Second Life Forums Archive - Lockups again with 1.10.3

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-14-2006 10:50

I just installed the new version and I'm getting the same error as before. The following comes up, and then the client totally locks on me:

2006-06-14T17:51:17Z INFO: GL_VENDOR NVIDIA Corporation
2006-06-14T17:51:17Z GL_RENDERER GeForce4 MX Integrated GPU/AGP/SSE/3DNOW!
2006-06-14T17:51:17Z GL_VERSION 1.5.6 NVIDIA 87.62
2006-06-14T17:51:17Z

I've re-installed the drivers that worked just fine in 1.10.2, and even reconficuted sax2 again. The result? I'm back to the same old problem, except now there seems to be even fewer ways to resolve it. I've even restarted several times.

This is really beginning to frustrate me.

Drake Bacon

Linux is Furry

Join date: 13 Jul 2005

Posts: 443

06-14-2006 12:30

From: Kelseigh Miranda

I just installed the new version and I'm getting the same error as before. The following comes up, and then the client totally locks on me:

2006-06-14T17:51:17Z INFO: GL_VENDOR NVIDIA Corporation
2006-06-14T17:51:17Z GL_RENDERER GeForce4 MX Integrated GPU/AGP/SSE/3DNOW!
2006-06-14T17:51:17Z GL_VERSION 1.5.6 NVIDIA 87.62
2006-06-14T17:51:17Z

I've re-installed the drivers that worked just fine in 1.10.2, and even reconficuted sax2 again. The result? I'm back to the same old problem, except now there seems to be even fewer ways to resolve it. I've even restarted several times.

This is really beginning to frustrate me.

Which Xorg version are you running? Distro?

Marco Spoonhammer

Registered User

Join date: 23 Feb 2006

Posts: 42

FC5 running fine

06-14-2006 12:37

I am running FC5 fine, just got DC/ed but that was due to a sim reboot at their end.

Havent noticed many updates to the client tho. Specifically I think we are all after:

Audio
Cut n Paste
Fullscreen

but correct me if I am wrong. Seems as stable for me as 1.10.2

BTW we now have a preview in the preview section - recognition at last!!

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-14-2006 12:39

I'm running OpenSuSe 10.0. XOrg..uh...not sure. How do I check?

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-14-2006 12:57

Mmm...looks like XOrg is 6.8.2-100.2

I'm running an integrated GeForce4.

Drake Bacon

Linux is Furry

Join date: 13 Jul 2005

Posts: 443

06-14-2006 13:21

Make sure you've pulled the latest NVidia drivers. You may have to compile new modules.

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-14-2006 13:29

*sigh*

I *DID* check, as I said. The version I've got is the current version. I did the re-install thing from the command prompt. But it still gives me the exact same problem it did when 1.10.2 came out. Updating the driver worked fine then, but there's nothing to install it *to* now. I'm as up to date as I can get.

Which, I assume, leaves me SOL.

Bones Detritus

Registered User

Join date: 14 Dec 2005

Posts: 23

06-14-2006 20:31

Kelseigh,

I'm getting the same freeze up. It's happening every other time I open the client. It's got me mystified.

-Bones

Darkside Eldrich

Registered User

Join date: 10 Feb 2006

Posts: 200

06-14-2006 21:17

Are we talking just the SL client freezing, or your entire machine?

If it's the entire machine, are any other opengl games freezing the same way? That would indicate driver or hardware issues, not a problem with the SL client.

Just food for thought.

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-15-2006 02:44

No, it's only the SL client. I tried running GL-117 (Open GL flight sim) and it runs just fine. So it's definitely a SL issue.

Bones Detritus

Registered User

Join date: 14 Dec 2005

Posts: 23

06-15-2006 04:33

Ditto. The SL client freezes and needs to be killed. No Ctrl+C from the terminal.

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-15-2006 06:21

Yes, I forgot to mention that. When I run it from terminal, ^C and ^X don't work, only closing the terminal window (and thus the SL client) will work.

If I run it from the taskbar, I need to go into the system monitor to shut it down.

Bones Detritus

Registered User

Join date: 14 Dec 2005

Posts: 23

06-15-2006 21:59

I still have this client freeze in 1.10.3.4 Still happens only on every other run of secondlife.

Edit: Every time I've moved to a newer version client, I've untarred the archive into a new folder instead of overwriting old clients. There's no issue with incompatabilities caused by that.

-Bones

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-16-2006 04:57

1.10.3-4 gives me no improvement. Same lockup as before, although unlike Bones it's EVERY time I try to run the client. I restarted, I re-re-re-installed the most up to date drivers, I reconfigured sax2 as recommended on the nVidia page, I did everything but sacrifice a goat, just to be sure. And still, SL is the *only* GL application that won't run. I get a bit over 1700 fps from GLXgears, which is my usual test.

If anyone has any idea of what's going on here, and why SL won't play nice, I'd really appreciate it. I'm a paid member, and it bothers me that I'm currently paying for, well, nothing.

Bones Detritus

Registered User

Join date: 14 Dec 2005

Posts: 23

06-16-2006 11:45

My client seems to lock in the same place as Kelseigh's.

CODE

2006-06-16T18:35:04Z INFO: GL_VENDOR      NVIDIA Corporation
2006-06-16T18:35:04Z GL_RENDERER    GeForce 6600/AGP/SSE/3DNOW!
2006-06-16T18:35:04Z GL_VERSION     2.0.2 NVIDIA 87.62
2006-06-16T18:35:04Z

I was thinking that it was my imagination that the client locks exactly every other time, but the pattern has held steady now.

More instructive of the source of the problem might be to see what the client isn't getting to when it crashes. Looking at my logs of known good client runs, the lines that are never reached when my client crashes are these:

CODE

2006-06-16T18:35:05Z INFO: Viewer Digest: abcdefg1-2hij-k3l4-mnop-5678901q2r34
2006-06-16T18:35:05Z INFO: Couldn't open pilot.txt, aborting agentpilot load!
2006-06-16T18:35:05Z INFO: Resend: 75000
2006-06-16T18:35:05Z INFO: Land: 85000
2006-06-16T18:35:05Z INFO: Wind: 17000
2006-06-16T18:35:05Z INFO: Cloud: 17000
2006-06-16T18:35:05Z INFO: Task: 223000
2006-06-16T18:35:05Z INFO: Texture: 223000
2006-06-16T18:35:05Z INFO: Asset: 110000
2006-06-16T18:35:05Z INFO: Total: 750000

and so on and so forth....

Note: Digest code changed because, well, I'm not sure if it's a client specific identifier or what precise function it serves. It's the Viewer Digest line that is never reached when the client crashes though.

Validus Bishop

Registered User

Join date: 24 Feb 2006

Posts: 81

06-16-2006 12:29

Just FYI, not sure it will change much. My main PC uses a 7800GS and I actually had a whole slew of issues with driver 8762 and the one prior. I am using 8178 and it runs perfectly. Browsing through the nvidia forums, I found quite a few people with issues with the last 2 linux nvidia driver releases.

--Val

Bones Detritus

Registered User

Join date: 14 Dec 2005

Posts: 23

06-16-2006 12:32

Edit: Update taken away. pKilling a good run of SL instead of Quitting out properly does make a difference in the pattern with which I experience the bug, but it doesn't do so reliably. No diagnostic conclusions I can draw from it.

My system, if it matters (I can't see how as all versions of SL linux up till 1.10.3.3 worked wonderfully.)
Ubuntu 6.06 (Same install has been dist-upgraded all the way from 4.10->5.04->5.10->6.06)
Athlon XP processor, running k7 Ubuntu kernel

nvidia-glx from Ubuntu repos. Acceleration via AGPGART module instead of nvidia's own. (It made a noticable improvement in my SL framerate)
SL lists OpenGL version as: 2.0.2 NVIDIA 87.62
Graphics card: Nvidia GeForce 6600, 256 MB of RAM on card.

No problems with any of this in previous SL versions.

Theora Aquitaine

Registered User

Join date: 12 Feb 2006

Posts: 266

06-17-2006 10:04

Don't have a fix I'm afraid, although the suggestion to downgrade the drivers sounds worth a try (or using the nvidia agpgart?)

Do you get the same lockup under wine?

Edit: I notice Kelseigh has already had the same lockups with the earlier driver.
Is it possible for you to run SL in gdb and then get a backtrace?
Also, is there anything in syslog or dmesg|tail after the lockup?

Angel Sunset

Linutic

Join date: 7 Apr 2005

Posts: 636

06-17-2006 12:04

The newer SL Versions really hammer AGP, much more than before.

I had to patch the nvidia driver source to only use 4x agp, instead of 8x. The lockup didn't only affect SL, though; KDE would not start at all. So it may not be what you have.

Instructions for this are in appendix F (I think) of the NVidia Readme for this version.

_____________________

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Kubuntu Intrepid 8.10, KDE, linux 2.6.27-11, X.Org 11.0, server glx vendor: NVIDIA Corporation, server glx version: 1.5.2, OpenGL vendor: NVIDIA Corporation, OpenGL renderer: GeForce 9800 GTX+/PCI/SSE2, OpenGL version: 3.0.0 NVIDIA 180.29, glu version: 1.3, NVidia GEForce 9800 GTX+ 512 MB, Intel Core 2 Duo, Mem: 3371368k , Swap: 2570360k

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-17-2006 13:03

From: Theora Aquitaine

Don't have a fix I'm afraid, although the suggestion to downgrade the drivers sounds worth a try (or using the nvidia agpgart?)

Do you get the same lockup under wine?

Edit: I notice Kelseigh has already had the same lockups with the earlier driver.
Is it possible for you to run SL in gdb and then get a backtrace?
Also, is there anything in syslog or dmesg|tail after the lockup?

I'm not very proficient with that stuff. How do I do that with gdb, and where to I find the logs?

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-17-2006 13:05

From: Theora Aquitaine

Don't have a fix I'm afraid, although the suggestion to downgrade the drivers sounds worth a try (or using the nvidia agpgart?)

Do you get the same lockup under wine?

Edit: I notice Kelseigh has already had the same lockups with the earlier driver.
Is it possible for you to run SL in gdb and then get a backtrace?
Also, is there anything in syslog or dmesg|tail after the lockup?

Ah, I figured out one of them. Here's what I get from dmesg|tail:

end_request: I/O error, dev fd0, sector 0
end_request: I/O error, dev fd0, sector 0
end_request: I/O error, dev fd0, sector 0
end_request: I/O error, dev fd0, sector 0
subfs: unsuccessful attempt to mount media (256)
ISO 9660 Extensions: Microsoft Joliet Level 3
ISO 9660 Extensions: RRIP_1991A
kjournald starting. Commit interval 5 seconds
EXT3 FS on hda1, internal journal
EXT3-fs: mounted filesystem with ordered data mode.

Darkside Eldrich

Registered User

Join date: 10 Feb 2006

Posts: 200

06-17-2006 14:20

From: Kelseigh Miranda

I'm not very proficient with that stuff. How do I do that with gdb, and where to I find the logs?

with SL running and frozen, run "ps ax", and note the number at the far left of the column for the sl-bin process. That's the "process ID" for the process. Now, run gdb, and you'll get a gdb prompt. Type "attach <PID>", where <PID> is the process ID you noted earlier. Now try "bt" to print out a backtrace. That shows where the sl process is stuck. Post it here, as requested earlier.

Kelseigh Miranda

Pet Mousey

Join date: 15 Feb 2006

Posts: 33

06-17-2006 14:28

Okay, here's the backtrace:

From: someone

#0 0xffffe410 in ?? ()
#1 0xbf8bfb4c in ?? ()
#2 0x00000002 in ?? ()
#3 0x00000000 in ?? ()
#4 0x4077845e in __lll_mutex_lock_wait () from /lib/tls/libc.so.6
#5 0x40715b24 in _L_mutex_lock_3518 () from /lib/tls/libc.so.6
#6 0x0a87b4e8 in ?? ()
#7 0x00000000 in ?? ()
#8 0x407c8ff4 in ?? () from /lib/tls/libc.so.6
#9 0x00000000 in ?? ()
#10 0xbf8bfcf4 in ?? ()
#11 0xbf8bfb6c in ?? ()
#12 0x40702931 in fclose@@GLIBC_2.1 () from /lib/tls/libc.so.6
#13 0x40702931 in fclose@@GLIBC_2.1 () from /lib/tls/libc.so.6
#14 0x09f7fb58 in close_debug ()
#15 0x09f6f148 in viewer_crash_callback ()
#16 0x09f86362 in signal_handlers ()
#17 <signal handler called>
#18 0x40711937 in malloc_consolidate () from /lib/tls/libc.so.6
#19 0x407ca8b8 in main_arena () from /lib/tls/libc.so.6
#20 0x407ca8c4 in main_arena () from /lib/tls/libc.so.6
#21 0x00000000 in ?? ()
#22 0x407ca8d8 in main_arena () from /lib/tls/libc.so.6
#23 0x000002a0 in ?? ()
#24 0x00003800 in ?? ()
#25 0x407c8ff4 in ?? () from /lib/tls/libc.so.6
#26 0x407ca8a0 in mp_ () from /lib/tls/libc.so.6
#27 0x00000000 in ?? ()
#28 0xbf8c00bc in ?? ()
#29 0x40712534 in _int_malloc () from /lib/tls/libc.so.6
Previous frame inner to this frame (corrupt stack?)

It's all gibberish to me, but hopefully somebody can make sense of it.

Bones Detritus

Registered User

Join date: 14 Dec 2005

Posts: 23

06-17-2006 15:24

Thanks for the help everyone! Gotta love the power of a good community.

Nothing in dmesg|tail. (Verified because the last entry in it was about me unplugging my USB thumb drive, which I'd done 3 minutes before running SL.)

Here's the gdb trace of secondlife-bin:

CODE

#0  0xffffe410 in __kernel_vsyscall ()
#1  0x43fe482e in pthread_setcanceltype () from /lib/tls/i686/cmov/libc.so.6
#2  0x43f617d6 in fclose () from /lib/tls/i686/cmov/libc.so.6
#3  0x09f7fb58 in close_debug ()
#4  0x09f6f148 in viewer_crash_callback ()
#5  0x09f86362 in signal_handlers ()
#6  <signal handler called>
#7  0x43f706b0 in malloc_usable_size () from /lib/tls/i686/cmov/libc.so.6
#8  0x43f71653 in free () from /lib/tls/i686/cmov/libc.so.6
#9  0x43f73411 in malloc () from /lib/tls/i686/cmov/libc.so.6
#10 0xb7a602fe in ?? ()
#11 0x00000470 in ?? ()
#12 0x5f007645 in ?? ()
#13 0x33324e5a in ?? ()
#14 0x61544c4c in ?? ()
#15 0x61436b73 in ?? ()
#16 0x6e696c6c in ?? ()
#17 0x72614367 in ?? ()
#18 0xb7a81e64 in ?? ()
#19 0xb7a700b4 in ?? ()
#20 0x00000470 in ?? ()
#21 0xbf9b0908 in ?? ()
#22 0xb7a4c609 in ?? ()
#23 0x00000470 in ?? ()
#24 0x726f746e in ?? ()
#25 0x364b5279 in ?? ()
#26 0x55554c4c in ?? ()
#27 0x4b524449 in ?? ()
#28 0x5f007353 in ?? ()
#29 0x6d39315a in ?? ()
#30 0x61696465 in ?? ()
#31 0x6c6f765f in ?? ()
#32 0x5f656d75 in ?? ()
#33 0x6e616863 in ?? ()
#34 0xb7a81e64 in ?? ()
#35 0x00000463 in ?? ()
#36 0xbf9b0eb4 in ?? ()
#37 0xbf9b0938 in ?? ()
#38 0xb7a52008 in ?? ()
#39 0x00000470 in ?? ()
#40 0x53656d61 in ?? ()
#41 0x56746174 in ?? ()
#42 0x31776569 in ?? ()
#43 0x6e616831 in ?? ()
#44 0x48656c64 in ?? ()
#45 0x7265766f in ?? ()
#46 0x6a696945 in ?? ()
#47 0xb7a81e64 in ?? ()
#48 0x00000400 in ?? ()
#49 0xbf9b0968 in ?? ()
#50 0xb7a520de in ?? ()
#51 0x00000400 in ?? ()
#52 0xbf9b0990 in ?? ()
#53 0xbf9b0968 in ?? ()
#54 0x08066784 in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string<char*> ()
Previous frame inner to this frame (corrupt stack?)

Theora Aquitaine

Registered User

Join date: 12 Feb 2006

Posts: 266

06-18-2006 09:35

I am afraid I am no expert at this, but I think the above info should help someone to work out what is going on.

The intermittent nature of bone's problem makes it sound more like a hardware problem to me.. Perhaps it could be bad RAM??

A couple of things to try in the meantime:

1) disable AGP:

in your xorg.conf file set Option "NvAGP" "integer" to 0

2) check the output of

/var/log/Xorg.o.log for any errors

3) uninstall the manually installed nvidia drivers with the --uninstall option, purge the packaged drivers (aptitude purge nvidia-glx), try reinstalling either the packaged drivers or drivers from the nvidia site...

4) upgrade the kernel

5) run memtest86+

Lockups again with 1.10.3
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-14-2006 10:50 I just installed the new version and I'm getting the same error as before. The following comes up, and then the client totally locks on me: 2006-06-14T17:51:17Z INFO: GL_VENDOR NVIDIA Corporation 2006-06-14T17:51:17Z GL_RENDERER GeForce4 MX Integrated GPU/AGP/SSE/3DNOW! 2006-06-14T17:51:17Z GL_VERSION 1.5.6 NVIDIA 87.62 2006-06-14T17:51:17Z I've re-installed the drivers that worked just fine in 1.10.2, and even reconficuted sax2 again. The result? I'm back to the same old problem, except now there seems to be even fewer ways to resolve it. I've even restarted several times. This is really beginning to frustrate me.
Drake Bacon Linux is Furry Join date: 13 Jul 2005 Posts: 443	06-14-2006 12:30 From: Kelseigh Miranda I just installed the new version and I'm getting the same error as before. The following comes up, and then the client totally locks on me: 2006-06-14T17:51:17Z INFO: GL_VENDOR NVIDIA Corporation 2006-06-14T17:51:17Z GL_RENDERER GeForce4 MX Integrated GPU/AGP/SSE/3DNOW! 2006-06-14T17:51:17Z GL_VERSION 1.5.6 NVIDIA 87.62 2006-06-14T17:51:17Z I've re-installed the drivers that worked just fine in 1.10.2, and even reconficuted sax2 again. The result? I'm back to the same old problem, except now there seems to be even fewer ways to resolve it. I've even restarted several times. This is really beginning to frustrate me. Which Xorg version are you running? Distro?
Marco Spoonhammer Registered User Join date: 23 Feb 2006 Posts: 42	FC5 running fine 06-14-2006 12:37 I am running FC5 fine, just got DC/ed but that was due to a sim reboot at their end. Havent noticed many updates to the client tho. Specifically I think we are all after: Audio Cut n Paste Fullscreen but correct me if I am wrong. Seems as stable for me as 1.10.2 BTW we now have a preview in the preview section - recognition at last!!
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-14-2006 12:39 I'm running OpenSuSe 10.0. XOrg..uh...not sure. How do I check?
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-14-2006 12:57 Mmm...looks like XOrg is 6.8.2-100.2 I'm running an integrated GeForce4.
Drake Bacon Linux is Furry Join date: 13 Jul 2005 Posts: 443	06-14-2006 13:21 Make sure you've pulled the latest NVidia drivers. You may have to compile new modules.
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-14-2006 13:29 sigh I DID check, as I said. The version I've got is the current version. I did the re-install thing from the command prompt. But it still gives me the exact same problem it did when 1.10.2 came out. Updating the driver worked fine then, but there's nothing to install it to now. I'm as up to date as I can get. Which, I assume, leaves me SOL.
Bones Detritus Registered User Join date: 14 Dec 2005 Posts: 23	06-14-2006 20:31 Kelseigh, I'm getting the same freeze up. It's happening every other time I open the client. It's got me mystified. -Bones
Darkside Eldrich Registered User Join date: 10 Feb 2006 Posts: 200	06-14-2006 21:17 Are we talking just the SL client freezing, or your entire machine? If it's the entire machine, are any other opengl games freezing the same way? That would indicate driver or hardware issues, not a problem with the SL client. Just food for thought.
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-15-2006 02:44 No, it's only the SL client. I tried running GL-117 (Open GL flight sim) and it runs just fine. So it's definitely a SL issue.
Bones Detritus Registered User Join date: 14 Dec 2005 Posts: 23	06-15-2006 04:33 Ditto. The SL client freezes and needs to be killed. No Ctrl+C from the terminal.
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-15-2006 06:21 Yes, I forgot to mention that. When I run it from terminal, ^C and ^X don't work, only closing the terminal window (and thus the SL client) will work. If I run it from the taskbar, I need to go into the system monitor to shut it down.
Bones Detritus Registered User Join date: 14 Dec 2005 Posts: 23	06-15-2006 21:59 I still have this client freeze in 1.10.3.4 Still happens only on every other run of secondlife. Edit: Every time I've moved to a newer version client, I've untarred the archive into a new folder instead of overwriting old clients. There's no issue with incompatabilities caused by that. -Bones
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-16-2006 04:57 1.10.3-4 gives me no improvement. Same lockup as before, although unlike Bones it's EVERY time I try to run the client. I restarted, I re-re-re-installed the most up to date drivers, I reconfigured sax2 as recommended on the nVidia page, I did everything but sacrifice a goat, just to be sure. And still, SL is the only GL application that won't run. I get a bit over 1700 fps from GLXgears, which is my usual test. If anyone has any idea of what's going on here, and why SL won't play nice, I'd really appreciate it. I'm a paid member, and it bothers me that I'm currently paying for, well, nothing.
Bones Detritus Registered User Join date: 14 Dec 2005 Posts: 23	06-16-2006 11:45 My client seems to lock in the same place as Kelseigh's. CODE 2006-06-16T18:35:04Z INFO: GL_VENDOR NVIDIA Corporation 2006-06-16T18:35:04Z GL_RENDERER GeForce 6600/AGP/SSE/3DNOW! 2006-06-16T18:35:04Z GL_VERSION 2.0.2 NVIDIA 87.62 2006-06-16T18:35:04Z I was thinking that it was my imagination that the client locks exactly every other time, but the pattern has held steady now. More instructive of the source of the problem might be to see what the client isn't getting to when it crashes. Looking at my logs of known good client runs, the lines that are never reached when my client crashes are these: CODE 2006-06-16T18:35:05Z INFO: Viewer Digest: abcdefg1-2hij-k3l4-mnop-5678901q2r34 2006-06-16T18:35:05Z INFO: Couldn't open pilot.txt, aborting agentpilot load! 2006-06-16T18:35:05Z INFO: Resend: 75000 2006-06-16T18:35:05Z INFO: Land: 85000 2006-06-16T18:35:05Z INFO: Wind: 17000 2006-06-16T18:35:05Z INFO: Cloud: 17000 2006-06-16T18:35:05Z INFO: Task: 223000 2006-06-16T18:35:05Z INFO: Texture: 223000 2006-06-16T18:35:05Z INFO: Asset: 110000 2006-06-16T18:35:05Z INFO: Total: 750000 and so on and so forth.... Note: Digest code changed because, well, I'm not sure if it's a client specific identifier or what precise function it serves. It's the Viewer Digest line that is never reached when the client crashes though.
Validus Bishop Registered User Join date: 24 Feb 2006 Posts: 81	06-16-2006 12:29 Just FYI, not sure it will change much. My main PC uses a 7800GS and I actually had a whole slew of issues with driver 8762 and the one prior. I am using 8178 and it runs perfectly. Browsing through the nvidia forums, I found quite a few people with issues with the last 2 linux nvidia driver releases. --Val
Bones Detritus Registered User Join date: 14 Dec 2005 Posts: 23	06-16-2006 12:32 Edit: Update taken away. pKilling a good run of SL instead of Quitting out properly does make a difference in the pattern with which I experience the bug, but it doesn't do so reliably. No diagnostic conclusions I can draw from it. My system, if it matters (I can't see how as all versions of SL linux up till 1.10.3.3 worked wonderfully.) Ubuntu 6.06 (Same install has been dist-upgraded all the way from 4.10->5.04->5.10->6.06) Athlon XP processor, running k7 Ubuntu kernel nvidia-glx from Ubuntu repos. Acceleration via AGPGART module instead of nvidia's own. (It made a noticable improvement in my SL framerate) SL lists OpenGL version as: 2.0.2 NVIDIA 87.62 Graphics card: Nvidia GeForce 6600, 256 MB of RAM on card. No problems with any of this in previous SL versions.
Theora Aquitaine Registered User Join date: 12 Feb 2006 Posts: 266	06-17-2006 10:04 Don't have a fix I'm afraid, although the suggestion to downgrade the drivers sounds worth a try (or using the nvidia agpgart?) Do you get the same lockup under wine? Edit: I notice Kelseigh has already had the same lockups with the earlier driver. Is it possible for you to run SL in gdb and then get a backtrace? Also, is there anything in syslog or dmesg\|tail after the lockup?
Angel Sunset Linutic Join date: 7 Apr 2005 Posts: 636	06-17-2006 12:04 The newer SL Versions really hammer AGP, much more than before. I had to patch the nvidia driver source to only use 4x agp, instead of 8x. The lockup didn't only affect SL, though; KDE would not start at all. So it may not be what you have. Instructions for this are in appendix F (I think) of the NVidia Readme for this version. _____________________ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Kubuntu Intrepid 8.10, KDE, linux 2.6.27-11, X.Org 11.0, server glx vendor: NVIDIA Corporation, server glx version: 1.5.2, OpenGL vendor: NVIDIA Corporation, OpenGL renderer: GeForce 9800 GTX+/PCI/SSE2, OpenGL version: 3.0.0 NVIDIA 180.29, glu version: 1.3, NVidia GEForce 9800 GTX+ 512 MB, Intel Core 2 Duo, Mem: 3371368k , Swap: 2570360k
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-17-2006 13:03 From: Theora Aquitaine Don't have a fix I'm afraid, although the suggestion to downgrade the drivers sounds worth a try (or using the nvidia agpgart?) Do you get the same lockup under wine? Edit: I notice Kelseigh has already had the same lockups with the earlier driver. Is it possible for you to run SL in gdb and then get a backtrace? Also, is there anything in syslog or dmesg\|tail after the lockup? I'm not very proficient with that stuff. How do I do that with gdb, and where to I find the logs?
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-17-2006 13:05 From: Theora Aquitaine Don't have a fix I'm afraid, although the suggestion to downgrade the drivers sounds worth a try (or using the nvidia agpgart?) Do you get the same lockup under wine? Edit: I notice Kelseigh has already had the same lockups with the earlier driver. Is it possible for you to run SL in gdb and then get a backtrace? Also, is there anything in syslog or dmesg\|tail after the lockup? Ah, I figured out one of them. Here's what I get from dmesg\|tail: end_request: I/O error, dev fd0, sector 0 end_request: I/O error, dev fd0, sector 0 end_request: I/O error, dev fd0, sector 0 end_request: I/O error, dev fd0, sector 0 subfs: unsuccessful attempt to mount media (256) ISO 9660 Extensions: Microsoft Joliet Level 3 ISO 9660 Extensions: RRIP_1991A kjournald starting. Commit interval 5 seconds EXT3 FS on hda1, internal journal EXT3-fs: mounted filesystem with ordered data mode.
Darkside Eldrich Registered User Join date: 10 Feb 2006 Posts: 200	06-17-2006 14:20 From: Kelseigh Miranda I'm not very proficient with that stuff. How do I do that with gdb, and where to I find the logs? with SL running and frozen, run "ps ax", and note the number at the far left of the column for the sl-bin process. That's the "process ID" for the process. Now, run gdb, and you'll get a gdb prompt. Type "attach <PID>", where <PID> is the process ID you noted earlier. Now try "bt" to print out a backtrace. That shows where the sl process is stuck. Post it here, as requested earlier.
Kelseigh Miranda Pet Mousey Join date: 15 Feb 2006 Posts: 33	06-17-2006 14:28 Okay, here's the backtrace: From: someone #0 0xffffe410 in ?? () #1 0xbf8bfb4c in ?? () #2 0x00000002 in ?? () #3 0x00000000 in ?? () #4 0x4077845e in __lll_mutex_lock_wait () from /lib/tls/libc.so.6 #5 0x40715b24 in _L_mutex_lock_3518 () from /lib/tls/libc.so.6 #6 0x0a87b4e8 in ?? () #7 0x00000000 in ?? () #8 0x407c8ff4 in ?? () from /lib/tls/libc.so.6 #9 0x00000000 in ?? () #10 0xbf8bfcf4 in ?? () #11 0xbf8bfb6c in ?? () #12 0x40702931 in fclose@@GLIBC_2.1 () from /lib/tls/libc.so.6 #13 0x40702931 in fclose@@GLIBC_2.1 () from /lib/tls/libc.so.6 #14 0x09f7fb58 in close_debug () #15 0x09f6f148 in viewer_crash_callback () #16 0x09f86362 in signal_handlers () #17 <signal handler called> #18 0x40711937 in malloc_consolidate () from /lib/tls/libc.so.6 #19 0x407ca8b8 in main_arena () from /lib/tls/libc.so.6 #20 0x407ca8c4 in main_arena () from /lib/tls/libc.so.6 #21 0x00000000 in ?? () #22 0x407ca8d8 in main_arena () from /lib/tls/libc.so.6 #23 0x000002a0 in ?? () #24 0x00003800 in ?? () #25 0x407c8ff4 in ?? () from /lib/tls/libc.so.6 #26 0x407ca8a0 in mp_ () from /lib/tls/libc.so.6 #27 0x00000000 in ?? () #28 0xbf8c00bc in ?? () #29 0x40712534 in _int_malloc () from /lib/tls/libc.so.6 Previous frame inner to this frame (corrupt stack?) It's all gibberish to me, but hopefully somebody can make sense of it.
Bones Detritus Registered User Join date: 14 Dec 2005 Posts: 23	06-17-2006 15:24 Thanks for the help everyone! Gotta love the power of a good community. Nothing in dmesg\|tail. (Verified because the last entry in it was about me unplugging my USB thumb drive, which I'd done 3 minutes before running SL.) Here's the gdb trace of secondlife-bin: CODE #0 0xffffe410 in __kernel_vsyscall () #1 0x43fe482e in pthread_setcanceltype () from /lib/tls/i686/cmov/libc.so.6 #2 0x43f617d6 in fclose () from /lib/tls/i686/cmov/libc.so.6 #3 0x09f7fb58 in close_debug () #4 0x09f6f148 in viewer_crash_callback () #5 0x09f86362 in signal_handlers () #6 <signal handler called> #7 0x43f706b0 in malloc_usable_size () from /lib/tls/i686/cmov/libc.so.6 #8 0x43f71653 in free () from /lib/tls/i686/cmov/libc.so.6 #9 0x43f73411 in malloc () from /lib/tls/i686/cmov/libc.so.6 #10 0xb7a602fe in ?? () #11 0x00000470 in ?? () #12 0x5f007645 in ?? () #13 0x33324e5a in ?? () #14 0x61544c4c in ?? () #15 0x61436b73 in ?? () #16 0x6e696c6c in ?? () #17 0x72614367 in ?? () #18 0xb7a81e64 in ?? () #19 0xb7a700b4 in ?? () #20 0x00000470 in ?? () #21 0xbf9b0908 in ?? () #22 0xb7a4c609 in ?? () #23 0x00000470 in ?? () #24 0x726f746e in ?? () #25 0x364b5279 in ?? () #26 0x55554c4c in ?? () #27 0x4b524449 in ?? () #28 0x5f007353 in ?? () #29 0x6d39315a in ?? () #30 0x61696465 in ?? () #31 0x6c6f765f in ?? () #32 0x5f656d75 in ?? () #33 0x6e616863 in ?? () #34 0xb7a81e64 in ?? () #35 0x00000463 in ?? () #36 0xbf9b0eb4 in ?? () #37 0xbf9b0938 in ?? () #38 0xb7a52008 in ?? () #39 0x00000470 in ?? () #40 0x53656d61 in ?? () #41 0x56746174 in ?? () #42 0x31776569 in ?? () #43 0x6e616831 in ?? () #44 0x48656c64 in ?? () #45 0x7265766f in ?? () #46 0x6a696945 in ?? () #47 0xb7a81e64 in ?? () #48 0x00000400 in ?? () #49 0xbf9b0968 in ?? () #50 0xb7a520de in ?? () #51 0x00000400 in ?? () #52 0xbf9b0990 in ?? () #53 0xbf9b0968 in ?? () #54 0x08066784 in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string<char*> () Previous frame inner to this frame (corrupt stack?)
Theora Aquitaine Registered User Join date: 12 Feb 2006 Posts: 266	06-18-2006 09:35 I am afraid I am no expert at this, but I think the above info should help someone to work out what is going on. The intermittent nature of bone's problem makes it sound more like a hardware problem to me.. Perhaps it could be bad RAM?? A couple of things to try in the meantime: 1) disable AGP: in your xorg.conf file set Option "NvAGP" "integer" to 0 2) check the output of /var/log/Xorg.o.log for any errors 3) uninstall the manually installed nvidia drivers with the --uninstall option, purge the packaged drivers (aptitude purge nvidia-glx), try reinstalling either the packaged drivers or drivers from the nvidia site... 4) upgrade the kernel 5) run memtest86+

Welcome to the Second Life Forums Archive

Lockups again with 1.10.3