00:12:30 -!- slyrus [~chatzilla@173-228-44-88.dsl.static.sonic.net] has quit [Ping timeout: 250 seconds] 00:44:49 hargettp [~hargettp@96.237.121.111] has joined #sbcl 01:37:29 echo-area [~user@114.251.86.0] has joined #sbcl 01:39:34 -!- tcr [~tcr@217-162-131-235.dynamic.hispeed.ch] has quit [Quit: Leaving.] 01:44:06 Fare [~Fare@64.119.159.126] has joined #sbcl 01:58:33 redline6561 [~user@c-66-56-55-169.hsd1.ga.comcast.net] has joined #sbcl 01:58:46 -!- hargettp [~hargettp@96.237.121.111] has quit [Quit: Leaving...] 02:32:11 -!- echo-area [~user@114.251.86.0] has quit [Ping timeout: 276 seconds] 03:03:47 echo-area [~user@114.251.86.0] has joined #sbcl 03:28:03 -!- antgreen [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has quit [Remote host closed the connection] 04:19:07 -!- gor[e] [~svr@79.165.187.105] has quit [Read error: Operation timed out] 05:05:13 slyrus [~chatzilla@adsl-76-254-45-145.dsl.pltn13.sbcglobal.net] has joined #sbcl 05:22:39 -!- ignotus [~ignotus@unaffiliated/ignotus] has quit [Quit: Coyote finally caught me] 05:27:33 flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has joined #sbcl 05:27:33 -!- flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has quit [Changing host] 05:27:33 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 05:31:18 flip215 [~marek@h081217084238.dyn.cm.kabsi.at] has joined #sbcl 05:31:18 -!- flip215 [~marek@h081217084238.dyn.cm.kabsi.at] has quit [Changing host] 05:31:18 flip215 [~marek@unaffiliated/flip214] has joined #sbcl 05:34:06 -!- flip215 [~marek@unaffiliated/flip214] has quit [Read error: Operation timed out] 05:34:57 -!- flip214 [~marek@unaffiliated/flip214] has quit [Ping timeout: 276 seconds] 05:47:23 flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has joined #sbcl 05:47:23 -!- flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has quit [Changing host] 05:47:23 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 05:51:37 -!- luis [~luis@nhop.r42.eu] has quit [Quit: ZNC - http://znc.sourceforge.net] 05:52:09 luis [~luis@nhop.r42.eu] has joined #sbcl 06:39:18 -!- echo-area [~user@114.251.86.0] has quit [Ping timeout: 276 seconds] 06:41:38 -!- flip214 [~marek@unaffiliated/flip214] has quit [Ping timeout: 250 seconds] 06:41:56 flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has joined #sbcl 06:41:56 -!- flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has quit [Changing host] 06:41:56 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 06:44:58 echo-area [~user@114.251.86.0] has joined #sbcl 06:57:41 gor[e] [~svr@gw1.masterhost.ru] has joined #sbcl 07:17:29 tcr [~tcr@217-162-131-235.dynamic.hispeed.ch] has joined #sbcl 07:28:49 nikodemus [~nikodemus@cs181199216.pp.htv.fi] has joined #sbcl 07:28:49 -!- ChanServ has set mode +o nikodemus 07:53:09 woudshoo [~user@ipleiden.intellimagic.net] has joined #sbcl 07:53:18 -!- homie [~levgue@xdsl-78-35-170-253.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 08:03:15 homie [~levgue@xdsl-78-35-170-253.netcologne.de] has joined #sbcl 08:03:30 -!- Krystof [~csr21@csrhodes.plus.com] has quit [Ping timeout: 260 seconds] 08:09:02 attila_lendvai [~attila_le@145.38.broadband3.iol.cz] has joined #sbcl 08:09:02 -!- attila_lendvai [~attila_le@145.38.broadband3.iol.cz] has quit [Changing host] 08:09:02 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 08:25:05 -!- nikodemus [~nikodemus@cs181199216.pp.htv.fi] has quit [Ping timeout: 240 seconds] 08:25:36 stassats [~stassats@wikipedia/stassats] has joined #sbcl 08:39:57 -!- stassats [~stassats@wikipedia/stassats] has quit [Remote host closed the connection] 09:03:39 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has left #sbcl 09:06:44 scymtym [~user@2001:638:504:2093:21a:a0ff:fe34:2d7d] has joined #sbcl 09:18:33 Krystof [~csr21@158.223.161.59] has joined #sbcl 09:18:33 -!- ChanServ has set mode +o Krystof 09:26:55 woudshoo` [~user@ipleiden.intellimagic.net] has joined #sbcl 09:30:47 -!- woudshoo [~user@ipleiden.intellimagic.net] has quit [Ping timeout: 276 seconds] 09:36:28 -!- scymtym [~user@2001:638:504:2093:21a:a0ff:fe34:2d7d] has quit [Ping timeout: 264 seconds] 09:38:23 -!- homie [~levgue@xdsl-78-35-170-253.netcologne.de] has quit [Ping timeout: 240 seconds] 09:39:50 homie [~levgue@xdsl-78-35-174-74.netcologne.de] has joined #sbcl 09:55:07 stassats [~stassats@wikipedia/stassats] has joined #sbcl 10:25:51 nikodemus [~nikodemus@cs181199216.pp.htv.fi] has joined #sbcl 10:25:51 -!- ChanServ has set mode +o nikodemus 10:28:43 -!- Krystof [~csr21@158.223.161.59] has quit [Ping timeout: 240 seconds] 10:31:17 Krystof [~csr21@158.223.51.76] has joined #sbcl 10:31:17 -!- ChanServ has set mode +o Krystof 10:46:04 hargettp [~hargettp@96.237.121.111] has joined #sbcl 11:19:35 -!- hargettp [~hargettp@96.237.121.111] has quit [Quit: Leaving...] 11:21:52 hargettp [~hargettp@96.237.121.111] has joined #sbcl 11:49:15 -!- hargettp [~hargettp@96.237.121.111] has quit [Quit: Leaving...] 12:27:19 -!- flip214 [~marek@unaffiliated/flip214] has quit [Read error: Connection reset by peer] 12:27:31 flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has joined #sbcl 12:27:31 -!- flip214 [~marek@h081217084238.dyn.cm.kabsi.at] has quit [Changing host] 12:27:31 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 13:10:03 antgreen [~user@nat/redhat/x-ptnpustgwdmxqnsn] has joined #sbcl 13:16:12 -!- Krystof [~csr21@158.223.51.76] has quit [Read error: Operation timed out] 13:24:07 -!- woudshoo` [~user@ipleiden.intellimagic.net] has quit [Read error: Connection reset by peer] 13:46:52 Krystof [~csr21@158.223.51.76] has joined #sbcl 13:46:52 -!- ChanServ has set mode +o Krystof 13:59:21 kaek [~b@c-c3cae253.97-16-64736c12.cust.bredbandsbolaget.se] has joined #sbcl 13:59:55 -!- Krystof [~csr21@158.223.51.76] has quit [Ping timeout: 260 seconds] 14:32:21 -!- redline6561 [~user@c-66-56-55-169.hsd1.ga.comcast.net] has quit [Ping timeout: 240 seconds] 14:39:14 redline6561 [~user@c-66-56-55-169.hsd1.ga.comcast.net] has joined #sbcl 14:55:37 stassats` [~stassats@wikipedia/stassats] has joined #sbcl 14:56:49 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 246 seconds] 14:59:26 -!- Fare [~Fare@64.119.159.126] has quit [Ping timeout: 248 seconds] 15:11:18 woudshoo [~user@ipleiden.intellimagic.net] has joined #sbcl 15:14:50 -!- homie [~levgue@xdsl-78-35-174-74.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 15:41:15 When disassembling functions in SBCL, I often see long sequences of NOP's. Why is that? 15:41:33 alignment? 15:42:01 Of code? And if that's needed, why do I never see C-compilers generate any nops? 15:42:42 In this case I see 14 NOP's in a row. That makes no sense to me 15:42:48 does it matter what C-compilers you've seen do? 15:43:12 loke: alignment. 15:43:16 and C compilers do generate nops 15:43:29 Well, not really. I just used it as a justification for something that I understand was true 15:43:57 what they might do is generate smarter nops (no-op instructions with longer encodings) 15:44:03 pkhuong: they do, rarely. For code, however, why would alignment be needed? x64 has no alignment requirements for code. 15:44:04 nops are quite short, so it can take some number of them 15:44:27 along cache lines? 15:44:33 loke: Cache effects are a major reason. 15:44:37 loke: there's no requirement, but it helps eprformance. 15:44:39 there are ways to get multi-byte nops too, but simple 0x90 (on x86) get filtered out in the pipeline anyway - so the "cpu" never sees them 15:45:09 flip214: it still takes decoding overhead; for now, we always jump over them anyway. 15:45:35 yeah, I did notice they're always jumped over 15:45:38 perhaps the alignment is for tagged pointers to the functions and not the function calls? 15:45:52 pkhuong: TBH I'd be more worried about L1-cache and memory bandwidth 15:46:02 than about the overhead of NOPs 15:46:12 nyef: cache, but mostly BTB effects, since we can't align on cache lines. 15:46:28 Ah, okay. 15:46:30 I was thinking for a while that it had something to do with jump-tables, but I can see no computed jumps to the code that follows the nops 15:46:51 loke: i'm not speculating here. 15:46:55 pkhuong: what are BTB effects? 15:47:10 pkhuong: I know. I took a while to type that, didn't see your comments. 15:47:47 Did anyone do the work to change the alignment granularity for arrays so that SIMD instructions can reliably be used on x86? 15:48:29 aligning jump targets helps with performance; you can try and look at documentation for -falign-{jumps,loops} (which is enabled at -O >= 2 in gcc) 15:48:59 pkhuong: Ah, I will do that. Thanks 15:50:11 it's been a while, but the effect is that the destination of the jump is less likely to exhibit a random slowdown, and padding conditional/indirect jumps a bit helps reserve enough storage for branch predictors. 15:50:42 -!- echo-area [~user@114.251.86.0] has quit [Ping timeout: 246 seconds] 15:51:50 pkhuong: is this something that is employed for all optimisation levels? In particular, (optimize (size 3)) could remove it? 15:52:49 you need speed > space, iirc. 15:53:04 Ahh... Makes sense. Let me try :-) 15:54:37 homie [~levgue@xdsl-78-35-174-74.netcologne.de] has joined #sbcl 15:54:50 were the effects ever measured? 15:55:03 Hmm... (speed 0) (space 3) still generates it. Oh well, it's not like space-optimisations are very useful anyway 15:55:05 loop alignment is unconditional -- src/compiler/codegen.lisp 15:56:17 it didn't have a huge effect except for making many benchmarks less noisy 15:56:48 sometimes you "got lucky" when you compiled a function, then you recompiled it and the alignment went bad and results changed 15:57:04 results being the timings, that is 15:57:28 In general, I've been quite happy with the performance of SBCL. In fact, I'm sometimes surprised at just how good it is. However, I never actually tried many different lisps (mostly CLISP besides SBCL, which of course is painfully slow). My question is, how is SBCL performance compared to in particular the commercial offerings? 15:57:51 loke: very different. 15:58:29 different as in, faster, slower or faster on some, slower on some things? 15:58:57 loke: I have been told, repeatedly, that the commercial offerings are horrible, performancewise, compared to SBCL. I have never tested such things myself, however. 15:59:30 faster on some things, slower on some is almost certainly the case, though. 16:00:04 eg. i suspect that IO is likely to be faster with commercial offerings, out external formats being pretty slow and all 16:00:36 sorry, "character IO" -- no reason why binary IO would be wastly different 16:22:04 -!- hargettp_ [~hargettp_@dhcp-162.mirrorimage.net] has quit [Remote host closed the connection] 16:22:25 hargettp_ [~hargettp_@dhcp-162.mirrorimage.net] has joined #sbcl 16:33:31 -!- woudshoo [~user@ipleiden.intellimagic.net] has quit [Ping timeout: 250 seconds] 17:06:42 -!- gor[e] [~svr@gw1.masterhost.ru] has quit [Read error: Operation timed out] 17:40:41 rmarynch [~roman@88.135.194.233] has joined #sbcl 17:40:52 Good evening 17:41:08 Hello rmarynch. 17:53:21 -!- slyrus [~chatzilla@adsl-76-254-45-145.dsl.pltn13.sbcglobal.net] has quit [Quit: ChatZilla 0.9.86 [SeaMonkey 2.1b3pre/20110310003232]] 18:10:20 -!- homie [~levgue@xdsl-78-35-174-74.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 18:13:54 -!- pchrist [~spirit@gentoo/developer/pchrist] has quit [Quit: leaving] 18:33:23 Fare [~Fare@ita4fw1.itasoftware.com] has joined #sbcl 18:43:39 I have some problems with this issue https://bugs.launchpad.net/sbcl/+bug/573747 18:43:50 EVAL part is all clear and done 18:44:38 regarding the compiler, I decided to fix it by introducing IR1 translator for DECLARE 18:45:29 this translator signals compiler-error only 18:46:07 this approach works fine, except that DECLARE becomes a special operator 18:46:41 and the test (assert (not (special-operator-p 'declare))) fails (all other tests pass fine) 18:48:37 this is because def-ir-translator macro marks DECLARE (and other things which are defined using def-ir1-translator) as a special operator 18:49:12 since DECLARE is very much a special case, i think just hardcoding an (eq 'declare op) in ir1-convert-functoid or thereabouts would be ok 18:49:50 pchrist [~spirit@gentoo/developer/pchrist] has joined #sbcl 18:50:30 nikodemus: I am thinking about hardcoding the exceptional case in def-ir1-translator macro definition, so it does not marks DECLARE as a special operator 18:50:44 like (unless (eq name 'DECLARE)) 18:50:49 bah 18:50:51 looks ugly :) 18:51:47 rmarynch: don't 18:52:04 def-ir1-translater is far too big a hammer 18:53:49 granted, it would make M-. work on DECLARE -- but the result would not be useful, so i don't really see much benefit in using it 18:54:02 in principle, yes 19:11:26 gor[e] [~svr@79.165.187.105] has joined #sbcl 19:11:57 homie [~levgue@xdsl-78-35-174-74.netcologne.de] has joined #sbcl 19:12:13 -!- rmarynch [~roman@88.135.194.233] has quit [Quit: Leaving] 19:28:16 -!- homie [~levgue@xdsl-78-35-174-74.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 19:30:22 -!- flip214 [~marek@unaffiliated/flip214] has quit [Ping timeout: 248 seconds] 19:30:32 homie [~levgue@xdsl-78-35-174-74.netcologne.de] has joined #sbcl 19:35:13 mrSpec [~Spec@unaffiliated/mrspec] has joined #sbcl 19:35:15 Hello 20:04:15 scymtym [~user@2001:638:504:2093:21a:a0ff:fe34:2d7d] has joined #sbcl 20:15:40 rpg [~rpg@216.243.156.16.real-time.com] has joined #sbcl 20:17:35 prxq [~mommer@mnhm-5f75dd87.pool.mediaWays.net] has joined #sbcl 20:30:08 -!- hargettp_ [~hargettp_@dhcp-162.mirrorimage.net] has quit [Remote host closed the connection] 20:30:31 hargettp_ [~hargettp_@dhcp-162.mirrorimage.net] has joined #sbcl 20:31:29 -!- danlentz [~danlentz@c-68-46-98-23.hsd1.nj.comcast.net] has quit [Ping timeout: 246 seconds] 20:39:43 danlentz [~danlentz@c-68-46-98-23.hsd1.nj.comcast.net] has joined #sbcl 21:10:50 -!- gor[e] [~svr@79.165.187.105] has quit [Ping timeout: 276 seconds] 21:44:55 -!- antgreen [~user@nat/redhat/x-ptnpustgwdmxqnsn] has quit [Remote host closed the connection] 21:50:52 -!- mrSpec [~Spec@unaffiliated/mrspec] has quit [Quit: mrSpec] 22:10:36 -!- homie [~levgue@xdsl-78-35-174-74.netcologne.de] has quit [Read error: Operation timed out] 22:26:05 homie [~levgue@xdsl-78-35-166-226.netcologne.de] has joined #sbcl 22:33:42 -!- stassats` [~stassats@wikipedia/stassats] has quit [Ping timeout: 240 seconds] 22:39:14 -!- tcr [~tcr@217-162-131-235.dynamic.hispeed.ch] has quit [Quit: Leaving.] 22:39:34 hargettp [~hargettp@96.237.121.111] has joined #sbcl 22:55:12 -!- tsuru [~charlie@adsl-87-47-213.bna.bellsouth.net] has quit [Remote host closed the connection] 23:33:56 -!- homie [~levgue@xdsl-78-35-166-226.netcologne.de] has quit [Remote host closed the connection] 23:43:29 homie [~levgue@xdsl-78-35-166-226.netcologne.de] has joined #sbcl