Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 11-05-10, 01:49 PM   #1
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

Here's an attempt to relate my own experiences with the latest driver. I've tried to make it as detailed as possible, and hopefully you can stand to read through it all.

I recently put together a machine with an EVGA GTX 465 (s/n 01g-p3-1467-ar), and have not had any success getting this card working in Ubuntu. To install the drivers, I used the Hardware Drivers application in Ubuntu, after adding the ubuntu-x-swat PPA to my software sources. From what I have read, this should properly install the latest drivers.

When X starts, the screen goes black, and will often shift into power-saving mode (lost signal), and then return again to a black screen. Sometimes this cycle repeats several times. Every once in a while, I will get what look like text-mode rectangle/box characters spread on the screen. Sometimes they have color, sometimes they blink. The system becomes completely unresponsive from the keyboard at this point: ctrl+alt+del doesn't work, and neither do the SysRq combos. Curiously, I've been able to ctrl+alt+del out of the older drivers. In any case, I am able to log in via another machine over SSH and reboot that way.

It's hard to say if this is the same sort of issue other folks have been having or not: I've been entirely unable to get the card to load using any number of different combinations of hacks, whereas others have claimed that they had theirs working at least at one point. At least one other 465 user running 195.36.24 has it working (http://www.linuxquestions.org/hcl/sh...p?product=4466). I'm currently running Ubuntu 10.04 (lucid) 64-bit, but I originally installed 10.10 64-bit without any success, either. So far, I've attempted these "workarounds":
- Manually-installed other, older drivers, including 195.36.24, 195.36.31, 256.35, 256.53. These have all been uninstalled prior to trying the latest.
- Using X updates from the ubuntu-x-swat PPA
- Tried both kernel 2.6.32-25 and 2.6.36 from the Kernel PPA
- Uninstalled, blacklisted nouveau; used both nomodeset kernel parameter and "options nouveau modeset=0" in a modprobe.d config.
- Blacklisted vga16fb (grasping at straws here...)
- Probably some other magical voodoo hacks that I simply lost track of.

The card appears to work in Windows 7 Ultimate 64-bit using the 258.96 drivers, although admittedly I haven't stressed the card more than Half-Life 1 for about 30 minutes of play.

Other than the EVGA GTX 465, I am running:
Mainboard: Asus Crosshair IV Formula
RAM: 2 x 2GB G.Skill Ripjaws DDR3-1600 (running 7-7-7-24 at 1.65v)
CPU: AMD Phenom II x4 970 (stock -- no tweaks here).
PSU: Corsair TX750 W
HDDs: 2 x 1TB WD1002FAEX 6.0Gb/s in RAID 1/AES-XTS encrypted/LVM; 1 x 160GB Seagate (forgot the exact model -- it's the windows drive)
Optical drive: Lite-On IHAS424-98

I have also attached the output of nvidia-bug-report after the latest attempt to run X (using the 2.6.36 kernel, boot to single user mode, telinit 3, start ssh server, then startx -- -logverbose 6).

Never had any problems with Nvidia cards in Linux in the past. In fact, they always worked so easily, especially compared with pulling my hair out trying to get the ATI cards running. Now I've been stuck with the VESA driver for a while and naturally it's very disappointing. I don't believe the card to be defective, so is this card simply too new? Hasn't it been supported for at least a few months now? I'm aware that this latest driver release has some regressions, but I'm curious why I can't even get the older drivers working! Is this just a matter of waiting around for the N-th driver revision until the card is supported?

Thanks for reading and any input provided.
Attached Files
File Type: gz nvidia-bug-report.log.gz (85.1 KB, 89 views)

Last edited by Glajvect; 11-06-10 at 12:30 PM. Reason: Fixed copy/paste mischief
Glajvect is offline   Reply With Quote
Old 11-10-10, 06:49 PM   #2
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

A quick update for anyone who might be following this:
I re-tested everything with the 260.19.21 "pre-release". I had to revert back to the 2.6.32-25 kernel in order to build the driver this time (some GCC version incompatibility I had to work around).

The upshot is that I had no luck with this driver, either; but at least I got some more information. I hope it's useful to someone (or if at least someone could tell me what to do with it -- no Kernel debugging experience but I am a programmer).

Attached is the standard bug report. I also noticed these interesting lines:

Quote:
[ 74.608820] NVRM: Xid (0006:00): 31, Ch 00000000, engmask 00000081, intr 10000000
[ 76.880245] NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context
[ 86.992360] NVRM: Xid (0006:00): 31, Ch 00000001, engmask 00000081, intr 10000000
...And something else of interest: the first time I booted up, I missed the recovery mode and Ubuntu went straight into X. Strangely, the PC speaker was beeping at intervals roughly corresponding to the "Xid" messages above, and the screen briefly flashed to a different VT and back to black (though not power-saving black). I invoked "sudo stop gdm" from a SSH session, and got this stack trace in kern.log:

Quote:
Nov 10 19:06:16 zombywoof kernel: [ 267.369988] BUG: soft lockup - CPU#0 stuck for 61s! [Xorg:2577]
Nov 10 19:06:16 zombywoof kernel: [ 267.369993] Modules linked in: binfmt_misc ppdev snd_hda_codec_via ipt_MASQUERADE iptable_nat bridge stp kvm_amd kvm ipt_REJECT ipt_LOG xt_limit xt_tcpudp ipt_addrtype xt_state snd_hda_intel snd_hda_codec snd_hwdep snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_dummy snd_seq_oss ip6table_filter snd_seq_midi ip6_tables nf_nat_irc snd_rawmidi nf_conntrack_irc snd_seq_midi_event nf_nat_ftp nf_nat snd_seq nf_conntrack_ipv4 nf_defrag_ipv4 nf_conntrack_ftp snd_timer snd_seq_device nf_conntrack iptable_filter snd ip_tables x_tables usblp asus_atk0110 edac_core soundcore edac_mce_amd snd_page_alloc nvidia(P) i2c_piix4 xhci k10temp lp parport cryptd aes_x86_64 aes_generic xts gf128mul dm_crypt raid10 raid456 async_pq async_xor xor async_memcpy async_raid6_recov usbhid hid raid6_pq async_tx raid0 multipath linear raid1 usb_storage ohci1394 pata_jmicron sky2 ieee1394 ahci
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] CPU 0:
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] Modules linked in: binfmt_misc ppdev snd_hda_codec_via ipt_MASQUERADE iptable_nat bridge stp kvm_amd kvm ipt_REJECT ipt_LOG xt_limit xt_tcpudp ipt_addrtype xt_state snd_hda_intel snd_hda_codec snd_hwdep snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_dummy snd_seq_oss ip6table_filter snd_seq_midi ip6_tables nf_nat_irc snd_rawmidi nf_conntrack_irc snd_seq_midi_event nf_nat_ftp nf_nat snd_seq nf_conntrack_ipv4 nf_defrag_ipv4 nf_conntrack_ftp snd_timer snd_seq_device nf_conntrack iptable_filter snd ip_tables x_tables usblp asus_atk0110 edac_core soundcore edac_mce_amd snd_page_alloc nvidia(P) i2c_piix4 xhci k10temp lp parport cryptd aes_x86_64 aes_generic xts gf128mul dm_crypt raid10 raid456 async_pq async_xor xor async_memcpy async_raid6_recov usbhid hid raid6_pq async_tx raid0 multipath linear raid1 usb_storage ohci1394 pata_jmicron sky2 ieee1394 ahci
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] Pid: 2577, comm: Xorg Tainted: P 2.6.32-25-generic #45-Ubuntu System Product Name
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] RIP: 0010:[<ffffffffa069e4fc>] [<ffffffffa069e4fc>] os_io_read_byte+0xc/0x10 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] RSP: 0018:ffff88011d01bce8 EFLAGS: 00000292
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] RAX: 0000000000000058 RBX: ffff88011d01bce8 RCX: ffffffffa0a33600
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] RDX: 00000000000003d5 RSI: 00000000000c32b1 RDI: 00000000000003d5
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] RBP: ffffffff81012cae R08: 0000000000000000 R09: 0000000000000001
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] R10: 0000000000000001 R11: 0000000000000000 R12: ffff88011d78df20
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] R13: ffffffff81012cae R14: ffff88011d78df20 R15: ffffffff81012cae
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] FS: 00007ff91d632700(0000) GS:ffff880028200000(0000) knlGS:0000000000000000
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] CR2: 00007ff91737bf50 CR3: 000000012976a000 CR4: 00000000000006f0
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Nov 10 19:06:16 zombywoof kernel: [ 267.370008] Call Trace:
Nov 10 19:06:16 zombywoof kernel: [ 267.370900] [<ffffffffa0683e9d>] ? _nv000985rm+0xe/0x22 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.371762] [<ffffffffa0683dcc>] ? _nv028185rm+0xc/0x14 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.371762] [<ffffffffa068eaa0>] ? _nv027982rm+0x11/0x26 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa06857a6>] ? _nv002440rm+0x83/0xa4 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa06843a7>] ? _nv020773rm+0x1cb/0x208 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa067565c>] ? _nv021796rm+0xb0/0x364 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa067c949>] ? _nv002180rm+0x410/0x6d3 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa067c904>] ? _nv002180rm+0x3cb/0x6d3 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa067d3ad>] ? _nv002073rm+0x61f/0x6bd [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa06793c8>] ? _nv002084rm+0xf0/0x18a [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa067ea15>] ? rm_disable_adapter+0x78/0xcb [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffffa069af0b>] ? nv_kern_close+0x37b/0x420 [nvidia]
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffff811449a5>] ? __fput+0xf5/0x210
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffff81144ae5>] ? fput+0x25/0x30
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffff81140bdd>] ? filp_close+0x5d/0x90
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffff81140cc7>] ? sys_close+0xb7/0x120
Nov 10 19:06:16 zombywoof kernel: [ 267.372588] [<ffffffff810121b2>] ? system_call_fastpath+0x16/0x1b
Hope this helps nail down this nasty bug. All other factors remain the same from the last try. Again, thanks for looking and any suggestions on what to do are eagerly awaited!
Attached Files
File Type: gz nvidia-bug-report.log.gz (78.7 KB, 75 views)
Glajvect is offline   Reply With Quote
Old 11-10-10, 10:53 PM   #3
Milena
Registered User
 
Join Date: Mar 2009
Posts: 36
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

The "black screen freeze" or freeze with artifacts upon login/driver load with the 400 series, especially evga cards, is well known for windows users. nvidia.com and evga.com forums are flooding with threads about it. Your card is most likely defect and need to be RMA'd. I have the same problem with 2 different evga gtx470 cards on both windows 7 and arch linux. Most of the people with this problem have an ASUS MB. Some say increasing the NB voltage a bit helps with stability as these cards are stressing the northbridge a lot but i can't confirm that. I get the freeze every 5-10 restarts, its totally random and happens on both OS. I will get rid of the 400 series asap, my old 285 works perfectly fine in the same system.

My system is kinda similar to yours, ASUS MB and AMD CPU.
Milena is offline   Reply With Quote
Old 11-11-10, 06:17 PM   #4
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

Thanks for the heads up, Milena! Indeed, I haven't totally dismissed the possibility that the card is defective; however, it seems to work well enough in Windows 7 Ultimate. Maybe the drivers are different enough to trigger a hardware failure in one OS, but not the other? Let me try loading some more "stressful" applications in Windows and I'll give it another go there.

Adjusting the NB voltage is an idea I hadn't considered. I'll bump it up a bit and see if that helps any in Linux.
Glajvect is offline   Reply With Quote
Old 11-13-10, 06:50 AM   #5
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

It turns out that there *was* a RAM stability issue in Windows that I didn't notice earlier (hadn't messed around in Windows much up until now. It's telling that the problems never showed up under Linux!). I increased the DRAM voltage to around 1.75 and NB voltage to around 1.20. That seemed to stop the random blue screens in Windows and any errors as reported by Memtest86+. Not so much as one problem, even after running hours of glitzy, modern FPS demos that I downloaded off of Steam; and so I've now essentially ruled out the possibility of a defective card.

As much as I would like to say that fixing the stability issue in Windows also corrected my Linux driver problem, it has not. Guess I'll be waiting around for that auspicious driver release which makes my card start working...
Glajvect is offline   Reply With Quote
Old 12-15-10, 03:20 PM   #6
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

Still the same results with the 260.19.29 driver and 2.6.32-26-generic Lucid kernel.
Attached Files
File Type: gz nvidia-bug-report.log.gz (104.9 KB, 68 views)
Glajvect is offline   Reply With Quote
Old 01-23-11, 03:12 AM   #7
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

260.19.36: Sweet! I can't believe it! Amazing!

...

Ah, hell -- I can't lie to you guys. *Still* no luck with this driver, either. In the time since my last report, I've upgraded to the back-ported Natty kernel, 2.6.37-12, but this seems to make no difference in my ability to use 260.19.29 or .36.

This card works great in Windows. The built-in temperature-based fan speed could use some better calibration (gets too hot, e.g. Mafia 2), but that's nothing that can't be fixed with software fan control. After hours of Windows gaming, I'm quite sure that it's not defective. Don't let me discourage you from suggesting other ways of verifying this if you think it might still be an issue.

Can someone (NVidia perhaps?!) please do one of the following for me:
- Tell me how to grab the stack trace that I expect would come after the "Backtrace:" line in Xorg.0.log. With that, I expect that I can help you easily figure out where things are coming unraveled
- Offer up some other ideas as to kernel settings, Xorg.conf settings, etc. that might serve to at *least* get me into the X login without locking up my machine at the local terminal (really sucks having to reboot via SSH between all the testing I've had to do to figure out what could be wrong)
- Point out the phenomenally bone-headed mistake I'm making in how my system is set up (e.g. kernel or Xorg version is completely unreasonable, possible bad BIOS settings to avoid using with this motherboard's chipset and the GTX 465), how the drivers are installed, or my testing procedure. Other common pitfalls that might result in this misbehavior are welcome. Let's leave no stone unturned!

I'm stuck in low-res VESA land and it sucks. I use this machine day-to-day for software development, and I miss the hell out of the screen real-estate I had with my previous PC. I'd also love to make use of my second monitor, but it seems that's totally out of the question. I haven't a care about 3D acceleration in Linux (but Compiz is nice at times...) so long as it works in Windows as I do all of my gaming there anyway. "But Glajvect, just use another open-source driver", you might offer. Well, Nouveau is also not a good solution at this time since their support matrix shows that they are currently implementing support for the Fermi cards. Who knows how long that will be in the works! Whatever the true status is, I've been there, done that, and I didn't have any luck getting that driver working. Thing is, I can't be annoyed at that team since they certainly don't have any sort of advantage in completing their task over the people who made the card!

The only other option I see is to buy a new (but older) card. I'd really prefer to avoid doing that as I spent enough as it is on this one. There's no real guarantee it would work, either. All the other old cards I have sitting around are -- naturally -- AGP. Been a while since the last upgrade...

So where do I go from here?
Attached Files
File Type: gz nvidia-bug-report.log.gz (113.9 KB, 64 views)
Glajvect is offline   Reply With Quote
Old 01-23-11, 04:56 AM   #8
Licaon
Registered User
 
Licaon's Avatar
 
Join Date: Nov 2004
Location: Between the keyboard and the chair.
Posts: 490
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

Quote:
Originally Posted by Glajvect View Post
Can someone (NVidia perhaps?!) please do one of the following for me:
what other distros have you tried?
Licaon is offline   Reply With Quote

Old 01-23-11, 10:14 AM   #9
Glajvect
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

None yet, but I'm certainly willing to try some different LiveCDs. I'll try a few of the major distros to see where they're at. If you know of one in particular that has a good track record, I'm open to suggestions!
Glajvect is offline   Reply With Quote
Old 01-25-11, 07:01 AM   #10
zooz
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

Same here.. with Nvidia 8800GT

I tried pretty much everything I could, but still getting xorg freezes and errors in the logs.

It is so annoying that no one was able to fix very well known issue since probably 2008.

If NVIDIA keeps its specs closed - fair enough, but please do fix the bloody bugs.

Things I tried:

* Different kernel versions - starting from 2.6.30 to 2.6.37
* Different Xorg server versions - starting from 1.8.0 to 1.9.3 (freezes are less frequent with 1.8.x)
* Various NVIDIA driver versions
* Flashed my mainboard's bios to the newest one
* Tried various options in BIOS which might be related to video card / memory management
* Tried various different options in xorg.conf (suggestions from million other forums)
* Tried various different kernel options (same - as suggested by other people to address this issue)
* Prayed
* Broke a couple of keyboards
* etc

And guess what? - NOTHING HELPED!!!

I am still experiencing xorg freezes and errors.

I have always been a big fan of NVIDIA, but unfortunately I can only see one solution - spend some money on AMD video card or similar.

/me sighs..
zooz is offline   Reply With Quote
Old 01-25-11, 12:50 PM   #11
Licaon
Registered User
 
Licaon's Avatar
 
Join Date: Nov 2004
Location: Between the keyboard and the chair.
Posts: 490
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

@zooz: you're not helping by spamming the forums

try other distros than Ubuntu too, FFS
Licaon is offline   Reply With Quote
Old 01-25-11, 06:08 PM   #12
zooz
Registered User
 
Join Date: Nov 2010
Posts: 8
Default Re: 260.19.12 black screen on GTX 465, Ubuntu 10.04 64

Sorry for spamming.

Please see my other reply to similar/same problem:

http://www.nvnews.net/vbulletin/showthread.php?t=158233

I think I found a solution for this.
zooz is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 06:11 PM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.