IRC logs for #opendune on OFTC at 2010-03-11
⏴ go to previous day
00:40:53 <DorpsGek> SVN: glx (r976) -Add: C-ified emu_Unknown_B483_0156() and emu_Unknown_B483_0823()
13:56:12 <glx> changes nothing for opendune btw ;)
14:46:44 <glx> maybe need an emu_flags.inf check
14:48:04 <glx> like if (_pic_irq_service == 0 && emu_flags.inf) {
14:53:26 <glx> I don't know if I should skip timers if an interrupt ISR is running
15:21:03 <glx> interrupt is triggered but uninstalled before it's executed
15:23:44 <TrueBrain> so you have to protect against those cases :)
15:24:11 <glx> I'd say the problem is emu_io_read_005 :)
15:24:28 <glx> because the driver waits for a cycle before disabling the IRQ
15:25:47 <glx> and it's almost instant with the current implementation
15:30:38 <SmatZ> emulators can unhide some race conditions in programs :)
16:35:05 <glx> hmm ok it should be emu_io_read_3DA, but we have wrong (uninitialised) data in bios memory
16:35:07 <TrueBrain> all data is initalized in the bios memory :)
16:46:41 *** TrueBrain has joined #openDune
17:58:44 <glx> grr synchronising on vga is silly
18:00:46 <glx> and that's my timing issue
18:13:48 <glx> because our fake 3DA is probably too fast
18:47:24 <TrueBrain> that is a slowdown ... lol
18:47:49 <TrueBrain> it is fine by me :)
18:48:01 <glx> kind of 50Hz vertical sync ;)
18:48:34 <glx> of course faster than 50Hz as the CPU goes way faster
20:01:43 <DorpsGek> SVN: glx (r977) [LibEMU] -Fix: incorrect data in BIOS variable 0x63
20:03:08 <DorpsGek> SVN: glx (r978) [LibEMU] -Codechange: improved IRQ support
21:37:59 *** ChanServ sets mode: +v Yexo
22:15:37 <DorpsGek> SVN: glx (r979) -Fix (r973): forgot to remove a reference
22:20:13 <glx> I was wondering why toc readded an unused function ;)
22:43:26 <glx> I forced tox to add it in r966 (via an extern in driver.c)
22:44:23 <glx> but I removed it in r973 because it was just temporary to resolve a jump
22:44:39 <glx> and now it keeps adding it
22:45:34 <glx> f__1DD7_14C5_000A_A995() to be precise :)
22:49:41 * TrueBrain pushes a few buttons
22:50:08 <TrueBrain> when I run clean, it removes f__01F7_3BF1_0014_1524
22:50:16 <TrueBrain> f__2756_0D59_0006_C5AC
22:50:21 <TrueBrain> f__2756_0E7E_0006_82F7
22:50:33 <TrueBrain> and doens't add 1DD7
22:50:50 <TrueBrain> dunno .. that happens when I run on a clean txt set
22:51:18 <TrueBrain> check if yours is clean :)
22:51:42 <glx> well my txt is not clean as I had a JIT run
22:52:08 <TrueBrain> ah, that might explain why I don't get 14C5 :p
22:52:25 <glx> but 14C5 is available in txt
22:53:48 <TrueBrain> that is what I get on clean trunk + txt
22:54:08 <TrueBrain> notice the one 01F7 function comes from the 14C5 most likely :)
22:54:53 <TrueBrain> should I commit this?
22:55:37 <TrueBrain> doing a full clean again
22:57:14 <glx> ok trying rm and svn up in txt
22:57:22 <DorpsGek> SVN: truebrain (r980) -Cleanup: remove unneeded functions and temporary fixes
22:59:17 <glx> hmm a clean run still adds it
22:59:27 <TrueBrain> then let me make a new patch of my work, and lets see if that helps ;)
23:05:18 <glx> updated trunk it now readds what you removed
23:05:35 <TrueBrain> random errors in toc I guess?
23:05:51 <TrueBrain> valgrind doesn't tell me ...
23:06:42 <glx> the txt are present of course, but as there's no references to these functions it should not readd them
23:07:06 <TrueBrain> and it doesn't here :p
23:09:10 <TrueBrain> can't find anything to show what the issue can be ...
23:13:16 <glx> I run it with "./toc --static ."
23:14:08 <TrueBrain> how ever you run it, it either works, or not :p
23:14:26 <TrueBrain> but okay, too many functions isn't a real issue :)
23:15:03 <glx> true but it's weird it readds them for me and not for you
23:16:39 <glx> how can I see which libs are loaded at runtime?
23:17:13 <TrueBrain> ldd shows you what will be loaded
23:18:51 <glx> anyway toc uses only standard libs
23:19:15 <TrueBrain> so just ignore it for now
23:19:25 <TrueBrain> remove them manually if you want to, ignore them and commit it if you like to
23:19:33 <TrueBrain> they will remove theirselves in the end :)
23:19:53 <TrueBrain> either way, off to bed
23:22:04 <glx> hmm "rm -rf decompiled/* src/* && svn up" and now it works
continue to next day ⏵