#nuvie@irc.freenode.net logs for 6 Sep 2010 (GMT)

Archive Today Yesterday Tomorrow
Nuvie homepage


[01:27:32] <dongs> ok!
[01:38:21] <dongs> lets find out
[02:39:01] <dongs> woohoo
[02:39:12] <dongs> rebasing properly works now and it steps trhough right stuff
[02:39:25] <dongs> after debug session it rebases it back to normal but forgets all reerences to strings etc
[02:39:32] <dongs> though thats not too big of a deal (yet)
[06:04:43] --> llm has joined #nuvie
[06:06:05] <llm> wouldn't it be easier to ship directly the svn-root of dosbox - this way it is simple to merge with current/changed versions
[06:20:13] <dongs> heh
[06:23:22] --> Kirben_ has joined #nuvie
[06:25:23] <-- Kirben has left IRC (Ping timeout: 265 seconds)
[06:25:36] <llm> bbl
[06:25:38] <-- llm has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100723085541])
[07:35:37] --- Kirben_ is now known as Kirben
[07:36:23] <-- Dominus|away has left IRC (Ping timeout: 245 seconds)
[07:41:19] --> Dominus has joined #nuvie
[08:31:48] --> Yuv422 has joined #nuvie
[08:47:36] <dongs> Yuv422: after debugging is done and it rebases program back to normal, any code that was loaded at program runtime can't be seen. though I doubt this is a dosbox issue..
[08:48:28] <Yuv422> yeah I think ida trashes any debug values
[08:48:39] <dongs> well it has them back if I debug again
[08:48:46] <Yuv422> hmm
[08:48:53] <dongs> not "values" i mean actual code segments
[08:48:59] <dongs> also: what determines the segment list in your plugin?
[08:49:01] <dongs> bios/dos/etc
[08:49:11] <dongs> do you just base that on some general thing or do you pull that from dosbox
[08:49:55] <Yuv422> yeah the segment list is done on the dosbox side
[08:50:01] <Yuv422> by the idados code
[08:50:25] <Yuv422> and it is pretty crude
[08:51:54] <Yuv422> have a look at
[08:51:56] <Yuv422> int idaapi dosbox_debmod_t::dbg_get_memory_info(meminfo_vec_t &miv)
[08:52:06] <Yuv422> in dosbox_debmod.cpp
[08:52:53] <Yuv422> also that code I gave you is running on dosbox 0.73
[08:53:06] <dongs> ya i just used the giant debug .exe that was in the file.
[08:53:22] <Yuv422> hehe
[08:53:26] <Yuv422> yeah I didn't strip it
[08:54:21] <dongs> bios segment in x400?
[08:54:38] <Yuv422> end address
[08:54:57] <Yuv422> whoops
[08:54:58] <Yuv422> yeah
[08:55:01] <dongs> still. sbase = 0x40?
[08:55:20] <dongs> i thought pc bios was more like f000 or something
[08:55:22] <Yuv422> I'm not sure where I got the memory map from
[08:56:12] <Yuv422> 400h BIOS communication area
[08:56:31] <Yuv422> http://www.osdata.com/system/physical/memmap.htm
[08:57:15] <dongs> oh, thats like r/w shit, registers and junk
[09:06:54] <dongs> i wonder if dosbox has good way to query this
[09:08:13] <dongs> theres dos_memory stuff in dosbox
[09:08:20] <dongs> that creates a few memory regions etc
[09:09:09] <dongs> i guess it treats everything from DOS_MEM_START to 0x9FFE as accessible.
[09:09:20] <Yuv422> yeah
[09:09:34] <Yuv422> I tried to keep the segment fairly generic
[09:09:39] <Yuv422> the .text one
[09:19:36] <Yuv422> the dosbox side should be cleaned up a bit
[09:19:48] <Yuv422> I just hacked it together to see if it would work
[09:31:59] <Yuv422> bbl
[09:32:01] <-- Yuv422 has left IRC (Quit: Yuv422)
[10:50:02] <-- Kirben has left IRC (Ping timeout: 276 seconds)
[11:24:51] --> Yuv422 has joined #nuvie
[11:58:28] <Yuv422> I'm not sure what the stack view is showing
[12:41:27] <Yuv422> I think the rebasing still needs a bit of work
[12:41:46] <Yuv422> I've just tried debugging an idb file that was based correctly for dosbox
[12:41:57] <Yuv422> and all the strings and offsets appear correctly
[12:58:40] <-- Yuv422 has left IRC (Quit: Yuv422)
[13:36:24] <dongs> nice
[16:18:28] <-- Dominus has left IRC (Quit: Leaving.)
[16:20:06] --> Dominus has joined #nuvie
[20:57:31] --> Yuv422 has joined #nuvie
[21:48:48] <-- Dominus has left IRC (Quit: Leaving.)
[21:49:52] --> Dominus has joined #nuvie
[22:21:06] <-- Dominus has left IRC (Quit: Leaving.)
[22:39:37] <-- Yuv422 has left IRC (Quit: Yuv422)