View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0001175 | FSSCP | --------- | public | 2006-12-19 04:41 | 2008-02-22 18:33 |
| Reporter | Mp-Ryan | Assigned To | taylor | ||
| Priority | normal | Severity | crash | Reproducibility | always |
| Status | resolved | Resolution | no change required | ||
| Product Version | 3.6.9 | ||||
| Summary | 0001175: Random crash to desktop accompanied by disk-access light | ||||
| Description | A recent error has started cropping up when playing some missions in FS2 and Derelict (appears to be random in the mission; no specific missions, though Dunkerque is particularly bad for it). During gameplay, the screen will suddenly freeze, sound will begin to loop, and the IDE-access light comes on and blinks continuously until the program is shut down by a Windows error. This only occurs when the 3.6.8 Zeta MediaVPs plus patches are in use. It has not occurred without them set as the active mod. Lightspeed's nebulas are installed in the same mod folder, however disabling them has no effect if the entire mediavp set is not disabled. In addition, disabling individual command-line features has no effect (except, of course, disabling the mediavp mod). Build: 3.6.9 RC7.9 MediaVPs in use: Those in the zpack (no adveffects or cell) Flags: C:\StrcApps\FreeSpace2\fs2_open_3_6_9-7dot9x.exe -mod Mods\mediavps -glow -jpgtga -mipmap -img2dds -no_vsync -cache_bitmaps -dualscanlines -targetinfo -snd_preload -fps I ran the debug build and managed to get the error information, as well as Microsoft's error report information. I am attaching them in a ZIP file. Incident 1 contains the Microsoft error information, the .rpt file from the debug, and errorlog.txt for the first time it crashed with debug enabled. The Microsoft error report is missing in Incident 2. The crash occurs quite regularly (these are simply example crashes, by no means are they the first ones). I have no idea what's causing it. Hopefully one of you gentlemen can figure it out. | ||||
| Tags | No tags attached. | ||||
|
2006-12-19 04:41
|
|
|
|
The problem appears to be in process_sobjects() in aicode.cpp, pss->system_info would appear to be an invalid ptr. Apparently one of a ships subsystem ptrs isn't initialized properly, or has been cleared by mistake by the code for some reason. But without having this happen to me while running the game through gdb, I doubt I'll be able to track it down any better than that. Try to narrow the problem down to a particular mission, a particular ship (or best guess at least) in one or more missions, and also try using a debug build to see if it still happens there. Without more info there just isn't quite enough to go on though. |
|
|
Taylor, Those posted files came from running it in a debug build - the log files are included in it. It nearly ALWAYS happens on Dunkerque, so it's got to be due to one of the models in that mission. Oddly enough, disabling motion debris reduced the frequency of the problem, so perhaps it relates to debris from one of the ships. If there are more files you need, let me know. However, I'm on vacation until january 7th so I can't post anything more until then. |
|
|
Bah, I was looking at the wrong thing! ;) ... Ok, it appears to actually be hitting and Int3() somewhere, but that info is stored in another log file. Please attach the data/fs2_open.log file from one of the sessions where it crashed. That should have a more precise account of what's going on. I'll run through Dunkerque a few times just to find out if I can trigger it as well. |
|
|
So, it appears I deleted my fs2_open.log file somewhere along the way while I was trying to figure out the crash. I'm going to upgrade to RC8 and see if I can generate this same issue... if so I'll post up everything again (and fs2_open.log). Otherwise I'll try with RC7.9 again. |
|
|
Couldn't generate the crash again in Dunkerque, but I did get it to occur in the first Awakenings mission when I enabled motion debris again. However... it didn't generate a fs2_open.log file (I was using build 3.6.9 final, not the debug build). Does it only generate this in debug mode? |
|
|
Yep, only in debug builds. |
|
|
OK, I'll try debug mode. Problem is the bug doesn't occur everytime the mission is played (except on Dunkerque, for RC7) with motion debris enabled. I'll see if I can get it to crash tomorrow and post it up. |
|
|
So far I can only reproduce the crash in 3.6.9 Final in the Awakenings campaign, and unfortunately the debug build freaks out at some of these missions to begin with. One mission where the crash has occurred several times gives me this debug error when I attempt to run it in the debug build (from jan. 12th). Warning: Couldn't fix up turret indices in spline path Model: cruiser01x.pof Path: $path02 Vertex: 1 Turret model id:13 This probably means that the turret was not specified in the ship table(s). File:J:\src\cvs\fs2_open_3_6_9.final\code\ship\Ship.cpp Line: 8737 [This filename points to the location of a file on the computer that built this executable] Call stack: ------------------------------------------------------------------ parse_create_object_sub() parse_create_object() mission_parse_maybe_create_parse_object() post_process_ships_wings() post_process_mission() parse_mission() parse_main() mission_load() game_start_mission() game_enter_state() gameseq_set_state() game_process_event() gameseq_process_events() game_main() WinMain()------------------------------------------------------------------ I will continue trying to get an f2s_open.log file from the crash, it just may take quite a while. 3.6.9 appears to have resolved the issue in Dunkerque, or at least lessened it so the crash does not occur everytime it is played with motion debris on. |
|
2007-01-24 05:07
|
|
|
|
Finally, I've gotten a crash in debug (this time, during a Transcend mission) that generated the fs2_open.log file you were looking for. I'm uploading it now. Hopefully this should shed some light on the issue. As I said, it doesn't happen as frequently in 3.6.9 final, but it's still often enough to be annoying. |
|
|
Reminder sent to: taylor Hey taylor, made progress and finally got you a copy of fs2_open.log for that crash issue. Hope it helps. |
|
2007-01-24 17:54
|
|
|
|
I've uploaded another fs2_open log file, this one from the crash occurring during the load sequence. |
|
|
I figured I'd add some more information I've noticed along the way. First off, it's occurring quite frequently in the Transcend campaign - generally I can only get through 2 or 3 missions (sometimes not even one) before the crash occurs. Obviously something in this campaign is tapping directly into whatever the source of the problem is. However, it's not predictable - some missions crash more often than others, but they do not ALWAYS crash. Furthermore, the crash most frequently occurs when multiple sounds are playing, often as not shortly after hitting the afterburner. My video card has been stably overclocked for some time (It's a Radeon 9500 Pro, clocked up to 312/300), however, I'm going to try defaulting the clocks again and see if that reduces the errors. All diagnostics on the card fail to produce artifacts or errors of any kind with the overclocking, but on the off chance this is affecting the crash rate (if not the generation of the crash itself) I'm going to try running it with the lower clocks. I'll let you know the results. I have a few hours tonight to play with all of this. |
|
|
Reducing clock speed reduced crash frequency, but did not dispense with the problem altogether. |
|
|
Are you sure you haven't got a faulty GPU or system? I mean, if it is a code bug, it shouldn't be dependent on which clock rates you set through your system. (I have a XFX nVidia 7800 GS Extreme (factory overclocked) which seems to run fine, (no artifacts or whatever test you do), even if I push it up a little more. BUT Farcry (and only this game till now, I don't know why) tends to crash to a black screen if I increase the overclocking. So I've just returned it to its default settings as I don't want to risk anything). |
|
|
I won't dismiss a system issue, since it's always possible, but it is unlikely. The card has been overclocked for months, and 3.6.7 always ran fine. FSO is also the only game that crashes with regular frequency - it's runs Half-Life 2 (to pick one example) quite well. |
|
|
* BUMP * Is this still a problem with 3.6.9-final? |
|
|
Yessir; though testing with the build you provided in Mantis report 1255 changed the crash behaviour (the gamma change disappeared for one; I can't remember the other differences offhand). Do you have any other recent experimental builds I can try? It does not appear to be solely related to GPU overclocking. |
|
|
Bumping again, since it's been a while. A lot of builds have been released since the last update, so do any of those fix the problem? |
|
|
I've been using the Xt1207 build for weeks now and haven't experienced a single crash. Video card is clocked higher than it was with 3.6.9. This appears to be resolved. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2006-12-19 04:41 | Mp-Ryan | New Issue | |
| 2006-12-19 04:41 | Mp-Ryan | Status | new => assigned |
| 2006-12-19 04:41 | Mp-Ryan | Assigned To | => Da_Brain |
| 2006-12-19 04:41 | Mp-Ryan | File Added: FS2CTD.zip | |
| 2006-12-20 11:02 | taylor | Note Added: 0007313 | |
| 2006-12-20 11:02 | taylor | Assigned To | Da_Brain => taylor |
| 2006-12-20 11:02 | taylor | Category | mediaVP => --------- |
| 2006-12-21 06:49 | Mp-Ryan | Note Added: 0007314 | |
| 2006-12-21 08:26 | taylor | Note Added: 0007315 | |
| 2007-01-08 02:04 | Mp-Ryan | Note Added: 0007422 | |
| 2007-01-09 03:28 | Mp-Ryan | Note Added: 0007428 | |
| 2007-01-09 04:00 | taylor | Note Added: 0007429 | |
| 2007-01-09 08:05 | Mp-Ryan | Note Added: 0007430 | |
| 2007-01-12 17:35 | Mp-Ryan | Note Added: 0007451 | |
| 2007-01-24 05:07 | Mp-Ryan | File Added: fs2_open.log | |
| 2007-01-24 05:07 | Mp-Ryan | Note Added: 0007536 | |
| 2007-01-24 05:08 | Mp-Ryan | Note Added: 0007537 | |
| 2007-01-24 17:54 | Mp-Ryan | File Added: fs2_open_2.log | |
| 2007-01-24 17:55 | Mp-Ryan | Note Added: 0007539 | |
| 2007-01-24 21:18 | Mp-Ryan | Note Added: 0007540 | |
| 2007-01-25 07:06 | Mp-Ryan | Note Added: 0007544 | |
| 2007-01-25 07:39 | ARSPR | Note Added: 0007546 | |
| 2007-01-25 07:41 | ARSPR | Note Edited: 0007546 | |
| 2007-01-25 07:41 | ARSPR | Note Edited: 0007546 | |
| 2007-01-25 16:13 | Mp-Ryan | Note Added: 0007548 | |
| 2007-05-08 04:26 | taylor | Note Added: 0008044 | |
| 2007-05-08 05:01 | Mp-Ryan | Note Added: 0008058 | |
| 2008-02-22 09:53 | taylor | Note Added: 0008896 | |
| 2008-02-22 18:28 | Mp-Ryan | Note Added: 0008899 | |
| 2008-02-22 18:33 | taylor | Status | assigned => resolved |
| 2008-02-22 18:33 | taylor | Resolution | open => no change required |
