00:07:11 -!- fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has quit [Quit: Leaving] 00:19:52 fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has joined #sbcl 00:53:51 LiamH [~none@pool-173-73-124-231.washdc.east.verizon.net] has joined #sbcl 01:06:22 prxq_ [~mommer@mnhm-590c1c6b.pool.mediaWays.net] has joined #sbcl 01:09:32 pranavrc [~pranavrc@122.164.185.180] has joined #sbcl 01:09:43 -!- pranavrc [~pranavrc@122.164.185.180] has quit [Changing host] 01:09:43 pranavrc [~pranavrc@unaffiliated/pranavrc] has joined #sbcl 01:09:50 -!- prxq [~mommer@mnhm-5f75e29f.pool.mediaWays.net] has quit [Ping timeout: 264 seconds] 01:18:51 -!- davazp [~user@48.Red-79-150-239.dynamicIP.rima-tde.net] has quit [Ping timeout: 260 seconds] 01:43:37 echo-area [~user@182.92.247.2] has joined #sbcl 02:08:01 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 02:12:13 pkhuong: Are you online? 02:35:04 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 260 seconds] 02:39:20 -!- christoph_debian [~christoph@ppp-188-174-86-178.dynamic.mnet-online.de] has quit [Ping timeout: 256 seconds] 02:42:31 -!- LiamH [~none@pool-173-73-124-231.washdc.east.verizon.net] has quit [Quit: Leaving.] 02:52:50 christoph_debian [~christoph@ppp-188-174-43-98.dynamic.mnet-online.de] has joined #sbcl 03:03:14 attila_lendvai [~attila_le@92.47.225.171] has joined #sbcl 03:03:14 -!- attila_lendvai [~attila_le@92.47.225.171] has quit [Changing host] 03:03:14 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 03:23:48 psilord [~psilord@dyn-72-33-151-67.uwnet.wisc.edu] has joined #sbcl 03:57:37 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Read error: Connection reset by peer] 03:57:43 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 04:00:05 -!- yacks [~py@103.6.159.103] has quit [Quit: Leaving] 04:15:51 -!- fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has quit [Ping timeout: 276 seconds] 04:35:06 -!- pranavrc [~pranavrc@unaffiliated/pranavrc] has quit [Quit: Ping timeout: ] 05:20:40 pranavrc [~pranavrc@122.164.185.180] has joined #sbcl 05:20:40 -!- pranavrc [~pranavrc@122.164.185.180] has quit [Changing host] 05:20:40 pranavrc [~pranavrc@unaffiliated/pranavrc] has joined #sbcl 05:22:57 -!- eeezkil [~eeezkil@unaffiliated/eeezkil] has quit [Quit: ^D] 05:54:09 milosn_ [~milosn@user-5af50ad3.broadband.tesco.net] has joined #sbcl 05:56:30 -!- milosn [~milosn@user-5af5053c.broadband.tesco.net] has quit [Ping timeout: 264 seconds] 06:14:25 benkard [~benkard@2001:4ca0:0:f230:214c:6d34:d98c:454e] has joined #sbcl 06:15:00 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 245 seconds] 06:15:56 attila_lendvai [~attila_le@92.47.225.171] has joined #sbcl 06:15:56 -!- attila_lendvai [~attila_le@92.47.225.171] has quit [Changing host] 06:15:56 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 06:18:09 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Client Quit] 06:18:46 momo-reina [~user@112.198.192.129] has joined #sbcl 06:35:41 angavrilov [~angavrilo@217.71.227.190] has joined #sbcl 06:50:47 -!- Vivitron [~Vivitron@50.172.44.193] has quit [Ping timeout: 260 seconds] 07:20:54 -!- benkard [~benkard@2001:4ca0:0:f230:214c:6d34:d98c:454e] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 07:28:19 benkard [~benkard@2001:4ca0:0:f230:d59f:116d:ab63:1f4d] has joined #sbcl 07:34:08 attila_lendvai [~attila_le@87.247.13.98] has joined #sbcl 07:34:08 -!- attila_lendvai [~attila_le@87.247.13.98] has quit [Changing host] 07:34:08 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 07:45:28 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 264 seconds] 07:52:45 momo-rei` [~user@180.190.181.101] has joined #sbcl 07:55:01 -!- momo-reina [~user@112.198.192.129] has quit [Ping timeout: 245 seconds] 07:57:54 attila_lendvai [~attila_le@87.247.61.240] has joined #sbcl 07:57:54 -!- attila_lendvai [~attila_le@87.247.61.240] has quit [Changing host] 07:57:54 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 08:17:34 -!- momo-rei` [~user@180.190.181.101] has quit [Remote host closed the connection] 08:22:55 -!- ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has quit [Remote host closed the connection] 08:23:09 ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has joined #sbcl 08:28:40 -!- ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has quit [Ping timeout: 264 seconds] 09:12:28 -!- naryl [~weechat@46.182.24.168] has quit [Quit: WeeChat 0.4.0] 09:20:21 -!- pranavrc [~pranavrc@unaffiliated/pranavrc] has quit [Quit: Ping timeout: ] 09:31:02 -!- kludge` [~comet@unaffiliated/espiral] has quit [Ping timeout: 240 seconds] 09:36:25 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 09:42:21 -!- benkard [~benkard@2001:4ca0:0:f230:d59f:116d:ab63:1f4d] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 10:22:16 -!- echo-area [~user@182.92.247.2] has quit [Remote host closed the connection] 11:10:16 benkard [~benkard@2001:4ca0:0:f230:8cba:f56a:fe9f:fa6b] has joined #sbcl 11:40:31 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Quit: Leaving.] 11:45:55 ASau` [~user@p5797F31C.dip0.t-ipconnect.de] has joined #sbcl 11:49:28 -!- ASau [~user@p5797EFEE.dip0.t-ipconnect.de] has quit [Ping timeout: 260 seconds] 12:19:51 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 12:28:38 stassats [~stassats@wikipedia/stassats] has joined #sbcl 12:41:10 drmeister [~drmeister@166.137.84.213] has joined #sbcl 12:48:31 teggi [~teggi@113.173.4.29] has joined #sbcl 12:49:08 tylergoza [~tylergoza@72.29.34.246] has joined #sbcl 13:01:04 segv- [~mb@95-91-241-18-dynip.superkabel.de] has joined #sbcl 13:36:56 -!- drmeister [~drmeister@166.137.84.213] has quit [Remote host closed the connection] 13:39:52 drmeister [~drmeister@166.137.84.213] has joined #sbcl 14:00:14 -!- drmeister [~drmeister@166.137.84.213] has quit [Remote host closed the connection] 14:13:18 attila_lendvai [~attila_le@92.47.225.171] has joined #sbcl 14:13:18 -!- attila_lendvai [~attila_le@92.47.225.171] has quit [Changing host] 14:13:18 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 14:16:42 -!- teggi [~teggi@113.173.4.29] has quit [Remote host closed the connection] 14:16:43 drmeister [~drmeister@farnsworth.chem.temple.edu] has joined #sbcl 14:55:00 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 245 seconds] 15:04:19 who would like to review an sxhash-related patch? 15:04:22 http://paste.lisp.org/display/139003 15:05:09 lp 309443 15:05:09 https://bugs.launchpad.net/bugs/309443 15:09:15 attila_lendvai [~attila_le@92.46.5.54] has joined #sbcl 15:09:15 -!- attila_lendvai [~attila_le@92.46.5.54] has quit [Changing host] 15:09:15 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 15:10:15 pranavrc [~pranavrc@122.164.209.113] has joined #sbcl 15:10:15 -!- pranavrc [~pranavrc@122.164.209.113] has quit [Changing host] 15:10:15 pranavrc [~pranavrc@unaffiliated/pranavrc] has joined #sbcl 15:11:26 Krystof: xorred constant could be expanded to 64 bit as well, I guess. 15:12:01 true. There's a lot of 32-bitism in sxhash generally 15:12:41 as far as we understand modarith these days, the automatic shrinkening of the constants should Just Work 15:12:45 right? 15:13:04 yes. 15:13:21 but there needs to be a modarith around the mul-add as well 15:13:26 -!- gko [gko@2400:8900::f03c:91ff:fe70:e605] has quit [Ping timeout: 240 seconds] 15:14:02 nicdev` [user@2600:3c03::f03c:91ff:fedf:4986] has joined #sbcl 15:14:10 I put one there, yes 15:14:35 reb`` [user@nat/google/x-jpalidtvifpgpyww] has joined #sbcl 15:14:35 ok. That ought to fix the comments on needing (safety 0). The code was just wrong. 15:14:47 -!- reb` [user@nat/google/x-lvcotyijafqrznps] has quit [Ping timeout: 240 seconds] 15:15:29 -!- foom [~jknight@2620:15c:6:14:be30:5bff:fedf:6db6] has quit [Ping timeout: 240 seconds] 15:15:36 -!- nicdev [user@2600:3c03::f03c:91ff:fedf:4986] has quit [Ping timeout: 240 seconds] 15:15:41 yeah 15:15:45 good good 15:17:55 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 245 seconds] 15:18:13 bletch. Of course it turns out that the new sb-int:mix is four times slower 15:18:26 -!- reb`` [user@nat/google/x-jpalidtvifpgpyww] has quit [Remote host closed the connection] 15:18:37 reb`` [user@nat/google/x-cxaixefsvwzzcgmm] has joined #sbcl 15:22:30 multiplication by three is converted into a LEA 15:23:00 and a 32 bit constant will be a bit faster than a RIP-relative 64 bit one, in logxor. 15:25:02 yeah, the difference is probably lea vs imul/and 15:25:23 I suppose I could test the difference on that well known "build-sbcl" benchmark 15:26:45 the lack of logand around the muladd in HEAD makes code a bit worse than it could be, as well. 15:27:29 I suppose the logand around muladd could be word-sized 15:28:55 foom [~jknight@2620:15c:6:14:be30:5bff:fedf:6db6] has joined #sbcl 15:29:45 also with the better constant maybe we don't need the (ash xy -5) any more? 15:30:09 not with mod power-of-two, I don't think we can skip that. 15:30:11 the better multiply, that is 15:30:55 84.68user 3.47system vs 84.65user 3.54system # at least this benchmark is basically unaffected 15:31:19 without ash, bits only spread upward. 15:32:27 right 15:32:57 -!- Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has quit [Quit: Reconnecting] 15:33:01 and *sigh* I have been beaten by output/build-config 15:33:04 better rebenchmark 15:33:14 Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has joined #sbcl 15:36:04 (still about 85s user 3s sys, though) 15:39:46 but something is very weird 15:42:28 the version with the new MIX conses 5 times as much as the old one on the test code from lp 309443 15:44:00 rpg [~rpg@216.243.156.16.real-time.com] has joined #sbcl 15:44:42 conses or collides? 15:46:16 conses. There are no collisions from the new one 15:46:23 and the first test code or the second? 15:47:19 the test snippet in the description will make bad hash functions that are almost the identity work very nicely. 15:47:34 the first test code, but with 16 replaced by 1024 15:47:58 I suppose if there are no collisions the hash table has to grow more 15:48:04 it's kind of expected. 15:49:40 also, do you think it's worth the trouble to make mix work on signed fixnums (which entails using mask-signed-field instead of logand/ldb), and only mask the sign bit away when returning from sxhash? 15:51:14 ok, yes, consing mystery solved (was indeed hash table expansion) 15:51:52 latest measurement indicates the new version is only 2.8x slower 15:52:03 "only" 15:52:49 stassats [~stassats@wikipedia/stassats] has joined #sbcl 15:52:52 that's not a bad idea 15:53:16 bit tedious to implement 15:54:40 precomputing the conses and shuffling them may make the slow down disappear. Bad hash functions have really nice access patterns on sequential-ish queries. 15:55:27 (good hash function are equally mediocre on everything ;) 15:58:45 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 245 seconds] 16:04:41 Krystof: I bet removing the two useless logand in sxhash of fixnum can help compensate for the slowdown. 16:06:29 stassats [~stassats@wikipedia/stassats] has joined #sbcl 16:11:36 -!- reb`` [user@nat/google/x-cxaixefsvwzzcgmm] has quit [Remote host closed the connection] 16:12:54 reb`` [user@nat/google/x-lnbvpzyfjwtxfdku] has joined #sbcl 16:13:53 with http://paste.lisp.org/display/139003#1 to try to get nonlocality going, I get new: 0.40s, old: 0.14s 16:14:19 (without trying to make sxhash fixnum better) 16:15:25 (have run out of time, must travel) 16:17:39 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 260 seconds] 16:18:28 oh well. 16:25:54 -!- benkard [~benkard@2001:4ca0:0:f230:8cba:f56a:fe9f:fa6b] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 16:38:31 -!- ASau` is now known as ASau 16:44:37 gko [gko@2400:8900::f03c:91ff:fe70:e605] has joined #sbcl 16:51:23 -!- reb`` [user@nat/google/x-lnbvpzyfjwtxfdku] has quit [Remote host closed the connection] 16:51:37 reb`` [user@nat/google/x-ktleldpzxepwgddb] has joined #sbcl 16:55:40 -!- Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has quit [Ping timeout: 246 seconds] 16:58:36 ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has joined #sbcl 17:05:02 Bike [~Glossina@69.166.47.109] has joined #sbcl 17:35:40 cmack [~charlie@adsl-74-179-31-54.bna.bellsouth.net] has joined #sbcl 17:56:40 indeed at least one of the logands is useless and yet preserved 17:58:30 -!- pranavrc [~pranavrc@unaffiliated/pranavrc] has quit [Remote host closed the connection] 18:00:05 -!- Bike [~Glossina@69.166.47.109] has quit [Ping timeout: 248 seconds] 18:13:07 Vivitron [~Vivitron@c-50-172-44-193.hsd1.il.comcast.net] has joined #sbcl 18:26:33 yacks [~py@103.6.159.103] has joined #sbcl 18:45:52 looks like a -sb-unicode can't compile a +sb-unicode 18:46:36 looks like because using the host's char-code-limit 18:49:36 -!- tylergoza [~tylergoza@72.29.34.246] has quit [Ping timeout: 245 seconds] 18:52:41 -!- rpg [~rpg@216.243.156.16.real-time.com] has quit [Quit: rpg] 18:53:21 not just that, replacing it everywhere with sb!xc:char-code-limit didn't help 18:55:34 this is presumably a relatively new thing 18:55:37 (last year or so) 18:55:51 maybe related to the equalp things? 18:56:03 target-char uses char-code in the cross compiler 18:56:40 code-char, rather 18:57:31 i'm not actually exactly what, the backtrace is hosed by the long arrays 18:57:36 not sure 18:59:56 huh, setting the value of the **global**? 19:00:28 yeah, this frob 19:01:30 but what's it doing trying to evaluate the value form at compile-time? 19:02:37 i'll try redirecting output to a file and then looking at the backtrace 19:03:03 no, it's clearly that 19:03:31 i know that, but not which part of it 19:04:03 blimey, the debugger uses *terminal-io*, no redirection 19:04:28 **character-primary-compositions** 19:04:34 the initialization form for 19:06:27 you can redirect terminal-io, if you avoid having a tty 19:06:35 setsid sbcl ... > /tmp/foo 19:06:53 i just did (with-open-file (stream "/tmp/back" :direction :output :if-exists :overwrite :if-does-not-exist :create) (sb-debug:backtrace 1000 stream)) 19:07:13 and it's indeed the **character-primary-compositions** part 19:07:39 should it use sb!impl:code-char? 19:07:46 or whichever the package it's in 19:07:53 no, it should avoid evaluating that initialization form at xc time 19:08:08 something like (defglobal **character-primary-compositions** 'do-not-use) 19:08:35 no, wait 19:08:37 hold on 19:09:27 it should be put into the !charter-database-cold-init 19:09:58 and this whole form is scary 19:11:18 http://paste.lisp.org/display/139005 19:13:33 the comp.dat file encodes compositions as ,,, where each element is three octets being the codepoint in some endianness 19:13:53 the eventual lisp representation of that is a hash table with 42-bit keys and character values 19:15:00 i think you meant setf **character-primary-compositions**, not setf **character-long-decompositions**? 19:16:12 and can the hash table be #'eq on 64-bit? 19:17:38 I did mean **c-p-c**, yes. In principle it could be an EQ hash table on 64-bit (or indeed some cleverer data structure like a cuckoo hash or something) but at the moment, I wouldn't bother messing with it 19:18:10 and why is it present on -sb-unicode too, albeit empty 19:21:53 no characters compose on -sb-unicode 19:22:13 but you can still write (normalize ) 19:22:33 the functions and variables are still defined 19:23:01 and normalize-string is a no-op on sb-unicode 19:24:14 and what do you know, crashes in the warm-init 19:25:53 sorry 19:26:07 i'm now not sure whether it's due to this, or my replacing with sb!xc:char-code-limit 19:26:35 I can see one reason why it would crash 19:26:49 it's a segfault, at #x2da 19:27:29 *Krystof* tries again 19:28:18 http://paste.lisp.org/display/139005#1 19:31:21 Bike [~Glossina@wl-nat99.it.wsu.edu] has joined #sbcl 19:34:24 internal error #23 (An attempt was made to use an undefined FDEFINITION.) 19:34:37 oh, that's mine 19:34:44 ok, good :-) 19:34:54 that's what i was trying to build 19:39:56 and used a fast-sbcl for that, built with speed 3, safety 0, no sb-unicode, no sb-thread 19:44:19 *stassats* manages to lose another 32KB from the core 19:44:31 -!- angavrilov [~angavrilo@217.71.227.190] has quit [Remote host closed the connection] 19:44:32 each time i remove a bit of space, some new feature is added, and it's back 19:44:37 at least it doesn't grow 19:45:03 why have you been trying to reduce core size? do you need it for an application or something? 19:45:20 l_ [~l@84.233.246.170] has joined #sbcl 19:46:18 because smaller is better than larger, all else being equal 19:46:23 it's also usually faster 19:47:19 but without a proper tree shaker, it's nothing 19:48:38 the one significant, ~1MB, reduction is too complicated too commit 19:48:45 (the define-vop one) 19:51:36 -!- l_ [~l@84.233.246.170] has left #sbcl 19:57:38 i feel a rehaul of the assembler/disassembler could yield similar results 20:01:42 tylergoza [~tylergoza@72.29.34.246] has joined #sbcl 20:06:13 l_ [~l@84.233.246.170] has joined #sbcl 20:09:03 -!- drmeister [~drmeister@farnsworth.chem.temple.edu] has quit [Read error: Connection reset by peer] 20:09:09 edgar-rft [~GOD@HSI-KBW-109-193-013-113.hsi7.kabel-badenwuerttemberg.de] has joined #sbcl 20:09:13 drmeister [~drmeister@farnsworth.chem.temple.edu] has joined #sbcl 20:09:59 -!- l_ [~l@84.233.246.170] has left #sbcl 20:10:37 l_ [~l@84.233.246.170] has joined #sbcl 20:13:23 -!- l_ [~l@84.233.246.170] has left #sbcl 20:29:14 Fare [fare@nat/google/x-jfxgibowwjfjaeid] has joined #sbcl 20:40:14 Sagane [~Sagane@177.100-226-89.dsl.completel.net] has joined #sbcl 20:55:11 -!- drmeister [~drmeister@farnsworth.chem.temple.edu] has quit [Remote host closed the connection] 21:04:36 -!- Sagane [~Sagane@177.100-226-89.dsl.completel.net] has quit [Read error: Connection reset by peer] 21:46:22 -!- edgar-rft [~GOD@HSI-KBW-109-193-013-113.hsi7.kabel-badenwuerttemberg.de] has quit [Quit: session lost because continuation discontinued] 21:46:52 -!- Bike [~Glossina@wl-nat99.it.wsu.edu] has quit [Ping timeout: 246 seconds] 21:47:28 Bike [~Glossina@69.166.47.101] has joined #sbcl 22:47:39 rpg [~rpg@23-25-144-217-static.hfc.comcastbusiness.net] has joined #sbcl 22:57:02 -!- Bike [~Glossina@69.166.47.101] has quit [Ping timeout: 240 seconds] 23:10:12 -!- segv- [~mb@95-91-241-18-dynip.superkabel.de] has quit [Remote host closed the connection] 23:12:12 Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has joined #sbcl 23:35:45 -!- xymox [lechuck@unaffiliated/contempt] has quit [Ping timeout: 256 seconds] 23:42:29 xymox [lechuck@unaffiliated/contempt] has joined #sbcl 23:48:53 -!- xymox [lechuck@unaffiliated/contempt] has quit [Ping timeout: 248 seconds] 23:52:05 xymox [lechuck@unaffiliated/contempt] has joined #sbcl 23:58:28 -!- xymox [lechuck@unaffiliated/contempt] has quit [Ping timeout: 245 seconds] 23:58:59 -!- stassats [~stassats@wikipedia/stassats] has quit [Remote host closed the connection]