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)"
<kenaan>
stian math-improvements bbeb35d80f5f /rpython/rlib/rbigint.py: Undo this move, it is actully 50% slower for power of two calculations on single digit longs
yuyichao has joined #pypy
carljm has joined #pypy
adamholmberg has quit [Remote host closed the connection]
<saulll>
Hi, I have a pypy .whl installed on my debian 9 and want to generate the documentation using sphinx. Now in sphinx conf.py my module is imported and that fails. How can I make sphinx use the modules from pypy?
__jae__ has quit [Read error: Connection reset by peer]
adamholmberg has joined #pypy
<arigato>
saulll: please give more details
<arigato>
mattip: sorry, can't right now
adamholmberg has quit [Read error: Connection reset by peer]
adamholmberg has joined #pypy
<arigato>
saulll: maybe you're asking "if I run 'make html' then it starts sphynx using CPython but I'd like pypy"
<arigato>
saulll: then editing the Makefile, I guess?
<arigato>
saulll: there might be a more official way, but you are asking in the wrong place --- see sphynx's channel
pilne has joined #pypy
adamholm_ has joined #pypy
<saulll>
arigato: I have a .whl file created by colleagues. This has been built for pypy. I have acces to the documentation src. When calling sphinx-build to generate the documentation I receive this particular issue. ZipImportError: can't find module MyModule. This is because conf.py contains MyModule and MyModule is only in the site-packages of pypy. What do you propose me doing?
<arigato>
or maybe, from inside a virtualenv, 'make html' will pick the current python interpreter; then you only need virtualenv
<arigato>
if none of the suggestions above work, you should ask the sphynx people
exarkun has quit [Ping timeout: 240 seconds]
<saulll>
ok, ty
adamholmberg has quit [Ping timeout: 240 seconds]
exarkun has joined #pypy
<arigato>
if "sphynx-build" is a python script, then you can also try "pypy sphynx-build"
yuyichao has quit [Ping timeout: 248 seconds]
__Yiota has joined #pypy
yuyichao has joined #pypy
john51 has quit [Read error: Connection reset by peer]
john51 has joined #pypy
__Yiota has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
adamholm_ has quit [Remote host closed the connection]
adamholmberg has joined #pypy
<mattip>
arigato: np, maybe later
realitix has quit [Ping timeout: 240 seconds]
adamholm_ has joined #pypy
oberstet has quit [Ping timeout: 255 seconds]
adamhol__ has joined #pypy
adamholmberg has quit [Ping timeout: 252 seconds]
adamholm_ has quit [Ping timeout: 255 seconds]
adamhol__ has quit [Ping timeout: 255 seconds]
Guest50343 has quit [Ping timeout: 246 seconds]
oberstet has joined #pypy
__Yiota has joined #pypy
yuyichao has quit [Ping timeout: 260 seconds]
yuyichao has joined #pypy
marvin has joined #pypy
realitix has joined #pypy
marvin is now known as Guest31504
Guest31504 has quit [Ping timeout: 240 seconds]
exarkun has quit [Read error: Connection reset by peer]
__Yiota has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
marky1991 has joined #pypy
realitix has quit [Ping timeout: 252 seconds]
forgottenone has joined #pypy
exarkun has joined #pypy
<mattip>
bye
mattip has left #pypy ["bye"]
cstratak has quit [Ping timeout: 248 seconds]
marvin_ has joined #pypy
marky1991 has quit [Ping timeout: 248 seconds]
pilne has quit [Quit: Quitting!]
marky1991 has joined #pypy
cstratak has joined #pypy
__Yiota has joined #pypy
marvin_ has quit [Read error: Connection reset by peer]
<mjacob>
i asked in the #holyjit channel on mozilla, which now consists of two people: the main developer and me :)
<antocuni>
ah, the slides even mention that it's a tracing JIT
<antocuni>
my main concern is that rust might be a bit too low level to be able the same thing we do in RPython, but I might be wrong
<mjacob>
antocuni: i think that, except for garbage collection, rpython and rust are on a similar level
<Alex_Gaynor>
antocuni: what kinds of things are you thinking about?
adamholmberg has joined #pypy
<antocuni>
actually, I don't know the rust type system well enough; I thought it was closer to C than RPython, but if you say it's not, I just believe in it
<Alex_Gaynor>
It's more structured than C; e.g. you have Box<T> to say "heap ptr". And then like rpython if you look in corners there's lltypesystem for having raw ptrs with no rules :-)
<antocuni>
cool
<Alex_Gaynor>
I think the only big difference I can think of, in terms of the raw type system, is that rust has references to stack allocs, which RPython doesn't.
<mjacob>
which we may should add anyway :)
<antocuni>
the other big difference is the role of the GC. But maybe if can write it in a way which is transparent to HolyJIT
<Alex_Gaynor>
Perhaps; in Rust they are safe because of the lifetime tracking. In RPython they'd have to either be unsafe, or be limited to, e.g. not cross function call boundaries
<Alex_Gaynor>
antocuni: besides virtualizable, and fastpath in assembler, does any part of the JIT care about GC details?
<antocuni>
yes, but at the end of the day, if you are implementing JS, you need a GC somewhere
<Alex_Gaynor>
err, virtuals*
<Alex_Gaynor>
yes, you need a GC, but maybe it can just be yet-another-thing which gets inlined
<antocuni>
Alex_Gaynor: well, you need to emit memory barriers, for example
<Alex_Gaynor>
that's "just some code to be emitted" the only reason for the JIT to care is to make some optimizations easier I think?
<antocuni>
and you have to teach the JIT where are the roots on the stack
<Alex_Gaynor>
ah, indeed.
<antocuni>
I'm not saying you can't do it of course
<antocuni>
it's just different than what we do
alex_ has quit [Ping timeout: 240 seconds]
<mjacob>
maybe it's a good time to revive our plans to do a sprint / workshop at mozilla?
<Alex_Gaynor>
If I can be helpful, let me know. I work there these days (so does pjenvey)
<antocuni>
I am not sure how the two projects can cooperate, apart from a general "tell me what you did"
<antocuni>
but it might be funny, yes :)
<antocuni>
Alex_Gaynor: what do you do there?
<mjacob>
Alex_Gaynor: it would probably be in the paris office, though
<Alex_Gaynor>
I work on sandboxing for firefox security; no JITs for me :-)
<antocuni>
boring, HolyJit sounds more fun :)
<Alex_Gaynor>
:D
<mjacob>
rguillebert: what did you have in mind last time you considered a sprint at mozilla?
* antocuni
off
<antocuni>
see you, guys
<mjacob>
bye
<mjacob>
i could imagine that one day of presentations, one day of discussing things together and a few days of general pypy sprinting could work
<mjacob>
it wouldn't be the first time that a company or organization offers pypy a sprinting location without really benefitting from the sprint
<mjacob>
and in this case we would actually bring in some value :)
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
antocuni has quit [Ping timeout: 252 seconds]
alex_ has joined #pypy
tbodt has joined #pypy
<Alex_Gaynor>
mozilla does use pypy, so we benefit even if it's different people from people who write JITs :-)
exarkun has quit [Ping timeout: 252 seconds]
exarkun has joined #pypy
<rguillebert>
mjacob: they were eager to host us, no specific plans in mind
__Yiota has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<mjacob>
rguillebert: what was the main reason that it didn't happen in the end?
<rguillebert>
Couldn't find a date that worked for all of us I think
<rguillebert>
I don't remember who my contact at Mozilla Paris was though
rubdos has quit [Quit: WeeChat 1.9]
rubdos has joined #pypy
nunatak has joined #pypy
marky1991_2 has quit [Quit: Saliendo]
marky1991_2 has joined #pypy
marky1991_2 has quit [Remote host closed the connection]
marky1991 has joined #pypy
forgottenone has joined #pypy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
tbodt has joined #pypy
<kenaan>
rlamy py3.5 680d99dade82 /lib-python/3/ctypes/test/test_bitfields.py: @xfail a test: supporting bitfields in other-endian Structures is too messy
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
marky1991 has quit [Remote host closed the connection]
jimbaker_ has quit [Ping timeout: 248 seconds]
tbodt has joined #pypy
marky1991 has joined #pypy
jimbaker has joined #pypy
jimbaker has quit [Ping timeout: 260 seconds]
jimbaker has joined #pypy
exarkun has quit [Ping timeout: 246 seconds]
exarkun has joined #pypy
adamholmberg has quit [Remote host closed the connection]