IRC logs for #opendune on OFTC at 2009-10-10
⏴ go to previous day
08:02:33 *** Alberth has joined #openDune
08:37:52 <Alberth> Program Termination: jumped to 0AEC:01D0, which is not decompiled.
08:37:52 <Alberth> The jump was triggered at decompiled/cs__0AEC.c:550
08:37:52 <Alberth> The jump appears to originate from 0AEC:018F.
08:54:45 <Alberth> (I fear I am starting to report duplicates)
09:21:18 <TrueBrain> nope, all unique so far
10:32:26 <TrueBrain> so today again we are going to battle unifying functions :)
10:32:33 <TrueBrain> but today, I have a plan
10:35:41 <Alberth> a plan to conqueror Dune :)
10:36:15 <TrueBrain> that not yet: lvl5 is HARD! :p
10:36:30 <TrueBrain> but this plan should greatly reduce crashes :)
10:37:31 <Alberth> I seem to remember difficulty settings when starting the game, but I haven't seen them yet.
10:37:43 <TrueBrain> there is no difficult setting
10:37:57 <TrueBrain> it is said that Harkonnen is easy, and Atreidus is Hard, but I don't believe that :p
10:38:40 <Alberth> maybe I am confused with Dune2k
10:39:34 <TrueBrain> never played Dune2k :$
10:43:54 <Xaroth> mainly because of the last missions
10:44:10 <Xaroth> Death Hand > Fremen > Saboteur
10:44:22 <Xaroth> also, Devestator > Sonic Tank > Deviator
10:47:33 <Alberth> I never found the Fremen of much use tbh
10:50:31 <Xaroth> they are quite hard hitting
10:56:37 <Alberth> Xaroth: but quickly killed with any tank driving over them.
10:56:47 <Xaroth> yeh, AI isn't -that- smart :P
10:58:08 <Xaroth> would be more fun if they were any form of accurate :P
10:58:36 <SmatZ> save -> shoot -> didn't hit? load -> shoot ...
10:59:43 <TrueBrain> damn, I need to compare 2 nodes, if they are identical
10:59:46 <TrueBrain> turns out to be impossible :p
10:59:52 <TrueBrain> and I no longer have the bytecode which created them :(
12:08:18 <TrueBrain> step by step I am getting closer ... but one problem after the other appears
12:13:59 <TrueBrain> and again I have a big case of layer violation ... argh .........
12:14:25 <TrueBrain> well, it is my personal problem :)
12:14:33 <TrueBrain> I don't want to add to libtoc information about the caller-tree
12:14:39 <TrueBrain> that I want to leave to toc/static
12:14:56 <TrueBrain> so there needs to be some kind of communication back from libtoc to tell which functions became a bigger function
12:15:03 <TrueBrain> in order for him to produce the correct header information
12:49:44 <TrueBrain> fprintf(stderr, "ERROR: unhandled scenario where data is not linear in time. Expect errors.\n");
12:49:48 <TrueBrain> I love such solutions :p
12:53:31 <SmatZ> "data is not linear in time"?
12:53:45 <TrueBrain> yeah .. my way of saying something changed the memory on runtime :p
12:53:51 <TrueBrain> and at the same byte are 2 different values in time
12:53:57 <TrueBrain> linear is not the word :p
12:56:33 <SmatZ> so one global variable has two different meanings, ok?
12:56:47 <TrueBrain> for example: 1 time it jumps to A, the other time to B
12:57:35 * SmatZ doesn't want to waste your time by explaining that :)
12:57:43 <TrueBrain> you already know about that problem
12:57:51 <TrueBrain> as it was one of the big problems getting this to work in the first place ;)
12:57:57 <TrueBrain> VLAK .. PushOver ...
12:58:08 <TrueBrain> either protections or unpacking stuff
12:58:25 <TrueBrain> which makes bytes at, for example, cs:0100 different, depending on when it jumps to it
13:00:44 <TrueBrain> I solved most of it by introducing CRC and jumping to that
13:01:06 <TrueBrain> but now I want to combine functions, and it turns out to be very tricky to keep that in mind too .. and as Dune2 doesn't have that problem, it is easier for now to ignore it, and just warn like above :p
13:01:30 <TrueBrain> if I ever want to do a project with the problem, I need to rewrite a portion of the decompiler :p
13:02:58 <glx> valgrind may help to find what modifies it
13:03:12 <TrueBrain> euh, no, not really :)
13:03:24 <TrueBrain> valgrind runs on host-level .. we talk about emulation here
13:03:39 <TrueBrain> and I know exactly what modifies it, that is not the issue/problem here
13:03:51 <TrueBrain> a UPX executable for example
13:04:07 <TrueBrain> or any random executable with a tiny bit of decompiler-protection
13:04:36 <TrueBrain> the solution only requires something that counts time to the solution .. and that makes the problem 2D instead of 1D
13:04:47 <TrueBrain> and I don't want to waste my time solving that now, as it has no benefit to Dune2 :)
13:23:40 <TrueBrain> I should have made the stats file one per instruction, instead of per block tothe next jump :p
13:23:45 <TrueBrain> that would have created a nice amount of files :)
13:24:41 <Xaroth> around.. 560000000000000000000000000?
13:24:42 <TrueBrain> but at least it would be easier to solve a situation like: A B C (jump F) D E F G H I ... now I have a function A B C (jump F), A B C D E F G H I and F G H I .. and no sane way to tell if they have something in common .. in result it creates duplicated code
13:25:17 <glx> yeah I found the location of some strings :)
13:27:17 <TrueBrain> and I reverted all the work I did today :p
13:28:14 <TrueBrain> glx: how did you find them? and where? :)
13:28:54 <TrueBrain> 353F is global.h, so that is easy/good ;)
13:29:02 <TrueBrain> C:\DUNE2.EXE is generated at startup
13:29:34 <TrueBrain> glx: so you found the functions handling files ;)
13:29:58 <TrueBrain> I soon should make something simple, like 'function_name.txt' or what ever, which lists names of known functions, without C-ifying them :)
13:31:15 <TrueBrain> now first handling Alberth's crash-logs ..
13:31:44 <glx> anyway opendune doesn't like printfs :)
13:31:56 <glx> causes stack overflow after some time
13:32:05 <TrueBrain> threading problem :)
13:32:14 <TrueBrain> printf gives control back to the OS, and allows a thread to schedule
13:32:31 <TrueBrain> I had that problem a lot in the begin :)
13:32:34 <glx> oh like the interrupt thread
13:35:09 <DorpsGek> SVN: truebrain (r210) [JIT] -Add: 4 more problems found by Alberth
13:35:28 <DorpsGek> SVN: truebrain (r211) -Update (r210): update decompiled code
13:35:34 <TrueBrain> tnx again Alberth :)
13:36:36 <glx> so some things in GlobalData are indeed static char*
13:38:43 <TrueBrain> well, dunno about the 'static' part, but char * for sure ;)
13:42:00 <glx> hmm some strings are in 3eee
13:42:59 <TrueBrain> yes, 0x3EEE is one weird piece of memory
13:43:05 <TrueBrain> it is also related to overlays
13:43:24 <TrueBrain> but I guess there it stores which scenario it is working on
13:43:40 <TrueBrain> 44AF also passes by .. this is inside an overlay
13:43:50 <TrueBrain> (like a static const char * in a .c file)
13:45:34 <TrueBrain> no, most likely exactly that .. instead of reading from some global space, a local constant ..
13:45:47 <TrueBrain> most is global, but there are also big sets of local variables all over the place
13:46:13 <glx> hmm fartr.wsa is in 353f and 44af
13:46:30 <TrueBrain> there are plenty of bugs and ugly-ness ;)
13:47:28 <DorpsGek> SVN: truebrain (r212) -Fix: @init lines were wrong
13:48:09 <TrueBrain> so ... now a place for such naming function file ... euh ...
13:48:47 <TrueBrain> src/decompiled_name.hpp, with in it only names?
13:50:17 <TrueBrain> src/decompiled_names.txt
13:50:20 <TrueBrain> src/function_names.txt
13:50:58 *** Xaroth sets mode: +v Alberth
13:50:59 <glx> hmm src/function_names.txt seems good
13:51:05 *** Xaroth sets mode: +vv blathijs Yexo
13:51:10 *** Xaroth sets mode: +v proyvind
13:53:51 <Alberth> Xaroth: what does that v mean?
13:54:06 <TrueBrain> that you can now still speak :p
13:54:14 <glx> so you can talk when moderated :)
13:54:49 <TrueBrain> Alberth: it is mostly used to show to new users that the people with +v are respected users of the channel
13:55:03 <Xaroth> +voice overrides the mute of +b or +m
13:55:27 <TrueBrain> but the only real use I ahve seen it, is channel where +m is on by default, and when you join you get +v. Now when you start to go weird, you get -v
13:55:35 <TrueBrain> those channels have a strict no-ban policy :)
13:55:57 <Xaroth> +o for devs, +v for people who can ask a question
13:56:03 <Xaroth> -v when they are done asking
13:56:13 <Xaroth> like giving somebody the mic at a roundtable talk :P
14:00:29 <TrueBrain> 01F7:28B8:0024:83C9 emu_Tools_Memmove # (count, from_csip, to_csip)
14:00:31 <TrueBrain> 01F7:2947:0014:02B8 emu_Tools_Memset # (count, value, csip)
14:01:07 <glx> yes, though some times the args are not easy to find
14:01:19 <TrueBrain> 0C3A:1ABD:004E:E6F7 emu_GUI_Construction_Unknown # Something to do with the consturction panel on the right
14:02:02 <TrueBrain> the spaces between is a \t
14:02:19 <TrueBrain> and the comment is optional :)
14:10:53 <TrueBrain> I really don't know about naming yet, like the _Tools_ or _GUI_ part
14:10:56 <TrueBrain> just guessing there ;)
14:12:01 <TrueBrain> now the main.c is no longer generated
14:18:16 <Alberth> Program Termination: jumped to 4141:0115, which is not decompiled.
14:18:16 <Alberth> The jump was triggered at decompiled/cs__B50B.c:299
14:18:16 <Alberth> The jump appears to originate from B50B:010F.
14:19:45 <TrueBrain> you shouldn't be trying to do that :p :p
14:22:01 <Alberth> as my mentor says 'any enemy should be crushed as a matter of principle'
14:26:51 <TrueBrain> ==12591== All heap blocks were freed -- no leaks are possible.
14:27:32 <DorpsGek> SVN: truebrain (r213) -Add: allow a src/function_names.txt, which names function of which we know what they do (give or take), but not yet have made them C-like to be put in src/
14:28:05 <TrueBrain> any additions to src/function_names.txt, please send them to me, so I can generate new code out of it :)
14:31:09 <glx> like 01F7:2B75:002A:45D8,...emu_Tools_Open_File ?
14:31:21 <glx> not sure about the naming
14:33:52 <TrueBrain> any idea about the parameters?
14:34:38 <glx> for now I add what I found :)
14:36:26 <TrueBrain> that instruction is used exactly once ..
14:37:40 <DorpsGek> SVN: truebrain (r214) -Add: function-name for 01F7:2B75:002A:45D8 (glx)
14:39:42 <TrueBrain> hmm .. let me add 'native' first :p
14:40:10 <DorpsGek> SVN: truebrain (r215) -Add: SVN propset
14:40:15 <TrueBrain> can you make me a new patch? :p
14:40:38 <TrueBrain> Alberth: I tried loading your crash log, and it acts very weird :p
14:41:53 <Alberth> Let's forget that one then.
14:42:07 <TrueBrain> nah, it is not that .. it just suprised me that it did
14:42:21 <TrueBrain> either way, don't forget to instruct your harvester :p It was now doing nothing :p
14:43:17 <Alberth> I was winning a war, one cannot bother about everything :)
14:43:59 <TrueBrain> haha :) Yeah .. you took the full frontal approach :p
14:44:15 <glx> hmm it still adds extra lines
14:45:06 <DorpsGek> SVN: truebrain (r216) [JIT] -Add: 2 more problems found by Alberth
14:45:18 <DorpsGek> SVN: truebrain (r217) -Update (r216): update decompiled code
14:45:34 <Alberth> I sent one soldier in their direction, and then they all came to meet me :p
14:46:55 <TrueBrain> glx: ah, some whitespaces were added by me :)
14:47:17 <DorpsGek> SVN: truebrain (r218) -Add: a few more function names related to files (glx)
14:47:23 <TrueBrain> I btw renamed the functions to emu_File_XXX
14:47:57 <TrueBrain> dunno why .. it happened :p
14:49:36 <glx> oh extra lines are intended :)
14:55:43 <glx> hmm there are at least 2 write calls
14:55:45 <TrueBrain> yeah, I forgot I added them :p Ghehe :)
14:57:00 <TrueBrain> btw, FYI, what I gathered from fed2k forum, opening a file in dune2 is given to a wrapper, which first tries to load it from disk, then looks inside PAK files
14:58:35 <DorpsGek> SVN: truebrain (r219) -Add: even more file related function names (glx)
15:00:07 <glx> anyway I didn't check all int21 calls
15:00:26 <TrueBrain> we have to start somewhere .. any place as good as any other :)
15:01:40 <glx> 3 other files to check when I'm done with 01F7
15:09:38 <TrueBrain> hmm .. I tried to replace something in the libemu layer, but now it keeps on crashing on me :(
15:14:28 <TrueBrain> the small differences between lseek and fseek ...
15:17:38 <TrueBrain> it makes replacing the emu_File_XXX stuff much easier I guess ;)
15:25:57 <DorpsGek> SVN: truebrain (r220) [LibEMU] -Add: replace write/read/lseek/fileno and friends with fwrite/fread/fseek
16:03:14 <Alberth> Program Termination: jumped to 433B:06FA, which is not decompiled.
16:03:14 <Alberth> The jump was triggered at decompiled/cs__B511.c:1491
16:03:14 <Alberth> The jump appears to originate from B511:04D6.
16:03:14 <Alberth> loading a save game from the inital menu: game over in less than a second :)
16:04:10 <Alberth> at r220 both for libemu and trunk
16:04:58 <Alberth> the other savegame does work though
16:13:06 <TrueBrain> with that error, or a real game over?
16:47:53 <TrueBrain> Alberth: when I load that crashlog, the game exists cleanly?
16:52:32 <DorpsGek> SVN: truebrain (r221) [JIT] -Add: 1 more problem found by Alberth
16:52:47 <DorpsGek> SVN: truebrain (r222) -Update (r221): update decompiled code
17:03:23 <Alberth> I guess you already decided. I don't know, it crashses with that error without swapping to the game display, it seems
17:11:56 <TrueBrain> well, I did not decide anything, but your crashlog was valid :p
17:12:05 <TrueBrain> the functions in question resolve to what they should
17:13:16 <Alberth> Program Termination: jumped to 0C3A:270D, which is not decompiled.
17:13:16 <Alberth> The jump was triggered at decompiled/cs__0C3A.c:10582
17:13:16 <Alberth> The jump appears to originate from 0C3A:25FA.
17:14:22 <Alberth> loading the savegame now results in '[EMU] INT 0x21 AL 0x4C application termination', so I guess the file was broken.
17:14:36 <TrueBrain> maybe during saving it once crashed?
17:14:52 <Alberth> that may have happened yeah
17:15:10 <Alberth> the above report is when selecting the factory
17:15:30 <Alberth> from a valid save game just after starting harkonnen level 2 :)
17:16:28 <TrueBrain> I don't see a factory :p
17:19:04 <TrueBrain> lol, it is a 'failure' of something, that crashlog of yours :)
17:19:10 <TrueBrain> it failed to allocate a pointer
17:20:02 <TrueBrain> caused a nice cascade :)
17:20:31 <DorpsGek> SVN: truebrain (r223) [JIT] -Add: 4 more problems found by Alberth (cascade effect from 1 problem)
17:20:48 <DorpsGek> SVN: truebrain (r224) -Update (r223): update decompiled code
17:21:01 <Alberth> 4 for the price of 1 :)
17:21:08 <TrueBrain> yeah :p It was a nice one :)
17:35:02 <Alberth> Program Termination: jumped to 4343:027D, which is not decompiled.
17:35:02 <Alberth> The jump was triggered at decompiled/cs__B4F2.c:644
17:35:02 <Alberth> The jump appears to originate from B4F2:01DC.
17:35:02 <Alberth> never try to save a running game :p
17:35:35 <TrueBrain> I did a lot of saving, but somehow under certain conditions it keeps finding problem
17:49:39 <TrueBrain> bah bah bah .. I made the system so automated, that I now have a hard time faking data :p
17:55:47 <Alberth> Program Termination: jumped to 1391:05AF, which is not decompiled.
17:55:47 <Alberth> The jump was triggered at decompiled/cs__1391.c:1817
17:55:47 <Alberth> The jump appears to originate from 1391:05A0.
18:15:24 <TrueBrain> k, now wrote a piece of code which minimalizes the code, by removing duplicated trees .. but if the function it calls has something not resolved which is resolved when it would run down its own code, it fails on runtime ...
18:15:32 <TrueBrain> if you solve one problem, another immediatly shows :(
18:24:39 <Alberth> opendune: src/pool/unit.c:37: Unit_Get_ByMemory: Assertion `g_global->unitStartPos.csip <= address.csip && address.csip < g_global->unitStartPos.csip + sizeof(Unit) * UNIT_INDEX_MAX' failed.
18:24:39 <Alberth> zsh: abort (core dumped) ./opendune
18:24:55 <TrueBrain> so there is memory corruption going on :p
18:25:34 <TrueBrain> this time it clearly hit something critical :p
18:26:12 <TrueBrain> NULL POINTER ACCESS! :p
18:27:05 <TrueBrain> okay, that might really be me which wrongly translated commands
18:27:14 <TrueBrain> somehow in the cache-array there now is a null pointer
18:28:33 <TrueBrain> I will take it under review .. dunno if there really is anythig I can do about it :)
18:28:46 <TrueBrain> it is wrong to have a 0000:0000 in the cache array, but there is .. so who wrote it, and why ;)
18:31:46 <TrueBrain> bah, in my test-app my new code worked .. but with OpenDUNE it doesn't :(
18:32:09 <TrueBrain> - f__B4CD_009A_0003_2E57(); return;
18:32:10 <TrueBrain> + f__B4CD_0081_0002_CBBA(); return;
18:32:13 <TrueBrain> haha :) Sounds .... wrong?
18:32:55 <Alberth> unless functions move :)
18:33:40 <TrueBrain> funny enough, it is not wrong :)
18:33:47 <TrueBrain> function 81 has 1 job: jump to 9a :p
18:34:37 <TrueBrain> ERROR: jump goes to 261F:0008, but instruction suggested 3FF1:0289.
18:34:41 <TrueBrain> overlay overlay overlay ..
18:39:44 <TrueBrain> if (cs == 0x261F && ip != 0x0) continue; /* XXX -- Hardcoded Dune2 overlay manager */
18:39:47 <TrueBrain> the solution is always simple :p
18:40:33 <TrueBrain> reduced the unresolved entries from 4000+ to 1800 :p
18:40:57 <TrueBrain> well .. it is not a solution that works for all projects :p
18:41:08 <TrueBrain> bah ... OpenDUNE crashes .. hmm ..
18:41:55 <TrueBrain> I start to dislike overlay managers :p
18:43:51 <TrueBrain> oh wait, I already did
18:48:03 <Alberth> Program Termination: jumped to 2B6C:0276, which is not decompiled.
18:48:03 <Alberth> The jump was triggered at decompiled/cs__2B6C.c:1379
18:48:03 <Alberth> The jump appears to originate from 2B6C:0245.
18:48:03 <Alberth> another failed save attempt
18:54:53 <Alberth> Program Termination: jumped to 0AEC:1165, which is not decompiled.
18:54:53 <Alberth> The jump was triggered at decompiled/cs__0AEC.c:5142
18:54:53 <Alberth> The jump appears to originate from 0AEC:10BA.
18:54:53 <Alberth> starting the second light vehicle build
19:02:39 <Alberth> Program Termination: jumped to 1391:05AF, which is not decompiled.
19:02:39 <Alberth> The jump was triggered at decompiled/cs__1391.c:1817
19:02:39 <Alberth> The jump appears to originate from 1391:05A0.
19:02:39 <Alberth> in the heat of the battle
19:14:58 <Alberth> Program Termination: jumped to 0972:11B9, which is not decompiled.
19:14:58 <Alberth> The jump was triggered at decompiled/cs__15C2.c:1959
19:14:58 <Alberth> The jump appears to originate from 15C2:0932.
19:14:58 <Alberth> I seemed to get additional forces
19:17:31 <TrueBrain> ARGH! FUCKING RETARTED OVERLAY! I am working on you for an hour now ..
19:17:40 <TrueBrain> the only bug I have is there .. and I can't find a clean way around it
19:18:52 <Alberth> Program Termination: jumped to 1391:05AF, which is not decompiled.
19:18:52 <Alberth> The jump was triggered at decompiled/cs__1391.c:1817
19:18:52 <Alberth> The jump appears to originate from 1391:05A0.
19:18:52 <Alberth> this is a duplicate crash, judging from the crash text
19:19:04 <TrueBrain> I will run your files in a sec :)
19:19:24 <Alberth> Nickman87 wants something from me :)
19:26:57 <TrueBrain> finally the overlay is playing nice
19:28:30 <DorpsGek> SVN: truebrain (r225) -Update: avoid as much code duplication as possible, to join branches as soon as they are detected, at function level
19:29:14 <TrueBrain> I believe it only removes 10,000 C lines .. sad :p
19:32:41 <TrueBrain> Alberth: the 3rd, the one with 1391:05AF, is not what crash.bin says :s
19:32:52 <TrueBrain> oh, it is .. I was reading with my ass :(
19:33:20 <Alberth> otherwise try the last one :)
19:33:29 <TrueBrain> the 4rd is the one which is a bit odd
19:33:31 <Alberth> I got a backup just in case :)
19:34:33 <TrueBrain> when I resume thatone, it doesn't give the wanted jump
19:35:23 <TrueBrain> I had 6 crash-logs from you, not 4 :p
19:35:46 <Alberth> I lost count a long time ago ;)
19:40:34 <Xaroth> <@TrueBrain> I believe it only removes 10,000 C lines .. sad :p << only?
19:40:34 <TrueBrain> problem is that I manually need to confirm new links he created .. sucky job :p
19:41:14 <TrueBrain> from the 150,000? yes, only
19:41:30 <Xaroth> well, 1/15th is still noticable :P
19:49:27 <DorpsGek> SVN: truebrain (r226) [JIT] -Add: 6 more problems found by Alberth
19:49:44 <DorpsGek> SVN: truebrain (r227) -Update (r226): update decompiled code
19:49:52 <TrueBrain> there was one really pretty one in there :)
19:50:14 <Alberth> I seem to bring new problems only :p
19:50:51 <TrueBrain> well, the new code should be vastly more stable
19:50:59 <TrueBrain> still, it will give problems :) At least now problems will result in real new code :p
19:51:37 <TrueBrain> visiting Mentat is still broken :p
19:52:00 <TrueBrain> either way, code should now be more readable
19:52:15 <TrueBrain> now combining should be much easier :)
19:52:29 <Alberth> whoow you touched a lot of files :)
19:53:15 <TrueBrain> why do I have the idea it didn't work as expected in all cases .. hmm ..
20:04:49 <TrueBrain> slashed away another 5000 lines
20:05:04 <TrueBrain> I still expected much more, but okay .. about 10% was duplicated code
20:06:10 <DorpsGek> SVN: truebrain (r228) -Fix (r225): the decompiler didn't take 'nop', 'segment override' and other commands like that into account when calculating the length of the function that might share parts of the current function
20:06:31 <TrueBrain> Xaroth: now in 0C3A it is immediatly clear what a jump skips :)
20:06:38 <TrueBrain> if (emu_flags.zf) { f__0C3A_0095_0085_1C53(); return; }
20:06:40 <TrueBrain> emu_lfp(&emu_es, &emu_bx, &emu_get_memory16(emu_ss, emu_bp, -0x4));
20:06:41 <TrueBrain> emu_movw(&emu_get_memory16(emu_es, emu_bx, 0x54), 0x0);
20:06:43 <TrueBrain> f__0C3A_0095_0085_1C53(); return;
20:17:52 <Alberth> I get a lot of failures to respond to keys, mouse, etc now. It just seems to ignore all my input.
20:18:40 <TrueBrain> hmm .. I seem to never have compiled the latest version .... ghehe :)
20:19:10 <Alberth> possibly when I give a lot of commands
20:19:12 <TrueBrain> all works fine here
20:19:21 <TrueBrain> or you went to Mentat ;)
20:19:44 <TrueBrain> move over Options, if the button pressed, but doesn't operate, you are fucked :p
20:19:44 <Alberth> sometimes the mentat comes to me :)
20:19:49 <TrueBrain> still have no clue what happens ...
20:20:02 <TrueBrain> it has to do with the hardware interrupt ... but that is all I understand :)
20:21:07 <Alberth> ok. Just making sure it is a known problem.
20:21:18 <TrueBrain> sadly enough it is :(
20:23:33 <Xaroth> aw man, now i can almost re-do my thing :P
20:23:49 <Xaroth> (makes it easier for me to correct my mistakes)
20:23:50 <TrueBrain> Xaroth: it will get worse in the next week I hope :p Neverthelss, thereis no real need to do so
20:27:07 <TrueBrain> tnx for all the reports :)
20:27:16 <TrueBrain> if (emu_get_memory16(emu_ds, 0x00, 0x37A0) == 0x0) {} else {
20:27:18 <TrueBrain> emu_lfp(&emu_es, &emu_bx, &emu_get_memory16(emu_ss, emu_bp, -0x4));
20:27:19 <TrueBrain> emu_get_memory16(emu_es, emu_bx, 0x54) = 0x0;
20:27:22 <TrueBrain> f__0C3A_0095_0085_1C53(); return;
20:27:24 <TrueBrain> the above piece of code even more 'optimized'
20:27:42 <Alberth> eh, it looks like a program :p
20:28:30 <TrueBrain> I also ened to find a way to add \t ...
20:45:42 <TrueBrain> just added 150 'prefix'
20:45:50 <TrueBrain> it couldn't be done automated :p
20:48:01 <TrueBrain> if (emu_get_memory16(emu_ds, 0x00, 0x37A0) != 0x0) {
20:48:02 <TrueBrain> emu_lfp(&emu_es, &emu_bx, &emu_get_memory16(emu_ss, emu_bp, -0x4));
20:48:04 <TrueBrain> emu_get_memory16(emu_es, emu_bx, 0x54) = 0x0;
20:48:07 <TrueBrain> f__0C3A_0095_0085_1C53(); return;
20:53:28 <TrueBrain> lets see how OpenDUNE likes it ..
20:56:59 <TrueBrain> hmm .. then it no longer starts :'(
21:27:07 <TrueBrain> solved another 700 functions ... just 1086 left, which can result in a crash :)
21:28:38 <DorpsGek> SVN: truebrain (r229) -Update: the decompiler can now detect simple if-statements and show them in a readable form
21:28:53 <TrueBrain> please test OpenDUNE, the stability should have increased dramaticly
21:30:16 <TrueBrain> and the readability of the code is really really improved :)
21:31:17 <TrueBrain> I haven't enabled 2 other optimizations; one because I simply didn't do it .. the other because I am not 100% sure it works correctly :p
21:31:49 <TrueBrain> now combining functions and using labels ....
21:33:24 <TrueBrain> k, I can finish A1 without any crash :)
21:34:39 <TrueBrain> whoho, I had the security question correct :)
21:34:56 <TrueBrain> Xaroth: you too: please test!
21:35:11 <TrueBrain> if this is stable, tomorrow I will introduce a few other optimizations
21:35:13 <TrueBrain> for now: good night
22:42:56 <Xaroth> 1>libemu.lib(register.obj) : warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/OPT:ICF' specification
22:43:00 <Xaroth> 1>LINK : warning LNK4098: defaultlib 'MSVCRTD' conflicts with use of other libs; use /NODEFAULTLIB:library
22:49:21 <Xaroth> Program Termination: jumped to 4343:046B, which is not decompiled. The jump was triggered at ..\decompiled\cs__B4F2.c:1478 The jump appears to originate from B4F2:03EC.
22:50:29 <Xaroth> Program Termination: jumped to 10E4:01AE, which is not decompiled. The jump was triggered at ..\decompiled\cs__10E4.c:407 The jump appears to originate from 10E4:0135.
22:52:00 <Xaroth> Program Termination: jumped to 3FF1:0436, which is not decompiled. The jump was triggered at ..\decompiled\cs__B4B8.c:1359 The jump appears to originate from B4B8:0431.
22:52:09 <Xaroth> all 3 when starting, or when just playing first mission
22:52:18 <Xaroth> picked different house for the 10E4 one
22:53:22 <Xaroth> almost looks like it's old code
22:54:26 <Xaroth> that might be, because it is, in fact, OLD CODE
22:54:32 <Xaroth> the output dir got moved -_-
22:55:23 <Xaroth> Assertion failed: g_global->unitStartPos.csip <= address.csip && address.csip <
22:55:26 <Xaroth> g_global->unitStartPos.csip + sizeof(Unit) * UNIT_INDEX_MAX, file ..\src\pool\un
22:56:35 <Xaroth> something wierd is happening though
22:56:49 <Xaroth> if i click options it instantly tried to go to load a game
22:57:09 <Xaroth> then i clicked cancel and it went back to the menu, and a second later to 'pick another house'
22:58:13 <Xaroth> random gfx glitch in between, preceeded by a random popup of the options screen :o
22:59:02 <Xaroth> I -think- there's some random memory corruption
22:59:08 <Xaroth> seeing i just randomly won A1 o-O
22:59:32 <Xaroth> like, random things being clicked
23:00:40 <Xaroth> might have to do with pressing the push-to-talk button for vent :o
23:00:49 <Xaroth> and just failed the questsions :(
23:29:48 <Xaroth> tb: ignore the termination jumps.... the other stuff does apply to the new compile
continue to next day ⏵