00:07:11 Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has joined #sbcl 00:12:08 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 00:54:50 -!- kwmiebach [~kwmiebach@xdsl-87-79-134-109.netcologne.de] has quit [Quit: Leaving] 01:17:02 -!- oleo [~oleo@xdsl-78-35-156-198.netcologne.de] has quit [Ping timeout: 264 seconds] 01:17:45 fisxoj [~fisxoj@24-212-142-77.cable.teksavvy.com] has joined #sbcl 01:17:47 oleo [~oleo@xdsl-78-35-178-228.netcologne.de] has joined #sbcl 01:17:54 -!- fisxoj [~fisxoj@24-212-142-77.cable.teksavvy.com] has quit [Read error: Connection reset by peer] 02:13:20 ASau` [~user@p5083DB6A.dip0.t-ipconnect.de] has joined #sbcl 02:16:42 -!- ASau [~user@p54AFFBA5.dip0.t-ipconnect.de] has quit [Ping timeout: 252 seconds] 02:38:31 teggi [~teggi@123.21.195.226] has joined #sbcl 02:43:38 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 02:44:16 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 02:48:37 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 248 seconds] 02:59:29 teggi_ [~teggi@123.21.195.226] has joined #sbcl 02:59:52 -!- teggi [~teggi@123.21.195.226] has quit [Read error: Connection reset by peer] 03:08:17 -!- yacks [~py@103.6.159.103] has quit [Read error: Connection reset by peer] 03:14:51 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 03:19:41 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 268 seconds] 03:22:51 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 03:31:46 switch/case helps optional didpatch, but makes it even clearer that jump tensioning must happen late 03:32:14 we currently have cmp/jne next test/jmp entry point 03:32:53 because there's a bunch of noop MOVE VOPs between jne and jmp 03:33:27 with switch case, we have a jump table into... a series of jmp short. :/ 03:33:54 it's not worse than what we have right now, but more obviously bad. 03:36:54 -!- teggi_ [~teggi@123.21.195.226] has quit [Read error: Connection reset by peer] 03:37:21 teggi [~teggi@123.21.195.226] has joined #sbcl 03:39:43 -!- christoph_debian [~christoph@ppp-88-217-61-111.dynamic.mnet-online.de] has quit [Ping timeout: 272 seconds] 03:53:20 christoph_debian [~christoph@ppp-46-244-235-213.dynamic.mnet-online.de] has joined #sbcl 04:19:40 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 04:20:19 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 04:24:59 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 260 seconds] 04:26:28 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 04:31:53 one solution is to implement jump-jump elimination in the assembler. the best I can say is that it works, with a tiny patch :\ 04:33:55 ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has joined #sbcl 04:44:53 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 04:45:31 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 04:49:45 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 240 seconds] 05:29:52 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 05:54:54 -!- ASau` is now known as ASau 06:13:16 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 06:48:09 sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has joined #sbcl 07:24:04 -!- ASau [~user@p5083DB6A.dip0.t-ipconnect.de] has quit [Remote host closed the connection] 07:26:10 ASau [~user@p5083DB6A.dip0.t-ipconnect.de] has joined #sbcl 07:55:25 -!- sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 245 seconds] 07:57:42 prxq [~mommer@x2f6cba8.dyn.telefonica.de] has joined #sbcl 08:05:09 stassats [~stassats@wikipedia/stassats] has joined #sbcl 08:20:42 yacks [~py@103.6.159.103] has joined #sbcl 08:29:58 attila_lendvai [~attila_le@5.251.151.176] has joined #sbcl 08:29:58 -!- attila_lendvai [~attila_le@5.251.151.176] has quit [Changing host] 08:29:58 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 08:42:53 -!- angavrilov [~angavrilo@217.71.227.190] has quit [Ping timeout: 246 seconds] 08:50:50 -!- ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has quit [Ping timeout: 268 seconds] 08:53:52 ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has joined #sbcl 08:55:08 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 245 seconds] 08:55:38 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 08:58:12 -!- ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has quit [Ping timeout: 240 seconds] 09:08:55 sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has joined #sbcl 09:10:02 edgar-rft [~GOD@HSI-KBW-109-193-013-113.hsi7.kabel-badenwuerttemberg.de] has joined #sbcl 09:17:49 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 272 seconds] 09:34:29 attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has joined #sbcl 09:41:31 angavrilov [~angavrilo@217.71.227.190] has joined #sbcl 10:34:28 -!- teggi [~teggi@123.21.195.226] has quit [Remote host closed the connection] 10:34:56 -!- kludge` [~comet@unaffiliated/espiral] has quit [Ping timeout: 265 seconds] 10:37:12 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 10:51:16 holy deadlock batman 11:06:36 looks like concurrent access to hashtables is broke on windows 11:14:05 gko_ [~gko@114-32-172-194.HINET-IP.hinet.net] has joined #sbcl 11:20:14 InvalidCo [~invalidco@dsl-lhtbrasgw2-54f824-20.dhcp.inet.fi] has joined #sbcl 11:24:31 -!- sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 260 seconds] 11:25:58 in fact, pretty printing to a string stream causes a deadlock 11:28:22 and each thread printing into its own stream 11:30:14 but only conses 11:48:46 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 11:58:03 -!- gko_ [~gko@114-32-172-194.HINET-IP.hinet.net] has quit [Ping timeout: 252 seconds] 11:59:00 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 12:04:12 oh wow, it deadlocks when allocating structures 12:12:46 -!- echo-area [~user@114.254.99.176] has quit [Remote host closed the connection] 12:13:27 echo-area [~user@114.254.99.176] has joined #sbcl 12:13:29 eudoxia [~eudoxia@r190-135-69-243.dialup.adsl.anteldata.net.uy] has joined #sbcl 12:16:42 -!- eudoxia [~eudoxia@r190-135-69-243.dialup.adsl.anteldata.net.uy] has quit [Read error: No route to host] 12:27:38 sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has joined #sbcl 12:28:18 -!- echo-area [~user@114.254.99.176] has quit [Remote host closed the connection] 12:28:58 echo-area [~user@114.254.99.176] has joined #sbcl 12:30:36 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Ping timeout: 245 seconds] 12:47:00 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 252 seconds] 12:53:20 -!- fikusz [~fikusz@catv-89-132-137-62.catv.broadband.hu] has quit [Quit: Leaving] 12:59:15 fikusz [~fikusz@catv-89-132-137-62.catv.broadband.hu] has joined #sbcl 13:02:15 stassats [~stassats@wikipedia/stassats] has joined #sbcl 13:04:23 leuler [~user@p548F894B.dip0.t-ipconnect.de] has joined #sbcl 13:09:45 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 13:14:21 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 245 seconds] 13:31:35 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 13:37:13 -!- oleo [~oleo@xdsl-78-35-178-228.netcologne.de] has quit [Ping timeout: 245 seconds] 13:37:32 oleo [~oleo@xdsl-78-35-189-219.netcologne.de] has joined #sbcl 13:47:33 hargettp [~hargettp@c-65-96-162-255.hsd1.ma.comcast.net] has joined #sbcl 14:54:45 leoc [~leoc.git@p5DDBBFFD.dip0.t-ipconnect.de] has joined #sbcl 15:12:23 -!- hargettp [~hargettp@c-65-96-162-255.hsd1.ma.comcast.net] has quit [Quit: Linkinus - http://linkinus.com] 15:20:04 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 15:26:07 -!- yacks [~py@103.6.159.103] has quit [Ping timeout: 268 seconds] 15:31:29 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 15:39:14 yacks [~py@103.6.159.103] has joined #sbcl 15:56:21 -!- Vivitron [~Vivitron@c-50-172-44-193.hsd1.il.comcast.net] has quit [Ping timeout: 248 seconds] 16:12:09 oh hello, "There is no applicable method for the generic function # when called with arguments (# steps to reproduce? 16:49:34 it lost accessor-method-p check 16:50:34 wait, my rewrite. Oops 16:51:59 sorry. Yes, not careful enough, but whatever it is you did it isn't in the test suite 17:01:08 Krystof: http://paste.lisp.org/display/139968 17:02:12 i can fix it, if you want 17:14:35 -!- sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 252 seconds] 17:20:16 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 17:21:58 -!- leoc [~leoc.git@p5DDBBFFD.dip0.t-ipconnect.de] has quit [Remote host closed the connection] 17:42:14 ltt_ [~ltt_@189-19-120-48.dsl.telesp.net.br] has joined #sbcl 17:57:19 -!- kludge` [~comet@unaffiliated/espiral] has quit [Quit: leaving] 17:58:15 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 18:47:15 can it be true that the runtime locks on win32 are broken? 18:47:50 i see two threads locking succesfully the gc lock at the same time 18:48:16 Sagane [~Sagane@177.100-226-89.dsl.completel.net] has joined #sbcl 18:48:39 -!- edgar-rft [~GOD@HSI-KBW-109-193-013-113.hsi7.kabel-badenwuerttemberg.de] has quit [Quit: connection closed into permanent something] 18:55:41 hargettp [~hargettp@c-65-96-162-255.hsd1.ma.comcast.net] has joined #sbcl 18:56:19 -!- hargettp [~hargettp@c-65-96-162-255.hsd1.ma.comcast.net] has quit [Client Quit] 19:02:50 either the log is bullshitting me, or it's a free reign for all threads 19:29:53 -!- Sagane [~Sagane@177.100-226-89.dsl.completel.net] has quit [Ping timeout: 252 seconds] 20:09:21 So, getting jump-jump elimination right seems hopeless to me. We do some best effort stuff by ordering basic blocks just so, but it's impossible to tell when a basic block will definitely be empty until after regalloc 20:18:14 pkhuong: That's unfortunate. 20:22:12 -!- ltt_ [~ltt_@189-19-120-48.dsl.telesp.net.br] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 20:23:01 heddwch [~yoshi@76.8.3.189] has joined #sbcl 20:33:43 that, and we have VOPs (type tests, mostly) with internal conditional jumps 20:34:22 that's the only reason I'm even considering straightening some of these out in the assembler[D. 20:51:02 ltt_ [~ltt_@189-19-120-48.dsl.telesp.net.br] has joined #sbcl 21:00:18 -!- ltt_ [~ltt_@189-19-120-48.dsl.telesp.net.br] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 21:09:02 hargettp [~hargettp@c-65-96-162-255.hsd1.ma.comcast.net] has joined #sbcl 21:16:32 -!- samskulls [~user@S0106001111de1fc8.cg.shawcable.net] has quit [Read error: Connection reset by peer] 21:27:46 ltt_ [~ltt_@189-19-120-48.dsl.telesp.net.br] has joined #sbcl 21:34:20 stassats: if you can commit the obvious fix for the accessor-method that would be helpful 21:34:28 I'm half-way through adapting defstruct 21:34:51 if you can't get to it then send mail / open launchpad bug 21:35:05 will do 21:37:16 thank you 21:42:02 -!- hargettp [~hargettp@c-65-96-162-255.hsd1.ma.comcast.net] has quit [Quit: Leaving...] 22:02:39 Vivitron [~Vivitron@c-50-172-44-193.hsd1.il.comcast.net] has joined #sbcl 22:05:54 huh, callbacks from foreign threads out of nowhere 22:09:25 that's probably my C-c C-c 22:10:35 it is 22:11:42 lol 22:15:34 leuler: what do you think of https://github.com/pkhuong/sbcl/tree/sb-assem-jump-jump ? 22:16:11 looks like it actually locks in sub-gc 22:17:24 it is, and that's for tomorrow 22:17:26 Or everyone else... On one hand, I don't think it's possible for SBCL to avoid all jump to unconditional jumps without something like this; on the other, it doesn't feel right (but less wrong than attempting to handle that in a peephole pass like dougk suggested). 22:22:30 sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has joined #sbcl 22:22:49 pkhuong: I will take a look tomorrow. I am currently still investigating the regalloc changes. 22:23:02 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 264 seconds] 22:23:03 leuler: thanks. 22:24:09 I feel like the regalloc changes are more of an enabler for smarter hacks so far. Targeting could/should mostly be handled by coalescing vertices, for example. 22:29:19 OK. Regarding the current state of the new register allocator, I have now made some observations with code compiled with (speed 3) optimizations which I would like to share. 22:29:53 I'm all ears (for 5 minutes, then I'll have to read the logs while waiting for my popcorn ;) 22:30:02 That is: three observations about code size and the same as yesterday about compilation speed (presumably). 22:32:51 I have compiled the regalloc branch, with proclaim-target-optimization changed to speed 3, compilation speed 0, safety 1, space 0, once with *register-allocation-method* set to :greedy, once with :iterative. 22:33:10 The core size of :iterative is 32 KiB less than for :greedy. 22:34:02 (that's exactly one page, so actual savings may be arbitrarily lower) 22:34:02 Some functions are larger in one build, some in the other. (I will put up a paste with examples later). 22:34:56 (I thought a page was 4 KiB. Did I miss a change there?) 22:35:35 Depends on architecture and kernel 22:36:41 My observation about cl-bench's CLOS/complex-methods stands: This time it's more than five times as slow in :iterative than in :greedy. I fear I don't want to dig into this but if some CLOS stuff compiles so much slower that would be bad IMO. 22:39:45 An additional observation: The run time of several tests is 5 or 10 percent slower with :iterative, presumably again due to compilation times increasing. 22:40:03 allocation pages are 32k 22:41:11 I appreciate the hint towards enabling smarter hacks. Besides this, where would I currently need to look to see immediate advantages? Higher speed in some inner loops in complicated functions? 22:44:33 abarch reportes nice gains on arrayprocessing loops for cl bench 22:44:45 samskulls [~user@S0106001111de1fc8.cg.shawcable.net] has joined #sbcl 22:44:50 i'm waiting for confirmation i didn't break that 23:05:58 pkhuong: See http://paste.lisp.org/display/139973 23:16:45 -!- sdemarre [~serge@31.103-64-87.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 252 seconds] 23:24:21 interesing... i'll see if i can think of anything simple/obvious. 23:30:28 I need to quit now. I'll be back tomorrow. 23:30:38 -!- leuler [~user@p548F894B.dip0.t-ipconnect.de] has quit [Quit: Good night!] 23:33:29 -!- brucem [~bmitchene@waywardmonkeys.com] has quit [Quit: ZNC - http://znc.sourceforge.net] 23:34:34 brucem [~bmitchene@waywardmonkeys.com] has joined #sbcl