00:45:23 -!- tsuru [~charlie@adsl-74-179-25-40.bna.bellsouth.net] has quit [Remote host closed the connection] 01:42:45 -!- loke [~elias@bb220-255-108-49.singnet.com.sg] has quit [Ping timeout: 264 seconds] 01:48:56 loke [~elias@bb121-6-159-164.singnet.com.sg] has joined #sbcl 02:12:43 -!- attila_lendvai [~attila_le@unaffiliated/attila-lendvai/x-3126965] has quit [Quit: Leaving.] 02:48:09 -!- loke [~elias@bb121-6-159-164.singnet.com.sg] has quit [Ping timeout: 276 seconds] 02:48:58 loke [~elias@bb219-74-244-83.singnet.com.sg] has joined #sbcl 03:51:51 -!- loke [~elias@bb219-74-244-83.singnet.com.sg] has quit [Ping timeout: 276 seconds] 04:00:15 loke [~elias@bb219-74-244-83.singnet.com.sg] has joined #sbcl 04:00:27 -!- loke [~elias@bb219-74-244-83.singnet.com.sg] has quit [Read error: Connection reset by peer] 04:05:33 -!- cmm [~cmm@bzq-79-177-212-194.red.bezeqint.net] has quit [Ping timeout: 264 seconds] 04:05:55 cmm [~cmm@bzq-79-177-212-194.red.bezeqint.net] has joined #sbcl 04:30:07 -!- cmm [~cmm@bzq-79-177-212-194.red.bezeqint.net] has quit [Ping timeout: 258 seconds] 04:30:47 cmm [~cmm@bzq-79-177-212-194.red.bezeqint.net] has joined #sbcl 05:02:19 -!- gor[e] [~svr@gw2.masterhost.ru] has quit [Ping timeout: 258 seconds] 05:27:09 tcr1 [~tcr@89.108.255.33] has joined #sbcl 05:27:25 -!- tcr1 [~tcr@89.108.255.33] has quit [Client Quit] 06:17:35 flip214 [~marek@2001:858:107:1:7a2b:cbff:fed0:c11c] has joined #sbcl 06:17:35 -!- flip214 [~marek@2001:858:107:1:7a2b:cbff:fed0:c11c] has quit [Changing host] 06:17:35 flip214 [~marek@unaffiliated/flip214] has joined #sbcl 07:03:36 nikodemus [~nikodemus@cs181063174.pp.htv.fi] has joined #sbcl 07:03:37 -!- ChanServ has set mode +o nikodemus 07:10:45 -!- Quadrescence [~Quad@unaffiliated/quadrescence] has quit [Ping timeout: 276 seconds] 07:25:38 homie` [~levgue@xdsl-84-44-155-168.netcologne.de] has joined #sbcl 07:27:35 -!- homie [~levgue@xdsl-84-44-210-13.netcologne.de] has quit [Ping timeout: 260 seconds] 07:40:09 -!- nikodemus [~nikodemus@cs181063174.pp.htv.fi] has quit [Quit: Leaving] 08:34:01 hlavaty [~user@91-65-223-81-dynip.superkabel.de] has joined #sbcl 08:35:15 -!- pchrist [~spirit@gentoo/developer/pchrist] has quit [Quit: leaving] 09:18:57 misterncw [~misterncw@82.71.241.25] has joined #sbcl 10:18:24 daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has joined #sbcl 10:19:02 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #sbcl 10:19:02 -!- ChanServ has set mode +o nikodemus 13:32:14 -!- misterncw [~misterncw@82.71.241.25] has quit [Remote host closed the connection] 13:37:55 misterncw [~misterncw@82.71.241.25] has joined #sbcl 13:43:37 -!- flip214 [~marek@unaffiliated/flip214] has quit [Remote host closed the connection] 14:42:12 -!- homie` [~levgue@xdsl-84-44-155-168.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 14:46:43 homie [~levgue@xdsl-84-44-155-168.netcologne.de] has joined #sbcl 15:14:24 pdlogan [~patrick@174-25-37-137.ptld.qwest.net] has joined #sbcl 15:43:46 tsuru [~charlie@adsl-74-179-25-40.bna.bellsouth.net] has joined #sbcl 16:26:56 -!- misterncw [~misterncw@82.71.241.25] has quit [Remote host closed the connection] 16:57:10 <_8david> nikodemus: %symbol-value-in-thread has turned into an awfully tricky functions. Why not just use without-gcing? 17:01:40 i think i had a reason, but i can't remember offhand 17:01:51 isn't there a comment about that? 17:03:54 hm. maybe with-all-threads-lock should imply without-gcing 17:15:20 <_8david> OK. I think I'm just going to assume that both your and Anton's version of the function are overly careful (and also possibly buggy :-)), and replace the whole thing with a much simpler definition. 17:15:41 <_8david> Conditional on #!+sb-safepoint which depends on the fact that there is no safepoint between sap-ref-word and make-lisp-obj, so without-gcing isn't even needed. Nor is pinning. 17:15:49 <_8david> The non-safepoint code will stay unchanged. 17:20:25 _8david: at some point a no-safepoint optimisation quality might be a good idea. 17:21:51 *_8david* adds that to his to-do list 17:22:23 I could see a point in inserting safe points in the middle of very long straight-line code. 17:22:46 <_8david> At the memont we add safepoints for function calls and loop edges. The optimization quality would prevent insertion in the latter case. Should it also assert that function calls don't happen? 17:24:00 <_8david> Hah, I've always thought that SBCL doesn't scale well enough to compile long straight-line code anyway :-). But perhaps that's not true (anymore). 17:24:02 don't know (yet). 17:24:08 well, it's getting better. 17:35:48 whee, a whopping 5% improvement on (time (ql:quickload "mcclim")) with all fasls deleted 17:37:47 https://github.com/nikodemus/SBCL/tree/globaldb # could of contribs need fixing, and i'm not sure if something should be done about non-symbol names 17:39:50 <_8david> cool 17:41:06 s/could of/couple of/ 17:48:41 -!- hlavaty [~user@91-65-223-81-dynip.superkabel.de] has quit [Ping timeout: 240 seconds] 17:54:21 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Ping timeout: 258 seconds] 18:25:52 <_8david> observations: 1. it seems a little messy to litter ir2tran with #!+sb-safepoint policy checks for all the different call types (I'm counting four places that need the check). 18:26:16 <_8david> 2. that still won't catch calls to VOPs, because we need to black-list VOPs and assembly routines individually (like generic+) which can trip safepoints 18:27:17 <_8david> 3. if skip any of these checks, the optimization quality loses much of its appeal, because accidental full calls are the most likely way code can accidentally trip safepoints 18:28:06 <_8david> Of course, being able to prevent safepoints in a loop alone might be worth adding the policy anyway. It just doesn't help with my current use case (%symbol-value-in-thread). 18:44:43 yeah, just prevent automagic insertion, and if there's a full call, that's your problem. 18:56:37 -!- daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has quit [Read error: Connection reset by peer] 19:00:11 <_8david> OK. 19:09:01 tcr1 [~tcr@89.108.255.33] has joined #sbcl 19:22:50 tcr2 [~tcr@89.108.255.33] has joined #sbcl 19:23:50 -!- tcr1 [~tcr@89.108.255.33] has quit [Read error: Connection reset by peer] 19:23:50 -!- tcr2 [~tcr@89.108.255.33] has quit [Read error: Connection reset by peer] 19:24:29 tcr1 [~tcr@89.108.255.33] has joined #sbcl 19:29:09 -!- tcr1 [~tcr@89.108.255.33] has quit [Ping timeout: 255 seconds] 19:30:58 prevent automagic insertion and compile-time warn if there's a full call? 19:41:35 daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has joined #sbcl 19:41:58 -!- daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has quit [Client Quit] 19:43:17 daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has joined #sbcl 19:43:49 -!- daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has quit [Client Quit] 19:44:32 daimrod [~daimrod@ANantes-556-1-326-23.w90-59.abo.wanadoo.fr] has joined #sbcl 19:46:45 homie` [~levgue@xdsl-78-35-179-10.netcologne.de] has joined #sbcl 19:49:29 -!- homie [~levgue@xdsl-84-44-155-168.netcologne.de] has quit [Ping timeout: 276 seconds] 20:19:15 <_8david> Krystof: yeah, well, I didn't feel good about the patch for the function calling changes. But I guess I can do it in a separate commit and invite comments before pushing. 20:40:21 -!- cmm [~cmm@bzq-79-177-212-194.red.bezeqint.net] has quit [Ping timeout: 264 seconds] 20:41:08 cmm [~cmm@bzq-79-177-212-194.red.bezeqint.net] has joined #sbcl 21:12:28 tcr1 [~tcr@155-dom-3.acn.waw.pl] has joined #sbcl 21:43:32 redline6561_ [~redline65@li69-162.members.linode.com] has joined #sbcl 21:50:24 -!- tWip [u258@gateway/web/irccloud.com/x-ewvkhhtzlbrjfscf] has quit [*.net *.split] 21:50:24 -!- redline6561 [~redline65@li69-162.members.linode.com] has quit [*.net *.split] 21:50:24 -!- pepone [~jose@84.246.218.87.dynamic.jazztel.es] has quit [*.net *.split] 21:50:24 -!- deepfire [~deepfire@80.92.100.69] has quit [*.net *.split] 21:50:25 -!- foom [~jknight@ita4fw1.itasoftware.com] has quit [*.net *.split] 21:57:25 foom [~jknight@ita4fw1.itasoftware.com] has joined #sbcl 21:59:33 tWip [u258@gateway/web/irccloud.com/x-sggvdvowstlusmcb] has joined #sbcl 22:23:08 slyrus [~chatzilla@dsl081-060-042.sfo1.dsl.speakeasy.net] has joined #sbcl 22:43:18 -!- slyrus [~chatzilla@dsl081-060-042.sfo1.dsl.speakeasy.net] has quit [Ping timeout: 252 seconds] 22:44:58 I'm pretty sure we generate code that could use some jump tensioning (jmp/jmp elimination), but I can't craft any right now. 22:45:04 Do you have one at hand? 23:15:50 -!- pdlogan [~patrick@174-25-37-137.ptld.qwest.net] has left #sbcl 23:49:22 deepfire [~deepfire@80.92.100.69] has joined #sbcl