#pypy IRC log for Monday, 2009-07-27

gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) left irc: Remote closed the connection00:44
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) joined #pypy.00:45
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) left irc: Read error: 110 (Connection timed out)01:03
jbaker (n=jbaker@67.176.103.236) joined #pypy.01:07
fuzzyman (n=fuzzyman@87-194-212-65.bethere.co.uk) left irc: 01:22
Nick change: magcius_ -> magcius01:37
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) joined #pypy.01:50
lsoto_ (n=lsoto@pc-148-86-104-200.cm.vtr.net) left irc: Read error: 110 (Connection timed out)01:54
Trundle (n=andy@p5B134805.dip.t-dialin.net) left irc: "..."02:02
yangkun (n=wayhome@58.249.41.164) joined #pypy.02:20
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) left irc: Remote closed the connection02:23
ousado_ (n=johnny@p5B3C32DE.dip0.t-ipconnect.de) joined #pypy.03:03
panni_ (i=hannes@ip-95-222-19-21.unitymediagroup.de) left irc: "( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )"03:05
panni_ (i=hannes@ip-95-222-19-21.unitymediagroup.de) joined #pypy.03:05
ousado (n=johnny@p5B3C1884.dip0.t-ipconnect.de) left irc: Read error: 110 (Connection timed out)03:17
panni_ (i=hannes@ip-95-222-19-21.unitymediagroup.de) left irc: "( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )"03:37
verte-work (i=9665ddc5@gateway/web/freenode/x-fa0fa023f5b949f7) joined #pypy.04:27
verte-work (i=9665ddc5@gateway/web/freenode/x-fa0fa023f5b949f7) left irc: Ping timeout: 180 seconds04:35
kursor (n=kursor@p4FC09FB8.dip0.t-ipconnect.de) joined #pypy.05:06
lsoto_ (n=lsoto@pc-148-86-104-200.cm.vtr.net) joined #pypy.05:07
verte-work (i=9665ddc5@gateway/web/freenode/x-05cf553d107341f9) joined #pypy.05:11
kursor (n=kursor@p4FC09FB8.dip0.t-ipconnect.de) left irc: 05:47
cwillu (n=cwillu@cwillu.com) left irc: Remote closed the connection06:17
stakkars (n=tismer@rrcs-64-183-124-227.west.biz.rr.com) left irc: Read error: 110 (Connection timed out)06:21
cwillu (n=cwillu@cwillu.com) joined #pypy.06:26
pypybot (n=pypybot@wyvern.cs.uni-duesseldorf.de) got netsplit.06:34
regulate (i=regulate@notchill.com) got netsplit.06:34
atob (n=atob@blackwall.inajar.co.uk) got netsplit.06:34
cwillu (n=cwillu@cwillu.com) got netsplit.06:34
cwillu (n=cwillu@cwillu.com) returned to #pypy.06:39
pypybot (n=pypybot@wyvern.cs.uni-duesseldorf.de) returned to #pypy.06:39
regulate (i=regulate@notchill.com) returned to #pypy.06:39
atob (n=atob@blackwall.inajar.co.uk) returned to #pypy.06:39
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) joined #pypy.06:49
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) left irc: Client Quit06:51
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) joined #pypy.06:51
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) left irc: Remote closed the connection06:52
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) joined #pypy.06:52
mishok13 (n=gdmfsob@dm.sonopia.com) joined #pypy.07:18
hpk (n=chatzill@HSI-KBW-078-042-106-051.hsi3.kabel-badenwuerttemberg.de) joined #pypy.07:27
jbaker (n=jbaker@67.176.103.236) left irc: 07:46
verte-work (i=9665ddc5@gateway/web/freenode/x-05cf553d107341f9) left irc: "Page closed"07:52
stakkars (n=tismer@rrcs-64-183-124-227.west.biz.rr.com) joined #pypy.08:00
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) left irc: Remote closed the connection08:03
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) joined #pypy.08:04
lsoto_ (n=lsoto@pc-148-86-104-200.cm.vtr.net) left irc: Read error: 113 (No route to host)08:04
RealNitro (n=jens@nat/sun/x-c8e91bd80fb7b61e) joined #pypy.08:18
jewel (n=jewel@dsl-242-130-65.telkomadsl.co.za) joined #pypy.08:37
mwhudson (n=mwh@canonical/launchpad/mwhudson) left irc: Read error: 104 (Connection reset by peer)09:01
RealNitro (n=jens@nat/sun/x-c8e91bd80fb7b61e) left #pypy ("Leaving").09:07
antocuni (n=antocuni@host221-124-dynamic.7-79-r.retail.telecomitalia.it) joined #pypy.09:17
mwhudson (n=mwh@canonical/launchpad/mwhudson) joined #pypy.09:19
verte (n=verte@dsl-220-235-92-135.vic.westnet.com.au) joined #pypy.09:20
mwhudson_ (n=mwh@118-92-164-253.dsl.dyn.ihug.co.nz) joined #pypy.09:49
mwhudson (n=mwh@canonical/launchpad/mwhudson) left irc: Read error: 110 (Connection timed out)09:50
cfbolz (n=cfbolz@fwstups.cs.uni-duesseldorf.de) joined #pypy.09:54
cfbolzhi09:55
antocunihi!09:55
vertehi!09:59
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) left irc: 10:09
cfbolzhi anto, hi verte10:10
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) joined #pypy.10:20
antocunihi armin10:20
antocuniarigato, cfbolz: do you know the exact sprint dates?10:21
arigato"probably 18th-25th"...10:22
Action: antocuni would like something more official to book the flight :-/10:23
arigatoI suppose that you have to ask on pypy-Z10:23
Action: antocuni writes an email10:23
verte_ (n=verte@dsl-124-150-106-11.vic.westnet.com.au) joined #pypy.10:29
verte (n=verte@dsl-220-235-92-135.vic.westnet.com.au) left irc: Nick collision from services.10:30
Nick change: verte_ -> verte10:30
kursor (n=kursor@dslb-088-072-224-184.pools.arcor-ip.net) joined #pypy.10:38
fuzzyman (n=fuzzyman@87-194-212-65.bethere.co.uk) joined #pypy.10:38
cfbolzfuzzyman: hey!10:39
fuzzymanHello Carl10:40
fuzzymanMorning10:40
arigatohi10:40
fuzzymanHello Armin10:42
fuzzymanEverything going well?10:42
arigatoroughly, yes :-)10:42
fuzzyman:-)10:43
fuzzymangood10:43
Trundle (n=andy@p5B13508E.dip.t-dialin.net) joined #pypy.10:48
kenaan03antocuni 10r6664814 pypy/branch/pyjitpl5/pypy/jit/backend/cli/test 07M(test_basic.py test_runner.py): skip tests that cannot be run on top of pythonnet ...10:48
antocuniarigato: why new_with_vtable no longer take a descr?10:51
arigatobecause it's too much work to keep two constants around all optimizations, instead of just one10:53
arigatoso I just picked the one that stores the most information (from an lltype point of view)10:53
arigatobut I realized later that it was the wrong one from an ootype point of view...10:54
antocuniso I should basically rely on self.class_sizes to fish the typedescr?10:54
cfbolz (n=cfbolz@fwstups.cs.uni-duesseldorf.de) left irc: "Leaving"10:58
arigatoyes11:00
antocuniok (it works :-))11:00
arigatook :-)11:00
cfbolz (n=cfbolz@fwstups.cs.uni-duesseldorf.de) joined #pypy.11:09
kenaan03antocuni 10r6664914 pypy/branch/pyjitpl5/pypy/jit/backend/cli 07M(method.py runner.py): fix most of the failing cli jit tests: new_with_vtable no longer takes a ...11:15
cfbolz (n=cfbolz@fwstups.cs.uni-duesseldorf.de) left irc: "Leaving"11:15
arigatofound out the reason for a crash of pypy-c-jit11:30
arigatoit's because there can be Consts in live_arg_boxes in reached_can_enter_jit()11:31
atob (n=atob@blackwall.inajar.co.uk) got netsplit.11:35
pypybot (n=pypybot@wyvern.cs.uni-duesseldorf.de) got netsplit.11:35
regulate (i=regulate@notchill.com) got netsplit.11:35
jewel (n=jewel@dsl-242-130-65.telkomadsl.co.za) got netsplit.11:35
cwillu (n=cwillu@cwillu.com) got netsplit.11:35
jewel (n=jewel@dsl-242-130-65.telkomadsl.co.za) returned to #pypy.11:36
cwillu (n=cwillu@cwillu.com) returned to #pypy.11:36
-christel (i=christel@freenode/staff/exherbo.christel) to $*- [Global Notice] Hi all, we appear to be having some routing issues causing the network which services reside on to be unreachable, we're looking into it now -- please be aware that services (chanserv, nickserv etc) are unresponsive currently. Please set yourself +w for further information as we look into this. Thank you for using freenode.11:36
pypybot (n=pypybot@wyvern.cs.uni-duesseldorf.de) returned to #pypy.11:36
regulate (i=regulate@notchill.com) returned to #pypy.11:36
atob (n=atob@blackwall.inajar.co.uk) returned to #pypy.11:36
zejn (n=hruske@89-212-104-197.dynamic.dsl.t-2.net) left irc: Read error: 110 (Connection timed out)12:03
cfbolz (n=cfbolz@fwstups.cs.uni-duesseldorf.de) joined #pypy.12:07
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) left irc: "See you"12:14
verte_ (n=verte@dsl-220-235-93-82.vic.westnet.com.au) joined #pypy.12:28
yangkun (n=wayhome@58.249.41.164) left irc: "Leaving."12:36
verte (n=verte@dsl-124-150-106-11.vic.westnet.com.au) left irc: Read error: 110 (Connection timed out)12:40
Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) left irc: 12:41
verte_ (n=verte@dsl-220-235-93-82.vic.westnet.com.au) left irc: Read error: 60 (Operation timed out)12:43
verte (n=verte@dsl-220-235-93-82.vic.westnet.com.au) joined #pypy.12:48
kursor (n=kursor@dslb-088-072-224-184.pools.arcor-ip.net) left irc: 12:53
zejn (n=hruske@217.72.91.237) joined #pypy.12:58
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) joined #pypy.12:58
lizardo (n=lizardo@189.2.128.130) joined #pypy.12:59
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) joined #pypy.13:08
kenaan03arigo 10r6665014 07M(pypy/branch/pyjitpl5/pypy/jit/metainterp/test/test_basic.py): A failing test, showing the same problem as test_zlib when run with ...13:10
jewel (n=jewel@dsl-242-130-65.telkomadsl.co.za) left irc: Read error: 113 (No route to host)13:13
[mathias] (i=mathias@client-82-27-233-90.brnt.adsl.virgin.net) joined #pypy.13:39
pedronis (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) joined #pypy.13:48
cfbolzhi samuele!13:49
pedronishi13:49
arigatohi!13:49
mwhudson_ (n=mwh@118-92-164-253.dsl.dyn.ihug.co.nz) left irc: Read error: 110 (Connection timed out)14:01
kenaan03arigo 10r6665114 pypy/branch/pyjitpl5/pypy/jit 07M(llimpl.py runner.py compile.py runner_test.py test_random.py assembler.py regalloc.py test_runner.py executor.py pyjitpl.py resoperation.py test_basic.py): Add rop.SAME_AS.  Norm ...14:05
arigatoantocuni: ^^^ sorry, that's yet another detail that I didn't fix in the cli backend...14:05
arigatoI think this fixes "pypy-c-jit test_zlib.py"14:06
panni_ (i=hannes@ip-95-222-19-21.unitymediagroup.de) joined #pypy.14:08
antocuniarigato: ok, should be easy to fix :-)14:12
antocuniarigato: I'm currently fighting with this: http://codespeak.net:8099/summary/longrepr?testname=TestVList().test_cannot_be_virtual&builder=own-linux-x86-32&build=452&mod=pypy.jit.backend.cli.test.test_zrpy_vlist14:13
cfbolzarigato: why does adding same_as fix this test?14:13
antocuniany clue what it could be?14:13
arigatoyes14:13
arigatothe helper ll_listslice_startonly is annotated as taking a non-negative argument14:14
lykki (n=mallasko@fw90.netikka.fi) joined #pypy.14:14
arigatothen as a 2nd phase, we annotate the JIT14:14
arigatoit crashes become this time it needs the same helper with a possibly-negative argument14:14
arigatos/become/because14:14
antocuniuhm14:15
antocunishould we annotate it to take a maybe-negative arg in the first place?14:15
arigatoyes, but I forgot how we solved this problem14:16
arigatoit's a known issue that was solved already for some other helpers14:16
arigatoah yes14:16
arigatosee rpython/lltypesystem/rlist.py14:17
arigato_ll_list_resize_really._annenforceargs_ = (None, int)14:17
antocuniah!14:17
arigatoI guess you need to add14:17
arigatoll_listslice_startonly._annenforceargs_ = (None, None, int)14:18
Action: antocuni tries14:18
arigatocfbolz: this test fails because the loop has unexpectedly a Const in the inputargs14:19
antocuniuhm, no14:19
arigatogrumble14:20
arigatoI guess the difference is that ll_listslice_startonly is gendirectcall()ed14:20
arigatothen the cleanest solution is to add support for _annenforceargs_ in gendirectcall() too14:21
antocuniyak shaving...14:22
arigatonot really14:22
arigatoit is really a known problem whose solution we delayed14:22
antocuniok14:22
arigatowait, I'm tracking...14:23
arigatoobscure, it should already work14:24
antocuniarigato: why? I don't see any reference to _annenforceargs_ in the call chain from gendirectcall14:25
arigato=> annotate_helper => get_call_parameters => pycall => specialize14:26
antocuniarigato: but gendirectcall calls annotate_lowlevel_helper, not annotate_helper14:28
arigatoyes, sorry, so it's14:29
arigato=> annotate_lowlevel_helper => annotate_helper => get_call_parameters => pycall => specialize14:29
[mathias] (i=mathias@client-82-27-233-90.brnt.adsl.virgin.net) left irc: Remote closed the connection14:29
antocuniah14:29
antocunisorry, I was confused by all these methods with almost similar names :-)14:29
arigato:-)14:30
cfbolzarigato: ah, and you get around that by making a same_as(const), as usual14:30
arigatoyes14:30
[mathias] (i=mathias@client-82-27-233-90.brnt.adsl.virgin.net) joined #pypy.14:30
Action: antocuni afk for some minutes14:30
kursor (n=kursor@p4FC0B619.dip0.t-ipconnect.de) joined #pypy.14:33
arigatoah14:44
arigatothe problem is actually another unrelated graph14:44
arigatoI don't know why it's printing that14:44
arigatoit's actually about ootype:690:ll_newlist14:44
kenaan03arigo 10r6665214 pypy/branch/pyjitpl5/pypy/rpython 07M(ootype.py rlist.py): Add a few _annenforceargs_, to fix jit/backend/cli/test_zrpy_vlist. ...14:48
arigatopfff, building pypy-c-jit gives a segfault14:49
arigatohttp://codespeak.net:8099/summary?category=lib-python&branch=branch/pyjitpl514:50
arigatoa segfault while building, so of CPython, of course14:50
[mathias] (i=mathias@client-82-27-233-90.brnt.adsl.virgin.net) left irc: Read error: 60 (Operation timed out)14:50
cfbolzarigato: great work14:52
kenaan03arigo 10r6665314 07M(pypy/branch/pyjitpl5/pypy/jit/metainterp/TODO): A note: we are still missing threads. ...14:53
[mathias] (n=mathias@client-82-27-233-90.brnt.adsl.virgin.net) joined #pypy.14:54
arigatocfbolz: what exactly? :-)14:55
cfbolzarigato: that you manage to find (probably completely undebugabble) segfaults in CPython :-)14:56
arigatobah14:56
cfbolzis it repeatable?14:56
arigatoI don't know, it triggered the last two runs14:56
arigatobut at different places14:56
cfbolzgrrrr14:56
arigatoit looks actually a bit more like the machine is suddently a bit unrealiable14:57
arigatounreliable14:57
cfbolzwhich one is that?14:58
arigatobigdogvm114:58
[mathias] (n=mathias@client-82-27-233-90.brnt.adsl.virgin.net) left irc: 15:01
victorpoluceno (n=victorpo@189.73.83.248) joined #pypy.15:02
Action: antocuni back15:03
antocuniarigato: thanks!15:03
arigatoseems to have helped.  I really think we should at some point do some trick like15:04
arigatoreverse-forcing the annotation of all rtyped functions15:04
cfbolzreverse-forcing?15:04
arigatoto turn SomeInteger(nonneg=True) into SomeInteger(nonneg=False)15:04
arigatoafter all, these two annotations have the same rtyping15:05
arigatohum, but it may be wrong15:05
arigatoe.g. "def f(x): return x"15:05
lsoto_ (n=lsoto@pc-148-86-104-200.cm.vtr.net) joined #pypy.15:05
arigatogeneralizing its arg annotation also means we have to generalize the result...15:06
arigatogrumble15:06
arigatook, that's all nonsense15:06
vertedo what you do with objspace method variants that can raise; copy the function, reuse the code object15:08
vertesorry. you know that, of course.15:08
victorpoluceno_ (n=victorpo@189.73.101.31) joined #pypy.15:09
antocuniarigato: I cannot find a single test that fails because the cli backend doesn't implement rop.SAME_AS15:10
arigatoyes, I guess so15:11
arigatoI just added a test to test_basic15:11
arigatoand even if you import it and run it too15:11
arigatothe truth is that optimize.py will remove SAME_AS again15:11
antocunican't we disable the optimizer for the single test?15:11
arigatoyes15:12
Action: antocuni does15:12
arigatoright, I forgot that we can do that for a single test :-)15:12
arigato(the test was originally about testing the front-end, before optimize.py)15:12
antocuniuh? now the loops contains the same_as, but it still passes15:13
antocunipypy has finally reached the AI state, where code writes itself15:14
antocuniah, I know15:14
antocunithe cli backend is smart enough to deduce the implementation of most operations from translation/cli/opcodes.py15:15
cfbolzhah15:15
antocuniso same_as is automatically added :-)15:15
arigatoah :-)15:15
arigatogreat15:15
cfbolzarigato: so you didn't even break the cli backend15:16
arigatothe AI state is indeed close15:16
kenaan03antocuni 10r6665414 07M(pypy/branch/pyjitpl5/pypy/jit/metainterp/test/test_basic.py): use simple_optimize for this test, so that the generated same_as is not removed away ...15:17
Topic changed on #pypy by verte!n=verte@dsl-220-235-93-82.vic.westnet.com.au: PyPy the flexible snake (irc-logs: http://wyvern.cs.uni-duesseldorf.de/irc-logs/pypy/ ) | <arigato> the AI state is indeed close15:18
setanta (n=setanta@200.184.118.130) joined #pypy.15:19
lmoura (n=lauromou@200.184.118.130) joined #pypy.15:20
victorpoluceno (n=victorpo@189.73.83.248) left irc: Read error: 110 (Connection timed out)15:26
lykki (n=mallasko@fw90.netikka.fi) left #pypy.15:28
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) left irc: Remote closed the connection15:30
ericvrp_ (n=ericvrp_@ericvrp.demon.nl) joined #pypy.15:33
jbaker_ (n=jbaker@67.176.103.236) joined #pypy.15:42
kursor_ (n=kursor@p4FC0B619.dip0.t-ipconnect.de) joined #pypy.15:46
pedronis (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) got netsplit.15:48
mishok13 (n=gdmfsob@dm.sonopia.com) got netsplit.15:48
ousado__ (n=johnny@p5B3C32DE.dip0.t-ipconnect.de) joined #pypy.15:48
ousado_ (n=johnny@p5B3C32DE.dip0.t-ipconnect.de) left irc: Remote closed the connection15:48
jbaker_ (n=jbaker@67.176.103.236) left irc: Read error: 60 (Operation timed out)15:50
pedronis` (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) joined #pypy.15:50
pedronis` (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) left irc: Remote closed the connection15:51
victorpoluceno_ (n=victorpo@189.73.101.31) left irc: Read error: 110 (Connection timed out)15:51
cfbolzarigato: in the C backend we don't have processor-specific assembler somewhere for the ovf versions of operations, right?15:52
pedronis (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) got netsplit.15:53
mishok13 (n=gdmfsob@dm.sonopia.com) returned to #pypy.15:53
arigatothere is, it's actually not processor-specific but gcc-specific code15:53
ericvrp_hi guys15:54
arigatouh, no15:54
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) joined #pypy.15:54
arigatoit's really x86-specific15:54
arigatofile asm_gcc_x86.h15:54
cfbolzarigato: ah, where is that?15:54
cfbolzok15:54
arigatoericvrp_: hey, hi!15:54
victorpoluceno (n=victorpo@189.73.101.31) joined #pypy.15:55
cfbolzarigato: I see. I was staring at some assembler and was already impressed by GCC figuring out that it can use the overflow bit :-)15:55
arigato:-)15:55
arigatoI wonder if there is a better way to write this, nowadays15:56
arigatolike a builtin function of GCC15:56
cfbolzno clue15:56
pedronis (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) returned to #pypy.15:56
cfbolzarigato: it's anyway far from optimal15:56
arigatoyes15:56
cfbolzarigato: as the return value is still put into the exception struct15:56
arigatobut it's not extremely important to have the most optimal code, I suppose15:56
arigatoit's more important in the context of the jit15:57
arigatobecause there is much less overhead-ish code15:57
cfbolzarigato: I am looking a bit at what happens to tagged pointers15:57
cfbolzusing microbenchmarks for a tiny object system15:58
cfbolzand one problem is the int_lshift_ovf that is needed to construct a new instance of the small int15:58
arigatoit's a problem because it is slowish?15:59
ericvrp_long time no see. I've been trying to read some pypy code again the last few days and there is a lot of terminology to catch up with.15:59
arigatoericvrp_: :-)15:59
arigatoI guess "x<<1" could be rewritten into "x+x"...16:00
cfbolzah, it's worth a try16:00
arigato(maybe automatically)16:00
arigato(at least for int_lshift_ovf)16:00
kursor (n=kursor@p4FC0B619.dip0.t-ipconnect.de) left irc: Read error: 110 (Connection timed out)16:00
cfbolzarigato: yes. it is definitely not much faster than the two compares you need for a semispace allocation :-)16:01
arigato:-)16:01
ericvrp_arigato: I couldn't get the llvm jit backend to work with LLVM svn head. Is that a known problem?16:01
arigatoericvrp_: hum.  I also have llvm 2.6svn16:02
arigator7261016:02
arigatothat's already a few months old16:03
ericvrp_My version is r7706516:03
arigatoI'm updating, I will tell you if the tests still pass for me16:03
ericvrp_But the machine I tried it on is Gentoo that was not updates for a looooonnnng time16:03
ericvrp_ok, thanks16:04
arigatoah :-)16:04
arigatowell, not "ah:-)" in the sense that this is no obvious problem16:04
dialtone_ (n=dialtone@unaffiliated/dialtone) left irc: Connection timed out16:04
ericvrp_I thought that perhaps there would have been a 'simple' jit c backend that would generate code not just in time, but only after a longer duration.16:05
ericvrp_But that is probably just one of those things that might be slightly faster but not general enough for the linking of anyone else but me. ;-)16:08
ericvrp_s/linking/liking     (picking up the lingo :-) )16:08
arigatoah, no, there is no such thing for now16:09
arigatoyou remind me of the 'ivm' backend of psyco16:09
arigatoI wonder if we could use similar tricks16:09
ericvrp_Can you say in one line what that 'ivm' psyco backend was about?16:12
ericvrp_Perhaps qwe 16:12
dialtone (n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net) joined #pypy.16:12
ericvrp_we could use llvm-gcc to generate the llvm code for us and so avoiding an llvm backend.16:12
arigatowell, the llvm backend is mostly done already :-)16:15
arigatoah, sorry, I mean the JIT llvm backend16:15
arigatobigdog1: the pseudomachine bigdog-vm1 seems to get random segfaults16:16
arigatoit crashes on the last two runs of:16:16
arigatohttp://codespeak.net:8099/summary?category=lib-python&branch=branch/pyjitpl516:16
arigatobut the build is successful on another machine (wyvern)16:17
arigatoyay, the bug I thought I fixed really went away16:18
bigdog1arigato:  The ubuntu under the  VM has not been updated in quite some time.  Would you like me to update the packages on the VM instance?16:37
arigatoI've no clue what the error is16:38
bigdog1that would also bring up the latest kernel version16:38
arigatoit's strange that a CPython segfault suddenly appeared16:38
bigdog1arigato: hard to say16:38
arigatobut I suppose that you can upgrade it -- if nothing else, it will shuffle randomness again :-/16:38
ericvrp_Someone suggests to remove pypy from Debian unstable: http://groups.google.com/group/linux.debian.bugs.dist/browse_thread/thread/efa0347b7f504075?pli=1 16:40
ericvrp_I have no clue if we care/mind/arehappy but it looks relevant to this channel.16:41
arigatothe comment is right: anyone with a non-trivial interest should use HEAD instead16:47
cfbolzarigato: of course switching to int_add_ovf(x, x) has no measurable effect :-)16:48
arigatoah16:48
bigdog1arigato: the update will take a couple of hours16:48
arigatook, thanks16:48
bigdog1by tomorrow,  I will update all VM instances16:48
Action: cfbolz confused and unhappy16:48
bigdog1I don't know if this will impact anything, but we will be running with the best available16:49
ericvrp_cfbolz: hi! and sorry for not replying to your last email. Truth be told, I forgot all about it...16:49
cfbolzericvrp_: np :-)16:49
arigatobah, the assert at optimizeopt.py:125 can actually fail16:50
arigatoafter running pypy-c-jit for some time :-(16:50
stakkars (n=tismer@rrcs-64-183-124-227.west.biz.rr.com) left irc: 16:56
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) left irc: Remote closed the connection16:56
Nick change: ousado__ -> ousado17:11
ericvrp_ (n=ericvrp_@ericvrp.demon.nl) left irc: 17:17
stakkars (n=tismer@72.37.205.4) joined #pypy.17:18
stakkars (n=tismer@72.37.205.4) left irc: Connection reset by peer17:18
stakkars (n=tismer@72.37.205.4) joined #pypy.17:19
amaury_ (i=5bc34844@gateway/web/freenode/x-c6ec8e403f885417) joined #pypy.17:21
victorpoluceno (n=victorpo@189.73.101.31) left irc: Read error: 110 (Connection timed out)17:40
kursor_ (n=kursor@p4FC0B619.dip0.t-ipconnect.de) left irc: 17:49
verte (n=verte@dsl-220-235-93-82.vic.westnet.com.au) left irc: "~~~ Crash in JIT!"18:03
zejn (n=hruske@217.72.91.237) left irc: Read error: 113 (No route to host)18:08
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) joined #pypy.18:17
stakkarsarigato: Hi Armin18:19
arigatohi Christian18:19
stakkarssorry, wrong channel, I meant psyco18:19
gonsor (n=gon@kel30.kel.stud.uni-goettingen.de) joined #pypy.18:20
zejn (n=hruske@89-212-104-197.dynamic.dsl.t-2.net) joined #pypy.18:32
cfbolz (n=cfbolz@fwstups.cs.uni-duesseldorf.de) left irc: Read error: 113 (No route to host)18:33
amaury_ (i=5bc34844@gateway/web/freenode/x-c6ec8e403f885417) left irc: "Page closed"18:36
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) left irc: "bye"18:38
jewel (n=jewel@dsl-242-130-65.telkomadsl.co.za) joined #pypy.19:06
hpk (n=chatzill@HSI-KBW-078-042-106-051.hsi3.kabel-badenwuerttemberg.de) left irc: Read error: 110 (Connection timed out)19:12
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) joined #pypy.19:24
kenaan03arigo 10r6665514 07M(pypy/branch/pyjitpl5/pypy/jit/metainterp/test/test_optimizefindnode.py): A skipped test corresponding to a bug seen on pypy-c-jit. ...19:41
ericvrp_ (n=ericvrp_@ericvrp.demon.nl) joined #pypy.19:55
victorpoluceno (n=victorpo@189-72-71-47.bnut3700.dsl.brasiltelecom.net.br) joined #pypy.19:56
panni_ (i=hannes@ip-95-222-19-21.unitymediagroup.de) left irc: Read error: 104 (Connection reset by peer)20:09
kenaan03arigo 10r6665614 pypy/branch/pyjitpl5/pypy/jit/metainterp 07M(history.py optimize.py): An RPython-friendly way to dump a loop. ...20:14
victorpoluceno_ (n=victorpo@189.72.83.208) joined #pypy.20:19
panni_ (i=hannes@ip-95-222-19-21.unitymediagroup.de) joined #pypy.20:19
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) left irc: 20:28
pjenvey_ (n=pjenvey@astound-69-42-4-166.ca.astound.net) joined #pypy.20:30
victorpoluceno (n=victorpo@189-72-71-47.bnut3700.dsl.brasiltelecom.net.br) left irc: Read error: 110 (Connection timed out)20:38
magcius (n=jstpierr@c-24-61-240-155.hsd1.ma.comcast.net) left irc: Read error: 104 (Connection reset by peer)20:40
magcius_ (n=jstpierr@c-24-61-240-155.hsd1.ma.comcast.net) joined #pypy.20:40
ykk (n=arigo@adsl-188-155-110-188.adslplus.ch) joined #pypy.20:42
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) left irc: Read error: 104 (Connection reset by peer)20:42
ericvrp_ (n=ericvrp_@ericvrp.demon.nl) left irc: 20:43
jewel (n=jewel@dsl-242-130-65.telkomadsl.co.za) left irc: Read error: 60 (Operation timed out)20:50
Nick change: magcius_ -> magcius21:01
antocuni (n=antocuni@host221-124-dynamic.7-79-r.retail.telecomitalia.it) left irc: "Leaving"21:09
-tomaw (i=tom@freenode/staff/tomaw) to $*- [Global Notice] Hi all, one of our sponsors is having connectivity problems resulting in a loss of services (ChanServ, NickServ etc). Please keep with us while we investigate. Thanks.21:09
victorpoluceno_ (n=victorpo@189.72.83.208) left irc: "Leaving"21:15
victorpoluceno (n=victorpo@189.72.83.208) joined #pypy.21:15
victorpoluceno (n=victorpo@189.72.83.208) left irc: "Leaving"21:24
kenaan03arigo 10r6665714 pypy/branch/pyjitpl5/pypy/jit/metainterp 07M(optimizefindnode.py optimizeopt.py test_optimizefindnode.py test_optimizeopt.py): Bug fix in optimize*.py, originally shown by 'pypy-c-jit test_decimal.py'. ...21:29
Nick change: ykk -> arigato21:30
mwhudson (n=mwh@canonical/launchpad/mwhudson) joined #pypy.21:32
jbaker (n=jbaker@c-75-71-73-18.hsd1.co.comcast.net) joined #pypy.21:32
arigato (n=arigo@adsl-188-155-110-188.adslplus.ch) left irc: "See you"21:35
pedronis (n=user@c-c348e353.022-54-67626719.cust.bredbandsbolaget.se) left irc: Remote closed the connection21:36
fuzzyman (n=fuzzyman@87-194-212-65.bethere.co.uk) left irc: 21:50
dialtone (n=dialtone@unaffiliated/dialtone) left irc: Read error: 104 (Connection reset by peer)22:22
jbaker (n=jbaker@c-75-71-73-18.hsd1.co.comcast.net) left irc: 22:22
lizardo (n=lizardo@189.2.128.130) left irc: "Leaving"22:23
dialtone (n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net) joined #pypy.22:27
lesshaste (n=lesshast@87-194-206-189.bethere.co.uk) joined #pypy.22:29
lesshastehi all22:29
lesshasteor.. anyone :)22:32
Rhamphoryncus (n=rhamph@unaffiliated/rhamphoryncus) joined #pypy.22:39
lesshastehi Rhamphoryncus 22:40
Rhamphoryncushiya lesshaste22:40
lesshasteare you a pypy dev by any chance?22:41
setanta (n=setanta@200.184.118.130) left irc: "Leaving."23:05
Nick change: foottone -> armtone23:37
--- Tue Jul 28 200900:00

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