error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start

BarbuDreadMon · 05-13-10, 10:33 AM

I'm running Gentoo Linux. Laptop is Asus N61Vn with nvidia GT240M.

I got the laptop for over 6 months, everything was working fine up to kernel 2.6.31-gentoo-r10 and nvidia-drivers-190.42-r3 (included).
Got the problem when i upgraded my kernel to latest gentoo stable version aka 2.6.32-gentoo-r7.
I tried upgrading my nvidia-drivers but it doesn't work either, only thing working is returning to my old 2.6.31 kernel.

System hang 5 second after i try launching X, it become unreachable even with a ssh session opened from my desktop to my laptop before crash. Error is "NVIDIA(0): WAIT: (E, 0, 0x857d, 0)" in my Xorg.0.log.

PS : i got almost the same config (system and hardware) on my desktop except card is GT240 instead of GT240M, no problem on it.

BarbuDreadMon · 05-17-10, 04:41 AM

I tried with kernel 2.6.33, same problem. Seems like it's related to new features in kernel > 2.6.31.
Problem seems to occur just after he enable acpi hotkeys events and just before he initialize OpenGL acceleration.

dave_mwi · 05-20-10, 01:41 AM

I'm running fedora, but the way I got around this was to add acpi=off as one of the kernel arguments.

After I did that no more problems booting and no more WAIT error.

You can add this value to the end of the 'kernel' lines in /etc/grub.conf (remember this is fedora)

or you can add them to the kernel lines when your grub.conf starts, if you're using grub.

hope this helps.

kernel 2.6.32

BarbuDreadMon · 05-27-10, 12:32 PM

Thanks for the tip but i will stay on my kernel 2.6.31 with acpi until the bug with acpi and mobile nvidia is fixed in 2.6.32+ kernel

olifre · 05-27-10, 06:21 PM

Hello,

I'm running fine here with an 8600M GT and Gentoo, currently on kernel 2.6.33.
Maybe the driver-version does not fit with the newer kernel? I am using 256.25 now, but 195.36.24 (needs to be keyworded, but is in portage) worked fine for me, too.
Maybe just try that one?

Good luck!

aljo666 · 05-28-10, 01:09 PM

I also have similar problems in fedora 12 and 13 (kernel > 2.6.30). No problems with fedora 11 an 2.6.30 kernel. The screen starts flickering or the xserver doesn't even start. I've attached the xorg.log files with errors like
[ 273.523] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x00007c48, 0x00007e38)
[ 274.852] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[ 274.852]
Backtrace:
[ 274.874] 0: /usr/bin/Xorg (xorg_backtrace+0x28) [0x49ecb8]
[ 274.874] 1: /usr/bin/Xorg (mieqEnqueue+0x1f4) [0x49e664]
[ 274.874] 2: /usr/bin/Xorg (xf86PostMotionEventP+0xc4) [0x477e24]
[ 274.874] 3: /usr/lib64/xorg/modules/input/evdev_drv.so (0x7fbec02b9000+0x3dbf) [0x7fbec02bcdbf]
[ 274.874] 4: /usr/bin/Xorg (0x400000+0x6aae7) [0x46aae7]
[ 274.874] 5: /usr/bin/Xorg (0x400000+0x1180f3) [0x5180f3]
[ 274.874] 6: /lib64/libc.so.6 (0x3fba000000+0x32a40) [0x3fba032a40]
[ 274.874] 7: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7fbec0b21000+0x78140) [0x7fbec0b99140]
[ 274.874] 8: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (_nv001056X+0x289) [0x7fbec0b99d79]
[ 274.874] 9: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7fbec0b21000+0xd5206) [0x7fbec0bf6206]
[ 274.874] 10: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7fbec0b21000+0x383c7e) [0x7fbec0ea4c7e]
[ 274.874] 11: /usr/bin/Xorg (0x400000+0xce777) [0x4ce777]
[ 274.874] 12: /usr/bin/Xorg (0x400000+0x2c32c) [0x42c32c]
[ 274.874] 13: /usr/bin/Xorg (0x400000+0x219ca) [0x4219ca]
[ 274.874] 14: /lib64/libc.so.6 (__libc_start_main+0xfd) [0x3fba01ec5d]
[ 274.874] 15: /usr/bin/Xorg (0x400000+0x21579) [0x421579]
[ 280.523] (WW) NVIDIA(0): WAIT (1, 6, 0x8000, 0x00007c48, 0x00007e38)

**danix** · 05-29-10, 08:49 AM

BarbuDreadMon, do you still experience this crash if you pass option intel_iommu=off to the kernel at boot?

If intel_iommu=off prevents the crash, then this is a bug which should be fixed in the next release.

BarbuDreadMon · 05-30-10, 02:29 AM

No more crash with intel_iommu=off at boot time. I'll wait for next release

.

Scramblejams · 04-23-11, 03:51 AM

Last several weeks I've noticed my laptop screen sometimes not coming up out of suspend, though I was changing kernel versions a lot so I thought that was causing it. Just had the opportunity to ssh into the laptop during one of those moments, and I see the exact same error in Xorg.0.log.

I'm on 2.6.32 amd64 with driver 270.41.06. I'll give the suggested intel_iommu=off tip a try. If that doesn't work, I'm going to go back to one of the 190 series, that was the last time I was free of this other problem: http://www.nvnews.net/vbulletin/showthread.php?t=158091

Steve

Scramblejams · 04-24-11, 03:09 AM

Does anybody know whether disabling VT-d is identical to specifying intel_iommu=off?

(BTW, forgot to add on the previous post that I'm on a Thinkpad T410.)

Scramblejams · 05-10-11, 05:00 PM

Just had it occur again waking from sleep. Thinkpad T410 with NVS 3100M, kernel 2.6.32-21, driver 270.41.06. Didn't boot with intel_iommu=off, but I've had VT-d disabled from the BIOS since my last post. Gonna see if I can possibly revert to 2.6.31, based on BarbuDreadMon's experience.

Edit: Unable to revert since 2.6.31 is incompatible with my wifi card.

05-17-10, 04:41 AM	#2
BarbuDreadMon Registered User Join Date: May 2010 Posts: 4	X doesn't start with GT240M and kernel > 2.6.31 I tried with kernel 2.6.33, same problem. Seems like it's related to new features in kernel > 2.6.31. Problem seems to occur just after he enable acpi hotkeys events and just before he initialize OpenGL acceleration.

05-20-10, 01:41 AM	#3
dave_mwi Registered User Join Date: Sep 2007 Posts: 3	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start I'm running fedora, but the way I got around this was to add acpi=off as one of the kernel arguments. After I did that no more problems booting and no more WAIT error. You can add this value to the end of the 'kernel' lines in /etc/grub.conf (remember this is fedora) or you can add them to the kernel lines when your grub.conf starts, if you're using grub. hope this helps. kernel 2.6.32

05-27-10, 12:32 PM	#4
BarbuDreadMon Registered User Join Date: May 2010 Posts: 4	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start Thanks for the tip but i will stay on my kernel 2.6.31 with acpi until the bug with acpi and mobile nvidia is fixed in 2.6.32+ kernel

05-27-10, 06:21 PM	#5
olifre Registered User Join Date: May 2008 Posts: 50	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start Hello, I'm running fine here with an 8600M GT and Gentoo, currently on kernel 2.6.33. Maybe the driver-version does not fit with the newer kernel? I am using 256.25 now, but 195.36.24 (needs to be keyworded, but is in portage) worked fine for me, too. Maybe just try that one? Good luck!

05-28-10, 01:09 PM	#6
aljo666 Registered User Join Date: Apr 2006 Posts: 2	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start I also have similar problems in fedora 12 and 13 (kernel > 2.6.30). No problems with fedora 11 an 2.6.30 kernel. The screen starts flickering or the xserver doesn't even start. I've attached the xorg.log files with errors like [ 273.523] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x00007c48, 0x00007e38) [ 274.852] [mi] EQ overflowing. The server is probably stuck in an infinite loop. [ 274.852] Backtrace: [ 274.874] 0: /usr/bin/Xorg (xorg_backtrace+0x28) [0x49ecb8] [ 274.874] 1: /usr/bin/Xorg (mieqEnqueue+0x1f4) [0x49e664] [ 274.874] 2: /usr/bin/Xorg (xf86PostMotionEventP+0xc4) [0x477e24] [ 274.874] 3: /usr/lib64/xorg/modules/input/evdev_drv.so (0x7fbec02b9000+0x3dbf) [0x7fbec02bcdbf] [ 274.874] 4: /usr/bin/Xorg (0x400000+0x6aae7) [0x46aae7] [ 274.874] 5: /usr/bin/Xorg (0x400000+0x1180f3) [0x5180f3] [ 274.874] 6: /lib64/libc.so.6 (0x3fba000000+0x32a40) [0x3fba032a40] [ 274.874] 7: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7fbec0b21000+0x78140) [0x7fbec0b99140] [ 274.874] 8: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (_nv001056X+0x289) [0x7fbec0b99d79] [ 274.874] 9: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7fbec0b21000+0xd5206) [0x7fbec0bf6206] [ 274.874] 10: /usr/lib64/xorg/modules/drivers/nvidia_drv.so (0x7fbec0b21000+0x383c7e) [0x7fbec0ea4c7e] [ 274.874] 11: /usr/bin/Xorg (0x400000+0xce777) [0x4ce777] [ 274.874] 12: /usr/bin/Xorg (0x400000+0x2c32c) [0x42c32c] [ 274.874] 13: /usr/bin/Xorg (0x400000+0x219ca) [0x4219ca] [ 274.874] 14: /lib64/libc.so.6 (__libc_start_main+0xfd) [0x3fba01ec5d] [ 274.874] 15: /usr/bin/Xorg (0x400000+0x21579) [0x421579] [ 280.523] (WW) NVIDIA(0): WAIT (1, 6, 0x8000, 0x00007c48, 0x00007e38)

05-29-10, 08:49 AM	#7
danix NVIDIA Corporation Join Date: Feb 2010 Location: Santa Clara, CA Posts: 237	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start BarbuDreadMon, do you still experience this crash if you pass option intel_iommu=off to the kernel at boot? If intel_iommu=off prevents the crash, then this is a bug which should be fixed in the next release.

05-30-10, 02:29 AM	#8
BarbuDreadMon Registered User Join Date: May 2010 Posts: 4	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start No more crash with intel_iommu=off at boot time. I'll wait for next release .

04-23-11, 03:51 AM	#9
Scramblejams Registered User Join Date: Jan 2008 Location: USA Posts: 11	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start Last several weeks I've noticed my laptop screen sometimes not coming up out of suspend, though I was changing kernel versions a lot so I thought that was causing it. Just had the opportunity to ssh into the laptop during one of those moments, and I see the exact same error in Xorg.0.log. I'm on 2.6.32 amd64 with driver 270.41.06. I'll give the suggested intel_iommu=off tip a try. If that doesn't work, I'm going to go back to one of the 190 series, that was the last time I was free of this other problem: http://www.nvnews.net/vbulletin/showthread.php?t=158091 Steve

04-24-11, 03:09 AM	#10
Scramblejams Registered User Join Date: Jan 2008 Location: USA Posts: 11	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start Does anybody know whether disabling VT-d is identical to specifying intel_iommu=off? (BTW, forgot to add on the previous post that I'm on a Thinkpad T410.)

05-10-11, 05:00 PM	#11
Scramblejams Registered User Join Date: Jan 2008 Location: USA Posts: 11	Re: error "WAIT: (E, 0, 0x857d, 0)" => X doesn't start Just had it occur again waking from sleep. Thinkpad T410 with NVS 3100M, kernel 2.6.32-21, driver 270.41.06. Didn't boot with intel_iommu=off, but I've had VT-d disabled from the BIOS since my last post. Gonna see if I can possibly revert to 2.6.31, based on BarbuDreadMon's experience. Edit: Unable to revert since 2.6.31 is incompatible with my wifi card.