2014-12-31T16:08:17Z ccl-logbot joined #sbcl
2014-12-31T16:08:17Z 
2014-12-31T16:08:17Z names: ccl-logbot drmeister_ nikki93 daviid rpg pchrist attila_lendvai Bicyclidine nyef oleo__ karswell milosn angavrilov hzp stassats krzysz00 @Krystof scymtym drichard` specbot |3b| PuercoPop jdz edgar-rft redline6561 Posterdati DeadTrickster ams undata akkad antoszka flip214_ jsnell_ loke luis irsol Intensity jackdaniel brucem sobel pkhuong kanru snafuchs nicdev White_Flame fitzsim njmurphy 1JTABHVVP ferada ivan4th weissschloss tmh_ ivan`` fikusz Hydan joshe
2014-12-31T16:08:17Z names: les fe[nl]ix minion jasom alchemis7 qbit phf` Blkt carvite mood p_l Perlboy yauz
2014-12-31T16:11:53Z psy joined #sbcl
2014-12-31T16:12:17Z nikki93 quit (Read error: Connection reset by peer)
2014-12-31T16:12:31Z nikki93 joined #sbcl
2014-12-31T16:13:51Z oleo__ quit (Quit: Verlassend)
2014-12-31T16:16:58Z nikki93 quit (Ping timeout: 244 seconds)
2014-12-31T16:22:33Z oleo joined #sbcl
2014-12-31T16:23:18Z nikki93 joined #sbcl
2014-12-31T16:30:36Z stassats quit (Ping timeout: 244 seconds)
2014-12-31T16:35:10Z stassats joined #sbcl
2014-12-31T16:39:45Z stassats quit (Ping timeout: 265 seconds)
2014-12-31T16:43:35Z nikki93 quit (Remote host closed the connection)
2014-12-31T16:44:09Z nikki93 joined #sbcl
2014-12-31T16:45:52Z nikki93_ joined #sbcl
2014-12-31T16:45:55Z nikki93 quit (Read error: Connection reset by peer)
2014-12-31T16:52:53Z LiamH joined #sbcl
2014-12-31T16:56:49Z nikki93_ quit (Remote host closed the connection)
2014-12-31T17:01:37Z Krystof: well total suck
2014-12-31T17:01:43Z Krystof: arm build fails
2014-12-31T17:02:59Z Krystof: similar failure: failing to get 512MB at 0x4f00000000
2014-12-31T17:03:36Z Krystof: though oddly make-target-2 succeeds on my pi, and it's just that loading that dumped core while building contribs fails
2014-12-31T17:04:36Z nyef: Something different between the cold-load environment and the contrib-build environment?
2014-12-31T17:04:50Z nyef: (And by "environment", I mean the set of environment variables.)
2014-12-31T17:06:08Z Krystof: I suppose there's an outside chance that stasssats' reordering has fixed this
2014-12-31T17:06:21Z Krystof tries on the "fast" arm board
2014-12-31T17:59:52Z stassats joined #sbcl
2014-12-31T18:01:10Z |3b|: 1.2.6.61 built on whatever sort of arm i have
2014-12-31T18:13:40Z drichard`: :)
2014-12-31T18:19:45Z Adlai joined #sbcl
2014-12-31T18:20:56Z Adlai: I have reproduced (by accident) this SBCL weirdness where suddenly parts of the internet doesn't exist
2014-12-31T18:21:11Z Adlai: I'm leaving the lisp image open, whenever an SBCL hacker wants to investigate vicariously, please speak up!
2014-12-31T18:28:28Z pkhuong: problems with name resolution?
2014-12-31T18:28:33Z attila_lendvai quit (Quit: Leaving.)
2014-12-31T18:28:42Z pkhuong: I think that's a known bug wrt threads and some libc cache
2014-12-31T18:28:55Z pkhuong: Adlai: what's the OS?
2014-12-31T18:30:18Z eudoxia joined #sbcl
2014-12-31T18:30:36Z rpg: I just build sbcl-1.2.6 on linux, and find that it cannot compile ITERATE. Is this a known issue?
2014-12-31T18:31:06Z pkhuong: rpg: what's the error?
2014-12-31T18:31:15Z Adlai: 3.16.1-1-ARCH #1 SMP PREEMPT Thu Aug 14 07:48:39 CEST 2014 i686 GNU/Linux
2014-12-31T18:32:06Z rpg: pkhuong: will pastebin.. Hang on...
2014-12-31T18:32:38Z pkhuong: Adlai: and what's the problem? I'm pretty sure we switched to re-entrant resolution on glibc.
2014-12-31T18:32:42Z Adlai: (lisp-implementation-version) → 1.2.2
2014-12-31T18:33:06Z Adlai: some calls to http-request hang indefinitely. it seems to be for certain domains only.
2014-12-31T18:33:40Z pkhuong: Adlai: http-request isn't part of SBCL. a backtrace might be useful.
2014-12-31T18:33:40Z Adlai: specifically, the domain I was hammering for the past two months is dead, but ones that I have never called before - work OK.
2014-12-31T18:33:48Z rpg: pastebin.com/evR4BTHF
2014-12-31T18:34:37Z |3b|: compare-and-swap.impure.lisp still gets unhandled error on arm due to a missing symbol i think (it wouldn't try to run the test in question even if it did READ, but other tests in that file list success)
2014-12-31T18:34:41Z rpg: Had this problem with 1.2.4, too, if that helps. Trying to use from quicklisp gave same error. Surprised that Xach didn't stub his toe on it.
2014-12-31T18:35:28Z Adlai: odd, the backtrace is very short, and ends before the call to http-request: http://paste.lisp.org/+33WH
2014-12-31T18:35:46Z pkhuong: rpg: break-on-signals might help you figure out where the error happens in iterate.
2014-12-31T18:35:55Z Adlai: this was from calling http-request on the problematic domain, waiting a minute, then C-cC-c
2014-12-31T18:36:14Z rpg: pkhuong: isn't that going to trip on every single STYLE-WARNING?
2014-12-31T18:36:27Z pkhuong: rpg: with an appropriate type ;)
2014-12-31T18:36:44Z pkhuong: Adlai: I think you'll have to compile C libraries with frame pointers or cl+ssl at a high debug level.
2014-12-31T18:37:14Z Adlai: can I dispose of this image, or is there any other useful info you could glean from it?
2014-12-31T18:38:18Z Adlai: this is only the second time I've ever encountered this bug; both after roughly a month of uptime, during which three threads make requests to various URLs in the one domain
2014-12-31T18:40:22Z pkhuong: I don't have the relevant bits paged in
2014-12-31T18:40:27Z rpg: pkhuong: I can get it to error out in SLIME by trying to expand that WITH-TEMPORARY form; trying to winkle out of SLDB some information about how things are getting called.
2014-12-31T18:40:29Z pkhuong: someone else might have ideas..
2014-12-31T18:41:12Z Adlai: the previous time this happened, fe[nl]ix suggested patching drakma to use iolib as a workaround, based on what he thought was the problem
2014-12-31T18:41:28Z Adlai: but I don't remember what his hunch was
2014-12-31T18:43:17Z rpg: pkhuong: Looks like it might be a problem in the #L reader macro.
2014-12-31T18:49:11Z rpg: pkhuong: Yes, that's it. When I replace the #L reader macro with something equivalent, it compiles successfully.
2014-12-31T18:50:32Z rpg: pkhuong: That's *definitely* it.  After that fix, we move on to a place where we get a different error that's much more clearly from #L.
2014-12-31T18:51:50Z rpg: I'll rebuild from HEAD and make sure that I get the same error.
2014-12-31T18:54:14Z stassats: Adlai: when name getaddrinfo is interrupted it may break its state
2014-12-31T18:54:24Z stassats: Adlai: usocket uses timeouts around, and timeouts use interrupts
2014-12-31T18:54:33Z Adlai: and I sometimes kill all my threads in anger
2014-12-31T18:55:13Z rpg: pkhuong: btw, I also got one test failure: Failure:            interface.impure.lisp / DISASSEMBLE-CLOSURE -- could that be related?
2014-12-31T18:55:34Z stassats: rpg: is that current sbcl? i believe that was fixed
2014-12-31T18:55:47Z rpg: stassats: No it's the released one (1.2.6)
2014-12-31T18:56:00Z Adlai: iirc last time this happened was just randomly, but this time it happened after I was mucking about and killed the threads several times while trying to restart a certain part
2014-12-31T18:56:18Z nyef: We're at less than five hours to go in code-freeze, aren't we?
2014-12-31T18:56:29Z stassats: rpg: yes, it was fixed this cycle
2014-12-31T18:56:45Z rpg is having one of those days when every piece of software he touches, breaks.
2014-12-31T18:57:58Z stassats: rpg: this tells you that should be drinking and celebrating instead
2014-12-31T18:58:09Z rpg: stassats: amen!
2014-12-31T18:59:26Z rpg: I pretty religiously use  ITERATE instead of LOOP, so am a bit taken aback by this. My first guess, though, is that somehow I have done something wrong....
2014-12-31T19:01:46Z rpg: Yes, with HEAD it blew up on me, too.  Seems like something in SBCL for a while (since 1.2.4 is breaking the build of ITERATE).
2014-12-31T19:02:19Z rpg: aha. builds from quicklisp.
2014-12-31T19:02:30Z rpg: So probably I have an ITERATE that needs updating....
2014-12-31T19:03:11Z nyef: ... which would probably make it "backquotegate" again.
2014-12-31T19:03:53Z Krystof: //build started:  Wed Dec 31 17:05:48 UTC 2014
2014-12-31T19:03:53Z Krystof: //build finished: Wed Dec 31 17:51:13 UTC 2014
2014-12-31T19:03:58Z Krystof: what do you know?  Thank you, stassats
2014-12-31T19:04:02Z nyef: Krystof: Congratulations.
2014-12-31T19:04:16Z Krystof: it appears that the reordering of validate/init fixes everything
2014-12-31T19:04:48Z stassats: Krystof: oh, it did break things for you too?
2014-12-31T19:05:25Z rpg: nyef: yup.
2014-12-31T19:05:32Z Krystof: stassats: yeah, it put the locales mmap in the middle of our dynamic space
2014-12-31T19:05:40Z rpg: please tell me that I don't need to get ITERATE from darcs....
2014-12-31T19:05:42Z Krystof: (which I think we could afford to move down a bit, but it's fine for now)
2014-12-31T19:05:59Z Krystof: ok, setting off the build process if I remember how to chroot after enough bubbly
2014-12-31T19:05:59Z nyef: rpg: Of course you don't! That's what quicklisp is for. d-:
2014-12-31T19:06:20Z stassats: i think validate should be done first thing, actually
2014-12-31T19:06:21Z rpg: nyef:  I know, but I need to get a new copy and push it into my company's SVN repo....
2014-12-31T19:06:28Z stassats: before any malloc breaks anything
2014-12-31T19:06:34Z rpg: nyef: I was hoping someone had moved this to git.
2014-12-31T19:10:10Z pkhuong: stassats: I think foom had linker magic at some point.
2014-12-31T19:10:35Z nyef: ... linker magic? Like the Alpha backend?
2014-12-31T19:11:01Z pkhuong: some link script to reserve address space.
2014-12-31T19:11:38Z nyef: Sounds about right, yes.
2014-12-31T19:12:03Z rpg: nyef: No, I was wrong: it is busted. :-(
2014-12-31T19:13:02Z rpg: I think I need to take stassats' suggestion: time to go and have a drink.
2014-12-31T19:15:19Z rpg: if anyone has a chance to try building iterate (probably with :force t), please drop me a line w/ the results.  I'll follow up and file a bug report.
2014-12-31T19:15:43Z stassats: where did you get iterate from?
2014-12-31T19:16:57Z stassats: is this Wed Apr 23 08:47:59 MSK 2014  Douglas Katzman <dougk@google.com>
2014-12-31T19:16:57Z stassats:   * make the #L reader macro standard conformant by not assuming anything about the representation of quasiquoted forms.
2014-12-31T19:16:57Z stassats:  
2014-12-31T19:16:57Z stassats:  
2014-12-31T19:18:37Z nyef: Hence, backquotegate.
2014-12-31T19:26:12Z rpg: stassats: thanks!
2014-12-31T19:26:20Z rpg:  I have an out-of-date copy, obviously.
2014-12-31T19:26:40Z rpg: and when I loaded from quicklisp, the local copy overrode.... :-/
2014-12-31T19:27:34Z rpg_ joined #sbcl
2014-12-31T19:31:28Z rpg quit (Ping timeout: 255 seconds)
2014-12-31T19:32:16Z rpg_ quit (Ping timeout: 244 seconds)
2014-12-31T19:59:19Z rpg joined #sbcl
2014-12-31T20:36:40Z Krystof: gah, totally incapable of driving the release script
2014-12-31T20:36:46Z Krystof: last time I try to release on New Year's Eve
2014-12-31T20:49:09Z rpg quit (Quit: rpg)
2014-12-31T20:55:51Z nyef: stassats: Did you end up figuring out that logcount bounds thing?
2014-12-31T20:56:23Z stassats: nope
2014-12-31T21:03:47Z psy quit (Remote host closed the connection)
2014-12-31T21:21:34Z krzysz00 quit (Ping timeout: 250 seconds)
2014-12-31T21:34:54Z stassats quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2014-12-31T21:44:41Z psilord joined #sbcl
2014-12-31T22:06:14Z pacon joined #sbcl
2014-12-31T22:08:51Z eudoxia quit (Quit: Leaving)
2014-12-31T22:23:25Z Adlai quit (Quit: WeeChat 1.0.1)
2014-12-31T22:28:54Z angavrilov quit (Remote host closed the connection)
2014-12-31T22:29:09Z White__Flame joined #sbcl
2014-12-31T22:34:20Z nyef quit (*.net *.split)
2014-12-31T22:34:21Z |3b| quit (*.net *.split)
2014-12-31T22:34:21Z PuercoPop quit (*.net *.split)
2014-12-31T22:34:22Z 1JTABHVVP quit (*.net *.split)
2014-12-31T22:34:23Z njmurphy quit (*.net *.split)
2014-12-31T22:34:23Z fitzsim quit (*.net *.split)
2014-12-31T22:34:23Z White_Flame quit (*.net *.split)
2014-12-31T22:34:24Z nicdev quit (*.net *.split)
2014-12-31T22:40:50Z nyef joined #sbcl
2014-12-31T22:44:22Z rpg joined #sbcl
2014-12-31T22:44:22Z rpg quit (Remote host closed the connection)
2014-12-31T22:44:32Z rpg joined #sbcl
2014-12-31T22:48:50Z fitzsim joined #sbcl
2014-12-31T22:48:57Z PuercoPop joined #sbcl
2014-12-31T22:49:40Z |3b| joined #sbcl
2014-12-31T22:49:41Z njmurphy joined #sbcl
2014-12-31T22:50:18Z nicdev joined #sbcl
2014-12-31T23:09:54Z krzysz00 joined #sbcl
2014-12-31T23:22:28Z nikki93_ joined #sbcl
2014-12-31T23:26:31Z krzysz00 quit (Ping timeout: 244 seconds)
2014-12-31T23:41:32Z Bicyclidine quit (Ping timeout: 250 seconds)