3.15-rc1: Endless "Disabling PPGTT because VT-d is on" messages in /var/log/messages

3.15-rc1: Endless "Disabling PPGTT because VT-d is on" messages in /var/log/messages

[Alessandro Suardi]

[hoping I persuaded gmail to get my old text-only email settings back, grrr]


Noticed that my root fs was filling up, and found out that /var/log/messages was
 approaching 2GB of space, most of which were as per $subject. To give an idea
 of the magnitude, here's a snippet from a saved sample of /var/log/messages:

[root@xbox log]# grep  "Disabling PP" messages.save.head |head -1
Apr 15 03:44:10 xbox kernel: [    3.115126] [drm] Disabling PPGTT
because VT-d is on
[root@xbox log]# grep  "Disabling PP" messages.save.head |tail -1
Apr 15 03:45:50 xbox kernel: [  130.165219] [drm] Disabling PPGTT
because VT-d is on
[root@xbox log]# grep -c "Disabling PP" messages.save.head
610

Or from the dmesg ring:

[root@xbox log]# dmesg|head
[22446.690572] [drm] Disabling PPGTT because VT-d is on
[22446.690616] [drm] Disabling PPGTT because VT-d is on
[22446.704560] [drm] Disabling PPGTT because VT-d is on
[22446.706747] [drm] Disabling PPGTT because VT-d is on
[22446.720202] [drm] Disabling PPGTT because VT-d is on
[22446.722447] [drm] Disabling PPGTT because VT-d is on
[22446.724850] [drm] Disabling PPGTT because VT-d is on
[22446.727268] [drm] Disabling PPGTT because VT-d is on
[22446.733052] [drm] Disabling PPGTT because VT-d is on
[22446.735319] [drm] Disabling PPGTT because VT-d is on



Are there patches for this issue, or is a workaround available to prevent a
 filesystem full condition on my root fs?


This is on a Dell Latitude E6420 laptop, Fedora 19 x86_64, onboard
 Intel video card... yes, I have an old A08 BIOS, but I've never seen
 this issue while regularly updating my mainline kernel weekly or so...


Available for more details if needed, thanks in advance,

--alessandro

"don't underestimate the things that I will do"

 (Adele, "Rolling In The Deep")

Re: [Intel-gfx] Fwd: 3.15-rc1: Endless "Disabling PPGTT because VT-d is on" messages in /var/log/messages

[Daniel Vetter]

On Fri, Apr 18, 2014 at 06:29:45PM +0200, Alessandro Suardi wrote:
> forwarding from my lkml post as probably more appropriate to send this here...

Yup, but just adding more cc's is preferred. I've re-added lkml for
reference. Bug is fairly obvious regression in 3.15, patch is on the way.
-Daniel

> 
> thanks,
> 
> ---------- Forwarded message ----------
> From: Alessandro Suardi <alessandro.suardi@gmail.com>
> Date: Fri, Apr 18, 2014 at 12:31 AM
> Subject: 3.15-rc1: Endless "Disabling PPGTT because VT-d is on"
> messages in /var/log/messages
> To: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
> 
> 
> [hoping I persuaded gmail to get my old text-only email settings back, grrr]
> 
> 
> Noticed that my root fs was filling up, and found out that /var/log/messages was
>  approaching 2GB of space, most of which were as per $subject. To give an idea
>  of the magnitude, here's a snippet from a saved sample of /var/log/messages:
> 
> [root@xbox log]# grep  "Disabling PP" messages.save.head |head -1
> Apr 15 03:44:10 xbox kernel: [    3.115126] [drm] Disabling PPGTT
> because VT-d is on
> [root@xbox log]# grep  "Disabling PP" messages.save.head |tail -1
> Apr 15 03:45:50 xbox kernel: [  130.165219] [drm] Disabling PPGTT
> because VT-d is on
> [root@xbox log]# grep -c "Disabling PP" messages.save.head
> 610
> 
> Or from the dmesg ring:
> 
> [root@xbox log]# dmesg|head
> [22446.690572] [drm] Disabling PPGTT because VT-d is on
> [22446.690616] [drm] Disabling PPGTT because VT-d is on
> [22446.704560] [drm] Disabling PPGTT because VT-d is on
> [22446.706747] [drm] Disabling PPGTT because VT-d is on
> [22446.720202] [drm] Disabling PPGTT because VT-d is on
> [22446.722447] [drm] Disabling PPGTT because VT-d is on
> [22446.724850] [drm] Disabling PPGTT because VT-d is on
> [22446.727268] [drm] Disabling PPGTT because VT-d is on
> [22446.733052] [drm] Disabling PPGTT because VT-d is on
> [22446.735319] [drm] Disabling PPGTT because VT-d is on
> 
> 
> 
> Are there patches for this issue, or is a workaround available to prevent a
>  filesystem full condition on my root fs?
> 
> 
> This is on a Dell Latitude E6420 laptop, Fedora 19 x86_64, onboard
>  Intel video card... yes, I have an old A08 BIOS, but I've never seen
>  this issue while regularly updating my mainline kernel weekly or so...
> 
> 
> Available for more details if needed, thanks in advance,
> 
> --alessandro
> 
> "don't underestimate the things that I will do"
> 
>  (Adele, "Rolling In The Deep")
> 
> 
> -- 
> --alessandro
> 
> "don't underestimate the things that I will do"
> 
>  (Adele, "Rolling In The Deep")
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch