cfbolz changed the topic of #pypy to: #pypy PyPy, the flexible snake | IRC logs: and | the pypy angle is to shrug and copy the implementation of CPython as closely as possible, and staying out of design decisions
mjacob has quit [Server closed connection]
mjacob has joined #pypy
mattip has joined #pypy
<mattip> hexology: what is the real question? Have you tried something and it failed?
ruth2345345 has quit [*.net *.split]
mjacob has quit [*.net *.split]
sthalik has quit [*.net *.split]
glyph has quit [*.net *.split]
marvin_ has quit [*.net *.split]
tazle has quit [*.net *.split]
jryans has quit [*.net *.split]
shodan45 has quit [*.net *.split]
xcm_ has quit [*.net *.split]
dmalcolm has quit [*.net *.split]
[Arfrever] has quit [*.net *.split]
mgorny has quit [*.net *.split]
atomizer has quit [*.net *.split]
pjenvey has quit [*.net *.split]
mwhudson has quit [*.net *.split]
agronholm has quit [*.net *.split]
pvital has quit [*.net *.split]
jinsun has quit [*.net *.split]
habnabit_ has quit [*.net *.split]
fish1 has quit [*.net *.split]
pbsds has quit [*.net *.split]
the_rat has quit [*.net *.split]
luckydonald has quit [*.net *.split]
graingert has quit [*.net *.split]
jerith has quit [*.net *.split]
samth has quit [*.net *.split]
cfbolz has quit [*.net *.split]
Alex_Gaynor has quit [*.net *.split]
krono has quit [*.net *.split]
nimaje has quit [*.net *.split]
Techcable has quit [*.net *.split]
exarkun has quit [*.net *.split]
ruth2345345 has joined #pypy
tazle has joined #pypy
marvin_ has joined #pypy
shodan45 has joined #pypy
jryans has joined #pypy
sthalik has joined #pypy
glyph has joined #pypy
[Arfrever] has joined #pypy
xcm_ has joined #pypy
cfbolz has joined #pypy
dmalcolm has joined #pypy
agronholm has joined #pypy
pvital has joined #pypy
mgorny has joined #pypy
atomizer has joined #pypy
habnabit_ has joined #pypy
jinsun has joined #pypy
pbsds has joined #pypy
pjenvey has joined #pypy
the_rat has joined #pypy
luckydonald has joined #pypy
mwhudson has joined #pypy
graingert has joined #pypy
krono has joined #pypy
Alex_Gaynor has joined #pypy
Techcable has joined #pypy
exarkun has joined #pypy
samth has joined #pypy
nimaje has joined #pypy
fish1 has joined #pypy
jerith has joined #pypy
mjacob has joined #pypy
rwb has quit [Max SendQ exceeded]
rowbee has joined #pypy
rowbee is now known as rwb
mgorny has quit [Quit: No Ping reply in 60 seconds.]
mgorny has joined #pypy
lehmrob has joined #pypy
<mgorny> weird rpython assertion:
<mgorny> unfortunately, i don't know how to proceed with debugging that further
<mgorny> i suppose i could try creduce on the .c extension
<mgorny> s/creduce/cvise/
<cfbolz> wow, that's impressive
<cfbolz> mgorny: did you try to run this with a pypy binary built with "make lldebug"?
habnabit_ has quit [Read error: Connection reset by peer]
habnabit_ has joined #pypy
<mattip> I see this in the backtrace
<mattip> memcpy(posi, PyBytes_AS_STRING(block), Py_SIZE(block))
<mattip> the memory of PyBytes_AS_STRING is messed up, or Py_SIZE(block) is bogus, or something else is off
<mgorny> cfbolz: no, just release gentoo build and the one from buildbot
<mgorny> mattip: so this could be an upstream bug?
<mattip> dunno. Probably not, it works on CPython
<mattip> storing bytes in a PyBytesObject is fragile, we have to keep the c-char array and the internal memory sychronzed
<mgorny> mattip: should i try creduce to build a minimal C repro?
<mgorny> i'm a bit worried it could yield some nonsense — but i suppose any nonsense triggering rpython assertion should be a good reproducer? xP
<mattip> not sure it is worth the effort: they have a cffi version that would work better on PyPy anyway
<mgorny> if you mean brotlicffi, that's a separate project
<mgorny> and unfortunately not all projects support both
nimaje has quit [Server closed connection]
nimaje has joined #pypy
Dejan has joined #pypy
lehmrob has quit [Ping timeout: 245 seconds]
lehmrob has joined #pypy
dmalcolm_ has joined #pypy
dmalcolm has quit [Ping timeout: 240 seconds]
lehmrob has quit [Ping timeout: 245 seconds]
exarkun has quit [Server closed connection]
exarkun has joined #pypy
<Alex_Gaynor> Does cffi use alloca or dynamic stack allocation in generated modules? (See CVE-2023-4039)
<mattip> I see some usages of alloca. It seems the CVE is not yet generally available to the public
<tumbleweed> /41/41
<Alex_Gaynor> has the details on the CVE
Techcable has quit [Server closed connection]
Techcable has joined #pypy
Dejan has quit [Quit: Leaving]
<hexology> mattip: the question is as i asked it. the context is that i installed the latest pypy 3.10 with pyenv, then created a venv, then did `pip install ...` of which pyarrow was included, it tried to build pyarrow from source, and compilation failed with some highly technical C++ errors that are beyond my knowledge to understand or diagnose
<hexology> so before i even tried to figure out the problem, i wanted to know if it was expected to work smoothly and something is weird on my end, or if it was never expected to work well, in which case i'd just continue using cpython and worry about it in future projects
lehmrob has joined #pypy
lehmrob has quit [Ping timeout: 245 seconds]