2015-01-10T00:04:18Z slyrus joined #sbcl
2015-01-10T00:20:41Z rpg quit (Quit: rpg)
2015-01-10T01:05:54Z Bike joined #sbcl
2015-01-10T01:16:07Z slyrus quit (Ping timeout: 244 seconds)
2015-01-10T01:24:16Z attila_lendvai quit (Quit: Leaving.)
2015-01-10T02:15:18Z Bike quit (Quit: restart)
2015-01-10T02:16:30Z Bike joined #sbcl
2015-01-10T02:31:33Z echo-area joined #sbcl
2015-01-10T02:38:18Z Quadrescence joined #sbcl
2015-01-10T02:43:15Z Bike quit (Quit: one more restart)
2015-01-10T02:49:04Z Bike joined #sbcl
2015-01-10T03:02:58Z Quadrescence quit (Quit: This computer has gone to sleep)
2015-01-10T03:39:25Z christoph_debian quit (Ping timeout: 264 seconds)
2015-01-10T03:49:48Z Quadrescence joined #sbcl
2015-01-10T03:52:00Z christoph_debian joined #sbcl
2015-01-10T03:58:33Z nikki93 joined #sbcl
2015-01-10T04:54:53Z adlai quit (Ping timeout: 250 seconds)
2015-01-10T04:55:05Z nikki93 quit (Remote host closed the connection)
2015-01-10T04:55:40Z nikki93 joined #sbcl
2015-01-10T04:56:12Z adlai joined #sbcl
2015-01-10T04:57:55Z pacon joined #sbcl
2015-01-10T05:00:09Z nikki93 quit (Ping timeout: 256 seconds)
2015-01-10T05:02:49Z nikki93 joined #sbcl
2015-01-10T05:03:35Z echo-area quit (Remote host closed the connection)
2015-01-10T05:22:38Z edgar-rft quit (Quit: mental process lost because happening stopped)
2015-01-10T05:51:29Z oleo is now known as Guest71432
2015-01-10T05:52:21Z oleo__ joined #sbcl
2015-01-10T05:54:24Z Guest71432 quit (Ping timeout: 245 seconds)
2015-01-10T05:58:39Z pacon quit (Read error: Connection reset by peer)
2015-01-10T05:59:42Z pacon joined #sbcl
2015-01-10T06:04:08Z nyef quit (Quit: G'night all)
2015-01-10T06:16:34Z pacon quit (Quit: Leaving)
2015-01-10T06:31:04Z pacon joined #sbcl
2015-01-10T06:35:46Z nikki93 quit (Remote host closed the connection)
2015-01-10T06:45:17Z nikki93 joined #sbcl
2015-01-10T06:50:20Z nikki93 quit (Remote host closed the connection)
2015-01-10T07:24:23Z nikki93 joined #sbcl
2015-01-10T07:24:25Z nikki93 quit (Remote host closed the connection)
2015-01-10T07:39:12Z slyrus joined #sbcl
2015-01-10T07:45:31Z pacon quit (Read error: Connection reset by peer)
2015-01-10T07:45:59Z pacon joined #sbcl
2015-01-10T07:59:55Z angavrilov joined #sbcl
2015-01-10T08:42:16Z Quadrescence quit (Quit: This computer has gone to sleep)
2015-01-10T08:43:41Z nikki93 joined #sbcl
2015-01-10T08:44:12Z nikki93 quit (Remote host closed the connection)
2015-01-10T08:50:16Z nikki93 joined #sbcl
2015-01-10T08:55:15Z nikki93 quit (Remote host closed the connection)
2015-01-10T08:55:50Z nikki93 joined #sbcl
2015-01-10T09:00:05Z nikki93 quit (Ping timeout: 244 seconds)
2015-01-10T09:00:31Z stassats joined #sbcl
2015-01-10T09:07:05Z karswell quit (Remote host closed the connection)
2015-01-10T09:52:00Z nikki93_ joined #sbcl
2015-01-10T09:52:30Z stassats quit (Ping timeout: 264 seconds)
2015-01-10T09:56:42Z nikki93_ quit (Ping timeout: 264 seconds)
2015-01-10T10:44:16Z psy_ joined #sbcl
2015-01-10T11:05:05Z psy_ quit (Ping timeout: 256 seconds)
2015-01-10T11:16:55Z pacon quit (Ping timeout: 255 seconds)
2015-01-10T11:21:50Z psy_ joined #sbcl
2015-01-10T11:39:08Z angavrilov quit (Remote host closed the connection)
2015-01-10T11:45:47Z edgar-rft joined #sbcl
2015-01-10T12:20:16Z stassats joined #sbcl
2015-01-10T12:44:56Z gingerale quit (Ping timeout: 240 seconds)
2015-01-10T13:04:14Z stassats: interrupting ppc and arm before the stack pointer is adjusted is not good for backtraces
2015-01-10T13:06:03Z stassats: ppc goes completely bonkers, and arm just shows the same frame over an over
2015-01-10T13:06:04Z stassats: and the precise mode depends on the number of mismatched arguments, really weird
2015-01-10T13:33:56Z stassats: on arm: (cons) => the backtrace is (cons) (cons) (cons), while (cons 1) gives a normal backtrace
2015-01-10T13:33:59Z stassats: why the hell?
2015-01-10T14:30:50Z pacon joined #sbcl
2015-01-10T14:45:30Z Krystof: I seem to remember all sbcl sessions breaking last time there was a leap second
2015-01-10T14:45:42Z Krystof: does anyone else remember anything more?
2015-01-10T14:46:53Z LiamH joined #sbcl
2015-01-10T14:47:27Z Krystof: https://bugs.launchpad.net/sbcl/+bug/1028026
2015-01-10T14:48:49Z stassats: i usually say: just don't reset the clock
2015-01-10T14:51:38Z pkhuong: Krystof: going forward should be fine...
2015-01-10T14:52:13Z stassats: shouldn't going backward be handled by slowing down time?
2015-01-10T14:52:27Z stassats: so the clock is monotonous, at least that how ntpd adjusts time
2015-01-10T14:53:58Z stassats: and if you don't have ntp, your clock is going to be out of sync anyway, what's the point of adjusting the leap second?
2015-01-10T14:55:04Z Krystof: stassats: only if we actually use CLOCK_MONOTONIC
2015-01-10T14:55:17Z Krystof: actually from that bug report I see that there's only a window of a second or so where things will go wrong
2015-01-10T14:55:47Z Krystof: but if that's still a real bug report there's an hour or so's window at the autumn clock change where things go wrong
2015-01-10T14:56:12Z stassats: yeah, ntp doesn't handle that
2015-01-10T14:59:06Z oleo__ quit (Quit: Verlassend)
2015-01-10T14:59:57Z oleo joined #sbcl
2015-01-10T15:03:03Z pkhuong: stassats: I've had people tell me the same about ntp. "if you're using ntp, why are you looking at milliseconds?" (:
2015-01-10T15:16:21Z pacon quit (Read error: Connection reset by peer)
2015-01-10T15:21:25Z stassats: seems like only applying functions is broken
2015-01-10T15:22:19Z minion quit (Remote host closed the connection)
2015-01-10T15:22:20Z specbot quit (Remote host closed the connection)
2015-01-10T15:25:07Z minion joined #sbcl
2015-01-10T15:25:08Z specbot joined #sbcl
2015-01-10T15:50:25Z gingerale joined #sbcl
2015-01-10T15:57:10Z stassats: tail-call-variable works, call-variable doesn't
2015-01-10T16:04:22Z stassats: i think i see it
2015-01-10T16:04:57Z stassats: in (defun bar (a b) (apply a b) 1)
2015-01-10T16:05:24Z stassats: it first does apply-list, which puts the list on the stack, but when b is NIL, nothing is put on the stack and the stack pointer is not extended
2015-01-10T16:05:51Z stassats: but it thinks that it is, and puts the CFP and LRA in those register locations
2015-01-10T16:06:02Z stassats: so, when called with '(1), the stack is extended and it works
2015-01-10T16:07:08Z stassats: i.e. it was always broken
2015-01-10T16:08:26Z stassats: and it uses the OLD-CSP - CSP to calculate NARGS too, so just extending CSP before copying the list wouldn't work
2015-01-10T16:10:17Z slyrus quit (Ping timeout: 245 seconds)
2015-01-10T16:13:51Z stassats: x86-64 has this comment
2015-01-10T16:13:51Z stassats:  The variable args are on the stack and become the frame, but there may be <3 args and 3 stack slots are assumed allocate on the call. So need to ensure there are at least 3 slots. This hack just adds 3 more.
2015-01-10T16:40:38Z stassats: that was it
2015-01-10T16:52:16Z stassats: the same trick doesn't exactly fix PPC
2015-01-10T16:53:14Z stassats: since it stores the old frame pointer and lra in the function prologue
2015-01-10T16:53:49Z stassats: this hack is getting too complicated
2015-01-10T16:54:31Z stassats: now i'm thinking of only enabling it on x86oids and ARM, and to hell with everything else
2015-01-10T17:02:51Z stassats: of course ppc can be fixed if i know that the PC is before OCFP and LRA are saved and still alive in registers
2015-01-10T17:03:04Z stassats: but that's more reader conditionals than i care about
2015-01-10T17:14:49Z stassats: ppc, x86 and arm do three different things
2015-01-10T17:15:18Z stassats: ppc reserves  the stack space and stores OCFP in the callee, arm does that in the caller
2015-01-10T17:15:32Z stassats: x86 reserves space in the caller, saves OCFP in the callee
2015-01-10T17:17:31Z stassats: and i don't want to add copy-more-arg and interrupt safety to that mental model
2015-01-10T17:19:13Z stassats: when will SBCL crumble with all the diverging hacks?
2015-01-10T17:19:32Z stassats: i guess it already kinda did, mips doesn't work
2015-01-10T17:21:45Z stassats: on that note, got a working cold-init on mips
2015-01-10T17:38:36Z stassats: and found the commit, happens to be mine...
2015-01-10T17:56:27Z scymtym_ joined #sbcl
2015-01-10T18:12:56Z oleo__ joined #sbcl
2015-01-10T18:13:52Z oleo__ quit (Remote host closed the connection)
2015-01-10T18:15:05Z oleo quit (Ping timeout: 252 seconds)
2015-01-10T18:16:48Z oleo joined #sbcl
2015-01-10T18:25:13Z DeadTrickster quit (Read error: No route to host)
2015-01-10T18:32:23Z stassats: just a stupid typo, damn wasted me too much time
2015-01-10T18:38:49Z psy__ joined #sbcl
2015-01-10T18:39:04Z psy__ quit (Client Quit)
2015-01-10T18:39:28Z psy_ quit (Quit: Leaving)
2015-01-10T18:39:49Z nikki93_ joined #sbcl
2015-01-10T18:44:10Z psy_ joined #sbcl
2015-01-10T18:57:45Z nikki93_ quit (Remote host closed the connection)
2015-01-10T19:00:32Z nikki93__ joined #sbcl
2015-01-10T19:14:08Z nikki93__ quit (Remote host closed the connection)
2015-01-10T19:53:05Z nikki93_ joined #sbcl
2015-01-10T20:22:54Z slyrus joined #sbcl
2015-01-10T20:24:44Z karswell` joined #sbcl
2015-01-10T21:31:53Z nikki93_ quit (Remote host closed the connection)
2015-01-10T21:35:42Z slyrus quit (Ping timeout: 245 seconds)
2015-01-10T21:41:07Z slyrus joined #sbcl
2015-01-10T21:50:14Z Quadrescence joined #sbcl
2015-01-10T22:04:52Z slyrus quit (Ping timeout: 245 seconds)
2015-01-10T23:02:28Z dtm` joined #sbcl
2015-01-10T23:04:33Z attila_lendvai joined #sbcl
2015-01-10T23:04:34Z attila_lendvai quit (Changing host)
2015-01-10T23:04:34Z attila_lendvai joined #sbcl
2015-01-10T23:08:58Z stassats quit (Ping timeout: 244 seconds)
2015-01-10T23:10:00Z dtm` quit (Quit: ERC Version 5.3 (IRC client for Emacs))