[Lenovo ThinkPad T440p][10de:1290] switching nvidia graphic card raises fwts high failure

[Taihsiang Ho]

Lenovo ThinkPad T440p ( CID 201309-14169 ) on Ubuntu 14.04
3.13.0-24-generic

----------------------------------------

fwts high failures were found as the following:

klog: HIGH Kernel message: [ 27.371715] ACPI Warning: \_SB_.PCI0.PEG_.VID_._DSM: Argument #4 type mismatch - Found [Buffer], ACPI requires [Package] (20131115/nsarguments-95)
klog: HIGH Kernel message: [ 29.882167] ACPI Error: Field [TBF3] at 270336 exceeds Buffer [NULL] size 262144 (bits) (20131115/dsopcode-236)
klog: HIGH Kernel message: [ 29.882172] ACPI Error: Method parse/execution failed [\_SB_.PCI0.PEG_.VID_.GETB] (Node ffff8802333d1960), AE_AML_BUFFER_LIMIT (20131115/psparse-536)
klog: HIGH Kernel message: [ 29.882179] ACPI Error: Method parse/execution failed [\_SB_.PCI0.PEG_.VID_._ROM] (Node ffff8802333d1938), AE_AML_BUFFER_LIMIT (20131115/psparse-536)

----------------------------------------

Steps to reproduce the bug:
1.on the power-saving mode (intel chip),
fwts found a klog high failure

klog: HIGH Kernel message: [ 26.704702] ACPI Warning: \_SB_.PCI0.PEG_.VID_._DSM: Argument #4 type mismatch - Found [Buffer], ACPI requires [Package] (20131115/nsarguments-95)

by the command
fwts -q --stdout-summary -r <log file name> klog

2. switch to the nvidia graphic card via dash → nvidia setting → select nvidia
3. reboot to make the switching work
4. use the same fwts command and it will show the high failures. There are 3 more high failures found.

reproducing rate: 5 out of 5

----------------------------------------

More information:
This high failiure

klog: HIGH Kernel message: [ 27.371715] ACPI Warning: \_SB_.PCI0.PEG_.VID_._DSM: Argument #4 type mismatch - Found [Buffer], ACPI requires [Package] (20131115/nsarguments-95)

was explained in
LP: #1281966/comments/4

-----------------------------------------

ProblemType: Bug
DistroRelease: Ubuntu 14.04
Package: linux-image-3.13.0-24-generic 3.13.0-24.46 [modified: boot/vmlinuz-3.13.0-24-generic]
ProcVersionSignature: Ubuntu 3.13.0-24.46-generic 3.13.9
Uname: Linux 3.13.0-24-generic x86_64
NonfreeKernelModules: nvidia
ApportVersion: 2.14.1-0ubuntu3
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC1: ubuntu 1790 F.... pulseaudio
 /dev/snd/controlC0: ubuntu 1790 F.... pulseaudio
CurrentDesktop: Unity
Date: Wed Sep 3 00:15:16 2014
HibernationDevice: RESUME=UUID=b437ec5c-4a11-4108-a4b7-900f2ea9f312
InstallationDate: Installed on 2014-08-27 (6 days ago)
InstallationMedia: Ubuntu 14.04 LTS "Trusty Tahr" - Release amd64 (20140417)
MachineType: LENOVO 20ANZ03CUS
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.13.0-24-generic root=UUID=cb520008-8859-45fa-b8dd-d66a667b1be2 ro rootdelay=60 quiet splash initcall_debug
RelatedPackageVersions:
 linux-restricted-modules-3.13.0-24-generic N/A
 linux-backports-modules-3.13.0-24-generic N/A
 linux-firmware 1.127
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 08/20/2013
dmi.bios.vendor: LENOVO
dmi.bios.version: GLET34WW (1.09.T05 )
dmi.board.asset.tag: Not Available
dmi.board.name: 20ANZ03CUS
dmi.board.vendor: LENOVO
dmi.board.version: Not Defined
dmi.chassis.asset.tag: No Asset Information
dmi.chassis.type: 10
dmi.chassis.vendor: LENOVO
dmi.chassis.version: Not Available
dmi.modalias: dmi:bvnLENOVO:bvrGLET34WW(1.09.T05):bd08/20/2013:svnLENOVO:pn20ANZ03CUS:pvrThinkPadT440p:rvnLENOVO:rn20ANZ03CUS:rvrNotDefined:cvnLENOVO:ct10:cvrNotAvailable:
dmi.product.name: 20ANZ03CUS
dmi.product.version: ThinkPad T440p
dmi.sys.vendor: LENOVO

[Brad Figg (brad-figg)]

This change was made by a bot.

[penalvch (penalvch)]

Taihsiang Ho, thank you for reporting this and helping make Ubuntu better. As per http://download.lenovo.com/express/ddfm.html an update to your BIOS is available (2.25). If you update to this following https://help.ubuntu.com/community/BiosUpdate does it change anything? If it doesn't, could you please both specify what happened, and provide the output of the following terminal command:
sudo dmidecode -s bios-version && sudo dmidecode -s bios-release-date

Thank you for your understanding.

[Alex Hung (alexhung)]

@Taihsiang,

Can you please attach fwts's results.log and acpidump from "sudo acpidump > acpi.log"

[Taihsiang Ho (tai271828)]

@ Christopher: Thanks for your comment.
I am certificating the machine so I have to target on the same BIOS.
Thus I am not going to update the BIOS.

Please refer to the attachment dmidecode.log for your request of the comment #3.

[Taihsiang Ho (tai271828)]

@ Alex
Please refer to the attachment 201309-14169-lp1364748.tar.gz for your comment in this thread ( the comment #4)
It includes
a. 01-no-proprietary-driver.log - fwts output with open source driver.
b. 02-nvidia-performance.log - fwts output with performance mode.
c. 03-nvidia-power-saving.log - fwts output with power-saving mode.
d. acpi.log - your suggestion in the comment #4
e. system-conf.log - more information for your reference

[Taihsiang Ho (tai271828)]

201309-14169
3.13.0-39-generic and updated system
ubuntu-drivers-common 1:0.2.91.7

still has this issue.

Please see the attachment
201309-14169-lp1364748-comment07.tar.gz
for more details.

[Alex Hung (alexhung)]

ACPI spec defines _ROM as "This method is used to get a copy of the display devices’ ROM data". In this case, it is used to get VBIOS (ex. VGA's configuration) from ACPI BIOS AML code. As it fails, OS is not able to get such configuration information.

I tried nvidia-prime and switched between nvidia and intel VGA, and the switching did happen according to its system indicator, but I am running glmark2 in both VGA modes in order to see whether the switching is indeed occurred.

[Alex Hung (alexhung)]

There are two signs that shows switching occurred:

1. lspci shows that nvidia geforce 730m's rev becomes 0xFF (ex. unreadable) when using Intel's VGA.
2. glmark2 shows an increase in score (from ~1950 -> ~2015, averages are taken for 4 measurement).

"NVIDIA X Server Settings" also shows GPU frequency and utilization changes with glmark2 running. It seems no problem with hardware switching.

[Alex Hung (alexhung)]

From the error messages itself, it seems that BIOS did not perform boundary check when it tries to refer to a memory address. As a result, Field [TBF3] is out-of-boundary. The disassemabled code is as below:

                    OperationRegion (ATRP, SystemMemory, \ATRB, 0x00010000)
                    Field (ATRP, AnyAcc, Lock, Preserve)
                    {
                        IDX0, 262144,
                        IDX1, 262144
                    }

                    Method (_ROM, 2, Serialized) // _ROM: Read-Only Memory
                    {
                        If (LGreaterEqual (Arg0, 0x8000))
                        {
                            Return (GETB (Subtract (Arg0, 0x8000), Arg1, IDX1))
                        }

                        If (LGreater (Add (Arg0, Arg1), 0x8000))
                        {
                            Subtract (0x8000, Arg0, Local0)
                            Subtract (Arg1, Local0, Local1)
                            Store (GETB (Arg0, Local0, IDX0), Local3)
                            Store (GETB (0x00, Local1, IDX1), Local4)
                            Concatenate (Local3, Local4, Local5)
                            Return (Local5)
                        }

                        Return (GETB (Arg0, Arg1, IDX0))
                    }

                    Method (GETB, 3, Serialized)
                    {
                        Multiply (Arg0, 0x08, Local0)
                        Multiply (Arg1, 0x08, Local1)
                        CreateField (Arg2, Local0, Local1, TBF3)
                        Return (TBF3)
                    }

As I am not able to do a BIOS update (I have the system but not its battery that is required for BIOS update), I am not able to confirm whether new BIOS fixes this problem.

[Alex Hung (alexhung)]

The BIOS updates are available @ http://support.lenovo.com/en/downloads/ds037576.
The system is currently running on 1.09 which is oldest one, and 2.18 added a hybrid graphic fix (see below and http://download.lenovo.com/ibmdl/pub/pc/pccbbs/mobiles/gluj15uc.txt), but again I do not have the required battery pack for updating the BIOS.

<2.18>
 UEFI: 2.18 / ECP: 1.08
- (New) Added support for Hybrid Graphics on Windows 8.1 for NVIDIA graphics
        models.
        Note:
        To support this, uninstall and re-install the NVIDIA graphics driver
        after updating the UEFI BIOS from the version 1.18 or lower.

If someone in community can help comment or upload acpidump file, I can do more analysis. If that's not the case, I would suggest to "won't fix" this bug as it seems not affect any function.

[Anthony Wong (anthonywong)]

@taihsiang
Alex has analyzed this issue and does not think it really affects hybrid graphics. Could you discuss with your team and see if we can close this issue, or should we try to update the BIOS for further investigation?

[Taihsiang Ho (tai271828)]

@Anthony
In this case we are not going to update the BIOS.
According to the comments, please close this bug.
Thanks for your help.