plan_rich has quit [Ping timeout: 260 seconds]
tmarkovich has joined #pypy
panni_ has quit [Ping timeout: 268 seconds]
lukasa has quit [Ping timeout: 268 seconds]
plan_rich has joined #pypy
bgola has quit [Ping timeout: 260 seconds]
ebarrett has quit [Ping timeout: 268 seconds]
gutworth has quit [Ping timeout: 268 seconds]
benbangert has quit [Ping timeout: 268 seconds]
vishesh has quit [Ping timeout: 260 seconds]
tote has quit [Ping timeout: 260 seconds]
WGH has quit [Ping timeout: 268 seconds]
amontalenti has quit [Ping timeout: 268 seconds]
ebarrett has joined #pypy
gutworth has joined #pypy
panni_ has joined #pypy
WGH has joined #pypy
john51_ has quit [Read error: Connection reset by peer]
tote has joined #pypy
lukasa has joined #pypy
john51 has joined #pypy
amontalenti has joined #pypy
vishesh has joined #pypy
vkirilichev has joined #pypy
bgola has joined #pypy
gbutnaru has quit [Ping timeout: 260 seconds]
gbutnaru has joined #pypy
gbutnaru has quit [Ping timeout: 240 seconds]
realitix has joined #pypy
gbutnaru has joined #pypy
vkirilichev has quit [Remote host closed the connection]
jamesaxl has quit [Ping timeout: 240 seconds]
jamesaxl has joined #pypy
mattip has quit [Ping timeout: 240 seconds]
mattip has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
oberstet2 has quit [Ping timeout: 245 seconds]
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
oberstet2 has joined #pypy
antocuni has joined #pypy
<kenaan> arigo py3.5 d944ef8c8aa0 /lib-python/3/: pathlib.mkdir(): apply the fixes done in CPython > 3.5, and apply on top of that the fix for CPython issue #29694
arigato has joined #pypy
marr has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
jacob22_ has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
sven__ has quit [Remote host closed the connection]
sven__ has joined #pypy
sven__ has quit [Remote host closed the connection]
sven__ has joined #pypy
kurtbrose[m] has quit [Remote host closed the connection]
agates has quit [Read error: Connection reset by peer]
bendlas has quit [Read error: Connection reset by peer]
kurtbrose[m] has joined #pypy
OOPMan has joined #pypy
agates has joined #pypy
bendlas has joined #pypy
antocuni has quit [Ping timeout: 240 seconds]
arigato has quit [Quit: Leaving]
vkirilichev has joined #pypy
OOPMan has quit [Quit: Konversation terminated!]
girish946 has joined #pypy
jamadden has joined #pypy
squeaky_pl has joined #pypy
bgola_ has joined #pypy
vkirilichev has quit [Remote host closed the connection]
vkirilichev has joined #pypy
bgola_ has quit [Quit: leaving]
jacob22_ has quit [Quit: Konversation terminated!]
antocuni has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
jamesaxl has quit [Ping timeout: 260 seconds]
jamesaxl has joined #pypy
girish946 has quit [Quit: Leaving]
[0__0] has quit [Remote host closed the connection]
<mattip> issue 2530 has it all: using multiprocess.ThreadPool, cython, numpy object arrays, _csv, StringIO
<mattip> using valgrind and gdb hasn't been productive, neither has trying to cut the code down to a smaller example
squeaky_pl has quit [Ping timeout: 260 seconds]
Rhy0lite has joined #pypy
adamholmberg has joined #pypy
<cfbolz> mattip: that's terrifying
<mattip> that's after I took out cython's optimization to use the buffer interface for fast access to the ndarray data
lritter has joined #pypy
marky1991 has joined #pypy
jacob22_ has joined #pypy
[0__0] has joined #pypy
capuk has joined #pypy
capuk has quit [Remote host closed the connection]
capuk has joined #pypy
arigato has joined #pypy
vkirilic_ has joined #pypy
yuyichao has quit [Ping timeout: 240 seconds]
vkirilichev has quit [Ping timeout: 260 seconds]
ceridwen has quit [Ping timeout: 246 seconds]
<kenaan> arigo py3.5 a25d530e6773 /lib-python/3/test/test_pathlib.py: Update from feedback
derwolfe has joined #pypy
ceridwen has joined #pypy
yuyichao has joined #pypy
ceridwen has quit [Changing host]
ceridwen has joined #pypy
<derwolfe> good morning :) I've deployed several services with pypy and would love to try to get them all moved over to pypy3. I've noticed that the pypy3-5.7.1 build doesn't currently have a binary build for OSX available. Is this something with which I can help or are there other reasons why it might not have been built with which I'm not familiar.
<LarstiQ> derwolfe: you can likely help
<LarstiQ> derwolfe: iirc the buildbot has had some issues, and perhaps some porting work
<LarstiQ> derwolfe: so first would be to figure out if the osx buildbot is online again, and see if it has produced a workable nightly lately
<derwolfe> LarstiQ: thanks! I'll take a look :)
<mattip> derwolfe: the build is failing AFAICT because it cannot find the needed ssl headers
<mattip> it might just be the OS version, or the way the environment is set up or ...
<dstufft> arigato: Can cffi itself be compiled with Py_Limited_API defined, or is that something that only things using CFFI can have?
tbodt has joined #pypy
<mattip> derwolfe: no, that is a newer error, but looking back over the last builds it seems to have crept in a while ago
<mattip> derwolfe: looking at the man page for posix_fadvise http://man7.org/linux/man-pages/man2/posix_fadvise.2.html
<mattip> they seem to suggest using a macro to test it will work
<mattip> _POSIX_C_SOURCE >= 200112L
* mattip off
mattip has left #pypy ["bye"]
sven__ has quit [Remote host closed the connection]
jamesaxl has quit [Read error: Connection reset by peer]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
benbange1 is now known as benbangert
vkirilic_ has quit [Remote host closed the connection]
yuyichao has quit [Ping timeout: 240 seconds]
yuyichao has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
arigato has quit [Ping timeout: 252 seconds]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
realitix has quit [Quit: Leaving]
oberstet2 has quit [Ping timeout: 240 seconds]
sven__ has joined #pypy
yuyichao has quit [Remote host closed the connection]
nimaje has quit [Remote host closed the connection]
yuyichao has joined #pypy
nimaje has joined #pypy
yuyichao_ has joined #pypy
yuyichao has quit [Ping timeout: 260 seconds]
oberstet2 has joined #pypy
necaris has joined #pypy
brechtm has joined #pypy
<brechtm> Hi. Is it possible to somehow pass CFLAGS and LDFLAGS to build_cffi_imports.py?
necaris has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<brechtm> Hmm, build_cffi_imports.py doesn't show the original error about the openssl include anymore...
pxwang has joined #pypy
<brechtm> Don't tell me I have to redo the whole translation to fix this...
arigato has joined #pypy
<arigato> dstufft: no, Py_LIMITED_API is not compatible with cffi itself, I tried
<dstufft> arigato: ah bummer :(
<cfbolz> brechtm: no, you can call lib_pypy/_ssl_build.py without translating again, I think
<brechtm> cfbolz: even without setting CFLAGS, LDFLAGS or PKG_CONFIG_PATH, it doesn't result in the same error about the missing openssl header
<brechtm> cfbolz: so something must have been generated that is now being skipped or something
<brechtm> From http://doc.pypy.org/en/latest/build.html#build-cffi-import-libraries-for-the-stdlib it is not very clear, but it seems openssl needs to be present (and discoverable) at translate-time.
<cfbolz> brechtm: this is pypy3, right?
<brechtm> cfbolz: yes
<brechtm> Building on macOS
<cfbolz> right
<cfbolz> the build products of this script are .so files in lib_pypy, I think? but not 100% familiar with that code
<brechtm> I started translation again, prefixing the command with PKG_CONFIG_PATH="-I$(brew --prefix openssl)/lib/pkgconfig" as suggested in https://bitbucket.org/pypy/pypy/issues/2174/pypy-cannot-be-translated-on-os-x-1011-el
<brechtm> I'll find out tomorrow whether it worked or not...
<cfbolz> brechtm: as I said, you should really not have to retranslate completely
<brechtm> cfbolz: yup, but I don't know what else to do, so ...
<cfbolz> brechtm: the issue is about pypy2, which had the _ssl module as a compile in module, not in cffi
<brechtm> ah
<cfbolz> brechtm: maybe poking around lib_pypy/_cffi_ssl/_cffi_src/build_openssl.py can shed some light
<brechtm> cfbolz: already did, but no light :-)
<brechtm> cfbolz: I have the feeling I need to delete some intermediary file that was built the first time
<brechtm> oh... perhaps lib_pypy/_pypy_openssl.o
<cfbolz> brechtm: yes
antocuni has quit [Ping timeout: 240 seconds]
<brechtm> nope
<cfbolz> brechtm: wait, so it manages to build a .so?
<cfbolz> but then the import fails, or what?
sven__ has quit [Remote host closed the connection]
<cfbolz> looking around, it seems the support needed for mac was not at all ported from the pypy2 way of building things to the pypy3 way
<brechtm> there have been pypy3 binary releases for mac though
<cfbolz> no?
<cfbolz> not since 3.5 support landed, I think
<brechtm> cfbolz: This was the output at the end of the translation: https://paste.pound-python.org/show/Ou30wAgQev6gUZM3hJKr/
<brechtm> cfbolz: indeed
<brechtm> cfbolz: notice the error on #include... don't get that anymore
<cfbolz> so something is weirdly different, for a bit unknown reasons. but the fact remains that it's probably not at all looking into the right paths
<cfbolz> so it's probably a matter of teaching the new code about pkg-config
<cfbolz> plan_rich: ^^ if you have any input on this?
<brechtm> cfbolz: thanks for the help in any case... I have to run now
<cfbolz> brechtm: thanks for digging
<cfbolz> brechtm: if you're interested, it's probably not super hard
<cfbolz> no rpython magic involved or anything
brechtm_ has joined #pypy
brechtm has quit [Ping timeout: 252 seconds]
yuyichao has joined #pypy
brechtm_ has quit [Ping timeout: 240 seconds]
yuyichao_ has quit [Ping timeout: 268 seconds]
necaris has joined #pypy
necaris is now known as necaris[away]
necaris[away] is now known as necaris
<necaris> Hey folks -- CFFI question: how do I create a C `char **`? I have a library function `FooGetInfo(char **info)` that will allocate as necessary to return data in `**info` and requires me to call `FooFree(*info)` after I'm done with it. All the examples for `ffi.new()` involve initializing data to pass into C functions, but I can't see how to do so without any initializing data...
<kenaan> rlamy PyBuffer 23f969f08eb4 /pypy/: Move tolist() implementation from W_MemoryObject to Buffer
<kenaan> rlamy PyBuffer 96ffa9bf2352 /pypy/: Rename Buffer to PyBuffer
<kenaan> rlamy PyBuffer a2ed7ebe5863 /pypy/: clean up ArrayBuffer
jamesaxl has joined #pypy
<arigato> necaris: that's simply ffi.new("char **")
<arigato> it gets you one "char *" in memory, and returns a "char **" that points to it
squeaky_pl has joined #pypy
<necaris> arigato: thanks! so I should be able to then dereference it with `ffi.string(ptr)`?
<arigato> no, if you do ``info = ffi.new("char **")`` then info is of type "char **"
<arigato> but you can do ffi.string(info[0])
<arigato> i.e. first read the "char *" from memory
<necaris> aha, perfect, that works -- thank you arigato!
<necaris> so I can then later free it happily with `FooFree(info[0])` as well?
<nanonyme> arigato, I've occasionally difficulties remembering the dereference syntax despite having used CFFI a lot :/
<nanonyme> It's just somehow not entirely intuitive to me even though it's in a way logical
<dash> nanonyme: that's how I feel about C syntax in general
<arigato> yes, I see the point
<arigato> I considered but rejected the alternative that feels really magical:
<arigato> "ffi *p"
<arigato> (as a binary multiplication, of course)
<nanonyme> Yeah
jamesaxl has quit [Ping timeout: 240 seconds]
jamesaxl has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
sven__ has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
jamesaxl has quit [Ping timeout: 240 seconds]
necaris is now known as necaris[away]
jamesaxl has joined #pypy
necaris[away] has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
necaris has joined #pypy
DragonSA has joined #pypy
DragonSA has quit [Changing host]
DragonSA has joined #pypy
sven__ has quit [Ping timeout: 260 seconds]
brechtm has joined #pypy
pxwang has quit [Quit: Page closed]
<kenaan> rlamy PyBuffer 2a790c6d7613 /pypy/: fix translation
<bbot2_> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/4530 [ronan: force build, PyBuffer]
<bbot2_> Started: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/5855 [ronan: force build, PyBuffer]
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
brechtm has quit [Remote host closed the connection]
DragonSA has quit [Quit: Konversation terminated!]
brechtm has joined #pypy
kipras`away is now known as kipras
vkirilichev has joined #pypy
brechtm has quit [Ping timeout: 240 seconds]
jamesaxl has quit [Ping timeout: 240 seconds]
jamesaxl has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
danchr has joined #pypy
sven__ has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
marky1991 has quit [Read error: Connection reset by peer]
vkirilichev has quit [Remote host closed the connection]
vkirilichev has joined #pypy
jamesaxl has quit [Read error: Connection reset by peer]
vkirilichev has quit [Read error: Connection reset by peer]
jamesaxl has joined #pypy
vkirilichev has joined #pypy
tbodt has joined #pypy
<danchr> I posted a patch in a bug (seems too trivial for a pull request) — would be grateful if someone could take a peak
<arigato> ffi._embedding? you're not supposed to access internal attributes of the ffi object
<danchr> yeah, it's kind of horrible
<arigato> I'm not complaining (well, I am), but I would like to understand first
<danchr> just wondering; I see some references to building with commoncrypto rather than OpenSSL — would that work as a replacement?
jamesaxl has quit [Ping timeout: 260 seconds]
<arigato> never heard of it, so no opinion
<danchr> commoncrypto is Apple's own crypto library; the question is whether the SSL support really needs to be OpenSSL?
jamesaxl has joined #pypy
<arigato> I'd like to understand why this hack fixes openssl on pypy3, and why it's not needed for package X on pypy2/cpython
<danchr> afaict that code doesn't exist on pypy2
<danchr> s/on/in
<arigato> for openssl, no, but for any other project written using cffi
jamesaxl has quit [Read error: Connection reset by peer]
<danchr> my best guess is linker semantics: the Linux linkers don't care about undefined symbols in shared libraries, wheras the Darwin linker does
<tumbleweed> I'm pretty sure linux linkers do
<tumbleweed> it depends on how you use them
<arigato> you'll have to be more precise---if a shared library uses another shared library, then that's also undefined symbols, right?
jamesaxl has joined #pypy
<danchr> perhaps the issue is that the FFI modules are built as dylibs?
<danchr> I'm going from memory here, but I remeber fixing bugs like this years ago…
<arigato> ah
<arigato> so the issue would be .dylib versus .so, and the default (when not using embedding) is the wrong one?
<arigato> well, nevermind, the real problem is that this version of openssl uses the CPython C API when it's clearly recommended not to do so on cffi
<arigato> e.g. it makes pypy3's openssl unusable if pypy3 is translated without support for cpyext
marky1991 has joined #pypy
<arigato> so a more proper fix would be to fix lib_pypy/_cffi_ssl/_cffi_src/openssl/callbacks.py:
<arigato> to remove the "#include <pythread.h>" line
jamesaxl has quit [Read error: Connection reset by peer]
<arigato> I'll do that tomorrow if you don't get to it first
<danchr> AFAICT that doesn't fix the bug
<arigato> not *just* the include, of course
<arigato> changing the rest of the file to not call any PyXxx
<danchr> I've never actually written any cffi code…
<arigato> it's pure C
<arigato> we need to fix that pure C code so that it doesn't depend on PyXxx
<arigato> i.e. make it more pure C
<nanonyme> Will that also make it faster? :)
<arigato> yes, but I doubt it changes anything
<nanonyme> Maybe not but I like the idea that things become faster a side-effect of fixing them anyhow
<danchr> should the code just use pthread locks instead?
<arigato> yes, pthread mutexes or Windows critical sections, with an #ifdef
<arigato> no need for the more complicated kind of locks behind the PyThread_Xxx functions
* arigato zzz
<bbot2_> Failure: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/5855 [ronan: force build, PyBuffer]
arigato has quit [Quit: Leaving]
Rhy0lite has quit [Quit: Leaving]
capuk_ has joined #pypy
squeaky_pl has quit [Ping timeout: 260 seconds]
tbodt has quit [Ping timeout: 240 seconds]
vkirilichev has quit [Remote host closed the connection]
capuk_ has quit [Quit: Leaving]
vkirilichev has joined #pypy
vkirilichev has quit [Ping timeout: 240 seconds]
sven__ has quit [Remote host closed the connection]
capuk_ has joined #pypy
jacob22 has joined #pypy
jacob22_ has quit [Ping timeout: 260 seconds]
ronan__ has joined #pypy
ronan has quit [Ping timeout: 240 seconds]
nimaje1 has joined #pypy
nimaje1 is now known as nimaje
nimaje is now known as Guest18636
marky1991 has quit [Remote host closed the connection]
brechtm has joined #pypy
Guest18636 has quit [Ping timeout: 240 seconds]
john51 has quit [Ping timeout: 255 seconds]
john51 has joined #pypy
john51 has quit [Read error: Connection reset by peer]
john51 has joined #pypy
necaris has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
brechtm has quit [Remote host closed the connection]
brechtm has joined #pypy
brechtm has quit [Ping timeout: 240 seconds]
vkirilichev has joined #pypy
<bbot2_> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/4530 [ronan: force build, PyBuffer]
jcea has joined #pypy
adamholmberg has quit [Remote host closed the connection]
adamholmberg has joined #pypy
adamholmberg has quit [Ping timeout: 240 seconds]
vkirilichev has quit [Remote host closed the connection]
vkirilichev has joined #pypy
vkirilichev has quit [Ping timeout: 260 seconds]
kipras is now known as kipras`away
ronan__ is now known as ronan
necaris has joined #pypy
adamholmberg has joined #pypy
Civil has quit [Remote host closed the connection]
adamholmberg has quit [Ping timeout: 260 seconds]
Civil has joined #pypy
<kenaan> rlamy PyBuffer 2fbdc25dcb1b /: hg merge py3.5
Civil has quit [Remote host closed the connection]
Civil has joined #pypy