[00:18:21] <dingo__> i think i've figured it out
[00:18:39] <dingo__> read4() reads over a 4byte integer
[00:18:43] <dingo__> smashing the stack
[00:18:54] <dingo__> and pooping on my return address
[00:19:10] <dingo__> on openbsd using propolice on arm architecture
[00:20:04] <dingo__> 720 in.seek(0x54); // Get to where file count sits.
[00:20:11] <dingo__> 721 int numfiles = in.read4();
[00:20:20] <dingo__> btw, 12: numfiles = 300
[00:20:28] <dingo__> is 300 appropriate?
[00:20:53] <dingo__> the comments do not make it eentirely clear what this represents... 300 files??
[00:23:43] <dingo__> anyway, my stack gets smmashed soon afeter
[00:23:50] <dingo__> inside that loop
[00:24:00] <dingo__> i=0; i<numfiles
[00:24:06] <dingo__> in.read4()
[00:38:13] <dingo__> maybe
[00:38:32] <dingo__> idk..... its being very difficult to track down
[00:39:19] <dingo__> pthread gets involved at some point
[00:40:02] <dingo__> and now im reading some functions "fail to work corrrectly" when using -g, and -ggdb should be used
[00:40:24] <dingo__> which means complete recompile again, holding off for another 5 hours
[09:13:14] --- servus_ is now known as servus
[13:12:23] --- Lord_Nightmare is now known as LordNAway
[21:35:09] <dingo__> hey
[21:35:19] <dingo__> got it to work on openbsd/zaurus
[21:35:26] <dingo__> even mostly through standard configure.ac
[21:35:49] <dingo__> im going to try to patch it to differentiate between zaurus's qtopia linux and just plain arm
[21:35:58] <dingo__> of the #ifdef's in the code
[21:36:06] <dingo__> #ifdef arm and #ifdef qtopia_zaurus
[21:36:32] <dingo__> since theres some #ifdef's that work fine on arm but i guess not on qtopia
[21:36:34] <dingo__> like 2x scaling
[21:36:47] <dingo__> i'll do that for cvs
[21:36:56] <dingo__> i gotta make real patches against configure.ac and so on for an openbsd port
[21:37:03] <dingo__> maybe add the french patch as a flavor
[21:37:21] <dingo__> openbsd port would only be able to use 1.2, since its a release
[21:37:59] <dingo__> openbsd has the option to use any autoconf or automake version
[21:38:06] <dingo__> by exporting AUTOCONF_VERSION and AUTOMAKE_VERSION
[21:38:14] <dingo__> i picked 2.54 for autoconf and 1.4 for automake
[21:38:25] <dingo__> hope thats reasonable, as some warnings still come out
[21:40:19] <dingo__> and the strange crashes i had earlier were pthread
[21:40:21] <dingo__> due to using -g
[21:40:33] <dingo__> pthreads manpage on openbsd says -g can cause instability
[21:41:03] <dingo__> and sometimes it was hard to tell in gdb, but a thread would aquire attention, and thats when the stack got smashed and openbsd stepped in and knocked it down
[21:41:09] <dingo__> -ggdb fixes that