Crash when entering hyperspace

For discussion of ports to POSIX based systems, especially using GNUStep.

Moderators: winston, another_commander, Getafix

User avatar
Getafix
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 979
Joined: Tue Apr 01, 2008 12:55 pm
Location: A small ice asteroid, orbiting Oresrati in Galaxy 8 (a.k.a. northwest Armorica).
Contact:

Crash when entering hyperspace

Post by Getafix »

@Bugbear

Please, increase the "FBTexPercent" value in the /etc/X11/xorg.conf file to "80":
Code:
Section "Device"
Identifier "Configured Video Device"
Option "FBTexPercent" "80" # add this line as shown here
EndSection
Restart the X-server, as described in my previous post, for
the changes to take effect. To check if the changes are applied,
look into the /var/log/Xorg.0.log for "FBTexPercent" value "80".

Don't forget to post the terminal output and
e-mail me the Xorg.0.log following a crash.



EDIT 06/12/2010 13:36
Please, also post the result of the following command:

Code: Select all

$ ls -l /usr/lib/Oolite/oolite-deps/lib/
or, in case this does not exist post the result of the following command:

Code: Select all

$ ls -l ~/.local/lib/Oolite/oolite-deps/lib/
EDIT 06/12/2010 14:23
a. In case you have "Compiz" installed, please, deactivate it and check if that eliminates the crashes. :!:

b. If /etc/X11/xorg.conf has the following line:

Code: Select all

Section "ServerLayout"
...
  Option      "AIGLX" "true"
...
EndSection
revert it to

Code: Select all

Section "ServerLayout"
...
  Option      "AIGLX" "false"
...
EndSection
and restart the X-server (as described in my previous post).
To check if the AIGLX changes are in effect, check the /var/log/Xorg.0.log file.
There shouldn't be any line like "(==) AIGLX enabled".


In any case, please:
1. keep any OXP's inactive for the moment (rename AddOns folder to AddOns_)
2. post the contents of the Oolite/oolite-deps/lib/ folder, as described in my first "Edit"
3. post the Latest.log file and
4. e-mail me the latest /var/log/Xorg.0.log file and the /etc/X11/xorg.conf file, following a crash.
"Any sufficiently advanced information is indistinguishable from noise." [Newman, Lachmann, Moore]
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

Code: Select all

ls -l /usr/lib/Oolite/oolite-deps/lib/
total 5572
-rwxr-xr-x 1 root root  265726 2010-10-04 21:47 libespeak.so.1
-rwxr-xr-x 1 root root  219112 2010-10-04 21:47 libespeak.so.1.pulseaudio
-rw-r--r-- 1 root root   22308 2010-10-04 21:47 libffi.so.4
-rwxr-xr-x 1 root root 3955224 2010-10-04 21:47 libgnustep-base.so.1.18
-rw-r--r-- 1 root root   98944 2010-10-04 21:47 libobjc.so.2
-rwxr-xr-x 1 root root  139080 2010-10-04 21:47 libpng14.so.14
-rwxr-xr-x 1 root root  360068 2010-10-04 21:47 libportaudio.so.2
-rwxr-xr-x 1 root root  431392 2010-10-04 21:47 libSDL-1.2.so.0
-rwxr-xr-x 1 root root  165756 2010-10-04 21:47 libSDL_mixer-1.2.so.0
I checked xorg.conf for the ServerLayout section but that section does not exist.
Looking at the Xorg.0.log file, however, I did find an entry with:

Code: Select all

(II) Loading extension DOUBLE-BUFFER
(II) LoadModule: "glx"
(II) Loading /usr/lib/xorg/modules/extensions//libglx.so
(II) Module glx: vendor="X.Org Foundation"
        compiled for 1.4.2, module version = 1.0.0
        ABI class: X.Org Server Extension, version 0.3
(==) AIGLX enabled
(II) Loading extension GLX
(II) LoadModule: "freetype"
(II) Loading /usr/lib/xorg/modules//fonts/libfreetype.so
I've had one crash since changing FBTexPercent to 65 but I didn't capture Latest.log. I'll keep "testing" and get you the file when I can.
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

Waddaya know. Three jumps later and we hit paydirt.

I've sent you the files via email.
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

Another three jumps, another crash

Code: Select all

/usr/bin/oolite
2010-12-09 23:31:17.808 oolite[6763] Unknown time zone name `WST'.
2010-12-09 23:31:17.808 oolite[6763] Using time zone with absolute offset 0.
2010-12-09 23:31:17.792 oolite[6763] File NSUserDefaults.m: 626. In [NSUserDefaults +standardUserDefaults] Improper installation: No language locale found
ALSA lib pcm.c:7050:(snd_pcm_recover) underrun occured
oolite: tnl/t_draw.c:203: bind_inputs: Assertion `inputs[i]->BufferObj->Pointer' failed.
/usr/lib/Oolite/oolite.app/oolite-wrapper: line 60:  6763 Aborted                 ./oolite.app/oolite $@

Erk. It looks like Oolite died with an error. When making an error
report, please copy + paste the log above into the report.

(Press Ctrl-C to continue)
Not saying there's a pattern yet as I've had longer runs of consecutive jumps without crashing.
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

Possibly related...every now and then Oolite hangs when I enter the navigation screens (f6).

Looking at the terminal output, I now see the following:

Code: Select all

2010-12-09 23:43:53.233 oolite[7232] Unknown time zone name `WST'.
2010-12-09 23:43:53.233 oolite[7232] Using time zone with absolute offset 0.
2010-12-09 23:43:53.217 oolite[7232] File NSUserDefaults.m: 626. In [NSUserDefaults +standardUserDefaults] Improper installation: No language locale found
ALSA lib pcm.c:7050:(snd_pcm_recover) underrun occured
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
 /* snip - this goes on for a minute or so, probably a couple of hundred lines like this */
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
[driAllocateTexture:636] unable to allocate texture
oolite: tnl/t_draw.c:203: bind_inputs: Assertion `inputs[i]->BufferObj->Pointer' failed.
/usr/lib/Oolite/oolite.app/oolite-wrapper: line 60:  7232 Aborted                 ./oolite.app/oolite $@

Erk. It looks like Oolite died with an error. When making an error
report, please copy + paste the log above into the report.

(Press Ctrl-C to continue)
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
another_commander
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 6776
Joined: Wed Feb 28, 2007 7:54 am

Post by another_commander »

I am by no means a Linux expert, but the file tnl/t_draw.c is definitely not a part of Oolite's code. Something else in the system must be messing things up.
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

A quick Google search reveals that tnl/t_draw.c is part of the Mesa3D graphics package.

And other people have been reporting the same symptoms as what I've been seeing.

I'll try updating my Mesa3D package and see if that makes any difference.

Everyone, thanks for your help! I'll update when I have further news.
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
User avatar
Getafix
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 979
Joined: Tue Apr 01, 2008 12:55 pm
Location: A small ice asteroid, orbiting Oresrati in Galaxy 8 (a.k.a. northwest Armorica).
Contact:

Crash when entering hyperspace

Post by Getafix »

@Bugbear
A. Please execute the following and post the result:

Code: Select all

$ dmesg | grep agpgart
B. Also execetute the following:

Code: Select all

$ lspci -vvv
You should look for two things here:
B1. a line like nn:nn.n VGA compatible controller: ...
Please, post the whole paragraph output related to this line for example:

Code: Select all

01:00.0 VGA compatible controller: ATI Technologies Inc M6P
Subsystem: Dell Unknown device 2006
Flags: bus master, fast devsel, latency 0, IRQ 16
Memory at d0000000 (32-bit, prefetchable) [size=16K]
I/O ports at de00 [size=256]
Memory at dfdf0000 (32-bit, non-prefetchable) [size=64K]
B2. something like agp host bridge (the whole paragraph again):

Code: Select all

Memory at 10000000 (32-bit, non-prefetchable) [size=8M]
Capabilities: [c0] AGP version 2.0
"Any sufficiently advanced information is indistinguishable from noise." [Newman, Lachmann, Moore]
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

Tried reinstalling Mesa but it hasn't made any difference.

Getafix, here's the output from the commands you've requested:

Code: Select all

dmesg | grep agpgart
--------------------
[   14.219296] Linux agpgart interface v0.103
[   14.307420] agpgart: Detected an Intel 830M Chipset.
[   14.323215] agpgart: AGP aperture is 256M @ 0xd0000000
[   50.988000] agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0.
[   50.988000] agpgart: Putting AGP V2 device at 0000:00:00.0 into 4x mode
[   50.988000] agpgart: Putting AGP V2 device at 0000:01:00.0 into 4x mode
[ 2666.527766] agpgart: Found an AGP 2.0 compliant device at 0000:00:00.0.
[ 2666.527802] agpgart: Putting AGP V2 device at 0000:00:00.0 into 4x mode
[ 2666.527832] agpgart: Putting AGP V2 device at 0000:01:00.0 into 4x mode

lspci -vvv
----------
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility M6 LY (prog-if 00 [VGA controller])
	Subsystem: Dell Device 00e3
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop+ ParErr- Stepping+ SERR+ FastB2B- DisINTx-
	Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 32 (2000ns min), Cache Line Size: 32 bytes
	Interrupt: pin A routed to IRQ 11
	Region 0: Memory at e0000000 (32-bit, prefetchable) [size=128M]
	Region 1: I/O ports at c000 [size=256]
	Region 2: Memory at fcff0000 (32-bit, non-prefetchable) [size=64K]
	[virtual] Expansion ROM at fc000000 [disabled] [size=128K]
	Capabilities: <access denied>
	Kernel modules: radeonfb
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
User avatar
Getafix
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 979
Joined: Tue Apr 01, 2008 12:55 pm
Location: A small ice asteroid, orbiting Oresrati in Galaxy 8 (a.k.a. northwest Armorica).
Contact:

Crash when entering hyperspace

Post by Getafix »

Bugbear. thank you for your feedback.

Before proceeding, make sure you still have the /etc/X11/xorg.conf.beforeoolite backup file.

Please, update your /etc/X11/xorg.conf from

Code: Select all

Section "Device"
	Identifier "Configured Video Device"
	Option "FBTexPercent" "65"	# add this line as shown here
EndSection 
to

Code: Select all

Section "Device"
	Identifier "Configured Video Device"
	Option "FBTexPercent" "80"	# increased from 65 to 80
	Option "GARTSize" "128"		# add this line as shown here
EndSection 
You'll have to restart the X server (i.e. Ctrl+Alt+Backspace) for the changes to take effect.

Before giving up :oops: and if the crashes persist, e-mail me one last time
the /var/log/Xorg.0.log
the /etc/X11/xorg.conf and
post here the terminal output.


If at any point you are prompted that the X server cannot start,
you logon in text mode and you execute

Code: Select all

$ su
# cp -p /etc/X11/xorg.conf.beforeoolite /etc/X11/xorg.conf
# /etc/init.d/gdm start
"Any sufficiently advanced information is indistinguishable from noise." [Newman, Lachmann, Moore]
User avatar
Bugbear
---- E L I T E ----
---- E L I T E ----
Posts: 415
Joined: Sun Sep 17, 2006 1:30 am

Post by Bugbear »

Another day, another crash.

I made the changes you recommended, unfortunately I had another crash.

Code: Select all

2010-12-21 11:35:17.985 oolite[4917] Unknown time zone name `WST'.
2010-12-21 11:35:17.986 oolite[4917] Using time zone with absolute offset 0.
2010-12-21 11:35:17.939 oolite[4917] File NSUserDefaults.m: 626. In [NSUserDefaults +standardUserDefaults] Improper installation: No language locale found
ALSA lib pcm.c:7050:(snd_pcm_recover) underrun occured
oolite: tnl/t_draw.c:203: bind_inputs: Assertion `inputs[i]->BufferObj->Pointer' failed.
/usr/lib/Oolite/oolite.app/oolite-wrapper: line 60:  4917 Aborted                 ./oolite.app/oolite $@

Erk. It looks like Oolite died with an error. When making an error
report, please copy + paste the log above into the report.

(Press Ctrl-C to continue)
^C
I've emailed to you the various log and config files.


Out of interest, what was the intended effect of the changes you suggested?

At any rate, I noticed a new Debian release, and this one's been going strong since Sept 2009 which I suppose ain't too bad so maybe it's time for a rebuild (something to do over the Christmas break...or perhaps Ill investigate the Mesa3D lead a bit further...)

Either way, thanks so much for your interest in getting this behaviour sorted, even if we haven't got a happy ending yet!
Commander Bugbear
Cruising chart 5 in a Boa Class Criuser: Quantum Pelican I
Vigilante, trader, gems and precious metals hoarder.
Black Monks bothering performed at no extra charge.
User avatar
Getafix
Quite Grand Sub-Admiral
Quite Grand Sub-Admiral
Posts: 979
Joined: Tue Apr 01, 2008 12:55 pm
Location: A small ice asteroid, orbiting Oresrati in Galaxy 8 (a.k.a. northwest Armorica).
Contact:

Crash when entering hyperspace

Post by Getafix »

Well, my purpose was to force Linux to use more memory for OpenGL textures,
than the memory dictated by the default configuration you were using.

The FBTexPercent tries to address the "unable to allocate texture" error message.
The GARTSize tries to address possible AGP memory mapping issues.
Anyhow, we did not manage to splash the
oolite: tnl/t_draw.c:203: bind_inputs: Assertion `inputs->BufferObj->Pointer' failed.
/usr/lib/Oolite/oolite.app/oolite-wrapper: line 60: 3143 Aborted ./oolite.app/oolite $@


I would not bother for

Code: Select all

2010-12-21 11:35:17.985 oolite[4917] Unknown time zone name `WST'.
2010-12-21 11:35:17.986 oolite[4917] Using time zone with absolute offset 0.
2010-12-21 11:35:17.939 oolite[4917] File NSUserDefaults.m: 626. In [NSUserDefaults +standardUserDefaults] Improper installation: No language locale found
ALSA lib pcm.c:7050:(snd_pcm_recover) underrun occured 
I also get them in my VBox Debian and oolite still goes well with no crash.

Tough luck! I think I will have to withdraw from that one. :(

I suggest you leave your /etc/X11/xorg.conf as it is now,
if it does not generate any strange behavior.

Your case has something to do with the driver configuration and, therefore,
I can only recommend to address it to your Linux distro community,
for more specific expertise there.

You can always try the new Debian release, of course, and post any updates on that fat bugger. :wink:
"Any sufficiently advanced information is indistinguishable from noise." [Newman, Lachmann, Moore]
Post Reply