cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end ) | use cffi for calling C | if a pep adds a mere 25-30 [C-API] functions or so, it's a drop in the ocean (cough) - Armin
<antocuni>
ronan: since you are the expert in pytest, app-level tests, etc: do you think it is possible to write a conftest which automatically collects pypy/module/_hpy_universal/_vendored/test and turns them into app-level tests automatically?
<antocuni>
basically, I'd like to do automatically what we are doing now manually in _hpy_universal/test/test_hpy*.py
<antocuni>
because e.g. I just realized that I forgot to port many tests files, and I fear that the situation is going to get worse over time
<mattip>
strange, using the same msvc compiler on default and py3.6 in the latest nightly builds
<mattip>
os.fdopen(10) crashes on default, raises an exception on py3.6.
<mattip>
Now if I could only remember how I made that work on py3.6
<mattip>
both are supposed to use the same error handling code in rposix (around exit_ and enter_suppress_iph )
<mattip>
background: windows wants to catch invalid fd access, and provides a per-process or per-thread hook to override the default behaviour
<mattip>
which is to crash the process
<mattip>
the hook seems to be be working on py3.6, but not on default
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
<nimaje>
the version detection in _curses_build.py is only in find_library() and that is only called if find_curses_dir_and_name() doesn't find anything, is that intended
mvantellingen has quit [Ping timeout: 246 seconds]
YannickJadoul has joined #pypy
<mattip>
nimaje: that sounds right, it is meant for people who have a non-standard ncurses build they want to use
<mattip>
and have set up environment varialbes so that compilation Just Works for them
<nimaje>
mattip: what is meant for people who have a non-standard ncurses build? find_library()? that is used on freebsd where ncurses is /usr/include/ncurses.h /usr/include/panel.h /lib/libncursesw.so.8 /usr/lib/libpanelw.so.5
<mattip>
hmm. On my Ubuntu, there is a /usr/include/ncursesw/ directory, and I think it works that way in the CentOS6-based buildbots as well
<mattip>
nimaje: a patch welcome to change the logic to better handle FreeBSD
jcea has joined #pypy
xcm has quit [Ping timeout: 256 seconds]
<mattip>
I guess I didn't think it through too much when changing it, sorry
<mattip>
cool. Seems reasonable. I will put it in, thanks
<nimaje>
ah, no that would use libncurses instead of libncursesw
<mattip>
so just switch the order 1, 3, 2, 4
<mattip>
grr, nevermind
<mattip>
where does FreeBSD put ncursesw.h ?
adamholmberg has joined #pypy
dddddd has joined #pypy
<nimaje>
there is non, seems like you should #define NCURSES_WIDECHAR 1 https://0x0.st/iBqb.txt
<mattip>
"if the standard feature test macros indicate that it should be defined" - any idea how to check those?
xcm has quit [Remote host closed the connection]
<mattip>
I guess we can always define it, then the check at the end of the file will always be true, whther we use ncursesw or not?
jvesely has joined #pypy
xcm has joined #pypy
<mattip>
I don't understand why there is ncursesw then
lritter has joined #pypy
Dejan has joined #pypy
<cfbolz>
ok, finally understood the reason for all the zipimport failures on 3.7! (pycs got an extra field in the header, zipimporter is still expecting the old layout)
adamholmberg has quit [Remote host closed the connection]
adamholmberg has joined #pypy
adamholmberg has quit [Remote host closed the connection]
adamholmberg has joined #pypy
marky1991 has quit [Ping timeout: 264 seconds]
<mattip>
whoohoo!
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
BPL has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
marky1991 has quit [Ping timeout: 256 seconds]
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
<arigo>
can't resist: snow, snow!
<antocuni>
arigo: is the ski resort still open?
<arigo>
no, everything is closed in switzerland
<antocuni>
ah, I read somewhere that Switzerland was trying to apply the same strategy as UK
<antocuni>
well, "good" then
asmeurer has joined #pypy
<mattip>
arigato: on win32, untranslated, it seems that _set_thread_local_invalid_parameter_handler does not work correctly
<mattip>
but _set_invalid_parameter_handler does work.
<mattip>
for instance in rpython/rlib/test/test_rwin32.test_get_osfhandle crashes, but if I change to the second function it works
<mattip>
is there anything about calling out external functions untranslated and threads?
<antocuni>
so, I am reading some cpyext code, when I encounter a function called "hack_for_result_often_existing_obj"
<antocuni>
and I think "uh, interesting, you never stop to learn about new hacks in cpyext, I wonder what is it"
<antocuni>
and then, guess who wrote it? Me :)
<cfbolz>
of course :-)
<mattip>
that's one for the quotes
<cfbolz>
antocuni: I recently managed to write some complicated logic twice. once last year on the 3.6 branch. then this year on default. discovered it when trying to merge
<antocuni>
cfbolz: this is even better :)
<cfbolz>
for some value of "better"
Dejan has quit [Quit: Leaving]
<mattip>
cfbolz: was 6ac071bf5627 (zipimporter) meant to be on default?
<cfbolz>
mattip: yep, found a segfault while reading code
marky1991 has quit [Ping timeout: 256 seconds]
inhahe_ has joined #pypy
zmt01 has joined #pypy
zmt00 has quit [Ping timeout: 246 seconds]
inhahe has quit [Ping timeout: 268 seconds]
wleslie has quit [Ping timeout: 250 seconds]
wleslie has joined #pypy
XorSwap has joined #pypy
marky1991 has joined #pypy
marky1991 has quit [Read error: Connection reset by peer]
marky1991 has joined #pypy
asmeurer has quit [Quit: asmeurer]
camelCaser has quit [Ping timeout: 240 seconds]
camelCaser has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
lritter has quit [Quit: Leaving]
adamholmberg has quit [Remote host closed the connection]