2014-09-01T00:19:30Z ccl-logbot joined #sbcl
2014-09-01T00:19:30Z 
2014-09-01T00:19:30Z names: ccl-logbot ^Posterdati^ sav redline6561_ luis` jackdani1l jsnell_ krzysz00_ angavrilov scymtym_ oleo attila_lendvai segv- irsol drmeister karswell` pkhuong nyef LiamH yacks ivan`` carvite weissschloss christoph_debian les foom2 heddwch otwieracz |3b| reb` milosn akkad slyrus psilord p_l antoszka wmarvel Perlboy cracauer_ specbot leo2007 nicdev qbit yauz_2 ferada alchemis7 faheem_ jdz phf thoto_ lacedaemon White__Flame Blkt minion arrdem joshe @Krystof
2014-09-01T00:19:30Z names: p_l|backup hzp pchrist Bike leoc flip214 brucem mood
2014-09-01T00:57:04Z ivan`` quit (Remote host closed the connection)
2014-09-01T01:08:59Z ivan`` joined #sbcl
2014-09-01T01:17:04Z ivan`` quit (Remote host closed the connection)
2014-09-01T01:23:08Z ivan`` joined #sbcl
2014-09-01T01:23:43Z echo-area joined #sbcl
2014-09-01T01:31:49Z kanru joined #sbcl
2014-09-01T01:56:37Z wmarvel: Is there something I have to do to get a defknown that's in vm-fndb.lisp (and showing it compiling) to be around later on?
2014-09-01T01:57:18Z nyef: Not really, you probably have package issues more than anything else.
2014-09-01T01:59:08Z wmarvel: well, the defknown was there, I just added 'armv6' to the read-time conditional, it compiles the vop
2014-09-01T01:59:52Z wmarvel: (which I copied and modified from ppc)
2014-09-01T02:00:18Z nyef: Mmm. Both load-linked/store-conditional architecture, aren't they?
2014-09-01T02:00:35Z nyef: FWIW, I did the original PPC threading work.
2014-09-01T02:00:41Z wmarvel: :)
2014-09-01T02:00:46Z wmarvel: I'm trying to get there :)
2014-09-01T02:00:55Z nyef: ... And the original ARM work, for that matter.
2014-09-01T02:01:05Z wmarvel: yeah, I saw you had a branch on github
2014-09-01T02:01:12Z nyef: ... I did?
2014-09-01T02:01:19Z wmarvel: well
2014-09-01T02:01:24Z nyef: Can't have, my  branches were on repo.or.cz.
2014-09-01T02:02:08Z wmarvel: maybe I'm just assuming it was github
2014-09-01T02:02:52Z wmarvel: but I know I saw an arm sbcl repository somewhere with 'nyef' in the url
2014-09-01T02:03:13Z nyef: It'll be good to have someone interested in pressing forward on the ARM stuff, especially since my interest is largely finished for the time being.
2014-09-01T02:03:39Z wmarvel: I'm kind of pondering how to deal w/ v5/v6/v7
2014-09-01T02:03:49Z wmarvel: from a make-config and having stuff available standpoint
2014-09-01T02:04:00Z wmarvel: I've got a pi, v6
2014-09-01T02:04:16Z wmarvel: I've got a v7 quad-core coming because the pi is painfully slow compiling
2014-09-01T02:04:35Z wmarvel: so I should be able to push for as long as I stay interested :)
2014-09-01T02:04:39Z nyef: I have an RPi, a BBB, a GN10.1, a Galaxy Nexus, and a (bricked) NSLU2.
2014-09-01T02:05:01Z nyef: Yeah, the Pi is slow. You might be better off with a cross-compiler for much of your work.
2014-09-01T02:05:13Z wmarvel: I do the host compiling on my mac
2014-09-01T02:05:24Z wmarvel: (which is where this is currently blowing up, actually)
2014-09-01T02:05:33Z nyef: I was figuring that I'd ignore the ARM until at least the point where I could get a good multi-core 64-bit 1U rackmount.
2014-09-01T02:05:37Z wmarvel: but I tried it without crossing and it bombs that way too
2014-09-01T02:07:13Z nyef: What's the actual error that you're getting?
2014-09-01T02:07:55Z nyef: (Lisppaste or similar preferred if you can't describe it concisely in-channel, of course.)
2014-09-01T02:09:18Z wmarvel: http://pastebin.com/S8kZmpLf
2014-09-01T02:09:31Z wmarvel: It looks like it's blowing up trying to load the fasl it just compiled
2014-09-01T02:11:06Z wmarvel: the file the defknown is in has sb!c as the package, this file sb!vm, but I don't understand why all the other vops worked
2014-09-01T02:11:32Z wmarvel: or rather, the package difference doesn't all by itself
2014-09-01T02:11:36Z pkhuong: wmarvel: you can define new defknowns and vops in a running image.
2014-09-01T02:11:59Z nyef: have a look for "%RAW-INSTANCE-ATOMIC-INCF/WORD" in package-data-list.lisp-expr.
2014-09-01T02:12:08Z nyef: There's a similar conditional there.
2014-09-01T02:12:15Z nyef: (as the one in vm-fndb.)
2014-09-01T02:12:57Z nyef: (See? Package issues.)
2014-09-01T02:12:57Z wmarvel: pk: right - I did that, no errors.
2014-09-01T02:12:57Z pkhuong: interactive development will cut the edit/compile/test cycle down to a couple seconds ;)
2014-09-01T02:13:12Z pkhuong: wmarvel: in the same file?
2014-09-01T02:13:48Z wmarvel: in the slime I have running in this emacs
2014-09-01T02:14:39Z pkhuong: wmarvel: ... while editing the same in-tree file?
2014-09-01T02:14:40Z wmarvel: so, my problem seems to be I did all my finds for the symbol in src - thanks :)
2014-09-01T02:15:31Z nyef: Glad I could help. Hope to see another backend support threads soon. (-:
2014-09-01T02:16:13Z wmarvel: pk: yes, unless I've somehow got the wrong one in this buffer ;)
2014-09-01T02:17:14Z nyef: (With a bit of luck, I should have working PPC, SPARC, HPPA and possibly Alpha systems to try to build on this week. Not that I expect the SPARC to support threading or the HPPA or Alpha to build at all.)
2014-09-01T02:17:25Z pkhuong: wmarvel: I don't see how that define-vop could have succeeded interactively but fail that way on a full build.
2014-09-01T02:19:42Z wmarvel: I can try it again, but I didn't go for a full build until I'd built it interactively ..
2014-09-01T02:20:04Z wmarvel: I mean, slime just figures the package out and dumps the form at the repl, essentially, right?
2014-09-01T02:21:25Z pkhuong: yes.
2014-09-01T02:31:14Z wmarvel: *shrug* I dunno, probably my fault. :D
2014-09-01T02:31:18Z wmarvel: (somehow)
2014-09-01T02:32:38Z pkhuong quit (Remote host closed the connection)
2014-09-01T02:38:37Z christoph_debian quit (Ping timeout: 245 seconds)
2014-09-01T02:44:41Z wmarvel: nyef: so I have compile-and-swap-vops, memory-barrier-vops, will be running tests on this atomic-incf - off the top of your head, anything else I should go poke at before diving at threads?
2014-09-01T02:46:35Z LiamH quit (Quit: Leaving.)
2014-09-01T02:47:38Z nyef: Those are the big things, IIRC.
2014-09-01T02:47:50Z heddwch quit (Quit: ZNC - http://znc.in)
2014-09-01T02:48:10Z wmarvel: Apparenly I have 'compile' on the brain :)
2014-09-01T02:49:01Z nyef: And you're likely to run into far fewer hassles than I did, simply because I ran into a great number of places where the runtime wasn't actually thread-safe.
2014-09-01T02:49:41Z nyef: Umm... You may have trouble scaring up somewhere to hold a pointer to the TLS block (thread structure). ARM is dreadfully tight on registers.
2014-09-01T02:50:47Z heddwch joined #sbcl
2014-09-01T02:50:48Z nyef: The number of times that I had to alter the register partitioning in order to scare up just one more register for one purpose or another...
2014-09-01T02:51:31Z wmarvel: I read something-or-other somewhere about there being a cp15 register for tls but I haven't looked into whether or not the pi supports it
2014-09-01T02:51:33Z nyef: Storing the control stack pointer in a register was one of the more memorable WTF moments there.
2014-09-01T02:51:43Z wmarvel: (or if that's even how I should do it if it does)
2014-09-01T02:51:44Z christoph_debian joined #sbcl
2014-09-01T02:52:19Z wmarvel: basically everything I know about arm and sbcl internals I learned in the last week and half to two weeks-ish
2014-09-01T02:52:30Z nyef: You're doing well, then.
2014-09-01T02:53:57Z brucem: wmarvel: It is much easier to be effective when you know less.
2014-09-01T02:54:34Z wmarvel: It's good that I forget stuff, then
2014-09-01T02:55:33Z pkhuong joined #sbcl
2014-09-01T02:56:26Z brucem: wmarvel: knowing less, you're less likely to get paralyzed by "But that might break this or that!" ... you don't know how hard something might be ... etc.
2014-09-01T02:57:44Z nyef: On the upside, knowing more means you're more likely to look at an error and say "oh, I know what's going on here" or to look at a feature that you want to implement and know what steps to take to get it done.
2014-09-01T02:58:11Z nyef: Also means making mistakes with great confidence, so it's not all upside. (-:
2014-09-01T02:59:59Z wmarvel: :)
2014-09-01T03:01:52Z nyef: Maybe you'll be the one to tackle some of our outstanding issues and opportunities for excellence simply because you either don't have an idea of how hard they are or you realize that the issues aren't actually where everyone keeps saying they are.
2014-09-01T03:02:39Z nyef: (Or because you realize that everybody is right, and that the ARE hard, but that you can afford to spend the time and effort on them anyway.)
2014-09-01T03:06:48Z wmarvel: that second one is probably not it (job, kids, etc ...)
2014-09-01T03:06:55Z wmarvel: but I can try ;)
2014-09-01T03:17:27Z attila_lendvai quit (Quit: Leaving.)
2014-09-01T03:20:42Z nyef quit (Quit: Sleep beckons)
2014-09-01T04:03:06Z op12345 joined #sbcl
2014-09-01T04:07:57Z op12345: Hello all, I'm new here and to LISP in general and have recently fallen in love with SBCL! Thanks for the awesome job!. And now a quick question -- I'd like to build a web app and was wondering if someone could recommend a list of performant packages for the following -- postgresql, json, image manipulation, html generation and serving capabilities.
2014-09-01T04:09:03Z Bike: you probably want #lisp
2014-09-01T04:09:40Z op12345: Oh I see, thank you. I do have a dev question however...
2014-09-01T04:10:19Z Bike: just cos most libraries for such things aren't sbcl-specific
2014-09-01T04:13:53Z op12345: Ah got it. I don't know if my next question makes sense in that case but I'll try anyways -- I see most Schemes come with a neat set of packages, how come Common LISP implementations don't? Most packages on Quicklisp are poorly documented...
2014-09-01T04:14:20Z op12345: and doesn't give me all that confidence. That said does SBCL have any plans for a standard package?
2014-09-01T04:14:28Z Bike: http://quickdocs.org/
2014-09-01T04:14:36Z op12345: If not, are there any pointers for someone willing to take on this job?
2014-09-01T04:15:49Z op12345: Thanks Bike, I shall take a look at it
2014-09-01T04:16:47Z Bike: i don't think sbcl's working on anything like that (though i'm not a dev). i've never heard of language implementations doing that, anyway. python and gcc don't come with image processing libraries or anything
2014-09-01T04:21:42Z op12345: That's true and I understand. But what got me thinking was this -- I was browsing the SBCL source and saw a contrib directory there; was curious as to what kind of code gets promoted to here?
2014-09-01T04:23:02Z Bike: stuff that's pretty closely wed to the particular implementation, like gmp support and concurrency
2014-09-01T04:23:05Z Bike: plus asdf.
2014-09-01T04:25:18Z op12345: I see
2014-09-01T04:26:41Z op12345: Btw is there an automated way to generate FFI wrappers? I saw the example of sbcl.org but it doesn't mention automation of such.
2014-09-01T04:27:22Z Bike: the cffi groveller, i think. again, not really an sbcl-specific question
2014-09-01T04:27:38Z Bike: http://common-lisp.net/project/cffi/manual/html_node/The-Groveller.html
2014-09-01T04:28:08Z op12345: my apologies, I don't really know the whole ecosystem at the moment. Thanks for your patience.
2014-09-01T04:28:56Z Bike: "more dev than help", as the topic says, is all
2014-09-01T04:29:17Z drmeister quit (Remote host closed the connection)
2014-09-01T04:30:52Z op12345: Thanks for your help.
2014-09-01T04:30:59Z op12345 quit (Quit: Leaving)
2014-09-01T04:41:22Z ASau joined #sbcl
2014-09-01T05:11:29Z sdemarre joined #sbcl
2014-09-01T05:38:14Z sdemarre quit (Quit: Leaving.)
2014-09-01T06:03:16Z oleo quit (Quit: Verlassend)
2014-09-01T06:18:08Z drmeister joined #sbcl
2014-09-01T06:19:21Z pranavrc joined #sbcl
2014-09-01T06:22:21Z drmeister quit (Ping timeout: 240 seconds)
2014-09-01T06:57:23Z gingerale joined #sbcl
2014-09-01T06:57:43Z karswell` quit (Remote host closed the connection)
2014-09-01T06:57:57Z karswell` joined #sbcl
2014-09-01T07:18:22Z ASau quit (Ping timeout: 240 seconds)
2014-09-01T08:06:51Z drmeister joined #sbcl
2014-09-01T08:11:36Z drmeister quit (Ping timeout: 260 seconds)
2014-09-01T08:13:35Z edgar-rft joined #sbcl
2014-09-01T08:58:59Z erikvarga joined #sbcl
2014-09-01T08:59:34Z hlavaty joined #sbcl
2014-09-01T09:05:51Z DGASAU joined #sbcl
2014-09-01T09:40:40Z karswell` is now known as karswell
2014-09-01T09:44:35Z attila_lendvai joined #sbcl
2014-09-01T09:44:35Z attila_lendvai quit (Changing host)
2014-09-01T09:44:35Z attila_lendvai joined #sbcl
2014-09-01T09:55:36Z drmeister joined #sbcl
2014-09-01T09:59:52Z drmeister quit (Ping timeout: 240 seconds)
2014-09-01T10:51:15Z otwieracz quit (Quit: Reconnecting)
2014-09-01T10:51:30Z otwieracz joined #sbcl
2014-09-01T10:52:53Z echo-area quit (Remote host closed the connection)
2014-09-01T10:54:56Z otwieracz quit (Client Quit)
2014-09-01T11:34:45Z yacks quit (Quit: Leaving)
2014-09-01T11:44:25Z drmeister joined #sbcl
2014-09-01T11:49:04Z drmeister quit (Ping timeout: 260 seconds)
2014-09-01T12:02:09Z drmeister joined #sbcl
2014-09-01T12:22:28Z yacks joined #sbcl
2014-09-01T12:23:19Z LiamH joined #sbcl
2014-09-01T12:31:18Z flip214 quit (Remote host closed the connection)
2014-09-01T12:37:52Z slyrus quit (Ping timeout: 240 seconds)
2014-09-01T12:51:39Z flip214 joined #sbcl
2014-09-01T13:33:44Z pranavrc quit
2014-09-01T13:38:44Z jackdani1l is now known as jackdaniel
2014-09-01T13:45:54Z davazp joined #sbcl
2014-09-01T13:47:06Z eudoxia joined #sbcl
2014-09-01T13:48:37Z psilord quit (*.net *.split)
2014-09-01T13:49:40Z psilord joined #sbcl
2014-09-01T13:51:21Z eudoxia quit (Client Quit)
2014-09-01T13:53:15Z eudoxia joined #sbcl
2014-09-01T14:04:03Z attila_lendvai quit (Quit: Leaving.)
2014-09-01T14:04:15Z wbooze joined #sbcl
2014-09-01T14:06:21Z oleo joined #sbcl
2014-09-01T14:27:05Z drmeister quit (Remote host closed the connection)
2014-09-01T14:27:36Z loke_ joined #sbcl
2014-09-01T14:59:25Z DGASAU quit (Remote host closed the connection)
2014-09-01T15:06:06Z DGASAU` joined #sbcl
2014-09-01T15:07:05Z davazp quit (Remote host closed the connection)
2014-09-01T15:12:41Z LiamH quit (Remote host closed the connection)
2014-09-01T15:13:27Z LiamH joined #sbcl
2014-09-01T15:19:08Z DGASAU` is now known as DGASAU
2014-09-01T15:34:10Z flip214 quit (Ping timeout: 250 seconds)
2014-09-01T15:40:15Z flip214 joined #sbcl
2014-09-01T16:06:37Z scymtym_ quit (Remote host closed the connection)
2014-09-01T16:07:29Z scymtym joined #sbcl
2014-09-01T16:15:57Z drmeister joined #sbcl
2014-09-01T16:20:40Z drmeister quit (Ping timeout: 260 seconds)
2014-09-01T16:29:57Z loke_ quit (Remote host closed the connection)
2014-09-01T16:46:56Z edgar-rft quit (Quit: you better not read this)
2014-09-01T16:47:45Z drmeister joined #sbcl
2014-09-01T16:53:52Z LiamH quit (Read error: Connection reset by peer)
2014-09-01T16:55:08Z ASau joined #sbcl
2014-09-01T18:12:49Z eudoxia quit (Quit: Lost terminal)
2014-09-01T18:21:17Z ubii joined #sbcl
2014-09-01T18:29:53Z slyrus joined #sbcl
2014-09-01T19:23:04Z LiamH joined #sbcl
2014-09-01T19:49:24Z sdemarre joined #sbcl
2014-09-01T19:52:16Z scymtym_ joined #sbcl
2014-09-01T20:11:31Z erikvarga left #sbcl
2014-09-01T20:35:02Z edgar-rft joined #sbcl
2014-09-01T20:49:22Z nyef joined #sbcl
2014-09-01T20:49:43Z nyef: Hello all.
2014-09-01T20:55:08Z pkhuong: ni nyef
2014-09-01T20:55:10Z pkhuong: *hi too
2014-09-01T20:55:37Z segv- quit (Remote host closed the connection)
2014-09-01T20:55:37Z fikusz joined #sbcl
2014-09-01T20:58:14Z pkhuong: random ARM q: I have this system that only needs to work on x86-64, but might one day be ported to SPARC or ARM(64). I'm considering a trivial form of runtime codepatching, where I conditionally disable some jumps by rewriting them in test eax, [displacement].
2014-09-01T20:59:12Z pkhuong: That way toggling between test and jmp is a single byte write. Does arm have similar pairs of jmp/almost-nop instructions?
2014-09-01T21:02:15Z pkhuong: on further thought, that's probably a non-issue on ISAs with aligned instructions
2014-09-01T21:04:25Z nyef: Yeah, sounds about right. Aligned instructions more-or-less means that you can do that without too much damage.
2014-09-01T21:13:16Z sdemarre quit (Ping timeout: 260 seconds)
2014-09-01T21:13:49Z pkhuong: for extra fun, I'm partly doing this to elide rarely-contended locks in a multithreaded application ;)
2014-09-01T21:33:37Z Krystof: that cannot possibly end badly
2014-09-01T21:51:20Z attila_lendvai joined #sbcl
2014-09-01T21:51:20Z attila_lendvai quit (Changing host)
2014-09-01T21:51:20Z attila_lendvai joined #sbcl
2014-09-01T21:56:15Z wmarvel: would you even need to toggle on arm? or could you just have cond set up so you float through the jmp bits?
2014-09-01T21:57:07Z wmarvel: (this might be a dumb question, I dunno)
2014-09-01T21:57:11Z eudoxia joined #sbcl
2014-09-01T21:59:56Z wbooze quit (Ping timeout: 260 seconds)
2014-09-01T22:04:33Z pkhuong: wmarvel: the goal is to avoid the "load, test, conditional jump" sequence.
2014-09-01T22:05:25Z wmarvel: and you're avoiding it when? you do some test to check if you want to avoid it, right?
2014-09-01T22:06:57Z wmarvel: and then you're patching the jump so it's a tst?
2014-09-01T22:09:38Z pkhuong: Say you have a loop that executes millions of times per second. load/test/conditional jump in the middle of that loop has a noticeable impact, even if the jump is never taken.
2014-09-01T22:10:06Z wmarvel: right
2014-09-01T22:10:44Z pkhuong: instead, you can have a pseudo-noop (in my case, test eax, #x...) and that'll be negligible even for a trivial loop body
2014-09-01T22:11:08Z wmarvel: will you have a free flag that you can set? if so, you can probably just have the load/test/conditional-jump be nops by whether or not that flag is set/not set
2014-09-01T22:11:33Z wmarvel: unless I'm nuts
2014-09-01T22:11:37Z pkhuong: flags tend to be affected by arithmetic operations
2014-09-01T22:11:50Z wmarvel: right, which is why 'do you have a free ..'
2014-09-01T22:13:00Z wmarvel: (not saying it's actually possible, just thinking about it)
2014-09-01T22:13:46Z pkhuong: anyway, not really. I have a real, large pre-existing code base to work from ;)
2014-09-01T22:14:35Z pkhuong: I expect 64 bit ARM has also specified most instructions to affect all or none of the flags to help out of order execution.
2014-09-01T22:16:27Z wmarvel: I might have read somewhere that 64 bit arm doesn't do condition codes. I might be remembering wrong :)
2014-09-01T22:19:41Z oleo quit (Read error: Connection reset by peer)
2014-09-01T22:20:59Z oleo joined #sbcl
2014-09-01T22:21:02Z fikusz quit (Ping timeout: 245 seconds)
2014-09-01T22:21:39Z pkhuong: no. there are much fewer conditional operations.
2014-09-01T22:22:53Z pkhuong: mostly, they removed stuff from the original ISA that was "free" on older microarchitecture but isn't on superscalar and out of order chips.
2014-09-01T22:23:22Z pkhuong: so, predicated everything, PC as almost a GPR, and bit rotates everywhere.
2014-09-01T22:26:41Z nyef: ... You mean, we finally get a sixteenth GPR?
2014-09-01T22:27:10Z alexander-01 joined #sbcl
2014-09-01T22:27:46Z pkhuong: 31 GPRs.
2014-09-01T22:28:46Z nyef: Luxury!
2014-09-01T22:28:48Z pkhuong: one of the hard FP/SIMD extension baked in.
2014-09-01T22:29:16Z pkhuong: can't remember if it's the mostly useless or almost useful one ;)
2014-09-01T22:29:22Z nyef: Heh.
2014-09-01T22:29:45Z nyef: Well, either way, I'm proposing to not do anything about an arm64 backend any time soon.
2014-09-01T22:29:54Z pkhuong: seconded (:
2014-09-01T22:30:31Z nyef: (After all, we already have enough trouble maintaining the current set of backends.)
2014-09-01T22:34:10Z alexander-01: please excuse me if I'm butting in but I came here looking for help in installing sbcl ob linux deploy which is running on my nexus tablet. Is this even possible?
2014-09-01T22:36:47Z wmarvel: have you tried one of the binary arm builds? (it looks like the nexus is all armv7?)
2014-09-01T22:37:34Z Intensity joined #sbcl
2014-09-01T22:39:01Z alexander-01: the nexus is arm 7. didn't realise there was a binary will go look thanks
2014-09-01T22:43:39Z alexander-01: Thanks for the help. I was only looking at the linux and not reading across the top
2014-09-01T22:44:00Z wmarvel: no worries
2014-09-01T22:48:07Z alexander-01 quit (Ping timeout: 245 seconds)
2014-09-01T22:48:17Z wmarvel ponders the odds that vops won't be short registers if I pull t8 as thread
2014-09-01T22:48:30Z wmarvel: I am guessing "probably not good"
2014-09-01T22:50:29Z wmarvel: err, r8
2014-09-01T23:10:31Z davazp joined #sbcl
2014-09-01T23:18:46Z eudoxia quit (Quit: Lost terminal)
2014-09-02T00:08:49Z nyef: wmarvel: On ARM? You're not likely to be able to scare up ANY registers anywhere.
2014-09-02T00:11:54Z wmarvel: yeah
2014-09-02T00:11:59Z wmarvel: that's what it looks like
2014-09-02T00:22:00Z nyef: ARM has number and control stacks, unlike x86, so you need at least three arg-passing registers (boxed), and there are a couple of other boxed registers required for call-out.
2014-09-02T00:22:07Z nyef: Err... for funcall.
2014-09-02T00:22:15Z nyef: And then you have a similar requirement for call-out.
2014-09-02T00:25:29Z wmarvel: and we have to have or be able to get a pointer to a thread-local environment to have threads, right? Or am I assuming that falsely?
2014-09-02T00:26:02Z wmarvel: so I need to do something like, stash it somewhere, and make sure when we actually need it, it gets put in a reg?
2014-09-02T00:26:26Z wmarvel: rather than just letting it float in a reg all the time like x86 / ppc look like they do?
2014-09-02T00:28:32Z wmarvel: or maybe I'm misunderstanding something
2014-09-02T00:32:33Z nyef: You either need to have the TLS block pointer in a register (ideal) or to be able to easily pull it from thread-specific storage (less ideal).
2014-09-02T00:33:02Z nyef: If necessary, push something else that's currently in a register to a slot on the thread structure... If you can.
2014-09-02T00:39:36Z yacks quit (Ping timeout: 255 seconds)
2014-09-02T00:42:51Z yacks joined #sbcl
2014-09-02T00:57:01Z pkhuong: I bet we could have csp/cfp/nsp/nfp in the thread struct... with safe points (:
2014-09-02T00:57:13Z attila_lendvai quit (Quit: Leaving.)
2014-09-02T01:04:11Z pkhuong: all right, perhaps not cfp.
2014-09-02T01:05:24Z nyef: Also not NSP, because it's the C stack pointer, and having it anywhere else is just begging for trouble.
2014-09-02T01:06:52Z nyef: (One of the last major changes I made to the register assignment for arm-port-2 was to switch CSP into a static symbol and NSP into the register, so we didn't have anything crazy going on with the altstack and stack switching and whatnot, which stabilized things immensely.)
2014-09-02T01:07:52Z pkhuong: safe points!
2014-09-02T01:09:47Z nyef: Altstack only kicks in if the stack pointer value isn't already within the altstack range.
2014-09-02T01:10:04Z pkhuong: well... that sucks :\
2014-09-02T01:10:18Z nyef: Only having 15 registers sucks as well.
2014-09-02T01:52:36Z davazp quit (Remote host closed the connection)
2014-09-02T01:53:51Z scymtym_ quit (Ping timeout: 255 seconds)
2014-09-02T02:11:33Z nicdev quit (Ping timeout: 260 seconds)
2014-09-02T02:12:08Z nyef quit (Quit: G'night all.)
2014-09-02T02:15:25Z echo-area joined #sbcl
2014-09-02T02:34:42Z Bike quit (Quit: leaving)
2014-09-02T02:37:59Z Bike joined #sbcl
2014-09-02T02:38:57Z christoph_debian quit (Ping timeout: 245 seconds)
2014-09-02T02:52:37Z christoph_debian joined #sbcl
2014-09-02T03:02:54Z LiamH quit (Read error: Connection reset by peer)
2014-09-02T04:00:31Z White__Flame quit (Read error: Connection reset by peer)
2014-09-02T04:40:53Z edgar-rft quit (Quit: activity disappeared because all hope lost)
2014-09-02T04:42:10Z echo-area quit (Remote host closed the connection)
2014-09-02T04:43:31Z echo-area joined #sbcl
2014-09-02T04:51:10Z echo-area quit (Remote host closed the connection)
2014-09-02T04:52:04Z echo-area joined #sbcl
2014-09-02T05:01:19Z oleo quit (Quit: Verlassend)
2014-09-02T05:05:53Z echo-area quit (Remote host closed the connection)
2014-09-02T05:21:46Z echo-area joined #sbcl
2014-09-02T05:41:03Z echo-area quit (Remote host closed the connection)
2014-09-02T06:00:15Z sdemarre joined #sbcl
2014-09-02T06:06:35Z pranavrc joined #sbcl
2014-09-02T06:06:35Z pranavrc quit (Changing host)
2014-09-02T06:06:35Z pranavrc joined #sbcl
2014-09-02T06:09:28Z sdemarre quit (Ping timeout: 260 seconds)
2014-09-02T06:35:04Z scymtym_ joined #sbcl
2014-09-02T07:00:03Z edgar-rft joined #sbcl
2014-09-02T07:21:33Z p_l quit (Ping timeout: 240 seconds)
2014-09-02T07:38:17Z echo-area joined #sbcl
2014-09-02T07:39:33Z p_l joined #sbcl
2014-09-02T07:49:48Z ASau quit (Ping timeout: 260 seconds)
2014-09-02T08:41:34Z ASau joined #sbcl
2014-09-02T08:52:21Z luis` is now known as luis
2014-09-02T09:06:51Z erikvarga joined #sbcl
2014-09-02T09:08:23Z zmyrgel joined #sbcl
2014-09-02T09:10:30Z zmyrgel left #sbcl
2014-09-02T09:41:48Z attila_lendvai joined #sbcl
2014-09-02T09:47:57Z ASau quit (Ping timeout: 264 seconds)
2014-09-02T10:28:25Z fikusz joined #sbcl
2014-09-02T10:32:46Z echo-area quit (Remote host closed the connection)
2014-09-02T10:58:21Z scymtym_ quit (Ping timeout: 255 seconds)
2014-09-02T11:04:52Z karswell quit (Remote host closed the connection)
2014-09-02T11:04:57Z karswell` joined #sbcl
2014-09-02T11:07:28Z fikusz quit (Quit: Leaving)
2014-09-02T11:13:49Z fikusz joined #sbcl
2014-09-02T11:29:08Z davazp joined #sbcl
2014-09-02T11:59:34Z davazp quit (Ping timeout: 268 seconds)
2014-09-02T12:05:09Z irsol quit (Ping timeout: 240 seconds)
2014-09-02T12:05:10Z pranavrc quit (Read error: Connection reset by peer)
2014-09-02T12:05:16Z pranavrc joined #sbcl
2014-09-02T12:05:16Z pranavrc quit (Changing host)
2014-09-02T12:05:16Z pranavrc joined #sbcl
2014-09-02T12:15:06Z nyef joined #sbcl
2014-09-02T12:15:17Z nyef: G'morning all.
2014-09-02T12:29:54Z drmeister quit (Remote host closed the connection)
2014-09-02T12:42:29Z yacks quit (Read error: Connection reset by peer)
2014-09-02T13:13:48Z psilord quit (Quit: Leaving.)
2014-09-02T13:28:27Z eudoxia joined #sbcl
2014-09-02T13:37:44Z pranavrc quit
2014-09-02T13:37:51Z loke_ joined #sbcl
2014-09-02T13:45:52Z krzysz00_ quit (Quit: leaving)
2014-09-02T13:49:57Z drmeister joined #sbcl
2014-09-02T14:00:53Z redline6561_ is now known as redline6561
2014-09-02T14:11:10Z psilord joined #sbcl
2014-09-02T14:19:19Z oleo joined #sbcl
2014-09-02T14:29:11Z yacks joined #sbcl
2014-09-02T14:43:27Z asedeno joined #sbcl
2014-09-02T14:46:03Z wmarvel: morning nyef
2014-09-02T15:10:29Z nicdev joined #sbcl
2014-09-02T15:15:50Z irsol joined #sbcl
2014-09-02T15:48:24Z yacks quit (Ping timeout: 246 seconds)
2014-09-02T15:51:53Z yacks joined #sbcl
2014-09-02T15:53:10Z attila_lendvai quit (Quit: Leaving.)
2014-09-02T16:07:17Z slyrus quit (Ping timeout: 245 seconds)
2014-09-02T16:12:42Z rpg joined #sbcl
2014-09-02T16:49:33Z milosn quit (Ping timeout: 240 seconds)
2014-09-02T16:50:24Z milosn joined #sbcl
2014-09-02T16:55:27Z loke_ quit (Remote host closed the connection)
2014-09-02T16:59:04Z psilord quit (Ping timeout: 252 seconds)
2014-09-02T17:00:36Z rpg quit (Ping timeout: 255 seconds)
2014-09-02T17:11:02Z eudoxia quit (Ping timeout: 245 seconds)
2014-09-02T17:15:50Z psilord joined #sbcl
2014-09-02T17:17:11Z eudoxia joined #sbcl
2014-09-02T17:18:39Z sdemarre joined #sbcl
2014-09-02T17:52:09Z sdemarre quit (Ping timeout: 264 seconds)
2014-09-02T18:09:57Z sdemarre joined #sbcl
2014-09-02T18:17:11Z eudoxia quit (Quit: Lost terminal)
2014-09-02T18:24:51Z karswell` quit (Read error: Connection reset by peer)
2014-09-02T18:50:33Z foom2 is now known as foom
2014-09-02T19:00:57Z prxq joined #sbcl
2014-09-02T19:16:18Z ASau joined #sbcl
2014-09-02T19:18:25Z sdemarre quit (Read error: Connection reset by peer)
2014-09-02T19:23:52Z sdemarre joined #sbcl
2014-09-02T19:50:24Z erikvarga left #sbcl
2014-09-02T19:50:27Z slyrus joined #sbcl
2014-09-02T19:52:16Z drmeister quit (Remote host closed the connection)
2014-09-02T19:56:10Z edgar-rft quit (Quit: mental process terminated into eternal corruption)
2014-09-02T19:59:28Z attila_lendvai joined #sbcl
2014-09-02T19:59:28Z attila_lendvai quit (Changing host)
2014-09-02T19:59:28Z attila_lendvai joined #sbcl
2014-09-02T20:18:32Z gingerale quit (Ping timeout: 245 seconds)
2014-09-02T20:23:06Z sdemarre quit (Ping timeout: 255 seconds)
2014-09-02T20:34:29Z karswell` joined #sbcl
2014-09-02T20:40:20Z slyrus_ joined #sbcl
2014-09-02T20:41:16Z slyrus quit (Ping timeout: 252 seconds)
2014-09-02T20:41:30Z slyrus_ is now known as slyrus
2014-09-02T20:46:31Z scymtym_ joined #sbcl
2014-09-02T21:17:04Z psilord quit (Quit: Leaving.)
2014-09-02T21:19:49Z drmeister joined #sbcl
2014-09-02T21:54:47Z alexander-01 joined #sbcl
2014-09-02T21:55:39Z drmeister quit (Remote host closed the connection)
2014-09-02T22:05:11Z ASau` joined #sbcl
2014-09-02T22:07:01Z ASau` quit (Remote host closed the connection)
2014-09-02T22:07:41Z ASau` joined #sbcl
2014-09-02T22:08:57Z ASau quit (Ping timeout: 264 seconds)
2014-09-02T22:09:05Z drmeister joined #sbcl
2014-09-02T22:10:51Z ASau` is now known as ASau
2014-09-02T22:14:48Z alexander-01 quit (Ping timeout: 246 seconds)
2014-09-02T22:32:58Z drmeister quit (Read error: Connection reset by peer)
2014-09-02T22:41:04Z drmeister joined #sbcl
2014-09-02T23:03:17Z psilord joined #sbcl
2014-09-02T23:03:20Z oleo is now known as Guest50585
2014-09-02T23:04:08Z oleo__ joined #sbcl
2014-09-02T23:06:28Z Guest50585 quit (Ping timeout: 252 seconds)
2014-09-02T23:58:22Z slyrus quit (Ping timeout: 260 seconds)
2014-09-03T00:00:27Z scymtym_ quit (Ping timeout: 255 seconds)
2014-09-03T00:08:53Z wmarvel: What exactly does OAOOM stand for? From context it seems roughly/vaguely 'done something like the same thing somewhere else'
2014-09-03T00:09:17Z wmarvel: and I google it and all I get is ref's to sbcl commits using it :)
2014-09-03T00:12:05Z nyef: Once And Only Once More.
2014-09-03T00:12:44Z nyef: The original being "Once And Only Once", but we have several places where we define something once for C and once for the runtime, or once for the cross-compiler and once for the target.
2014-09-03T00:15:29Z wmarvel: now that you've said it I feel like I ought to have just read it that way in the first place ;)
2014-09-03T00:16:25Z nyef: I think it got defined in sbcl-internals.
2014-09-03T00:17:00Z wmarvel: I know I've seen 'once and only once' in the tree a few places
2014-09-03T00:17:29Z wmarvel: I think my brain locked into reading OOM as 'Out Of Memory'
2014-09-03T00:19:02Z nyef: So, any luck on scaring up some way to hold onto an ARM TLS block?
2014-09-03T00:21:43Z wmarvel: Not yet, been work-working
2014-09-03T00:22:10Z wmarvel: I have a notes file with a list of things
2014-09-03T00:22:51Z wmarvel: like, in a multithread build I probably need to stick at least some of the things you defined as static due to lack of registers as something on the thread struct, right?
2014-09-03T00:23:22Z wmarvel: Having different threads have different control stacks seems important, for instance. Unless I'm dumb.
2014-09-03T00:24:19Z wmarvel: and I want to smack myself right now for using 'like' like that in actual text.
2014-09-03T00:24:23Z wmarvel: ;)
2014-09-03T00:27:08Z nyef: Yes, you'd promote the control stack pointer to the thread structure. Or, more accurately, you'd TLS-bind it to a slot in the thread structure, which might already exist.
2014-09-03T01:11:47Z prxq_ joined #sbcl
2014-09-03T01:15:32Z prxq quit (Ping timeout: 276 seconds)
2014-09-03T01:50:02Z echo-area joined #sbcl
2014-09-03T02:13:15Z ubii quit (Quit: Leaving)
2014-09-03T02:27:54Z attila_lendvai quit (Quit: Leaving.)
2014-09-03T02:39:36Z christoph_debian quit (Ping timeout: 260 seconds)
2014-09-03T02:52:50Z christoph_debian joined #sbcl
2014-09-03T03:18:00Z ubii joined #sbcl
2014-09-03T03:24:19Z Fare joined #sbcl
2014-09-03T03:27:29Z Fare quit (Remote host closed the connection)
2014-09-03T04:35:05Z nyef quit (Quit: Sleep well, all.)
2014-09-03T05:34:33Z sdemarre joined #sbcl
2014-09-03T05:39:19Z yacks quit (Quit: Leaving)
2014-09-03T05:39:33Z ASau quit (Ping timeout: 240 seconds)
2014-09-03T05:39:53Z ASau` joined #sbcl
2014-09-03T05:44:52Z sdemarre quit (Ping timeout: 245 seconds)
2014-09-03T06:02:25Z pranavrc joined #sbcl
2014-09-03T06:02:26Z pranavrc quit (Changing host)
2014-09-03T06:02:26Z pranavrc joined #sbcl
2014-09-03T06:15:58Z edgar-rft joined #sbcl
2014-09-03T06:19:24Z yacks joined #sbcl
2014-09-03T06:25:52Z pranavrc quit (Ping timeout: 240 seconds)
2014-09-03T06:26:27Z pranavrc joined #sbcl
2014-09-03T06:26:27Z pranavrc quit (Changing host)
2014-09-03T06:26:27Z pranavrc joined #sbcl
2014-09-03T06:31:32Z ASau` is now known as ASau
2014-09-03T07:04:21Z pkhuong quit (Ping timeout: 255 seconds)
2014-09-03T07:23:01Z pkhuong joined #sbcl
2014-09-03T07:27:46Z ASau quit (Ping timeout: 260 seconds)
2014-09-03T08:20:56Z Krystof: committers: did you get my mail about GSoC students?
2014-09-03T08:31:33Z yacks quit (Ping timeout: 240 seconds)
2014-09-03T08:48:44Z erikvarga joined #sbcl
2014-09-03T08:52:18Z yacks joined #sbcl
2014-09-03T08:53:20Z scymtym_ joined #sbcl
2014-09-03T08:55:44Z prxq_: Krystof: if you mean the mail to sbcl-governance, yes, I saw it.
2014-09-03T08:57:49Z scymtym_: Krystof: i saw it, too. i was going to reply soon.
2014-09-03T08:59:36Z Krystof: fine. :)
2014-09-03T09:10:01Z luis quit (Quit: ZNC - http://znc.sourceforge.net)
2014-09-03T09:10:09Z luis joined #sbcl
2014-09-03T09:11:03Z luis quit (Client Quit)
2014-09-03T09:14:18Z luis joined #sbcl
2014-09-03T09:57:21Z attila_lendvai joined #sbcl
2014-09-03T10:21:10Z PuercoPop joined #sbcl
2014-09-03T10:26:17Z loke joined #sbcl
2014-09-03T11:02:07Z echo-area quit (Remote host closed the connection)
2014-09-03T11:15:05Z scymtym_ quit (Ping timeout: 268 seconds)
2014-09-03T11:30:26Z attila_lendvai quit (Ping timeout: 260 seconds)
2014-09-03T11:36:44Z oleo joined #sbcl
2014-09-03T11:39:07Z davazp joined #sbcl
2014-09-03T11:39:08Z oleo__ quit (Ping timeout: 268 seconds)
2014-09-03T12:11:49Z yacks quit (Quit: Leaving)
2014-09-03T12:15:37Z nyef joined #sbcl
2014-09-03T12:18:06Z erikvarga left #sbcl
2014-09-03T12:35:33Z drmeister quit (Remote host closed the connection)
2014-09-03T12:45:50Z davazp quit (Ping timeout: 276 seconds)
2014-09-03T12:46:31Z drmeister joined #sbcl
2014-09-03T12:53:30Z drmeister quit (Read error: Connection reset by peer)
2014-09-03T12:55:42Z drmeister joined #sbcl
2014-09-03T13:00:33Z yacks joined #sbcl
2014-09-03T13:09:27Z pranavrc quit
2014-09-03T13:17:05Z drmeister quit (Read error: Connection reset by peer)
2014-09-03T13:33:15Z eudoxia joined #sbcl
2014-09-03T13:36:32Z loke_ joined #sbcl
2014-09-03T13:45:02Z eudoxia quit (Quit: Lost terminal)
2014-09-03T13:47:05Z loke_: Hello guys
2014-09-03T13:49:16Z eudoxia joined #sbcl
2014-09-03T14:05:03Z leo2007 quit (Ping timeout: 246 seconds)
2014-09-03T14:23:20Z gingerale joined #sbcl
2014-09-03T14:25:48Z eudoxia quit (Quit: Lost terminal)
2014-09-03T14:41:46Z leo2007 joined #sbcl
2014-09-03T15:38:32Z ubii quit (Ping timeout: 245 seconds)
2014-09-03T15:40:44Z ubii joined #sbcl
2014-09-03T15:46:06Z wheelsucker joined #sbcl
2014-09-03T15:50:02Z ubii quit (Ping timeout: 252 seconds)
2014-09-03T15:56:24Z ubii joined #sbcl
2014-09-03T16:02:22Z gale joined #sbcl
2014-09-03T16:03:11Z joshee joined #sbcl
2014-09-03T16:03:54Z luis` joined #sbcl
2014-09-03T16:04:30Z drmeister joined #sbcl
2014-09-03T16:04:49Z _8hzp joined #sbcl
2014-09-03T16:04:58Z thoto joined #sbcl
2014-09-03T16:05:56Z asedeno_ joined #sbcl
2014-09-03T16:06:08Z Blkt_ joined #sbcl
2014-09-03T16:06:53Z pchrist_ joined #sbcl
2014-09-03T16:06:56Z fe[nl]ix joined #sbcl
2014-09-03T16:07:03Z ferada_ joined #sbcl
2014-09-03T16:08:05Z jdz_ joined #sbcl
2014-09-03T16:08:10Z qbit_ joined #sbcl
2014-09-03T16:08:30Z joshe quit (Ping timeout: 260 seconds)
2014-09-03T16:08:31Z luis quit (Ping timeout: 260 seconds)
2014-09-03T16:08:31Z phf quit (Ping timeout: 260 seconds)
2014-09-03T16:08:31Z asedeno quit (Ping timeout: 260 seconds)
2014-09-03T16:08:32Z gingerale quit (Ping timeout: 260 seconds)
2014-09-03T16:08:32Z qbit quit (Ping timeout: 260 seconds)
2014-09-03T16:08:33Z thoto_ quit (Ping timeout: 260 seconds)
2014-09-03T16:08:33Z hzp quit (Ping timeout: 260 seconds)
2014-09-03T16:08:33Z pchrist quit (Ping timeout: 260 seconds)
2014-09-03T16:08:34Z lacedaemon quit (Ping timeout: 260 seconds)
2014-09-03T16:08:34Z Blkt quit (Ping timeout: 260 seconds)
2014-09-03T16:08:34Z ferada quit (Ping timeout: 260 seconds)
2014-09-03T16:08:34Z jdz quit (Ping timeout: 260 seconds)
2014-09-03T16:08:34Z yacks quit (Ping timeout: 260 seconds)
2014-09-03T16:08:35Z leoc quit (Ping timeout: 260 seconds)
2014-09-03T16:08:51Z gale is now known as gingerale
2014-09-03T16:09:12Z qbit_ is now known as qbit
2014-09-03T16:09:13Z phf joined #sbcl
2014-09-03T16:10:47Z yacks joined #sbcl
2014-09-03T16:11:30Z leoc joined #sbcl
2014-09-03T16:11:58Z jdz_ is now known as jdz
2014-09-03T16:13:10Z joshee is now known as joshe
2014-09-03T16:26:26Z eudoxia joined #sbcl
2014-09-03T16:36:22Z luis` is now known as luis
2014-09-03T16:58:15Z sdemarre joined #sbcl
2014-09-03T17:14:07Z loke_ quit (Remote host closed the connection)
2014-09-03T17:53:57Z scymtym_ joined #sbcl
2014-09-03T18:21:45Z eudoxia quit (Quit: Lost terminal)
2014-09-03T18:25:42Z davazp joined #sbcl
2014-09-03T18:36:47Z slyrus joined #sbcl
2014-09-03T18:46:16Z erikvarga joined #sbcl
2014-09-03T19:24:01Z ASau joined #sbcl
2014-09-03T19:33:58Z Bicyclidine joined #sbcl
2014-09-03T19:34:27Z gingerale quit (Ping timeout: 245 seconds)
2014-09-03T19:55:09Z pkhuong: Krystof: nope
2014-09-03T19:56:48Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-03T19:59:45Z slyrus quit (Remote host closed the connection)
2014-09-03T20:04:27Z Krystof: pkhuong: hm, curses
2014-09-03T20:04:48Z sdemarre quit (Ping timeout: 268 seconds)
2014-09-03T20:05:19Z Bicyclidine joined #sbcl
2014-09-03T20:06:21Z Krystof: ok, added you.  https://sourceforge.net/p/sbcl/mailman/sbcl-governance/ has archives
2014-09-03T20:08:26Z edgar-rft quit (Quit: conversation discontinued by sudden death)
2014-09-03T20:13:59Z attila_lendvai joined #sbcl
2014-09-03T20:19:22Z ferada_ is now known as ferada
2014-09-03T20:30:23Z davazp quit (Ping timeout: 240 seconds)
2014-09-03T20:32:14Z krzysz00 joined #sbcl
2014-09-03T20:34:37Z erikvarga left #sbcl
2014-09-03T20:50:37Z wmarvel: Hmm. Armv7 multicore system-on-a-board just showed up ;)
2014-09-03T21:03:07Z Bicyclidine quit (Ping timeout: 245 seconds)
2014-09-03T21:06:08Z karswell` is now known as karswell
2014-09-03T21:16:13Z attila_lendvai quit (Quit: Leaving.)
2014-09-03T21:17:22Z attila_lendvai joined #sbcl
2014-09-03T21:17:22Z attila_lendvai quit (Changing host)
2014-09-03T21:17:22Z attila_lendvai joined #sbcl
2014-09-03T21:20:16Z Bicyclidine joined #sbcl
2014-09-03T21:22:41Z attila_lendvai quit (Quit: Leaving.)
2014-09-03T21:25:11Z attila_lendvai joined #sbcl
2014-09-03T21:29:22Z ASau quit (Remote host closed the connection)
2014-09-03T21:33:59Z ASau joined #sbcl
2014-09-03T21:45:58Z wheelsucker quit (Quit: Client Quit)
2014-09-03T21:53:52Z Bicyclidine quit (Ping timeout: 240 seconds)
2014-09-03T22:03:40Z Bicyclidine joined #sbcl
2014-09-03T22:17:52Z krzysz00 quit (Quit: leaving)
2014-09-03T22:32:24Z davazp joined #sbcl
2014-09-03T22:54:16Z prxq_ quit (Ping timeout: 252 seconds)
2014-09-03T23:03:48Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-03T23:14:48Z Bicyclidine joined #sbcl
2014-09-03T23:52:52Z Krystof quit (Ping timeout: 250 seconds)
2014-09-04T00:05:02Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-04T00:05:11Z carvite quit (Ping timeout: 272 seconds)
2014-09-04T00:05:44Z carvite joined #sbcl
2014-09-04T00:08:30Z oleo is now known as Guest31080
2014-09-04T00:09:18Z oleo__ joined #sbcl
2014-09-04T00:09:25Z flip214_ joined #sbcl
2014-09-04T00:10:07Z angavrilov_ joined #sbcl
2014-09-04T00:11:15Z Guest31080 quit (Ping timeout: 246 seconds)
2014-09-04T00:11:54Z p_l_ joined #sbcl
2014-09-04T00:12:10Z leo2007 quit (Ping timeout: 255 seconds)
2014-09-04T00:12:11Z flip214 quit (Ping timeout: 243 seconds)
2014-09-04T00:12:12Z p_l quit (Ping timeout: 255 seconds)
2014-09-04T00:12:12Z Intensity quit (Ping timeout: 255 seconds)
2014-09-04T00:12:12Z ^Posterdati^ quit (Ping timeout: 255 seconds)
2014-09-04T00:12:12Z fikusz quit (Ping timeout: 255 seconds)
2014-09-04T00:12:12Z angavrilov quit (Ping timeout: 255 seconds)
2014-09-04T00:12:13Z Intensity joined #sbcl
2014-09-04T00:12:13Z Intensity quit (Changing host)
2014-09-04T00:12:13Z Intensity joined #sbcl
2014-09-04T00:12:18Z fikusz_ joined #sbcl
2014-09-04T00:12:46Z ^Posterdati^ joined #sbcl
2014-09-04T00:13:18Z p_l_ is now known as p_l
2014-09-04T00:19:02Z scymtym_ quit (Ping timeout: 260 seconds)
2014-09-04T00:24:03Z leo2007 joined #sbcl
2014-09-04T00:27:35Z attila_lendvai quit (Quit: Leaving.)
2014-09-04T00:27:55Z drmeister quit (Remote host closed the connection)
2014-09-04T00:51:04Z drmeister joined #sbcl
2014-09-04T00:52:12Z Intensity quit (Ping timeout: 255 seconds)
2014-09-04T01:00:07Z drmeister quit (Read error: Connection reset by peer)
2014-09-04T01:08:00Z drmeister joined #sbcl
2014-09-04T01:17:24Z davazp quit (Remote host closed the connection)
2014-09-04T01:58:05Z wmarvel: weird. On this new armv7 box I have, clearing the current/accrued exceptions works ... but setting the traps doesn't.
2014-09-04T01:58:32Z wmarvel: (I should have said floating point in there somewhere)
2014-09-04T01:58:44Z wmarvel: I'll dig at that at some point
2014-09-04T02:22:38Z nyef: There's such a profusion of FPU hardware options on ARM, some of which is badly crippled in one way or another...
2014-09-04T02:39:28Z christoph_debian quit (Ping timeout: 268 seconds)
2014-09-04T02:48:35Z wmarvel: /proc/cpuinfo says vfp, vfpv3 and neon
2014-09-04T02:48:55Z wmarvel: sbcl comes up with :TRAPS nil
2014-09-04T02:49:00Z Intensity joined #sbcl
2014-09-04T02:50:32Z wmarvel: I'm mildly wondering if it's something to do w/ having neon on there and flags being slightly different or something silly like that. Or maybe it wants vmsr instead of fmsr or something like that.
2014-09-04T02:52:36Z christoph_debian joined #sbcl
2014-09-04T02:58:00Z |3b| has vfp neon vfpv3 vfpv4 (and other random stuff), and can't set traps on 1.2.1
2014-09-04T02:59:35Z |3b|: and google says those don't support traps, so i guess that makes sense ( https://sourceware.org/ml/libc-ports/2012-06/msg00021.html )
2014-09-04T03:30:43Z krzysz00 joined #sbcl
2014-09-04T03:41:16Z White_Flame joined #sbcl
2014-09-04T04:40:13Z wmarvel: ahh, yeah, that's probably it.
2014-09-04T04:44:46Z nyef quit (Quit: G'night all)
2014-09-04T05:03:58Z oleo__ quit (Quit: Verlassend)
2014-09-04T05:14:02Z pranavrc joined #sbcl
2014-09-04T05:14:05Z pranavrc quit (Changing host)
2014-09-04T05:14:05Z pranavrc joined #sbcl
2014-09-04T05:33:12Z drmeister quit (Remote host closed the connection)
2014-09-04T05:42:49Z slyrus joined #sbcl
2014-09-04T05:49:38Z sdemarre joined #sbcl
2014-09-04T05:50:45Z yacks quit (Ping timeout: 246 seconds)
2014-09-04T05:51:53Z yacks joined #sbcl
2014-09-04T05:56:26Z sdemarre quit (Ping timeout: 260 seconds)
2014-09-04T06:02:54Z slyrus quit (Ping timeout: 252 seconds)
2014-09-04T06:07:24Z prxq joined #sbcl
2014-09-04T06:14:42Z sdemarre joined #sbcl
2014-09-04T06:15:04Z slyrus joined #sbcl
2014-09-04T06:22:57Z sdemarre quit (Ping timeout: 246 seconds)
2014-09-04T06:34:12Z edgar-rft joined #sbcl
2014-09-04T07:05:44Z Krystof joined #sbcl
2014-09-04T07:05:45Z ChanServ has set mode +o Krystof
2014-09-04T07:22:10Z drmeister joined #sbcl
2014-09-04T07:26:39Z drmeister quit (Ping timeout: 246 seconds)
2014-09-04T07:37:14Z ASau quit (Ping timeout: 260 seconds)
2014-09-04T07:56:16Z flip214_ is now known as flip214
2014-09-04T07:56:24Z flip214 quit (Changing host)
2014-09-04T07:56:25Z flip214 joined #sbcl
2014-09-04T08:17:19Z erikvarga joined #sbcl
2014-09-04T09:10:52Z drmeister joined #sbcl
2014-09-04T09:15:33Z drmeister quit (Ping timeout: 260 seconds)
2014-09-04T10:59:41Z drmeister joined #sbcl
2014-09-04T11:03:52Z drmeister quit (Ping timeout: 240 seconds)
2014-09-04T11:53:08Z ccl-logbot joined #sbcl
2014-09-04T11:53:08Z 
2014-09-04T11:53:08Z names: ccl-logbot les_ scymtym_ erikvarga @Krystof edgar-rft slyrus prxq yacks pranavrc White_Flame krzysz00 christoph_debian Intensity leo2007 ^Posterdati^ fikusz_ p_l angavrilov_ flip214 carvite leoc phf qbit jdz ferada fe[nl]ix pchrist_ Blkt_ asedeno_ thoto _8hzp luis joshe ubii loke PuercoPop pkhuong psilord karswell milosn irsol nicdev Bike DGASAU hlavaty heddwch ivan`` sav redline6561 jackdaniel jsnell_ weissschloss foom |3b| reb` akkad antoszka wmarvel
2014-09-04T11:53:08Z names: Perlboy cracauer_ specbot yauz_2 alchemis7 faheem_ minion arrdem p_l|backup mood brucem
2014-09-04T12:35:27Z Krystof: ;; WARNING: Not %100 correct. Removes forms like (... if-match ...) from the
2014-09-04T12:35:28Z Krystof: ;; sub-lexical scope even though this isn't an invocation of the macro.
2014-09-04T12:35:29Z Krystof: haha
2014-09-04T12:36:31Z Krystof: looks like I have fodder for my next backquote-related blog
2014-09-04T12:42:46Z pranavrc quit
2014-09-04T13:06:58Z psilord quit (Quit: Leaving.)
2014-09-04T13:09:05Z InfusoElAmbulant joined #sbcl
2014-09-04T13:19:02Z eudoxia joined #sbcl
2014-09-04T13:33:42Z oleo joined #sbcl
2014-09-04T13:34:45Z sav quit (Ping timeout: 240 seconds)
2014-09-04T13:35:53Z tmh_ joined #sbcl
2014-09-04T13:58:12Z drmeister joined #sbcl
2014-09-04T14:02:57Z erikvarga left #sbcl
2014-09-04T14:08:25Z angavrilov joined #sbcl
2014-09-04T14:09:53Z psilord joined #sbcl
2014-09-04T14:11:00Z les joined #sbcl
2014-09-04T14:11:00Z les quit (Changing host)
2014-09-04T14:11:00Z les joined #sbcl
2014-09-04T14:11:02Z gingerale joined #sbcl
2014-09-04T14:13:13Z slyrus_ joined #sbcl
2014-09-04T14:14:41Z leoc_ joined #sbcl
2014-09-04T14:17:03Z tmh_ quit (*.net *.split)
2014-09-04T14:17:03Z les_ quit (*.net *.split)
2014-09-04T14:17:03Z slyrus quit (*.net *.split)
2014-09-04T14:17:03Z ^Posterdati^ quit (*.net *.split)
2014-09-04T14:17:03Z angavrilov_ quit (*.net *.split)
2014-09-04T14:17:03Z leoc quit (*.net *.split)
2014-09-04T14:17:08Z leoc_ is now known as leoc
2014-09-04T14:17:33Z krzysz00 quit (Ping timeout: 246 seconds)
2014-09-04T14:18:24Z Krystof:     #+sbcl
2014-09-04T14:18:25Z Krystof:     (if (string-lessp (lisp-implementation-version) "1.2.2")
2014-09-04T14:18:58Z Krystof: oh dear oh dear oh dear
2014-09-04T14:18:58Z slyrus_ is now known as slyrus
2014-09-04T14:20:51Z ^Posterdati^ joined #sbcl
2014-09-04T14:21:11Z Krystof: wait, this is just all wrong
2014-09-04T14:24:06Z ams joined #sbcl
2014-09-04T14:32:58Z pkhuong: ... didn't I add a function to sb-ext exactly for that use case?
2014-09-04T14:33:27Z 17SAA4BHF joined #sbcl
2014-09-04T14:33:27Z angavrilov_ joined #sbcl
2014-09-04T14:33:27Z 17SAA33JG joined #sbcl
2014-09-04T14:33:33Z tmh_ joined #sbcl
2014-09-04T14:34:00Z 17SAA4BHF quit (Ping timeout: 244 seconds)
2014-09-04T14:34:01Z angavrilov_ quit (Ping timeout: 244 seconds)
2014-09-04T14:34:07Z 17SAA33JG quit (Ping timeout: 244 seconds)
2014-09-04T14:38:59Z Krystof: pkhuong: I have identified four defects in three lines of code :(
2014-09-04T14:44:29Z Krystof: pkhuong: you had a codewalking experiment somewhere, didn't you?
2014-09-04T14:44:40Z pkhuong: I did.
2014-09-04T14:45:12Z pkhuong: I think I pushed that branch to github; do you think I should try and polish it?
2014-09-04T14:46:48Z pkhuong: https://github.com/pkhuong/sbcl/tree/composable-codewalking
2014-09-04T14:47:59Z Krystof: I don't know yet
2014-09-04T14:48:26Z Krystof: codewalking is on my mind because of a possibly-reckless desire to find out how hard it is to use correctly
2014-09-04T14:49:07Z pkhuong: the interface in that branch is a lot weaker than full codewalking
2014-09-04T14:49:16Z pkhuong: but I believe it's also as hard to use incorrectly as macros.
2014-09-04T14:59:25Z wheelsucker joined #sbcl
2014-09-04T15:10:44Z scymtym joined #sbcl
2014-09-04T16:08:16Z krzysz00 joined #sbcl
2014-09-04T16:13:22Z krzysz00 quit (Ping timeout: 240 seconds)
2014-09-04T16:28:25Z Skrylar joined #sbcl
2014-09-04T16:28:27Z Skrylar: meep
2014-09-04T16:50:32Z krzysz00 joined #sbcl
2014-09-04T16:57:47Z erikvarga joined #sbcl
2014-09-04T17:11:39Z krzysz00 quit (Quit: leaving)
2014-09-04T17:51:20Z hlavaty left #sbcl
2014-09-04T17:56:07Z slyrus quit (Ping timeout: 245 seconds)
2014-09-04T17:57:42Z eudoxia quit (Quit: Lost terminal)
2014-09-04T18:12:58Z DGASAU quit (Remote host closed the connection)
2014-09-04T18:13:46Z DGASAU joined #sbcl
2014-09-04T18:26:59Z angavrilov quit (Remote host closed the connection)
2014-09-04T18:30:55Z p_l quit (Remote host closed the connection)
2014-09-04T18:40:35Z DGASAU quit (Remote host closed the connection)
2014-09-04T18:41:29Z DGASAU joined #sbcl
2014-09-04T18:43:21Z attila_lendvai joined #sbcl
2014-09-04T18:43:21Z attila_lendvai quit (Changing host)
2014-09-04T18:43:21Z attila_lendvai joined #sbcl
2014-09-04T18:58:24Z DGASAU quit (Ping timeout: 252 seconds)
2014-09-04T19:02:30Z DGASAU joined #sbcl
2014-09-04T19:06:02Z PuercoPop left #sbcl
2014-09-04T19:12:05Z drmeister quit (Remote host closed the connection)
2014-09-04T19:28:16Z DGASAU quit (Remote host closed the connection)
2014-09-04T19:29:20Z DGASAU joined #sbcl
2014-09-04T19:45:42Z slyrus joined #sbcl
2014-09-04T19:53:22Z ASau joined #sbcl
2014-09-04T20:13:36Z drmeiste_ joined #sbcl
2014-09-04T20:15:14Z drmeiste_ is now known as drmeister_
2014-09-04T20:16:10Z erikvarga left #sbcl
2014-09-04T20:27:44Z drmeister_ quit (Read error: Connection reset by peer)
2014-09-04T20:31:02Z gingerale quit (Ping timeout: 245 seconds)
2014-09-04T20:49:53Z rpg joined #sbcl
2014-09-04T20:57:49Z Bicyclidine joined #sbcl
2014-09-04T21:41:00Z prxq quit (Remote host closed the connection)
2014-09-04T21:42:39Z wheelsucker quit (Quit: Client Quit)
2014-09-04T21:52:04Z psilord quit (Quit: Leaving.)
2014-09-04T22:16:44Z oleo__ joined #sbcl
2014-09-04T22:18:08Z oleo is now known as Guest35756
2014-09-04T22:18:48Z Guest35756 quit (Ping timeout: 246 seconds)
2014-09-04T22:58:02Z edgar-rft quit (Quit: activity closed by sudden death)
2014-09-04T23:01:26Z rpg quit (Quit: rpg)
2014-09-04T23:16:53Z psilord joined #sbcl
2014-09-04T23:56:27Z attila_lendvai quit (Quit: Leaving.)
2014-09-05T00:48:07Z scymtym quit (Ping timeout: 245 seconds)
2014-09-05T01:02:00Z InfusoElAmbulant quit (Ping timeout: 272 seconds)
2014-09-05T01:20:21Z LiamH joined #sbcl
2014-09-05T01:31:47Z Bicyclidine quit (Ping timeout: 255 seconds)
2014-09-05T01:48:27Z slyrus quit (Ping timeout: 255 seconds)
2014-09-05T02:10:11Z krzysz00 joined #sbcl
2014-09-05T02:32:21Z LiamH quit (Read error: Connection reset by peer)
2014-09-05T02:34:57Z kanru joined #sbcl
2014-09-05T02:39:10Z christoph_debian quit (Ping timeout: 260 seconds)
2014-09-05T02:52:06Z christoph_debian joined #sbcl
2014-09-05T03:35:16Z p_l joined #sbcl
2014-09-05T03:54:40Z krzysz00 quit (Quit: leaving)
2014-09-05T04:03:07Z echo-area joined #sbcl
2014-09-05T04:06:23Z wmarvel quit (Quit: Lost terminal)
2014-09-05T04:20:28Z angavrilov joined #sbcl
2014-09-05T05:26:30Z oleo__ quit (Quit: Verlassend)
2014-09-05T05:51:28Z pranavrc joined #sbcl
2014-09-05T05:51:28Z pranavrc quit (Changing host)
2014-09-05T05:51:28Z pranavrc joined #sbcl
2014-09-05T05:59:19Z kanru quit (Remote host closed the connection)
2014-09-05T06:00:50Z kanru joined #sbcl
2014-09-05T07:06:09Z ASau quit (Ping timeout: 255 seconds)
2014-09-05T07:41:08Z drmeiste_ joined #sbcl
2014-09-05T07:47:03Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-05T07:51:13Z drmeiste_ joined #sbcl
2014-09-05T07:53:59Z attila_lendvai joined #sbcl
2014-09-05T07:54:05Z attila_lendvai quit (Changing host)
2014-09-05T07:54:05Z attila_lendvai joined #sbcl
2014-09-05T07:56:15Z edgar-rft joined #sbcl
2014-09-05T08:10:32Z carvite quit (Ping timeout: 268 seconds)
2014-09-05T08:11:17Z carvite joined #sbcl
2014-09-05T08:13:34Z attila_lendvai quit (Quit: Leaving.)
2014-09-05T08:48:21Z drmeiste_ quit (Remote host closed the connection)
2014-09-05T09:03:52Z DGASAU quit (Ping timeout: 240 seconds)
2014-09-05T09:05:48Z DGASAU joined #sbcl
2014-09-05T09:36:25Z loke_ joined #sbcl
2014-09-05T09:43:34Z milanj joined #sbcl
2014-09-05T10:15:20Z attila_lendvai joined #sbcl
2014-09-05T10:37:12Z heddwch quit (Ping timeout: 250 seconds)
2014-09-05T10:43:37Z heddwch joined #sbcl
2014-09-05T10:49:14Z drmeiste_ joined #sbcl
2014-09-05T10:49:34Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-05T11:02:11Z echo-area quit (Read error: Connection reset by peer)
2014-09-05T11:50:45Z fridim_ joined #sbcl
2014-09-05T12:52:22Z erikvarga joined #sbcl
2014-09-05T13:04:50Z irsol quit (Ping timeout: 276 seconds)
2014-09-05T13:07:46Z joshe quit (Ping timeout: 268 seconds)
2014-09-05T13:11:54Z irsol joined #sbcl
2014-09-05T13:22:02Z eudoxia joined #sbcl
2014-09-05T13:29:07Z oleo joined #sbcl
2014-09-05T13:34:49Z pranavrc quit
2014-09-05T14:17:52Z leo2007 quit (Ping timeout: 240 seconds)
2014-09-05T14:23:32Z gingerale joined #sbcl
2014-09-05T14:37:30Z kanru` joined #sbcl
2014-09-05T14:39:12Z wheelsucker joined #sbcl
2014-09-05T15:21:32Z kanru` quit (Ping timeout: 245 seconds)
2014-09-05T15:27:57Z drmeiste_ joined #sbcl
2014-09-05T15:38:31Z attila_lendvai quit (Quit: Leaving.)
2014-09-05T16:15:25Z loke_ quit (Quit: Leaving)
2014-09-05T16:43:15Z scymtym joined #sbcl
2014-09-05T16:51:57Z fridim_ quit (Ping timeout: 245 seconds)
2014-09-05T17:06:00Z eudoxia quit (Quit: wow lunchtime)
2014-09-05T17:15:58Z slyrus joined #sbcl
2014-09-05T17:17:09Z reb` quit (Ping timeout: 240 seconds)
2014-09-05T17:28:31Z pchrist_ is now known as pchrist
2014-09-05T17:29:01Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-05T17:30:18Z drmeiste_ joined #sbcl
2014-09-05T17:53:37Z slyrus quit (Ping timeout: 245 seconds)
2014-09-05T18:04:51Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-05T18:09:13Z drmeiste_ joined #sbcl
2014-09-05T18:19:44Z eudoxia joined #sbcl
2014-09-05T18:30:05Z carvite quit (Ping timeout: 260 seconds)
2014-09-05T18:31:17Z carvite joined #sbcl
2014-09-05T18:45:34Z oleo quit (Quit: Verlassend)
2014-09-05T18:49:40Z eudoxia quit (Quit: Lost terminal)
2014-09-05T18:49:54Z oleo joined #sbcl
2014-09-05T18:52:54Z brown joined #sbcl
2014-09-05T18:54:47Z gingerale quit (Ping timeout: 245 seconds)
2014-09-05T18:56:21Z brown quit (Disconnected by services)
2014-09-05T19:02:21Z scymtym quit (Ping timeout: 264 seconds)
2014-09-05T19:18:53Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-05T19:20:29Z milanj quit (Quit: Leaving)
2014-09-05T19:37:52Z slyrus joined #sbcl
2014-09-05T19:39:34Z slyrus_ joined #sbcl
2014-09-05T19:43:40Z slyrus quit (Ping timeout: 268 seconds)
2014-09-05T19:43:43Z slyrus_ is now known as slyrus
2014-09-05T19:51:15Z ccl-logbot joined #sbcl
2014-09-05T19:51:15Z 
2014-09-05T19:51:15Z names: ccl-logbot slyrus oleo carvite wheelsucker irsol erikvarga heddwch DGASAU edgar-rft kanru angavrilov p_l christoph_debian psilord Skrylar tmh_ ams ^Posterdati^ leoc les scymtym_ @Krystof yacks White_Flame Intensity fikusz_ flip214 phf qbit jdz ferada fe[nl]ix pchrist Blkt_ asedeno_ thoto _8hzp luis ubii loke pkhuong karswell milosn nicdev Bike ivan`` redline6561 jackdaniel jsnell_ mood brucem p_l|backup arrdem minion faheem_ alchemis7 yauz_2 specbot
2014-09-05T19:51:15Z names: cracauer_ Perlboy antoszka akkad |3b| foom weissschloss
2014-09-05T19:55:01Z oleo__ joined #sbcl
2014-09-05T19:55:08Z slyrus quit (Ping timeout: 252 seconds)
2014-09-05T19:55:17Z oleo is now known as Guest1627
2014-09-05T19:55:35Z oleo__ quit (Read error: Connection reset by peer)
2014-09-05T19:56:02Z Guest1627 quit (Ping timeout: 245 seconds)
2014-09-05T19:57:33Z Hydan joined #sbcl
2014-09-05T19:58:14Z oleo__ joined #sbcl
2014-09-05T19:58:29Z oleo__ is now known as oleo
2014-09-05T20:04:40Z Bicyclidine joined #sbcl
2014-09-05T20:04:59Z erikvarga left #sbcl
2014-09-05T20:07:05Z Hydan quit (Quit: -a- IRC for Android 2.1.19)
2014-09-05T20:07:28Z slyrus joined #sbcl
2014-09-05T20:53:30Z wmarvel joined #sbcl
2014-09-05T21:02:42Z Bicyclidine quit (Ping timeout: 255 seconds)
2014-09-05T21:09:34Z Bicyclidine joined #sbcl
2014-09-05T21:14:38Z slyrus_ joined #sbcl
2014-09-05T21:15:33Z Bicyclidine quit (Ping timeout: 264 seconds)
2014-09-05T21:16:27Z slyrus quit (Ping timeout: 245 seconds)
2014-09-05T21:16:39Z slyrus_ is now known as slyrus
2014-09-05T21:18:56Z Bicyclidine joined #sbcl
2014-09-05T21:44:20Z wheelsucker quit (Quit: Client Quit)
2014-09-05T21:46:00Z edgar-rft quit (Quit: no meaningful message available)
2014-09-05T21:48:07Z slyrus quit (Ping timeout: 245 seconds)
2014-09-05T22:00:00Z slyrus joined #sbcl
2014-09-05T22:09:25Z ASau joined #sbcl
2014-09-05T22:16:46Z slyrus_ joined #sbcl
2014-09-05T22:19:41Z slyrus quit (Ping timeout: 268 seconds)
2014-09-05T22:19:47Z slyrus_ is now known as slyrus
2014-09-05T22:28:08Z drmeiste_ joined #sbcl
2014-09-05T22:41:42Z ASau quit (Ping timeout: 255 seconds)
2014-09-05T22:50:17Z slyrus quit (Ping timeout: 245 seconds)
2014-09-05T22:56:05Z Bicyclidine quit (Ping timeout: 260 seconds)
2014-09-05T23:06:15Z Bicyclidine joined #sbcl
2014-09-05T23:15:11Z slyrus joined #sbcl
2014-09-05T23:20:22Z heddwch quit (Ping timeout: 240 seconds)
2014-09-05T23:22:46Z heddwch joined #sbcl
2014-09-05T23:35:36Z slyrus_ joined #sbcl
2014-09-05T23:38:17Z slyrus__ joined #sbcl
2014-09-05T23:38:37Z slyrus quit (Ping timeout: 268 seconds)
2014-09-05T23:41:39Z slyrus_ quit (Ping timeout: 246 seconds)
2014-09-05T23:42:49Z drmeiste_ quit (Remote host closed the connection)
2014-09-05T23:43:36Z slyrus joined #sbcl
2014-09-05T23:47:14Z slyrus__ quit (Ping timeout: 252 seconds)
2014-09-05T23:50:34Z slyrus_ joined #sbcl
2014-09-05T23:53:12Z slyrus quit (Ping timeout: 245 seconds)
2014-09-05T23:53:13Z slyrus_ is now known as slyrus
2014-09-05T23:55:16Z heddwch quit (Ping timeout: 268 seconds)
2014-09-05T23:55:49Z heddwch joined #sbcl
2014-09-06T00:00:12Z heddwch quit (Ping timeout: 246 seconds)
2014-09-06T00:03:50Z heddwch joined #sbcl
2014-09-06T00:04:28Z slyrus quit (Ping timeout: 252 seconds)
2014-09-06T00:04:28Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-06T00:29:05Z ubii quit (Quit: Leaving)
2014-09-06T00:59:11Z drmeiste_ joined #sbcl
2014-09-06T01:06:48Z leo2007 joined #sbcl
2014-09-06T01:23:51Z heddwch quit (Ping timeout: 246 seconds)
2014-09-06T01:28:28Z heddwch joined #sbcl
2014-09-06T01:29:36Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T01:53:04Z drmeiste_ joined #sbcl
2014-09-06T02:39:12Z christoph_debian quit (Ping timeout: 252 seconds)
2014-09-06T02:53:22Z christoph_debian joined #sbcl
2014-09-06T02:54:58Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T03:17:10Z _8hzp` joined #sbcl
2014-09-06T03:21:14Z _8hzp quit (Ping timeout: 268 seconds)
2014-09-06T03:24:13Z wmarvel: interesting. How terrible would it be to have a vop that would slam the tls block pointer into a reg in one instruction (on arm?)
2014-09-06T03:24:33Z wmarvel: I'm not sure if I what I'm thinking will work but if it will I could do that.
2014-09-06T03:32:12Z wmarvel: maybe set the tn up so it does that or something? (don't know how doable that is but maybe I'll figure it out soon)
2014-09-06T04:13:32Z drmeiste_ joined #sbcl
2014-09-06T04:14:06Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T04:32:13Z drmeiste_ joined #sbcl
2014-09-06T04:41:05Z gingerale joined #sbcl
2014-09-06T05:54:09Z edgar-rft joined #sbcl
2014-09-06T06:08:34Z slyrus joined #sbcl
2014-09-06T06:12:41Z _8hzp`: I like how the read-n-bytes issue sounds almost identical to CMUCL problems from 10 years ago.
2014-09-06T07:12:58Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T07:13:38Z drmeiste_ joined #sbcl
2014-09-06T07:18:03Z drmeiste_ quit (Ping timeout: 246 seconds)
2014-09-06T07:46:09Z leo2007 quit (Ping timeout: 264 seconds)
2014-09-06T08:30:41Z erikvarga joined #sbcl
2014-09-06T08:34:43Z oleo is now known as Guest7881
2014-09-06T08:35:28Z oleo__ joined #sbcl
2014-09-06T08:37:56Z leo2007 joined #sbcl
2014-09-06T08:38:44Z Guest7881 quit (Ping timeout: 276 seconds)
2014-09-06T09:08:27Z weissschloss quit (Remote host closed the connection)
2014-09-06T09:11:21Z weissschloss joined #sbcl
2014-09-06T09:13:55Z drmeiste_ joined #sbcl
2014-09-06T09:18:54Z drmeiste_ quit (Ping timeout: 268 seconds)
2014-09-06T11:01:41Z attila_lendvai joined #sbcl
2014-09-06T11:01:41Z attila_lendvai quit (Changing host)
2014-09-06T11:01:41Z attila_lendvai joined #sbcl
2014-09-06T11:14:40Z drmeiste_ joined #sbcl
2014-09-06T11:19:02Z drmeiste_ quit (Ping timeout: 245 seconds)
2014-09-06T13:15:26Z drmeiste_ joined #sbcl
2014-09-06T13:19:52Z drmeiste_ quit (Ping timeout: 245 seconds)
2014-09-06T13:32:38Z Shaftoe___ joined #sbcl
2014-09-06T13:33:22Z Shaftoe___: hi all. I'm running sbcl 1.2.2 on osx 10.9.2
2014-09-06T13:33:56Z Shaftoe___: I've run into a little problem with 'sb-posix:termios. The class definition seems to not match the OS definition
2014-09-06T13:34:21Z Shaftoe___: so sb-posix:tcsetattr with the current definition of termios gives me the following error:
2014-09-06T13:34:32Z Shaftoe___: "Error in SB-POSIX:TCSETATTR: Invalid argument (22)"
2014-09-06T13:35:02Z Shaftoe___: http://paste.lisp.org/display/143629 for rough sample code.
2014-09-06T13:45:13Z LiamH joined #sbcl
2014-09-06T13:46:46Z Shaftoe___: so the problem, as far as I can tell, is the definition of 'termios in  contrib/sb-posix/interface.lisp does not match that of struct termios from /usr/include/sys/termios.h
2014-09-06T13:46:55Z Shaftoe___: I shall await for word.
2014-09-06T14:14:39Z les quit (Remote host closed the connection)
2014-09-06T14:20:36Z les joined #sbcl
2014-09-06T14:21:41Z Shaftoe___: is there any way to just rebuild a particular contrib without rebuilding the entire sbcl source?
2014-09-06T14:38:36Z LiamH quit (Read error: Connection reset by peer)
2014-09-06T14:39:18Z LiamH joined #sbcl
2014-09-06T14:40:31Z yacks quit (Read error: Connection reset by peer)
2014-09-06T14:45:31Z kanru` joined #sbcl
2014-09-06T14:46:24Z eudoxia joined #sbcl
2014-09-06T14:47:09Z yacks joined #sbcl
2014-09-06T15:05:40Z wmarvel: you can probably tell asdf to rebuild it.
2014-09-06T15:06:08Z wmarvel: you probably have to link to the .asd in .sbcl/systems
2014-09-06T15:19:08Z Shaftoe___: well, it's a contrib, so it's already in sbcl/systems
2014-09-06T15:19:12Z Shaftoe___: it's sb-posix
2014-09-06T15:19:46Z wmarvel: actually, looking at the script you could try removing the fasls from obj/asdf-cache/<thecontribyouwanttobuild>/ and rerunning make-target-contribs.sh with DONT_CLEAN_SBCL_CONTRIB set in your environment
2014-09-06T15:20:11Z wmarvel: (shell environment)
2014-09-06T15:21:57Z Shaftoe___: ahh. thanks. Let me try
2014-09-06T15:23:08Z pkhuong: wmarvel: seems like a function that can be used within several VOPs would be easier and solve most of the issues, re TLS
2014-09-06T15:23:59Z pkhuong: the only reason I'd use a VOP is if I wanted really clever spilling/loading, and I feel that's unnecessary complexity for a first stab at threading
2014-09-06T15:24:03Z pkhuong: once it works, perhaps.
2014-09-06T15:25:15Z wmarvel: So there's a user rw register in cp15 that is defined as 'User Read/Write Thread and Process ID RegisterUser Read/Write Thread and Process ID Register'
2014-09-06T15:25:38Z wmarvel: and there's code in the linux kernel in switch_tls that looks like it's supposed to save/restore it on context switch
2014-09-06T15:26:03Z wmarvel: except at least on my box it gets set to zero by the kernel somewhere when I set it :D
2014-09-06T15:27:16Z wmarvel: so I was thinking, if I could grab the pointer from pthreads and store it there, it's almost like a register, one-instruction mcr/mrc cp15 to get it.
2014-09-06T15:27:55Z wmarvel: although I don't know if depending on that existing is really the way to go.
2014-09-06T15:34:24Z wmarvel: and my little asm/c pthreads test shows it not surviving through context switches anyway (which looks like a bug but I might be doing it wrong)
2014-09-06T15:43:37Z Shaftoe___: so anyone have any ideas on my termios predicament?
2014-09-06T15:44:00Z attila_lendvai quit (Quit: Leaving.)
2014-09-06T15:44:09Z Shaftoe___: alternatively, maybe I can ask for some advice on hacking sb-posix internals (specifically sb-grovel related stuff).
2014-09-06T15:46:14Z Shaftoe___: it would seem that :structure alien-termios as defined in contrib/sb-posix/constants.lisp is missing two fields at the end of the definition. Also, it would seem that the array defined in that struct is actually a fixed length array.
2014-09-06T15:47:18Z Shaftoe___: but the length of the array is defined in the header file so I don't know exactly how to do that
2014-09-06T15:47:55Z Shaftoe___: (i.e. looking into sb-bsd-sockets to see how array is used, I only see hard coded array sizes)
2014-09-06T15:48:23Z pkhuong: Shaftoe___: it's generated at compile-time
2014-09-06T15:48:59Z pkhuong: wmarvel: we depend on way worse on other platforms.
2014-09-06T15:49:08Z Shaftoe___: it is. however, constants.lisp seems to tell sb-grovel what to generate, right?
2014-09-06T15:49:38Z pkhuong: missing fields should be taken care of with padding
2014-09-06T15:49:57Z pkhuong: there may be an issue when trying to copy a termios struct to/from lisp
2014-09-06T15:50:19Z wmarvel: pkhuong: Ok. In that case I'll try to figure out if I'm doing it wrong and/or if I'm hitting a kernel bug.
2014-09-06T15:50:47Z pkhuong: if rebuilding that contrib (including grovelling) doesn't help, it's probably an old bug where we replace unknown fields with 0
2014-09-06T15:52:25Z Shaftoe___: pkhuong: http://paste.lisp.org/display/143632
2014-09-06T15:53:17Z Shaftoe___: is what I would like to do even possible? or does it have issues like padding/alignment etc?
2014-09-06T15:54:12Z Shaftoe___: pkhuong: that paste has an error in it. The "as it currently is" should not have the two speed-t fields after the array. Sorry for mistype
2014-09-06T15:55:44Z Shaftoe___: http://paste.lisp.org/display/143633 is what I meant
2014-09-06T16:34:22Z kanru` quit (Ping timeout: 240 seconds)
2014-09-06T16:37:59Z Shaftoe___: pkhuong: I've been scratching my head a lot over this. I'm not sure how to proceed. Can you give me any pointers?
2014-09-06T16:38:19Z Shaftoe___: I've attempted to simply hard code the structure to see if that would work but it doesn't.
2014-09-06T16:47:46Z drmeiste_ joined #sbcl
2014-09-06T17:23:10Z edgar-rft quit (Quit: conversation closed because existence stopped)
2014-09-06T17:23:53Z eudoxia quit (Quit: leaving)
2014-09-06T17:37:32Z edgar-rft joined #sbcl
2014-09-06T17:58:08Z wmarvel: oh, hell. The kernel on this thing reports a particular version, the distro kernel code is what I was looking at, the actual kernel is coming from a git tree that doesn't have the patch that put what I'm looking at in.
2014-09-06T18:02:08Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T18:20:19Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-06T18:24:12Z drmeiste_ joined #sbcl
2014-09-06T18:26:17Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T18:40:02Z oleo__ quit (Quit: Verlassend)
2014-09-06T18:40:19Z drmeiste_ joined #sbcl
2014-09-06T18:43:52Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T18:46:17Z drmeiste_ joined #sbcl
2014-09-06T18:54:18Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T18:57:15Z drmeiste_ joined #sbcl
2014-09-06T18:59:53Z oleo joined #sbcl
2014-09-06T19:05:14Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T19:05:51Z drmeiste_ joined #sbcl
2014-09-06T19:11:59Z drmeiste_ quit (Ping timeout: 252 seconds)
2014-09-06T19:12:33Z Krystof quit (Ping timeout: 252 seconds)
2014-09-06T19:14:45Z drmeiste_ joined #sbcl
2014-09-06T19:15:04Z kanru quit (Ping timeout: 252 seconds)
2014-09-06T19:16:47Z edgar-rft quit (*.net *.split)
2014-09-06T19:16:48Z White_Flame quit (*.net *.split)
2014-09-06T19:20:30Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T19:23:24Z attila_lendvai joined #sbcl
2014-09-06T19:23:24Z attila_lendvai quit (Changing host)
2014-09-06T19:23:24Z attila_lendvai joined #sbcl
2014-09-06T19:29:17Z edgar-rft joined #sbcl
2014-09-06T19:29:18Z White_Flame joined #sbcl
2014-09-06T19:50:40Z drmeiste_ joined #sbcl
2014-09-06T19:54:59Z sobel joined #sbcl
2014-09-06T19:58:38Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T20:01:16Z angavrilov quit (Remote host closed the connection)
2014-09-06T20:10:36Z attila_lendvai quit (Quit: Leaving.)
2014-09-06T20:59:40Z erikvarga left #sbcl
2014-09-06T21:09:20Z oleo is now known as Guest9870
2014-09-06T21:10:09Z oleo__ joined #sbcl
2014-09-06T21:12:23Z Guest9870 quit (Ping timeout: 268 seconds)
2014-09-06T21:12:52Z drmeiste_ joined #sbcl
2014-09-06T21:14:05Z edgar-rft quit (Quit: connection terminated into eternal nothing)
2014-09-06T21:27:24Z drmeiste_ quit (Remote host closed the connection)
2014-09-06T21:44:54Z Krystof joined #sbcl
2014-09-06T21:44:54Z ChanServ has set mode +o Krystof
2014-09-06T22:06:43Z oleo__ quit (Quit: Verlassend)
2014-09-06T22:09:04Z oleo joined #sbcl
2014-09-06T22:17:38Z drmeiste_ joined #sbcl
2014-09-07T00:35:42Z DGASAU quit (Remote host closed the connection)
2014-09-07T00:36:31Z DGASAU joined #sbcl
2014-09-07T01:42:11Z Shaftoe___ joined #sbcl
2014-09-07T01:45:35Z LiamH quit (Quit: Leaving.)
2014-09-07T01:49:10Z gingerale quit (Read error: Connection reset by peer)
2014-09-07T02:03:07Z drmeiste_ quit (Remote host closed the connection)
2014-09-07T02:09:47Z pkhuong: https://www.kernel.org/doc/Documentation/vm/soft-dirty.txt
2014-09-07T02:10:06Z pkhuong: linux grew the solaris thingy
2014-09-07T02:10:17Z pkhuong: without the scalability issue!
2014-09-07T02:14:14Z pkhuong: mprotect-less gengc. this could be awesome.
2014-09-07T02:16:34Z wmarvel: Nice.
2014-09-07T02:17:15Z wmarvel: In other news, I patched my kernel up on the box that actually has multiple cores, and am about to break my sbcl into tiny little pieces. ;)
2014-09-07T02:22:25Z Bicyclidine joined #sbcl
2014-09-07T02:25:53Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-07T02:34:47Z Bicyclidine quit (Ping timeout: 245 seconds)
2014-09-07T02:39:02Z christoph_debian quit (Ping timeout: 245 seconds)
2014-09-07T02:46:49Z drmeiste_ joined #sbcl
2014-09-07T02:50:30Z drmeist__ joined #sbcl
2014-09-07T02:51:33Z drmeiste_ quit (Ping timeout: 264 seconds)
2014-09-07T02:52:07Z christoph_debian joined #sbcl
2014-09-07T03:39:55Z Shaftoe___ joined #sbcl
2014-09-07T03:41:09Z milosn quit (Ping timeout: 240 seconds)
2014-09-07T03:42:20Z milosn joined #sbcl
2014-09-07T03:49:17Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-07T04:15:01Z drmeist__ quit (Remote host closed the connection)
2014-09-07T04:41:18Z Shaftoe___ joined #sbcl
2014-09-07T04:52:38Z Shaftoe___: pkhuong: you were mentioning padded structs for posix calls earlier. You have any examples of this?
2014-09-07T04:52:38Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-07T04:57:32Z Shaftoe___ joined #sbcl
2014-09-07T05:09:14Z pkhuong: Shaftoe___: the groveller should handle that part automatically
2014-09-07T05:10:10Z Shaftoe___: I'm just trying to find a mechanism to induce it. But maybe that's the wrong approach. Whatever the solution may be, my problem is that the structure being created by the groveler is too small for the actual sys call
2014-09-07T05:10:14Z pkhuong: like i said, the issue is likely that we leave unknown fields seroed out/uninitialised
2014-09-07T05:10:26Z Shaftoe___: I think zeroed out would be fine.
2014-09-07T05:10:41Z Shaftoe___: are those unknown fields not visible in inspect?
2014-09-07T05:24:28Z Shaftoe___: ok. Well, I've finally found what I was missing. it actually works now. only question is whether this "fix" works on other systems.
2014-09-07T05:25:03Z Shaftoe___: Is there anyone that would be willing to guide me as far as submitting a patch is concerned? I know I'd be unable to test on anything other than osx at the present moment.
2014-09-07T05:48:47Z drmeiste_ joined #sbcl
2014-09-07T06:22:48Z angavrilov joined #sbcl
2014-09-07T06:24:37Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-07T06:31:31Z Shaftoe___ joined #sbcl
2014-09-07T06:51:41Z gingerale joined #sbcl
2014-09-07T07:12:54Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-07T07:15:23Z erikvarga joined #sbcl
2014-09-07T07:17:55Z edgar-rft joined #sbcl
2014-09-07T07:36:22Z Shaftoe___ joined #sbcl
2014-09-07T07:48:42Z Shaftoe___ quit (Quit: Shaftoe___)
2014-09-07T07:58:19Z Shaftoe___ joined #sbcl
2014-09-07T08:00:20Z Shaftoe___ quit (Client Quit)
2014-09-07T08:08:34Z drmeiste_ quit (Quit: Leaving...)
2014-09-07T09:01:50Z Shaftoe___ joined #sbcl
2014-09-07T09:10:45Z Shaftoe___ quit (Ping timeout: 264 seconds)
2014-09-07T09:22:35Z Shaftoe___ joined #sbcl
2014-09-07T09:23:30Z Shaftoe____ joined #sbcl
2014-09-07T09:27:33Z Shaftoe___ quit (Ping timeout: 264 seconds)
2014-09-07T09:27:33Z Shaftoe____ is now known as Shaftoe___
2014-09-07T09:27:53Z leo2007 quit (Ping timeout: 276 seconds)
2014-09-07T09:28:07Z Shaftoe___ quit (Client Quit)
2014-09-07T09:43:45Z oleo__ joined #sbcl
2014-09-07T09:45:08Z oleo is now known as Guest26340
2014-09-07T09:46:21Z Guest26340 quit (Ping timeout: 255 seconds)
2014-09-07T09:52:18Z attila_lendvai joined #sbcl
2014-09-07T09:54:00Z milosn_ joined #sbcl
2014-09-07T09:54:20Z milosn_ quit (Client Quit)
2014-09-07T09:54:52Z milosn_ joined #sbcl
2014-09-07T09:55:00Z milosn_ quit (Client Quit)
2014-09-07T10:31:43Z leo2007 joined #sbcl
2014-09-07T11:07:47Z DGASAU quit (Ping timeout: 245 seconds)
2014-09-07T13:38:43Z fikusz_ quit (Remote host closed the connection)
2014-09-07T15:05:57Z leo2007 quit (Ping timeout: 264 seconds)
2014-09-07T15:28:01Z pkhuong: my mprotect less gencgc is compiling contribs :|
2014-09-07T16:37:07Z nyef joined #sbcl
2014-09-07T17:06:13Z pkhuong: and all tests are green
2014-09-07T17:06:31Z pkhuong: https://github.com/sbcl/sbcl/tree/gengc-soft-dirty if anyone wants to play with this.
2014-09-07T17:28:28Z davazp joined #sbcl
2014-09-07T17:45:23Z wmarvel: "In conclusion, there is no reliable way to extract the FP state from a ucontext_t on ARM"
2014-09-07T17:49:53Z pkhuong: sounds about par for FP on ARM
2014-09-07T17:51:14Z nyef: Sounds about par for signal contexts on most non-x86 systems.
2014-09-07T17:54:36Z jsnell_: oh, wow. linux exported the page table dirty bits?
2014-09-07T17:54:38Z jsnell_: that's really nice
2014-09-07T17:58:15Z pkhuong: jsnell_: yes. and they're separated clearing from reading the bits
2014-09-07T17:58:27Z pkhuong: so we can choose what address range we want to read
2014-09-07T18:16:27Z oleo__ quit (Read error: Connection reset by peer)
2014-09-07T18:18:48Z oleo joined #sbcl
2014-09-07T18:23:18Z oleo quit (Read error: Connection reset by peer)
2014-09-07T18:41:47Z davazp quit (Remote host closed the connection)
2014-09-07T18:47:12Z oleo joined #sbcl
2014-09-07T18:52:06Z oleo quit (Remote host closed the connection)
2014-09-07T18:54:29Z pkhuong: .. we should probably remove the #define for FUTEX_FD
2014-09-07T18:55:24Z oleo joined #sbcl
2014-09-07T19:18:57Z drmeiste_ joined #sbcl
2014-09-07T19:28:13Z prxq joined #sbcl
2014-09-07T19:30:55Z pkhuong: interesting. GC is much slower. I'll have to check. Hopefully it's not just the overhead of reading dirty bits :\
2014-09-07T19:33:07Z nyef: Even if the GC is slower, is the overall system faster or slower?
2014-09-07T19:36:58Z pkhuong: I can probably cook up a microbenchmark that points whichever way I want. Loading cl-ppcre is still slower.
2014-09-07T19:37:10Z pkhuong: I could look at compiling, but I want to understand where the slowdown comes from first
2014-09-07T19:46:33Z pkhuong: 3 ms to open the pagemap file and read it, .5ms to clear all dirty bits
2014-09-07T19:46:59Z pkhuong: total gc time: ~400 ms instead of ~100
2014-09-07T19:50:08Z nyef: Something weird there.
2014-09-07T19:50:54Z pkhuong: it's all in scavenge and collect garbage
2014-09-07T19:51:50Z pkhuong: upside: it's probably one of the first time someone's looked at our GC's instruction-level profile
2014-09-07T19:55:37Z pkhuong: ~40%of collect_garbage is in update_page_write_prot!
2014-09-07T19:56:31Z nyef: Which... if we're using the kernel dirty-bits tracking, should be a lot simpler?
2014-09-07T19:57:14Z pkhuong: it's also badly named
2014-09-07T19:57:27Z nyef: A *LOT* of things in the GC are badly named.
2014-09-07T19:57:37Z pkhuong: because it mostly scans for pointers to younger generations
2014-09-07T19:58:03Z nyef: ... The "remembered set"?
2014-09-07T19:58:39Z pkhuong: we're not that clever
2014-09-07T19:58:55Z nyef: But would that be what it is if we WERE?
2014-09-07T19:59:03Z pkhuong: yes (:
2014-09-07T19:59:17Z pkhuong: I guess we have page-level remembered sets :\
2014-09-07T19:59:28Z nyef: Yeah, ouch.
2014-09-07T20:00:30Z pkhuong: where a page is remembered because we leave its write protected bit off
2014-09-07T20:00:31Z nyef: At some point I really do want to spend some more time hammering on the GC, but right now I barely have the energy to hack on free-time Lisp stuff at all.
2014-09-07T20:02:20Z pkhuong: ah. I was too eagerly marking pages as unprotected.
2014-09-07T20:02:29Z pkhuong: so we'd end up spending all our time figuring if we should reprotect pages.
2014-09-07T20:03:06Z attila_lendvai quit (Read error: Connection reset by peer)
2014-09-07T20:03:09Z pkhuong: (there are two reasons we unprotect pages: 1) mprotect write barrier 2) they're in oldspace, so we're about to scribble all over them)
2014-09-07T20:04:16Z attila_lendvai joined #sbcl
2014-09-07T20:04:17Z attila_lendvai quit (Changing host)
2014-09-07T20:04:17Z attila_lendvai joined #sbcl
2014-09-07T20:04:43Z pkhuong: this hack + madvise instead of munmap/mmap to release memory means that our memory map finally looks like that of a normal C program.
2014-09-07T20:05:08Z pkhuong: might be worth revisiting fork for concurrent major GC
2014-09-07T20:05:45Z pkhuong: that's more like it
2014-09-07T20:06:45Z nyef: Hrm. Can we also get discontiguous and relocatable heap segments?
2014-09-07T20:09:05Z pkhuong: I only have a plan up to and exlucing saving a core for discontiguous heap
2014-09-07T20:10:00Z pkhuong: I have a pretty good idea how to run actual GCs. saving and loading a discontiguous core... I don't see how to make that work nicely.
2014-09-07T20:10:37Z nyef: Relocate pages on load if they don't get their preferred locations?
2014-09-07T20:10:58Z nyef: The relocatable-heap stuff should really have been merged years ago. /-:
2014-09-07T20:11:07Z nyef: It's probably all bit-rotted by now.
2014-09-07T20:12:11Z pkhuong: right, so if we had relocation, discontiguous core would be easier
2014-09-07T20:12:45Z pkhuong: as it is, it would probably be useful for regular-sized cores that need huge heaps (e.g., for data analysis)
2014-09-07T20:14:47Z oleo quit (Ping timeout: 245 seconds)
2014-09-07T20:16:54Z pkhuong: this is insane. 30% is in scavenge itself.
2014-09-07T20:18:46Z pkhuong: more like 50%+ of GC is in scavenge.
2014-09-07T20:19:06Z nyef: Scavenge alone, or scavenge and the functions it calls?
2014-09-07T20:19:16Z pkhuong: scavenge alone
2014-09-07T20:20:36Z nyef: There's plausibly a good bit of inlined code here.
2014-09-07T20:22:42Z pkhuong: yeah, but none of the type-specific scavenging functions we'd expect to use time
2014-09-07T20:23:00Z pkhuong: we spend most of our time discovering that there's nothing to scavenge
2014-09-07T20:23:56Z nyef: Have a look at space_patches_p() in gencgc_internal.h. It's called by scavenge on every pointerish thing it sees.
2014-09-07T20:24:50Z nyef: Err... space_matches_p().
2014-09-07T20:25:06Z oleo joined #sbcl
2014-09-07T20:26:22Z pkhuong: it also doesn't help that our asserts are compiled with the error condition in the middle of our inner loop
2014-09-07T20:28:42Z nyef: We scavenge() oldspace pages, don't we?
2014-09-07T20:29:14Z nyef: Not oldspace... older-generation pages.
2014-09-07T20:31:10Z pkhuong: yes, when they might point to young data
2014-09-07T20:32:03Z nyef: So, the remembered sets should deal with that much, once they're implemented.
2014-09-07T20:33:54Z pkhuong: or just switching to a two-generation scheme
2014-09-07T20:34:09Z nyef: I'm not sure if it makes sense to check for forwarders before oldspace or not.
2014-09-07T20:36:27Z pkhuong: I bet there's some locality to pointers too
2014-09-07T20:36:47Z nyef: There's supposedly locality to object lifetimes, at least.
2014-09-07T20:37:30Z pkhuong: pointers that are next to the other likely point in the same generation
2014-09-07T20:37:52Z pkhuong: (i.e., not in old space ;)
2014-09-07T20:38:24Z nyef: ... Too easy to come up with plausible counter-examples, that's something to instrument for.
2014-09-07T20:42:35Z gingerale quit (Ping timeout: 276 seconds)
2014-09-07T20:42:38Z pkhuong: the slow part seems to be the random accesses to whatever we're pointing to
2014-09-07T20:43:54Z pkhuong: could try prefetching
2014-09-07T20:58:52Z Skrylar: reading this backlog i begin to wonder how insane the insides of sbcl look right now o_O
2014-09-07T21:00:16Z erikvarga left #sbcl
2014-09-07T21:02:59Z nyef: Too bad we can't encode a simple space-id as part of the pointer, and use that as a cheaper oldspace detector or something.
2014-09-07T21:03:53Z nyef: Hrm, no... The primary use-case for scavenge is actually newspace pages, isn't it?
2014-09-07T21:04:22Z pkhuong: yeah.
2014-09-07T21:04:53Z pkhuong: the hot inner loop is when we're trying to fixup pointer to from_space
2014-09-07T21:04:54Z nyef: Older generations should have the remembered sets (in theory), and stack uses separate scavenge logic...
2014-09-07T21:05:04Z pkhuong: we already know that they're in from_space
2014-09-07T21:05:22Z pkhuong: so whatever happens, we'll have to look at the pointee
2014-09-07T21:05:27Z pkhuong: and that relaly hurts
2014-09-07T21:06:30Z nyef: Wait, we already know that it's a fromspace pointer?
2014-09-07T21:06:55Z pkhuong: yeah
2014-09-07T21:06:59Z angavrilov quit (Remote host closed the connection)
2014-09-07T21:07:00Z pkhuong: that's not too bad
2014-09-07T21:07:08Z nyef: And we're checking for a forwarder or arranging transport?
2014-09-07T21:07:20Z pkhuong: yup
2014-09-07T21:08:16Z pkhuong:     6.55 :        40fc19:       mov    %rsi,%rdx
2014-09-07T21:08:16Z pkhuong:     0.00 :        40fc1c:       and    $0xfffffffffffffff0,%rdx
2014-09-07T21:08:16Z pkhuong:     0.87 :        40fc20:       cmpq   $0x1,(%rdx)
2014-09-07T21:08:16Z pkhuong:    24.45 :        40fc24:       jne    40fc58 <scavenge+0xe8>
2014-09-07T21:08:39Z nyef: forwarding_pointer_p(), I presume?
2014-09-07T21:08:45Z pkhuong: that's turning our lisp pointer into a native pointer and following that pointer to see if it's a broken heart
2014-09-07T21:09:30Z nyef: Can we blow a widetag on a forwarding_pointer_widetag?
2014-09-07T21:09:53Z pkhuong: how would that work?
2014-09-07T21:10:52Z nyef: Hrm. Not sure, actually. I've paged too much of this stuff out already. /-:
2014-09-07T21:11:39Z pkhuong: btw, does anyone else find object_ptr += (scavtab[widetag_of(object)])(object_ptr, object); really confusing when the object is a pointer?
2014-09-07T21:11:45Z pkhuong: we should just increment object_ptr
2014-09-07T21:12:47Z pkhuong: the scavtab is setup so that these cases always return 1
2014-09-07T21:13:03Z pkhuong: it might also be a good idea to have a separate table for lowtag dispatch, since it's the common case.
2014-09-07T21:16:18Z Perlboy quit (Ping timeout: 250 seconds)
2014-09-07T21:16:22Z pkhuong: cheap hack: thread this up with a sequence lock around writes to the heap/allocations.
2014-09-07T21:16:39Z nyef: Would having a more-compact indication that an object has already been transported help?
2014-09-07T21:16:53Z yauz_2 quit (Ping timeout: 240 seconds)
2014-09-07T21:16:57Z nyef: Say, one bit per allocation-unit in from-space?
2014-09-07T21:17:10Z pkhuong: if it were out of band, perhaps
2014-09-07T21:17:17Z nyef: Yeah, that's what I'm thinking.
2014-09-07T21:17:48Z nyef: When you mark a page as being from-space, you allocate a bitmap for the evicted objects, and use that to mark your broken hearts.
2014-09-07T21:18:07Z pkhuong: how do you know how to fix up your pointer?
2014-09-07T21:18:53Z nyef: All from-space pointers need fixing up, don't they? The question is if the object has been evicted yet or not.
2014-09-07T21:19:01Z nyef: Hrm.
2014-09-07T21:19:18Z nyef: Damnit, that doesn't actually gain anything unless you already know where it's being evicted TO. /-:
2014-09-07T21:19:21Z pkhuong: right
2014-09-07T21:19:47Z Perlboy joined #sbcl
2014-09-07T21:19:50Z Perlboy is now known as Guest45219
2014-09-07T21:20:00Z nyef: Which gets right back to markbits processing.
2014-09-07T21:20:34Z pkhuong: now, if only the nursery was ever evicted (:
2014-09-07T21:20:46Z nyef: Heh.
2014-09-07T21:20:55Z drmeiste_ is now known as drmeister_
2014-09-07T21:20:57Z nyef: That's crazy talk. (-:
2014-09-07T21:22:42Z nyef: What if we don't evict objects from a page until it's mostly-dead anyway?
2014-09-07T21:22:59Z nyef: (As determined by markbits.)
2014-09-07T21:24:02Z nyef: Possibly blank dead objects in a page with a freespace widetag or something, but leave the live objects as-is until they only comprise, say, a third of the page or so.
2014-09-07T21:24:33Z pkhuong: that would be pretty cool
2014-09-07T21:24:48Z nyef: It'd require a separate mark pass, though.
2014-09-07T21:24:52Z yauz joined #sbcl
2014-09-07T21:25:59Z pkhuong: we could do some lossy refcounting-y stuff that scavenges + marks in a single pass
2014-09-07T21:26:03Z BitPuffin joined #sbcl
2014-09-07T21:28:36Z nyef: How about a two-phase system? We evict objects from condemned pages *first*, then we clear markbits and scavenge?
2014-09-07T21:29:19Z nyef: And at the start of the next GC, we evacuate the pages that are newly condemned based on the markbits from the previous GC.
2014-09-07T21:30:08Z pkhuong: yeah.
2014-09-07T21:30:09Z yauz quit (Ping timeout: 255 seconds)
2014-09-07T21:30:25Z BitPuffin: #_#
2014-09-07T21:30:30Z pkhuong: what's the point of evicting eagerly?
2014-09-07T21:31:10Z yauz joined #sbcl
2014-09-07T21:31:17Z pkhuong: might as well do it on demand (like cheneygc currently does) and leave dead objects where they are?
2014-09-07T21:32:02Z pkhuong: that would also unify how we handle mostly-empty and nursery pages (and pinned pages).
2014-09-07T21:34:12Z pkhuong: n.b., you still need mark bits, because that's how we approximate live objects in pages that aren't being condemned
2014-09-07T21:38:15Z nyef: Right, and also how we can compute the address of an object in a page that the evicted objects are being transported to (logcount of the markbits prior to the object being transported, times the allocation granularity).
2014-09-07T21:40:18Z nyef: Oh! And this also means that pinned objects no longer hold the entire page worth of objects live, it merely prevents condemning the page.
2014-09-07T21:40:40Z pkhuong: yes (:
2014-09-07T21:40:56Z pkhuong: I'm just not sure how to get mark bits for the whole heap :\
2014-09-07T21:41:43Z nyef: Why would we need them if we have the remembered sets?
2014-09-07T21:42:48Z pkhuong: to determine which objects must be blanked out
2014-09-07T21:43:35Z nyef: But we're only doing that for the current generation's "oldspace", surely?
2014-09-07T21:44:19Z nyef: ... I'm just about out of time for tonight, unfortunately.
2014-09-07T21:44:54Z Guest45219 quit (Ping timeout: 250 seconds)
2014-09-07T21:44:55Z pkhuong: ah yeah. implicit assumption that we won't need 7 generations if we stop copying all the things when we GC
2014-09-07T21:47:02Z pkhuong: I might try and parallelise GC soonish. Our inner loop is stupid.
2014-09-07T21:47:10Z francogrex joined #sbcl
2014-09-07T21:47:25Z francogrex left #sbcl
2014-09-07T21:47:27Z yauz quit (Ping timeout: 268 seconds)
2014-09-07T21:56:11Z yauz joined #sbcl
2014-09-07T22:00:21Z Perlboy_ joined #sbcl
2014-09-07T22:02:28Z pkhuong: great. benign race condition in the way we scrub the control stack
2014-09-07T22:02:42Z pkhuong: we first mark ourselves as stopped, and then scrub
2014-09-07T22:02:57Z pkhuong: but GC could start reading our control stack before we get around to scrubbing it.
2014-09-07T22:03:38Z pkhuong: (I also don't see where safepoint builds scrub the control stack)
2014-09-07T22:09:02Z yauz quit (Ping timeout: 268 seconds)
2014-09-07T22:10:59Z krzysz00 joined #sbcl
2014-09-07T22:12:29Z oleo is now known as Guest74721
2014-09-07T22:15:42Z Guest74721 quit (Ping timeout: 245 seconds)
2014-09-07T22:23:16Z attila_lendvai quit (Quit: Leaving.)
2014-09-07T22:24:18Z oleo__ joined #sbcl
2014-09-07T22:24:38Z oleo__ is now known as oleo
2014-09-07T22:25:07Z drmeister_ quit (Remote host closed the connection)
2014-09-07T22:29:08Z leo2007 joined #sbcl
2014-09-07T22:32:34Z brucem quit (Ping timeout: 250 seconds)
2014-09-07T22:34:54Z Perlboy_ is now known as Perlboy
2014-09-07T22:36:46Z yauz joined #sbcl
2014-09-07T22:46:33Z nyef quit (Quit: G'night all)
2014-09-07T22:52:21Z wmarvel: So - we have a thread struct in non-thread builds - and in arm we've got static *control-stack-pointer*, etc
2014-09-07T22:52:49Z wmarvel: I am pondering whether it makes sense to just promote those to the thread struct in the ordinary non-thread build
2014-09-07T22:54:27Z wmarvel: in arm-assem.S the address of the control stack and frame pointers is being stashed into words .frame-pointer-address and .stack-pointer-address
2014-09-07T22:54:58Z pkhuong: it would
2014-09-07T22:58:27Z wmarvel: ok
2014-09-07T22:58:33Z wmarvel: I'll take a stab at that then
2014-09-07T23:13:14Z eudoxia joined #sbcl
2014-09-07T23:30:09Z brucem joined #sbcl
2014-09-07T23:31:09Z brucem quit (Changing host)
2014-09-07T23:31:09Z brucem joined #sbcl
2014-09-08T00:15:55Z eudoxia quit (Quit: Lost terminal)
2014-09-08T00:16:39Z edgar-rft quit (Quit: continuation closed by death)
2014-09-08T00:43:21Z drmeiste_ joined #sbcl
2014-09-08T01:10:18Z pkhuong: bah. soft dirty bits just don't work with huge pages
2014-09-08T01:10:22Z pkhuong: always dirty
2014-09-08T01:14:43Z prxq_ joined #sbcl
2014-09-08T01:18:50Z prxq quit (Ping timeout: 276 seconds)
2014-09-08T01:19:37Z BitPuffin quit (Ping timeout: 260 seconds)
2014-09-08T02:12:02Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T02:38:38Z christoph_debian quit (Ping timeout: 252 seconds)
2014-09-08T02:52:43Z christoph_debian joined #sbcl
2014-09-08T03:23:53Z weissschloss quit (Remote host closed the connection)
2014-09-08T04:34:07Z BitPuffin joined #sbcl
2014-09-08T04:56:08Z oleo quit (Quit: Verlassend)
2014-09-08T04:58:45Z BitPuffin quit (Ping timeout: 264 seconds)
2014-09-08T05:00:58Z slyrus quit (Ping timeout: 268 seconds)
2014-09-08T05:05:02Z slyrus joined #sbcl
2014-09-08T05:20:42Z pranavrc joined #sbcl
2014-09-08T05:24:26Z gingerale joined #sbcl
2014-09-08T05:28:15Z drmeiste_ joined #sbcl
2014-09-08T05:50:45Z les quit (Quit: leaving)
2014-09-08T05:51:34Z les joined #sbcl
2014-09-08T05:51:34Z les quit (Changing host)
2014-09-08T05:51:34Z les joined #sbcl
2014-09-08T05:52:25Z pranavrc_ joined #sbcl
2014-09-08T05:54:04Z pranavrc quit (Ping timeout: 252 seconds)
2014-09-08T06:53:53Z BitPuffin joined #sbcl
2014-09-08T07:19:01Z yacks quit (Quit: Leaving)
2014-09-08T07:20:44Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-08T07:24:01Z drmeiste_ joined #sbcl
2014-09-08T07:31:06Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-08T07:32:33Z yacks joined #sbcl
2014-09-08T07:54:05Z angavrilov joined #sbcl
2014-09-08T08:20:37Z edgar-rft joined #sbcl
2014-09-08T08:37:48Z attila_lendvai joined #sbcl
2014-09-08T09:34:31Z DGASAU joined #sbcl
2014-09-08T09:44:37Z attila_lendvai quit (Quit: Leaving.)
2014-09-08T09:50:03Z fe[nl]ix quit (Remote host closed the connection)
2014-09-08T09:50:03Z Blkt_ quit (Remote host closed the connection)
2014-09-08T09:50:11Z Blkt joined #sbcl
2014-09-08T09:50:13Z fe[nl]ix joined #sbcl
2014-09-08T10:26:43Z Puffin joined #sbcl
2014-09-08T10:30:38Z zacts joined #sbcl
2014-09-08T11:55:34Z Puffin quit (Remote host closed the connection)
2014-09-08T12:33:47Z pranavrc_ quit (Remote host closed the connection)
2014-09-08T12:34:04Z pranavrc joined #sbcl
2014-09-08T12:34:04Z pranavrc quit (Changing host)
2014-09-08T12:34:04Z pranavrc joined #sbcl
2014-09-08T12:52:53Z psilord quit (Quit: Leaving.)
2014-09-08T12:58:49Z eudoxia joined #sbcl
2014-09-08T13:05:43Z eudoxia_ joined #sbcl
2014-09-08T13:05:43Z eudoxia quit (Read error: Connection reset by peer)
2014-09-08T13:09:53Z eudoxia_ quit (Client Quit)
2014-09-08T13:12:39Z eudoxia joined #sbcl
2014-09-08T13:24:22Z gale joined #sbcl
2014-09-08T13:26:32Z gingerale quit (Ping timeout: 252 seconds)
2014-09-08T13:26:32Z gale is now known as gingerale
2014-09-08T13:29:17Z oleo joined #sbcl
2014-09-08T13:42:22Z pranavrc quit
2014-09-08T13:42:51Z drmeiste_ joined #sbcl
2014-09-08T13:47:54Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T14:00:39Z eudoxia quit (Quit: Lost terminal)
2014-09-08T14:23:40Z eudoxia joined #sbcl
2014-09-08T14:47:12Z drmeiste_ joined #sbcl
2014-09-08T14:49:56Z luis quit (Quit: ZNC - http://znc.sourceforge.net)
2014-09-08T14:56:11Z luis joined #sbcl
2014-09-08T14:58:26Z momo-reina joined #sbcl
2014-09-08T14:58:51Z momo-reina quit (Remote host closed the connection)
2014-09-08T15:01:24Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-08T15:13:20Z nyef joined #sbcl
2014-09-08T15:19:56Z psilord joined #sbcl
2014-09-08T15:30:39Z krzysz00 quit (Quit: leaving)
2014-09-08T15:46:27Z brown joined #sbcl
2014-09-08T15:46:51Z brown is now known as Guest22974
2014-09-08T16:12:04Z BitPuffin quit (Ping timeout: 272 seconds)
2014-09-08T16:23:39Z drmeiste_ joined #sbcl
2014-09-08T16:59:21Z Guest22974 quit (Remote host closed the connection)
2014-09-08T17:09:57Z brown joined #sbcl
2014-09-08T17:10:21Z brown is now known as Guest91470
2014-09-08T17:10:25Z Guest91470 quit (Remote host closed the connection)
2014-09-08T17:11:17Z reb joined #sbcl
2014-09-08T17:12:06Z reb quit (Remote host closed the connection)
2014-09-08T17:14:03Z reb joined #sbcl
2014-09-08T17:23:38Z slyrus quit (Ping timeout: 272 seconds)
2014-09-08T17:32:27Z eudoxia quit (Quit: Lost terminal)
2014-09-08T17:48:05Z BitPuffin joined #sbcl
2014-09-08T17:59:52Z drmeiste_ quit (Ping timeout: 245 seconds)
2014-09-08T18:05:50Z drmeiste_ joined #sbcl
2014-09-08T18:12:27Z reb quit (Remote host closed the connection)
2014-09-08T18:20:37Z reb` joined #sbcl
2014-09-08T18:43:01Z slyrus joined #sbcl
2014-09-08T18:45:46Z slyrus_ joined #sbcl
2014-09-08T18:48:13Z slyrus quit (Ping timeout: 260 seconds)
2014-09-08T18:48:26Z slyrus_ is now known as slyrus
2014-09-08T18:50:49Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T18:55:07Z drmeiste_ joined #sbcl
2014-09-08T18:58:43Z slyrus_ joined #sbcl
2014-09-08T18:59:22Z drmeiste_ quit (Ping timeout: 240 seconds)
2014-09-08T19:00:22Z slyrus quit (Ping timeout: 240 seconds)
2014-09-08T19:00:37Z slyrus_ is now known as slyrus
2014-09-08T19:02:16Z attila_lendvai joined #sbcl
2014-09-08T19:02:16Z attila_lendvai quit (Changing host)
2014-09-08T19:02:16Z attila_lendvai joined #sbcl
2014-09-08T19:13:30Z drmeiste_ joined #sbcl
2014-09-08T19:30:41Z slyrus_ joined #sbcl
2014-09-08T19:32:22Z slyrus quit (Ping timeout: 245 seconds)
2014-09-08T19:33:08Z slyrus joined #sbcl
2014-09-08T19:36:52Z slyrus_ quit (Ping timeout: 252 seconds)
2014-09-08T19:40:42Z gingerale quit (Ping timeout: 245 seconds)
2014-09-08T19:46:24Z Puffin joined #sbcl
2014-09-08T19:47:10Z Bicyclidine joined #sbcl
2014-09-08T19:47:55Z slyrus_ joined #sbcl
2014-09-08T19:49:22Z slyrus quit (Ping timeout: 245 seconds)
2014-09-08T19:49:23Z slyrus_ is now known as slyrus
2014-09-08T19:51:26Z eudoxia joined #sbcl
2014-09-08T19:51:43Z eudoxia quit (Client Quit)
2014-09-08T19:51:58Z BitPuffin quit (Ping timeout: 252 seconds)
2014-09-08T19:53:47Z slyrus_ joined #sbcl
2014-09-08T19:55:45Z slyrus quit (Ping timeout: 264 seconds)
2014-09-08T19:55:47Z slyrus_ is now known as slyrus
2014-09-08T20:02:32Z _8hzp joined #sbcl
2014-09-08T20:05:38Z angavrilov quit (Remote host closed the connection)
2014-09-08T20:06:07Z _8hzp` quit (Ping timeout: 245 seconds)
2014-09-08T20:13:52Z arrdem quit (Ping timeout: 250 seconds)
2014-09-08T20:14:47Z arrdem joined #sbcl
2014-09-08T20:26:05Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T20:27:54Z slyrus_ joined #sbcl
2014-09-08T20:29:33Z slyrus quit (Ping timeout: 246 seconds)
2014-09-08T20:29:36Z slyrus_ is now known as slyrus
2014-09-08T20:48:32Z faheem_ quit (Ping timeout: 250 seconds)
2014-09-08T20:49:23Z faheem_ joined #sbcl
2014-09-08T20:56:29Z drmeiste_ joined #sbcl
2014-09-08T21:02:48Z Bicyclidine quit (Ping timeout: 246 seconds)
2014-09-08T21:11:46Z Bicyclidine joined #sbcl
2014-09-08T21:12:54Z nyef quit (Quit: G'night all.)
2014-09-08T21:14:03Z slyrus_ joined #sbcl
2014-09-08T21:15:41Z slyrus quit (Ping timeout: 260 seconds)
2014-09-08T21:15:57Z slyrus_ is now known as slyrus
2014-09-08T21:24:54Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T21:32:59Z drmeiste_ joined #sbcl
2014-09-08T21:35:19Z slyrus_ joined #sbcl
2014-09-08T21:37:17Z slyrus__ joined #sbcl
2014-09-08T21:37:30Z slyrus quit (Ping timeout: 252 seconds)
2014-09-08T21:39:29Z slyrus__ is now known as slyrus
2014-09-08T21:40:22Z Bicyclidine quit (Ping timeout: 240 seconds)
2014-09-08T21:41:09Z slyrus_ quit (Ping timeout: 252 seconds)
2014-09-08T21:41:18Z _8hzp` joined #sbcl
2014-09-08T21:45:23Z _8hzp quit (Ping timeout: 276 seconds)
2014-09-08T21:52:47Z psilord quit (Quit: Leaving.)
2014-09-08T21:58:51Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T22:00:35Z drmeiste_ joined #sbcl
2014-09-08T22:05:18Z attila_lendvai quit (Quit: Leaving.)
2014-09-08T22:10:45Z drmeiste_ quit (Remote host closed the connection)
2014-09-08T22:15:44Z slyrus_ joined #sbcl
2014-09-08T22:18:34Z slyrus quit (Ping timeout: 252 seconds)
2014-09-08T22:18:38Z slyrus_ is now known as slyrus
2014-09-08T23:15:36Z Bicyclidine joined #sbcl
2014-09-08T23:21:13Z psilord joined #sbcl
2014-09-09T00:00:49Z slyrus_ joined #sbcl
2014-09-09T00:01:32Z edgar-rft quit (Quit: no meaningful message available)
2014-09-09T00:02:42Z slyrus quit (Ping timeout: 255 seconds)
2014-09-09T00:02:46Z slyrus_ is now known as slyrus
2014-09-09T00:04:57Z Bicyclidine quit (Ping timeout: 255 seconds)
2014-09-09T00:09:50Z oleo is now known as Guest19128
2014-09-09T00:10:35Z oleo__ joined #sbcl
2014-09-09T00:13:35Z Guest19128 quit (Ping timeout: 276 seconds)
2014-09-09T00:25:12Z momo-reina joined #sbcl
2014-09-09T00:27:05Z Bicyclidine joined #sbcl
2014-09-09T00:45:00Z slyrus quit (Ping timeout: 255 seconds)
2014-09-09T00:50:46Z drmeiste_ joined #sbcl
2014-09-09T01:05:32Z drmeiste_ quit (Remote host closed the connection)
2014-09-09T01:09:56Z drmeiste_ joined #sbcl
2014-09-09T01:15:16Z prxq joined #sbcl
2014-09-09T01:18:32Z prxq_ quit (Ping timeout: 245 seconds)
2014-09-09T01:19:45Z momo-reina quit (Remote host closed the connection)
2014-09-09T01:30:12Z Bicyclidine quit (Ping timeout: 246 seconds)
2014-09-09T01:43:57Z Bicyclidine joined #sbcl
2014-09-09T01:49:15Z Blkt quit (Read error: Connection reset by peer)
2014-09-09T01:49:21Z Blkt joined #sbcl
2014-09-09T01:50:04Z lacedaemon joined #sbcl
2014-09-09T01:50:17Z fe[nl]ix quit (Read error: Connection reset by peer)
2014-09-09T01:55:20Z alchemis7 quit (Quit: @)
2014-09-09T01:57:49Z alchemis7 joined #sbcl
2014-09-09T02:04:22Z Bicyclidine quit (Ping timeout: 245 seconds)
2014-09-09T02:29:42Z Puffin quit (Ping timeout: 246 seconds)
2014-09-09T02:39:33Z christoph_debian quit (Ping timeout: 264 seconds)
2014-09-09T02:43:02Z drmeiste_ quit (Remote host closed the connection)
2014-09-09T02:43:33Z jackdaniel quit (Ping timeout: 240 seconds)
2014-09-09T02:44:30Z jackdaniel joined #sbcl
2014-09-09T02:46:07Z drmeiste_ joined #sbcl
2014-09-09T02:53:34Z christoph_debian joined #sbcl
2014-09-09T03:17:54Z krzysz00 joined #sbcl
2014-09-09T03:35:07Z brucem_ joined #sbcl
2014-09-09T03:35:11Z scymtym__ joined #sbcl
2014-09-09T03:36:36Z p_l_ joined #sbcl
2014-09-09T03:36:49Z brucem quit (Disconnected by services)
2014-09-09T03:36:55Z brucem_ is now known as brucem
2014-09-09T03:37:48Z oleo joined #sbcl
2014-09-09T03:38:44Z drmeiste_ quit (*.net *.split)
2014-09-09T03:38:44Z oleo__ quit (*.net *.split)
2014-09-09T03:38:44Z reb` quit (*.net *.split)
2014-09-09T03:38:44Z yauz quit (*.net *.split)
2014-09-09T03:38:45Z Krystof quit (*.net *.split)
2014-09-09T03:38:45Z White_Flame quit (*.net *.split)
2014-09-09T03:40:33Z krzysz00 quit (*.net *.split)
2014-09-09T03:40:36Z irsol quit (*.net *.split)
2014-09-09T03:40:36Z p_l quit (*.net *.split)
2014-09-09T03:40:37Z ams quit (*.net *.split)
2014-09-09T03:40:37Z scymtym_ quit (*.net *.split)
2014-09-09T03:42:23Z kanru joined #sbcl
2014-09-09T03:42:42Z krzysz00 joined #sbcl
2014-09-09T03:42:42Z drmeiste_ joined #sbcl
2014-09-09T03:42:42Z reb` joined #sbcl
2014-09-09T03:42:42Z yauz joined #sbcl
2014-09-09T03:42:42Z Krystof joined #sbcl
2014-09-09T03:42:42Z White_Flame joined #sbcl
2014-09-09T03:42:42Z cameron.freenode.net has set mode +o Krystof
2014-09-09T03:45:26Z gingerale joined #sbcl
2014-09-09T03:45:51Z irsol joined #sbcl
2014-09-09T03:46:26Z irsol quit (Max SendQ exceeded)
2014-09-09T03:48:01Z irsol joined #sbcl
2014-09-09T04:17:40Z slyrus joined #sbcl
2014-09-09T04:18:22Z Blkt quit (Quit: No Ping reply in 180 seconds.)
2014-09-09T04:18:31Z Blkt joined #sbcl
2014-09-09T04:30:05Z Blkt quit (Read error: Connection reset by peer)
2014-09-09T04:30:11Z Blkt joined #sbcl
2014-09-09T04:31:05Z lacedaemon quit (Ping timeout: 260 seconds)
2014-09-09T04:31:24Z fe[nl]ix joined #sbcl
2014-09-09T04:34:58Z fe[nl]ix quit (Read error: Connection reset by peer)
2014-09-09T04:35:14Z fe[nl]ix joined #sbcl
2014-09-09T04:35:16Z Blkt quit (Read error: Connection reset by peer)
2014-09-09T04:35:22Z Blkt joined #sbcl
2014-09-09T04:49:48Z gingerale quit (Ping timeout: 252 seconds)
2014-09-09T05:27:29Z pranavrc joined #sbcl
2014-09-09T05:27:29Z pranavrc quit (Changing host)
2014-09-09T05:27:29Z pranavrc joined #sbcl
2014-09-09T05:28:18Z oleo quit (Quit: Verlassend)
2014-09-09T08:00:38Z momo-reina joined #sbcl
2014-09-09T08:01:58Z attila_lendvai joined #sbcl
2014-09-09T08:01:58Z attila_lendvai quit (Changing host)
2014-09-09T08:01:58Z attila_lendvai joined #sbcl
2014-09-09T08:15:10Z ams joined #sbcl
2014-09-09T08:29:01Z yacks quit (Remote host closed the connection)
2014-09-09T08:44:56Z pranavrc_ joined #sbcl
2014-09-09T08:47:02Z pranavrc quit (Ping timeout: 252 seconds)
2014-09-09T08:47:41Z Krystof quit (Read error: Connection reset by peer)
2014-09-09T08:47:44Z Perlboy quit (Ping timeout: 276 seconds)
2014-09-09T08:53:40Z Perlboy joined #sbcl
2014-09-09T08:53:51Z Krystof joined #sbcl
2014-09-09T08:53:51Z ChanServ has set mode +o Krystof
2014-09-09T08:54:04Z Perlboy is now known as Guest54788
2014-09-09T08:54:34Z angavrilov joined #sbcl
2014-09-09T09:00:45Z Guest54788 quit (Changing host)
2014-09-09T09:00:46Z Guest54788 joined #sbcl
2014-09-09T09:01:04Z Guest54788 is now known as Perlman
2014-09-09T09:01:07Z Perlman is now known as Perlboy
2014-09-09T09:05:24Z BitPuffin joined #sbcl
2014-09-09T09:35:49Z leo2007 quit (Ping timeout: 260 seconds)
2014-09-09T09:37:49Z yacks joined #sbcl
2014-09-09T09:37:51Z edgar-rft joined #sbcl
2014-09-09T09:45:36Z leo2007 joined #sbcl
2014-09-09T09:55:17Z momo-reina quit (Ping timeout: 245 seconds)
2014-09-09T10:03:47Z p_l_ is now known as p_l
2014-09-09T10:27:42Z BitPuffin quit (Ping timeout: 245 seconds)
2014-09-09T11:11:01Z White__Flame joined #sbcl
2014-09-09T11:11:08Z White_Flame quit (Ping timeout: 252 seconds)
2014-09-09T11:25:14Z BitPuffin joined #sbcl
2014-09-09T11:58:26Z yauz quit (Ping timeout: 252 seconds)
2014-09-09T11:58:33Z yauz joined #sbcl
2014-09-09T12:01:15Z reb` quit (Ping timeout: 252 seconds)
2014-09-09T12:01:23Z krzysz00 quit (*.net *.split)
2014-09-09T12:01:23Z drmeiste_ quit (*.net *.split)
2014-09-09T12:05:55Z krzysz00 joined #sbcl
2014-09-09T12:05:55Z drmeiste_ joined #sbcl
2014-09-09T12:17:56Z drmeiste_ quit (Remote host closed the connection)
2014-09-09T12:27:55Z Krystof quit (Write error: Broken pipe)
2014-09-09T12:34:56Z drmeiste_ joined #sbcl
2014-09-09T12:35:10Z drmeiste_ quit (Remote host closed the connection)
2014-09-09T12:45:41Z Krystof joined #sbcl
2014-09-09T12:45:42Z ChanServ has set mode +o Krystof
2014-09-09T12:45:50Z drmeiste_ joined #sbcl
2014-09-09T12:51:25Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-09T12:51:59Z eudoxia joined #sbcl
2014-09-09T12:53:26Z drmeiste_ joined #sbcl
2014-09-09T13:08:43Z psilord quit (Quit: Leaving.)
2014-09-09T13:15:23Z pranavrc_ quit
2014-09-09T13:16:59Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-09T13:23:09Z attila_lendvai quit (Quit: Leaving.)
2014-09-09T13:28:41Z oleo joined #sbcl
2014-09-09T13:41:32Z drmeiste_ joined #sbcl
2014-09-09T13:44:30Z Puffin joined #sbcl
2014-09-09T13:48:00Z BitPuffin quit (Ping timeout: 246 seconds)
2014-09-09T13:59:43Z psilord joined #sbcl
2014-09-09T14:18:14Z reb` joined #sbcl
2014-09-09T14:19:36Z krzysz00 quit (Ping timeout: 252 seconds)
2014-09-09T14:22:16Z nicdev quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2014-09-09T14:24:14Z nicdev joined #sbcl
2014-09-09T14:42:42Z Puffin is now known as BitPuffin
2014-09-09T15:10:23Z gingerale joined #sbcl
2014-09-09T15:48:53Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-09T15:49:20Z drmeiste_ joined #sbcl
2014-09-09T16:11:37Z White__Flame: disassemblies seem to always start at the no-arg parsing entry point.  Out of curiosity, is there way to disassemble the arg parsing as well?
2014-09-09T16:13:28Z White__Flame is now known as White_Flame
2014-09-09T16:17:16Z Krystof: from memory, something like (sb-disassem:disassemble-code-component #'+)
2014-09-09T16:18:26Z White_Flame: thanks, that seems to be it
2014-09-09T16:51:13Z BitPuffin quit (Ping timeout: 260 seconds)
2014-09-09T16:58:41Z drmeiste_ quit (Ping timeout: 260 seconds)
2014-09-09T17:24:39Z eudoxia quit (Quit: Lost terminal)
2014-09-09T17:30:01Z drmeiste_ joined #sbcl
2014-09-09T17:45:52Z krzysz00 joined #sbcl
2014-09-09T17:49:05Z carvite quit (Ping timeout: 260 seconds)
2014-09-09T18:04:52Z yacks quit (Ping timeout: 240 seconds)
2014-09-09T18:13:17Z yacks joined #sbcl
2014-09-09T18:14:57Z jdz quit (Quit: ZNC - http://znc.in)
2014-09-09T18:15:12Z jdz_ joined #sbcl
2014-09-09T18:29:29Z psilord quit (Ping timeout: 276 seconds)
2014-09-09T18:33:26Z carvite joined #sbcl
2014-09-09T18:34:41Z drmeiste_ quit (Ping timeout: 276 seconds)
2014-09-09T18:35:27Z psilord joined #sbcl
2014-09-09T18:36:02Z krzysz00 quit (Ping timeout: 245 seconds)
2014-09-09T18:37:11Z carvite quit (Remote host closed the connection)
2014-09-09T18:37:50Z carvite joined #sbcl
2014-09-09T18:54:19Z attila_lendvai joined #sbcl
2014-09-09T18:54:19Z attila_lendvai quit (Changing host)
2014-09-09T18:54:19Z attila_lendvai joined #sbcl
2014-09-09T19:05:00Z krzysz00 joined #sbcl
2014-09-09T19:05:13Z drmeiste_ joined #sbcl
2014-09-09T19:19:54Z krzysz00 quit (Ping timeout: 268 seconds)
2014-09-09T19:21:18Z drmeiste_ is now known as drmeister
2014-09-09T19:21:21Z BitPuffin joined #sbcl
2014-09-09T19:54:59Z Bicyclidine joined #sbcl
2014-09-09T20:28:14Z angavrilov quit (Remote host closed the connection)
2014-09-09T20:35:57Z gingerale quit (Ping timeout: 264 seconds)
2014-09-09T21:04:19Z drmeister quit (Read error: Connection reset by peer)
2014-09-09T21:06:24Z drmeister joined #sbcl
2014-09-09T21:49:33Z krzysz00 joined #sbcl
2014-09-09T21:52:52Z logand` joined #sbcl
2014-09-09T21:56:40Z logand`: (sb-kernel:make-double-float 3225976832 0) gives #<TYPE-ERROR expected-type: (SIGNED-BYTE 32) datum: 3225976832;  what is the best way to convert 3225976832 to signed-byte 32 for making double float with the same bits?
2014-09-09T22:00:41Z pkhuong: sb-c::mask-signed-field
2014-09-09T22:07:28Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-09T22:19:20Z logand`` joined #sbcl
2014-09-09T22:23:15Z logand` quit (Ping timeout: 255 seconds)
2014-09-09T22:26:06Z drmeister quit (Remote host closed the connection)
2014-09-09T22:36:39Z drmeister joined #sbcl
2014-09-09T22:46:47Z attila_lendvai quit (Quit: Leaving.)
2014-09-09T22:56:10Z psilord quit (Quit: Leaving.)
2014-09-09T22:59:45Z drmeister quit (Read error: Connection reset by peer)
2014-09-09T23:26:59Z edgar-rft quit (Quit: experience disappeared because of unknown reasons)
2014-09-09T23:49:52Z leo2007 quit (Ping timeout: 245 seconds)
2014-09-09T23:53:05Z psilord joined #sbcl
2014-09-10T00:00:50Z leo2007 joined #sbcl
2014-09-10T00:03:36Z slyrus quit (Ping timeout: 255 seconds)
2014-09-10T00:38:44Z oleo is now known as Guest3107
2014-09-10T00:39:37Z oleo__ joined #sbcl
2014-09-10T00:41:50Z Guest3107 quit (Ping timeout: 255 seconds)
2014-09-10T00:43:35Z drmeister joined #sbcl
2014-09-10T00:50:33Z BitPuffin quit (Ping timeout: 246 seconds)
2014-09-10T01:05:49Z slyrus joined #sbcl
2014-09-10T01:10:47Z prxq_ joined #sbcl
2014-09-10T01:13:57Z prxq quit (Ping timeout: 245 seconds)
2014-09-10T01:30:32Z LiamH joined #sbcl
2014-09-10T02:21:07Z slyrus quit (Ping timeout: 245 seconds)
2014-09-10T02:39:33Z christoph_debian quit (Ping timeout: 264 seconds)
2014-09-10T02:46:55Z LiamH quit (Quit: Leaving.)
2014-09-10T02:53:25Z christoph_debian joined #sbcl
2014-09-10T03:07:03Z leo2007 quit (Ping timeout: 246 seconds)
2014-09-10T03:14:22Z leo2007 joined #sbcl
2014-09-10T03:35:50Z leo2007 quit (Quit: rcirc on GNU Emacs 24.4.50.1)
2014-09-10T03:48:19Z slyrus joined #sbcl
2014-09-10T03:53:58Z gingerale joined #sbcl
2014-09-10T04:59:15Z gingerale quit (Ping timeout: 255 seconds)
2014-09-10T05:00:02Z oleo__ quit (Quit: Verlassend)
2014-09-10T05:05:39Z leo2007 joined #sbcl
2014-09-10T05:06:06Z pranavrc joined #sbcl
2014-09-10T05:32:15Z wmarvel quit (Ping timeout: 268 seconds)
2014-09-10T05:43:51Z angavrilov joined #sbcl
2014-09-10T05:54:57Z wmarvel joined #sbcl
2014-09-10T06:42:36Z logand``: pkhuong: thanks a lot!
2014-09-10T07:13:32Z kanru left #sbcl
2014-09-10T07:15:23Z kanru joined #sbcl
2014-09-10T07:44:25Z yacks quit (Ping timeout: 260 seconds)
2014-09-10T07:51:30Z prxq_ quit (Remote host closed the connection)
2014-09-10T08:18:59Z kanru quit (Read error: Connection reset by peer)
2014-09-10T08:20:05Z kanru joined #sbcl
2014-09-10T08:24:22Z pkhuong quit (Ping timeout: 245 seconds)
2014-09-10T08:26:02Z kanru quit (Read error: Connection reset by peer)
2014-09-10T08:27:02Z kanru joined #sbcl
2014-09-10T08:27:24Z kanru quit (Remote host closed the connection)
2014-09-10T08:32:06Z pkhuong joined #sbcl
2014-09-10T08:34:05Z kanru joined #sbcl
2014-09-10T08:35:41Z momo-reina joined #sbcl
2014-09-10T08:42:20Z pranavrc_ joined #sbcl
2014-09-10T08:42:47Z pranavrc quit (Ping timeout: 272 seconds)
2014-09-10T08:45:06Z BitPuffin joined #sbcl
2014-09-10T08:47:58Z hlavaty joined #sbcl
2014-09-10T08:58:44Z alchemis7 quit (Remote host closed the connection)
2014-09-10T09:07:33Z momo-reina quit (Ping timeout: 246 seconds)
2014-09-10T09:15:29Z edgar-rft joined #sbcl
2014-09-10T10:06:09Z echo-area joined #sbcl
2014-09-10T10:22:08Z echo-area quit (Remote host closed the connection)
2014-09-10T10:32:39Z yacks joined #sbcl
2014-09-10T11:34:47Z stassats joined #sbcl
2014-09-10T11:34:50Z stassats quit (Changing host)
2014-09-10T11:34:50Z stassats joined #sbcl
2014-09-10T11:51:57Z alchemis7 joined #sbcl
2014-09-10T11:54:25Z attila_lendvai joined #sbcl
2014-09-10T12:06:02Z drmeister quit (Remote host closed the connection)
2014-09-10T12:06:29Z drmeister joined #sbcl
2014-09-10T12:09:14Z drmeister quit (Remote host closed the connection)
2014-09-10T12:22:25Z drmeister joined #sbcl
2014-09-10T12:33:14Z drmeister quit (Remote host closed the connection)
2014-09-10T12:50:28Z drmeister joined #sbcl
2014-09-10T12:50:49Z drmeister quit (Read error: Connection reset by peer)
2014-09-10T13:16:55Z eudoxia joined #sbcl
2014-09-10T13:22:41Z oleo joined #sbcl
2014-09-10T13:50:36Z drmeister joined #sbcl
2014-09-10T13:51:53Z pranavrc_ quit
2014-09-10T14:09:05Z stassats` joined #sbcl
2014-09-10T14:14:57Z stassats`: is (declaim (hash-table *x*)) (loop for *x* = (make-hash-table)) really supposed to type-conflict?
2014-09-10T14:15:16Z stassats` quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2014-09-10T14:16:45Z stassats: i'll use a LET alright, but that's a tad inconvenient
2014-09-10T14:18:56Z Krystof: uh, doesn't that make a new hash table each iteration?  Is that what you meant?
2014-09-10T14:19:06Z stassats: it is
2014-09-10T14:19:10Z Krystof: oh ok
2014-09-10T14:19:21Z stassats: the problem is that it's set to NIL in the prologue
2014-09-10T14:19:35Z Krystof: yeah.  I was going to suggest with *x* but not if that's really what you mean
2014-09-10T14:19:35Z stassats: i remember something to that case in the clhs
2014-09-10T14:19:39Z Krystof: our loop expansion sucks
2014-09-10T14:19:50Z stassats: our loop sucks
2014-09-10T14:20:22Z pkhuong: yup, our loop sucks.
2014-09-10T14:21:19Z flip214: isn't ITERATE already standard in CL21? ;)
2014-09-10T14:21:34Z stassats: who said that iterate is any better?
2014-09-10T14:21:44Z eudoxia: it isn't but i think it has for and while
2014-09-10T14:22:56Z ams: eudoxia: LOOP also has for and while, at least with the original LOOP ... (you could add your own keywords)
2014-09-10T14:24:07Z eudoxia: i know, i mean a dedicated (for (item list) body) (while cond body) macro
2014-09-10T14:24:35Z flip214: stassats: how about (loop for *x* hash-table = (make-hash-table)) ?
2014-09-10T14:24:51Z stassats: what did i write that declaim for?
2014-09-10T14:25:46Z stassats: a) it doesn't work, b) i don't want to specify the types everywhere instead of declaiming it
2014-09-10T14:27:51Z pkhuong: flip214: the extensions examples in iterate's manual are broken, so I would hope not.
2014-09-10T14:30:05Z flip214: pkhuong: I already defined own extensions, so they're not _fully_ defunct, are they?
2014-09-10T14:36:19Z pkhuong: I don't know about defunct. Defective, perhaps.
2014-09-10T14:39:13Z fridim_ joined #sbcl
2014-09-10T14:46:44Z stassats: it does say that the variables are initialized in the prologue
2014-09-10T14:47:05Z stassats: maybe they should be initialized to the specified form?
2014-09-10T14:48:56Z foom: I expect it's done this way so as not to have to repeat the same code in the "let" and within the loop?
2014-09-10T14:49:26Z stassats: sometimes it does need to be repeated anyway
2014-09-10T14:50:53Z stassats: e.g. (loop for *x* hash-table = (make-hash-table) for x = 10 then 20 ), but yeah, sometimes it saves things
2014-09-10T14:52:01Z stassats: but being correct is usually more important than saving space
2014-09-10T14:53:04Z flip214: stassats: but calling it once _before_ the loop might be wrong, too!
2014-09-10T14:53:17Z stassats: that is the question
2014-09-10T14:53:29Z flip214: (loop while nil for *x* = (`rm -rf /*`))
2014-09-10T14:53:49Z stassats: that's something pjb would write
2014-09-10T14:54:07Z stassats: besides, that's not a legal loop form
2014-09-10T14:54:11Z flip214: the only way that I can see is to provide a "cheap" initial value (if possible), and the updates.
2014-09-10T14:54:39Z flip214: "for *x* hash-table initially (make-hash-table) then (make-hash-table)" is what you want/need.
2014-09-10T14:54:58Z flip214: then others can write
2014-09-10T14:55:06Z stassats: is it?
2014-09-10T14:55:17Z flip214: "for *x* initially nil then (expensive-op)"
2014-09-10T14:55:27Z flip214: I guess so. YMMV ;)
2014-09-10T14:55:49Z stassats: are we still talking about cl:loop?
2014-09-10T14:57:03Z flip214: yes!?
2014-09-10T14:58:09Z flip214: ah, sorry. in LOOP it's "for var [type-spec] = expr1 [then expr2]
2014-09-10T14:58:17Z flip214: not "initially"
2014-09-10T14:58:23Z pkhuong: stassats: I guess we might have a special "undef" value that always passes casts and if flow analysis can show that it's never read, we don't warn
2014-09-10T14:58:23Z stassats: it's also of-type hash-table
2014-09-10T14:58:53Z pkhuong: but that won't play nicely with closures until we're smarter about the scope of captures
2014-09-10T14:59:40Z stassats: on a similar note, i still haven't fixed the destructuring-bind order of binding
2014-09-10T15:00:04Z stassats: got it mostly working, mostly converting everything into a single LET*
2014-09-10T15:00:31Z stassats: (as in, (destructuring-bind (&optional y ((&optional x) y)) '((1)) x) should work)
2014-09-10T15:00:52Z fridim_ quit (Ping timeout: 240 seconds)
2014-09-10T15:15:59Z gingerale joined #sbcl
2014-09-10T15:18:08Z drmeister quit (Ping timeout: 276 seconds)
2014-09-10T15:25:19Z drmeister joined #sbcl
2014-09-10T15:28:04Z drmeiste_ joined #sbcl
2014-09-10T15:32:26Z drmeister quit (Ping timeout: 276 seconds)
2014-09-10T15:43:28Z krzysz00 quit (Ping timeout: 252 seconds)
2014-09-10T16:33:09Z stassats quit (Ping timeout: 255 seconds)
2014-09-10T17:44:01Z drmeiste_ is now known as drmeister
2014-09-10T17:47:46Z slyrus quit (Ping timeout: 252 seconds)
2014-09-10T17:58:37Z eudoxia quit (Quit: leaving)
2014-09-10T18:00:03Z drmeister quit (Remote host closed the connection)
2014-09-10T18:03:11Z drmeister joined #sbcl
2014-09-10T18:03:41Z krzysz00 joined #sbcl
2014-09-10T18:09:33Z sdemarre joined #sbcl
2014-09-10T18:14:51Z angavrilov quit (Remote host closed the connection)
2014-09-10T18:22:25Z yacks quit (Quit: Leaving)
2014-09-10T18:30:27Z yacks joined #sbcl
2014-09-10T18:37:53Z krzysz00 quit (Ping timeout: 268 seconds)
2014-09-10T18:42:21Z krzysz00 joined #sbcl
2014-09-10T18:56:25Z krzysz00 quit (Ping timeout: 260 seconds)
2014-09-10T19:02:35Z slyrus joined #sbcl
2014-09-10T19:14:30Z Skrylar is now known as ASkrylarA
2014-09-10T19:14:40Z BitPuffin is now known as ABitPuffinA
2014-09-10T19:15:44Z ASkrylarA is now known as Skrylar
2014-09-10T19:16:33Z ABitPuffinA is now known as BitPuffin
2014-09-10T19:33:59Z drmeister quit (Remote host closed the connection)
2014-09-10T19:38:34Z drmeister joined #sbcl
2014-09-10T19:40:21Z drmeiste_ joined #sbcl
2014-09-10T19:40:37Z drmeiste_ is now known as drmeister_
2014-09-10T19:41:34Z slyrus_ joined #sbcl
2014-09-10T19:43:45Z drmeister quit (Ping timeout: 272 seconds)
2014-09-10T19:43:50Z slyrus__ joined #sbcl
2014-09-10T19:44:27Z slyrus quit (Ping timeout: 245 seconds)
2014-09-10T19:44:36Z slyrus__ is now known as slyrus
2014-09-10T19:46:52Z slyrus_ quit (Ping timeout: 245 seconds)
2014-09-10T20:08:10Z theethicalegoist joined #sbcl
2014-09-10T20:11:13Z Bicyclidine joined #sbcl
2014-09-10T20:24:34Z sdemarre quit (Ping timeout: 268 seconds)
2014-09-10T21:01:33Z Bicyclidine quit (Ping timeout: 246 seconds)
2014-09-10T21:07:12Z gingerale quit (Ping timeout: 255 seconds)
2014-09-10T21:07:37Z Bicyclidine joined #sbcl
2014-09-10T21:07:37Z arrdem is now known as grimoire
2014-09-10T21:08:25Z grimoire is now known as GENSYM_9001
2014-09-10T21:08:43Z GENSYM_9001 is now known as arrdem
2014-09-10T21:32:09Z fridim_ joined #sbcl
2014-09-10T21:37:59Z Bicyclidine quit (Ping timeout: 272 seconds)
2014-09-10T21:38:45Z DGASAU quit (Remote host closed the connection)
2014-09-10T21:39:36Z DGASAU joined #sbcl
2014-09-10T21:46:01Z drmeister_ quit (Read error: Connection reset by peer)
2014-09-10T21:46:26Z drmeister joined #sbcl
2014-09-10T22:00:06Z attila_lendvai quit (Quit: Leaving.)
2014-09-10T22:03:58Z drmeister quit (Remote host closed the connection)
2014-09-10T22:09:32Z Bicyclidine joined #sbcl
2014-09-10T22:09:34Z drmeister joined #sbcl
2014-09-10T22:11:57Z fridim_ quit (Ping timeout: 264 seconds)
2014-09-10T22:19:06Z logand``` joined #sbcl
2014-09-10T22:22:48Z logand`` quit (Ping timeout: 255 seconds)
2014-09-10T22:25:11Z slyrus quit (Ping timeout: 276 seconds)
2014-09-10T22:34:59Z drmeister quit (Read error: Connection reset by peer)
2014-09-10T22:40:23Z slyrus joined #sbcl
2014-09-10T22:45:01Z drmeister joined #sbcl
2014-09-10T22:49:06Z theethicalegoist quit (Quit: Leaving)
2014-09-10T23:05:24Z Bicyclidine quit (Ping timeout: 272 seconds)
2014-09-10T23:10:28Z Bicyclidine joined #sbcl
2014-09-10T23:29:52Z slyrus quit (Ping timeout: 252 seconds)
2014-09-10T23:35:35Z slyrus joined #sbcl
2014-09-11T00:00:57Z gingerale joined #sbcl
2014-09-11T00:01:08Z Bicyclidine quit (Quit: Reconnecting)
2014-09-11T00:01:25Z Bicyclidine joined #sbcl
2014-09-11T00:17:52Z Bicyclidine quit (Ping timeout: 240 seconds)
2014-09-11T00:21:03Z BitPuffin quit (Ping timeout: 246 seconds)
2014-09-11T00:34:03Z gingerale quit (Read error: Connection reset by peer)
2014-09-11T00:38:35Z Bicyclidine joined #sbcl
2014-09-11T00:38:48Z DGASAU quit (Remote host closed the connection)
2014-09-11T00:39:47Z DGASAU joined #sbcl
2014-09-11T00:44:47Z Bicyclidine quit (Ping timeout: 245 seconds)
2014-09-11T00:57:29Z slyrus quit (Ping timeout: 272 seconds)
2014-09-11T01:21:56Z echo-area joined #sbcl
2014-09-11T01:39:49Z oleo__ joined #sbcl
2014-09-11T01:41:18Z oleo is now known as Guest82040
2014-09-11T01:41:33Z Guest82040 quit (Ping timeout: 246 seconds)
2014-09-11T02:38:22Z christoph_debian quit (Ping timeout: 240 seconds)
2014-09-11T02:51:59Z christoph_debian joined #sbcl
2014-09-11T03:19:23Z Bike quit (Quit: restart)
2014-09-11T03:20:31Z Bike joined #sbcl
2014-09-11T03:30:39Z dto joined #sbcl
2014-09-11T03:31:17Z slyrus joined #sbcl
2014-09-11T03:45:04Z dto-00 joined #sbcl
2014-09-11T03:47:23Z dto quit (Remote host closed the connection)
2014-09-11T03:49:36Z dto-00 is now known as dto
2014-09-11T04:34:37Z yacks quit (Ping timeout: 260 seconds)
2014-09-11T05:00:48Z oleo__ quit (Quit: Verlassend)
2014-09-11T05:26:00Z Skrylar quit (Ping timeout: 268 seconds)
2014-09-11T05:35:36Z Skrylar joined #sbcl
2014-09-11T05:51:24Z sdemarre joined #sbcl
2014-09-11T06:07:01Z sdemarre quit (Ping timeout: 260 seconds)
2014-09-11T06:27:34Z stassats joined #sbcl
2014-09-11T06:35:10Z fridim_ joined #sbcl
2014-09-11T06:41:53Z kanru quit (Remote host closed the connection)
2014-09-11T06:42:12Z dto quit (Remote host closed the connection)
2014-09-11T06:42:47Z BitPuffin joined #sbcl
2014-09-11T06:44:41Z kanru joined #sbcl
2014-09-11T06:51:38Z momo-reina joined #sbcl
2014-09-11T07:30:31Z angavrilov joined #sbcl
2014-09-11T08:03:54Z momo-reina quit (Remote host closed the connection)
2014-09-11T08:06:32Z stassats quit (Remote host closed the connection)
2014-09-11T08:06:38Z stassats joined #sbcl
2014-09-11T08:08:05Z yacks joined #sbcl
2014-09-11T08:12:27Z stassats` joined #sbcl
2014-09-11T08:12:27Z stassats` quit (Remote host closed the connection)
2014-09-11T08:30:29Z attila_lendvai joined #sbcl
2014-09-11T10:19:10Z echo-area quit (Remote host closed the connection)
2014-09-11T10:55:58Z Krystof: ouch
2014-09-11T10:55:59Z Krystof: bugs bugs bugs
2014-09-11T11:12:19Z yacks quit (Remote host closed the connection)
2014-09-11T11:23:32Z BitPuffin quit (Ping timeout: 245 seconds)
2014-09-11T11:34:29Z nyef joined #sbcl
2014-09-11T11:34:47Z nyef: G'morning all.
2014-09-11T12:06:58Z BitPuffin joined #sbcl
2014-09-11T12:10:11Z yacks joined #sbcl
2014-09-11T12:19:45Z drmeister quit (Remote host closed the connection)
2014-09-11T12:21:53Z DGASAU quit (Read error: Connection reset by peer)
2014-09-11T12:22:29Z DGASAU joined #sbcl
2014-09-11T12:31:38Z DGASAU quit (Read error: Connection reset by peer)
2014-09-11T12:32:14Z DGASAU joined #sbcl
2014-09-11T12:42:23Z DGASAU quit (Read error: Connection reset by peer)
2014-09-11T12:43:06Z DGASAU joined #sbcl
2014-09-11T12:55:21Z psilord quit (Quit: Leaving.)
2014-09-11T13:04:30Z attila_lendvai quit (Quit: Leaving.)
2014-09-11T13:24:23Z oleo joined #sbcl
2014-09-11T13:50:42Z oleo quit (Ping timeout: 245 seconds)
2014-09-11T13:51:12Z oleo joined #sbcl
2014-09-11T13:56:56Z Krystof: is it obvious to anyone why I get results from walk.impure.lisp printed twice when I do ./run-tests.sh?
2014-09-11T13:57:17Z stassats: no with-test?
2014-09-11T13:57:32Z Krystof: there is a with-test
2014-09-11T13:57:46Z Krystof: but not when I specify ./run-tests.sh walk.impure.lisp
2014-09-11T13:58:02Z stassats: maybe it finds to execute it twice
2014-09-11T13:58:06Z stassats: let me see
2014-09-11T14:00:19Z stassats: but is it run twice? or just the results?
2014-09-11T14:01:06Z stassats: do you maybe have an emacs created backup file?
2014-09-11T14:01:51Z stassats: though that would be .#walk.impure.lisp
2014-09-11T14:02:14Z |3b|: hmm, looks like i got that last time i ran tests too
2014-09-11T14:02:42Z stassats: i had that with some tests, but i didn't bother figuring out, just fixed the failures
2014-09-11T14:04:28Z drmeister joined #sbcl
2014-09-11T14:05:28Z stassats: and i can't artificially reproduce it now
2014-09-11T14:10:43Z Krystof: can I leave that with someone else to sort out because I am knee-deep in walker changes
2014-09-11T14:12:01Z psilord joined #sbcl
2014-09-11T14:21:29Z |3b|: looks like it only ran the file once
2014-09-11T14:23:42Z attila_lendvai joined #sbcl
2014-09-11T14:28:02Z |3b|: prints twice with sh ./run-tests.sh walk.impure.lisp win32-foreign-stack-unwind.impure.lisp
2014-09-11T14:28:21Z eudoxia joined #sbcl
2014-09-11T14:29:13Z stassats: indeed
2014-09-11T14:42:01Z stassats: i see the problem
2014-09-11T14:46:20Z attila_lendvai quit (Quit: Leaving.)
2014-09-11T14:48:44Z stassats: fixed
2014-09-11T14:51:54Z Krystof: thank
2014-09-11T14:51:55Z Krystof: s
2014-09-11T14:59:44Z gingerale joined #sbcl
2014-09-11T15:06:58Z drmeiste_ joined #sbcl
2014-09-11T15:09:45Z drmeister quit (Ping timeout: 260 seconds)
2014-09-11T15:20:36Z eudoxia quit (Quit: leaving)
2014-09-11T15:21:29Z eudoxia joined #sbcl
2014-09-11T15:27:56Z stassats: what's with the latest bug ticket?
2014-09-11T15:28:11Z stassats: there's too much java
2014-09-11T15:32:16Z drmeiste_ is now known as drmeister
2014-09-11T15:32:17Z jsnell_: clearly somebody found the wrong bug reporting page
2014-09-11T15:34:27Z Krystof: I'm so glad they specified their version of sbcl
2014-09-11T15:35:18Z stassats: and it's an older one, as usual
2014-09-11T15:45:29Z stassats quit (Read error: No route to host)
2014-09-11T15:45:35Z stassats` joined #sbcl
2014-09-11T15:46:21Z Krystof: right, now that's all the bugs that I can find just by reading the code.
2014-09-11T15:46:25Z Krystof: wtf is a bogofun?
2014-09-11T15:47:01Z Krystof: and why is a walker environment different from a normal environment.  PCL heritage, I guess
2014-09-11T15:53:19Z eudoxia quit (Quit: Lost terminal)
2014-09-11T15:53:47Z eudoxia joined #sbcl
2014-09-11T15:59:50Z pkhuong: yes
2014-09-11T16:03:29Z Krystof: gaah ";;; There are some semantic problems with this facility."
2014-09-11T16:03:45Z Krystof: ok, I can be thankful that we are not using all of PCL's heritage
2014-09-11T16:14:45Z Krystof: this is crying for a random form generation tester
2014-09-11T16:23:22Z stassats` quit (Ping timeout: 240 seconds)
2014-09-11T16:50:20Z pkhuong: nyef: jemalloc 4 lets you feed it chunks of memory
2014-09-11T16:52:38Z pkhuong: don't let it pretend it's regular malloc, one arena for boxed objects and another for unboxed, and only get chunks dynamic space. perfect for a mark/sweep collector.
2014-09-11T16:53:54Z nyef: Hrm... And was your reaction to the suggestion of always bundling array headers with their data vectors the same as mine?
2014-09-11T16:53:57Z pkhuong: not sure how it allocates space for its out of band data structures, though. saving a core miht be interesting (:
2014-09-11T16:54:15Z pkhuong: nyef: uh?
2014-09-11T16:54:31Z pkhuong: for non-vector simple arrays?
2014-09-11T16:54:51Z nyef: Nikodemus' suggestion in the array-name thread.
2014-09-11T16:55:44Z pkhuong: kills with-array-data
2014-09-11T16:55:52Z pkhuong: oh, no I see.
2014-09-11T16:56:12Z pkhuong: mm...
2014-09-11T16:56:42Z pkhuong: I'd try and get the GC & allocator to keep them contiguous w/o any assumption in the compiler
2014-09-11T16:57:13Z pkhuong: that'd almost surely be a win, just because of improved locality
2014-09-11T16:57:33Z nyef: Yeah, that'd be a good first step, though you still end up with a mix of boxed and unboxed data when dealing with specialized array types.
2014-09-11T16:58:45Z scymtym joined #sbcl
2014-09-11T16:59:18Z pkhuong: interior pointers really suck in our current scheme
2014-09-11T16:59:33Z pkhuong: a smarter allocator would help.
2014-09-11T17:02:30Z nyef: Smarter in what sense?
2014-09-11T17:03:00Z pkhuong: in the sense that it supports more use cases than just incrementing a pointer
2014-09-11T17:03:08Z pkhuong: so it has to determine where objects begin/end
2014-09-11T17:03:30Z pkhuong: or at least whether an address is or isn't allocated
2014-09-11T17:05:03Z pkhuong: still not sure how we'd support pointers to the interior simple array
2014-09-11T17:05:08Z pkhuong: pinning? that sucks
2014-09-11T17:06:04Z nyef: We... might need to change the array representation a bit for this.
2014-09-11T17:06:45Z nyef: We support pointers to SIMPLE-FUN and LRA objects by being able to find the surrounding object given the embedded object pointer.
2014-09-11T17:07:33Z pkhuong: if we go down that route
2014-09-11T17:07:59Z pkhuong: we might as well implement david's struct thing too
2014-09-11T17:08:10Z pkhuong: (unboxed slots at negative offsets)
2014-09-11T17:15:34Z nyef: At the same time, I think there's a lot more that could be done with the GC, and plausibly to better general effect.
2014-09-11T17:17:53Z sdemarre joined #sbcl
2014-09-11T17:20:35Z nyef: Do we still have the GC race conditions with allocating from the runtime?
2014-09-11T17:29:08Z eudoxia quit (Quit: Lost terminal)
2014-09-11T17:37:51Z krzysz00 joined #sbcl
2014-09-11T17:40:49Z sdemarre quit (Ping timeout: 267 seconds)
2014-09-11T17:58:54Z sdemarre joined #sbcl
2014-09-11T18:15:39Z pkhuong: yes
2014-09-11T18:17:40Z nyef: Okay, I think I might have an angle for that, at least.
2014-09-11T18:18:21Z pkhuong: better than my lisppaste?
2014-09-11T18:18:43Z nyef: I don't know if I've seen your lisppaste. URL?
2014-09-11T18:20:53Z pkhuong: http://paste.lisp.org/display/142817#2
2014-09-11T18:22:19Z nyef: It's not sufficient for non-x86oids.
2014-09-11T18:22:26Z pkhuong: nope
2014-09-11T18:24:31Z nyef: This is ringing a bit of a bell now, and it's a good start, but we still need to cover the non-x86oid cases.
2014-09-11T18:25:21Z psilord quit (Ping timeout: 246 seconds)
2014-09-11T18:28:25Z psilord joined #sbcl
2014-09-11T18:28:45Z nyef: Umm... And AIUI, pa_alloc() is pseudo-atomic on any threaded platform, and ARM, as we need it to be P-A for threading purposes and ARM uses an x86oid-style P-A implementation.
2014-09-11T18:30:09Z pkhuong: it might be
2014-09-11T18:30:23Z pkhuong: but it's not enough to have the allocation itself be PA
2014-09-11T18:30:28Z pkhuong: we need initialisation as well
2014-09-11T18:31:07Z nyef: Yes, but for "precise" targets we also need to pin the object until it's either no longer used or passed back to Lisp.
2014-09-11T18:31:35Z pkhuong: right
2014-09-11T18:31:35Z nyef: For that matter, access to the lisp heap from the runtime generally should be audited for precisely this.
2014-09-11T18:31:50Z pkhuong: but for precise targets, the pin list should be explicit
2014-09-11T18:32:01Z nyef: Yes, and we should use it.
2014-09-11T18:32:10Z pkhuong: so it's not *as* bad
2014-09-11T18:32:22Z slyrus quit (Ping timeout: 240 seconds)
2014-09-11T18:32:38Z nyef: For that matter, we could also have an explicit pin list for the non-precise targets, as it would do no harm and would allow us to use the same code paths for both types of target.
2014-09-11T18:32:55Z pkhuong: sure. DX pin list?
2014-09-11T18:33:57Z nyef: DXing the pin list requires :stack-allocatable-fixed-objects (when pinning a single object at a time) and :stack-allocatable-lists (when pinning more than one object at once).
2014-09-11T18:34:18Z nyef: But we can DX the pin list entries on the C stack if we're in the runtime.
2014-09-11T18:34:39Z pkhuong: .. that sounds dangerous
2014-09-11T18:34:52Z nyef: The dangerous bit is the unwinds.
2014-09-11T18:35:18Z nyef: For which? Add a binding stack entry for the cleanup.
2014-09-11T18:36:09Z pkhuong: yeah, I'm concerned about unwinding in the right order
2014-09-11T18:38:44Z krzysz00 quit (Ping timeout: 276 seconds)
2014-09-11T19:04:23Z krzysz00 joined #sbcl
2014-09-11T19:05:33Z prxq joined #sbcl
2014-09-11T19:32:48Z krzysz00 quit (Ping timeout: 272 seconds)
2014-09-11T19:33:29Z krzysz00 joined #sbcl
2014-09-11T19:35:37Z BitPuffin quit (Ping timeout: 245 seconds)
2014-09-11T19:37:30Z psilord quit (Quit: Leaving.)
2014-09-11T19:40:14Z kanru quit (Ping timeout: 252 seconds)
2014-09-11T19:59:30Z sdemarre quit (Ping timeout: 246 seconds)
2014-09-11T20:20:22Z krzysz00 quit (Ping timeout: 240 seconds)
2014-09-11T20:20:42Z InfusoElAmbulant joined #sbcl
2014-09-11T20:41:50Z BitPuffin joined #sbcl
2014-09-11T20:47:48Z prxq quit (Ping timeout: 246 seconds)
2014-09-11T20:50:42Z gingerale quit (Ping timeout: 272 seconds)
2014-09-11T20:59:08Z slyrus joined #sbcl
2014-09-11T21:10:39Z krzysz00 joined #sbcl
2014-09-11T21:18:37Z fridim_ quit (Ping timeout: 245 seconds)
2014-09-11T21:20:47Z drmeister quit (Remote host closed the connection)
2014-09-11T21:31:46Z drmeister joined #sbcl
2014-09-11T21:37:59Z DGASAU quit (Remote host closed the connection)
2014-09-11T21:39:19Z DGASAU joined #sbcl
2014-09-11T21:56:43Z slyrus_ joined #sbcl
2014-09-11T21:58:56Z slyrus quit (Ping timeout: 276 seconds)
2014-09-11T21:59:09Z slyrus_ is now known as slyrus
2014-09-11T21:59:52Z drmeister quit (Read error: Connection reset by peer)
2014-09-11T22:00:18Z angavrilov quit (Remote host closed the connection)
2014-09-11T22:03:43Z InfusoElAmbulant quit (Remote host closed the connection)
2014-09-11T22:04:08Z InfusoElAmbulant joined #sbcl
2014-09-11T22:10:02Z drmeister joined #sbcl
2014-09-11T22:23:17Z logand``` quit (Ping timeout: 260 seconds)
2014-09-11T22:31:29Z scymtym_ joined #sbcl
2014-09-11T22:32:05Z slyrus_ joined #sbcl
2014-09-11T22:33:12Z slyrus quit (Ping timeout: 245 seconds)
2014-09-11T22:33:16Z slyrus_ is now known as slyrus
2014-09-11T22:34:26Z scymtym quit (Ping timeout: 252 seconds)
2014-09-11T22:37:28Z drmeister quit (Remote host closed the connection)
2014-09-11T22:38:43Z eudoxia joined #sbcl
2014-09-11T22:41:17Z InfusoElAmbulant quit (Remote host closed the connection)
2014-09-11T22:41:27Z psilord joined #sbcl
2014-09-11T22:53:40Z davazp joined #sbcl
2014-09-11T23:02:31Z attila_lendvai joined #sbcl
2014-09-11T23:28:17Z drmeister joined #sbcl
2014-09-11T23:29:08Z arrdem quit (Quit: leaving)
2014-09-11T23:34:00Z attila_lendvai quit (Quit: Leaving.)
2014-09-11T23:44:03Z drmeister quit (Read error: Connection reset by peer)
2014-09-12T00:29:42Z scymtym_ quit (Ping timeout: 246 seconds)
2014-09-12T00:34:07Z davazp quit (Remote host closed the connection)
2014-09-12T00:35:27Z eudoxia quit (Quit: Lost terminal)
2014-09-12T00:48:54Z drmeister joined #sbcl
2014-09-12T00:56:23Z BitPuffin quit (Ping timeout: 276 seconds)
2014-09-12T01:06:52Z akkad quit (Quit: Emacs must have died)
2014-09-12T01:07:23Z akkad joined #sbcl
2014-09-12T01:29:45Z nyef: ... What does the runtime allocate that can't be stack-allocated, anyway?
2014-09-12T01:30:07Z nyef: (Almost any SAPs allocated are used in D-X fashion, after all.)
2014-09-12T01:35:34Z nyef: I suppose interactions with ARTLF would be problematic.
2014-09-12T01:43:29Z slyrus quit (Ping timeout: 260 seconds)
2014-09-12T01:49:29Z Bicyclidine joined #sbcl
2014-09-12T02:04:13Z kanru joined #sbcl
2014-09-12T02:23:11Z edgar-rft quit (Quit: end of life during parsing)
2014-09-12T02:31:55Z oleo is now known as Guest94911
2014-09-12T02:32:47Z oleo__ joined #sbcl
2014-09-12T02:34:39Z Guest94911 quit (Ping timeout: 246 seconds)
2014-09-12T02:39:01Z christoph_debian quit (Ping timeout: 260 seconds)
2014-09-12T02:41:33Z drmeister quit (Remote host closed the connection)
2014-09-12T02:51:26Z Bicyclidine quit (Ping timeout: 276 seconds)
2014-09-12T02:52:04Z christoph_debian joined #sbcl
2014-09-12T03:01:09Z drmeister joined #sbcl
2014-09-12T04:12:08Z ams quit (Ping timeout: 272 seconds)
2014-09-12T05:16:20Z pranavrc joined #sbcl
2014-09-12T05:16:20Z pranavrc quit (Changing host)
2014-09-12T05:16:20Z pranavrc joined #sbcl
2014-09-12T05:16:30Z nyef quit (Quit: G'night all.)
2014-09-12T05:30:50Z oleo__ quit (Quit: Verlassend)
2014-09-12T05:58:10Z fridim_ joined #sbcl
2014-09-12T06:19:50Z BitPuffin joined #sbcl
2014-09-12T06:20:29Z logand joined #sbcl
2014-09-12T06:26:15Z ams joined #sbcl
2014-09-12T06:53:26Z gingerale joined #sbcl
2014-09-12T07:20:06Z prxq joined #sbcl
2014-09-12T07:23:10Z attila_lendvai joined #sbcl
2014-09-12T07:26:51Z attila_lendvai quit (Client Quit)
2014-09-12T07:29:13Z ivan`` quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2014-09-12T07:30:24Z fridim_ quit (Remote host closed the connection)
2014-09-12T07:33:59Z ivan`` joined #sbcl
2014-09-12T08:03:52Z angavrilov joined #sbcl
2014-09-12T08:20:18Z pranavrc quit (Remote host closed the connection)
2014-09-12T08:23:53Z DGASAU quit (Read error: Connection reset by peer)
2014-09-12T08:24:33Z DGASAU joined #sbcl
2014-09-12T09:20:15Z fridim_ joined #sbcl
2014-09-12T09:29:17Z pranavrc joined #sbcl
2014-09-12T09:29:17Z pranavrc quit (Changing host)
2014-09-12T09:29:17Z pranavrc joined #sbcl
2014-09-12T09:35:13Z stassats joined #sbcl
2014-09-12T09:40:35Z attila_lendvai joined #sbcl
2014-09-12T10:09:02Z BitPuffin quit (Read error: Connection reset by peer)
2014-09-12T10:09:17Z BitPuffin joined #sbcl
2014-09-12T10:55:39Z fridim_ quit (Ping timeout: 255 seconds)
2014-09-12T11:07:00Z edgar-rft joined #sbcl
2014-09-12T11:08:25Z segv- joined #sbcl
2014-09-12T11:20:03Z nyef joined #sbcl
2014-09-12T11:22:48Z attila_lendvai quit (Quit: Leaving.)
2014-09-12T11:48:45Z yacks quit (Read error: Connection reset by peer)
2014-09-12T12:26:43Z drmeister quit (Remote host closed the connection)
2014-09-12T12:38:51Z oleo joined #sbcl
2014-09-12T12:57:33Z fridim_ joined #sbcl
2014-09-12T12:59:57Z ebrasca joined #sbcl
2014-09-12T13:01:00Z ebrasca: have sbcl dtrace?
2014-09-12T13:01:15Z stassats: no
2014-09-12T13:01:56Z ebrasca: stassats: thx
2014-09-12T13:02:17Z stassats: isn't dtrace independent of things?
2014-09-12T13:10:10Z ebrasca: stassts: thx for help.DTRACE is a portable alternative to the Common Lisp TRACE and UNTRACE macros.
2014-09-12T13:10:40Z stassats: oh well, and not a solaris thingy?
2014-09-12T13:11:07Z |3b| assumed the OS thing also
2014-09-12T13:11:21Z nyef: It wouldn't surprise me if it were both. Look at the number of things called "allegro", after all. d-:
2014-09-12T13:11:41Z stassats: i rode on a train called "allegro"
2014-09-12T13:12:08Z |3b|: sbcl's TRACE is fairly configurable, so you might be able to get the information you want from it if you can't port dtrace
2014-09-12T13:12:23Z stassats: and ate some candies named "allegro"
2014-09-12T13:12:42Z pkhuong: and I couldn't live without allegra
2014-09-12T13:12:46Z nyef: But don't use (TRACE ... :ENCAPSULATE NIL) unless you know how and why it's broken.
2014-09-12T13:12:56Z jsnell_: I'm intending to get mbenkard's evaluator branch from last GSoC into a mergable state this weekend
2014-09-12T13:13:01Z pkhuong: jsnell_: awesome.
2014-09-12T13:13:06Z jsnell_: one part of that is probably renaming it from "eval2"...
2014-09-12T13:13:13Z pkhuong: heh.
2014-09-12T13:13:14Z |3b|: though stassats broke some of the configurability of TRACE :p
2014-09-12T13:13:22Z stassats: jsnell_: does it still hack its own destructuring-bind?
2014-09-12T13:13:39Z stassats: i had some changes lined up for full-eval which may hurt the merge
2014-09-12T13:13:43Z jsnell_: any opinions on what to call it (e.g. in package names)?
2014-09-12T13:13:53Z pkhuong: jsnell_: would we still want to keep the old sb-eval?
2014-09-12T13:14:01Z stassats: |3b|: which?
2014-09-12T13:14:09Z eudoxia joined #sbcl
2014-09-12T13:14:15Z jsnell_: pkhuong: that's the plan, I think
2014-09-12T13:14:19Z |3b|: stassats: https://bugs.launchpad.net/sbcl/+bug/1357826
2014-09-12T13:14:24Z jsnell_: stassats: the existing full-eval isn't touched at all
2014-09-12T13:14:48Z stassats: |3b|: ok, didn't see the bisection
2014-09-12T13:15:09Z |3b|: stassats: looked like it wouldn't be too hard to fix, but no idea which would be the 'right' way
2014-09-12T13:15:20Z pkhuong: I'm not a fan of adding more options (in addition to sb-eval and the internal fake eval), but OK :\
2014-09-12T13:15:20Z jsnell_: this adds a new :minimally-compile *evaluator-mode*
2014-09-12T13:16:06Z jsnell_: google have their own faster interpreting eval, apparently with conflicting requirements
2014-09-12T13:16:10Z stassats: i'll fix it sometime, but i guess i need to get some work actually done first
2014-09-12T13:18:51Z stassats: i somehow missed that arg-list usage
2014-09-12T13:22:49Z pranavrc quit
2014-09-12T13:23:44Z Krystof: pkhuong: my current thought is that we need an evaluator bake-off
2014-09-12T13:24:06Z Krystof: people need to say "but sb-eval2 chews through too much stack!" "but google-eval has sucky debugger integration"
2014-09-12T13:24:08Z Krystof: or whatever
2014-09-12T13:24:13Z stassats: (flet 1) is a bit broken
2014-09-12T13:24:28Z Krystof: and no-one is saying any of these things because they're all in branches with no visibility, or inside the Googly monolith
2014-09-12T13:25:02Z nyef: You're going to use "debugger integration" as an argument for *ANYTHING* in SBCL?  Really?
2014-09-12T13:25:18Z Krystof: I'm going to moderate the argument and declare a winner at the end
2014-09-12T13:25:27Z Krystof: I will be awarding gold, silver and bronze medals for evaluators
2014-09-12T13:25:32Z Krystof: there's still time for new entrants!
2014-09-12T13:25:52Z stassats: i nominate the compiler
2014-09-12T13:26:28Z Krystof: pkhuong: one possibility with google-eval is that we could lose the PCL walker
2014-09-12T13:26:37Z Krystof: so that's something
2014-09-12T13:27:13Z nyef: git rm src/pcl/walk.lisp; echo "what PCL walker?"
2014-09-12T13:27:51Z stassats: a walker competition too?
2014-09-12T13:27:54Z Krystof: nyef: tempting, but that would break 99% of the code out there
2014-09-12T13:28:18Z Krystof: and though I don't think abuse on twitter is a lot of fun, in this case it might be justified
2014-09-12T13:28:38Z nyef: stassats: Will someone nominate Johnny Walker?
2014-09-12T13:29:06Z stassats: hopefully
2014-09-12T13:29:29Z stassats: i also want better LOOP and FORMAT implementations
2014-09-12T13:29:45Z pkhuong: Krystof: that'd be cool.
2014-09-12T13:30:01Z nyef: Can we also get a version of PCL that doesn't require the compiler at runtime?
2014-09-12T13:30:16Z Krystof: sure!  Just run an evaluator :)
2014-09-12T13:30:18Z stassats: can we get a version of PCL which can be used by the compiler?
2014-09-12T13:30:20Z pkhuong: nyef: if we have a minimal compiler, it might be good enough to replace the compiler.
2014-09-12T13:30:28Z pkhuong: stassats: single dispatch OK?
2014-09-12T13:30:36Z Krystof: sb-eval2 is probably good enough to do all of PCL
2014-09-12T13:30:49Z Krystof: I would also expect google-eval to be, given that they've removed the PCL walker
2014-09-12T13:30:55Z pkhuong: oh nice.
2014-09-12T13:31:13Z Krystof: *I* want a make-method-lambda that preserves cons identities so that sb-cover works
2014-09-12T13:31:24Z Krystof: maybe composable-codewalker could make that happen?
2014-09-12T13:31:37Z pkhuong: I should pop by their offices and annoy them into releasing their evaluator
2014-09-12T13:31:53Z Krystof: pkhuong: this is why I've pushed jsnell_ into looking at sb-eval2 again
2014-09-12T13:32:03Z jsnell_ is now known as jsnell
2014-09-12T13:32:07Z stassats: wait, there are two new evals?
2014-09-12T13:32:08Z pkhuong: Krystof: it won't preserve cons identities, but it might play nice with sb-cover
2014-09-12T13:32:14Z nyef: Hunh. My problems with SB-COVER are more to do with things like the source files being weirdly broken-up in the toplevel output index.
2014-09-12T13:32:25Z pkhuong: sb-cover can be re-implemented as a composable codewalking extension
2014-09-12T13:32:32Z pkhuong: not all of what PCL does can, however.
2014-09-12T13:32:38Z Krystof: meanwhile, we have cracauer saying that we are about to have a scary new GC, and dougk wants to annotate reader information with line information
2014-09-12T13:32:56Z stassats: source locations are broken everywhere when it comes to top-levelness
2014-09-12T13:32:58Z pkhuong: we could also have a less-scary linux-only one ;)
2014-09-12T13:33:11Z Krystof: yes, I saw chatter about your experiments
2014-09-12T13:33:19Z karswell quit (Read error: Connection reset by peer)
2014-09-12T13:33:21Z Krystof: maybe someday we'll merge _8jean's steeldump
2014-09-12T13:33:25Z stassats: and i still have unknown GC crashes in production
2014-09-12T13:33:58Z pkhuong: I *think* PCL could get away with local codewalking + some form of compiler support for explicit common subexpression elimination
2014-09-12T13:34:22Z nyef: Maybe someday we'll have the ability to build cores without major chunks of the system, for application delivery purposes.
2014-09-12T13:34:32Z pkhuong: like (with-cse-scope scope-key (perhaps-cse-this scope-key value-key form) ...)
2014-09-12T13:35:11Z stassats: maybe someday we'll have relocatable cores, there's already an older branch
2014-09-12T13:35:12Z Krystof: nyef: yeah, if only there were some blog post we could start from
2014-09-12T13:35:25Z pkhuong: and if value-key occurs multiple times, the form gets hoisted up to with-cse-scope.
2014-09-12T13:35:36Z Krystof: stassats: oh yes, I knew there was one of lichteblau's things that I wanted more than steeldump: relocatable cores with growable heap
2014-09-12T13:35:37Z stassats: relocatable cores => sbcl.dll, yay
2014-09-12T13:35:55Z pkhuong: nyef: I made a tiny hello world *once*. It segfaulted on errors and was mostly ffi ;)
2014-09-12T13:36:10Z karswell joined #sbcl
2014-09-12T13:36:12Z nyef: I'm honestly surprised that relocatable cores hasn't been merged yet.
2014-09-12T13:36:27Z stassats: i couldn't built even with the same version?
2014-09-12T13:36:47Z pkhuong: (turns out it's not too hard to write a stupid copying that writes to a file)
2014-09-12T13:37:02Z nyef: ISTR the main issue being that it didn't work for cold-cores, and that should be a long-since solved problem.
2014-09-12T13:37:03Z stassats: and i ran out of steam working on ARM things, so, no threads, no sbcl.so from me
2014-09-12T13:38:49Z nyef: stassats: I'm still impressed by your ARM hacking. You took it from only getting partway through cold-init to something actually fairly polished.
2014-09-12T13:39:38Z stassats: well, i wouldn't be able to do that from scratch, so that was a more important part
2014-09-12T13:40:38Z Krystof: I sometimes think that a starter and a completer are both needed in software projects
2014-09-12T13:40:39Z nyef: Mmm. At some point I'd like to spend the time to go over my logs and write up a bit of a retrospective.
2014-09-12T13:40:42Z stassats: and there is still that pa_alloc thing
2014-09-12T13:41:24Z Krystof: I have worked with some people who are great at starting things and TERRIBLE at keeping them going.  (And I'm broadly the opposite)
2014-09-12T13:42:05Z yacks joined #sbcl
2014-09-12T13:42:15Z stassats: i need immediate gratification, getting the next form of cold-init is one
2014-09-12T13:42:48Z stassats: writing vops to no end without seeing them work wouldn't be as fun
2014-09-12T13:43:36Z pkhuong: same. I sometimes find ways to twist the project into giving me immediate feedback
2014-09-12T13:44:04Z nyef: Same here. That's why the change of strategy for arm-port-2 as compared with the original.
2014-09-12T13:45:13Z nyef: Rather than try to write all of the VOPs required to get IR2TRAN to build, I broke as much of IR2TRAN as required to get the compiler to build and then added enough things back to get it to build a single, simple test file (which got processed into a core file that I could step through with gdb).
2014-09-12T13:46:54Z Krystof: anyone up for a London SBCL15?
2014-09-12T13:47:12Z Krystof: around mid-December?
2014-09-12T13:47:33Z stassats: that, like, a year from now?
2014-09-12T13:47:36Z Krystof: we could laugh hollowly about all the things that we said we'd do at SBCL10 that haven't happened
2014-09-12T13:47:39Z Krystof: no, 3 months
2014-09-12T13:47:48Z Krystof: SBCL0 was December 1999
2014-09-12T13:47:54Z stassats: oh, right
2014-09-12T13:48:53Z pkhuong: I could probably get space in NYC, FWIW
2014-09-12T13:49:26Z reb`: Google may have a free conference room in NYC ...
2014-09-12T13:49:41Z Krystof: I sort-of don't want it to be too Googly
2014-09-12T13:50:45Z Krystof: also I hate travel so London is ideal from my point of view :)
2014-09-12T13:50:48Z pkhuong: AppNexus is super non googly ;)
2014-09-12T13:51:07Z Krystof: alternative plan: we could have an SBCL workshop before ELS 2015, which will be in London in April 2015
2014-09-12T13:51:22Z Krystof: state preferences
2014-09-12T13:52:09Z pkhuong: don't think I can do this december
2014-09-12T13:52:22Z pkhuong: unless it's on the (my) east coast
2014-09-12T13:55:47Z nyef: I... probably can't get to any such conference for the foreseeable future, though I hope to change that "soon".
2014-09-12T13:57:11Z stassats: if it were in the schengen area, that would've been easier
2014-09-12T13:58:45Z Krystof: an alternative Schengen location is Bielefeld
2014-09-12T13:59:06Z Krystof: which, frankly, is in the middle of nowhere :)
2014-09-12T13:59:21Z stassats: does it exist?
2014-09-12T14:00:52Z jsnell: Zurich is lovely that time of year, and in the Schengen zone ;-)
2014-09-12T14:01:50Z Krystof: clearly everyone wants to stay where they are, so let's just have a google hangout
2014-09-12T14:02:07Z Krystof: then we can spend hours trying to get acceptable audio, and each drink our own favourite coffee
2014-09-12T14:02:31Z stassats is at some startup accelerator in Estonia right now, using CL, no less
2014-09-12T14:08:39Z drmeister joined #sbcl
2014-09-12T14:09:45Z karswell quit (Read error: Connection reset by peer)
2014-09-12T14:13:01Z karswell joined #sbcl
2014-09-12T14:13:20Z pkhuong: ELS sounds doable. So do Zurich and Bielefeld later next year
2014-09-12T14:16:01Z drmeister quit (Read error: Connection reset by peer)
2014-09-12T14:16:25Z hlavaty: what is in Bielefeld?
2014-09-12T14:18:51Z drmeister joined #sbcl
2014-09-12T14:21:38Z stassats: nothing?
2014-09-12T14:25:33Z Krystof: a University
2014-09-12T14:26:48Z attila_lendvai joined #sbcl
2014-09-12T14:26:48Z attila_lendvai quit (Changing host)
2014-09-12T14:26:48Z attila_lendvai joined #sbcl
2014-09-12T14:27:14Z drmeister quit (Read error: Connection reset by peer)
2014-09-12T14:32:21Z hlavaty: Krystof: they are into sbcl?
2014-09-12T14:36:29Z hlavaty: stassats: it seems it's a place near Hannover, DE;  there are quite a few lispers in germany
2014-09-12T14:36:48Z stassats: there is a joke that it does not exist
2014-09-12T14:37:03Z Krystof: hlavaty: scymtym__ is based there
2014-09-12T14:37:19Z hlavaty: ok
2014-09-12T14:40:50Z hlavaty: stassats: i see https://en.wikipedia.org/wiki/Bielefeld_Conspiracy :-) i recently found that Bundesrepublik Deutschland doesnt really exist either;  and the Reich is still at war...
2014-09-12T14:41:50Z hlavaty: but that's not conspiracy but rather legal chaos :-)
2014-09-12T14:41:52Z stassats: and wwII didn't end either, russia is still at war with japan
2014-09-12T14:53:35Z drmeister joined #sbcl
2014-09-12T14:54:29Z jdz_ quit (Ping timeout: 268 seconds)
2014-09-12T14:54:52Z stassats quit (Ping timeout: 245 seconds)
2014-09-12T14:58:05Z scymtym joined #sbcl
2014-09-12T14:58:33Z ebrasca quit (Quit: ChatZilla 0.9.90.1 [Firefox 24.0/20131118140013])
2014-09-12T15:10:39Z Krystof:          ;; This form was a call to a macro. Maybe it expanded
2014-09-12T15:10:40Z Krystof:          ;; into a declare?  Recurse to find out.
2014-09-12T15:10:42Z Krystof: oh FFS
2014-09-12T15:14:29Z scymtym: Krystof: would you like to discuss next steps for pattern-specializers sometime? (IIRC, they basically work but are super-slow at compile-time)
2014-09-12T15:14:52Z Krystof: that would be good
2014-09-12T15:15:18Z pkhuong: :'( working hard to do it wrong.
2014-09-12T15:15:36Z scymtym: also, should i carefully inquire with our management whether hosting an SBCL is a possibility (would be here: http://www.campus-bielefeld.de/en/university-of-bielefeld/forschungsbau-interaktive-intelligente-systeme/)
2014-09-12T15:15:47Z Krystof: towards the end of next week might be good for me -- Thursday, for example
2014-09-12T15:16:24Z scymtym: s/an SBCL/an SBCL meeting/
2014-09-12T15:16:25Z Krystof: scymtym: I think there would be no harm in asking.  It's all very preliminary and hypothetical at the moment
2014-09-12T15:16:27Z stassats joined #sbcl
2014-09-12T15:16:40Z scymtym: sure
2014-09-12T15:17:08Z scymtym: pkhuong: we are doing it wrong?
2014-09-12T15:17:46Z |3b|: macros can't expand into declarations
2014-09-12T15:17:53Z stassats: i thought it was about the walker, ^
2014-09-12T15:18:06Z scymtym: oh, i see, sorry
2014-09-12T15:18:12Z Krystof: oh, it didn't even _occur_ to me that pkhuong could be disparaging our specializers :)
2014-09-12T15:18:35Z Krystof: now, let's see who's awake on a Friday afternoon
2014-09-12T15:19:39Z stassats relocated to a bar, so, awake, but not for long
2014-09-12T15:19:46Z Krystof: [ clearly not me because I'm having difficulty constructing the example ]
2014-09-12T15:22:20Z Krystof: (macrolet ((x () ''y)) (macrolet ((z () (x))) (macrolet ((frob (form &environment env))) `',(macroexpand-1 form env))) (frob (z)))
2014-09-12T15:22:50Z Krystof: d'ooh
2014-09-12T15:22:59Z Krystof: (macrolet ((x () ''y)) (macrolet ((z () (x))) (macrolet ((frob (form
2014-09-12T15:23:14Z Krystof:           &environment env) `',(macroexpand-1 form env))) (frob (z)))))
2014-09-12T15:23:16Z stassats: speaking of macrolet, sometimes macrolet is broken in some stage of the compiler
2014-09-12T15:23:22Z stassats: don't remember which and how
2014-09-12T15:23:37Z pkhuong: flet confusion?
2014-09-12T15:23:48Z Krystof: right, I think it's right now
2014-09-12T15:24:13Z stassats: something with a macro expanding into a macrolet, i think, or some other top-level macrolet
2014-09-12T15:24:23Z stassats: it was alright in the host, but gets lost in the sb-xc
2014-09-12T15:24:26Z Krystof: in my latest tower of macrolets, should it return a. (X); b. Y; c. something else
2014-09-12T15:26:57Z Krystof: d. unspecified ; e. depends on compiled/interpreted evaluation strategy
2014-09-12T15:27:11Z Krystof: (but specified if compiled)
2014-09-12T15:27:11Z |3b|: (x) sounds wrong, since it is evaluated in the body of Z
2014-09-12T15:27:33Z |3b|: clhs macroexpand-1
2014-09-12T15:27:33Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/f_mexp_.htm
2014-09-12T15:28:10Z stassats: all other implementations seem to return Y
2014-09-12T15:28:12Z stassats: fwiw
2014-09-12T15:28:37Z oleo is now known as Guest16902
2014-09-12T15:28:46Z stassats: i would think it should too
2014-09-12T15:29:16Z Krystof: I think you're right
2014-09-12T15:29:21Z Krystof: guess what the PCL code walker does?
2014-09-12T15:29:29Z oleo__ joined #sbcl
2014-09-12T15:29:31Z stassats: per |3b| , Z is just 'y, then it just macroexpands (z) in the environment
2014-09-12T15:32:20Z Guest16902 quit (Ping timeout: 272 seconds)
2014-09-12T15:37:17Z stassats quit (Ping timeout: 252 seconds)
2014-09-12T15:38:07Z attila_lendvai quit (Quit: Leaving.)
2014-09-12T15:40:29Z krzysz00 quit (Ping timeout: 273 seconds)
2014-09-12T15:45:17Z dkcl joined #sbcl
2014-09-12T15:58:32Z krzysz00 joined #sbcl
2014-09-12T16:08:12Z Krystof: huh, ok, what about this case?
2014-09-12T16:08:36Z Krystof: (locally (declare (special y)) (macrolet ((z () 'y)) (macrolet ((x () (z))))))
2014-09-12T16:10:09Z Krystof: is that a. undefined behaviour ; b. totally specified behaviour ; c. something else?  (Apart from the fact that it's nothing at all: the question is whether the lookup of Y in X is well-defined or not)
2014-09-12T16:11:42Z |3b|: hmm, you mean whether LOCALLY affects the variable in the body of a macrolet?
2014-09-12T16:11:47Z Krystof: yes
2014-09-12T16:12:23Z Krystof: well, actually it _has_ to affect it at least in part
2014-09-12T16:12:24Z Bike: i don't see why it wouldn't, the declare is part of the lexical environment, it's not like an actual binding
2014-09-12T16:13:23Z Krystof: (symbol-macrolet ((x 3)) (locally (declare (special x)) (macrolet ((y () x)) (y)))) ; had better not return 3
2014-09-12T16:13:39Z |3b|: Bike: isn't y evaluated during macroexpansion though, not in the lexical environment?
2014-09-12T16:13:59Z Krystof: CLHS MACROLET says "Declarations and macrolet and symbol-macrolet definitions affect the local macro definitions in a macrolet"
2014-09-12T16:14:04Z Bike: yeah, but i think the macro function bodies are affected by outside declarations
2014-09-12T16:14:42Z Krystof: gaaah "had better not return 3" -- what does SBCL do?
2014-09-12T16:14:43Z Bike: well, that form returns 3 in my sbcl. oops?
2014-09-12T16:14:57Z Krystof: yeah.  So maybe I am wrong, or maybe this is horrible
2014-09-12T16:16:11Z |3b|: hmm, yeah... macro-expansion functions ... are defined in the lexical environment"
2014-09-12T16:17:20Z |3b|: and the declaration would mask any local binding in the lexical environment
2014-09-12T16:17:49Z Bike: yeah. (symbol-macrolet ((x 3)) (locally (declare (special x)) x)) looks up the special binding, of course.
2014-09-12T16:19:52Z |3b| leans towards B, specified behavior
2014-09-12T16:22:55Z stassats joined #sbcl
2014-09-12T16:25:04Z psilord quit (Quit: Leaving.)
2014-09-12T16:28:27Z scymtym: how would you feel about something like http://paste.lisp.org/display/143697 to better support user-defined lazy sequences?
2014-09-12T16:28:53Z scymtym: (assuming a similar treatment of e.g. CONCATENATE)
2014-09-12T16:31:44Z krzysz00 quit (Ping timeout: 276 seconds)
2014-09-12T16:36:30Z krzysz00 joined #sbcl
2014-09-12T16:39:47Z hlavaty left #sbcl
2014-09-12T16:40:57Z jdz joined #sbcl
2014-09-12T16:43:07Z stassats quit (Ping timeout: 245 seconds)
2014-09-12T16:46:32Z ehaliewicz joined #sbcl
2014-09-12T16:53:30Z ehaliewicz quit (Remote host closed the connection)
2014-09-12T16:54:25Z krzysz00 quit (Ping timeout: 260 seconds)
2014-09-12T16:55:09Z ehaliewicz joined #sbcl
2014-09-12T16:58:18Z sdemarre joined #sbcl
2014-09-12T17:02:51Z sdemarre quit (Ping timeout: 255 seconds)
2014-09-12T17:06:00Z slyrus joined #sbcl
2014-09-12T17:15:18Z faheem__1 joined #sbcl
2014-09-12T17:15:47Z milosn_ joined #sbcl
2014-09-12T17:18:39Z milosn quit (Write error: Broken pipe)
2014-09-12T17:18:39Z faheem_ quit (Write error: Broken pipe)
2014-09-12T17:19:15Z wmarvel quit (Remote host closed the connection)
2014-09-12T17:19:24Z wmarvel joined #sbcl
2014-09-12T17:21:17Z sdemarre joined #sbcl
2014-09-12T17:30:37Z gingerale quit (Read error: Connection reset by peer)
2014-09-12T17:34:21Z slyrus quit (Ping timeout: 255 seconds)
2014-09-12T17:59:21Z krzysz00 joined #sbcl
2014-09-12T18:06:45Z pchrist quit (Ping timeout: 240 seconds)
2014-09-12T18:08:18Z pchrist joined #sbcl
2014-09-12T18:20:15Z krzysz00 quit (Ping timeout: 255 seconds)
2014-09-12T18:29:22Z sdemarre quit (Ping timeout: 240 seconds)
2014-09-12T18:32:12Z eudoxia quit (Quit: Lost terminal)
2014-09-12T18:51:20Z Krystof: scymtym: possible.  I'm a bit unsure about (class-name class) in the interface
2014-09-12T18:52:34Z scymtym: maybe just the prototype?
2014-09-12T18:54:07Z scymtym: as in (defgeneric sb-sequence:map (result-prototype function sequence &rest sequences))
2014-09-12T19:10:01Z sdemarre joined #sbcl
2014-09-12T19:10:38Z psilord joined #sbcl
2014-09-12T19:17:52Z ehaliewicz quit (Ping timeout: 240 seconds)
2014-09-12T19:31:17Z slyrus joined #sbcl
2014-09-12T19:47:12Z psilord quit (Quit: Leaving.)
2014-09-12T20:12:18Z slyrus quit (Ping timeout: 255 seconds)
2014-09-12T20:24:23Z krzysz00 joined #sbcl
2014-09-12T20:43:34Z sdemarre quit (Ping timeout: 252 seconds)
2014-09-12T20:47:23Z fridim_ quit (Ping timeout: 240 seconds)
2014-09-12T20:55:05Z slyrus joined #sbcl
2014-09-12T21:05:02Z drmeister quit (Remote host closed the connection)
2014-09-12T21:26:07Z drmeister joined #sbcl
2014-09-12T21:26:39Z oleo__ is now known as oleo
2014-09-12T21:33:22Z Bicyclidine joined #sbcl
2014-09-12T21:39:11Z kanru quit (Ping timeout: 276 seconds)
2014-09-12T21:39:30Z DGASAU` joined #sbcl
2014-09-12T21:41:59Z nyef_ joined #sbcl
2014-09-12T21:43:10Z nyef quit (Ping timeout: 276 seconds)
2014-09-12T21:43:11Z DGASAU quit (Ping timeout: 276 seconds)
2014-09-12T21:43:57Z ^Posterdati^ quit (Ping timeout: 268 seconds)
2014-09-12T21:43:58Z asedeno_ quit (Ping timeout: 268 seconds)
2014-09-12T21:45:50Z zacts quit (Remote host closed the connection)
2014-09-12T21:45:51Z zacts_ joined #sbcl
2014-09-12T21:45:57Z asedeno joined #sbcl
2014-09-12T21:46:25Z carvite quit (Ping timeout: 268 seconds)
2014-09-12T21:46:31Z DGASAU` is now known as DGASAU
2014-09-12T21:46:50Z ^Posterdati^ joined #sbcl
2014-09-12T21:47:03Z carvite joined #sbcl
2014-09-12T21:51:42Z zacts_ quit (Ping timeout: 272 seconds)
2014-09-12T21:52:31Z zacts joined #sbcl
2014-09-12T21:57:27Z slyrus_ joined #sbcl
2014-09-12T21:58:41Z zacts quit (Ping timeout: 260 seconds)
2014-09-12T21:59:19Z zacts joined #sbcl
2014-09-12T21:59:22Z slyrus quit (Ping timeout: 268 seconds)
2014-09-12T21:59:36Z slyrus_ is now known as slyrus
2014-09-12T22:00:10Z drmeister quit (Read error: Connection reset by peer)
2014-09-12T22:01:06Z Bicyclidine quit (Quit: Reconnecting)
2014-09-12T22:01:19Z Bicyclidine joined #sbcl
2014-09-12T22:02:27Z drmeister joined #sbcl
2014-09-12T22:03:33Z drmeister quit (Read error: Connection reset by peer)
2014-09-12T22:03:52Z zacts quit (Ping timeout: 252 seconds)
2014-09-12T22:05:02Z zacts joined #sbcl
2014-09-12T22:05:33Z oleo quit (Quit: Verlassend)
2014-09-12T22:33:06Z ccl-logbot joined #sbcl
2014-09-12T22:33:06Z 
2014-09-12T22:33:06Z names: ccl-logbot tmh_ jdz
2014-09-12T22:38:37Z ccl-logbot joined #sbcl
2014-09-12T22:38:37Z 
2014-09-12T22:38:37Z names: ccl-logbot thoto angavrilov_ scymtym milosn jackdaniel wmarvel_ irsol_ les_ zacts krzysz00 logand` yauz p_l faheem_ drmeister Bicyclidine asedeno nyef_ DGASAU yacks segv- BitPuffin ivan`` prxq christoph_debian akkad alchemis7 pkhuong White_Flame Perlboy Blkt fe[nl]ix scymtym__ brucem luis sobel heddwch leoc Intensity flip214 phf ferada loke redline6561 jsnell foom |3b| antoszka cracauer_ specbot minion mood p_l|backup
2014-09-12T22:39:48Z Skrylar joined #sbcl
2014-09-12T22:40:54Z Bike joined #sbcl
2014-09-12T22:41:13Z edgar-rft joined #sbcl
2014-09-12T22:41:59Z asedeno quit (Ping timeout: 246 seconds)
2014-09-12T22:41:59Z brucem quit (Ping timeout: 246 seconds)
2014-09-12T22:42:04Z prxq_ joined #sbcl
2014-09-12T22:42:32Z prxq quit (Ping timeout: 246 seconds)
2014-09-12T22:42:34Z DGASAU` joined #sbcl
2014-09-12T22:42:37Z scymtym_ joined #sbcl
2014-09-12T22:42:52Z nyef_ quit (Ping timeout: 246 seconds)
2014-09-12T22:42:52Z Intensity quit (Ping timeout: 246 seconds)
2014-09-12T22:43:54Z DGASAU quit (Ping timeout: 246 seconds)
2014-09-12T22:43:56Z angavrilov_ quit (Remote host closed the connection)
2014-09-12T22:44:17Z segv- quit (Ping timeout: 246 seconds)
2014-09-12T22:45:48Z irsol_ quit (Changing host)
2014-09-12T22:45:48Z irsol_ joined #sbcl
2014-09-12T22:45:50Z irsol_ is now known as irsol
2014-09-12T22:46:10Z pchrist joined #sbcl
2014-09-12T22:46:12Z DGASAU` is now known as DGASAU
2014-09-12T22:48:12Z Posterdati joined #sbcl
2014-09-12T22:48:14Z ams joined #sbcl
2014-09-12T22:48:30Z zacts quit (Ping timeout: 276 seconds)
2014-09-12T22:48:30Z krzysz00 quit (Ping timeout: 276 seconds)
2014-09-12T22:48:30Z scymtym quit (Ping timeout: 276 seconds)
2014-09-12T22:48:30Z logand` quit (Ping timeout: 276 seconds)
2014-09-12T22:49:22Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-12T22:50:44Z oleo joined #sbcl
2014-09-12T22:51:20Z carvite joined #sbcl
2014-09-12T22:55:01Z tmh_ joined #sbcl
2014-09-12T22:55:50Z jdz joined #sbcl
2014-09-12T22:58:12Z qbit joined #sbcl
2014-09-12T22:58:38Z qbit is now known as Guest53819
2014-09-12T23:01:10Z reb` joined #sbcl
2014-09-12T23:01:11Z Guest53819 is now known as qbit
2014-09-12T23:04:31Z slyrus joined #sbcl
2014-09-12T23:10:26Z nicdev joined #sbcl
2014-09-12T23:11:25Z faheem_ quit (Ping timeout: 252 seconds)
2014-09-12T23:12:47Z slyrus quit (Ping timeout: 276 seconds)
2014-09-12T23:14:03Z scymtym_ quit (Remote host closed the connection)
2014-09-12T23:15:31Z slyrus joined #sbcl
2014-09-12T23:16:51Z Posterdati quit (Ping timeout: 252 seconds)
2014-09-12T23:16:52Z yauz quit (Ping timeout: 252 seconds)
2014-09-12T23:20:07Z oleo quit (Ping timeout: 252 seconds)
2014-09-12T23:22:13Z jackdaniel quit (Ping timeout: 252 seconds)
2014-09-12T23:22:13Z slyrus quit (Ping timeout: 252 seconds)
2014-09-12T23:22:22Z Posterdati joined #sbcl
2014-09-12T23:23:01Z yauz joined #sbcl
2014-09-12T23:24:10Z DGASAU quit (Ping timeout: 252 seconds)
2014-09-12T23:28:15Z asedeno joined #sbcl
2014-09-12T23:30:57Z Bicyclidine joined #sbcl
2014-09-12T23:38:18Z slyrus joined #sbcl
2014-09-12T23:38:20Z zacts joined #sbcl
2014-09-12T23:38:20Z slyrus quit (Ping timeout: 268 seconds)
2014-09-12T23:38:21Z zacts quit (Client Quit)
2014-09-12T23:38:21Z brucem joined #sbcl
2014-09-12T23:38:21Z slyrus joined #sbcl
2014-09-12T23:38:29Z 6JTAAFVUB joined #sbcl
2014-09-12T23:38:29Z slyrus quit (Ping timeout: 268 seconds)
2014-09-12T23:38:29Z jackdaniel joined #sbcl
2014-09-12T23:39:38Z jdz is now known as Guest13222
2014-09-12T23:52:03Z DGASAU joined #sbcl
2014-09-12T23:54:15Z oleo joined #sbcl
2014-09-12T23:54:15Z nyef joined #sbcl
2014-09-12T23:54:15Z krzysz00 joined #sbcl
2014-09-12T23:54:39Z oleo quit (Max SendQ exceeded)
2014-09-12T23:56:11Z oleo joined #sbcl
2014-09-13T00:00:05Z reb` quit (Changing host)
2014-09-13T00:00:05Z reb` joined #sbcl
2014-09-13T00:05:58Z Bicyclidine quit (Ping timeout: 252 seconds)
2014-09-13T00:14:03Z brucem quit (Changing host)
2014-09-13T00:14:03Z brucem joined #sbcl
2014-09-13T00:18:08Z nicdev` joined #sbcl
2014-09-13T00:20:13Z carvite_ joined #sbcl
2014-09-13T00:21:31Z ams quit (Ping timeout: 272 seconds)
2014-09-13T00:21:31Z nicdev quit (Ping timeout: 272 seconds)
2014-09-13T00:21:31Z pchrist quit (Ping timeout: 272 seconds)
2014-09-13T00:21:31Z thoto quit (Ping timeout: 272 seconds)
2014-09-13T00:21:31Z scymtym__ quit (Ping timeout: 272 seconds)
2014-09-13T00:21:34Z carvite quit (Ping timeout: 272 seconds)
2014-09-13T00:21:37Z thoto joined #sbcl
2014-09-13T00:22:37Z pchrist joined #sbcl
2014-09-13T00:30:52Z carvite_ quit (Changing host)
2014-09-13T00:30:52Z carvite_ joined #sbcl
2014-09-13T00:31:02Z carvite_ is now known as carvite
2014-09-13T00:37:57Z thoto_ joined #sbcl
2014-09-13T00:45:17Z thoto quit (*.net *.split)
2014-09-13T00:55:23Z carvite quit (Write error: Broken pipe)
2014-09-13T00:55:23Z nicdev` quit (Write error: Broken pipe)
2014-09-13T00:55:24Z thoto_ quit (Remote host closed the connection)
2014-09-13T00:55:34Z carvite_ joined #sbcl
2014-09-13T00:55:58Z carvite_ is now known as carvite
2014-09-13T00:58:10Z asedeno_ joined #sbcl
2014-09-13T01:00:51Z thoto joined #sbcl
2014-09-13T01:02:59Z faheem_ joined #sbcl
2014-09-13T01:06:49Z nyef quit (Ping timeout: 245 seconds)
2014-09-13T01:06:52Z krzysz00 quit (*.net *.split)
2014-09-13T01:07:01Z asedeno quit (Ping timeout: 268 seconds)
2014-09-13T01:07:02Z tmh_ quit (Ping timeout: 268 seconds)
2014-09-13T01:07:06Z Guest13222 quit (Ping timeout: 268 seconds)
2014-09-13T01:07:06Z jackdaniel quit (Ping timeout: 268 seconds)
2014-09-13T01:07:06Z 6JTAAFVUB quit (Ping timeout: 268 seconds)
2014-09-13T01:07:06Z jdz joined #sbcl
2014-09-13T01:07:06Z pchrist quit (Ping timeout: 268 seconds)
2014-09-13T01:07:06Z brucem quit (Ping timeout: 268 seconds)
2014-09-13T01:07:06Z tmh_ joined #sbcl
2014-09-13T01:07:06Z pchrist joined #sbcl
2014-09-13T01:07:07Z pchrist quit (Changing host)
2014-09-13T01:07:07Z pchrist joined #sbcl
2014-09-13T01:07:08Z tmh_ quit (Read error: Connection reset by peer)
2014-09-13T01:07:14Z segv- joined #sbcl
2014-09-13T01:07:30Z jackdani1l joined #sbcl
2014-09-13T01:07:33Z tmh_ joined #sbcl
2014-09-13T01:08:03Z brucem joined #sbcl
2014-09-13T01:08:13Z jdz is now known as Guest22814
2014-09-13T01:15:35Z jackdaniel joined #sbcl
2014-09-13T01:15:42Z prxq_ quit (Ping timeout: 245 seconds)
2014-09-13T01:20:09Z edgar-rft quit (Quit: lifetime closed by computer contermination)
2014-09-13T01:20:34Z jackdani1l quit (*.net *.split)
2014-09-13T01:20:35Z segv- quit (*.net *.split)
2014-09-13T01:26:44Z DGASAU` joined #sbcl
2014-09-13T01:26:55Z reb`` joined #sbcl
2014-09-13T01:28:20Z segv- joined #sbcl
2014-09-13T01:31:59Z ^Posterdati^ joined #sbcl
2014-09-13T01:32:11Z yauz quit (Ping timeout: 276 seconds)
2014-09-13T01:32:15Z DGASAU quit (Ping timeout: 276 seconds)
2014-09-13T01:32:15Z Posterdati quit (Ping timeout: 276 seconds)
2014-09-13T01:32:19Z reb` quit (Ping timeout: 276 seconds)
2014-09-13T01:36:44Z brucem quit (Changing host)
2014-09-13T01:36:45Z brucem joined #sbcl
2014-09-13T02:25:49Z Bike quit (Quit: restarting)
2014-09-13T02:38:41Z christoph_debian quit (Ping timeout: 260 seconds)
2014-09-13T02:43:18Z Bike joined #sbcl
2014-09-13T02:45:25Z Intensity joined #sbcl
2014-09-13T02:51:47Z christoph_debian joined #sbcl
2014-09-13T03:02:56Z zacts joined #sbcl
2014-09-13T04:02:08Z oleo is now known as Guest43245
2014-09-13T04:03:05Z oleo__ joined #sbcl
2014-09-13T04:05:02Z Guest43245 quit (Ping timeout: 252 seconds)
2014-09-13T04:24:09Z BitPuffin quit (Ping timeout: 260 seconds)
2014-09-13T04:55:39Z sdemarre joined #sbcl
2014-09-13T04:58:32Z yauz joined #sbcl
2014-09-13T04:58:32Z krzysz00 joined #sbcl
2014-09-13T04:58:32Z prxq joined #sbcl
2014-09-13T05:04:15Z reb`` quit (Changing host)
2014-09-13T05:04:15Z reb`` joined #sbcl
2014-09-13T06:01:53Z slyrus joined #sbcl
2014-09-13T06:36:33Z ams joined #sbcl
2014-09-13T06:56:55Z fridim_ joined #sbcl
2014-09-13T07:04:20Z Krystof joined #sbcl
2014-09-13T07:04:20Z ChanServ has set mode +o Krystof
2014-09-13T07:20:56Z ^Posterdati^ quit (Quit: KVIrc 4.1.3 Equilibrium http://www.kvirc.net/)
2014-09-13T07:21:03Z Posterdati joined #sbcl
2014-09-13T07:42:06Z leo2007 joined #sbcl
2014-09-13T08:11:04Z fridim_ quit (Ping timeout: 252 seconds)
2014-09-13T08:12:00Z logand joined #sbcl
2014-09-13T08:24:12Z Posterdati quit (Ping timeout: 252 seconds)
2014-09-13T08:24:12Z faheem_ quit (Ping timeout: 252 seconds)
2014-09-13T08:24:17Z Posterdati joined #sbcl
2014-09-13T08:24:25Z faheem_ joined #sbcl
2014-09-13T08:24:49Z qbit quit (Ping timeout: 252 seconds)
2014-09-13T08:27:50Z qbit joined #sbcl
2014-09-13T08:27:55Z logand quit (Read error: Connection reset by peer)
2014-09-13T08:28:01Z logand joined #sbcl
2014-09-13T08:28:14Z qbit is now known as Guest76673
2014-09-13T09:03:05Z dkcl joined #sbcl
2014-09-13T09:04:45Z stassats joined #sbcl
2014-09-13T09:06:31Z angavrilov joined #sbcl
2014-09-13T09:11:27Z sdemarre quit (Ping timeout: 246 seconds)
2014-09-13T09:37:04Z oleo__ quit (Quit: Verlassend)
2014-09-13T09:44:04Z PuercoPop joined #sbcl
2014-09-13T09:57:08Z sdemarre joined #sbcl
2014-09-13T09:57:15Z oleo joined #sbcl
2014-09-13T10:06:02Z leo2007 quit (Ping timeout: 245 seconds)
2014-09-13T10:19:22Z DGASAU` quit (Ping timeout: 268 seconds)
2014-09-13T10:19:59Z DGASAU` joined #sbcl
2014-09-13T10:24:47Z krzysz00 quit (Ping timeout: 245 seconds)
2014-09-13T10:25:35Z krzysz00 joined #sbcl
2014-09-13T10:33:19Z edgar-rft joined #sbcl
2014-09-13T10:36:46Z BitPuffin joined #sbcl
2014-09-13T10:53:41Z scymtym joined #sbcl
2014-09-13T11:28:05Z krzysz00_ joined #sbcl
2014-09-13T11:29:38Z krzysz00 quit (Ping timeout: 252 seconds)
2014-09-13T11:47:34Z leo2007 joined #sbcl
2014-09-13T11:48:02Z hlavaty joined #sbcl
2014-09-13T11:49:52Z fridim_ joined #sbcl
2014-09-13T12:05:51Z fridim_ quit (Read error: Connection reset by peer)
2014-09-13T12:44:00Z stassats quit (Read error: No route to host)
2014-09-13T13:02:24Z LiamH joined #sbcl
2014-09-13T13:09:16Z stassats joined #sbcl
2014-09-13T13:27:07Z segv- quit (Ping timeout: 272 seconds)
2014-09-13T13:34:46Z nyef joined #sbcl
2014-09-13T13:35:06Z nyef: G'morning all.
2014-09-13T13:37:24Z oleo: morning nyef
2014-09-13T13:55:25Z ivan`` quit (*.net *.split)
2014-09-13T13:55:29Z redline6561 quit (*.net *.split)
2014-09-13T13:55:29Z foom quit (*.net *.split)
2014-09-13T13:55:29Z |3b| quit (*.net *.split)
2014-09-13T13:55:29Z antoszka quit (*.net *.split)
2014-09-13T13:55:29Z PuercoPop quit (*.net *.split)
2014-09-13T13:55:29Z cracauer_ quit (*.net *.split)
2014-09-13T13:55:33Z PuercoPop joined #sbcl
2014-09-13T13:55:33Z antoszka joined #sbcl
2014-09-13T13:55:41Z redline6561 joined #sbcl
2014-09-13T13:55:46Z Krystof quit (*.net *.split)
2014-09-13T13:55:49Z Bike quit (*.net *.split)
2014-09-13T13:55:49Z zacts quit (*.net *.split)
2014-09-13T13:55:49Z reb`` quit (*.net *.split)
2014-09-13T13:55:49Z jackdaniel quit (*.net *.split)
2014-09-13T13:55:50Z brucem quit (*.net *.split)
2014-09-13T13:55:50Z tmh_ quit (*.net *.split)
2014-09-13T13:55:50Z pchrist quit (*.net *.split)
2014-09-13T13:55:51Z Guest22814 quit (*.net *.split)
2014-09-13T13:55:53Z foom joined #sbcl
2014-09-13T13:55:53Z cracauer_ joined #sbcl
2014-09-13T13:55:53Z jackdaniel joined #sbcl
2014-09-13T13:55:57Z Bike joined #sbcl
2014-09-13T13:56:00Z reb`` joined #sbcl
2014-09-13T13:56:08Z jdz joined #sbcl
2014-09-13T13:56:09Z zacts joined #sbcl
2014-09-13T13:56:16Z tmh_ joined #sbcl
2014-09-13T13:56:35Z pchrist joined #sbcl
2014-09-13T13:56:51Z brucem joined #sbcl
2014-09-13T13:56:51Z brucem quit (Changing host)
2014-09-13T13:56:52Z brucem joined #sbcl
2014-09-13T13:57:38Z |3b|`` joined #sbcl
2014-09-13T13:58:57Z |3b|`` is now known as |3b|
2014-09-13T14:11:29Z nyef: Okay, has anyone had a chance to review cracauer_ 's GC patches yet?
2014-09-13T14:17:09Z nyef: It looks like there's a lot of debugging noise in here, and a fundamental confusion between conservative roots and pinned objects.
2014-09-13T14:26:02Z sdemarre quit (*.net *.split)
2014-09-13T14:26:02Z hlavaty quit (*.net *.split)
2014-09-13T14:26:02Z Guest76673 quit (*.net *.split)
2014-09-13T14:26:02Z faheem_ quit (*.net *.split)
2014-09-13T14:26:03Z slyrus quit (*.net *.split)
2014-09-13T14:26:07Z christoph_debian quit (*.net *.split)
2014-09-13T14:26:18Z hlavaty joined #sbcl
2014-09-13T14:26:20Z qbit joined #sbcl
2014-09-13T14:26:40Z christoph_debian joined #sbcl
2014-09-13T14:26:44Z qbit is now known as Guest17545
2014-09-13T14:27:08Z sdemarre joined #sbcl
2014-09-13T14:28:39Z slyrus joined #sbcl
2014-09-13T14:31:20Z faheem_ joined #sbcl
2014-09-13T14:50:52Z nyef: Hrm. And confusion between the idea of "dontmove" (which is a constraint) and "pinned object" (which is WHAT shouldn't be moved).
2014-09-13T14:55:12Z faheem_: what is the current official sbcl git repos? git://sbcl.git.sourceforge.net/gitroot/sbcl/sbcl seems quite unreliable. half the time it says there is nothing there
2014-09-13T14:55:35Z faheem_: should i use https://github.com/sbcl/sbcl instead?
2014-09-13T14:55:49Z nyef: cracauer_: Why evacuate pinned pages at all, if you're scrubbing the dead objects?
2014-09-13T14:56:28Z nyef: faheem_: AFAIK, sourceforge is still the official repository.
2014-09-13T14:57:19Z faheem_: nyef: ok. but pulling github is just as good?
2014-09-13T14:57:31Z nyef: I have no idea.
2014-09-13T14:57:36Z faheem_: nyef: ok
2014-09-13T14:58:12Z jsnell: nyef: doesn't everything on a pinned page get treated as a root?
2014-09-13T14:58:51Z nyef: jsnell: This is in the context of cracauer_ 's GC changes, where only the actually-pinned objects are treated as roots.
2014-09-13T15:01:32Z nyef: There's an amazing amount of noise, and some actual damage in the patchset, so I'm trying to figure out which bits are which, and what the actual changes do and if they make sense.
2014-09-13T15:17:11Z nyef: Oh, god. code_page_p(). Did we ever get that sorted out, or is it still a disaster of a concept?
2014-09-13T15:24:15Z zacts: is there anything that SBCL can't do on OpenBSD that it can on other OS?
2014-09-13T15:24:53Z nyef: ... Threads, maybe?
2014-09-13T15:25:06Z nyef: FP signal handling on PPC, IIRC.
2014-09-13T15:26:04Z nyef: ... Wha? If scavtab says the size is 1 and it's pointing to a fixnum, it's pointing to a pointer, or it's pointing to something with a lowtag of 9, then size is 2? WTF?
2014-09-13T15:26:30Z nyef: Err... sizetab, not scavtab, but the horror remains.
2014-09-13T15:28:01Z nyef: OTHER_IMMEDIATE_2_LOWTAG? /seriously/?
2014-09-13T15:36:43Z nyef: And of course there's prior art in gc_search_space() for precisely what this little horror is attempting to do.
2014-09-13T15:50:40Z slyrus quit (Ping timeout: 250 seconds)
2014-09-13T16:11:06Z nyef: cracauer_: Next question, what's a "consi" page?
2014-09-13T16:35:07Z drmeister quit (Read error: Connection reset by peer)
2014-09-13T16:35:18Z drmeister joined #sbcl
2014-09-13T16:35:21Z nyef: Hrm. Okay, from_space_p() is no longer a strictly correct name.  Or, possibly, "from space" isn't a strictly applicable concept anymore.
2014-09-13T16:49:09Z nyef: The latter, which implies the former. What we have here is a partial mark/sweep system, which has no from-space. And the question isn't if the object is in from_space, the question is if it's /evacuatable/.
2014-09-13T16:49:37Z nyef: For a two-space collector, everything in from-space is evacuatable.
2014-09-13T16:49:47Z nyef: (And must, in fact, be evacuated.)
2014-09-13T16:50:17Z nyef: For a mark/sweep system, nothing is evacuatable, because objects are never moved.
2014-09-13T16:57:32Z dandersen joined #sbcl
2014-09-13T17:00:06Z dkcl quit (Ping timeout: 246 seconds)
2014-09-13T17:00:41Z oleo is now known as Guest66776
2014-09-13T17:01:33Z oleo__ joined #sbcl
2014-09-13T17:02:21Z nyef: pkhuong: Ping?
2014-09-13T17:03:23Z nyef: pkhuong: Did you want a mark/sweep option for gencgc?
2014-09-13T17:03:53Z Guest66776 quit (Ping timeout: 260 seconds)
2014-09-13T17:04:00Z nyef: Because if so, it looks like cracauer_ did a good chunk of the initial work for us.
2014-09-13T17:09:31Z nyef: ... aww. trace-table went away? /-:
2014-09-13T17:10:45Z dandersen quit (Changing host)
2014-09-13T17:10:45Z dandersen joined #sbcl
2014-09-13T17:11:49Z dandersen is now known as dkcl
2014-09-13T17:21:37Z dkcl quit (Remote host closed the connection)
2014-09-13T17:22:05Z dkcl joined #sbcl
2014-09-13T17:32:12Z ivan`` joined #sbcl
2014-09-13T17:38:37Z nyef: ... The wipe_with value is 0x00424242? Isn't that an other-immediate on 32-bit targets?
2014-09-13T17:40:17Z dandersen joined #sbcl
2014-09-13T17:43:01Z dkcl quit (Ping timeout: 246 seconds)
2014-09-13T17:46:28Z dandersen is now known as dickle
2014-09-13T17:49:38Z nyef: Ah, probably value-cell-header-widetag.
2014-09-13T17:59:46Z pkhuong: nyef: I'd expect that to happen :) we'll see if it's easier to have a mark/sweep with compaction than finer grained pinning
2014-09-13T18:01:00Z stassats quit (Ping timeout: 246 seconds)
2014-09-13T18:08:01Z dickle is now known as earth
2014-09-13T18:08:13Z earth is now known as dkcl
2014-09-13T18:31:06Z dkcl quit (Ping timeout: 246 seconds)
2014-09-13T19:10:41Z scymtym: improved version of SB-SEQUENCE:{MAP,CONCATENATE,MERGE} patch set: http://paste.lisp.org/display/143705
2014-09-13T19:34:28Z oleo__ is now known as oleo
2014-09-13T19:45:24Z nyef: pkhuong: Does "evacuable_object_p" as a replacement for "from_space_p" sound right?
2014-09-13T19:45:33Z segv- joined #sbcl
2014-09-13T19:54:33Z pkhuong: more like evacuated_object_p
2014-09-13T19:54:56Z nyef: No, because it also needs to return true for objects that should be evacuated but haven't yet been.
2014-09-13T19:54:59Z pkhuong: in that it's not an option: evacuation *will* happen and you better fix up that pointer
2014-09-13T19:55:28Z nyef: (That is, it's true even before the forwarding pointer is set.)
2014-09-13T20:01:23Z pkhuong: ... we still need something like from_space_p
2014-09-13T20:01:46Z pkhuong: to denote objects that are undergoing GC, regardless of their evacuation status.
2014-09-13T20:02:03Z nyef: Oh?
2014-09-13T20:02:42Z pkhuong: there are two reasons an object wouldn't be evacuated
2014-09-13T20:02:55Z pkhuong: it might be in an older generation that we assume is 100% live
2014-09-13T20:03:06Z pkhuong: or it might be pinned in place
2014-09-13T20:03:31Z nyef: You need a concept of "this object has/has-not been scavenged" and a concept of "this object may/may-not/must be moved".
2014-09-13T20:03:43Z pkhuong: For the former, we simply disregard the pointer. For the latter, we still want to scavenge the object, without copying it
2014-09-13T20:04:35Z pkhuong: I think we can assume objects are may-not/must be moved
2014-09-13T20:04:35Z nyef: For the pinned objects, we have the markbits so that we can find the object.
2014-09-13T20:05:06Z nyef: Dead objects MAY be moved, but there's no MUST about it (we'd prefer that they not be moved).
2014-09-13T20:05:28Z pkhuong: but if we know that they're dead, we won't move them
2014-09-13T20:05:37Z pkhuong: so assuming that they're live, we'll consider them as must-move.
2014-09-13T20:05:40Z nyef: In fact, the entire point of the patches that I'm drawing from here is to NOT move certain dead objects.
2014-09-13T20:07:08Z nyef: Okay, how about this? "All live objects in from space must be evacuated to new-space" is the basic principle of a two-space collector, and gencgc currently acts this way.
2014-09-13T20:07:42Z nyef: But that "must be evacuated" is executed on a page level when there's a pinned object involved.
2014-09-13T20:08:16Z nyef: A mark/sweep collector specifically DOESN'T evacuate objects (until you start getting clever to reduce heap fragmentation).
2014-09-13T20:09:35Z nyef: And we're looking at setting up a combination of mark/sweep for pinned objects, while still evacuating non-pinned objects from the pinned page.
2014-09-13T20:10:17Z nyef: Either we're now effectively tracking from-space on a per-object granularity, or from-space isn't what we're really looking at.
2014-09-13T20:10:34Z pkhuong: I wonder if determining the mark/sweep v. copying status on a per-page granularity would make this easier
2014-09-13T20:10:46Z nyef: We're heading that way anyway.
2014-09-13T20:11:35Z pkhuong: IIUC, this patch's goal is to pin specific objects and let normal evacuation occur for other live objects in the page
2014-09-13T20:11:50Z angavrilov quit (Remote host closed the connection)
2014-09-13T20:12:02Z nyef: Yes, which is in some ways a complete waste of potential.
2014-09-13T20:12:43Z nyef: I'd love to have a tuning parameter that says "evacuate pages that are less than x% full, but leave them in-place until then".
2014-09-13T20:12:49Z pkhuong: which feels like it's paying for the complexity of per-page evacuation policy already.
2014-09-13T20:13:15Z nyef: It's not paying the full price yet.
2014-09-13T20:22:58Z nyef: Okay, how about evacuating_object_p()?
2014-09-13T20:23:49Z nyef: Oh, damn. We really DO have a three-way branch situation here. Ugh.
2014-09-13T20:25:35Z nyef: Since the main question from scavenge() is "what do I need to do about this pointer?", and the answer quickly becomes "move it and update to this", "update to this", "mark it as live", or "leave it alone". Guess that's four-way.
2014-09-13T20:25:58Z pkhuong: the first two are kind of the same
2014-09-13T20:27:11Z nyef: Yes and no. We're going to end up with the option of evacuating and releasing the page before updating inbound pointers.
2014-09-13T20:32:06Z nyef: I guess I'm just going to have to move on and come back to this bit. /-:
2014-09-13T20:50:26Z sdemarre quit (Ping timeout: 252 seconds)
2014-09-13T21:16:09Z Bike quit (Read error: Connection reset by peer)
2014-09-13T21:18:26Z pkhuong: issue with malloc during GC: we might interrupt a thread in foreign code that's in the middle of malloc/free and deadlock on some internal lock
2014-09-13T21:18:46Z nyef: Lovely.
2014-09-13T21:19:06Z nyef: You know, I've always wanted to be able to GC without stopping threads in foreign code.
2014-09-13T21:19:18Z pkhuong: that's called sb-safepoint ;)
2014-09-13T21:19:26Z nyef: No, even without that.
2014-09-13T21:25:36Z Bike joined #sbcl
2014-09-13T21:39:01Z hlavaty` joined #sbcl
2014-09-13T21:39:41Z les joined #sbcl
2014-09-13T21:39:41Z les quit (Changing host)
2014-09-13T21:39:41Z les joined #sbcl
2014-09-13T21:40:32Z asedeno joined #sbcl
2014-09-13T21:45:17Z faheem_ quit (*.net *.split)
2014-09-13T21:45:17Z hlavaty quit (*.net *.split)
2014-09-13T21:45:20Z asedeno_ quit (*.net *.split)
2014-09-13T21:45:21Z les_ quit (*.net *.split)
2014-09-13T21:46:11Z faheem_ joined #sbcl
2014-09-13T21:47:03Z pkhuong: ok. I think I see a way out
2014-09-13T21:47:33Z pkhuong: but we'd end up ripping out all but gc-common and the page allocation/deallocation logic
2014-09-13T21:47:46Z nyef: Oh?
2014-09-13T21:48:29Z pkhuong: minor GC:
2014-09-13T21:49:30Z pkhuong: assign copy vs mark policy to each page in the nursery
2014-09-13T21:49:55Z pkhuong: write barriers mean we know the (super)set of pages with refs to the nursery
2014-09-13T21:50:43Z pkhuong: walk all the roots + our "remembered" set and find pointers to the nursery
2014-09-13T21:51:22Z pkhuong: if it's on a copy page, look for a forwarding pointer, otherwise evict and overwrite with a forwarding pointer
2014-09-13T21:52:31Z pkhuong: if it's on a mark page, ensure that the "live" and "to scan" bitmaps are allocated, and set the right bits to 1
2014-09-13T21:53:03Z pkhuong: the only remaining part is maintaining a set of pages to scan for pointers
2014-09-13T21:53:31Z pkhuong: cheney makes this easy by letting newspace = worklist, but that already makes things weird in gengc.
2014-09-13T21:54:28Z nyef: This looks rather similar to what I had in mind, yes.
2014-09-13T21:54:53Z pkhuong: add an intrusive list pointer to each page struct. mark pages are added to the work list when we find a new object to scan and the page isn't already in that list
2014-09-13T21:55:13Z pkhuong: newspace pages are added when we close an allocation region
2014-09-13T21:56:24Z csziacobus joined #sbcl
2014-09-13T21:56:34Z pkhuong: weak references now need to stop looking for forwarding pointer. They should instead call object_marked_p that'll look for a forwarding pointer in copy pages and at the live bitmap in mark pages
2014-09-13T21:57:07Z pkhuong: major GCs are also easy (all the pages are up for GC)
2014-09-13T21:57:39Z pkhuong: we lose medium GCs that reclaim more than just the nursery but not the whole heap
2014-09-13T21:59:15Z nyef: Hrm. We can't malloc() during GC, right? Can we allocate markbits on the lisp heap instead?
2014-09-13T21:59:29Z pkhuong: I was going to use a mmap-backed region allocator
2014-09-13T21:59:45Z pkhuong: we only release mark bitmaps at the end of GC
2014-09-13T22:00:11Z nyef: That's fair, I guess.
2014-09-13T22:00:21Z pkhuong: (and we then release them all, in bulk)
2014-09-13T22:02:53Z pkhuong: for medium GCs, I guess we could do something like track an overapproximation of the set of pages that point to each page
2014-09-13T22:03:41Z pkhuong: clear the set before major GCs, otherwise it's append-only. If the set becomes too large, replace it with a special "everything refers here" marker.
2014-09-13T22:04:34Z nyef: Hrm.
2014-09-13T22:05:00Z nyef: I'm almost thinking that it might make sense to have pluggable GC options.
2014-09-13T22:05:12Z pkhuong: woah.
2014-09-13T22:05:37Z nyef: Possibly build-time-only, but still...
2014-09-13T22:06:14Z pkhuong: I don't want more configurations to maintain and debug ;)
2014-09-13T22:06:40Z nyef: Heh.
2014-09-13T22:08:16Z nyef: We have a hard enough time with two GCs, neither of which work on all the backends that the compiler supports, eight compiler backends, two of which are suspected to be so stale as to possibly not even build anymore...
2014-09-13T22:09:01Z nyef: Maybe I'll have time and attention span in October to try and do an HPPA/Linux build.
2014-09-13T22:09:27Z nyef: Or to set up an Alpha.
2014-09-13T22:11:58Z zacts_ joined #sbcl
2014-09-13T22:14:15Z zacts quit (Remote host closed the connection)
2014-09-13T22:15:21Z nyef: Okay, markbits. One (or two) bits per allocation unit, with allocation units being two "heap words" wide. "Heap words" are 32 or 64 bits, so we're looking at allocations as large as 1/32 the size of a heap page or as small as 1/128 the size of a heap page.
2014-09-13T22:15:47Z pkhuong: right
2014-09-13T22:15:50Z attila_lendvai joined #sbcl
2014-09-13T22:16:31Z nyef: And that depends on if we're also collecting marked-through bits and the width of a heap word.
2014-09-13T22:16:36Z |3b| quit (Ping timeout: 267 seconds)
2014-09-13T22:17:18Z Krystof joined #sbcl
2014-09-13T22:17:18Z ChanServ has set mode +o Krystof
2014-09-13T22:19:12Z logand` joined #sbcl
2014-09-13T22:19:34Z edgar-rfx joined #sbcl
2014-09-13T22:19:45Z sav joined #sbcl
2014-09-13T22:19:48Z csziacobus quit (Quit: csziacobus)
2014-09-13T22:20:45Z prxq_ joined #sbcl
2014-09-13T22:22:30Z brucem_ joined #sbcl
2014-09-13T22:23:33Z logand quit (Ping timeout: 260 seconds)
2014-09-13T22:24:17Z oleo__ joined #sbcl
2014-09-13T22:27:24Z oleo quit (*.net *.split)
2014-09-13T22:27:26Z christoph_debian quit (*.net *.split)
2014-09-13T22:27:26Z tmh_ quit (*.net *.split)
2014-09-13T22:27:26Z brucem quit (*.net *.split)
2014-09-13T22:27:26Z edgar-rft quit (*.net *.split)
2014-09-13T22:27:29Z yauz quit (*.net *.split)
2014-09-13T22:27:29Z prxq quit (*.net *.split)
2014-09-13T22:28:03Z yauz joined #sbcl
2014-09-13T22:29:07Z christoph_debian joined #sbcl
2014-09-13T22:46:32Z attila_lendvai quit (Disconnected by services)
2014-09-13T22:46:32Z attila_lendvai1 joined #sbcl
2014-09-13T22:46:32Z attila_lendvai1 quit (Changing host)
2014-09-13T22:46:32Z attila_lendvai1 joined #sbcl
2014-09-13T22:48:14Z antoszka_ joined #sbcl
2014-09-13T22:49:01Z drmeiste_ joined #sbcl
2014-09-13T22:49:28Z nyef_ joined #sbcl
2014-09-13T22:52:23Z slyrus joined #sbcl
2014-09-13T22:53:16Z antoszka quit (Ping timeout: 246 seconds)
2014-09-13T22:53:21Z scymtym quit (Ping timeout: 246 seconds)
2014-09-13T22:53:22Z Krystof quit (Ping timeout: 246 seconds)
2014-09-13T22:53:22Z nyef quit (Ping timeout: 246 seconds)
2014-09-13T22:53:23Z jackdaniel quit (Ping timeout: 246 seconds)
2014-09-13T22:53:25Z drmeister quit (Ping timeout: 246 seconds)
2014-09-13T22:53:26Z Intensity quit (Ping timeout: 246 seconds)
2014-09-13T22:55:23Z jackdaniel joined #sbcl
2014-09-13T22:55:29Z Intensity joined #sbcl
2014-09-13T23:00:33Z csziacobus joined #sbcl
2014-09-13T23:07:16Z segv- quit (Ping timeout: 272 seconds)
2014-09-13T23:21:22Z slyrus quit (Ping timeout: 240 seconds)
2014-09-13T23:22:22Z zacts_ quit (Quit: leaving)
2014-09-13T23:31:25Z LiamH1 joined #sbcl
2014-09-13T23:33:28Z LiamH quit (Ping timeout: 250 seconds)
2014-09-13T23:58:11Z attila_lendvai1 quit (Quit: Leaving.)
2014-09-14T00:02:20Z csziacobus quit (Quit: csziacobus)
2014-09-14T00:15:36Z drmeiste_ is now known as drmeister
2014-09-14T00:27:10Z drmeister is now known as drmeister_
2014-09-14T00:31:06Z drmeister_ is now known as drmeister__
2014-09-14T00:58:38Z antoszka_ is now known as antoszka
2014-09-14T01:10:03Z Skrylar: so if i was interested in peering at the sbcl internals, how many levels of cthulian horrors should be expected
2014-09-14T01:11:11Z Bike: i think it's pretty readable once you get past the (defun car (x) (car x)) stuff. i'm not a developer in the real deep parts, though.
2014-09-14T01:12:08Z nyef_: There's the NIL-is-a-CONS thing...
2014-09-14T01:12:26Z nyef_: And I haven't really dared to look into the guts of PCL.
2014-09-14T01:13:01Z prxq joined #sbcl
2014-09-14T01:13:21Z nyef_: Umm... And the breakpoint stuff is... ugh.
2014-09-14T01:13:36Z nyef_: And the debugger is a mess.
2014-09-14T01:13:55Z nyef_: ... And the type system.
2014-09-14T01:14:07Z nyef_: The type system, especially alien-types is twisty.
2014-09-14T01:16:07Z prxq_ quit (Ping timeout: 245 seconds)
2014-09-14T01:20:34Z Skrylar: i'll have to take a look at it when i've acclimated better to CL; i can understand a lot of the compiler voodoo but not what lisp calls everything. lol
2014-09-14T01:21:00Z nyef_: I still don't understand a good chunk of the front-end of the compiler.
2014-09-14T01:24:06Z nyef_: Actually, does anyone here understand how the compiler decides to issue a code-deletion note? I've got a branch where it's emitting them rather more frequently, and I don't believe that it should be.
2014-09-14T01:24:50Z nyef_: (Yes, it's plausibly deleting more code, but I don't believe that it should be emitting more notes.)
2014-09-14T01:32:38Z jsnell: there's a heuristic to suppress code that came from macros, inlining, etc. are they those kinds of notices?
2014-09-14T01:40:38Z pkhuong: the type system is complex, but only because CL's type system is.
2014-09-14T01:41:02Z nyef_: This is the bit that triggers the extra code-deletion notes: http://repo.or.cz/w/sbcl/nyef.git/commitdiff/a8c08c36b759d0ede74d5298a5cdadf301bf7d0b
2014-09-14T01:42:15Z nyef_: It nails a number of bugs to the wall, but there's this side-effect...
2014-09-14T01:42:42Z pkhuong: nyef_: do the deletion notes refer to arg parsing code?
2014-09-14T01:43:00Z nyef_: I don't remember, actually. That's a plausibility.
2014-09-14T01:43:53Z pkhuong: I think we already emit spurious notes for entry point code. Your patch probably deletes more of them.
2014-09-14T01:44:52Z pkhuong: nyef_: any reason you don't do a full mark and sweep on clambdas?
2014-09-14T01:46:06Z nyef_: pkhuong: I thought that that was what this DID?
2014-09-14T01:47:02Z pkhuong: your root set includes "It has a REF which has a NODE-COMPONENT other than the LAMBDA-COMPONENT of the CLAMBDA."
2014-09-14T01:47:10Z pkhuong: oh wait, I see.
2014-09-14T01:47:32Z pkhuong: we don't have access to other components when this runs
2014-09-14T01:47:35Z nyef_: Right.
2014-09-14T01:49:10Z nyef_: I don't quite understand the mess which is the toplevel logic and the DFO bit that splits out components, but my impression is that the separate components all have live external hooks.
2014-09-14T01:49:45Z nyef_: Or, well, if they get this far they suddenly become empty.
2014-09-14T01:49:48Z pkhuong: I assume you'll address the potential stack overflow if this works out
2014-09-14T01:50:06Z nyef_: ... potential stack overflow?
2014-09-14T01:50:43Z nyef_: DCE-ANALYZE-BLOCK executes early if the block has already been seen, and the first thing it does after that is mark the block as seen.
2014-09-14T01:51:10Z pkhuong: what if there's very many blocks?
2014-09-14T01:51:17Z pkhuong: or many lambdas.
2014-09-14T01:52:42Z nyef_: Worst-case scenario, we move to a heap-based worklist, but that feels somewhat unlikely at this point.
2014-09-14T01:53:33Z pkhuong: I think we do worklist everywhere but in ir1tran, and I've seen stack exhaustion warnings in ir1tran
2014-09-14T01:54:36Z nyef_: Well, if it turns out to be a problem, there's a fallback position that should be easy enough to implement.
2014-09-14T01:54:42Z pkhuong: yeah.
2014-09-14T01:54:54Z nyef_: Before then, however, there's this code-deletion-note thing.
2014-09-14T01:55:46Z nyef_: I guess I'll push it to my worklist of things to dig into as I find time. /-:
2014-09-14T01:56:54Z pkhuong: if it is arg parsing code, a declaration to muffle code deletion notes might be the simplest fix ;)
2014-09-14T02:03:53Z LiamH1 is now known as LiamH
2014-09-14T02:05:42Z brucem_ quit (Changing host)
2014-09-14T02:05:42Z brucem_ joined #sbcl
2014-09-14T02:05:47Z brucem_ is now known as brucem
2014-09-14T02:38:07Z christoph_debian quit (Ping timeout: 245 seconds)
2014-09-14T02:51:51Z christoph_debian joined #sbcl
2014-09-14T02:55:08Z karswell` joined #sbcl
2014-09-14T02:55:37Z DGASAU` quit (Ping timeout: 260 seconds)
2014-09-14T02:59:18Z DGASAU` joined #sbcl
2014-09-14T03:29:09Z edgar-rfx quit (Quit: lifeform experiment ended because all hope lost)
2014-09-14T03:57:51Z DGASAU` quit (Remote host closed the connection)
2014-09-14T03:59:22Z DGASAU` joined #sbcl
2014-09-14T04:14:57Z LiamH quit (Quit: Leaving.)
2014-09-14T04:16:07Z DGASAU` quit (Ping timeout: 245 seconds)
2014-09-14T04:24:56Z zacts joined #sbcl
2014-09-14T04:24:56Z DGASAU` joined #sbcl
2014-09-14T04:32:16Z nyef_ quit (Quit: G'night all.)
2014-09-14T04:34:00Z |3b|`` joined #sbcl
2014-09-14T04:35:24Z |3b|`` is now known as |3b|
2014-09-14T05:09:03Z leo2007 quit (Quit: rcirc on GNU Emacs 24.4.50.1)
2014-09-14T05:09:28Z leo2007 joined #sbcl
2014-09-14T05:31:56Z drmeister__ is now known as drmeister
2014-09-14T05:44:33Z yacks quit (Ping timeout: 260 seconds)
2014-09-14T05:50:19Z yacks joined #sbcl
2014-09-14T05:56:04Z oleo joined #sbcl
2014-09-14T05:58:16Z oleo__ quit (Ping timeout: 250 seconds)
2014-09-14T06:09:19Z sdemarre joined #sbcl
2014-09-14T06:44:32Z sdemarre quit (Ping timeout: 272 seconds)
2014-09-14T07:02:02Z angavrilov joined #sbcl
2014-09-14T07:02:32Z sdemarre joined #sbcl
2014-09-14T07:23:53Z leo2007 quit (Quit: rcirc on GNU Emacs 24.4.50.1)
2014-09-14T07:24:07Z Krystof joined #sbcl
2014-09-14T07:24:07Z ChanServ has set mode +o Krystof
2014-09-14T07:27:17Z gingerale joined #sbcl
2014-09-14T07:40:02Z momo-reina joined #sbcl
2014-09-14T07:45:11Z dkcl joined #sbcl
2014-09-14T08:35:17Z erikvarga joined #sbcl
2014-09-14T08:48:25Z momo-reina quit (Ping timeout: 272 seconds)
2014-09-14T09:09:08Z segv- joined #sbcl
2014-09-14T09:26:32Z leo2007 joined #sbcl
2014-09-14T10:56:55Z attila_lendvai joined #sbcl
2014-09-14T11:29:52Z sdemarre quit (Ping timeout: 245 seconds)
2014-09-14T12:33:21Z edgar-rft joined #sbcl
2014-09-14T12:48:12Z erikvarga quit (Ping timeout: 245 seconds)
2014-09-14T12:50:38Z erikvarga joined #sbcl
2014-09-14T13:04:11Z dkcl quit (Remote host closed the connection)
2014-09-14T13:04:56Z dkcl joined #sbcl
2014-09-14T13:05:06Z dkcl quit (Changing host)
2014-09-14T13:05:06Z dkcl joined #sbcl
2014-09-14T13:44:15Z loke_ joined #sbcl
2014-09-14T14:16:30Z nyef joined #sbcl
2014-09-14T14:16:42Z nyef: G'morning all.
2014-09-14T14:19:08Z pkhuong: good morning
2014-09-14T14:36:46Z scymtym joined #sbcl
2014-09-14T14:43:00Z sdemarre joined #sbcl
2014-09-14T14:52:59Z ebrasca joined #sbcl
2014-09-14T14:58:12Z sdemarre quit (Ping timeout: 250 seconds)
2014-09-14T15:16:59Z sdemarre joined #sbcl
2014-09-14T15:22:45Z gingerale- joined #sbcl
2014-09-14T15:24:57Z dkcl quit (Ping timeout: 246 seconds)
2014-09-14T15:24:57Z gingerale quit (Ping timeout: 246 seconds)
2014-09-14T15:24:57Z Intensity quit (Ping timeout: 246 seconds)
2014-09-14T15:24:57Z dkcl joined #sbcl
2014-09-14T15:24:58Z gingerale- is now known as gingerale
2014-09-14T15:25:08Z dkcl quit (Changing host)
2014-09-14T15:25:08Z dkcl joined #sbcl
2014-09-14T15:27:29Z Intensity joined #sbcl
2014-09-14T15:31:48Z DGASAU`` joined #sbcl
2014-09-14T15:33:22Z drmeiste_ joined #sbcl
2014-09-14T15:38:20Z sdemarre quit (*.net *.split)
2014-09-14T15:38:20Z edgar-rft quit (*.net *.split)
2014-09-14T15:38:20Z attila_lendvai quit (*.net *.split)
2014-09-14T15:38:21Z DGASAU` quit (*.net *.split)
2014-09-14T15:38:21Z drmeister quit (*.net *.split)
2014-09-14T15:39:55Z pkhuong: btw, if anyone is looking for an easy commit: s/0x%/%#/ would clean up the runtime.
2014-09-14T15:41:20Z attila_lendvai joined #sbcl
2014-09-14T15:42:00Z sdemarre joined #sbcl
2014-09-14T15:48:20Z attila_lendvai quit (Quit: Leaving.)
2014-09-14T15:55:49Z hlavaty` quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2014-09-14T16:22:14Z attila_lendvai joined #sbcl
2014-09-14T16:29:12Z segv- quit (Ping timeout: 250 seconds)
2014-09-14T16:39:51Z sdemarre quit (Ping timeout: 272 seconds)
2014-09-14T16:44:12Z nyef: pkhuong: I think that you might be right about the arg-parsing code being the cause of the extra code-deletion notes.
2014-09-14T16:45:14Z nyef: Hrm. Possibly-more-specifically, &KEY or maybe &REST and &KEY parsing.
2014-09-14T16:45:52Z nyef: Right, anything with &MORE processing, it looks like.
2014-09-14T16:50:21Z nyef does a build of HEAD so he can compare before-and-after disassembly of a couple of test functions.
2014-09-14T16:51:06Z dkcl quit (Remote host closed the connection)
2014-09-14T16:52:43Z dkcl joined #sbcl
2014-09-14T17:14:36Z attila_lendvai quit (Quit: Leaving.)
2014-09-14T17:17:33Z nyef: Ugh, typical. Identical code, both versions, but with the DCE pass it mentions deleting stuff.
2014-09-14T17:18:09Z loke_ quit (Remote host closed the connection)
2014-09-14T17:27:32Z slyrus joined #sbcl
2014-09-14T17:36:15Z yacks quit (Quit: Leaving)
2014-09-14T17:48:08Z scymtym: do we have something like ENSURE-CLASS-FINALIZED? i would add it to http://paste.lisp.org/display/143705 otherwise (and also use it at several places in CHANGE-CLASS)
2014-09-14T17:53:22Z nyef: Is FINALIZE-CLASS idempotent?
2014-09-14T17:59:17Z ebrasca quit (Remote host closed the connection)
2014-09-14T18:12:45Z Krystof: in MOP as specified, yes, but in principle user methods on it can be non-idempotent
2014-09-14T18:14:03Z drmeiste_ is now known as drmeister
2014-09-14T18:15:11Z Krystof: I could imagine something which hooked finalize-inheritance to keep track of class versions, and they might get annoyed if every call to a sequence function incremented the version
2014-09-14T18:32:47Z oleo quit (Ping timeout: 245 seconds)
2014-09-14T18:48:00Z scymtym: my motivation is reducing the many cases of (let ((c …)) (unless (c-f-p c) (f-i c)) (foo c)) to (foo (ensure-class-finalized c))
2014-09-14T18:48:47Z scymtym: so returning the finalized class would be second requirement
2014-09-14T18:50:54Z nyef: Sounds like a good factoring to me, even if there's only one instance of that pattern.
2014-09-14T18:51:25Z scymtym: i think there are many
2014-09-14T18:52:09Z scymtym: at least CHANGE-CLASS, some sequence functions and some places in PCL
2014-09-14T18:53:27Z scymtym: ok, thanks for the feedback, i will add ENSURE-CLASS-FINALIZED as the first commit in the patch
2014-09-14T18:54:20Z oleo joined #sbcl
2014-09-14T18:56:48Z scymtym: maybe one more question: i think certain uses of &rest + apply are efficient. in, for example, %concatenate-to-extended-sequence i basically do (defun %c-t-e-s (&rest seqs) (apply … seqs)) where the apply is a tail call. is this such efficient case?
2014-09-14T18:57:24Z scymtym: or should the function receive seqs as an explicit list in a single parameter?
2014-09-14T18:57:43Z nyef: It should be fairly efficient, IIRC. Disassemble such a case and see if it's consing and re-spreading SEQS or if it's just moving the stack packet around.
2014-09-14T18:59:49Z scymtym: i will do that. thanks
2014-09-14T19:00:19Z scymtym: any objections to the sequence protocol extension in general ?
2014-09-14T19:02:16Z nyef: I don't feel qualified to comment, as I'm not even a particularly heavy CLOS user, let alone familiar with the innards of PCL.
2014-09-14T19:08:14Z slyrus quit (Ping timeout: 250 seconds)
2014-09-14T19:14:52Z attila_lendvai joined #sbcl
2014-09-14T19:15:06Z attila_lendvai quit (Changing host)
2014-09-14T19:15:07Z attila_lendvai joined #sbcl
2014-09-14T19:19:11Z Krystof quit (Remote host closed the connection)
2014-09-14T19:44:22Z segv- joined #sbcl
2014-09-14T19:54:51Z dandersen joined #sbcl
2014-09-14T19:55:07Z csziacobus joined #sbcl
2014-09-14T19:55:13Z dkcl quit (Read error: Connection reset by peer)
2014-09-14T20:01:32Z oleo is now known as Guest4024
2014-09-14T20:02:24Z oleo__ joined #sbcl
2014-09-14T20:04:22Z Guest4024 quit (Ping timeout: 245 seconds)
2014-09-14T20:09:57Z sdemarre joined #sbcl
2014-09-14T20:11:42Z erikvarga left #sbcl
2014-09-14T20:14:33Z logand` left #sbcl
2014-09-14T20:41:24Z Bicyclidine joined #sbcl
2014-09-14T20:44:40Z sdemarre quit (Ping timeout: 260 seconds)
2014-09-14T20:59:07Z zacts: would add support for threads on OpenBSD x86-64 be trivial?
2014-09-14T20:59:16Z zacts: *adding
2014-09-14T21:06:34Z prxq: zacts: if it were...
2014-09-14T21:08:55Z zacts: prxq: I don't understand
2014-09-14T21:21:35Z nyef: zacts: Adding threading support for a new OS when there is already support for the same CPU in a similar OS should, in theory, be straightforward.
2014-09-14T21:22:23Z nyef: Theory is wonderful, everything works there.
2014-09-14T21:22:56Z nyef: ... And haven't you been asking that question for days now?
2014-09-14T21:23:13Z zacts: nyef: I only asked it once a couple of days ago
2014-09-14T21:23:30Z zacts: on #lisp, I figured I would ask the actual people who know sbcl well once also
2014-09-14T21:23:58Z zacts: I promise not to ask again, unless I get more specific questions from my own research. =)
2014-09-14T21:24:09Z zacts: ^ disclaimer
2014-09-14T21:24:19Z nyef: Honestly, if you have the machine and OS required, why not try it?
2014-09-14T21:24:53Z zacts: nyef: oh really? so just try it to see if it already works? I guess I just figured that it didn't work at all.
2014-09-14T21:25:22Z nyef: To see if it already works, and to see if you can figure out why not and how to fix it if it doesn't.
2014-09-14T21:25:57Z zacts: sure, I'll do that as I learn CL, and get to that point. I'll let you guys know what problems I'm finding.
2014-09-14T21:26:09Z zacts: first on #lisp I guess?
2014-09-14T21:26:14Z nyef: Let's see, x86-64, so you won't need to set up the segment register...
2014-09-14T21:26:27Z zacts: indeed I'm on OpenBSD-5.5 on x86-64
2014-09-14T21:26:52Z zacts: I just feel that a working SBCL on OpenBSD would be a killer combination
2014-09-14T21:27:01Z zacts: s/working/fully working/
2014-09-14T21:28:03Z nyef: Just looking at the makefile fragments for the runtime, it seems unlikely that threads will work out-of-the-box, as it were.
2014-09-14T21:29:33Z zacts: nyef: If there is anything I can do to help on my end, I'll test out patches or whatever. Note: I'm a newbie still. I only really learned what threads are like last week. Although, I know UNIX cli and vim.
2014-09-14T21:30:09Z Krystof joined #sbcl
2014-09-14T21:30:10Z ChanServ has set mode +o Krystof
2014-09-14T21:30:48Z nyef: It *might* be as simple as linking in a thread library, for all I know.
2014-09-14T21:31:01Z nyef: I'm not an OpenBSD user at all, so I have no idea on that side of things.
2014-09-14T21:47:48Z karswell` quit (Remote host closed the connection)
2014-09-14T21:48:43Z karswell joined #sbcl
2014-09-14T22:12:33Z edgar-rft joined #sbcl
2014-09-14T22:44:20Z |3b| quit (Read error: Connection reset by peer)
2014-09-14T22:45:21Z |3b| joined #sbcl
2014-09-14T23:14:30Z dandersen quit (Ping timeout: 246 seconds)
2014-09-14T23:20:52Z gingerale quit (Ping timeout: 240 seconds)
2014-09-14T23:34:45Z edgar-rft quit (Ping timeout: 258 seconds)
2014-09-14T23:35:51Z slyrus joined #sbcl
2014-09-14T23:50:35Z scymtym: hopefully final version: http://paste.lisp.org/display/143721
2014-09-14T23:50:36Z attila_lendvai quit (Quit: Leaving.)
2014-09-14T23:51:01Z scymtym: if nobody objects, i would like to commit this tomorrow or the day after tomorrow
2014-09-14T23:54:19Z pkhuong: the difference is once only and with-unique-names?
2014-09-14T23:54:42Z pkhuong: ah, missed the annotations
2014-09-14T23:56:57Z pkhuong: s/loose/lose/
2014-09-14T23:58:03Z pkhuong: Not sure I like labels around the two defmethod.
2014-09-14T23:58:38Z pkhuong: why can't it be another topevel function?
2014-09-14T23:59:29Z slyrus_ joined #sbcl
2014-09-14T23:59:36Z pkhuong: a type declaration on ensure-class-finalized might help improve some code.
2014-09-15T00:00:35Z PuercoPope joined #sbcl
2014-09-15T00:02:27Z pkhuong: why doesn't sequence:map call length?
2014-09-15T00:03:12Z thoto_ joined #sbcl
2014-09-15T00:03:40Z angavrilov_ joined #sbcl
2014-09-15T00:04:45Z pkhuong: sequence:merge doesn't look stable.
2014-09-15T00:04:47Z carvite_ joined #sbcl
2014-09-15T00:06:19Z pkhuong: It should probably do (cond ((funcall compare elt2 elt1) [enqueue elt2])) (t [enqueue elt1]))
2014-09-15T00:06:44Z pkhuong: (see sb!impl:merge-lists)
2014-09-15T00:10:23Z faheem__1 joined #sbcl
2014-09-15T00:13:25Z slyrus quit (*.net *.split)
2014-09-15T00:13:27Z csziacobus quit (*.net *.split)
2014-09-15T00:13:27Z segv- quit (*.net *.split)
2014-09-15T00:13:36Z angavrilov quit (*.net *.split)
2014-09-15T00:13:36Z sav quit (*.net *.split)
2014-09-15T00:13:36Z faheem_ quit (*.net *.split)
2014-09-15T00:13:36Z PuercoPop quit (*.net *.split)
2014-09-15T00:13:49Z thoto quit (*.net *.split)
2014-09-15T00:13:49Z carvite quit (*.net *.split)
2014-09-15T00:13:49Z carvite_ is now known as carvite
2014-09-15T00:13:49Z PuercoPope is now known as PuercoPop
2014-09-15T00:13:49Z scymtym: pkhuong: thanks, i make these changes
2014-09-15T00:14:16Z scymtym: map for extended sequences has been like this (not calling LENGTH) before
2014-09-15T00:15:26Z scymtym: maybe the intention is having it work on circular lists and infinite user-defined sequences as long as at least one argument sequence is finite?
2014-09-15T00:16:00Z pkhuong: perhaps. in theory, sequences must have a finite length ;)
2014-09-15T00:17:06Z scymtym: yeah, but i think useful things can be done in the grey-area
2014-09-15T00:17:39Z scymtym: and things like sb-sequence:emptyp hopefully improve the situation
2014-09-15T00:18:01Z pkhuong: Perhaps something like: sequence:length-below sequence limit => returns length of the sequence. May also return NIL if the length is > the limit.
2014-09-15T00:18:27Z scymtym: i actually have something like this in my sequence branch :)
2014-09-15T00:19:04Z pkhuong: it'd simplify some list code as well
2014-09-15T00:19:37Z tmh_ joined #sbcl
2014-09-15T00:19:45Z pkhuong: currently, everyone has to roll their own list of length at least/at most/exactly
2014-09-15T00:20:00Z scymtym: ok, i will try to make length-below next (i think mine was called length<)
2014-09-15T00:20:17Z pkhuong: length< looks like a predicate ;)
2014-09-15T00:20:45Z scymtym: if i remember correctly, it returned a boolean
2014-09-15T00:20:51Z scymtym: but your idea is more useful
2014-09-15T00:21:29Z scymtym: but regarding map and generic sequences: maybe when extended sequences can define their own map implementations, it is more acceptable to use length in the default implementation
2014-09-15T00:21:32Z pkhuong: the name is awful (:
2014-09-15T00:22:25Z pkhuong: otoh, map involves multiple sequence types, and it doesn't seem reasonable to have people implement the cross product of all the possibilities they want to handle
2014-09-15T00:23:03Z scymtym: and they cannot nicely dispatch one all argument sequences
2014-09-15T00:23:13Z scymtym: one problem with using length could be (map 'my-seq (circular-list 1 2 3) (list 1 2 3)) which works with ordinary map
2014-09-15T00:24:05Z scymtym: (assuming no method on sequence:map for my-seq)
2014-09-15T00:24:47Z pkhuong: right. I guess it's safer to keep the current %map-for-effect thing until we see if length-below (except with a better name) makes sense
2014-09-15T00:25:24Z Bicyclidine: Would you just pick some arbitrary limit for length-below for the first sequence?
2014-09-15T00:25:51Z pkhuong: Bicyclidine: geometric growth until one of the sequences returns non-NIL.
2014-09-15T00:26:44Z scymtym: ok, so i will implement your suggestions except for using length in the default sequence:map method
2014-09-15T00:26:46Z scymtym: i have to sleep now :)
2014-09-15T00:26:56Z pkhuong: cool, thanks for the patches!
2014-09-15T00:27:14Z scymtym: thanks for the review; will commit tomorrow or so
2014-09-15T00:32:20Z scymtym quit (Ping timeout: 272 seconds)
2014-09-15T00:35:28Z DGASAU`` quit (Write error: Connection reset by peer)
2014-09-15T00:36:09Z DGASAU`` joined #sbcl
2014-09-15T00:48:21Z krzysz00_ quit (Ping timeout: 260 seconds)
2014-09-15T00:55:03Z krzysz00 joined #sbcl
2014-09-15T00:55:37Z echo-area joined #sbcl
2014-09-15T01:10:57Z krzysz00 quit (Ping timeout: 272 seconds)
2014-09-15T01:13:37Z prxq_ joined #sbcl
2014-09-15T01:16:57Z prxq quit (Ping timeout: 245 seconds)
2014-09-15T01:17:24Z Bicyclidine quit (Ping timeout: 255 seconds)
2014-09-15T01:17:24Z echo-area quit (Remote host closed the connection)
2014-09-15T01:26:26Z krzysz00 joined #sbcl
2014-09-15T01:55:00Z krzysz00 quit (Ping timeout: 260 seconds)
2014-09-15T02:01:22Z segv- joined #sbcl
2014-09-15T02:06:17Z BitPuffin quit (Ping timeout: 260 seconds)
2014-09-15T02:15:14Z krzysz00 joined #sbcl
2014-09-15T02:16:56Z slyrus_ quit (Ping timeout: 260 seconds)
2014-09-15T02:19:42Z kanru joined #sbcl
2014-09-15T02:34:46Z slyrus joined #sbcl
2014-09-15T02:38:54Z christoph_debian quit (Ping timeout: 250 seconds)
2014-09-15T02:51:47Z christoph_debian joined #sbcl
2014-09-15T03:09:58Z nyef quit (Quit: G'night all.)
2014-09-15T03:33:27Z krzysz00 quit (Ping timeout: 272 seconds)
2014-09-15T03:37:05Z krzysz00 joined #sbcl
2014-09-15T03:54:52Z yacks joined #sbcl
2014-09-15T04:32:50Z kanru quit (Remote host closed the connection)
2014-09-15T04:33:30Z kanru joined #sbcl
2014-09-15T04:52:09Z yacks quit (Ping timeout: 258 seconds)
2014-09-15T04:54:54Z yacks joined #sbcl
2014-09-15T05:21:52Z prxq_ is now known as prxq
2014-09-15T05:24:46Z sdemarre joined #sbcl
2014-09-15T05:29:19Z sdemarre left #sbcl
2014-09-15T05:37:05Z BitPuffin joined #sbcl
2014-09-15T05:47:24Z segv- quit (Ping timeout: 260 seconds)
2014-09-15T05:50:32Z oleo__ quit (Quit: Verlassend)
2014-09-15T05:50:33Z BitPuffin quit (Ping timeout: 255 seconds)
2014-09-15T06:13:04Z gingerale joined #sbcl
2014-09-15T06:23:30Z angavrilov_ quit (Remote host closed the connection)
2014-09-15T06:23:53Z angavrilov_ joined #sbcl
2014-09-15T06:25:18Z prxq quit (Ping timeout: 258 seconds)
2014-09-15T06:25:52Z pkhuong quit (Ping timeout: 240 seconds)
2014-09-15T07:21:09Z BitPuffin joined #sbcl
2014-09-15T07:32:02Z loke: The ABCL bootstrap is broken. Is this a known issue?
2014-09-15T07:38:24Z DGASAU`` is now known as DGASAU
2014-09-15T07:48:20Z DGASAU quit (Read error: Connection reset by peer)
2014-09-15T07:57:46Z ams: loke: should you not ask that to the abcl people?
2014-09-15T07:58:05Z loke: ams: That too, I guess
2014-09-15T07:58:19Z loke: I just wanted to know if anyone heard of it yet?
2014-09-15T07:59:08Z DGASAU joined #sbcl
2014-09-15T07:59:08Z pkhuong joined #sbcl
2014-09-15T07:59:08Z prxq joined #sbcl
2014-09-15T07:59:08Z echo-area joined #sbcl
2014-09-15T08:00:10Z pkhuong is now known as Guest53341
2014-09-15T08:04:10Z DGASAU quit (Remote host closed the connection)
2014-09-15T08:06:41Z DGASAU joined #sbcl
2014-09-15T09:02:52Z DGASAU quit (Read error: Connection reset by peer)
2014-09-15T09:03:15Z DGASAU` joined #sbcl
2014-09-15T09:05:41Z stassats joined #sbcl
2014-09-15T09:06:02Z Krystof: loke: no.  What's the failure mode
2014-09-15T09:06:27Z loke: An undefined symbol SB-INT...something
2014-09-15T09:06:40Z loke: A colleague of mine did it
2014-09-15T09:07:41Z Krystof: please find and paste if possible
2014-09-15T09:14:10Z loke: Sure. As soon as he's back at his desk
2014-09-15T09:15:45Z Krystof: thanks
2014-09-15T09:20:00Z stassats quit (Ping timeout: 246 seconds)
2014-09-15T09:23:47Z momo-reina joined #sbcl
2014-09-15T09:38:12Z dkcl joined #sbcl
2014-09-15T10:01:17Z DGASAU` is now known as DGASAU
2014-09-15T10:10:55Z segv- joined #sbcl
2014-09-15T10:13:42Z stassats joined #sbcl
2014-09-15T10:21:45Z momo-reina quit (Remote host closed the connection)
2014-09-15T10:26:39Z echo-area quit (Remote host closed the connection)
2014-09-15T10:36:47Z BitPuffin quit (Remote host closed the connection)
2014-09-15T10:49:59Z scymtym joined #sbcl
2014-09-15T10:54:19Z scymtym: sequence:merge with required stability property (as pointed out by pkhuong) and fewer iterator function calls: http://paste.lisp.org/display/143731
2014-09-15T10:55:47Z dkcl: zacts: I have tried to run a multithreaded SBCL under OpenBSD in the past
2014-09-15T10:56:52Z dkcl: zacts: It worked for a while under x86, even though other tests failed. I don't think it can compile with --with-sb-ext anymore. The multithreading model for OpenBSD was changed (to rthreads, iirc), but should be mature enough by now.
2014-09-15T10:59:08Z dkcl: zacts: I also wish I could use OpenBSD as my main CL development OS, but the closest thing to it would be FreeBSD, under which SBCL builds with any and all options just fine (and also CCL, but that is off-topic to this channel).
2014-09-15T10:59:30Z dkcl: s/and also/and so does/
2014-09-15T11:02:58Z dkcl: s/--with-sb-ext/--with-sb-thread/, too. Derp.
2014-09-15T11:06:09Z DGASAU quit (Read error: Connection reset by peer)
2014-09-15T11:06:48Z DGASAU joined #sbcl
2014-09-15T11:26:55Z scymtym quit (Ping timeout: 258 seconds)
2014-09-15T11:41:41Z edgar-rft joined #sbcl
2014-09-15T11:42:42Z kanru quit (Remote host closed the connection)
2014-09-15T11:51:13Z scymtym joined #sbcl
2014-09-15T11:54:33Z kanru joined #sbcl
2014-09-15T11:58:02Z Guest53341 is now known as pkhuong
2014-09-15T12:04:39Z pkhuong: Krystof: Looking at the code for sb-sequence reminded me of tcr's iterator project. I think CPSing make-sequence-iterator might be the nicest way to allow DXification.
2014-09-15T12:17:06Z segv- quit (Ping timeout: 246 seconds)
2014-09-15T12:32:44Z scymtym: pkhuong: maybe something like that can be provided as an additional generic function (e.g. call-with-dx-sequence-iterator)? at least my uses of sb-sequence have heavyweight sequences and long-lived iterators which are passed around and stored.
2014-09-15T12:37:49Z Guest17545 is now known as qbit
2014-09-15T13:00:32Z drmeister quit (Remote host closed the connection)
2014-09-15T13:28:56Z oleo joined #sbcl
2014-09-15T13:35:05Z eudoxia joined #sbcl
2014-09-15T14:04:50Z drmeister joined #sbcl
2014-09-15T14:05:56Z drmeister quit (Remote host closed the connection)
2014-09-15T14:06:50Z nicdev joined #sbcl
2014-09-15T15:06:44Z erikvarga joined #sbcl
2014-09-15T15:35:24Z krzysz00 quit (Ping timeout: 260 seconds)
2014-09-15T16:03:28Z erikvarga left #sbcl
2014-09-15T16:06:22Z iqool joined #sbcl
2014-09-15T16:06:37Z iqool left #sbcl
2014-09-15T16:16:00Z slyrus quit (Ping timeout: 260 seconds)
2014-09-15T16:21:52Z krzysz00 joined #sbcl
2014-09-15T16:49:24Z stassats quit (Ping timeout: 246 seconds)
2014-09-15T16:55:39Z krzysz00 quit (Ping timeout: 255 seconds)
2014-09-15T17:06:12Z krzysz00 joined #sbcl
2014-09-15T17:16:09Z attila_lendvai joined #sbcl
2014-09-15T17:26:51Z krzysz00 quit (Ping timeout: 246 seconds)
2014-09-15T17:41:30Z yacks quit (Ping timeout: 272 seconds)
2014-09-15T17:44:46Z drmeister joined #sbcl
2014-09-15T17:45:39Z yacks joined #sbcl
2014-09-15T17:58:37Z segv- joined #sbcl
2014-09-15T18:06:14Z eudoxia quit (Read error: Connection reset by peer)
2014-09-15T18:17:53Z reb``: ls
2014-09-15T18:17:59Z reb``: oops
2014-09-15T18:25:48Z drmeister quit (Read error: Connection reset by peer)
2014-09-15T18:31:50Z foom quit (Read error: Connection reset by peer)
2014-09-15T18:41:27Z pkhuong: http://htmlpreview.github.io/?https://github.com/HowardHinnant/papers/blob/master/hashing.html <- good proposal
2014-09-15T18:41:43Z pkhuong: switching sxhash to that design might be a good idea
2014-09-15T18:42:08Z foom joined #sbcl
2014-09-15T18:56:45Z drmeister joined #sbcl
2014-09-15T19:05:20Z Posterdati quit (Remote host closed the connection)
2014-09-15T19:12:48Z Posterdati joined #sbcl
2014-09-15T19:40:09Z pkhuong: re malloc in the gc: I checked and glibc qsort still calls msort when it feels like it, so we can't use qsort in the gc until they fix that.
2014-09-15T19:52:35Z krzysz00 joined #sbcl
2014-09-15T23:54:06Z ccl-logbot joined #sbcl
2014-09-15T23:54:06Z 
2014-09-15T23:54:06Z names: ccl-logbot tmh_ foom2 nicdev` Bicyclidine drmeister krzysz00 oleo scymtym_ Posterdati yacks kanru scymtym edgar-rft DGASAU prxq pkhuong angavrilov_ christoph_debian faheem__1 carvite thoto_ PuercoPop |3b| @Krystof Intensity leo2007 zacts jackdaniel antoszka yauz brucem asedeno les Bike ivan`` qbit pchrist jdz cracauer_ redline6561 ams Skrylar milosn wmarvel_ irsol p_l akkad alchemis7 White_Flame Perlboy Blkt fe[nl]ix luis sobel heddwch leoc flip214 phf
2014-09-15T23:54:06Z names: ferada loke jsnell specbot mood p_l|backup minion
2014-09-15T23:54:46Z karswell joined #sbcl
2014-09-19T14:49:35Z ccl-logbot joined #sbcl
2014-09-19T14:49:35Z 
2014-09-19T14:49:35Z names: ccl-logbot reb`` Krystof_ Xach yauz_2 drmeiste_ DGASAU eudoxia oleo `JRG dkcl scymtym Hydan hlavaty segv- fridim__ slyrus leo2007 christoph_debian prxq psilord kanru p_l flip214 antoszka jdz Posterdati angavrilov PuercoPop luis akkad redline6561 pchrist fe[nl]ix Blkt alchemis7 foom nicdev pkhuong faheem__1 carvite thoto_ |3b| zacts jackdaniel brucem asedeno les ivan`` qbit cracauer_ ams irsol White_Flame Perlboy sobel heddwch leoc phf ferada loke jsnell
2014-09-19T14:49:35Z names: specbot minion p_l|backup mood karswell milosn Intensity Bike krzysz00
2014-09-19T14:51:06Z Krystof_ is now known as Krystof
2014-09-19T14:51:18Z ChanServ has set mode +o Krystof
2014-09-19T14:51:43Z Krystof: Xach: instructions to reproduce would be appreciated
2014-09-19T14:51:45Z hlavaty: iterate doesnt seem to work for me with SBCL 1.2.3.56-053d885 anymore
2014-09-19T14:52:08Z hlavaty: i havent got around investigating yet though
2014-09-19T14:54:26Z tmh_ joined #sbcl
2014-09-19T14:56:46Z Krystof: argh
2014-09-19T14:57:01Z `JRG quit (Ping timeout: 246 seconds)
2014-09-19T14:57:10Z Xach: Krystof: working on it.
2014-09-19T14:57:19Z Krystof: instructions for how to get my emacs out of this mess with old slime in the image but new slime on the filesystem would also be helpful
2014-09-19T14:59:35Z Krystof: ok, http://slime-tips.tumblr.com/post/11571143331/reloading-slime-without-restarting-emacs worked surprisingly well.  Thanks, slime-tips!
2014-09-19T15:00:28Z Krystof weeps as swankr is totally broken
2014-09-19T15:01:05Z karswell` joined #sbcl
2014-09-19T15:03:06Z |3b|: so that backtrace got an error, then got a BUG while trying to find the source?
2014-09-19T15:04:00Z Krystof: a style-warning, in fact
2014-09-19T15:04:06Z Krystof: a format control with more arguments than it needed
2014-09-19T15:04:22Z Krystof: but yes, then a BUG
2014-09-19T15:04:56Z Krystof: specifically a mismatch between the depth of the form and the length of the source-form numbers
2014-09-19T15:05:16Z Xach: Krystof: cd ~/quicklisp/local-projects/ && wget weitz.de/files/drakma.tar.gz && tar xzvf drakma.tar.gz for a first step
2014-09-19T15:05:18Z |3b|: Xach: root cause is that drakma changed, and drakma-async didn't adapt
2014-09-19T15:05:36Z Xach: Then with a recent quicklisp with a recent drakma-async (ql:quickload "drakma-async" :verbose t)
2014-09-19T15:05:47Z karswell quit (Ping timeout: 260 seconds)
2014-09-19T15:05:48Z krzysz00 quit (Ping timeout: 260 seconds)
2014-09-19T15:05:48Z Xach: I'm not sure if that's sufficient to reproduce, still investigating.
2014-09-19T15:06:05Z |3b|: https://github.com/edicl/drakma/commit/201d3638849e6bc2813a15e1f4e4c8bc60387817#diff-f3f3c870b53d40d654c2d9a00dca8b2cL141 called by https://github.com/orthecreedence/drakma-async/blob/0ed34802707241fda16d75d7a18a7085b06afed4/port/drakma-1.2.8.hijack.lisp#L378
2014-09-19T15:06:32Z krzysz00 joined #sbcl
2014-09-19T15:08:01Z loke_ joined #sbcl
2014-09-19T15:15:47Z |3b|: ok, reproduced it, but slime is apparently helpfully hiding all the interesting parts of the backtrace :/
2014-09-19T15:16:13Z Krystof: Xach: sorry, in the attempt to do this with a recent sbcl I've caused a cascade of software failure
2014-09-19T15:16:20Z Krystof: new sbcl -> new slime -> broken R
2014-09-19T15:16:35Z Krystof: need to fix that before I can look at drakma-async
2014-09-19T15:20:01Z Xach: no problem. i submitted a bug report to drakma-async. if i can help troubleshoot the sbcl bug let me know.
2014-09-19T15:21:05Z Krystof: ok, now I can get to recent quicklisp
2014-09-19T15:21:07Z loke_: Xach: Did you have some documentation on how to create a "stable" QL level?
2014-09-19T15:21:19Z Xach: loke_: what is a level?
2014-09-19T15:21:20Z loke_: I only found the post where you said you've made it easier, but now how it's done
2014-09-19T15:21:34Z loke_: Xach: "i want these versions, and not newer"
2014-09-19T15:21:35Z Xach: Which post?
2014-09-19T15:21:45Z loke_: One of your blog posts
2014-09-19T15:22:13Z Xach: loke_: you can specify a specific quicklisp dist version during installation, or go to that version afterwards.
2014-09-19T15:22:18Z loke_: http://blog.quicklisp.org/2014/01/consistent-configuration-with-quicklisp.html
2014-09-19T15:22:24Z eudoxia: "going back in dist time" i think
2014-09-19T15:23:08Z Xach: Going back in dist time is it.
2014-09-19T15:23:27Z Xach: loke_: it's not exactly what you're describing it, but it's what quicklisp offers.
2014-09-19T15:24:02Z Krystof: well, bother
2014-09-19T15:24:39Z yacks joined #sbcl
2014-09-19T15:24:43Z Krystof: after all of that, loading drakma-async worked
2014-09-19T15:25:12Z Xach: fooey
2014-09-19T15:25:29Z |3b| can get an error from it in slime, but not see the useful part of the backtrace :/
2014-09-19T15:25:45Z Krystof: |3b|: how did you get an error?
2014-09-19T15:26:06Z |3b|: (asdf:load-systems 'drakma-async) in *slime-scratch*
2014-09-19T15:27:00Z Krystof: and that's with drakma-1.3.10 in local-projects, and latest quicklisp otherwise?
2014-09-19T15:27:08Z |3b| doesn't use quicklisp :/
2014-09-19T15:27:33Z |3b|: fresh git pull of cffi, drakma and all the othercredence stuff
2014-09-19T15:27:50Z Xach: Krystof: are you using :verbose t?
2014-09-19T15:29:04Z Krystof: ah, no
2014-09-19T15:29:24Z Krystof: oh, if it muffles the style-warning.  Duh.
2014-09-19T15:29:46Z Krystof: thanks
2014-09-19T15:29:54Z |3b|: and same without slime, and sbcl debugger also hides the backtrace :p
2014-09-19T15:30:48Z qbit_ joined #sbcl
2014-09-19T15:30:59Z mood_ joined #sbcl
2014-09-19T15:31:15Z faheem joined #sbcl
2014-09-19T15:31:35Z mood quit (Ping timeout: 250 seconds)
2014-09-19T15:31:36Z faheem__1 quit (Ping timeout: 250 seconds)
2014-09-19T15:31:37Z jackdaniel quit (Ping timeout: 252 seconds)
2014-09-19T15:31:37Z brucem quit (Ping timeout: 252 seconds)
2014-09-19T15:31:37Z Perlboy quit (Ping timeout: 252 seconds)
2014-09-19T15:31:38Z qbit quit (Ping timeout: 250 seconds)
2014-09-19T15:31:39Z p_l|backup quit (Ping timeout: 250 seconds)
2014-09-19T15:31:43Z brucem_ joined #sbcl
2014-09-19T15:31:49Z jackdaniel joined #sbcl
2014-09-19T15:31:57Z p_l|backup joined #sbcl
2014-09-19T15:32:28Z Perlboy joined #sbcl
2014-09-19T15:32:55Z Krystof: |3b|: (sb-debug:print-backtrace :from :interrupted-frame)
2014-09-19T15:34:17Z Krystof: you probably also want :count 30
2014-09-19T15:35:07Z |3b|: that gets pretty much what Xach posted, is there any way to see arguments/locals for those frames though?
2014-09-19T15:35:31Z |3b|: or locals i guess
2014-09-23T23:06:50Z ccl-logbot joined #sbcl
2014-09-23T23:06:50Z 
2014-09-23T23:06:50Z names: ccl-logbot drmeiste_ scymtym_ Bicyclidine edgar-rft ebrasca oleo DGASAU fikusz Bike attila_lendvai nyef minion karswell` yacks prxq christoph_debian fitzsim nicdev specbot hlavaty alchemis7 leo2007 heddwch akkad Posterdati antoszka faheem__1 milosn weissschloss pkhuong p_l flip214 joshe irsol Perlboy p_l|backup jackdaniel brucem mood qbit tmh_ reb`` @Krystof yauz_2 scymtym Hydan` psilord Intensity jdz PuercoPop luis redline6561 pchrist fe[nl]ix Blkt foom
2014-09-23T23:06:50Z names: carvite thoto_ |3b| asedeno les ivan`` cracauer_ ams White_Flame sobel leoc phf ferada loke jsnell
2014-09-23T23:07:11Z Bicyclidine quit (Ping timeout: 244 seconds)
2014-09-23T23:10:17Z karswell` quit (Ping timeout: 272 seconds)
2014-09-23T23:14:50Z karswell` joined #sbcl
2014-09-23T23:23:35Z karswell` is now known as karswell
2014-09-23T23:29:48Z drmeiste_ quit (Remote host closed the connection)
2014-09-23T23:31:14Z leo2007 quit (Quit: rcirc on GNU Emacs 24.4.50.1)
2014-09-23T23:37:41Z drmeiste_ joined #sbcl
2014-09-23T23:41:28Z leo2007 joined #sbcl
2014-09-23T23:47:02Z prxq_ joined #sbcl
2014-09-23T23:50:37Z prxq quit (Ping timeout: 245 seconds)
2014-09-23T23:55:07Z ebrasca quit (Quit: ChatZilla 0.9.90.1 [Firefox 24.0/20131118140013])
2014-09-24T00:08:21Z attila_lendvai quit (Quit: Leaving.)
2014-09-24T00:47:36Z Antonia_Volkman joined #sbcl
2014-09-24T00:55:48Z leo2007 quit (Quit: rcirc on GNU Emacs 24.4.50.1)
2014-09-24T01:01:06Z leo2007 joined #sbcl
2014-09-24T01:01:39Z leo2007 quit (Client Quit)
2014-09-24T01:03:33Z momo-reina joined #sbcl
2014-09-24T01:04:07Z momo-reina quit (Remote host closed the connection)
2014-09-24T01:08:25Z leo2007 joined #sbcl
2014-09-24T01:08:38Z prxq joined #sbcl
2014-09-24T01:13:09Z prxq_ quit (Ping timeout: 272 seconds)
2014-09-24T01:25:29Z Antonia_Volkman quit (Remote host closed the connection)
2014-09-24T01:44:32Z Anastasia10 joined #sbcl
2014-09-24T01:58:31Z scymtym_ quit (Remote host closed the connection)
2014-09-24T02:20:27Z karswell quit (Read error: Connection reset by peer)
2014-09-24T02:22:39Z karswell joined #sbcl
2014-09-24T02:38:30Z christoph_debian quit (Ping timeout: 244 seconds)
2014-09-24T02:38:57Z Anastasia10 quit (Remote host closed the connection)
2014-09-24T02:49:11Z Amaya_Larson61 joined #sbcl
2014-09-24T02:52:42Z christoph_debian joined #sbcl
2014-09-24T03:16:08Z nyef quit (Quit: G'night all.)
2014-09-24T03:32:28Z gingerale joined #sbcl
2014-09-24T03:39:20Z Amaya_Larson61 quit (Remote host closed the connection)
2014-09-24T03:41:56Z Ana_Feil joined #sbcl
2014-09-24T04:11:35Z ccl-logbot joined #sbcl
2014-09-24T04:11:35Z 
2014-09-24T04:11:35Z names: ccl-logbot Intensity Ana_Feil gingerale christoph_debian karswell prxq leo2007 drmeiste_ edgar-rft oleo DGASAU fikusz Bike minion yacks fitzsim nicdev specbot hlavaty alchemis7 heddwch akkad Posterdati antoszka faheem__1 milosn weissschloss pkhuong p_l flip214 joshe irsol Perlboy p_l|backup jackdaniel brucem mood qbit tmh_ reb`` @Krystof yauz_2 scymtym Hydan` psilord jdz PuercoPop luis redline6561 pchrist fe[nl]ix Blkt foom jsnell loke ferada phf leoc
2014-09-24T04:11:35Z names: sobel White_Flame ams cracauer_ ivan`` les asedeno |3b| thoto_ carvite
2014-09-24T04:18:20Z Ana_Feil quit (K-Lined)
2014-09-24T04:48:52Z gingerale quit (Ping timeout: 260 seconds)
2014-09-24T05:15:28Z oleo quit (Quit: Verlassend)
2014-09-24T05:32:17Z sdemarre joined #sbcl
2014-09-24T05:35:39Z DGASAU quit (Remote host closed the connection)
2014-09-24T05:39:18Z DGASAU joined #sbcl
2014-09-24T05:45:05Z pranavrc joined #sbcl
2014-09-24T05:45:05Z pranavrc quit (Changing host)
2014-09-24T05:45:05Z pranavrc joined #sbcl
2014-09-24T06:00:59Z drmeiste_ quit (Remote host closed the connection)
2014-09-24T06:07:02Z sdemarre quit (Ping timeout: 250 seconds)
2014-09-24T06:26:27Z DGASAU quit (Remote host closed the connection)
2014-09-24T06:28:29Z DGASAU joined #sbcl
2014-09-24T06:49:44Z DGASAU quit (Ping timeout: 260 seconds)
2014-09-24T07:04:31Z angavrilov joined #sbcl
2014-09-24T07:28:35Z drmeiste_ joined #sbcl
2014-09-24T07:29:21Z fridim__ joined #sbcl
2014-09-24T07:33:24Z drmeiste_ quit (Ping timeout: 272 seconds)
2014-09-24T07:34:39Z segv- joined #sbcl
2014-09-24T07:54:33Z attila_lendvai joined #sbcl
2014-09-24T07:54:33Z attila_lendvai quit (Changing host)
2014-09-24T07:54:33Z attila_lendvai joined #sbcl
2014-09-24T08:06:24Z stassats joined #sbcl
2014-09-24T08:23:12Z attila_lendvai quit (Quit: Leaving.)
2014-09-24T08:23:42Z DGASAU joined #sbcl
2014-09-24T08:28:18Z leo2007 quit (Quit: rcirc on GNU Emacs 24.4.50.1)
2014-09-24T08:31:33Z stassats: regarding the problem in #lisp, what to do with bad slot options, but which are actually initargs for slot definitions?
2014-09-24T08:32:05Z stassats: (defclass pile () ((cards :initargs :cards))) doesn't signal an error
2014-09-24T08:33:23Z stassats: neither does (defclass pile () ((cards :class 10 :name 35 :readers t :writers -))), etc.
2014-09-24T08:33:42Z leo2007 joined #sbcl
2014-09-24T08:34:57Z stassats: for really internal initargs, non-keyword symbols can be used, but :writers, :readers and :initargs are supposed to be passed to ensure-class
2014-09-24T08:44:04Z stassats: so, convert :class, :allocate-class to sb-pcl::class, sb-pcl::allocation-class, disallow the appearance of :initargs :readers :writers :name :initfunction
2014-09-24T09:16:40Z drmeiste_ joined #sbcl
2014-09-24T09:21:29Z drmeiste_ quit (Ping timeout: 260 seconds)
2014-09-24T09:31:31Z edgar-rft quit (Quit: existence lost because of unknown reasons)
2014-09-24T10:40:12Z yacks quit (Ping timeout: 245 seconds)
2014-09-24T11:01:18Z fikusz quit (Ping timeout: 246 seconds)
2014-09-24T11:04:49Z drmeiste_ joined #sbcl
2014-09-24T11:09:22Z drmeiste_ quit (Ping timeout: 240 seconds)
2014-09-24T11:14:50Z fikusz joined #sbcl
2014-09-24T11:20:21Z drmeiste_ joined #sbcl
2014-09-24T11:29:19Z yacks joined #sbcl
2014-09-24T11:31:03Z nyef joined #sbcl
2014-09-24T11:31:53Z yacks quit (Client Quit)
2014-09-24T11:36:56Z yacks joined #sbcl
2014-09-24T12:20:32Z drmeiste_ quit (Remote host closed the connection)
2014-09-24T12:28:43Z drmeiste_ joined #sbcl
2014-09-24T12:37:05Z drmeiste_ quit (Remote host closed the connection)
2014-09-24T12:43:32Z drmeiste_ joined #sbcl
2014-09-24T12:52:09Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-24T12:56:03Z yacks quit (Read error: Connection reset by peer)
2014-09-24T12:56:13Z drmeiste_ joined #sbcl
2014-09-24T13:04:12Z yacks joined #sbcl
2014-09-24T13:26:39Z eudoxia joined #sbcl
2014-09-24T13:33:41Z oleo joined #sbcl
2014-09-24T13:38:40Z attila_lendvai joined #sbcl
2014-09-24T13:38:40Z attila_lendvai quit (Changing host)
2014-09-24T13:38:40Z attila_lendvai joined #sbcl
2014-09-24T13:41:42Z eudoxia_ joined #sbcl
2014-09-24T13:45:05Z eudoxia quit (Ping timeout: 260 seconds)
2014-09-24T13:59:49Z drmeiste_ quit (Read error: Connection reset by peer)
2014-09-24T14:06:10Z akkad quit (Ping timeout: 272 seconds)
2014-09-24T14:06:36Z drmeiste_ joined #sbcl
2014-09-24T14:11:48Z akkad joined #sbcl
2014-09-24T14:21:23Z pranavrc quit
2014-09-24T14:22:57Z drmeiste_ is now known as drmeister_
2014-09-24T14:41:23Z eudoxia_ quit (Quit: Lost terminal)
2014-09-24T14:44:46Z drmeister_ quit (Remote host closed the connection)
2014-09-24T14:50:20Z drmeiste_ joined #sbcl
2014-09-24T14:55:28Z drmeiste_ is now known as drmeister_
2014-09-24T14:59:28Z gingerale joined #sbcl