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
<bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/8355 [mattip: force build, py3.7]
<bbot2> Retry: http://buildbot.pypy.org/builders/own-win-x86-32/builds/2461 [mattip: force build, py3.7]
lritter has quit [Ping timeout: 240 seconds]
lritter has joined #pypy
<bbot2> Started: http://buildbot.pypy.org/builders/own-win-x86-32/builds/2462 [mattip: force build, py3.7]
mwhudson has quit [*.net *.split]
xcm has quit [*.net *.split]
<bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/5548 [mattip: force build, py3.7]
jcea has quit [Quit: jcea]
_whitelogger has joined #pypy
_whitelogger has joined #pypy
<bbot2> Failure: http://buildbot.pypy.org/builders/own-win-x86-32/builds/2462 [mattip: force build, py3.7]
<tumbleweed> PPA packages for 7.3.2 finally published :)
_whitelogger has joined #pypy
<mattip> we may have to do a 7.3.3 soon, to fix macos<=10.13 and ppc64 on all versions, venv on windows on py3.7
<mattip> .. or we could just tell people on MacOS<=10.13 to use 7.3.1
lritter has quit [Ping timeout: 246 seconds]
EWDurbin has quit [Ping timeout: 258 seconds]
DRMacIver has quit [Read error: Connection reset by peer]
DRMacIver has joined #pypy
EWDurbin has joined #pypy
<bbot2> Exception: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/5548 [mattip: force build, py3.7]
oberstet has joined #pypy
xcm has joined #pypy
mwhudson has joined #pypy
phlebas has quit [Ping timeout: 240 seconds]
phlebas has joined #pypy
ctismer has quit [Read error: Connection reset by peer]
Alex_Gaynor has quit [Read error: Connection reset by peer]
ctismer has joined #pypy
Alex_Gaynor has joined #pypy
Olorin_ has quit [Read error: Connection reset by peer]
Olorin_ has joined #pypy
<danchr> mattip: could you perhaps merge the release branch into default and py3.6? I think that should fix the Darwin build
YannickJadoul has joined #pypy
samth has quit [Ping timeout: 272 seconds]
samth has joined #pypy
<YannickJadoul> mattip: If I didn't mess up, that didn't seem to have fixed it :-(
<mattip> ok, thanks for trying
<mattip> danchr: any ideas how to get the xcode on the buildbot to build for MacOS<=10.13 ?
<danchr> are you getting the libstdc++ error?
<mattip> no, the "Symbol not found: ____chkstk_darwin" error when building for cffi
<mattip> ahh, hang on, my fix was not for cffi, I added it to the basic pypy exe
<danchr> my best guess is that something isn't setting MACOSX_DEPLOYMENT_TARGET
<danchr> I have a branch and MR just for that :)
<mattip> this is during the cffi build step, after translation
<mattip> so it needs the MACOSX_DEPLOYMENT_TARGET from sysconfig or so
<danchr> yeah, the MR does that
<danchr> essentially, it's saved as a translation option
<danchr> I just started four translations locally to test the changes
<danchr> two each in 2.7 & 3.7, and two targeting each of 10.7 and 10.15
<mattip> as far as I can tell, sysconfig should already be reporting 10.7
<danchr> which branch is this?
<mattip> default. See the travis-ci link above
<danchr> the one posted by YannickJadoul?
<danchr> anyway, got to go now
<mattip> maybe somehow cffi is not picking up that value?
<mattip> ok, bye.
<bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/5231 [danchr: force build of my MR, darwin-sendfile]
<mattip> fwiw, the command to try is
<mattip> cd lib_pypy; ../path/to/pypy _ssl_build.py
<mattip> which should verbosely (sic?) report the gcc command used, and it should include MACOSX_DEPLOYMENT_TARGET
<mgorny> mattip: when working on stdlib update, should i create the integration branch per instructions or just a topic and let you do the branches?
<YannickJadoul> mattip: Thanks for the quick help, as always :-) Should we update our tests to run on 10.14, then, or do you wish to try a few other things?
<mattip> mgorny: please do not use a topic branch, since we will want to run on the buildbots before merging
<mattip> you can use what I did around e888dd32ae03 to update py3.7 to 3.7.9
<mattip> I created a stdlib3.7.9 branch with a single meaningless commit so I could run the buildbots on it
<mattip> which then gave me a baseline for comparison to see how many new errors there were
<mattip> probably not needed for default, since there should be no errors in the first place
<mattip> so you can go ahead and update the vendor/stdlib-2.7 (accepting the new versions of the files)
<mattip> to sync it with CPython2.7,
<mattip> then create a branch off default, say stdlib-2.7.16, and merge vendor/stdlib-2.7 into it
<mattip> and then start running tests
<mattip> YannickJadoul: up to you. If your users are still on <=10.13, you might want to stick with the previous release until we work out what is wrong
<mgorny> ok
<mattip> whohoo, 15 minutes of glory on hacker new's front page
<YannickJadoul> mattip: Right, let's see what the others think :-)
<mattip> and even some nice comments too
<YannickJadoul> Congrats! :)
<mattip> I guess we _should_ migrate speed.pypy.org to show python3 benchmarks on the summary
oberstet has quit [Quit: Leaving]
<mgorny> mattip: should i mark wheel updates (i.e. from lib-python/2.7/ensurepip/_bundled/pip-18.1-py2.py3-none-any.whl to lib-python/2.7/ensurepip/_bundled/pip-19.2.3-py2.py3-none-any.whl) as copies or just remove/add?
<mgorny> i suppose copy doesn't give any real benefit here
<mattip> as for ensurepip specifically, I would prefer to stay with the versions we have anyway
<mattip> since the newer ones allows people to install manylinux2010/manylinxu2014 wheels
<mattip> although admittedly the chances of anyone creating modern wheels for python2.7 is quite slim
<mgorny> i suppose i can try undoing the update as part of integration branch
<mgorny> mattip: i'm a bit confused now. it seems that stdlib-2.7.16 branch hasn't been merged to master?
<mgorny> or at least in merge conflicts it seems that default is older in cProfile.py
<agronholm> where's the hacker news article?
<agronholm> nvm, found it
<mattip> ahh, we already have a stdlib-2.7.16 branch? So I am confused too
<mattip> someone (me?) must have started the job and never finished
<mattip> feel free to override it and start again if it is easier
* mattip bbl
mattip has quit [Remote host closed the connection]
<bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/5231 [danchr: force build of my MR, darwin-sendfile]
<bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/5232 [danchr: forced build of my other MR, macosx-deployment-target-2.7]
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
cjmcdonald has quit [Ping timeout: 260 seconds]
<bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/5232 [danchr: forced build of my other MR, macosx-deployment-target-2.7]
xqzhou has joined #pypy
<cfbolz> mattip: nice, I tweeted the two speed comments
YannickJadoul has quit [Remote host closed the connection]
YannickJadoul has joined #pypy
YannickJadoul has quit [Client Quit]
xqzhou has quit [Remote host closed the connection]
wilbowma has quit [Ping timeout: 256 seconds]
wilbowma has joined #pypy
cjmcdonald has joined #pypy
dnshane has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
dnshane has joined #pypy
_whitelogger has joined #pypy
_whitelogger has joined #pypy
epsilonKNOT has quit [Quit: ZNC - https://znc.in]
epsilonKNOT has joined #pypy
jcea has joined #pypy