00:13:56 The_Jon_Smith [~The_Jon_S@ip24-250-13-137.ri.ri.cox.net] has joined #sbcl 00:35:54 -!- attila_lendvai [~attila_le@catv-89-133-171-82.catv.broadband.hu] has quit [Quit: Leaving.] 00:42:58 *nyef* sighs. 00:43:36 Ever have that sudden realization that the build you started ten minutes earlier was a threaded build, and you needed a non-threaded build for the next phase of your project? 00:58:39 -!- tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Quit: Leaving.] 01:09:39 -!- hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has quit [Quit: hargettp] 01:11:18 hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has joined #sbcl 01:17:01 -!- Kaer [~b@c-cfcee253.97-16-64736c12.cust.bredbandsbolaget.se] has quit [Read error: Connection reset by peer] 01:40:37 rbarraud_ [~rbarraud@118-92-9-206.dsl.dyn.ihug.co.nz] has joined #sbcl 01:41:40 -!- rbarraud [~rbarraud@118-92-140-168.dsl.dyn.ihug.co.nz] has quit [Ping timeout: 265 seconds] 01:52:44 -!- nyef [~nyef@pool-70-109-134-127.cncdnh.east.myfairpoint.net] has quit [Quit: G'night all.] 02:00:35 -!- hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has quit [Quit: hargettp] 02:19:20 rbarraud__ [~rbarraud@118-92-153-145.dsl.dyn.ihug.co.nz] has joined #sbcl 02:20:17 -!- rbarraud_ [~rbarraud@118-92-9-206.dsl.dyn.ihug.co.nz] has quit [Ping timeout: 276 seconds] 03:55:55 -!- lisppaste2 [~lisppaste@common-lisp.net] has quit [Ping timeout: 240 seconds] 03:55:55 -!- minion [~minion@common-lisp.net] has quit [Ping timeout: 240 seconds] 06:46:17 mega1 [~quassel@catv4E5CABA2.pool.t-online.hu] has joined #sbcl 06:54:58 -!- Krystof [~csr21@78.146.236.83] has quit [Ping timeout: 240 seconds] 06:57:08 nikodemus [~nikodemus@cs181199216.pp.htv.fi] has joined #sbcl 06:57:08 -!- ChanServ has set mode +o nikodemus 07:42:19 tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #sbcl 07:51:11 nikodemus: thank you, my inbox was feeling a bit empty 07:52:58 always happy to oblidge ;) 07:53:09 (launchpad?) 07:54:02 Yes. 07:54:44 Also, are you really looking at dispatching more than one timer handler per signal? 07:55:33 joshe: yes 07:56:00 do you see an issue with that? i already have the fix pending commit 07:56:29 Neat. 07:57:19 I can't see any reason why it wouldn't work, just that I tried it several months ago and couldn't make it work for reasons which I never figured out. 07:57:30 http://github.com/nikodemus/SBCL # if you want to test on openbsd 07:58:18 Well once you commit then cron will let me know if it doesn't work on openbsd. 07:58:46 ok, cool 07:59:38 Well, on amd64 at least. 08:01:25 if someone wants to snipe an easy bug, https://bugs.launchpad.net/sbcl/+bug/561974 is a good candidate 08:22:40 Blkt [~user@93-33-130-201.ip44.fastwebnet.it] has joined #sbcl 08:41:29 attila_lendvai [~attila_le@catv-89-133-171-82.catv.broadband.hu] has joined #sbcl 08:49:22 cmm- [~cmm@bzq-79-180-201-14.red.bezeqint.net] has joined #sbcl 08:52:18 -!- cmm [~cmm@bzq-109-64-205-76.red.bezeqint.net] has quit [Ping timeout: 240 seconds] 09:02:11 commit frenzy over, cvs dongle free 09:09:28 cmpitg [~cmpitg@118.71.132.157] has joined #sbcl 09:15:58 -!- mega1 [~quassel@catv4E5CABA2.pool.t-online.hu] has quit [Read error: Operation timed out] 09:17:00 -!- cmpitg [~cmpitg@118.71.132.157] has quit [Quit: leaving] 09:20:01 cmpitg [~cmpitg@118.71.132.157] has joined #sbcl 09:20:50 -!- cmpitg [~cmpitg@118.71.132.157] has quit [Client Quit] 09:33:37 mega1 [~quassel@catv4E5CABA2.pool.t-online.hu] has joined #sbcl 10:08:03 hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has joined #sbcl 10:20:47 haha 10:21:58 nikodemus: you broke my nightly cvs build script beacuse, for the first time since I wrote my script, no patches are necessary for the build and tests to succeed 10:22:08 (well, maybe) 10:25:28 heh 10:35:41 -!- nikodemus [~nikodemus@cs181199216.pp.htv.fi] has quit [Quit: This computer has gone to sleep] 10:36:02 -!- rbarraud__ [~rbarraud@118-92-153-145.dsl.dyn.ihug.co.nz] has quit [Read error: Connection reset by peer] 10:37:44 \ 10:38:18 er 10:47:39 Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #sbcl 10:47:39 -!- ChanServ has set mode +o Krystof 10:50:17 -!- Blkt [~user@93-33-130-201.ip44.fastwebnet.it] has quit [Ping timeout: 265 seconds] 11:10:58 nyef [~nyef@pool-70-20-56-192.man.east.myfairpoint.net] has joined #sbcl 11:11:12 G'morning all. 11:41:25 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #sbcl 11:41:25 -!- ChanServ has set mode +o nikodemus 11:45:07 Is there an easy way to go from an (unsigned-byte 8) value to the name of a corresponding widetag? 11:45:16 -!- hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has quit [Quit: hargettp] 11:45:48 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Ping timeout: 240 seconds] 11:46:44 should be something for that in room.lisp 11:47:12 Ah. 11:47:14 *nyef* shudders. 11:47:15 Thanks. 11:50:52 hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has joined #sbcl 11:57:23 jennyf [~jennyf@212-198-248-35.rev.numericable.fr] has joined #sbcl 11:57:59 -!- jennyf [~jennyf@212-198-248-35.rev.numericable.fr] has quit [Client Quit] 12:08:00 nyef_ [~nyef@pool-70-20-56-192.man.east.myfairpoint.net] has joined #sbcl 12:09:33 -!- nyef [~nyef@pool-70-20-56-192.man.east.myfairpoint.net] has quit [Ping timeout: 265 seconds] 12:13:28 -!- nyef_ is now known as nyef 12:30:48 Blkt [~user@93-33-143-16.ip44.fastwebnet.it] has joined #sbcl 13:03:50 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #sbcl 13:03:50 -!- ChanServ has set mode +o nikodemus 13:22:17 Hello nikodemus. 13:25:43 minion [~minion@common-lisp.net] has joined #sbcl 13:26:16 specbot [~specbot@common-lisp.net] has joined #sbcl 13:27:43 minion: paste 115054, please? 13:27:43 Paste number 115054: "nikodemus: Does this disassembly work for you?" by nyef in None. http://paste.lisp.org/display/115054 13:34:43 (Just annotated with another disassembly, showing a feature probably not requested.) 13:41:46 KB [b@c-cfcee253.97-16-64736c12.cust.bredbandsbolaget.se] has joined #sbcl 13:45:12 nyef: wow, that looks really nice 13:46:04 Thanks. I've got an idea about what to do about TLS-block access on threaded SBCL as well. 13:47:45 nyef: looks excellent 13:51:34 nikodemus: thanks for closing the bugs. I managed to script pretty much everything else in the release process, but ran out of steam before figuring out how to send gpg signed email :-) 13:54:52 heh 13:57:35 oh, and of course a rather minimal amount of clicking in the sf.net interface of doom. I rather like the new system 14:00:35 the current new system or the beta? 14:01:31 the current one, I didn't dare to click on the beta link 14:16:43 stassats [~stassats@wikipedia/stassats] has joined #sbcl 14:20:15 -!- stassats [~stassats@wikipedia/stassats] has quit [Remote host closed the connection] 14:21:08 mbohun [~mbohun@ppp115-156.static.internode.on.net] has joined #sbcl 14:33:28 nikodemus: in your mail in "so each timer gets its own interrupt"... do you mean "so each THREAD gets its own interrupt"? 14:52:51 no, each timer that runs causes interrupt-thread to be called, no matter which thread the timer runs in 14:56:50 cmpitg [~cmpitg@118.71.132.157] has joined #sbcl 14:59:08 then i go back thinking more... :) 15:42:19 minion: paste 115062? 15:42:19 Paste number 115062: "more disassembly" by nyef in None. http://paste.lisp.org/display/115062 15:42:35 -!- cmpitg [~cmpitg@118.71.132.157] has left #sbcl 16:04:17 *nyef* sighs. 16:04:40 Okay, getting the x86 disassembler up to scratch is going to be harder than fixing the x86-64 disassembler was. 16:04:56 I'm tempted to just do the first couple fixes and ignore the last one. 16:12:24 hm, I wonder if any cmucl 20b improvements can be glommed for sbcl 16:13:00 Anything good? 16:13:24 fround improvements for single/double-float (non-x86) 16:13:36 faster utf8 input, though I don't know if that's easily grabbale 16:14:48 Sparc and PPC build improvements came from SBCL in the first place... 16:14:55 s/build/gc/ 16:17:28 I think we probably have or originated most of the interesting bugfixes. 16:18:28 Static-arrays might be interesting. 16:19:22 So... Just static-arrays, then? 16:21:53 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Quit: This computer has gone to sleep] 16:22:01 nikodemus: the new timer dispatch code seems to work fine for me 16:22:14 Oh, bye 16:27:01 joshe: you could ask minion to leave a memo for nikodemus 16:29:00 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #sbcl 16:29:12 -!- ChanServ has set mode +o nikodemus 16:32:13 drewc [~user@S01060013101b6ddb.vc.shawcable.net] has joined #sbcl 16:39:58 -!- attila_lendvai [~attila_le@catv-89-133-171-82.catv.broadband.hu] has quit [Quit: Leaving.] 16:43:50 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Quit: Leaving] 16:56:10 Oh thanks, I forgot about minion 16:57:11 minion: memo for nikodemus: the timer tests all pass for me on OpenBSD now 16:57:12 Remembered. I'll tell nikodemus when he/she/it next speaks. 17:16:13 -!- hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has quit [Quit: hargettp] 17:37:40 Fun discovery for today: At some point in the past, I went to the trouble of setting up a 32-bit debian chroot and making it obvious how to use it. 17:51:52 attila_lendvai [~attila_le@catv4E5CABA2.pool.t-online.hu] has joined #sbcl 17:59:41 apparently you didn't make it obvious where it was. :) 18:00:37 It was in ~root. 18:03:19 minion: paste 115066? 18:03:19 Paste number 115066: "more disassembly" by nyef in None. http://paste.lisp.org/display/115066 18:04:06 ... Something's not right with that. 18:06:20 0110009f and 0110000b... The second of those is NIL, clearly, and the first has to be *CURRENT-CATCH-BLOCK*, and then there's *ALIEN-STACK*... 18:07:32 Okay, so it -half- works. :-/ 18:08:19 ... Because I patched the wrong arg-type? 18:18:40 -!- angavrilov [~angavrilo@217.71.227.181] has quit [Remote host closed the connection] 18:28:51 -!- attila_lendvai [~attila_le@catv4E5CABA2.pool.t-online.hu] has quit [Quit: Leaving.] 18:56:09 unenana [~unenana@212-198-248-35.rev.numericable.fr] has joined #sbcl 18:58:26 Fare [~Fare@ita4fw1.itasoftware.com] has joined #sbcl 18:58:33 hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has joined #sbcl 18:59:07 -!- unenana [~unenana@212-198-248-35.rev.numericable.fr] has quit [Client Quit] 19:27:34 well, I guess the large number of failures in the threads tests on darwin/x86-64 is a testament to the quality of the threads tests :) 19:28:19 Heh. 19:28:52 that's positive thinking for you! 19:52:13 -!- Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 20:08:58 Heh. Amusing thing seen in a disassembly note: "; #" 20:10:33 -!- ASau [~user@83.69.227.32] has quit [Ping timeout: 245 seconds] 20:12:11 it's too bad sbcl doesn't support 32bit modular arith on x86-64. 20:12:41 wait, it does not? 20:13:12 tcr: Why would it? 64-bit registers, remember? 20:13:19 tcr: as fixnum arith that's then masked away. 20:13:40 it would because 32bit multiplies and adds are cheaper 20:13:54 Ideally, we would only use 32 bit registers. 20:14:46 Hrm... Maybe I should put that on my list of things to look into, then? 20:15:10 dunno, it seems like it'd be nearly impossible without fixing so much of the compiler 20:15:45 Okay, so I should put it on my list of things to look into on-spec, then? 20:15:57 foom: it's mostly a representation selection issue, actually. If the TNs are unboxed, then it's not too hard to look at the type when emitting VOPs. 20:15:57 right now, it prefers doing everything in fixnums if it fits, even if working in ub64 would've been better. 20:16:22 it's pretty easy to do, actually 20:16:32 rep selection seems to pretty much suck in sbcl. 20:16:46 *nyef* is knee-deep in the x86 disassembler right now, but should have some of his bandwidth back "soon". 20:17:18 you just need to set up the modular arith bits to convert * into sb-vm::*-mod32 or whatever 20:17:25 foom: pretty much. It's an interesting problem, but I'd never dream of integrating any of the approaches I would develop for a research project. 20:17:27 And speaking of... Should it be FS:[#x1C] or [FS:#x1C]? 20:17:29 then write the vop for that to take unsigned-reg arguments only 20:17:51 froydnj: and that'd suck, because it would immediately convert from and back to fixnum 20:18:06 since 32 < 60 20:18:15 foom: depending on the usage context, yes, it would 20:18:17 foom: depends. If you only performed mod arith on unboxed input, it wouldn't. 20:18:40 but you can't do anything on unboxed values of 32bits, it almost *always* puts them into fixnum rep 20:19:03 it'll choose fixnum vops, since they're cheaper. 20:19:05 but froydnj's right, rep selection is mostly orthogonal to 32 bit mod arith, actually. 20:19:21 *froydnj* has a patch for doing it lying about... 20:19:43 I doubt it'd actually make anything faster without fixing rep selection and vop selection first though 20:20:17 IIRC, it made ironclad's md5 and sha1 significantly faster 20:20:39 I didn't check something like twofish, but I imagine it'd have helped there, too 20:36:16 -!- hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has quit [Quit: hargettp] 20:53:14 ... and the votes (all one of them) are in, FS:[#x1C] it is. 20:55:38 hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has joined #sbcl 21:15:12 -!- hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has quit [Quit: hargettp] 21:19:03 hargettp [~anonymous@pool-71-184-188-218.bstnma.east.verizon.net] has joined #sbcl 21:26:44 -!- mega1 [~quassel@catv4E5CABA2.pool.t-online.hu] has quit [Ping timeout: 276 seconds] 21:27:36 Ah, that lovely moment of realization: You're about halfway through a build when you realize that you made some of your changes to the wrong backend. 21:28:31 ... and hurriedly move them from one backend to the other in the hopes that you can finish up before the file in question gets compiled. 21:33:05 kill -STOP the build until you finish? 21:33:29 Nah. I'm fairly sure I made it in time. 21:33:43 Failing that, I can redefine the function in question in the running system. 21:34:10 (Can, and will, most likely, as there's at least one more change to make after this.) 21:36:13 -!- Blkt [~user@93-33-143-16.ip44.fastwebnet.it] has quit [Read error: Connection reset by peer] 21:37:17 Hrm. I appear to have screwed up one or more of my instruction definitions. :-/ 21:38:12 Blkt [~user@93-33-143-16.ip44.fastwebnet.it] has joined #sbcl 21:40:55 Two of them, in fact. 21:41:01 ... wait, that shouldn't be possible. 21:41:43 Oh well. 21:48:41 -!- Blkt [~user@93-33-143-16.ip44.fastwebnet.it] has quit [Ping timeout: 255 seconds] 21:55:45 attila_lendvai [~attila_le@catv-89-133-171-82.catv.broadband.hu] has joined #sbcl 22:00:54 -!- mbohun [~mbohun@ppp115-156.static.internode.on.net] has quit [Quit: Leaving] 22:01:37 Kaer [b@c-cfcee253.97-16-64736c12.cust.bredbandsbolaget.se] has joined #sbcl 22:05:14 -!- KB [b@c-cfcee253.97-16-64736c12.cust.bredbandsbolaget.se] has quit [Ping timeout: 276 seconds] 22:26:23 -!- drewc [~user@S01060013101b6ddb.vc.shawcable.net] has quit [Remote host closed the connection] 22:26:45 drewc [~user@S01060013101b6ddb.vc.shawcable.net] has joined #sbcl 22:40:02 rbarraud [~rbarraud@118-92-153-145.dsl.dyn.ihug.co.nz] has joined #sbcl 22:44:00 -!- attila_lendvai [~attila_le@catv-89-133-171-82.catv.broadband.hu] has left #sbcl 22:59:42 -!- slyrus__ [~chatzilla@dsl092-019-253.sfo1.dsl.speakeasy.net] has quit [Remote host closed the connection] 23:00:40 -!- tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Quit: Leaving.] 23:10:09 minion: memo for nikodemus: http://repo.or.cz/w/sbcl/nyef.git/shortlog/refs/heads/x86oid-disassembler-fixes has a small pile of disassembler fixes, for both x86 and x86-64. I'm thinking to commit them this weekend, but wouldn't mind feedback beforehand. 23:10:09 Remembered. I'll tell nikodemus when he/she/it next speaks. 23:12:30 That goes for the rest of you x86oid-using people. Better disassembly is good, right? 23:18:53 -!- tsuru [~charlie@c-174-50-217-160.hsd1.tn.comcast.net] has quit [Remote host closed the connection] 23:58:20 -!- specbot [~specbot@common-lisp.net] has quit [Ping timeout: 240 seconds]