dmesg spam: "PnPBIOS: pnp_dock_thread: unexpected status 0x5 flooding"

dmesg spam: "PnPBIOS: pnp_dock_thread: unexpected status 0x5 flooding"

[devzero]

Hi, 

on an ancient GX1 Geode embedded board (Evo T20/Wyse wt3235le Thin Client, which are cheap (<$10) and low power x86 embeded platform) , dmesg is getting constantly spammed with "PnPBIOS: pnp_dock_thread: unexpected status 0x5" during and after boot. 

Any ideas what this means and how to stop that?

It seems it comes from drivers/pnp/pnpbios/bioscalls.c -> pnpbios_print_status()  

Tried adding pnpbios=off, but with that the system does not boot anymore.

Being curious, why a non-existing docking-station is being "polled at regular intervals" (i.e. every 2 seconds - see drivers/pnp/pnpbios/core.c )

>From http://books.google.de/books?id=ibLa4I5EnC4C&pg=PA231&lpg=PA231&dq=pnp+docking+bios&source=bl&ots=ekCFm34U_B&sig=9Z9L55IAL7_3NtuM10jT3serncM&hl=de&sa=X&ei=mAZcVILnJsviO-r5gNgI&ved=0CCsQ6AEwAQ#v=onepage&q=pnp%20docking%20bios&f=false   i read:  "Function 05h - Get Docking Status Information at page 251" 

it  seems this is being the result of a "docking event" ? Does this mean, my system is erroneusly generating such docking events, even if there is no docking station at all?

regards
roland

ps:
i`m booting via plopKexec Bootloader ( http://www.plop.at/de/plopkexec.html ) , as i can put kernel+rootfs on external usb and don`t need to reflash the device, which is complicated task.

pps:
some few old pointers from other people reporting it:
https://lists.debian.org/debian-kernel/2005/02/msg00238.html
http://sourceforge.net/p/open-evot20/mailman/message/20065770/

Re: dmesg spam: "PnPBIOS: pnp_dock_thread: unexpected status 0x5 flooding"

[One Thousand Gnomes]

On Fri, 7 Nov 2014 00:54:31 +0100
devzero@web.de wrote:

> Hi, 
> 
> on an ancient GX1 Geode embedded board (Evo T20/Wyse wt3235le Thin Client, which are cheap (<$10) and low power x86 embeded platform) , dmesg is getting constantly spammed with "PnPBIOS: pnp_dock_thread: unexpected status 0x5" during and after boot. 
> 
> Any ideas what this means and how to stop that?
> 
> It seems it comes from drivers/pnp/pnpbios/bioscalls.c -> pnpbios_print_status()  
> 
> Tried adding pnpbios=off, but with that the system does not boot anymore.
> 
> Being curious, why a non-existing docking-station is being "polled at regular intervals" (i.e. every 2 seconds - see drivers/pnp/pnpbios/core.c )

We poll for a dock, and if the BIOS reports that the function is not
supported we then exit the thread.

> 
> >From http://books.google.de/books?id=ibLa4I5EnC4C&pg=PA231&lpg=PA231&dq=pnp+docking+bios&source=bl&ots=ekCFm34U_B&sig=9Z9L55IAL7_3NtuM10jT3serncM&hl=de&sa=X&ei=mAZcVILnJsviO-r5gNgI&ved=0CCsQ6AEwAQ#v=onepage&q=pnp%20docking%20bios&f=false   i read:  "Function 05h - Get Docking Status Information at page 251" 

The spec is available at:

http://download.intel.com/support/motherboards/desktop/sb/pnpbiosspecificationv10a.pdf
 
> it  seems this is being the result of a "docking event" ? Does this mean, my system is erroneusly generating such docking events, even if there is no docking station at all?

Your BIOS appears to be making strange replies. Any error ought to have
the top bit set, so its reporting a nonsense value for some reason.

I've not seen other reports of pnp_dock_thread errors with a real dock.
The only reports at all are a couple of old ones in google that seem to
be similar and a Debian bug (294652) which appears to be your bug so I'd
suggest testing the following


Change the default to

              default:
                  pnpbios_print_status("pnp_dock..... etc)
	          printk(KERN_ERR "PnPBIOS: Disabling dock monitoring\n");
	          complete_and_exit(&unload_sem, 0);

that should produce you one error message, a warning that dock monitoring
is being disabled and then it should stfu.

If you can let us know if that does the trick then I can push it upstream.

Alan

Re: dmesg spam: "PnPBIOS: pnp_dock_thread: unexpected status 0x5 flooding"

[devzero]

> Change the default to
> 
>               default:
>                   pnpbios_print_status("pnp_dock..... etc)
> 	          printk(KERN_ERR "PnPBIOS: Disabling dock monitoring\n");
> 	          complete_and_exit(&unload_sem, 0);
> 
> that should produce you one error message, a warning that dock monitoring
> is being disabled and then it should stfu.
> 
> If you can let us know if that does the trick then I can push it upstream.

works perfectly - thanks very much !

[    9.712027] Switching to clocksource tsc
[    9.903305] udevd[42]: starting version 175
[   10.637309] natsemi dp8381x driver, version 2.1, Sept 11, 2006
[   10.637345]   originally by Donald Becker <becker@scyld.com>
[   10.637371]   2.4.x kernel port by Jeff Garzik, Tjeerd Mulder
[   10.700287] natsemi 0000:00:0f.0: setting latency timer to 64
[   10.880548] natsemi eth0: NatSemi DP8381[56] at 0x8010000 (0000:00:0f.0), 00:00:00:00:00:00, IRQ 10, port TP.
[   11.025143] PnPBIOS: pnp_dock_thread: unexpected status 0x5
>[  11.026800] PnPBIOS: Disabling dock monitoring
[   11.518269] SCSI subsystem initialized
[   11.550857] usbcore: registered new interface driver usbfs


> I've not seen other reports of pnp_dock_thread errors with a real dock.
> The only reports at all are a couple of old ones in google that seem to
> be similar and a Debian bug (294652) which appears to be your bug so I'd

indeed - seems a rather weird/rare/old one. 

glad that you fixed it, though. thanks again!

roland


> Gesendet: Montag, 10. November 2014 um 14:54 Uhr
> Von: "One Thousand Gnomes" <gnomes@lxorguk.ukuu.org.uk>
> An: devzero@web.de
> Cc: linux-kernel@vger.kernel.org, rafael.j.wysocki@intel.com
> Betreff: Re: dmesg spam: "PnPBIOS: pnp_dock_thread: unexpected status 0x5 flooding"
>
> On Fri, 7 Nov 2014 00:54:31 +0100
> devzero@web.de wrote:
> 
> > Hi, 
> > 
> > on an ancient GX1 Geode embedded board (Evo T20/Wyse wt3235le Thin Client, which are cheap (<$10) and low power x86 embeded platform) , dmesg is getting constantly spammed with "PnPBIOS: pnp_dock_thread: unexpected status 0x5" during and after boot. 
> > 
> > Any ideas what this means and how to stop that?
> > 
> > It seems it comes from drivers/pnp/pnpbios/bioscalls.c -> pnpbios_print_status()  
> > 
> > Tried adding pnpbios=off, but with that the system does not boot anymore.
> > 
> > Being curious, why a non-existing docking-station is being "polled at regular intervals" (i.e. every 2 seconds - see drivers/pnp/pnpbios/core.c )
> 
> We poll for a dock, and if the BIOS reports that the function is not
> supported we then exit the thread.
> 
> > 
> > >From http://books.google.de/books?id=ibLa4I5EnC4C&pg=PA231&lpg=PA231&dq=pnp+docking+bios&source=bl&ots=ekCFm34U_B&sig=9Z9L55IAL7_3NtuM10jT3serncM&hl=de&sa=X&ei=mAZcVILnJsviO-r5gNgI&ved=0CCsQ6AEwAQ#v=onepage&q=pnp%20docking%20bios&f=false   i read:  "Function 05h - Get Docking Status Information at page 251" 
> 
> The spec is available at:
> 
> http://download.intel.com/support/motherboards/desktop/sb/pnpbiosspecificationv10a.pdf
>  
> > it  seems this is being the result of a "docking event" ? Does this mean, my system is erroneusly generating such docking events, even if there is no docking station at all?
> 
> Your BIOS appears to be making strange replies. Any error ought to have
> the top bit set, so its reporting a nonsense value for some reason.
> 
> I've not seen other reports of pnp_dock_thread errors with a real dock.
> The only reports at all are a couple of old ones in google that seem to
> be similar and a Debian bug (294652) which appears to be your bug so I'd
> suggest testing the following
> 
> 
> Change the default to
> 
>               default:
>                   pnpbios_print_status("pnp_dock..... etc)
> 	          printk(KERN_ERR "PnPBIOS: Disabling dock monitoring\n");
> 	          complete_and_exit(&unload_sem, 0);
> 
> that should produce you one error message, a warning that dock monitoring
> is being disabled and then it should stfu.
> 
> If you can let us know if that does the trick then I can push it upstream.
> 
> Alan
>