Bug#1050352: linux-image-amd64: backside USB-ports stop working after some time (mouse, keyboard), VIA Technologies, Inc. VL805/806 xHCI USB 3.0 Controller (rev 01)

[Rolf Reintjes]

Package: linux-image-amd64
Version: 6.1.38-4
Severity: normal

Dear Maintainer,

*** Reporter, please consider answering these questions, where appropriate ***

* What led up to the situation?
I start the system and it runs for some time. After some time the backside USB-ports stop working.

* What exactly did you do (or not do) that was effective (or
ineffective)?
Use the mouse and keyboard for some time. Then after a while they stop working.

In dmesg there is this error:
xhci_hcd 0000:05:00.0: WARNING: Host System Error
which is this from lspci:
05:00.0 USB controller: VIA Technologies, Inc. VL805/806 xHCI USB 3.0 Controller (rev 01)

As a workaround I got the system running normal with
just disable iommu in bios and add "iommu=soft" in grub commandline
I have an Intel i7-5820K CPU and in Bios this is the VT-D in overclocking section.

My mainboard is MSI X99A SLI PLUS.

With bullseye I did not have this problems.

* What was the outcome of this action?
Without the workaround Mouse and keyboard of backside USB-ports stop working.
With this workaround the work normal.

* What outcome did you expect instead?
Mouse and keyboard of backside USB-ports work for complete uptime of the system.

*** End of the template - remove these template lines ***


-- System Information:
Debian Release: 12.1
APT prefers stable-updates
APT policy: (500, 'stable-updates'), (500, 'stable-security'), (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 6.1.0-11-amd64 (SMP w/12 CPU threads; PREEMPT)
Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled

Versions of packages linux-image-amd64 depends on:
ii linux-image-6.1.0-11-amd64 6.1.38-4

linux-image-amd64 recommends no packages.

linux-image-amd64 suggests no packages.

-- no debconf information

[Rolf Reintjes]

I have done some tests today to look if this problem does also occur
with other distributions and kernels:

Via Technologies VL805/806 xHCI USB 3.0 Controller working or not
working depended on Linux Kernel version

Distribution Kernel works or not
Debian 11.x bullseye 5.? yes
Debian 12.1 bookworm 6.1.0-10-amd64 no
CentOS-Stream 9 5.14.0-354.el9.x86_64 yes
Manjaro 22.1.3 20230529 6.1.30-1-MANJARO yes
Fedora-Workstation 38-1.6 6.2.9-300.fc38.x86_64 yes
Ubuntu 23.04 6.2.0-20-generic yes

I also tried to install testing. But it did not install grub and could
not solve this with rescue-mode. But during testing installation the
backside USB-port mouse and keyboard also stopped working after some
time. I continued with a second mouse and a second keyboard on front
USB-ports.

Best regards

Rolf Reintjes

[Rolf Reintjes]

Here are more of the near dmesg meassages:


[   28.747758] IPv6: ADDRCONF(NETDEV_CHANGE): eno1: link becomes ready
[  344.813681] DMAR: DRHD: handling fault status reg 2
[  344.813692] DMAR: [DMA Read NO_PASID] Request device [05:00.0] fault
addr 0xfffad000 [fault reason 0x06] PTE Read access is not set
[  344.813917] DMAR: DRHD: handling fault status reg 102
[  344.813923] xhci_hcd 0000:05:00.0: WARNING: Host System Error
[  344.813923] DMAR: [DMA Read NO_PASID] Request device [05:00.0] fault
addr 0xfffad000 [fault reason 0x06] PTE Read access is not set

[Diederik de Haas]

Control: reassign -1 src:linux/6.1.38-4

A page fault doesn't sound good and I wouldn't be surprised if that wes the
reason for the issues.
On https://snapshot.debian.org/binary/linux-image-amd64/ you'll find a whole
bunch of kernel version and it would be really helpful if you can determine
what the last version of the 6.1 series was where the USB ports work as
expected and then the first version where it stopped working.

You could start with (one of) the earliest 6.1 kernels to verify whether it
has worked properly in a 6.1 kernel at all.

[Rolf Reintjes]

I tried some older kernels. They all have the problem.

linux-image-6.1.0-0-amd64_6.1.1-1~exp2_amd64.deb no
linux-image-6.0.0-6-amd64_6.0.12-1_amd64.deb no
linux-image-6.0.0-1-amd64_6.0.2-1_amd64.deb no

[Rolf Reintjes]

I have just installed bullseye on another disk. It has the same problem.

I do not know why I did not run in this problem when bullseye was stable
and I used it.

[Rolf Reintjes]

On another disk I installed bookworm and upgraded to testing. With
testing the problem is also there.

[Rolf Reintjes]

rolf@i7-5820K-debian:~$ sudo lspci -s 05:00.0 -vv
[sudo] Passwort für rolf:

05:00.0 USB controller: VIA Technologies, Inc. VL805/806 xHCI USB 3.0

Controller (rev 01) (prog-if 30 [XHCI])
Subsystem: Micro-Star International Co., Ltd. [MSI] VL805/806 xHCI USB
3.0 Controller
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B- DisINTx+
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort-
<MAbort- >SERR- <PERR- INTx-
Latency: 0, Cache Line Size: 64 bytes
Interrupt: pin A routed to IRQ 34
NUMA node: 0
IOMMU group: 32
Region 0: Memory at fb200000 (64-bit, non-prefetchable) [size=4K]
Capabilities: [80] Power Management version 3
Flags: PMEClk- DSI- D1- D2- AuxCurrent=375mA
PME(D0+,D1+,D2+,D3hot+,D3cold+)
Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [90] MSI: Enable+ Count=1/4 Maskable- 64bit+
Address: 00000000fee003d8 Data: 0000
Capabilities: [c4] Express (v2) Endpoint, MSI 00
DevCap: MaxPayload 256 bytes, PhantFunc 0, Latency L0s <64ns, L1 <1us
ExtTag- AttnBtn- AttnInd- PwrInd- RBE+ FLReset- SlotPowerLimit 89W
DevCtl: CorrErr- NonFatalErr- FatalErr- UnsupReq-
RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+
MaxPayload 128 bytes, MaxReadReq 512 bytes
DevSta: CorrErr+ NonFatalErr- FatalErr- UnsupReq+ AuxPwr+ TransPend-
LnkCap: Port #0, Speed 5GT/s, Width x1, ASPM L0s L1, Exit Latency L0s
<2us, L1 <16us
ClockPM+ Surprise- LLActRep- BwNot- ASPMOptComp-
LnkCtl: ASPM Disabled; RCB 64 bytes, Disabled- CommClk+
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed 5GT/s, Width x1
TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
DevCap2: Completion Timeout: Range B, TimeoutDis+ NROPrPrP- LTR-
10BitTagComp- 10BitTagReq- OBFF Not Supported, ExtFmt- EETLPPrefix-
EmergencyPowerReduction Not Supported, EmergencyPowerReductionInit-
FRS- TPHComp- ExtTPHComp-
AtomicOpsCap: 32bit- 64bit- 128bitCAS-
DevCtl2: Completion Timeout: 50us to 50ms, TimeoutDis- LTR-
10BitTagReq- OBFF Disabled,
AtomicOpsCtl: ReqEn-
LnkCtl2: Target Link Speed: 5GT/s, EnterCompliance- SpeedDis-
Transmit Margin: Normal Operating Range, EnterModifiedCompliance-
ComplianceSOS-
Compliance Preset/De-emphasis: -6dB de-emphasis, 0dB preshoot
LnkSta2: Current De-emphasis Level: -6dB, EqualizationComplete-
EqualizationPhase1-
EqualizationPhase2- EqualizationPhase3- LinkEqualizationRequest-
Retimer- 2Retimers- CrosslinkRes: unsupported
Capabilities: [100 v1] Advanced Error Reporting
UESta: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- RxOF-
MalfTLP- ECRC- UnsupReq+ ACSViol-
UEMsk: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- RxOF-
MalfTLP- ECRC- UnsupReq- ACSViol-
UESvrt: DLP+ SDES+ TLP- FCP+ CmpltTO- CmpltAbrt- UnxCmplt- RxOF+
MalfTLP+ ECRC- UnsupReq- ACSViol-
CESta: RxErr- BadTLP- BadDLLP- Rollover- Timeout- AdvNonFatalErr+
CEMsk: RxErr- BadTLP- BadDLLP- Rollover- Timeout- AdvNonFatalErr+
AERCap: First Error Pointer: 14, ECRCGenCap- ECRCGenEn- ECRCChkCap-
ECRCChkEn-
MultHdrRecCap- MultHdrRecEn- TLPPfxPres- HdrLogCap-
HeaderLog: 00000000 05070000 0000000f 04000001
Kernel driver in use: xhci_hcd
Kernel modules: xhci_pci

[Rolf Reintjes]

I compiled and installed a kernel with

https://github.com/torvalds/linux/blob/master/drivers/iommu/intel/iommu.c

In this file I had to comment out:

/* case IOMMU_CAP_DEFERRED_FLUSH: */

With this self compiled kernel the problem is not there.

I guess that iommu.c patches in

https://sources.debian.org/patches/linux/6.4.11-1/

cause the problems I discovered.

How can I help further to solve the problem?

What to do next?

[Rolf Reintjes]

I am on testing.

On 28.08.23 10:35, Rolf Reintjes wrote:
> I compiled and installed a kernel with
>
> https://github.com/torvalds/linux/blob/master/drivers/iommu/intel/iommu.c
>
> In this file I had to comment out:
>
> /* case IOMMU_CAP_DEFERRED_FLUSH:  */
>
> With this self compiled kernel the problem is not there.

This file was from 6.5.

Now I took this one:

https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/drivers/iommu/intel/iommu.c?h=v6.4.11

and compiled and installed the kernel. The problem did not occur. The
debian patches were not applied here for my test.

> I guess that iommu.c patches in
>
> https://sources.debian.org/patches/linux/6.4.11-1/
>
> cause the problems I discovered.

I would appreciate further advice and guidance.

[Rolf Reintjes]

I am on testing.

I could isolate the problem causing code in the debian patches on file

drivers/iommu/intel/iommu.c

With this change

rolf@i7-5820K-debian-testing:~/kernel/linux-source-6.4/drivers/iommu/intel$
diff iommu.c.debian iommu.c
286c286
< int dmar_disabled = IS_ENABLED(CONFIG_INTEL_IOMMU_DEFAULT_OFF);
---
> int dmar_disabled = !IS_ENABLED(CONFIG_INTEL_IOMMU_DEFAULT_ON);

[Diederik de Haas]

Excellent, thanks for the analyses.

On Monday, 28 August 2023 16:46:02 CEST Rolf Reintjes wrote:
> I would appreciate further advice and guidance.

Have patience. When someone with the needed knowledge has time to look into
this issue, they will.
But we can't make claims on how people spend their free time.