00:25:08 -!- Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 00:29:53 -!- kclifton [~kclifton@s198-166-45-245.ab.hsia.telus.net] has quit [Quit: kclifton] 01:11:16 -!- redline6561 [~redline@adsl-232-119-3.asm.bellsouth.net] has quit [Ping timeout: 240 seconds] 01:21:30 -!- Fare [~Fare@ita4fw1.itasoftware.com] has quit [Quit: Leaving] 01:51:19 redline6561 [~redline@c-66-56-55-169.hsd1.ga.comcast.net] has joined #sbcl 02:21:19 Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #sbcl 02:21:19 -!- ChanServ has set mode +o Krystof 02:24:57 -!- homie [~user@xdsl-87-79-53-196.netcologne.de] has quit [Ping timeout: 265 seconds] 02:30:18 homie [~user@xdsl-84-44-252-66.netcologne.de] has joined #sbcl 02:45:44 tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #sbcl 03:38:10 -!- tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Quit: Leaving.] 03:45:54 antgreen [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has joined #sbcl 06:20:33 stassats [~stassats@wikipedia/stassats] has joined #sbcl 06:45:03 attila_lendvai [~attila_le@catv-89-132-188-47.catv.broadband.hu] has joined #sbcl 07:06:40 -!- attila_lendvai [~attila_le@catv-89-132-188-47.catv.broadband.hu] has left #sbcl 07:07:39 -!- antgreen [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has quit [Remote host closed the connection] 07:07:40 antgreen` [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has joined #sbcl 07:16:50 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 07:53:56 -!- cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has quit [Ping timeout: 255 seconds] 07:54:46 cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has joined #sbcl 08:10:44 -!- cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has quit [Ping timeout: 265 seconds] 08:11:30 cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has joined #sbcl 08:15:35 jeti [~user@p54B473A6.dip.t-dialin.net] has joined #sbcl 08:27:32 -!- ASau [~user@89-178-104-155.broadband.corbina.ru] has quit [Remote host closed the connection] 08:27:58 ASau [~user@89-178-104-155.broadband.corbina.ru] has joined #sbcl 09:19:39 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #sbcl 09:19:39 -!- ChanServ has set mode +o nikodemus 09:43:53 -!- Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 09:54:31 -!- slyrus [~chatzilla@adsl-75-36-222-218.dsl.pltn13.sbcglobal.net] has quit [Read error: Connection reset by peer] 10:01:31 slyrus [~chatzilla@adsl-75-36-222-218.dsl.pltn13.sbcglobal.net] has joined #sbcl 10:18:41 tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #sbcl 10:40:51 -!- cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has quit [Ping timeout: 265 seconds] 10:41:30 cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has joined #sbcl 10:46:52 Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #sbcl 10:46:52 -!- ChanServ has set mode +o Krystof 10:48:11 some questions. If you have *read-default-float-format* 'rational, what would you like to see on printing? 10:49:15 also, where is the time taken (about 2-3ms) in a gc of an empty nursery? 10:56:31 first - no idea 10:56:42 second: scavenging dirty pages 10:57:03 Krystof: (rational (let ((*read-default-float-format* 'double-float)) (read s))) 10:57:58 nikodemus: hm, what about no dirty pages also? 10:58:03 (two successive gcs) 10:58:49 you can still have dirty pages in old generations unless you run a full gc 10:59:28 fe[nl]ix: hm, that's reading. Though that is an interesting distinction from the base-10 rational/*r-d-f-f* I've implemented 10:59:35 can't mark them clean unless everything above them has been collected 11:00:15 (progn (gc :full t) (time (gc))) gives me run times of the order of 1-3ms 11:00:27 let me see 11:00:33 oh, no, sorry, real time 11:00:38 no, sometimes run time too 11:00:54 2,077,371 processor cycles 11:01:42 i see. i suspect that's the time it takes to walk the page-table and find out the nursery is empty 11:02:10 is there a good way of measuring this? 11:02:44 Krystof: the standard specifies that the value of *r-d-f-f* must be a type specifier, otherwise it would be perhaps more useful to support both 'rational and 'rationalize 11:02:53 beyond adding instrumentation to the runtime or using "standard" c profiling tools? no that i know 11:03:08 I don't know what standard c profiling tools are any good 11:03:11 gprof? 11:04:50 depends on what you need ... valgrind/cachegrind are very nice 11:05:14 i'm not sure there are any good standard tools. there was an lwn article about the situation a while back, plus a small controversy of course :) 11:05:23 last I tried valgrind on sbcl, admittedly years ago, it took longer than the age of the universe to set up its data structures to track sbcl's address space 11:06:06 huh, that has changed 11:06:46 grrr. i correctly split the commit, but left out the split part in the original commit message 11:09:18 Krystof, you can tell valgrind *not* to check from the start, but only from a certain point on 11:26:23 -!- tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 11:27:13 -!- Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 11:30:33 -!- ASau [~user@89-178-104-155.broadband.corbina.ru] has quit [Remote host closed the connection] 11:31:27 ASau [~user@89-178-104-155.broadband.corbina.ru] has joined #sbcl 11:42:14 -!- hefner [~hefner@ppp-58-9-182-3.revip2.asianet.co.th] has quit [Ping timeout: 255 seconds] 11:48:40 hefner [~hefner@ppp-58-9-119-114.revip2.asianet.co.th] has joined #sbcl 12:03:43 Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 12:03:43 -!- ChanServ has set mode +o Krystof 12:08:03 -!- Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 12:16:11 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 276 seconds] 12:24:03 Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 12:24:03 -!- ChanServ has set mode +o Krystof 12:24:27 -!- ASau [~user@89-178-104-155.broadband.corbina.ru] has quit [Remote host closed the connection] 12:25:17 ASau [~user@89-178-104-155.broadband.corbina.ru] has joined #sbcl 12:28:28 -!- Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 12:29:44 Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 12:29:44 -!- ChanServ has set mode +o Krystof 12:34:18 -!- Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 13:10:03 stassats [~stassats@wikipedia/stassats] has joined #sbcl 13:10:37 Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 13:10:37 -!- ChanServ has set mode +o Krystof 13:15:58 -!- Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 13:29:07 Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 13:29:17 -!- ChanServ has set mode +o Krystof 13:35:58 -!- Krystof [csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 13:49:46 Krystof [~csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 13:49:46 -!- ChanServ has set mode +o Krystof 13:51:24 -!- flip214 [~marek@unaffiliated/flip214] has quit [Remote host closed the connection] 13:57:14 -!- specbot [~specbot@common-lisp.net] has quit [Ping timeout: 255 seconds] 14:05:55 -!- hargettp [~hargettp@dhcp-174.mirrorimage.net] has quit [Remote host closed the connection] 14:06:17 hargettp [~hargettp@dhcp-174.mirrorimage.net] has joined #sbcl 14:29:41 51.33 8.09 8.09 59150583 0.00 0.00 scavenge 14:29:42 25.44 12.10 4.01 10007 0.40 1.43 garbage_collect_generation 14:30:16 this is one gc :full t and then a dotimes around (gc) 14:30:46 tcr [~tcr@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has joined #sbcl 14:46:13 -!- rpg [~rpg@216.243.156.16.real-time.com] has quit [Remote host closed the connection] 14:46:48 Krystof: do you have the cycle/call tree output as well? 14:49:32 yes 14:49:38 *Krystof* experiments with gc modifications 14:49:56 well, I don't any more, sorry 14:50:08 my rebuild has just clobbered the gprof output file 14:50:54 CORRUPTION WARNING in SBCL pid 31186(tid 3085153984): 14:50:54 Memory fault at 0 (pc=(nil), sp=0xb77a20c8) 14:50:54 The integrity of this image is possibly compromised. 14:50:57 well, there you go 14:52:19 let me put it back and show some more detailed profile output 14:55:33 writing 73728 bytes from the dynamic space at 0x09000000 14:55:33 done] 14:55:35 whoops 14:57:44 ok, an observation that I am utterly failing to exploit is that the page table starts with 6000-odd pages of pseudo-static generation 14:58:03 and iterating over those for every minor collection to check whether to do anything with them seems very pointless 15:00:50 Krystof pasted "gc loop profile" at http://paste.lisp.org/display/116840 15:02:23 Krystof annotated #116840 "call profiles" at http://paste.lisp.org/display/116840#1 15:05:06 -!- antgreen` [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has quit [Read error: Connection reset by peer] 15:06:24 antgreen` [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has joined #sbcl 15:10:12 udzinari [~opera@nat/ibm/x-xvvnfzjcsqduqaem] has joined #sbcl 15:11:18 -!- udzinari [~opera@nat/ibm/x-xvvnfzjcsqduqaem] has left #sbcl 15:11:20 udzinari [~opera@nat/ibm/x-xvvnfzjcsqduqaem] has joined #sbcl 15:16:43 homie` [~user@xdsl-84-44-252-124.netcologne.de] has joined #sbcl 15:19:18 -!- homie [~user@xdsl-84-44-252-66.netcologne.de] has quit [Ping timeout: 245 seconds] 15:24:57 -!- homie` [~user@xdsl-84-44-252-124.netcologne.de] has quit [Quit: none] 15:37:52 homie [~user@xdsl-84-44-252-124.netcologne.de] has joined #sbcl 15:38:16 Krystof: IIRC a (loop (gc)) benchmark also has the problem that we accumulate ridiculous amounts of pinned pages due to the forced gcs 15:54:52 nyef [~nyef@pool-64-222-179-143.man.east.myfairpoint.net] has joined #sbcl 15:59:38 that might be the problem 15:59:45 a problem 15:59:53 -!- gnooth [~gnooth@ip98-176-79-151.sd.sd.cox.net] has quit [Quit: Leaving] 16:01:00 Hello all. 16:01:01 Dare I ask what problem? 16:01:01 (And why don't we have a channel topic, anyway?) 16:01:38 was there a netsplit recently ? 16:03:41 doing (time (dotimes (i 10000) (dotimes (j 10000) (foo 3)))) with foo = (lambda (x) (cons x x)) gives 0.2s GC time, and 60 garbage collections 16:04:01 so, again, about 3 milliseconds per gc 16:04:29 this time not with forced gcs, and with essentially pure garbage in the nursery 16:06:10 gnooth [~gnooth@ip98-176-79-151.sd.sd.cox.net] has joined #sbcl 16:06:52 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Quit: This computer has gone to sleep] 16:08:25 -!- nyef [~nyef@pool-64-222-179-143.man.east.myfairpoint.net] has quit [Ping timeout: 245 seconds] 16:08:53 -!- Krystof [~csr21@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 16:11:47 nyef [~nyef@pool-64-222-179-143.man.east.myfairpoint.net] has joined #sbcl 16:23:31 -!- kae_ [~dsa@ext02.gsp.se] has quit [Ping timeout: 250 seconds] 16:42:57 rpg [~rpg@mpls.sift.info] has joined #sbcl 16:45:38 -!- tcr [~tcr@cpc1-bour5-2-0-cust921.15-1.cable.virginmedia.com] has quit [Quit: Leaving.] 16:51:58 kae_ [~dsa@ext02.gsp.se] has joined #sbcl 17:04:37 -!- cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has quit [Ping timeout: 265 seconds] 17:05:21 cmm [~cmm@bzq-79-181-201-133.red.bezeqint.net] has joined #sbcl 17:13:34 -!- jeti [~user@p54B473A6.dip.t-dialin.net] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 17:28:15 -!- antgreen` [~user@CPE00222d6c4710-CM00222d6c470d.cpe.net.cable.rogers.com] has quit [Remote host closed the connection] 18:52:48 Fare [~Fare@ita4fw1.itasoftware.com] has joined #sbcl 19:22:47 Blkt [~user@net-93-151-226-225.cust.dsl.teletu.it] has joined #sbcl 19:23:35 good evening everyone 19:23:43 Hello Blkt. 19:24:08 hi nyef 19:32:54 antgreen [~user@nat/redhat/x-nlcswrapkzfoljgr] has joined #sbcl 20:02:39 DanLentz [~danlentz@c-68-32-54-29.hsd1.nj.comcast.net] has joined #sbcl 20:03:18 -!- ASau [~user@89-178-104-155.broadband.corbina.ru] has quit [Remote host closed the connection] 20:04:53 ASau [~user@89-178-104-155.broadband.corbina.ru] has joined #sbcl 20:13:18 -!- DanLentz [~danlentz@c-68-32-54-29.hsd1.nj.comcast.net] has quit [Quit: Colloquy for iPhone - http://colloquy.mobi] 20:24:01 -!- hargettp [~hargettp@dhcp-174.mirrorimage.net] has quit [Ping timeout: 240 seconds] 20:38:42 DanLentz [~danlentz@c-68-32-54-29.hsd1.nj.comcast.net] has joined #sbcl 20:45:30 -!- nyef [~nyef@pool-64-222-179-143.man.east.myfairpoint.net] has quit [Ping timeout: 245 seconds] 20:47:37 nyef [~nyef@pool-64-222-179-143.man.east.myfairpoint.net] has joined #sbcl 20:48:35 -!- DanLentz [~danlentz@c-68-32-54-29.hsd1.nj.comcast.net] has quit [Quit: Colloquy for iPhone - http://colloquy.mobi] 20:50:02 -!- rpg [~rpg@mpls.sift.info] has quit [Remote host closed the connection] 20:59:32 rpg [~rpg@mpls.sift.info] has joined #sbcl 21:08:51 Krystof [~csr21@csrhodes.plus.com] has joined #sbcl 21:08:51 -!- ChanServ has set mode +o Krystof 21:33:14 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 276 seconds] 22:13:18 -!- antgreen [~user@nat/redhat/x-nlcswrapkzfoljgr] has quit [Remote host closed the connection] 22:17:01 -!- rpg [~rpg@mpls.sift.info] has quit [Remote host closed the connection]