#nuvie@irc.freenode.net logs for 30 Jan 2009 (GMT)

Archive Today Yesterday Tomorrow
Nuvie homepage

[00:26:59] --> Kirben has joined #nuvie
[00:48:53] <-- Kirben has left IRC (Read error: 110 (Connection timed out))
[03:14:09] --> Kirben has joined #nuvie
[03:39:54] <-- Dark-Star has left IRC (Read error: 60 (Operation timed out))
[03:40:05] --> Dark-Star has joined #nuvie
[07:23:20] --> Yuv422 has joined #nuvie
[08:27:33] <wjp> hi Yuv422
[08:27:43] <Yuv422> hi Willem
[08:27:46] <Yuv422> How's things?
[08:28:36] <wjp> pretty good :-)
[08:28:45] <wjp> and you?
[08:29:02] <Yuv422> yeah not bad, looking forward to the weekend
[08:29:11] <wjp> same here :-)
[08:29:34] <wjp> I guess you have a bit of a headstart there, though :-)
[08:29:46] <Yuv422> hehe yes
[08:29:55] <Yuv422> I'm already in weekend mode
[08:30:55] <Yuv422> I've been talking to another IDA user who is interested in making a gdb debug protocol interface to dosbox
[08:31:07] <Yuv422> which could be interesting
[08:32:08] <wjp> hm, yes
[08:32:36] <wjp> exciting things are happening :-)
[08:32:44] <Yuv422> yeah
[08:33:03] <Yuv422> the new SDK uses a map_address() function
[08:33:07] <wjp> pretty cool that Ilfak upgraded the debugger like he did
[08:33:14] <Yuv422> yeah :)
[08:33:19] <wjp> yeah, I saw the note in the changelog
[08:33:46] <wjp> I can't help but wonder if your prodding last year got him thinking about it :-)
[08:34:00] <Yuv422> hehe I think it might have
[08:34:27] <Yuv422> http://hexblog.com/2008/11/bochs_plugin_goes_alpha.html#comments
[08:35:23] <wjp> ah, very nice
[08:35:24] <Yuv422> The dosbox plugin is more stable now too on the new code base
[08:35:52] <wjp> so that's really something he added in the last two months
[08:36:02] <Yuv422> yeah
[08:36:50] <Yuv422> they've added bitness to the memory info segments too
[08:37:01] <wjp> bitness?
[08:37:06] <Yuv422> so you specify 16bit segments when returning memory info
[08:37:18] <Yuv422> hehe that's the field name
[08:47:14] <Yuv422> bbl
[10:37:08] <Yuv422> hmm I think I've found a problem with the new debugger code
[10:37:44] <Yuv422> it isn't handling breakpoints correctly using the 16bit mode
[10:38:19] <Yuv422> it is trying to set EIP to the ea address
[10:38:50] <Yuv422> which causes the display to shoot off to a weird address
[10:39:23] <Yuv422> I might be able to work around it
[10:40:36] <wjp> hm :/
[10:41:02] <Yuv422> probably a hangover from the old 32bit debugger code
[10:41:37] <Yuv422> I tried erroring the reg_write call
[10:41:50] <Yuv422> but IDA sets EIP anyway
[10:42:11] <Yuv422> hitting F7 steps and resyncs the display
[10:46:57] <wjp> can you make a call back to IDA from the reg_write function to correct EIP?
[10:47:40] <wjp> (memory is a bit hazy on the topic)
[10:47:54] <Yuv422> I can make a sync event
[10:48:08] <Yuv422> which should get IDA to reg reg vals again
[10:48:13] <Yuv422> I'll give that a go
[10:48:40] <Yuv422> IDA is asking to convert EIP to code when it hits the bad address
[10:48:50] <Yuv422> so you'd still have to hit no
[10:48:54] <Yuv422> to continue
[10:58:05] <Yuv422> that didn't seem to work
[10:58:23] <Yuv422> I wonder if I could issue an exception instead of a breakpoint
[12:12:15] <-- Kirben has left IRC (Read error: 60 (Operation timed out))
[13:09:13] <-- Yuv422 has left IRC ()
[13:17:52] <-- Dark-Star has left IRC (Read error: 60 (Operation timed out))
[13:20:09] --> Dark-Star has joined #nuvie
[21:36:40] --> Yuv422 has joined #nuvie
[21:55:42] --> Kirben has joined #nuvie
[23:37:50] <-- Yuv422 has left IRC ()