#pypy IRC log for Sunday, 2012-01-22

kenaan12amauryfa merge-2.7.2 11d74ec4c815b1 15/lib-python/2.7/: Install CPython 2.7.2 version of the std library00:02
kenaan12amauryfa merge-2.7.2 11725c5aa57558 15/lib-python/modified-2.7/: Merge 2.7/ recent changes to modified-2.7/00:02
amaury_Alex_Gaynor: 2.7.2 stdlib is merged00:02
Alex_Gaynoramaury_: cool, I guess we should hit the buildbots now00:02
Alex_GaynorI can do that00:02
Alex_Gaynor:)00:02
amaury_I did not change anything to pypy itself00:02
amaury_not even the version number00:03
Alex_Gaynorthat's fine I guess00:03
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/866 [12Alex, merge-2.7.2]00:03
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/1935 [12Alex, merge-2.7.2]00:03
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1204 [12Alex, merge-2.7.2]00:03
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1519 [12Alex, merge-2.7.2]00:03
Indecipherable (IceChat77@41.27.15.92) left #pypy.00:04
asmeurer (~asmeurer@dhcp-baca-230.resnet.nmt.edu) left irc: Ping timeout: 272 seconds00:07
papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) left irc: Quit: Computer has gone to sleep.00:11
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.00:15
mattip (~chatzilla@bzq-79-181-105-34.red.bezeqint.net) left irc: Quit: ChatZilla 0.9.88 [Firefox 9.0.1/20111221202246]00:16
Alex_Gaynoramaury_: btw, how did you merge 2.7/ into modified-2.7?00:16
amaury_before I started, I saved a diff between 2.7 and modified-2.700:17
amaury_after I drop the new version,00:17
amaury_I copy 2.7 onto modified-2.700:17
amaury_and apply the saved diff00:17
amaury_it's something that all VCS can do00:18
amaury_except hg and git00:18
amaury_so I did it manually00:18
Alex_Gaynorsmart00:18
amaury_a branch merge is not different00:18
Action: Alex_Gaynor is reading the full diff for fun00:18
arkx (~aku@eagleflow.fi) left irc: Read error: Operation timed out00:18
amaury_in hg you will only see CPython-related diffs00:19
Alex_Gaynorright00:20
amaury_differences between 2.7 and modified- are more interesting to me00:20
amaury_they show pypy differences00:20
amaury_many gc related items, and other "details"00:20
Alex_Gaynorright, well ATM I'm interested in what, if anything we'll need to change to merge this branch, so I'm seeing what changed between 7.1 and 2.700:20
Alex_Gaynorerr 7.2*00:21
Alex_Gaynoramaury_: while working on 3k I think we should try to make sure we *always* commit our changes to CPython so eventually we can kill -modified00:21
arkx (~aku@eagleflow.fi) joined #pypy.00:23
amaury_modified-3.2 is almost empty ATM00:23
Action: amaury_ -> bed00:23
Alex_Gaynorcool00:23
Alex_Gaynorgood night00:23
amaury_Alex_Gaynor: I'm afraid your tests will start in 3 hours00:24
Alex_Gaynoramaury_: that's fine I guess00:24
amaury_ok, see you00:24
Alex_Gaynorbye00:24
amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 276 seconds00:29
csenger (~user@dslb-084-058-150-061.pools.arcor-ip.net) left irc: Quit: ERC Version 5.3 (IRC client for Emacs)00:33
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds00:37
squiddy (~squiddy@f053084032.adsl.alicedsl.de) left irc: Quit: Leaving00:42
tilgovi (~randall@couchdb/developer/tilgovi) joined #pypy.00:49
ivan (~ivan@unaffiliated/ivan/x-000001) left irc: Quit: ERC Version 5.3 (IRC client for Emacs)00:56
ivan` (~ivan@unaffiliated/ivan/x-000001) left irc: Quit: ERC Version 5.3 (IRC client for Emacs)00:56
ivan` (~ivan@unaffiliated/ivan/x-000001) joined #pypy.00:57
ivan (~ivan@unaffiliated/ivan/x-000001) joined #pypy.00:57
nickik (~nick@217.150.234.222) left irc: Quit: Verlassend01:09
stakkars__ (~tismer@89.204.153.246) left irc: Quit: schnarch01:17
ousado (~ousado@unaffiliated/ousado) left irc: Ping timeout: 240 seconds01:25
ousado (~ousado@31-18-114-178-dynip.superkabel.de) joined #pypy.01:31
bbot23Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64-2/builds/3901:33
ousado (~ousado@31-18-114-178-dynip.superkabel.de) left irc: Read error: Operation timed out01:33
ousado (~ousado@31-18-114-178-dynip.superkabel.de) joined #pypy.01:35
bbot23Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/21201:36
papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) joined #pypy.01:36
ousado (~ousado@31-18-114-178-dynip.superkabel.de) left irc: Changing host01:40
ousado (~ousado@unaffiliated/ousado) joined #pypy.01:40
papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) left irc: Ping timeout: 248 seconds01:41
smithw (~ingsoc@189.38.230.73) joined #pypy.01:43
smithwHi, I installed pypy from the binaries on my Ubuntu 11.10 server and I get this error:http://pastebin.com/Y0vgwp10 . Any ideas on what might it be?01:46
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 252 seconds01:47
bbot24Failure: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/78301:48
nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services01:48
nedbat_ (~nedbat@python/psf/nedbat) joined #pypy.01:49
Tobu (~Tobu@2a01:e35:8a2c:b230:213:d3ff:fe85:5614) joined #pypy.01:52
Tobu (~Tobu@2a01:e35:8a2c:b230:213:d3ff:fe85:5614) left irc: Changing host01:52
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.01:52
driptonsmithw: try ./pypy -V to see what modules it's loading01:54
driptonalso maybe try strace ./pypy to see what file it's trying to load01:54
smithwdripton, ./pypy -V also gives the no such file or directory error, and as far as I can tell from strace output, it is trying to load the right file. do you want me to paste it too?01:56
driptonWhat does 'file ./pypy' say?01:57
driptonI'm wondering if you have 64-bit executable on 32-bit box01:57
smithwdripton, ./pypy: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, from '?', stripped01:58
smithwdripton, I might have the 32-bit exec on a 64-bit box, but I'm almost sure this is 32-bit01:59
dripton32-bit executable on 64-bit box is okay01:59
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/68602:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/33602:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/68302:00
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/34602:00
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/79702:00
driptonI just downloaded the pypy 1.7 32-bit tarball onto my Ubuntu 11.10 netbook, and untarred it into /opt, and it works fine.02:00
smithwno, it *is* a 64-bit box, this is news to me, let me try the 64-bit version02:00
driptonIn theory 32-bit exe on 64-bit box should work, but in practice it could be mixing 32- and 64-bit libs in the same process, which would break.  So I think switching to the 64-bit version might fix it for you.02:01
smithwdripton, yes, it did fix it. Thank you for your help!02:02
driptongreat.  You're welcome.02:02
timotimointeresting error message in that case02:03
timotimoshouldn't the dynamic linker have complained using its own name or something?02:03
driptonagreed that's a pretty bad error message02:03
timotimoyeah, just like you'd get "no such process" when you try to remove a network route :)02:04
DasIch_ (~DasIch@p3E99143F.dip.t-dialin.net) joined #pypy.02:05
DasIch (~DasIch@p3E99143F.dip.t-dialin.net) left irc: Ping timeout: 252 seconds02:08
Nick change: DasIch_ -> DasIch02:08
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 252 seconds02:26
mikefc (~mikefc@60-242-240-196.static.tpgi.com.au) left irc: Quit: mikefc02:29
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.02:40
justinpeel (~justinpee@c-76-23-44-72.hsd1.ut.comcast.net) left irc: Ping timeout: 245 seconds02:42
espes (~espes@123-243-221-248.tpgi.com.au) left irc: Ping timeout: 240 seconds02:45
espes (~espes@123-243-221-248.tpgi.com.au) joined #pypy.02:47
nedbat_ (~nedbat@python/psf/nedbat) left irc: Quit: Leaving02:53
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) joined #pypy.02:56
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) left irc: Remote host closed the connection02:58
Shanita (~John@osbk-4db17e97.pool.mediaWays.net) left irc: Ping timeout: 240 seconds03:04
Moku (~John@osbk-4db1688f.pool.mediaWays.net) joined #pypy.03:04
Dulak (~michael@unaffiliated/dulak) left irc: Remote host closed the connection03:15
bbot23Success: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/102203:23
tilgovi (~randall@couchdb/developer/tilgovi) left irc: Ping timeout: 240 seconds03:25
smithw (~ingsoc@189.38.230.73) left irc: Quit: War is Peace. Freedom is Slavery. Ignorance is Strength.03:41
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 252 seconds03:42
dripton (~dripton@pool-72-86-45-238.clppva.fios.verizon.net) left irc: Quit: Konversation terminated!03:46
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.03:49
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/866 [12Alex, merge-2.7.2]03:58
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/86703:58
mikefc (3cf2f0c4@gateway/web/freenode/ip.60.242.240.196) joined #pypy.04:10
mikefc__ (~mikefc@60-242-240-196.static.tpgi.com.au) joined #pypy.04:11
mikefc__ (~mikefc@60-242-240-196.static.tpgi.com.au) left irc: Client Quit04:11
justinpeel (~justinpee@c-76-23-44-72.hsd1.ut.comcast.net) joined #pypy.04:18
scitesy (~scitesy@d118-75-236-90.nap.wideopenwest.com) joined #pypy.04:26
scitesy (~scitesy@d118-75-236-90.nap.wideopenwest.com) left irc: Quit: Leaving04:41
bbot24Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/1935 [12Alex, merge-2.7.2]05:00
bbot2Started: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/193605:00
DasIch (~DasIch@p3E99143F.dip.t-dialin.net) left irc: Ping timeout: 240 seconds05:01
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.05:02
gtaylor (~gtaylor@108-196-160-156.lightspeed.tukrga.sbcglobal.net) joined #pypy.05:11
PiotrSikora (~none@nginx/adept/piotrsikora) left irc: Excess Flood05:18
PiotrSikora (~none@nginx/adept/piotrsikora) joined #pypy.05:19
papercrane (~papercran@c-76-103-172-115.hsd1.ca.comcast.net) joined #pypy.05:22
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) left irc: Read error: Connection reset by peer05:22
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) joined #pypy.05:23
justinpeel (~justinpee@c-76-23-44-72.hsd1.ut.comcast.net) left irc: Quit: Leaving05:31
nettok_ (~quassel@190.148.249.220) left irc: Remote host closed the connection05:39
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1519 [12Alex, merge-2.7.2]05:44
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/152005:44
papercraneHm, I'm seeing pypy be almost 2x slower than cpython with an async io test I'm running05:46
timotimoif you have a reproducable test case, a small one would be great, post it on the bugtracker, so it can be investigated and not forgotten about05:48
papercraneUnfortunately, it's while testing a simple gevent-like framework that uses greenlets to do asynchronous io&.05:49
papercranenot too much code but it's not exactly a simple script05:49
Da_Blitzthere is a slight issue iwth greenlets and the JIT at the moment 05:49
Da_Blitzthat prevents the jit from kicking in05:49
timotimooh, right05:50
papercraneah, that would explain it, then05:50
timotimothat would explain it05:50
Da_Blitzhas to do with swtiching int he middle of a trace/loop irc05:50
papercraneI'm also seeing an RPython assertion error having to do with asynchronous io when I run a bunch of tests05:51
Da_Blitzthat i would recomend reporting05:51
Da_Blitzhttp://bugs.pypy.org05:51
papercraneI tried to build a simple async io script to cause it but that goes on forever, so it appears to be related to the greenlet use as well05:52
papercraneok, I'll make a small tarball and post that05:53
whyking (~quassel@ip68-14-13-208.ri.ri.cox.net) left irc: Ping timeout: 272 seconds05:56
kAlug (~kalug@187.57.4.142) joined #pypy.05:58
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1204 [12Alex, merge-2.7.2]06:02
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/120506:02
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-64/builds/68306:02
stakkars (~tismer@p5DDB5C07.dip.t-dialin.net) left irc: Read error: Connection reset by peer06:12
stakkars_ (~tismer@p5DDB5C07.dip.t-dialin.net) joined #pypy.06:12
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/68606:21
timotimohuh, i can't seem to find any document detailing the changes from 2.7.1 to 2.7.206:23
kAlug (~kalug@187.57.4.142) left irc: Quit: Nettalk6 - www.ntalk.de06:27
mikefc"version bump"06:28
timotimooh, amazing06:29
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/152006:38
gutworth (~benjamin@dhcp-128-36-84-221.central.yale.edu) left irc: Ping timeout: 244 seconds06:40
gtaylor (~gtaylor@108-196-160-156.lightspeed.tukrga.sbcglobal.net) left irc: Quit: Konversation terminated!06:47
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 252 seconds06:48
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.06:49
tilgovi (~randall@couchdb/developer/tilgovi) joined #pypy.06:56
tilgovi (~randall@couchdb/developer/tilgovi) left irc: Remote host closed the connection06:57
CIA-1703papercrane 07roundup * 10#1016/RPython AssertionError while running async io greenlet test: (log message trimmed)06:59
CIA-17[new] I get:06:59
CIA-17RPython traceback:06:59
CIA-17File "jit_metainterp_compile.c", line 136, in force_now_106:59
CIA-17File "jit_metainterp_compile.c", line 200, in * 14https://bugs.pypy.org/issue101606:59
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/120507:03
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/34607:05
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Quit: DasIch07:45
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/33607:46
asmeurer (~asmeurer@dhcp-baca-230.resnet.nmt.edu) joined #pypy.08:01
bbot24Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-64/builds/79708:03
fijal (~fijal@196-210-199-8.dynamic.isadsl.co.za) joined #pypy.08:31
fijal (~fijal@196-210-199-8.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds08:43
squiddy (~squiddy@f053087141.adsl.alicedsl.de) joined #pypy.08:45
bbot23Success: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/86708:49
kvda (~kvda@124-169-134-221.dyn.iinet.net.au) joined #pypy.08:57
amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy.09:06
rguillebert (~rguillebe@89-40.5-85.cust.bluewin.ch) joined #pypy.09:07
kenaan12amauryfa merge-2.7.2 11ff154856cb9b 15/lib-python/conftest.py: List a test that was recently added, otherwise the test suite won't run :(09:17
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/868 [12amaury, merge-2.7.2]09:17
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1521 [12amaury, merge-2.7.2]09:18
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1206 [12amaury, merge-2.7.2]09:18
arigato (~arigo@89-40.5-85.cust.bluewin.ch) joined #pypy.09:23
arigatohi09:23
arigatoamaury_: bad you, you "hg push -f" 'ed09:24
amaury_arigato: yes, I created a new branch09:24
amaury_merge-2.7.209:24
arigatook, but default has still two heads09:24
amaury_huh?09:25
arigatothe command would rather have been "hg push --new-branch"09:25
lesshasteis there some way to get PIL to work with pypy?09:25
lesshasteI just get ImportError: No module named PIL09:25
amaury_arigato: sorry, I don't have this option09:26
arigatolesshaste: in PIL: "pypy setup.py install"09:26
lesshastearigato: oh I need to download PIL again09:26
arigatoamaury_: uh, I see, you're likely using an old hg then09:26
arigatolesshaste: well you need to compile from source, of course09:26
arigatoamaury_: I guess that "--new-branch" is a version of "--force" that still checks for two heads on the same branch09:27
lesshastearigato: ok.. is there an easy route for this sort of thing on ubuntu? Rather than going the module home page and trying to guess the right thing to get09:28
arigatolesshaste: I don't know, but I imagine getting the source is easy, maybe with easy_install or another tool09:28
lesshasteok thanks09:28
arigato"pypy easy_install pil" may just work?09:29
lesshastegot it to work09:29
lesshastethanks09:30
arigato:-)09:30
lesshastewas just testing was just testing to see what speed up you would get with http://paste.pocoo.org/show/538615/09:33
kenaan12amauryfa default 1101999b68d9cc 15/lib_pypy/datetime.py: merge heads09:33
lesshaste30s to 12s.. somehow I expected this simple code to get more of a speed up09:33
amaury_arigato: did I correctly merge heads?09:34
rguillebertthere's only one "default" head in my repo09:38
arigatoamaury_: ethanks09:39
arigatoyes :-)09:39
arigatolesshaste: already quite good, given the overhead of calling the CPython C extension module09:40
arigatoalso, did you test with the last night's build?09:40
arigatoI did yesterday fixes to improve another mandelbrot09:40
fijal (~fijal@196-210-199-8.dynamic.isadsl.co.za) joined #pypy.09:42
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.09:49
asmeurer (~asmeurer@dhcp-baca-230.resnet.nmt.edu) left irc: Quit: asmeurer09:49
lesshastearigato: oh! That's quite a coincidence :)09:50
lesshasteI will do that09:50
ronan (~ronan@host86-170-108-88.range86-170.btcentralplus.com) left irc: Ping timeout: 240 seconds09:51
bbot24Failure: 15http://buildbot.pypy.org/builders/own-linux-x86-32/builds/193609:51
fijalis sprint still on?09:52
arigatofijal: last morning of it :-)09:52
kenaan12arigo stm 11b8a1df61795a 15/pypy/: Add an argument to the callback invoked by stm_perform_transaction: a retry counter starting at 0.09:53
kenaan12arigo stm 11a18852e15aae 15/pypy/module/transaction/test/test_interp_transaction.py: Fix test_interp_transaction.09:53
fijalarigato: for what is worth, we have constantly people complaining about jti and stackless not being fast09:55
fijalis this enough of a motivation to fix it?09:55
arigatoI fear it will eventually be09:56
arigatobut I found yet another place that uses the asm frame pointer09:56
arigatothe other day09:56
arigatoI don't even remember where now :-(09:57
fijalI'm just saying it's not "an exotic feature noone uses"09:58
fijalapparently09:58
kenaan12rguillebert py3k 1199083d2bfef1 15/pypy/interpreter/: Fixes the combined use of kwonly arguments and default parameters09:58
arigatofrom my point of view there is first "write a gc for stm", "write jit support for stm", and "write stm for stackless", at which point I'll have to care again about stackless+jit speed :-)09:59
fijalhahaha10:00
fijalarigato: it would be very hard for me to blame you :-)10:00
arigatothank you :-)10:00
fijaldo we have any other release blockers except 2.7.2 merge?10:01
arigatowhat is merge-2.7.2 actually?10:02
arigatoah so far we have 2.7.110:03
arigatosorry10:03
fijala merge of 2.7.2 cpython stdlib10:03
fijalyes10:03
kenaan12rguillebert py3k 112457af5955f1 15/pypy/interpreter/test/test_interpreter.py: Add support for keyword arguments in the test suite, add a test for the order of kwonly arguments10:11
rguillebertamaury_, I fixed the case where there's only one keyword only argument but it doesn't look at the names of the variables so it depends on the dictionary order10:14
kenaan12fijal default 11444a1d6d447d 15/pypy/translator/c/src/profiling.c: fix newlines10:18
kenaan12fijal default 11eeb977421a88 15/lib_pypy/datetime.py: merge10:18
CIA-1703fijal 07roundup * 10#1014/numpypy sum() buggy with axis: [chatting] On Sun, Jan 22, 2012 at 12:10 PM, Dmitrey <dmitrey15@ukr.net> wrote: > On 01/21/2012 06:59 PM, Fijal wrote: >> >> Fijal<fijall@gmail.co ... * 14https://bugs.pypy.org/issue101410:19
amaury_rguillebert: did you write a failing test?10:20
rguillebertyes10:20
rguillebertamaury_, I don't have to put -s anymore when I put a pdb.set_trace() is it because of your patch ?10:22
amaury_no10:22
amaury_my patch just fixed some buggy imports10:22
amaury_(readline is disabled for now)10:23
rguillebertand I'm at interp-level so yes it can't be your patch10:23
rguillebertstrange10:23
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.10:27
kenaan12arigo stm 114ae4d737bd18 15/pypy/module/transaction/interp_transaction.py: Oups.10:29
kenaan12arigo stm 115ab80750fca1 15/pypy/module/transaction/: Add an explicit fifo queue implementation, instead of using list.pop(0).10:29
kenaan12arigo stm 1169e39e76c0d2 15/pypy/: Hacks to make checkmodule() work.10:29
csenger (~user@dslb-084-058-191-114.pools.arcor-ip.net) joined #pypy.10:33
CIA-1703Dmitrey 07roundup * 10#1014/numpypy sum() buggy with axis: [chatting] In today night build it's absent yet: >>>> N.sum([[0, 1], [0, 5]], axis=0) 6 * 14https://bugs.pypy.org/issue101410:36
fijal_ (~fijal@196-210-199-8.dynamic.isadsl.co.za) joined #pypy.10:37
fijal_arigato: how good would be the queue where you use two lists?10:37
fijal (~fijal@196-210-199-8.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds10:40
arigatofijal_: that might would work too10:40
JaRoel|4d (~jaroel|4d@2001:980:7133:1:4483:1e9a:552c:afb9) left irc: Remote host closed the connection10:40
JaRoel|4d (~jaroel|4d@sink.jaroel.nl) joined #pypy.10:40
arigatobut in this version it's really clear what can cause conflicts10:40
arigatophew, now two richards conflict with each other, and of course I don't know why10:44
fijal_you need hooks to figure out :)10:45
arigatoyes10:45
fijal_say jit abort hooks were done waaay too late10:45
arigato"the address 0x8ef6380 was read by transaction 1 but written by transaction 2"10:46
arigatothank you, what now? :-/10:46
fijal_which reminds me10:48
fijal_if we have a way to get back from memory location to the python object, we can do cool stuff10:49
arigatoah bah, there is at least the executioncontext.topframeref that would cause systematic conflicts10:49
fijal_can't you use something like gc.get_referrents?10:52
fijal_get_referrers10:52
fijal_some of those :)10:52
fijal_or to be more precise a similar hook10:53
arigatoyes, but note that it would be useless in my case10:53
fijal_ok\10:53
fijal_of course you need to get rid of all the systematic interpreter hooks10:53
fijal_hm10:53
fijal_interpreter sources of conflict10:53
arigatowhat breaks if we create one ExecutionContext for each thread?10:53
fijal_isn't that the default?10:54
Action: fijal_ thought ec is thread-local10:54
arigatoI'm talking about stm threads here, not user-visible ones10:54
arigatoreview time for the 82 callers of space.getexecutioncontext()? :-/10:55
kenaan12arigo stm 113c073f6abf1a 15/pypy/translator/stm/src_stm/et.c: print the numeric reason when aborting.10:56
fijal_ (~fijal@196-210-199-8.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds10:58
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1521 [12amaury, merge-2.7.2]10:58
lesshastearigato: tried pypy-c-jit-51616-5aa09e8483d3-linux which gives the same timings11:01
lesshastearigato: did you mean something even more recent than a nightly build?/11:01
arigatono11:01
lesshasteso im.putpixel is some annoying rffi call?11:02
amaury_28 failures in 2.7.2 branch11:04
kenaan12rguillebert py3k 110b136e560f89 15/pypy/interpreter/test/test_interpreter.py: The test was wrong...11:06
rguillebertamaury_, false alert, my test was wrong not the code...11:06
arigatolesshaste: indeed, it seems my hacks yesterday don't work11:07
arigatolesshaste: it's 10 times faster if you replace "a**2" with "a*a", but it should be done automatically now11:07
Action: arigato writes a test_pypy_c11:07
amaury_rguillebert: btw, I fixed the other failures in test_interpreter11:08
rguillebertthe prints and excepts ?11:08
amaury_yes11:09
amaury_there was a genuine bug in the compiler11:09
kenaan12amauryfa py3k 1152eaf273d8ed 15/pypy/interpreter/: Easy fixes in test_compiler11:09
kenaan12amauryfa py3k 11e27203d55679 15/pypy/interpreter/pyopcode.py: One more fix for test_compiler11:10
kenaan12amauryfa py3k 11e325e4d3227a 15/pypy/interpreter/astcompiler/: The exception handler target "except ValueError as exc" was always compiled as a global variable.  Test and fix.11:10
kenaan12amauryfa py3k 114ce0e0c33cd2 15/pypy/interpreter/test/test_interpreter.py: Fixes in test_interpreter11:10
arigatoaargh11:10
rguillebertamaury_, shouldn't it happen on default as well ?11:10
Action: arigato blames amaury_ very indirectly11:10
arigatothere was two heads on default, and amaury_'s head was more recent11:11
rguillebertah ah11:11
arigatobut that head didn't contain the latest changes done yesterday11:11
lesshastearigato: oh!11:12
lesshastearigato: that's quite a speed up :)11:12
arigato:-)11:12
lesshastearigato: goes from 12.5s to 5.5s on my computer 11:14
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) left irc: Read error: Connection reset by peer11:15
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-Ojit-no-jit-linux-x86-32/builds/868 [12amaury, merge-2.7.2]11:15
arigatolesshaste: ok, I'm testing by commenting out the im.* stuff11:16
arigatoso it likely takes something like 5s11:17
arigatothe "core" went down from 7.5s to 0.5s or something11:17
Telugodu (~rambaho@202.3.77.231) left irc: Ping timeout: 252 seconds11:21
amaury_arigato: sorry11:21
arigato:-)11:22
kenaan12arigo default 11937fb53e76ac 15/pypy/rpython/lltypesystem/module/ll_math.py: hg backout c5d041657831: I will move it to the stdobjspace instead, where it can be special-cased before we do all ...11:22
kenaan12arigo default 11f1fb181c13a3 15/pypy/objspace/std/floatobject.py: Special-case "x ** 2" here instead.11:22
Telugodu (~rambaho@202.3.77.231) joined #pypy.11:22
CIA-1703mattip 07roundup * 10#1014/numpypy sum() buggy with axis: 11:23
CIA-17Indeed, fromnumeric.py needs at the minimum this fix.11:23
CIA-17diff -r 5aa09e8483d3 lib_pypy/numpypy/fromnumeric.py11:23
CIA-17--- a/lib_pypy/numpypy/fromnumeric. ... * 14https://bugs.pypy.org/issue101411:23
Telugodu (~rambaho@202.3.77.231) left irc: Client Quit11:23
Telugodu (~rambaho@202.3.77.231) joined #pypy.11:24
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) joined #pypy.11:24
Telugodu (~rambaho@202.3.77.231) left irc: Read error: Connection reset by peer11:24
sacred_coder (~rambaho@202.3.77.231) joined #pypy.11:24
Telugodu (~rambaho@202.3.77.231) joined #pypy.11:25
sacred_coder (~rambaho@202.3.77.231) left irc: Client Quit11:26
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1206 [12amaury, merge-2.7.2]11:28
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1207 [12arigo]11:28
bbot213Exception: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1207 [12arigo]11:30
kenaan12arigo default 11e82f6594d309 15/pypy/module/pypyjit/test_pypy_c/test_math.py: Write a test_pypy_c checking that "f ** 2" gets us just a float_mul operation.  (Needs to check if the test really ...11:31
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1208 [12arigo]11:31
CIA-1703mattip 07roundup * 10#1014/numpypy sum() buggy with axis: 11:33
CIA-17There is work underway to correct the fromnumeric problems, in the mean time11:33
CIA-17perhaps try this patch, to completely remove it.11:33
CIA-17diff -r 5aa09e848 ... * 14https://bugs.pypy.org/issue101411:33
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Read error: Connection reset by peer11:33
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.11:33
kenaan12amauryfa merge-2.7.2 1170241d66bf5f 15/lib-python/modified-2.7/test/test_multibytecodec_support.py: Fix method used to find cjkencodings data files11:38
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1522 [12amaury, merge-2.7.2]11:38
lesshastearigato: aha11:42
lesshastearigato: thanks11:42
lesshastearigato: it's an interesting question if each call has a large overhead, whether this can be cleverly amortised11:43
lesshastearigato: particularly in the case where the return value of the function is not used11:44
arigatolesshaste: cpyext is not analyzed by the JIT at all so far, but that would be a bad idea if done uncarefully, I suppose11:45
lesshastearigato: ok.. makes sense11:45
arigatonot to mention that there is no chance to JIT the C code, of course11:45
lesshastearigato: right.. the problem is that I don't understand if the time is taken by a call overhead of the function being run itself11:46
lesshastearigato: do you have a feeling for this?11:46
lesshastearigato: ...  the time is taken by a call overhead *or* the function being run itself11:46
lesshastearigato: assuming it is the overhead of just doing a foreign function call I was trying to think how to minimise that cost11:47
arigatoI think that the overhead is generally big, whether to call the function or the function calling PyArg_ParseTuple() or other functions11:47
arigatothe overhead is not just the function call, but the conversions between CPython-like and PyPy objects11:47
lesshastearigato: ok.. so maybe there is some optimisation that can be done11:48
arigatoit requires some dict lookups in caches etc.11:48
arigatopossibly yes11:48
lesshastelookups can be made very fast if one is lucky :)11:48
lesshastewith some careful engineering11:48
lesshasteit is really great that it works at all of course11:49
lesshasteshedskin just gives up :)11:49
arigato:-)11:49
arigatothanks amaury_ :-)11:49
lesshaste (~lesshaste@87-194-206-189.bethere.co.uk) left irc: Quit: Leaving11:50
rguillebert (~rguillebe@89-40.5-85.cust.bluewin.ch) left irc: Quit: Quitte11:55
Action: arigato -> end sprint11:59
arigato (~arigo@89-40.5-85.cust.bluewin.ch) left irc: Quit: See you11:59
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Remote host closed the connection11:59
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.12:00
ronan (~ronan@host86-143-208-51.range86-143.btcentralplus.com) joined #pypy.12:04
voidspace (~anonymous@87-194-212-65.bethere.co.uk) joined #pypy.12:04
voidspace (~anonymous@87-194-212-65.bethere.co.uk) left irc: Changing host12:04
voidspace (~anonymous@python/psf/voidspace) joined #pypy.12:04
fijal_ (~fijal@196-210-150-33.dynamic.isadsl.co.za) joined #pypy.12:10
Nick change: fijal_ -> fijal12:11
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1522 [12amaury, merge-2.7.2]12:26
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/1208 [12arigo]12:31
kvda (~kvda@124-169-134-221.dyn.iinet.net.au) left irc: Quit: x___x12:37
nirbheek (~nirbheek@gentoo/developer/flyingspaghettimonster/nirbheek) left irc: Ping timeout: 240 seconds12:38
zain (~textual@bzq-79-182-196-50.red.bezeqint.net) joined #pypy.12:42
nedbat (~nedbat@python/psf/nedbat) joined #pypy.12:44
fijalfail12:49
fijalnice12:49
fijalwe have quite a few fatal rpython errors on the bugtracekrr :/12:51
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) left irc: Quit: pnguyen12:55
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Quit: DasIch12:58
mattip (c1a9466d@gateway/web/freenode/ip.193.169.70.109) joined #pypy.13:08
mattipfijal: ping13:09
Action: mattip apologize ahead of time, I only have a 10 minute break13:09
fijalmattip: pong13:10
fijalwhatsup?13:10
mattipI assume you saw the fix to the reshape/shape bug?13:10
fijalnope13:11
fijalI wasn't following closely13:11
fijalwhere did you check it?13:11
fijalnumpypy-shape-bug13:11
fijal?13:11
mattipyup.13:11
fijalI'm a bit worried - would that work with swapaxis?13:11
mattipit should, it no longer makes assumptions.13:12
fijalok13:12
mattipbut tests would help.13:12
fijalso why there is if on order?13:12
fijallike what does order change?13:12
mattipwell, it still assumes a memory layout for 'F' that is different than 'C'13:12
fijalyeah, but if you do T or swapaxis, your memory layout does not correspond to an order any more13:13
amaury_bah, issue with ABCMeta: http://paste.pocoo.org/show/538691/13:13
fijaldoes it?13:13
mattip'F' is layed out left-to-right, 'C' right-to-left13:13
fijalyes, I know13:13
fijalbut transposed 'C' is no longer layered that way13:13
mattipright. That's why calc_new_strides will fail more commonly after the fix.13:13
fijalso how about we fix it?13:13
fijalthere should be no assumptions13:13
mattipno, it needs to fail.13:14
fijalah13:14
fijalfail as in return None13:14
fijaland not fail as in explode?13:14
mattipthen when it fails, reshape copies and setshape bombs.13:14
fijalok, I'll review later13:14
mattipand all is good, since setshape should bomb.13:14
Action: mattip rambling a bit13:14
fijalok, cool13:15
fijalmattip: I'll review and merge later then13:15
fijalI have 171 unread pypy-svn mails :/13:15
mattipcalc_new_shape will allow switched axes to be split, but not combined.13:15
mattipcool.13:16
mattipjust wanted to take one thing off your list of todo, not add more. :)13:16
fijalhahaha13:17
fijalcool :)13:17
Action: mattip done rambling13:17
mattipbye now.13:17
mattip (c1a9466d@gateway/web/freenode/ip.193.169.70.109) left irc: Quit: Page closed13:17
fijalsee you13:17
Telugodu (~rambaho@202.3.77.231) left irc: Read error: Connection reset by peer13:27
kenaan12amauryfa merge-2.7.2 1124cea2c24401 15/pypy/: Respect __subclasscheck__ when rebinding a method to a more specific class. This change is needed to make Or...13:35
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) joined #pypy.13:37
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1523 [12amaury, merge-2.7.2]13:37
tito (~tito@pdpc/supporter/active/tito) left irc: Ping timeout: 244 seconds13:46
tito (~tito@pdpc/supporter/active/tito) joined #pypy.13:48
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.13:49
DasIch_ (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.13:49
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Ping timeout: 245 seconds13:53
DasIch_ (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Ping timeout: 255 seconds13:53
DasIch_ (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.13:58
Nick change: DasIch_ -> DasIch13:58
DasIch_ (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.14:00
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Read error: Connection reset by peer14:00
Nick change: DasIch_ -> DasIch14:00
nanonyme (nanonyme@unaffiliated/nanonyme) left irc: Remote host closed the connection14:00
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Quit: DasIch14:08
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) joined #pypy.14:09
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 248 seconds14:13
Arach (~arach@128-75-182-202.broadband.corbina.ru) left irc: Ping timeout: 276 seconds14:15
azanella (~azanella@189.6.80.131) joined #pypy.14:20
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) left irc: Remote host closed the connection14:21
Arach (~arach@128-75-172-73.broadband.corbina.ru) joined #pypy.14:27
gutworth (~benjamin@dhcp-128-36-73-23.central.yale.edu) joined #pypy.14:33
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.14:33
whyking (~quassel@ip68-14-13-208.ri.ri.cox.net) joined #pypy.14:35
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1523 [12amaury, merge-2.7.2]14:40
bolotov (~bolotov@27-10-179-94.pool.ukrtel.net) joined #pypy.14:42
DanKluev (~root@93-88-3-229-xdsl.vntc.ru) left irc: Ping timeout: 248 seconds14:48
Arfrever (~Arfrever@apache/committer/Arfrever) joined #pypy.14:54
ronan (~ronan@host86-143-208-51.range86-143.btcentralplus.com) left irc: Ping timeout: 240 seconds14:54
nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 276 seconds14:55
bolotov (~bolotov@27-10-179-94.pool.ukrtel.net) left irc: Quit: Leaving.15:00
DanKluev (~root@82-169-adsl.vntc.ru) joined #pypy.15:03
ronan (~ronan@host86-143-209-85.range86-143.btcentralplus.com) joined #pypy.15:07
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.15:08
dripton (~dripton@pool-72-86-45-238.clppva.fios.verizon.net) joined #pypy.15:26
nettok (~quassel@190.148.249.220) joined #pypy.15:26
elipsion (~elipsion@81-232-141-35-no217.tbcn.telia.com) left irc: Ping timeout: 252 seconds15:33
CIA-1703csenger 07roundup * 10#1015/Add build to run the benchmarks with python 2.7 tip: 15:37
CIA-17[chatting] I've extended the runner and saveresults scripts to accept parameters for url,15:37
CIA-17executable and project so it can be used to upload to m ... * 14https://bugs.pypy.org/issue101515:37
Nick change: csenger -> csenger|away15:44
elipsion (~elipsion@81-232-141-35-no217.tbcn.telia.com) joined #pypy.15:48
voidspace (~anonymous@python/psf/voidspace) left irc: Quit: voidspace15:49
setmeaway (~setmeaway@118.45.149.247) left irc: Read error: Connection reset by peer15:54
voidspace (~anonymous@87-194-212-65.bethere.co.uk) joined #pypy.15:54
voidspace (~anonymous@87-194-212-65.bethere.co.uk) left irc: Changing host15:54
voidspace (~anonymous@python/psf/voidspace) joined #pypy.15:54
voidspace (~anonymous@python/psf/voidspace) left irc: Client Quit15:58
nedbat (~nedbat@python/psf/nedbat) joined #pypy.16:18
arigato (~arigo@adsl-89-217-233-144.adslplus.ch) joined #pypy.16:24
kenaan12amauryfa merge-2.7.2 1189f238f17ecb 15/pypy/interpreter/: Undo the previous change, and just add "allow_override=True". This fixes the cases when an old-style class i...16:27
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1524 [12amaury, merge-2.7.2]16:27
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 272 seconds16:29
kenaan12arigo default 117cd209e0414e 15/pypy/module/pypyjit/test_pypy_c/test_math.py: fix the test.16:30
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 245 seconds16:34
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Quit: DasIch16:48
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.16:49
nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services16:49
nedbat (~nedbat@python/psf/nedbat) joined #pypy.16:50
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) joined #pypy.16:53
mattip (4fb769bd@gateway/web/freenode/ip.79.183.105.189) joined #pypy.16:57
nedbat (~nedbat@python/psf/nedbat) left irc: Read error: Connection reset by peer16:58
nedbat (~nedbat@python/psf/nedbat) joined #pypy.16:58
nedbat (~nedbat@python/psf/nedbat) left irc: Disconnected by services16:59
nedbat_ (~nedbat@python/psf/nedbat) joined #pypy.16:59
mattip (4fb769bd@gateway/web/freenode/ip.79.183.105.189) left irc: Client Quit17:01
mattip_ (~mattip@bzq-79-183-105-189.red.bezeqint.net) joined #pypy.17:01
Nick change: mattip_ -> mattip17:02
Nick change: nedbat_ -> nedbat17:14
jaffachief (~textual@host86-170-15-228.range86-170.btcentralplus.com) joined #pypy.17:22
Tobu (~Tobu@unaffiliated/tobu) left irc: Read error: Connection reset by peer17:30
rguillebert (5c5a172b@gateway/web/freenode/ip.92.90.23.43) joined #pypy.17:30
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.17:31
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1524 [12amaury, merge-2.7.2]17:35
rguillebert (5c5a172b@gateway/web/freenode/ip.92.90.23.43) left irc: Client Quit17:35
fijalarigato: we have 2-3 release blockers17:37
fijalwhich are hard JIT crashes17:37
fijalone is about virtualstate token not being correct17:37
fijalvirtualizable token, sorry17:37
fijalI'll make a list later17:37
CIA-1703fijal 07roundup * 10#1014/numpypy sum() buggy with axis: 17:38
CIA-17[resolved] Can you please not reopen this issue until it's really broken? Replying to mails17:38
CIA-17reopens it * 14https://bugs.pypy.org/issue101417:38
fijalhttps://bugs.pypy.org/issue99617:39
mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) left irc: Read error: Connection reset by peer17:40
voidspace (~anonymous@87-194-212-65.bethere.co.uk) joined #pypy.17:40
voidspace (~anonymous@87-194-212-65.bethere.co.uk) left irc: Changing host17:40
voidspace (~anonymous@python/psf/voidspace) joined #pypy.17:40
mcdonc (~mcdonc@ip72-209-213-54.dc.dc.cox.net) joined #pypy.17:40
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.17:49
Arach (~arach@128-75-172-73.broadband.corbina.ru) left irc: Quit: bye17:52
kenaan12alex_gaynor merge-2.7.2 11e5417ef18814 15/lib-python/modified-2.7/json/tests/: fix JSON tests17:53
mattipsetitem on a flatiter will repeat the input arguments. http://paste.pocoo.org/show/538844/ 17:54
mattipNon documented AFAIK. do we need to support it?17:54
mattipI guess it's easy enough, but just another feature to test and file bugs on.17:55
nettok (~quassel@190.148.249.220) left irc: Remote host closed the connection17:55
Arach (~arach@128-75-172-73.broadband.corbina.ru) joined #pypy.17:56
nedbat (~nedbat@python/psf/nedbat) left irc: Ping timeout: 245 seconds18:02
mattip (mattip@bzq-79-183-105-189.red.bezeqint.net) left #pypy ("Leaving").18:05
kenaan12alex_gaynor merge-2.7.2 11a0e53cbd4f7c 15/pypy/module/rctime/: Fix for time.mktime, -1 is a valid return value in some cases.18:09
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.18:10
kenaan12alex_gaynor merge-2.7.2 11f17c2e5a8629 15/pypy/objspace/std/: Change to the new exception for bytearray().pop()18:16
stakkars (~tismer@p5DCE205B.dip.t-dialin.net) joined #pypy.18:17
amaury_Alex_Gaynor: Hi18:22
Alex_Gaynoramaury_: hi!18:23
kenaan12amauryfa merge-2.7.2 119def34947826 15/pypy/rlib/: CPython issue11391: Fix a mmap crasher18:23
kenaan12amauryfa merge-2.7.2 11a49166fe832a 15/lib-python/modified-2.7/test/test_ssl.py: test_ssl: Correctly use the helper function to find data files18:23
kenaan12amauryfa merge-2.7.2 119f554debdd92 15/lib-python/modified-2.7/distutils/sysconfig.py: Fix our copy of "sysconfig.py", which I broke during the merge18:23
kenaan12amauryfa merge-2.7.2 1140000ebe74d7 15/pypy/objspace/std/: Fix edge cases in float.__mod__18:23
Alex_Gaynoramaury_: FYI I'm working on the first io failure ATM18:24
amaury_ok18:24
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) left irc: Remote host closed the connection18:24
amaury_I'm looking at the peepholer failure18:24
stakkars (~tismer@p5DCE205B.dip.t-dialin.net) left irc: Quit: schnarch18:24
Nick change: stakkars_ -> stakkars18:24
etrepum (~bob@71.21.40.181) left irc: Quit: etrepum18:25
Alex_Gaynoramaury_: you commented on the original issue FWIW18:28
amaury_which one?18:28
Tobu (~Tobu@unaffiliated/tobu) left irc: Ping timeout: 272 seconds18:28
Alex_Gaynorthe CPythone one18:28
Alex_Gaynorit has to do with the portability of pyc files18:28
amaury_I don't remember :)18:30
Alex_Gaynoramaury_: it's a narrow vs. wide build issue18:31
amaury_I understand now18:31
Alex_GaynorI guess it's not an issue for our pyc files?18:31
kenaan12alex_gaynor merge-2.7.2 1112b899970671 15/pypy/module/_io/: Added in the new behavior for RawIOBase.readall, if read() returns None then readall() also returns None18:32
kenaan12alex_gaynor merge-2.7.2 111119c3892bfc 15/: merged upstream18:32
amaury_win32 uses narrow builds18:32
Alex_Gaynorah right, do we tag our PYC files with that?18:32
mfoord (~voidspace@87-194-212-65.bethere.co.uk) joined #pypy.18:32
amaury_no18:32
amaury_we do have the issue, potentially18:33
voidspace (~anonymous@python/psf/voidspace) left irc: Quit: voidspace18:33
Nick change: mfoord -> voidspace18:33
voidspace (~voidspace@87-194-212-65.bethere.co.uk) left irc: Client Quit18:33
Alex_GaynorI suppose we should fix it then18:33
voidspace (~voidspace@87-194-212-65.bethere.co.uk) joined #pypy.18:33
voidspace (~voidspace@87-194-212-65.bethere.co.uk) left irc: Changing host18:33
voidspace (~voidspace@python/psf/voidspace) joined #pypy.18:33
Alex_Gaynorok I'm going to hit the buildbot and go for lunch18:36
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1525 [12Alex, merge-2.7.2]18:36
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 240 seconds18:39
fijalAlex_Gaynor: did yuo commit the numready thingie?18:43
fijalAlex_Gaynor, amaury_: awesome thanks18:44
fijalhttps://bugs.pypy.org/issue101618:48
fijalthis looks like a release blocker18:48
voidspace (~voidspace@python/psf/voidspace) left irc: Quit: Goodbye...18:54
mvt (~mvantelli@53530442.cm-6-4a.dynamic.ziggo.nl) left irc: Quit: Leaving18:57
fijalAlex_Gaynor: ping?19:00
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.19:01
CIA-1703fijal 07roundup * 10#567/Silence annrpython warnings for 1.5: [resolved] Done, annotator does not produce warnings any more (I think) * 14https://bugs.pypy.org/issue56719:05
CIA-1703fijal 07roundup * 10#580/Broken pipe error during sandbox installation?: [resolved] I think this is really working these days (or unreproducible) * 14https://bugs.pypy.org/issue58019:06
CIA-1703fijal 07roundup * 10#615/Errors when using urllib with SSL: [resolved] Works for me, probably improved stdlib version? * 14https://bugs.pypy.org/issue61519:07
CIA-1703fijal 07roundup * 10#617/range() slower than xrange(): Still valid, we should look at it (22 Jan 2012) * 14https://bugs.pypy.org/issue61719:09
fijalamaury_: https://bugs.pypy.org/issue28619:09
fijalcan you comment on the status of this?19:10
CIA-1703papercrane 07roundup * 10#1016/RPython AssertionError while running async io greenlet test: 19:10
CIA-17[chatting] I'm unable to reproduce with --opt=2 or --opt=3 (./pypy-c translate.py --19:10
CIA-17output=pypy-c.3 --opt=3 targetpypystandalone.py), so this se ... * 14https://bugs.pypy.org/issue101619:10
fijalAlex_Gaynor: can I close: https://bugs.pypy.org/issue65619:11
CIA-1703fijal 07roundup * 10#641/reading CSV files with csv module is much slower than CPython 2.6.6: This one is still pretty bad, pypy-c is ~3x slower than my CPython * 14https://bugs.pypy.org/issue64119:13
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.19:16
CIA-1703fijal 07roundup * 10#714/Translating on top of pypy sometimes screws up mandelbrot: [resolved] Failed to see that one in a while :) * 14https://bugs.pypy.org/issue71419:16
CIA-1703fijal 07roundup * 10#709/pypy 1.5 crashes and fails when compiling pypy on OSX: [resolved] Resolved long time ago? Please reopen if not * 14https://bugs.pypy.org/issue70919:17
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) joined #pypy.19:19
fijalwhat's bivab's email?19:21
fijalRhyolite: ping?19:21
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Ping timeout: 240 seconds19:21
Rhyolitefijal: pong19:21
Rhyolitefijal: pvt19:21
MostAwesomeDudeInteresting. Fedora carries PyPy 1.6. I did not know this.19:24
fijalMostAwesomeDude: I think they carry 1.7 these days19:25
mattip (~mattip@bzq-79-183-105-189.red.bezeqint.net) joined #pypy.19:26
amaury (~amaury_@46-127-23-192.dynamic.hispeed.ch) joined #pypy.19:26
MostAwesomeDudefijal: Maybe in Rawhide. I run stable though, and it's installing 1.6.19:26
MostAwesomeDudeI'm still pretty happy about it. Getting closer to world domination. :319:26
kenaan12arigo concurrent-marksweep 11d1c21fed7569 15/pypy/module/thread/ll_thread.py: Fix: conflict with global name 'error'.19:27
kenaan12arigo concurrent-marksweep 11edaf872702ab 15/pypy/module/thread/ll_thread.py: merge heads19:27
kenaan12arigo 32ptr-on-64bit 1135da68422201 15/: hg merge default19:27
kenaan12arigo 32ptr-on-64bit 116687f847a628 15/pypy/jit/backend/llsupport/rewrite.py: This part of the code lost during the merge belongs to rewrite.py. Save it there for now, commented out.19:28
kenaan12arigo 32ptr-on-64bit 1127de83186d13 15/pypy/rpython/lltypesystem/opimpl.py: fix.19:28
amaury_ (~amaury_@46-127-23-192.dynamic.hispeed.ch) left irc: Ping timeout: 252 seconds19:28
fijalamaury: ping19:28
amauryhi19:28
fijalamaury: https://bugs.pypy.org/issue28619:28
amauryI'm looking at it19:28
fijalcan you comment on the status?19:28
fijalcool19:28
amauryit's still valid I think19:29
amauryI got the same feeling some time ago19:29
fijalok19:29
fijalis this a release blocker or we don't care?19:30
fijalhow about https://bugs.pypy.org/issue73819:31
fijal?19:31
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) left irc: Read error: Connection reset by peer19:31
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) joined #pypy.19:32
CIA-1703fijal 07roundup * 10#726/django-cms tests fail using pypy 1.5 but pass under CPython 2.5, 2.6 and 2.7: 19:32
CIA-17This now fails with:19:32
CIA-17Error: The version option is deprecated. Read about the change on19:32
CIA-17http://pypi.python.org/pypi/djangorecipe/0.9919:32
CIA-17Can ... * 14https://bugs.pypy.org/issue72619:32
CIA-1703fijal 07roundup * 10#735/Unable to install Twisted via easy_install: [resolved] Fixed in twisted * 14https://bugs.pypy.org/issue73519:32
CIA-1703fijal 07roundup * 10#761/Cannot build an optional C extension in the Storm 1.8 package: 19:33
CIA-17[wontfix] Can anyone confirm/deny this? I'm going to close this ticket, if the problems19:33
CIA-17persist, please reopen * 14https://bugs.pypy.org/issue76119:33
CIA-1703fijal 07roundup * 10#762/Error trying to build a C extension in the zope.proxy 3.6.1 package: [wontfix] Closing as wontfix as we won't implement cl_dict * 14https://bugs.pypy.org/issue76219:34
fijalgarden garden garden19:34
CIA-1703fijal 07roundup * 10#790/SSL and bz2 lib paths are broken on RHEL 5: 19:36
CIA-17[wontfix] Closing as wontfix, linux does not support binary distributions. fortunately19:36
CIA-17fedora ships pypy, that hopefully sorts the problem. * 14https://bugs.pypy.org/issue79019:36
CIA-1703fijal 07roundup * 10#773/pyre2 fails import: [resolved] closing as it works * 14https://bugs.pypy.org/issue77319:37
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1525 [12Alex, merge-2.7.2]19:42
CIA-1703afa 07roundup * 10#738/utcfromtimestamp handles time before the epoch differently in pypy and cpython: [resolved] This was recently fixed with 235d8b8434a8 * 14https://bugs.pypy.org/issue73819:43
asmeurer (~asmeurer@dhcp-baca-230.resnet.nmt.edu) joined #pypy.19:49
fijal (~fijal@196-210-150-33.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds19:50
stakkars_ (~tismer@p5DCE205B.dip.t-dialin.net) joined #pypy.19:52
kenaan12amauryfa merge-2.7.2 11693b08144e00 15/pypy/interpreter/astcompiler/: Implement CPython issue5057: do not const-fold a unicode.__getitem__ operation which returns a non-BMP chara...19:57
kenaan12amauryfa merge-2.7.2 116e17d519d090 15/: hg merge default19:57
kenaan12amauryfa merge-2.7.2 11ead17cd9edbf 15/lib_pypy/_ctypes_test.c: Update _ctypes_test.c from CPython 2.7.219:57
kenaan12amauryfa merge-2.7.2 11c000c3332cb0 15/lib-python/modified-2.7/json/encoder.py: json: Rename some functions in our pure-python implementation: in CPython, encode_basestring() is supposed t...19:57
setmeaway (~setmeaway@118.45.149.247) joined #pypy.20:13
izibi (~izibi@unaffiliated/izibi) left irc: Quit: leaving20:17
stakkars__ (~tismer@p5DCE205B.dip.t-dialin.net) joined #pypy.20:32
lesshaste (~lesshaste@87-194-206-189.bethere.co.uk) joined #pypy.20:32
lesshastehi20:32
fijal (~fijal@196-215-121-137.dynamic.isadsl.co.za) joined #pypy.20:33
lesshastehi fijal 20:33
fijalhi20:33
lesshastehow's it going?20:34
fijalgood :)20:34
stakkars_ (~tismer@p5DCE205B.dip.t-dialin.net) left irc: Ping timeout: 240 seconds20:34
mattiplesshaste: does PIL work flawlessly in pypy or are there issues?20:35
stakkars_ (~tismer@p5DCE205B.dip.t-dialin.net) joined #pypy.20:35
lesshastemattip: seems pretty flawless.. just a little slow20:35
lesshastemattip: well.. very slow really20:35
voidspace (~voidspace@python/psf/voidspace) joined #pypy.20:35
lesshasteI was wondering what a plan could be with all this C in the python libraries20:35
Alex_Gaynorfijal: pong20:36
stakkars__ (~tismer@p5DCE205B.dip.t-dialin.net) left irc: Ping timeout: 245 seconds20:36
lesshasteI assume it makes no sense to write an rpython interpreter for C and just jit that :)20:36
antocuni_ (~antocuni@host155-125-dynamic.25-79-r.retail.telecomitalia.it) joined #pypy.20:36
MostAwesomeDudePIL sucks.20:37
MostAwesomeDudeI started working on fixing it up, but it's kinda difficult to handle 3000-line C files.20:37
lesshasteMostAwesomeDude: what were you trying to do with it?20:37
MostAwesomeDudelesshaste: Get it all pure-Python.20:38
lesshasteMostAwesomeDude: I mean were you converting it all to rpython20:38
lesshasteMostAwesomeDude: ah20:38
lesshasteMostAwesomeDude: this just seems an endless task20:38
lesshasteconverting all known python libraries from C to rpython20:38
MostAwesomeDudelesshaste: Well, no, there's a finite amount of PIL.20:38
MostAwesomeDudelesshaste: Actually, I was thinking about doing this for next PyCon.20:39
lesshasteMostAwesomeDude: right but a less finite amount of "all standard libraries commonly used"20:39
MostAwesomeDudeA talk about how I took various libraries with shitty inner code, and fixed them up.20:39
MostAwesomeDudeWell, I wouldn't touch Zope or Django. That cuts down a lot.20:39
lesshasteMostAwesomeDude: do you use any helpful conversion tools?20:39
lesshasteor to put it another way, would you like any helpful conversion tools :)20:39
MostAwesomeDudelesshaste: I just need unit tests, I guess. PIL is almost totally untested.20:40
lesshasteah20:41
lesshasteI am thinking of the whole of scipy for example20:41
lesshastedoes it make no sense to try to jit C ?20:41
lesshasteas in write an rpython interpreter for it20:41
Alex_GaynorNo, that would be a giant undertaking, the scope is just not commiserate with the benefit.20:43
MostAwesomeDudeSounds backwards. What do you do when you reach the Fortran sections?20:43
lesshasteMostAwesomeDude: :) panic20:43
lesshasteAlex_Gaynor: well.. ignoring the effort for a second.. would there actually be any benefit?20:44
Alex_GaynorI suppose20:44
lesshastethe only point would be speed as I can't tell by guessing if it would be any faster than an rffi call20:44
MostAwesomeDudeWell, here's my thinking: Python is terser than C.20:45
MostAwesomeDudeLots terser.20:45
lesshastego on...20:45
MostAwesomeDudeEspecially when you add in the overhead of all of the CPython API boilerplate.20:45
MostAwesomeDudeSo if I rewrite a bunch of PIL, I could make it smaller, *and* faster in PyPy.20:45
lesshasteyes20:45
Alex_GaynorWell, certain cases would be faster.20:46
Alex_GaynorIn the general case of C code, it would be slower20:46
Alex_Gaynorbut in theory C-API heavy things would be faster20:46
Alex_Gaynorit seems like a waste of time though20:46
lesshaste"CINT is written in C++ itself, with slightly less than 400,000 lines of code."20:46
lesshastethat's a C interpreter20:46
lesshastethat's a lot of lines of code :)20:46
lesshastec++ interpreter to be fair20:46
Alex_Gaynoramaury: hmm, I'm looking at these signals issues, it seems that you already added the checksignals() calls that antoine added in the commit that went with the tests20:47
lesshasteAlex_Gaynor: is it realistic for pypy devs to rewrite all libraries?  20:47
lesshastethey do seem to be doing a great job doing that :)20:47
Alex_GaynorNot for us, no, but I'm confident the community will do so as we get faster and better.20:47
MostAwesomeDudeAlex_Gaynor: Dude, if I don't do this, I'm gonna Twistify Ganeti. Would you believe they actually use the toy HTTP server in production?20:48
amauryAlex_Gaynor: yes, but there may be others that we completely missed20:48
amauryand are still needed20:48
lesshasteAlex_Gaynor: ok makes sense20:48
Alex_Gaynoramaury: indeed, I'm only looking at one commit ATM, I guess I'll go through and look for more antoine added20:48
Alex_Gaynoramaury: I wish we had a better story for CPython tests, I hate to copy them to module/_io/tests :/20:48
nedbat (~nedbat@python/psf/nedbat) joined #pypy.20:49
amauryI agree20:49
amauryI thought of just referencing them from our tests, by name20:50
amauryand run them untranslated20:50
amaurybut there are often tiny differences20:50
Alex_Gaynoramaury: true, that would be a good insentive to backport our fixes20:51
Alex_Gaynoror side-port I suppose20:51
Alex_GaynorMostAwesomeDude: heh, I'll believe anything these days20:51
MostAwesomeDudeAlex_Gaynor: I know a guy. He has bridges. Are you interested?20:52
Alex_GaynorMostAwesomeDude: you live in portland, there are tons of bridges!20:52
Action: Alex_Gaynor waits for fijal to make a joke about the JIT20:52
lesshasteAlex_Gaynor: well.. now I am intrigued by how small you can make a C interpreter :)20:52
MostAwesomeDudelesshaste: tcc is probably near that lower limit.20:52
lesshasteAlex_Gaynor: but that's OT20:53
lesshasteMostAwesomeDude: I was looking at http://code.google.com/p/picoc/20:53
kenaan12amauryfa merge-2.7.2 11c84666d936d2 15/lib-python/modified-2.7/test/test_descr.py: Allow both str.__add__ to raise TypeError or return NotImplemented.  IMO the point of the test is to check t...20:54
Rhyolitefijal: ping20:54
Alex_Gaynoramaury: I suppose we should forward-port that change to 3k?20:56
amauryyes, probably20:56
xorAxAx (~alexander@moinmoin/coreteam/alexander) left irc: Quit: Nobody expects the Spanish Inquisition!20:57
mitsuhiko (~mitsuhiko@ubuntu/member/mitsuhiko) left irc: Quit: ZNC - http://znc.sourceforge.net21:00
lesshaste (~lesshaste@87-194-206-189.bethere.co.uk) left irc: Quit: Leaving21:00
voidspace (~voidspace@python/psf/voidspace) left irc: Quit: Goodbye...21:05
mitsuhiko (~mitsuhiko@ubuntu/member/mitsuhiko) joined #pypy.21:06
fijalAlex_Gaynor: I want numready commited21:06
Alex_Gaynorfijal: pypy account or my personal one?21:07
fijalbuildbot maybe?21:08
Alex_Gaynorwell, I'm putting it in it's own repo21:08
Alex_GaynorI'll do it on the pypy one I guess21:08
fijalno, put it under the buildbot please21:08
Alex_Gaynorin the buildbot repo you mean?21:08
fijalyeah21:08
Alex_GaynorWhere in the repo?21:09
Alex_GaynorNew top-level dir?21:09
mikefc (3cf2f0c4@gateway/web/freenode/ip.60.242.240.196) left irc: Ping timeout: 264 seconds21:09
kenaan12alex_gaynor buildbot 114232c21cfab6 15/numready.py: Added the numready script.21:10
Alex_Gaynorfijal: ok added, feel free to move or refactor as you see fit21:10
fijalok21:11
fijalI'll21:11
arigato (~arigo@adsl-89-217-233-144.adslplus.ch) left irc: Quit: See you21:12
Alex_Gaynorfijal: thanks21:13
fijalnot *right now*21:13
fijalstarted tracker gardening :/21:13
Alex_Gaynorfijal: btw you can close that issue you linked I suppose (about max), we need to think some more about guards and such I guess21:13
fijalI want a way to like sort issues21:14
fijalI don't have one21:14
mattipsomeone put me out of my misery. shouldn't space.is_w work in this case http://pastie.org/3233012 ?21:15
mattipaccording to http://doc.pypy.org/en/latest/objspace.html#id21, I think so :(21:15
Alex_Gaynormattip: space.is_w is `0 is int`21:15
Alex_Gaynoryou want space.isinstance_w21:15
mattiphmm, couldn't find isinstance_w in the linked doc, but it appears copiously in numpypy code.21:16
mattips/copiously/freqeuntly/21:17
stakkars_ (~tismer@p5DCE205B.dip.t-dialin.net) left irc: Quit: schnarch21:17
xorAxAx (~alexander@ampleforth.srv.alexanderweb.de) joined #pypy.21:18
xorAxAx (~alexander@ampleforth.srv.alexanderweb.de) left irc: Changing host21:18
xorAxAx (~alexander@moinmoin/coreteam/alexander) joined #pypy.21:18
mikefc_ (~mike@60-242-240-196.static.tpgi.com.au) joined #pypy.21:20
kenaan12alex_gaynor merge-2.7.2 11b0b180ee94ea 15/pypy/: Fix for mmap when trying to specify an offset that's past the end of a file.21:26
Alex_Gaynoroof, bad me :(21:26
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) joined #pypy.21:31
amauryAlex_Gaynor: why?21:34
Alex_Gaynoramaury: I killed a ton of whitespace without noticing21:35
amauryas usual :;-)21:35
Alex_GaynorI'm getting better about it! (Or we have less whitespace :P)21:35
kenaan12amauryfa merge-2.7.2 110c697ef6b87f 15/pypy/: CPython Issue #12100: Don't reset incremental encoders of CJK codecs at each call to encode().21:35
amauryyes, because you touch every file :)21:36
Alex_Gaynorsome people would find that very impressive21:36
fijalAlex_Gaynor: we're not impressed here easily21:37
Alex_Gaynorfijal: too bad21:37
fijalAlex_Gaynor: can I make it write to /tmp?21:39
fijalI find python -c <a lot of stuff> grouse21:39
Alex_Gaynorfijal: fine with me21:39
Action: fijal is in the process of adding genshi benchmarks21:40
fijalwell21:40
Alex_Gaynorfijal: actually I have a better idea21:40
Alex_Gaynorfijal: make it two scripts :)21:40
fijalhodgestar is, I'm just watching over his shoulder21:40
Action: Alex_Gaynor away21:42
jaffachief (~textual@host86-170-15-228.range86-170.btcentralplus.com) left irc: Quit: Textual IRC Client: http://www.textualapp.com/21:49
jaffachief (~textual@host86-170-15-228.range86-170.btcentralplus.com) joined #pypy.21:51
kenaan12hodgestar benchmarks 11fad651ead401 15/: Add Genshi XML and text template benchmarks.21:55
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) left irc: Read error: Connection reset by peer21:57
`fox` (~fox@host19-111-dynamic.20-79-r.retail.telecomitalia.it) left irc: Quit: I didn't try, I succeded (cit.)22:00
Alex_GaynorHodgestar: so are we fast or slow at it?22:00
HodgestarAlex_Gaynor: About 2-3x faster.22:00
Alex_Gaynorcool22:00
pjenveyAlex_Gaynor - not that that script is that important anyway but you're the one who said it was frightening =P22:01
fijalAlex_Gaynor: a = numpy.array([1, 2, 3, 4]); a.strides = [4,] a[1] = 1222:01
Alex_Gaynorpjenvey: it is frightening, the real correct solution is to split it into 4 files or so :)22:01
Alex_Gaynorfijal: can we not implement writing to strides, it seems like a terrible idea :)22:01
fijalAlex_Gaynor: yes22:02
fijalAlex_Gaynor: I can't think about a usage22:02
fijalactually22:02
Alex_Gaynorpjenvey: rather than make one file do double duty based on a flag, just make it two files :)22:03
fijalpjenvey: two files sound much saner22:03
mattipit goes along with another bad idea: object dtypes.22:03
mattips/it/writable strides/22:04
fijaloriginally it was source because yuo have numpy vs numpypy22:04
fijalmattip: speaking about your fixes (which I did not have time to review)22:04
mattip?22:04
mattipstatus report: flatiter is still wip, reshape should be finished and waiting for review.22:05
fijalI wonder22:05
etrepum (~bob@71.21.40.181) joined #pypy.22:05
fijalwhat's the difference betwenn 'C' order and 'F' order22:05
fijalor more precisely22:05
fijalisn't transposed 'C' just 'F'?22:05
mattipyes and no.22:06
fijalok?22:06
mattipeventually we want to support swapaxis,22:06
pnguyen (~pnguyen@cpe-70-112-214-119.austin.res.rr.com) joined #pypy.22:06
fijalyes, precisely22:06
fijaland this is something that's neither22:06
mattipwhich would make things really confusing.22:06
fijalI'm just not understanding the if22:06
fijalshould there be any?22:06
mattipyes.22:07
fijalok, why?22:07
mattip'F' order means that stride[0] is smaller than stride[-1] for a contiguous array.22:07
fijalyes ok22:07
mattipbut for 'C' that's not true22:07
fijalnor for transposed22:08
fijalso what's the difference between transposed C and F?22:08
fijalI'm just repeating my question22:08
Action: mattip typing slowly, sorry22:08
mattipso when you get to reshape, it's important.22:08
Action: mattip talking in circles.22:09
fijalyes22:09
fijalwhat's the difference between reshaping F and transposed C22:09
fijalI don't see any22:09
fijalwould you iterate differently?22:09
mattiptransposed 'C' looks like 'F' but it's not contigous. That's the important thing.22:09
fijalAlex_Gaynor: I added contributors to pushers to pypy-benchmarks btw22:09
fijalok22:09
mattipso you can't just reshape an array based on strides, layout is important too.22:10
amauryAlex_Gaynor: actually the issue with _io is not about signals22:11
Alex_Gaynoramaury: einter?22:11
amauryyes22:11
mikefc (9665bcc5@gateway/web/freenode/ip.150.101.188.197) joined #pypy.22:11
amaury_trap_eintr in bufferedio.c22:11
Action: Alex_Gaynor got lazy and didn't implement it :/22:12
mattipfwiw, you do not iterate differently, but you must reshape differently.22:13
mattipsince iterators can jumpy around the memory layout, but to reshape you need contiguous memory.22:14
fijal (~fijal@196-215-121-137.dynamic.isadsl.co.za) left irc: Ping timeout: 240 seconds22:14
amauryAlex_Gaynor: I've I fix maybe22:14
amauryfixed it I mean22:14
Action: mattip mubling to myself "I even added a test to prove my claims"22:16
ronan (~ronan@host86-143-209-85.range86-143.btcentralplus.com) left irc: Ping timeout: 240 seconds22:18
ronan (~ronan@host86-172-60-8.range86-172.btcentralplus.com) joined #pypy.22:32
kenaan12amauryfa merge-2.7.2 11311b0271907b 15/pypy/module/_io/: Buffered IO retries read() and write() when the raw stream raises errno=EINTR.22:33
bbot2Started: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1526 [12amaury, merge-2.7.2]22:33
bgola (~bgola@c951826e.virtua.com.br) left irc: Ping timeout: 240 seconds22:36
Nick change: csenger|away -> csenger22:37
fijal (~fijal@196-210-150-33.dynamic.isadsl.co.za) joined #pypy.22:43
fijalmattip: ok, how about we have an attribute or a function that can determine "is_contiguous"?22:44
mattipI had a branch like that once22:45
mattipthe thing was, for reshape it doesn't HAVE to be contiguous.22:46
neunon (~snoonan@unaffiliated/neunon) joined #pypy.22:46
CIA-1703papercrane 07roundup * 10#1016/RPython AssertionError while running async io greenlet test: 22:46
CIA-17Running with --opt=0 (./pypy-c translate.py --22:46
CIA-17output=pypy-c.0 --opt=0 --gc=minimark targetpypystandalone.py) also did not22:46
CIA-17crash. Running with ... * 14https://bugs.pypy.org/issue101622:46
mattipif all you're doing is splitting a dimension, or combining two dimensions, then copy is unecessary22:46
fijalok22:47
mattipThat's why I calculate steps, and at each dimension change make sure the steps work out OK.22:47
fijalok, cool22:48
mattipI agree it's complicated and maybe over the top, 22:48
mattipbut it should save some copying here and there.22:48
mattiplots of tests are a good thing.22:48
fijalyes :)22:48
fijalwe should have more22:49
mattipso I put lots of them into test_calc_new_strides22:49
CIA-1703dripton 07roundup * 10#996/sqlite3 CREATE TABLE always fails: I think I have a fix for this. Pull request soon. * 14https://bugs.pypy.org/issue99622:50
mattipbut it's too complicated there to actually see what's going on.22:50
Action: mattip wondering how to stress test this stuff.22:51
mattipI wonder how numpy tests their stuff?22:52
Alex_Gaynornot especialyl well from what I've seen22:53
mattipthe strides/shape/reshape/swapaxis works flawlessly.22:53
mattipmaybe there is a matlab or mathematica test suite?22:55
driptonAnyone have a suggestion for testing changes to _sqlite3.py without translating?22:57
Alex_Gaynordripton: you don't need to, just ./pypy-c -mtest.test_sqlite322:57
Alex_Gaynorthat runs the tests22:57
driptonAlex_Gaynor: thanks, I'll try that22:58
bgola (~bgola@c951826e.virtua.com.br) joined #pypy.22:58
driptonAlex_Gaynor: that works but test_sqlite3.py only contains the one test I just added.  I was hoping for more regression tests.22:59
fijaloh22:59
fijalmattip: it's gonna be a mess to merge :/22:59
Alex_Gaynordripton: ah sorry -mtest.test_sqlite22:59
Alex_Gaynorthat'll run all the sqlite testws22:59
DasIch_ (~DasIch@p4FFDCB0C.dip.t-dialin.net) joined #pypy.22:59
bbot2Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/21323:00
bbot2Started: 15http://buildbot.pypy.org/builders/own-macosx-x86-32/builds/78423:00
bbot2Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-32/builds/102323:00
bbot2Started: 15http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64-2/builds/4023:00
mattipfijal: can I do something to help?23:00
fijalmattip: I think they're just carefully developing this over years23:00
fijalnot making it very fast23:00
rguillebert (~rguillebe@2a01:e34:eea7:c690:21f:c6ff:fe12:4dee) joined #pypy.23:01
fijalhttp://paste.pocoo.org/show/539006/23:01
fijalmattip: yes, answer how the merge of this looks like?23:01
driptonAlex_Gaynor: that worked, but since I'm using the pypy-c I built before my latest changes, am I actually testing my changes to lib_pypy/_sqlite3.py?23:01
Alex_Gaynordripton: If that pypy-c is in the same dir as your pypy checkout, yes23:02
rguilleberthi23:02
fijalskip("Those are broken, fix after removing Scalar")23:02
fijalmattip: what does that mean?23:02
Alex_Gaynorfijal: you can't reshape a scalar ATM, I suppose23:02
mattipsheez, you yung'uns type fast. Slow down23:03
fijalI don't think a is scalar there23:03
driptonAlex_Gaynor: thanks, that gives me much more confidence that my fix didn't break something else.23:03
Alex_Gaynordripton: that's what tests are for :)23:03
driptonyep23:03
mattipthe paste: take the bottom lines with the order argument.23:03
fijalmattip: why the top part has more tests?23:03
DasIch (~DasIch@p4FFDCB0C.dip.t-dialin.net) left irc: Ping timeout: 244 seconds23:03
Nick change: DasIch_ -> DasIch23:03
fijalor are they not that important?23:04
mattipI'm trying to recall why the test got involved with Scalar. hang on.23:04
fijalmattip: if you merge default into branch you can not answer my silly questions :)23:04
mattipI dunno why I got rid of the alast two test of the paste. Maybe cuz they were too simple.23:06
fijalmattip: I'm a bit worried because it seems the branch has net test removal :)23:07
antocuni_ (~antocuni@host155-125-dynamic.25-79-r.retail.telecomitalia.it) left irc: Ping timeout: 252 seconds23:07
fijalas in disables 3 asserts and removes 223:07
mattipwell, should I put the two tests back in?23:08
fijalI was kind of expecting you would write new which show buggy behavior and they fail on default but work on the branch23:08
fijaland yes, put the two back in23:08
fijaland what is this skip doing is another question23:08
mattipOk, gimme some time, everything here is slow, including me.23:09
fijalmattip: np23:09
fijalmattip: I'm just giving some feedback23:09
CIA-1703dripton 07roundup * 10#996/sqlite3 CREATE TABLE always fails: 23:09
CIA-17Here's a pull request with the fix (and some minor cleanups to _sqlite3.py)23:09
CIA-17https://bitbucket.org/pypy/pypy/pull-request/23/fix-_sqlite3py-fix-f ... * 14https://bugs.pypy.org/issue99623:09
mattipcool. You and Alex_Gaynor move so fast it makes me feel old. :)23:10
azanella (~azanella@189.6.80.131) left irc: Read error: Operation timed out23:11
Tobu (~Tobu@unaffiliated/tobu) left irc: Remote host closed the connection23:12
fijalAlex_Gaynor: the CSS of occupying only small fraction sucks23:13
Alex_Gaynorfijal: what?23:14
fijalI want more columns than one23:15
Alex_Gaynorok, perhaps multiple columns with shitespace in between?23:15
Alex_Gaynorwhitespace*23:15
Action: fijal goes for shitspace23:16
etrepum (~bob@71.21.40.181) left irc: Quit: etrepum23:18
CIA-1703dripton 07roundup * 10#934/subprocess hangs in 1.7 (but not in 1.6 or 1.5): (log message trimmed)23:19
CIA-17pypy 1.7, on Linux x86-32. Sometimes the process hangs. Sometimes I get a23:19
CIA-17"build failed" like23:19
CIA-17=========================23:19
CIA-17... * 14https://bugs.pypy.org/issue93423:19
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.23:25
mattipfijal: the problem I ran into when working on shape was what to do when a dimension is removed and we end up with a vector.23:29
mattipis it's size (1,105) or (105,)23:29
mattips/size/shape/23:29
mattipI think I can do better just now, but not tonight. It's too late already.23:30
mattipon that branch, this is the test I added that fails (did not raise an exception) on default:23:32
mattipa = array(range(6)).reshape(2,3).T; a.shape = 623:32
mattipOn the shape branch it does raise an AttributeError, as it should ( and numyp does too)23:33
amaurySometimes when I refresh the page http://buildbot.pypy.org/waterfall23:35
amauryI get strange text like this: http://paste.pocoo.org/show/539016/23:35
kvda (~kvda@124-169-134-221.dyn.iinet.net.au) joined #pypy.23:36
amauryit's transient, when I refresh again the usual waterfall appears23:36
Niedar (gfghfghf@host-173-230-2-23.vanodod.clients.pavlovmedia.com) joined #pypy.23:36
CIA-1703ita 07roundup * 10#934/subprocess hangs in 1.7 (but not in 1.6 or 1.5): 23:36
CIA-17Pypy 1.7 on Linux x86-64: same thing here, the main process can hang, or a few23:36
CIA-17child processes will report the exit code -11 too * 14https://bugs.pypy.org/issue93423:36
mattipamaury: I think it happens when the page is being updated, they're missing a threading lock on the filesystem.23:37
mattipI've seen it too.23:38
CIA-1703dripton 07roundup * 10#934/subprocess hangs in 1.7 (but not in 1.6 or 1.5): 23:38
CIA-17It behaves differently in a pypy-c built today. (Still x86 Linux.) Now the23:38
CIA-17attached test waf build succeeds some of the time, and makes it to 5 ... * 14https://bugs.pypy.org/issue93423:38
Alex_Gaynoramaury: I assume it's a buildbot bug23:40
bbot24Failure: 15http://buildbot.pypy.org/builders/pypy-c-app-level-linux-x86-32/builds/1526 [12amaury, merge-2.7.2]23:40
Tobu (~Tobu@unaffiliated/tobu) left irc: Remote host closed the connection23:40
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.23:45
Tobu (~Tobu@unaffiliated/tobu) left irc: Remote host closed the connection23:46
kenaan12fijal buildbot 11ba9d28fbab20 15/numready.py: some work on numready23:46
fijalAlex_Gaynor: have an opinion23:46
fijalmattip: I fail to see added tests though23:47
Alex_Gaynorfijal: probably, I'm doing 10 other things at the moment though, will reivew lter23:47
fijalmattip: where did you add it?23:47
fijalAlex_Gaynor: just have a quick look at the page23:47
fijalAlex_Gaynor: but fine23:47
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) joined #pypy.23:48
Alex_Gaynoris it online somwhre?23:48
setmeaway (~setmeaway@118.45.149.247) left irc: Ping timeout: 240 seconds23:49
setmeaway2 (~setmeaway@118.45.149.247) joined #pypy.23:49
fijalAlex_Gaynor: hg pull -u pypy numready.py23:49
fijalno23:49
setmeaway2 (~setmeaway@118.45.149.247) left irc: Read error: Connection reset by peer23:49
setmeaway (~setmeaway@118.45.149.247) joined #pypy.23:49
Tobu (~Tobu@unaffiliated/tobu) joined #pypy.23:50
Dulak (~michael@unaffiliated/dulak) joined #pypy.23:51
Alex_Gaynorlooks ok I guess23:51
cocoatomo (~cocoatomo@p2118-ipbf901souka.saitama.ocn.ne.jp) left irc: Remote host closed the connection23:54
mattipfijal: test_set_shape, line 391, last test23:55
fijalah23:55
fijalmattip: ok, I see23:55
fijalmattip: you made a failing test on default and then a branch that fixed it?23:55
fijalAlex_Gaynor: the word you were looking for is "awesome"23:56
Alex_Gaynorfijal: hehe, whatever :)23:56
fijalAlex_Gaynor: anyway, I'll glue it to buildbot23:56
kenaan12mattip numppy-flatitter 11ff368a6d0ff7 15/pypy/module/micronumpy/: redo, add lots of tests, some still fail23:56
kenaan12mattip numpypy-shape-bug 115a8fc969e644 15/pypy/module/micronumpy/: add more tests23:56
Alex_Gaynorfijal: great23:56
mattipuhh. I guess. I thought I did something else.23:56
fijalmattip: and what about skip?23:56
mattipskip is gone. all clear.23:57
fijalmattip: I don't know, the diff shows three disabled asserts and two removed :)23:57
fijalso that's what made me skeptical23:57
fijalif it fixes tests broken on trunk that's awesome23:57
fijalbut it's impossible for me to tell23:57
mattipthe test didn't exist on trunk. 23:57
fijalwell no23:57
fijalso wtf it does not show up on diff? hg merge fuckup?23:57
fijalman I hate hg :/23:57
fijalmattip: you did not merge default to branch did yuo?23:58
mattipnope.23:58
fijalok23:58
fijalI'll fix the conflicts23:58
mattipbut I only branched out of it with one changeset. (now two)23:58
fijalthey seem super dumb23:59
--- Mon Jan 23 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!