[SOLVED] xhci-hcd / USB3 issues with new motherboard

coralfang · 09-04-2015, 12:54 PM

Hi. Using -current on x84_64.

Few days ago, my motherboard died (a resistor blew out), and i replaced it this morning. The new motherboard is an "MSI 970 gaming"

When starting up the system with the generic kernel

Code:

$ uname -a
Linux darkstar 4.1.6 #1 SMP Fri Aug 21 01:21:18 CDT 2015 x86_64 AMD FX(tm)-8320 Eight-Core Processor AuthenticAMD GNU/Linux

I get spammed with an IO_PAGE_FAULT message after some xHCI host controllers are detected:

Code:

[    7.712830] xhci_hcd 0000:04:00.0: xHCI Host Controller
[    7.712873] xhci_hcd 0000:04:00.0: new USB bus registered, assigned bus number 7
[    7.712905] usb usb7: We don't know the algorithms for LPM for this host, disabling LPM.
[    7.712961] usb usb7: New USB device found, idVendor=1d6b, idProduct=0003
[    7.712963] usb usb7: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[    7.712965] usb usb7: Product: xHCI Host Controller
[    7.712966] usb usb7: Manufacturer: Linux 4.1.6 xhci-hcd
[    7.712967] usb usb7: SerialNumber: 0000:04:00.0
[    7.713025] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f000 flags=0x0030]
[    7.713091] hub 7-0:1.0: USB hub found
[    7.713104] hub 7-0:1.0: 4 ports detected
[    7.713178] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f040 flags=0x0030]
[    7.713317] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f080 flags=0x0030]
[    7.713486] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f0c0 flags=0x0030]
[    7.713629] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f100 flags=0x0030]
[    7.713770] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f140 flags=0x0030]
[    7.713911] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f180 flags=0x0030]
[    7.714053] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f1c0 flags=0x0030]
[    7.714195] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f200 flags=0x0030]
[    7.714336] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f240 flags=0x0030]
[    7.714477] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f280 flags=0x0030]
[    7.714617] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f2c0 flags=0x0030]
[    7.714787] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f300 flags=0x0030]
[    7.714927] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f340 flags=0x0030]
[    7.715067] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f380 flags=0x0030]
[    7.715207] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f3c0 flags=0x0030]
[    7.715346] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f400 flags=0x0030]
[    7.715483] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f440 flags=0x0030]
[    7.715622] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f480 flags=0x0030]
[    7.715760] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f4c0 flags=0x0030]
[    7.715927] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f500 flags=0x0030]
[    7.716065] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f540 flags=0x0030]
[    7.716205] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f580 flags=0x0030]
[    7.716343] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f5c0 flags=0x0030]
[    7.716481] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f600 flags=0x0030]
[    7.716618] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f640 flags=0x0030]
[    7.716756] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f680 flags=0x0030]
[    7.716918] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0015 address=0x000000010000f6c0 flags=0x0030]

I get these errors for 2 devices (02:00.0 and 04:00.0), from lspci output these are detected as:

Code:

02:00.0 USB controller: VIA Technologies, Inc. Device 3483 (rev 01)
04:00.0 USB controller: VIA Technologies, Inc. Device 3483 (rev 01)

Further down it has some failure messages about xhci_hcd:

Code:

[    9.130670] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  352.30  Tue Jul 21 18:53:45 PDT 2015
[    9.843073] Adding 7815616k swap on /dev/sda2.  Priority:-1 extents:1 across:7815616k 
[    9.913032] fuse init (API version 7.23)
[   10.885042] EXT4-fs (sda1): re-mounted. Opts: (null)
[   11.254619] ppdev: user-space parallel port driver
[   11.256521] lp: driver loaded but no devices found
[   21.244119] xhci_hcd 0000:04:00.0: Stopped the command ring failed, maybe the host is dead
[   21.272070] xhci_hcd 0000:04:00.0: Host not halted after 16000 microseconds.
[   21.272072] xhci_hcd 0000:04:00.0: Abort command ring failed
[   21.273758] xhci_hcd 0000:04:00.0: HC died; cleaning up                                                                                                                                     
[   21.275807] xhci_hcd 0000:04:00.0: Timeout while waiting for setup device command
[   21.275811] usb 6-1: hub failed to enable device, error -62
[   21.277655] usb usb6-port1: couldn't allocate usb_device                                                                                                                                    
[   21.448218] xhci_hcd 0000:02:00.0: Stopped the command ring failed, maybe the host is dead
[   21.477235] xhci_hcd 0000:02:00.0: Host not halted after 16000 microseconds.
[   21.477237] xhci_hcd 0000:02:00.0: Abort command ring failed
[   21.479226] xhci_hcd 0000:02:00.0: HC died; cleaning up                                                                                                                                     
[   21.481621] xhci_hcd 0000:02:00.0: Timeout while waiting for setup device command
[   21.481624] usb 4-1: hub failed to enable device, error -62
[   21.483687] usb usb4-port1: couldn't allocate usb_device                                                                                                                                    
[   25.769279] ip_tables: (C) 2000-2006 Netfilter Core Team
[   25.895221] nf_conntrack version 0.5.0 (65536 buckets, 262144 max)
[   26.942276] NET: Registered protocol family 10

The system works fine OTHER than USB3 ports. How can i resolve this?
All standard USB2 ports function correctly, and power can be drawn from the USB3 ports, although external drives are not detected when attached to a USB3 port. This is the case for both front panel and rear USB3 ports.

I previously had an "Asus M5A78L-M/Usb3" motherboard, which worked out of the box for USB3 support.

Possibly my new motherboard requires an extra kernel option to be enabled?
If anyone knows a solution, it would be much appreciated. Thanks.

coralfang · 09-04-2015, 01:17 PM

Just read about IOMMU being an issue with certain AMD chipsets, this seems to have solved it, all USB3 ports are working fine now.

SOLUTION:

I had to edit the bootloader options to add iommu=soft.

-- Step 1. (add iommu=soft to boot options)
In the case of /etc/lilo.conf on slackware;

Code:

append=" vt.default_utf8=1 resume=/dev/sda2 quiet iommu=soft"

-- Step 2. (update bootloader)

Code:

# /sbin/lilo

-- Step 3. (reboot)

-- Step 4. (enter bios)
Disable IOMMU. HEad to overclocking menu, and enter CPU features option to find it.
Save changes.

-- Step 5. (reboot)
USB3 should now be working on this board.

Hope this is useful to anyone else having the same issue.

273 · 05-06-2018, 09:10 AM

Many thanks, coralfang, I actually googled this error and am not surprised that the only answer I found was here.
I'm on Debian and some of the USB3 ports stopped working (need to find out why but this will hold me over) so I followed the guidance, adding the line about acpi to GRUB instead of LiLo, and they are now working.