00:31:34 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 00:36:04 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 00:38:54 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 256 seconds] 00:45:45 -!- edgar-rft [~GOD@HSI-KBW-109-193-013-113.hsi7.kabel-badenwuerttemberg.de] has quit [Quit: lifeform experiment terminated because connection lost] 00:53:21 -!- Bike [~Glossina@wl-nat99.it.wsu.edu] has quit [Ping timeout: 245 seconds] 00:57:55 Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has joined #sbcl 01:11:18 stassats [~stassats@wikipedia/stassats] has joined #sbcl 01:53:07 fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has joined #sbcl 02:38:54 -!- christoph_debian [~christoph@ppp-188-174-0-133.dynamic.mnet-online.de] has quit [Read error: Operation timed out] 02:52:55 christoph_debian [~christoph@ppp-188-174-71-230.dynamic.mnet-online.de] has joined #sbcl 03:27:06 -!- yacks [~py@103.6.159.103] has quit [Ping timeout: 264 seconds] 03:35:25 -!- scymtym [~user@ip-5-147-115-29.unitymediagroup.de] has quit [Ping timeout: 245 seconds] 03:42:56 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 260 seconds] 03:42:56 pranavrc [~pranavrc@unaffiliated/pranavrc] has joined #sbcl 03:46:44 -!- kludge` [~comet@unaffiliated/espiral] has quit [Read error: Operation timed out] 03:46:52 yacks [~py@103.6.159.103] has joined #sbcl 03:52:35 -!- Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has quit [Quit: don't feel like it] 04:10:24 -!- yacks [~py@103.6.159.103] has quit [Read error: Connection reset by peer] 04:11:01 yacks [~py@103.6.159.103] has joined #sbcl 04:14:08 Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has joined #sbcl 04:51:46 ASau` [~user@p5797F072.dip0.t-ipconnect.de] has joined #sbcl 04:54:58 -!- ASau [~user@p4FF97137.dip0.t-ipconnect.de] has quit [Read error: Operation timed out] 05:05:54 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 05:39:24 sdemarre [~serge@122.98-64-87.adsl-dyn.isp.belgacom.be] has joined #sbcl 05:55:03 -!- fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has quit [Ping timeout: 260 seconds] 06:27:29 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 06:45:25 -!- pranavrc [~pranavrc@unaffiliated/pranavrc] has quit [Ping timeout: 248 seconds] 07:05:25 -!- sdemarre [~serge@122.98-64-87.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 245 seconds] 07:41:21 mc40 [~mc@146.255.107.98] has joined #sbcl 07:50:04 -!- mc40 [~mc@146.255.107.98] has quit [Read error: No route to host] 08:04:30 benkard [~benkard@2001:4ca0:0:f240:2da1:423b:5e63:d618] has joined #sbcl 08:25:53 Kromitvs [~quassel@mcqueen.rnl.ist.utl.pt] has joined #sbcl 09:34:11 -!- kludge` [~comet@unaffiliated/espiral] has quit [Ping timeout: 245 seconds] 09:40:00 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 09:55:03 pranavrc [~pranavrc@122.164.253.237] has joined #sbcl 09:55:03 -!- pranavrc [~pranavrc@122.164.253.237] has quit [Changing host] 09:55:03 pranavrc [~pranavrc@unaffiliated/pranavrc] has joined #sbcl 10:20:07 -!- ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has quit [Ping timeout: 260 seconds] 10:26:16 -!- pranavrc [~pranavrc@unaffiliated/pranavrc] has quit [Ping timeout: 245 seconds] 10:27:21 pranavrc [~pranavrc@122.164.122.207] has joined #sbcl 10:27:21 -!- pranavrc [~pranavrc@122.164.122.207] has quit [Changing host] 10:27:21 pranavrc [~pranavrc@unaffiliated/pranavrc] has joined #sbcl 10:35:42 Sagane [~Sagane@177.100-226-89.dsl.completel.net] has joined #sbcl 10:57:06 -!- yacks [~py@103.6.159.103] has quit [Ping timeout: 245 seconds] 11:22:52 yacks [~py@103.6.159.103] has joined #sbcl 11:27:04 -!- yacks [~py@103.6.159.103] has quit [Max SendQ exceeded] 11:27:23 yacks [~py@103.6.159.103] has joined #sbcl 11:29:58 -!- segv- [~mb@95-91-243-225-dynip.superkabel.de] has quit [Read error: Connection reset by peer] 11:35:50 -!- benkard [~benkard@2001:4ca0:0:f240:2da1:423b:5e63:d618] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 11:35:55 -!- kludge` [~comet@unaffiliated/espiral] has quit [Ping timeout: 248 seconds] 11:40:17 stassats [~stassats@wikipedia/stassats] has joined #sbcl 11:45:51 scymtym [~user@ip-5-147-115-29.unitymediagroup.de] has joined #sbcl 11:50:43 kludge` [~comet@unaffiliated/espiral] has joined #sbcl 11:54:42 -!- yacks [~py@103.6.159.103] has quit [Ping timeout: 264 seconds] 11:55:26 l_ [~l_@84.233.246.170] has joined #sbcl 11:56:52 yacks [~py@103.6.159.103] has joined #sbcl 12:00:09 -!- l_ [~l_@84.233.246.170] has left #sbcl 12:25:01 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 245 seconds] 12:45:43 sdemarre [~serge@91.180.110.162] has joined #sbcl 13:31:40 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 13:33:13 LiamH [~none@pool-173-73-124-231.washdc.east.verizon.net] has joined #sbcl 13:50:12 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 13:52:21 pnpuff [~dioxin@unaffiliated/pnpuff] has joined #sbcl 13:52:49 -!- pnpuff [~dioxin@unaffiliated/pnpuff] has left #sbcl 14:09:58 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 14:16:43 ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has joined #sbcl 14:21:42 -!- ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has quit [Ping timeout: 264 seconds] 14:22:20 benkard [~benkard@2001:4ca0:0:f240:c027:2990:1020:df81] has joined #sbcl 15:30:30 stassats [~stassats@wikipedia/stassats] has joined #sbcl 15:37:44 -!- benkard [~benkard@2001:4ca0:0:f240:c027:2990:1020:df81] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 15:54:41 -!- Sagane [~Sagane@177.100-226-89.dsl.completel.net] has quit [Read error: Connection reset by peer] 16:12:47 benkard [~benkard@2001:4ca0:0:f240:8db7:291e:1c1f:59ca] has joined #sbcl 16:38:45 -!- benkard [~benkard@2001:4ca0:0:f240:8db7:291e:1c1f:59ca] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz] 17:04:58 Sagane [~Sagane@177.100-226-89.dsl.completel.net] has joined #sbcl 17:25:26 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 17:25:40 -!- Sagane [~Sagane@177.100-226-89.dsl.completel.net] has quit [Read error: Connection reset by peer] 17:25:51 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 17:33:15 -!- LiamH [~none@pool-173-73-124-231.washdc.east.verizon.net] has quit [Quit: Leaving.] 17:49:31 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 18:02:57 -!- pranavrc [~pranavrc@unaffiliated/pranavrc] has quit [Quit: Ping timeout: ] 18:16:13 -!- ASau` is now known as ASau 18:23:51 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 18:33:24 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 260 seconds] 18:35:31 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 18:41:47 -!- Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has quit [Ping timeout: 260 seconds] 18:43:46 drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 18:45:00 -!- drmeiste_ [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Remote host closed the connection] 18:48:07 -!- drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has quit [Ping timeout: 268 seconds] 19:08:31 LiamH [~none@pool-173-73-124-231.washdc.east.verizon.net] has joined #sbcl 19:12:06 fisxoj [~fisxoj@192.0.131.151] has joined #sbcl 19:16:03 drmeister [~drmeister@pool-173-59-25-58.phlapa.fios.verizon.net] has joined #sbcl 19:45:51 -!- fisxoj [~fisxoj@192.0.131.151] has quit [Ping timeout: 245 seconds] 19:47:38 -!- prxq_ [~mommer@x2f68c24.dyn.telefonica.de] has quit [Quit: Leaving] 19:57:18 ehaliewicz [~user@50-0-51-11.dsl.static.sonic.net] has joined #sbcl 20:36:56 Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has joined #sbcl 20:39:01 -!- sdemarre [~serge@91.180.110.162] has quit [Ping timeout: 248 seconds] 20:46:58 regarding interrupt safety and special binding, why not unconditionally bump the BSP before entering an interrupt handler? 20:47:25 those still possibly unitionalized entries won't overlap, and no zeroing required 20:49:37 except possibly for the "Binding stack exhausted." handler 20:50:06 though, it unprotects the guard page 20:51:55 needs working out what stop the world signals would do 20:54:23 and unwinding 20:55:24 looks like nothing, they'll just work 20:56:59 stop the world will work, but unwinding in a handler will be problematic 21:05:56 edgar-rft [~GOD@HSI-KBW-109-193-013-113.hsi7.kabel-badenwuerttemberg.de] has joined #sbcl 21:06:28 ok, the following scheme seems to work: binding will first store symbol and old-value on the binding stack and then increase BSP, an interrupt handler will save and zero the two entries past the binding stack, bump the BSP, call lisp, restore things upon return 21:18:36 awesome. Fixed copy-more-arg on x86oids seems to work, and eliminates the mystery of the initial stack frame size set to 4 slots. 21:19:04 pkhuong: does the above make sense? 21:23:23 what happens on unwind? 21:24:06 zeroing before handling the interrupt makes the unwind ignore it 21:24:22 zeroing, and then restoring upon exiting the interrupt 21:25:19 how do you know when to zero out? 21:25:35 always 21:25:58 so what happens when you zeroed out a valid entry? 21:26:03 -!- Bike [~Glossina@gannon-wless-gw.resnet.wsu.edu] has quit [Ping timeout: 260 seconds] 21:26:27 zeroing out past the BSP 21:26:48 and changing the sequence in unbind, first store entries, then increase BSP 21:26:56 so the entries don't go live until BSP is increased 21:27:13 the sequence in bind, that is 21:27:29 ok, so what happens when the binding sequence is interrupted in the middle of storing an entry? 21:28:27 that's what saving the entry before zeroing is for, and then restoring upon exiting 21:29:42 ok. so that means you have to get GC to scan the stashed values. 21:29:58 Bike [~Glossina@69.166.47.105] has joined #sbcl 21:30:43 won't they be accessible through the registers from which they were put there? 21:31:11 and it basically store the old value and a tls-index, the old value is still accessible in the tls space 21:31:55 the old value might be overwritten from TLS. 21:32:32 so you want to make how many things interact just right to save how many bytes? 21:32:50 fake_foreign_function_call should take care of it, i guess 21:33:16 it's already quite complicated with the sequence of zeroing etc 21:35:24 (and it's 10 bytes, and less memory access) 21:36:31 and if obsolete entries are never zeroed out, just not GCed, how can you tell whether what part if any of the entry in the "red zone" is valid, and which is a bogus pointer that's never been fixed up? 21:37:15 similar issue when scanning roots for GC. 21:37:36 why should it be ever examined? 21:39:31 ... because it may be pointer to the GCed heap? 21:39:51 and the entries should be saved somewhere where the GC sees them, for updates after moves 21:39:52 just that bind hasn't got around to bump BSP. 21:40:08 I'm not even considering interrupts here. 21:40:11 just plain GC. 21:41:14 safepoints solve that 21:43:52 fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has joined #sbcl 21:46:28 though that would mean safepoints by default everywhere, since i'm not a great fan of excessive conditionalization 21:46:57 but things like can offset the additional costs of safepoints 21:53:39 but for ordinary gc, it can speculatively update the entry past bsp 21:54:12 how do you know that the entry isn't bogus, from several GCs ago? 21:55:47 the only thing that needs to be done is adjusting its address, so, during scanning of tls space, if tls-index and the value are equal to the values past bsp, the address can be adjusted 21:56:50 this is almost as hairy as the magic interior pointers on precise platforms. 21:57:35 but yeah, this seems like a showstopper without safepoints 21:57:59 and with full safe points, you don't need any magic at all. 21:58:21 unless the signal is from outside, like sigchld 21:58:46 even then. with full safepoints, the signal should be recorded but only handled at a safe point. 21:59:12 ok then, will push for safer safepoints 22:08:14 I now see failures in ::: Running (:ASYNC-UNWIND :SPECIALS) 22:08:24 x86-64 only, not x86. 22:08:26 linux 22:11:27 pkhuong: that's without the safepoint things? 22:11:48 haven't tested without them 22:15:00 without safepoint 22:15:08 vanilla build 22:15:11 indeed, probably made a type or something 22:15:22 typo 22:26:40 indeed, the order of saves in bind got mixed up 22:28:57 right, C-M-t fixes it 23:07:07 -!- fisxoj [~fisxoj@192-0-131-151.cpe.teksavvy.com] has quit [Quit: Leaving] 23:29:53 -!- LiamH [~none@pool-173-73-124-231.washdc.east.verizon.net] has quit [Quit: Leaving.]