The Ur-Quan Masters Discussion Forum

The Ur-Quan Masters Re-Release => Technical Issues => Topic started by: v.v.b. on June 18, 2009, 07:58:53 pm



Title: UQM on Eee PC
Post by: v.v.b. on June 18, 2009, 07:58:53 pm
 Are there any plans to add an option to start UQM in 'wide-screen' resolutions?
 On Eee PC I have to play in 800x600 fullscreen, but 'coz of 1024x600 native resolution all seems SO wide...
 And every time after quitting is segfault. Is it normal? I have uqm-0.6.2-3 (fedora core 10). 'coz of this resolution is not switched back. And mouse hangs forever. So I have to restart X.


Title: Re: UQM on Eee PC
Post by: meep-eep on June 18, 2009, 10:14:34 pm
The images which are used are bitmaps, and they are created for a 4:3 screen ratio. To support another ratio, either new images would have to be created, or the layout of the screen would have to change. New images is unlikely to happen, because we don't have the graphics people for that. A different layout may be possible some day in the future, but the only realistic near-term "solution" would be black bars on the sides. It may be a good idea to create a feature enhancement request for this in our bug database (https://bugs.uqm.stack.nl/).

Segfaulting after quitting is definitely not normal. As the UQM core team has nothing to do with the Fedora packages, I suggest that you file a bug report with the Fedora package maintainer. Or, if you are able to compile UQM yourself, you could do that, and if the problem still occurs, we may be able to help.
For now, I can only offer you a possible way to restore your resolution after a crash: "xrandr -s 1024x600".


Title: Re: UQM on Eee PC
Post by: Novus on June 19, 2009, 09:54:33 am
Quote from: Changelog
- Added --keepaspectratio to keep correct aspect ratio when using
  custom resolutions in OpenGL mode - Mika
This feature already exists in the current SVN version. ;D

If you don't want to use the SVN version, you could try configuring your X server to maintain aspect ratio while scaling (this solves the aspect ratio issue for lots of software in one go). Also, disabling mode switching support in your X server could serve as a very crude workaround.


Title: Re: UQM on Eee PC
Post by: v.v.b. on June 20, 2009, 03:17:30 pm
 Thank you, Novus!
 I've compile uqm from svn and finally I've get what I want -- fullscreen on EeePC without distortion.
 But still there is a segfault after quitting. 8-(


Title: Re: UQM on Eee PC
Post by: Novus on June 20, 2009, 09:41:55 pm
I've compile uqm from svn and finally I've get what I want -- fullscreen on EeePC without distortion.
Good. This should also mean that your desktop resolution and UQM's are the same, so at least UQM leaves your screen in the right state when it crashes.

Quote
But still there is a segfault after quitting. 8-(
Not good. Can you tell us anything about what UQM is doing when it crashes? A good starting point would be the console output and a stack trace (produced, for example, by running UQM in gdb and using the "bt" command after UQM has crashed; I recommend running UQM in windowed mode when doing this).

You can often get your mouse cursor moving again by starting and exiting some other program (e.g. another game) that grabs the mouse and releases it.


Title: Re: UQM on Eee PC
Post by: v.v.b. on June 21, 2009, 03:02:37 pm
Here is console output:
Code:
[vvb@localhost ~]$ /usr/local/games/bin/uqm
The Ur-Quan Masters v0.6.6 (compiled Jun 20 2009 13:46:25)
This software comes with ABSOLUTELY NO WARRANTY;
for details see the included 'COPYING' file.

Netplay protocol version 0.3. Minimum required remote UQM version is 0.5.4.
Initializing base SDL functionality.
Using SDL version 1.2.13 (compiled with 1.2.13)
Using config dir '/home/vvb/.uqm/'
Using '/usr/local/games/share/uqm/content' as base content dir.
4 available addon packs.
    1. .svn
    2. 3domusic
    3. 3dovideo
    4. 3dovoice
Saved games are kept in /home/vvb/.uqm/save/.
Initializing SDL with OpenGL support.
SDL driver used: x11
SDL initialized.
Initializing Screen.
Set the resolution to: 1024x600x32 (surface reports 1024x600x32)
OpenGL renderer: Mesa DRI Intel(R) 915GM GEM 20080716 x86/MMX/SSE2 version: 1.4 Mesa 7.3-devel
Screen scalers are using SSE/MMX-Ext/MMX code
Initializing SDL audio subsystem.
SDL audio subsystem initialized.
Opening SDL audio device.
    using alsa at 44100 Hz 16 bit stereo, 940 samples audio buffer
Initializing mixer.
Mixer initialized.
Initializing sound decoders.
Sound decoders initialized.
Loading resouce index 'uqm.rmp'
Loading resouce index '3domusic.rmp'
Loading resouce index '3dovoice.rmp'
Loading resouce index '3dovideo.rmp'
'base/uqm.ct' -- 104466 bytes
We've loaded the Kernel
Trying to get type of undefined resource 'slides.logo'
'base/ui/title.ani' -- 20 bytes
Thread 'fade transform' done (returned 0).
'base/ships/arilou/skiff-icons.ani' -- 56 bytes
'base/ships/arilou/skiff-meleeicons.ani' -- 62 bytes
'base/ships/arilou/skiff.txt' -- 407 bytes
'base/ships/chmmr/avatar-icons.ani' -- 56 bytes
'base/ships/chmmr/avatar-meleeicons.ani' -- 66 bytes
'base/ships/chmmr/avatar.txt' -- 353 bytes
'base/ships/human/cruiser-icons.ani' -- 58 bytes
'base/ships/human/cruiser-meleeicons.ani' -- 67 bytes
'base/ships/human/cruiser.txt' -- 331 bytes
'base/ships/orz/nemesis-icons.ani' -- 60 bytes
'base/ships/orz/nemesis-meleeicons.ani' -- 68 bytes
'base/ships/orz/nemesis.txt' -- 389 bytes
'base/ships/pkunk/fury-icons.ani' -- 54 bytes
'base/ships/pkunk/fury-meleeicons.ani' -- 62 bytes
'base/ships/pkunk/fury.txt' -- 329 bytes
'base/ships/shofixti/scout-icons.ani' -- 56 bytes
'base/ships/shofixti/scout-meleeicons.ani' -- 62 bytes
'base/ships/shofixti/scout.txt' -- 395 bytes
'base/ships/spathi/eluder-icons.ani' -- 57 bytes
'base/ships/spathi/eluder-meleeicons.ani' -- 66 bytes
'base/ships/spathi/eluder.txt' -- 382 bytes
'base/ships/supox/blade-icons.ani' -- 55 bytes
'base/ships/supox/blade-meleeicons.ani' -- 63 bytes
'base/ships/supox/blade.txt' -- 321 bytes
'base/ships/thraddash/torch-icons.ani' -- 56 bytes
'base/ships/thraddash/torch-meleeicons.ani' -- 64 bytes
'base/ships/thraddash/torch.txt' -- 339 bytes
'base/ships/utwig/jugger-icons.ani' -- 58 bytes
'base/ships/utwig/jugger-meleeicons.ani' -- 66 bytes
'base/ships/utwig/jugger.txt' -- 315 bytes
'base/ships/vux/intruder-icons.ani' -- 62 bytes
'base/ships/vux/intruder-meleeicons.ani' -- 68 bytes
'base/ships/vux/intruder.txt' -- 249 bytes
'base/ships/yehat/terminator-icons.ani' -- 66 bytes
'base/ships/yehat/terminator-meleeicons.ani' -- 73 bytes
'base/ships/yehat/terminator.txt' -- 425 bytes
'base/ships/melnorme/trader-icons.ani' -- 57 bytes
'base/ships/melnorme/trader-meleeicons.ani' -- 66 bytes
'base/ships/melnorme/trader.txt' -- 369 bytes
'base/ships/druuge/mauler-icons.ani' -- 57 bytes
'base/ships/druuge/mauler-meleeicons.ani' -- 66 bytes
'base/ships/druuge/mauler.txt' -- 333 bytes
'base/ships/ilwrath/avenger-icons.ani' -- 58 bytes
'base/ships/ilwrath/avenger-meleeicons.ani' -- 68 bytes
'base/ships/ilwrath/avenger.txt' -- 365 bytes
'base/ships/mycon/podship-icons.ani' -- 60 bytes
'base/ships/mycon/podship-meleeicons.ani' -- 68 bytes
'base/ships/mycon/podship.txt' -- 329 bytes
'base/ships/slylandro/probe-icons.ani' -- 54 bytes
'base/ships/slylandro/probe-meleeicons.ani' -- 64 bytes
'base/ships/slylandro/probe.txt' -- 367 bytes
'base/ships/umgah/drone-icons.ani' -- 56 bytes
'base/ships/umgah/drone-meleeicons.ani' -- 62 bytes
'base/ships/umgah/drone.txt' -- 367 bytes
'base/ships/urquan/dreadnought-icons.ani' -- 67 bytes
'base/ships/urquan/dreadnought-meleeicons.ani' -- 76 bytes
'base/ships/urquan/dreadnought.txt' -- 425 bytes
'base/ships/zoqfotpik/stinger-icons.ani' -- 60 bytes
'base/ships/zoqfotpik/stinger-meleeicons.ani' -- 68 bytes
'base/ships/zoqfotpik/stinger.txt' -- 367 bytes
'base/ships/syreen/penetrator-icons.ani' -- 66 bytes
'base/ships/syreen/penetrator-meleeicons.ani' -- 73 bytes
'base/ships/syreen/penetrator.txt' -- 343 bytes
'base/ships/kohrah/marauder-icons.ani' -- 60 bytes
'base/ships/kohrah/marauder-meleeicons.ani' -- 70 bytes
'base/ships/kohrah/marauder.txt' -- 429 bytes
'base/ships/androsynth/guardian-icons.ani' -- 62 bytes
'base/ships/androsynth/guardian-meleeicons.ani' -- 69 bytes
'base/ships/androsynth/guardian.txt' -- 389 bytes
'base/ships/chenjesu/broodhome-icons.ani' -- 62 bytes
'base/ships/chenjesu/broodhome-meleeicons.ani' -- 72 bytes
'base/ships/chenjesu/broodhome.txt' -- 369 bytes
'base/ships/mmrnmhrm/xform-icons.ani' -- 54 bytes
'base/ships/mmrnmhrm/xform-meleeicons.ani' -- 64 bytes
'base/ships/mmrnmhrm/xform.txt' -- 395 bytes
'base/fonts/starcon.fon' -- 1 bytes
'base/fonts/tiny.fon' -- 1 bytes
'base/ui/activity.ani' -- 239 bytes
'base/ui/status.ani' -- 132 bytes
'base/gamestrings.txt' -- 15532 bytes
'base/fonts/micro.fon' -- 1 bytes
'base/ui/menu.snd' -- 56 bytes
_GetSoundBankData(): loading base/ui/menusnd01.wav
_GetSoundBankData(): decoded_bytes 1971
_GetSoundBankData(): loading base/ui/menusnd02.wav
_GetSoundBankData(): decoded_bytes 1955
_GetSoundBankData(): loading base/ui/menusnd03.wav
_GetSoundBankData(): decoded_bytes 1576
_GetSoundBankData(): loading base/ui/menusnd04.wav
_GetSoundBankData(): decoded_bytes 2399
'base/battle/stars.ani' -- 60 bytes
'base/battle/boom-big.ani' -- 209 bytes
'base/battle/boom-med.ani' -- 211 bytes
Thread 'fade transform' done (returned 0).
'base/battle/boom-sml.ani' -- 212 bytes
'base/battle/blast-big.ani' -- 192 bytes
'base/battle/blast-med.ani' -- 192 bytes
'base/battle/blast-sml.ani' -- 192 bytes
'base/battle/asteroid-big.ani' -- 618 bytes
'base/battle/asteroid-med.ani' -- 578 bytes
'base/battle/asteroid-sml.ani' -- 578 bytes
Thread 'fade transform' done (returned 0).
Thread 'fade transform' done (returned 0).
'base/ui/newgame.ani' -- 171 bytes
Trying to get undefined resource 'music.mainmenu'
Thread 'fade transform' done (returned 0).
Thread 'flash rectangle' done (returned 0).
'base/ui/playmenu.ani' -- 1177 bytes
Thread 'fade transform' done (returned 0).
Trying to get null resource
'base/ships/drone/drone-meleeicons.ani' -- 31 bytes
Trying to get null resource
'base/ui/flagshipstatus.ani' -- 789 bytes
'base/ui/miscdata.ani' -- 2548 bytes
'base/fonts/fontgradient.ani' -- 174 bytes
'base/nav/orbitbackground.ani' -- 698 bytes
'base/nav/flagship.ani' -- 390 bytes
'base/lander/orbit.ct' -- 6978 bytes
'base/nav/planets.ani' -- 2422 bytes
'base/nav/truespace.ct' -- 6978 bytes
'base/nav/truespacesun.ani' -- 144 bytes
'addons/3domusic/space.ogg' -- 2314527 bytes
_GetMusicData(): loading addons/3domusic/space.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'base/lander/lander.ani' -- 1488 bytes
'base/lander/hazard/quake.ani' -- 473 bytes
'base/lander/hazard/lightning.ani' -- 433 bytes
'base/lander/hazard/lavaspot.ani' -- 315 bytes
'base/lander/shield.ani' -- 342 bytes
'base/lander/launch.ani' -- 720 bytes
'base/lander/return.ani' -- 531 bytes
'base/lander/lander.snd' -- 169 bytes
_GetSoundBankData(): loading base/lander/land_bit.wav
_GetSoundBankData(): decoded_bytes 1433
_GetSoundBankData(): loading base/lander/land_qua.wav
_GetSoundBankData(): decoded_bytes 4811
_GetSoundBankData(): loading base/lander/land_lig.wav
_GetSoundBankData(): decoded_bytes 9918
_GetSoundBankData(): loading base/lander/land_hot.wav
_GetSoundBankData(): decoded_bytes 865
_GetSoundBankData(): loading base/lander/land_hrt.wav
_GetSoundBankData(): decoded_bytes 2696
_GetSoundBankData(): loading base/lander/land_sht.wav
_GetSoundBankData(): decoded_bytes 1087
_GetSoundBankData(): loading base/lander/land_hit.wav
_GetSoundBankData(): decoded_bytes 1888
_GetSoundBankData(): loading base/lander/land_can.wav
_GetSoundBankData(): decoded_bytes 1615
_GetSoundBankData(): loading base/lander/land_get.wav
_GetSoundBankData(): decoded_bytes 2129
_GetSoundBankData(): loading base/lander/land_ful.wav
_GetSoundBankData(): decoded_bytes 1804
_GetSoundBankData(): loading base/lander/land_lau.wav
_GetSoundBankData(): decoded_bytes 8341
_GetSoundBankData(): loading base/lander/land_awa.wav
_GetSoundBankData(): decoded_bytes 6396
_GetSoundBankData(): loading base/lander/land_die.wav
_GetSoundBankData(): decoded_bytes 6887
'base/nav/orbview.ani' -- 446 bytes
'addons/3domusic/orbit1.ogg' -- 1263226 bytes
_GetMusicData(): loading addons/3domusic/orbit1.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit2.ogg' -- 1191873 bytes
_GetMusicData(): loading addons/3domusic/orbit2.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit3.ogg' -- 1312469 bytes
_GetMusicData(): loading addons/3domusic/orbit3.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit4.ogg' -- 1320368 bytes
_GetMusicData(): loading addons/3domusic/orbit4.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit5.ogg' -- 1302891 bytes
_GetMusicData(): loading addons/3domusic/orbit5.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
Thread 'game clock' blocking on semaphore 'Clock'
Thread 'fade music' done (returned 1).
Thread 'fade transform' done (returned 0).
Thread 'game clock' awakens, released from semaphore 'Clock'
Thread 'game clock' blocking on semaphore 'Clock'
Thread 'game clock' awakens, released from semaphore 'Clock'
Thread 'game clock' done (returned 0).
'base/ui/newgame.ani' -- 171 bytes
Thread 'fade transform' done (returned 0).
Trying to get undefined resource 'music.mainmenu'
Thread 'fade transform' done (returned 0).
Thread 'fade transform' done (returned 0).
Thread 'flash rectangle' done (returned 0).
Thread 'Starcon2Main' done (returned 0).
Segmentation fault


Title: Re: UQM on Eee PC
Post by: v.v.b. on June 21, 2009, 03:13:11 pm
bt in gdb gives
Code:
#0  0x005c6e6d in ?? ()
#1  0x00e7f5ca in start_thread () from /lib/libpthread.so.0
#2  0x0026104e in clone () from /lib/libc.so.6

or should I post all gdb log?


Title: Re: UQM on Eee PC
Post by: Novus on June 22, 2009, 09:08:27 am
bt in gdb gives
Code:
#0  0x005c6e6d in ?? ()
#1  0x00e7f5ca in start_thread () from /lib/libpthread.so.0
#2  0x0026104e in clone () from /lib/libc.so.6

or should I post all gdb log?
This is what I requested. Thanks.

Tracking the underlying bug down is going to be a pain, though, since the only thing we can tell from this is that some thread is falling over just before returning. That return address looks suspiciously like ASCII text; we could have a stack corruption bug.

Edit: I just realised that you're using a release build. I should have requested that you compile UQM as a debug build (this is the default, isn't it?); otherwise, gdb can't give any meaningful information about what part of UQM is executing. In any case, judging by the console output, the audio decoder thread isn't stopping properly. Does turning sound off ("--sound=none") eliminate the segfault?


Title: Re: UQM on Eee PC
Post by: meep-eep on June 22, 2009, 01:46:44 pm
I'd also suggest "thread apply bt", to show the backtrace of all threads.


Title: Re: UQM on Eee PC
Post by: v.v.b. on June 22, 2009, 07:25:09 pm
If I turn sound to off (--sound=none), there is no segfault. I've test it several times and all is ok.
May be you'll find the end of gdb log slightly interesting:
Code:
................................................................................
_GetSoundBankData(): decoded_bytes 6887
'base/nav/orbview.ani' -- 446 bytes
'addons/3domusic/orbit1.ogg' -- 1263226 bytes
_GetMusicData(): loading addons/3domusic/orbit1.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit2.ogg' -- 1191873 bytes
_GetMusicData(): loading addons/3domusic/orbit2.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit3.ogg' -- 1312469 bytes
_GetMusicData(): loading addons/3domusic/orbit3.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit4.ogg' -- 1320368 bytes
_GetMusicData(): loading addons/3domusic/orbit4.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
'addons/3domusic/orbit5.ogg' -- 1302891 bytes
_GetMusicData(): loading addons/3domusic/orbit5.ogg
    decoder: Ogg Vorbis, rate 44100 format 17
Thread 'game clock' blocking on semaphore 'Clock'
[New Thread 0xb6f8bb90 (LWP 2553)]
Thread 'fade music' done (returned 1).
[Thread 0xb6f8bb90 (LWP 2553) exited]
[New Thread 0xb6f8bb90 (LWP 2554)]
Thread 'fade transform' done (returned 0).
[Thread 0xb6f8bb90 (LWP 2554) exited]
Thread 'game clock' awakens, released from semaphore 'Clock'
Thread 'game clock' blocking on semaphore 'Clock'
Thread 'game clock' awakens, released from semaphore 'Clock'
Thread 'game clock' done (returned 0).
[Thread 0xb2fcdb90 (LWP 2552) exited]
[New Thread 0xb6f8bb90 (LWP 2555)]
'base/ui/newgame.ani' -- 171 bytes
Thread 'fade transform' done (returned 0).
[Thread 0xb6f8bb90 (LWP 2555) exited]
Trying to get undefined resource 'music.mainmenu'
[New Thread 0xb2fcdb90 (LWP 2556)]
[New Thread 0xb6f8bb90 (LWP 2557)]
Thread 'fade transform' done (returned 0).
[Thread 0xb6f8bb90 (LWP 2557) exited]
[New Thread 0xb6f8bb90 (LWP 2558)]
Thread 'flash rectangle' done (returned 0).
[Thread 0xb2fcdb90 (LWP 2556) exited]
Thread 'Starcon2Main' done (returned 0).

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb79cfb90 (LWP 2539)]
0x00358e6d in ?? ()
Missing separate debuginfos, use: debuginfo-install SDL-1.2.13-7.fc10.i386 SDL_image-1.2.6-6.fc9.i386
 alsa-lib-1.0.20-1.fc10.i386 e2fsprogs-libs-1.41.3-2.fc10.i386 gdbm-1.8.0-29.fc10.i386 glibc-2.9-3.i686 libICE-1.0.4-4.fc10.i386 libSM-1.1.0-2.fc10.i386 libX11-1.1.5-4.fc10.i386 libXau-1.0.4-1.fc10.i386
 libXcursor-1.1.9-3.fc10.i386 libXdamage-1.1.1-4.fc9.i386 libXdmcp-1.0.2-6.fc10.i386 libXext-1.0.4-1.fc9.i386
libXfixes-4.0.3-4.fc10.i386 libXrandr-1.2.3-1.fc10.i386 libXrender-0.9.4-3.fc9.i386 libXxf86vm-1.0.2-1.fc10.i386
libasyncns-0.7-1.fc10.i386 libcap-2.10-2.fc10.i386 libdrm-2.4.0-0.21.fc10.i386 libgcc-4.3.2-7.i386 libjpeg-6b-43.fc10.i386
 libmikmod-3.2.0-3.beta2.fc9.i386 libogg-1.1.3-9.fc9.i386 libpng-1.2.35-1.fc10.i386 libtiff-3.8.2-11.fc10.i386 libvorbis-1.2.0-5.fc10.i386 libxcb-1.1.91-6.fc10.i386
mesa-libGL-7.2-0.15.fc10.i386
 pulseaudio-libs-0.9.13-6.fc10.i386 zlib-1.2.3-18.fc9.i386
(gdb) thread apply bt
(gdb) bt
#0  0x00358e6d in ?? ()
#1  0x00b7a5ca in start_thread () from /lib/libpthread.so.0
#2  0x0020804e in clone () from /lib/libc.so.6


Title: Re: UQM on Eee PC
Post by: v.v.b. on June 30, 2009, 07:34:00 pm
Are there any news about segfault?

I have one small question.
There is a little problem with slylandro. The voice is missing.
It seems that voice files must be placed in 'probe' folder, but they are located in slylandro folder now.
and slylandro 'home' files must be in 'slylandro' folder, but now they are located in slyhome folder.
If I copy needed files to slylandro folder, I get sync problem with voice and subtitles...
---
        'addons/3dovoice/slylandro/' -- voice clip directory
        'addons/3dovoice/slylandro/slylandro.ts' -- timestamps
Invalid timestamp data for 'HELLO_1'.  Disabling timestamps
SpliceTrack(): loading addons/3dovoice/slylandro/slyho000.ogg
SpliceTrack(): no voice ogg available so setting speech volume to zero
---
AND
---
_GetMusicData(): loading base/comm/probe/probe.mod
    decoder: MikMod, rate 44100 format 17
        'base/comm/probe/probe.txt' -- conversation phrases -- 4582 bytes
        No associated voice clips
        No associated timestamp file
SpliceTrack(): loading slyla004.ogg
SpliceTrack(): no voice ogg available so setting speech volume to zero
---



 And there are same problems with spathi SafeOnes, and second dialog with Shofixti (not first, but second one).
I's quite strange...


Title: Re: UQM on Eee PC
Post by: fossil on July 01, 2009, 09:54:04 pm
Are there any news about segfault?

It's not enough information to figure out what's wrong. However, tonight I will commit some exit cleanups which *might* solve the segfault. Please also let us know exactly how you exit the game, e.g. by selecting Quit from main menu, or by closing the window, or by hitting F10. Do you get the segfault with all three methods?

I have one small question.
There is a little problem with slylandro. The voice is missing.

You caught the SVN in the middle of some content violence. The voice addon has been updated since then. Please update to the latest SVN (do 'svn update' from sc2 dir).


Title: Re: UQM on Eee PC
Post by: v.v.b. on July 08, 2009, 07:11:28 pm
Quote
Do you get the segfault with all three methods?

 Yes. I get segfault with all three methods on svn 3167.


Title: Re: UQM on Eee PC
Post by: fossil on July 09, 2009, 12:55:34 am
It looks like you are using pthread for threading library. Please try using SDL (./build.sh uqm config, item 10) and see if that works out.
Make a debugging build, if you have not done so already. If it segfaults, please post a backtrace of all threads. By the way, the command to do so was misquoted here. It's actually 'thread apply all bt' ('all' was missing).



Title: Re: UQM on Eee PC
Post by: Cedric6014 on July 11, 2009, 02:20:03 am

This feature already exists in the current SVN version. ;D


I dot think I've ever seen Novus use a smiley face before. You in a particularly chipper mood this week Novus?


Title: Re: UQM on Eee PC
Post by: Novus on July 11, 2009, 11:18:10 am

This feature already exists in the current SVN version. ;D


I dot think I've ever seen Novus use a smiley face before. You in a particularly chipper mood this week Novus?

You (http://forum.uqm.stack.nl/index.php?topic=161.msg59768#msg59768) haven't (http://forum.uqm.stack.nl/index.php?topic=4347.msg56206#msg56206) been (http://forum.uqm.stack.nl/index.php?topic=4250.msg55384#msg55384) paying (http://forum.uqm.stack.nl/index.php?topic=4210.msg54970#msg54970) attention (http://forum.uqm.stack.nl/index.php?topic=4188.msg54723#msg54723). ;) Still, the summer sun does brighten up your day, doesn't it? :D


Title: Re: UQM on Eee PC
Post by: v.v.b. on July 26, 2009, 07:56:11 am
It looks like you are using pthread for threading library. Please try using SDL (./build.sh uqm config, item 10) and see if that works out.

 I've always use SDL AFAIR...

 Output of 'thread apply all bt':
Code:
...................... it's the end of 'main' log:....................................
'lbm/boosml.ani' -- 194 bytes
[Thread 0x7fffee138950 (LWP 4397) exited]
[New Thread 0x7fffee138950 (LWP 4398)]
[Thread 0x7fffee138950 (LWP 4398) exited]
[New Thread 0x7ffff2320950 (LWP 4399)]
Warning: Can't open 'lbm/mainmenu.ogg'
'lbm/newgame.ani' -- 177 bytes
Warning: Can't open 'lbm/mainmenu.ogg'
'lbm/newgame.ani' -- 177 bytes
[Thread 0x7ffff2320950 (LWP 4399) exited]
[New Thread 0x7ffff2320950 (LWP 4400)]
[New Thread 0x7fffee138950 (LWP 4401)]
[Thread 0x7fffee138950 (LWP 4401) exited]
[New Thread 0x7fffee138950 (LWP 4402)]
[Thread 0x7ffff2320950 (LWP 4400) exited]
Thread 'Unknown (probably renderer)' blocking on 'DCQ'
Thread 'Unknown (probably renderer)' blocking on 'DCQ'

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff2f5d950 (LWP 4390)]
0x00000037faa2e1d7 in ?? ()
(gdb) thread apply all bt

Thread 14 (Thread 0x7fffee138950 (LWP 4402)):
#0  0x00000037f3e0e851 in nanosleep () from /lib64/libpthread.so.0
#1  0x00000037ff05a4d4 in SDL_Delay () from /usr/lib64/libSDL-1.2.so.0
#2  0x000000000046a88f in SDL_EnableUNICODE ()
#3  0x00000000004b08db in SDL_EnableUNICODE ()
#4  0x00000037ff011017 in ?? () from /usr/lib64/libSDL-1.2.so.0
#5  0x00000037ff056fc9 in ?? () from /usr/lib64/libSDL-1.2.so.0
#6  0x00000037f3e073da in start_thread () from /lib64/libpthread.so.0
#7  0x00000037f32e62bd in clone () from /lib64/libc.so.6

Thread 7 (Thread 0x7fffecd36950 (LWP 4395)):
#0  0x00000037f3e0e851 in nanosleep () from /lib64/libpthread.so.0
#1  0x00000037ff05a4d4 in SDL_Delay () from /usr/lib64/libSDL-1.2.so.0
#2  0x00000000004a630c in SDL_EnableUNICODE ()
#3  0x00000000004b08db in SDL_EnableUNICODE ()
#4  0x00000037ff011017 in ?? () from /usr/lib64/libSDL-1.2.so.0
#5  0x00000037ff056fc9 in ?? () from /usr/lib64/libSDL-1.2.so.0
#6  0x00000037f3e073da in start_thread () from /lib64/libpthread.so.0
#7  0x00000037f32e62bd in clone () from /lib64/libc.so.6

Thread 6 (Thread 0x7fffed737950 (LWP 4394)):
#0  0x00000037f32dc886 in poll () from /lib64/libc.so.6
#1  0x0000003803852f96 in ?? () from /lib64/libasound.so.2
#2  0x0000003803853357 in ?? () from /lib64/libasound.so.2
---Type <return> to continue, or q <return> to quit---
#3  0x000000380385ea6a in snd_pcm_mmap_writei () from /lib64/libasound.so.2
#4  0x00000037ff036afd in ?? () from /usr/lib64/libSDL-1.2.so.0
#5  0x00000037ff0096de in ?? () from /usr/lib64/libSDL-1.2.so.0
#6  0x00000037ff011017 in ?? () from /usr/lib64/libSDL-1.2.so.0
#7  0x00000037ff056fc9 in ?? () from /usr/lib64/libSDL-1.2.so.0
#8  0x00000037f3e073da in start_thread () from /lib64/libpthread.so.0
#9  0x00000037f32e62bd in clone () from /lib64/libc.so.6

Thread 2 (Thread 0x7ffff2f5d950 (LWP 4390)):
#0  0x00000037faa2e1d7 in ?? ()
#1  0x00000037f3e07479 in start_thread () from /lib64/libpthread.so.0
#2  0x00000037f32e62bd in clone () from /lib64/libc.so.6

Thread 1 (Thread 0x7ffff7fce730 (LWP 4387)):
#0  0x00000037f3e0db14 in __lll_lock_wait () from /lib64/libpthread.so.0
#1  0x00000037f3e091cb in _L_lock_312 () from /lib64/libpthread.so.0
#2  0x00000037f3e08bd1 in pthread_mutex_lock () from /lib64/libpthread.so.0
#3  0x00000037ff05729e in SDL_mutexP () from /usr/lib64/libSDL-1.2.so.0
#4  0x00000000004b074b in SDL_EnableUNICODE ()
#5  0x00000000004a243f in SDL_EnableUNICODE ()
#6  0x00000000004a2721 in SDL_EnableUNICODE ()
#7  0x0000000000475872 in SDL_EnableUNICODE ()
#8  0x00000000004070e5 in SDL_EnableUNICODE ()
#9  0x00000037f321e576 in __libc_start_main () from /lib64/libc.so.6
#10 0x00000000004055f9 in SDL_EnableUNICODE ()
---Type <return> to continue, or q <return> to quit---
#11 0x00007fffffffe658 in ?? ()
#12 0x000000000000001c in ?? ()
#13 0x0000000000000001 in ?? ()
#14 0x00007fffffffe8a0 in ?? ()
#15 0x0000000000000000 in ?? ()
#0  0x00000037faa2e1d7 in ?? ()


 Hope it helps...


Title: Re: UQM on Eee PC
Post by: fossil on July 26, 2009, 06:57:47 pm
I don't know why the code ends up at a location that does not belong to any .so (at least according to gdb). Let's look at the process maps. Please make two map dumps: one while the game is running and one when it segfaults, and post them here. Assuming you are using a debug build and the executable is uqm-debug, dump the maps like so:
Code:
cat /proc/`pidof uqm-debug`/maps
(probably better to redirect the output to some file)
You will need to do the dumps in a separate terminal window. Of course, the game has to be under gdb when it segfaults, or you'll be dumping a map of a process that does not exist anymore  ;)


Title: Re: UQM on Eee PC
Post by: v.v.b. on July 26, 2009, 08:14:28 pm
Assuming you are using a debug build and the executable is uqm-debug

 Oh! Terribly sorry! I've start gdb with simply uqm! Many apologies!
 But now when I start gdb uqm-debug and 'start' after it, I get:
Code:
(gdb) start
Breakpoint 1 at 0x40663c: file src/starcon2.c, line 148.
Starting program: /home/vvb/rpmbuild/SOURCES/sc2/sc2/uqm-debug
[Thread debugging using libthread_db enabled]

[1]+  Stopped                 gdb uqm-debug

 and after 'run' I get:
Code:
(gdb) run
Starting program: /home/vvb/rpmbuild/SOURCES/sc2/sc2/uqm-debug
[Thread debugging using libthread_db enabled]

[2]+  Stopped                 gdb uqm-debug


Title: Re: UQM on Eee PC
Post by: fossil on July 27, 2009, 12:58:33 am
Don't do start, just do run from the get go.


Title: Re: UQM on Eee PC
Post by: v.v.b. on July 27, 2009, 04:43:22 pm
As I said, I can't run it.
After 'run' command gdb says:
Code:
(gdb) run
Starting program: /home/vvb/rpmbuild/SOURCES/sc2/sc2/uqm-debug
[Thread debugging using libthread_db enabled]

[1]+  Stopped                 gdb uqm-debug
And quit to console...


Title: Re: UQM on Eee PC
Post by: fossil on July 27, 2009, 04:55:46 pm
I don't know why this happens.  :-\  Perhaps someone else can tell why. You may need to rebuild the uqm-debug executable, especially if you recently upgraded the OS.


Title: Re: UQM on Eee PC
Post by: meep-eep on July 27, 2009, 05:45:16 pm
That sounds like a problem with gdb itself. I suspect that the suspend signal which is sent to the child somehow ends up at the gdb process itself. Are you running a recent gdb and kernel?
What happens if you continue gdb by typing 'fg' on your shell prompt?
Something else you could try is to run uqm, and then attach gdb to the running process:
Code:
gdb uqm-debug `pidof uqm-debug`