Crash Offsets

? Offline

adoxa wrote:
Looks to be a missing/mistyped formation.

Maybe. I have changed TTX of vanilla ships, have made pilotable Lifter and Repair ship from vanilla, have added some new ships. And all

SDK - yours, JFLP ))
FLScan 1.3 says 0 errors

Mystic…

adoxa

@Thaddeus: v2.01.

@Helloween: It really looks like it wants a formation - it’s the function to test if the position is for NPC or player. It’s crashing because it can’t read the position data (if you look at the report, ECX should be 0). It appears neither FLScan nor FLEC pick up this error. Here’s a perl script to test it (run it in DATA\MISSIONS).

open(FORMATIONS, 'formations.ini') || die('no formations.ini');
while (<formations>)
{
  if (/^nickname = (.*)/)
  {
    $formations{$1} = 1;
  }
}
close(FORMATIONS);

open(PROPS, 'faction_prop.ini') || die('no faction_prop.ini');
while (<props>)
{
  ++$line;
  if (/^formation = .*?, (.*)/)
  {
    if (!$formations{$1})
    {
      print "Missing formation '$1' at line $line.\n";
    }
  }
}</props></formations>

? Offline

Thx Adoxa! Now i know where to dig )

Thaddeus

adoxa wrote:
@Thaddeus: v2.01.

Thanks problem is cleared. Gracias amigo

? Offline

adoxa wrote:
Here’s a new plugin for the content 124bd problem. This one only logs on a bad address (also eliminating the crash), so you should have no problem on a running server. It will generate EXE\EngBase-0124BD-bad.txt, logging the values of the previous call.

2011-06-30 02:29:37.481: Common.0630E681
Bad address = CDCDCDCD
Previous call:
cmp -> equipment\models\weapons\ku_hornet_launcher.cmp
part -> k_hornetlauncher_base_lod1020829120522.3db

2011-06-30 02:45:49.832: Common.0630E681
Bad address = CDCDCDCD
Previous call:
part -> equipment\models\torpedoes\ge_torpedo.3db

2011-06-30 21:56:49.318: Common.0630E681
Bad address = CDCDCDCD
Previous call:
cmp -> equipment\models\turret\li_turret03.cmp
part -> base021119202809.3db

2011-07-01 21:59:10.152: Common.0630E681
Bad address = CDCDCDCD
Previous call:
cmp -> equipment\models\weapons\ku_hornet_launcher.cmp
part -> k_hornetlauncher_base_lod1020829120522.3db

2011-07-02 01:51:27.616: Common.0630E681
Bad address = CDCDCDCD
Previous call:
cmp -> equipment\models\turret\ku_turret04.cmp
part -> base021119202046.3db

Got crashes - where to dig?

adoxa

Wasn’t this resolved to the lack of hardpoints (hpid) in the sur?

? Offline

adoxa wrote:
Wasn’t this resolved to the lack of hardpoints (hpid) in the sur?

I dunno about it, aby prooflink?

adoxa

EngBase-0124BD was refined to GetRoot, which was further refined to GetCollisions. Adding hpid to the sur was found to be the solution.

? Offline

I get Crash at Fault offset: 0x00021117 in Common.DLL

it is appeared here:

.text:06281105 loc_6281105:			       ; CODE XREF: DamageList::operator=(DamageList const &)+91j
.text:06281105		       mov     [eax+4],	edi
.text:06281108		       mov     [esi+4],	eax
.text:0628110B		       mov     edx, [eax+4]
.text:0628110E		       mov     [edx], eax
.text:06281110		       add     eax, 8
.text:06281113		       test    eax, eax
.text:06281115		       jz      short loc_6281128
.text:06281117		       mov     ecx, [ebx+8] ; <-here
.text:0628111A		       mov     [eax], ecx
.text:0628111C		       mov     edx, [ebx+0Ch]
.text:0628111F		       mov     [eax+4],	edx
.text:06281122		       mov     ecx, [ebx+10h]
.text:06281125		       mov     [eax+8],	ecx

Whats wrong may be?

Lord of the Hell[X]

Crashes detected by me or our server:

content.dll
000db0f1 - bad patrol(i think)
0009100f - bad tradelane(i think)
0009566d - bad tradelane(i think)
000d8e14
00016bb3
000ab23d
000d8e14
00013bb3

freelancer.exe
0013d13a
0013b395
001545f7
00006100- bad wireframe(Adoxa)

remoteserver.dll
00012460

common.dll
0010960c
0005be63 - non existing exclusion zone, but is called in asteroid/nebuda file - look to FlSpew.txt for zone nickname
00025cc9 - bad surface file
000f2597
00103141
00025cc9

x86math.dll
0000189f

rendcomp.dll
00011163 - bad hash at <vmeshlibrary>example <li_gunboat.3db.lod0.vms include=“li_gunbship_cmp\lod0.vms.xml”>but hash is li_gunboat.lod0.vms

thorn.dll
0002c9ae</li_gunboat.3db.lod0.vms></vmeshlibrary>

MasterofSocks

Uh
not sure if i post this here or not, but FLServer crashes on me before it finishes even starting up. Don’;t know if this is correct, but
Offset 0x00003de4 in DACOM.dll

? Offline

Got strange offset at flmaterials.dll

0x00005b33

At flspew.txt:

C:\work\builds\dalibs\dalibs-build\build\Src\RendComp\VMesh\VMPartArch.cpp(697) : NOTICE:General: VMESH: couldnt find material 65536
C:\work\builds\dalibs\dalibs-build\build\Src\RendComp\VMesh\VMPartArch.cpp(697) : NOTICE:General: VMESH: couldnt find material 327684
C:\work\builds\dalibs\dalibs-build\build\Src\RendComp\VMesh\VMPartArch.cpp(697) : NOTICE:General: VMESH: couldnt find material 589832
C:\work\builds\dalibs\dalibs-build\build\Src\RendComp\VMesh\VMPartArch.cpp(697) : NOTICE:General: VMESH: couldnt find material 851980
C:\work\builds\dalibs\dalibs-build\build\Src\RendComp\VMesh\VMPartArch.cpp(697) : NOTICE:General: VMESH: couldnt find material 1114128
C:\work\builds\dalibs\dalibs-build\build\Src\RendComp\VMesh\VMPartArch.cpp(697) : NOTICE:General: VMESH: couldnt find material 1376276

Tried to find this values by FL Developer - no luck

Happens mostly at one system - bw08 (Tau-23)

Digging fl forums i see not only me had this problem…
Someone solved?

FR

Try to remove

….
[Object]
nickname = Bw08_suprise_ge_train_1
ids_name = 261761
pos = 21340, 0, 22673
rotate = 30, 0, 20
archetype = suprise_ge_train
visit = 16
ids_info = 66426
loadout = SECRET_train_bw08
……

and check if it still happens.

I encountered various texture related problems
using the HD texture mod.
Dragging that one down was real pain.

? Offline

Yeah, b!g thanks F!R!
Works ferroconcretely

? Offline

Nova wrote:
Help please… i’m in desperation…
Freel server (mod) crashes when jumping inside Kepler system…
000c465f in content.dll

It is because of mixed non-patched and 1.1 patched files.
P.S. Alows to run Itano Circus Server (Woohoo)

Thaddeus

Getting a reasonably consistant ctd on freelancer.exe at offset 0x00149e35

multiple characters, always at launch to space. Only change in the mod has been updating the jflp.dll to the new ver. 1.21. Restored jflp back to version 1.2 with same results, I now have less than 50/50 chance of getting into space without crashing.

Any ideas? I know there is a thread on this offset, but it doesn’t draw any real conclusions and the idea of deleting all character files on the off chance that it might work would not go down well

Vital

Thaddeus wrote:
Getting a reasonably consistant ctd on freelancer.exe at offset 0x00149e35

multiple characters, always at launch to space. Only change in the mod has been updating the jflp.dll to the new ver. 1.21. Restored jflp back to version 1.2 with same results, I now have less than 50/50 chance of getting into space without crashing.

Any ideas? I know there is a thread on this offset, but it doesn’t draw any real conclusions and the idea of deleting all character files on the off chance that it might work would not go down well

This topic:
topic

What I found: the crash seems to be random.

After I uninstalled FLHook 1.6.9(88Flak FLHook) and changed it to FLHook 2.0 –-> NO CRASHES.

I use 1.21 JFLP by Jason Hood.

So the reason in my case seems to be 88Flak FLHook?!

Hope this helps.

Thaddeus

Thanks Vital. I’m using flhook 2.0 already. The problem used to be quite random as you say, but now its more than ever other time I launch I get dumped to desktop

robocop

And the 1.21 jflp.dll didn’t fix it? We had some minor issues with the client CTD’ing but Adoxa was able to isolate the issue within the dll and published the 1.21 which so far seems to have fixed our problem.

We do NOT use FLHook.

Vital

Yeah, btw, will the crash persist without FLHook??