Solaris 11.1 host cannot connect USB to guests

[bq]

I know this is a duplicate of #11591, but that ticket has incomplete information I cannot edit.

On Solaris 11.1 hosts, USB cannot be successfully connected to any type of guest (tested XP, win7, win8, Ubuntu 11-12.x, Solaris 11.x guests). The messages file on the host always returns the entry below after connecting the USB device, whether a printer, storage, or USB WiFi adapter. All tested devices I know work with VB guests of Solaris 11.0 hosts.

Also, one host (Dell Optiplex 990) crashes every time I connect the USB to the guest, but not before. Other hosts do not crash, but still return the same error in the messages file. This does not happen on Solaris 11.0. Defect exists with all SRUs for Solaris 11.1 including latests SRU 6.4.

Messages file error:

Apr 23 13:24:37 wopr vboxdrv: [ID 275565 kern.notice] vboxusb:vboxUSBSolarisBulkXfer usb_pipe_bulk_xfer enmDir=0x2 Ep=0xd failed! rc=-6

Attaching compressed vbox log file from host that crashes when USB connected to a guest.

[bq]

Win 7 64 bit guest, attaching USB WiFi adapter. Gzip compressed.

[bq]

Attaching a log file from a different computer running the same host OS. This is for an Ubuntu 12.10 guest, attempting to attach a flash drive to the guest. Host can read & write the drive without complaint. This host does not crash when the USB device is captured. It is an Intel DQ77MK motherboard. Note it takes about 1 minute for the failure to occur as seen in /var/adm/messages:

Apr 23 22:36:38 fnog vboxdrv: [ID 555791 kern.notice] vboxusbmon: Capturing U3 Cruzer Micro 0x781:0x5406:/pci@0,0/pci8086,2035@14/storage

Apr 23 22:36:38 fnog vboxdrv: [ID 833392 kern.notice] vboxusb: Captured U3 Cruzer Micro 0x781:0x5406:512:/devices/pci@0,0/pci8086,2035@14/storage@3:vboxusb

Apr 23 22:37:41 fnog vboxdrv: [ID 861514 kern.notice] vboxusb:vboxUSBSolarisBulkXfer usb_pipe_bulk_xfer enmDir=0x2 Ep=0x2 failed! rc=-6

[bq]

Ubuntu 12.10 guest on Solaris 11.1 SRU 6.4 host. Attaching a USB flash drive.

[bq]

USB problem persists with SRU 8.4. Also tested with VB 4.1.26 with the same errors. See also the user forum for more reports:

​https://forums.virtualbox.org/viewtopic.php?f=11&t=55816

​https://forums.virtualbox.org/viewtopic.php?f=11&t=47304

​https://forums.virtualbox.org/viewtopic.php?f=11&t=53726

[lb]

I've had the same types of issues throughout the Oracle Solaris 11 releases - initial onward.

Lenovo T510 with Core i7 920M (1st gen) Lenovo T530 with Core i7 i7-3840QM (3rd gen)

The Lenovo T510 chipset has the bad VT-d, which has to be disabled. The T530 VT-d is enabled without error.

Both systems running Windows XP (32 Bit) VMs and Windows 7 (64 bit) VMs have the same problems with USB devices coming and going (look in device manager - it keeps refreshing with device(s) gone, then back) - and with the bulk-xfer messages.

[axlc]

My problem is similar except:

• I'm using VirtualBox 4.2.16
• I don't get that error:

Apr 23 22:37:41 fnog vboxdrv: [ID 861514 kern.notice] vboxusb:vboxUSBSolarisBulkXfer usb_pipe_bulk_xfer enmDir=0x2 Ep=0x2 failed! rc=-6

[Ramshankar Venkataraman]

Could you please try this build and see if it fixes anything? There were 2 fixes made to the VBoxUSB code on Solaris that are included here: ​http://www.virtualbox.org/download/testcase/VirtualBox-4.2.51-87810-SunOS.tar.gz (This link will expire in 14 days).

WARNING! This is a bleeding edge trunk build (not a 4.2.x maintenance release). You should not try this on production systems or where stability of the system is critical.

[Ramshankar Venkataraman]

Here is the extpack: ​http://www.virtualbox.org/download/testcase/Oracle_VM_VirtualBox_Extension_Pack-4.2.51-87722.vbox-extpack (link expires in 14 days)

[bq]

Sorry for the delay in responding. Didn't see the message right away.

First, I should mention that I noticed with 4.2.16 on a Windows 8 guest that the guest saw the USB storage I connected & loaded the driver in the first minute before the host logged the error at the beginning of this bug report. Then after the first minute, the usual results.

Now with this test build, the host no longer logs an error, but now the guest does.

With an Ubuntu 12.10 guest, the guest's syslog contains the complaint (vbox logs will also be attached, hostname is ub1204, not to be confusing):

Aug 25 23:50:11 ub1204 kernel: [   97.864265] usb 1-1: new high-speed USB device number 2 using ehci_hcd
Aug 25 23:50:11 ub1204 mtp-probe: checking bus 1, device 2: "/sys/devices/pci0000:00/0000:00:0b.0/usb1/1-1"
Aug 25 23:50:11 ub1204 mtp-probe: bus: 1, device: 2 was not an MTP device
Aug 25 23:50:11 ub1204 kernel: [   98.076873] usb 1-1: New USB device found, idVendor=0781, idProduct=5406
Aug 25 23:50:11 ub1204 kernel: [   98.076877] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
Aug 25 23:50:11 ub1204 kernel: [   98.076880] usb 1-1: Product: U3 Cruzer Micro
Aug 25 23:50:11 ub1204 kernel: [   98.076883] usb 1-1: Manufacturer: SanDisk Corporation
Aug 25 23:50:11 ub1204 kernel: [   98.076885] usb 1-1: SerialNumber: 0000184CF4736ABE
Aug 25 23:50:11 ub1204 kernel: [   98.133543] Initializing USB Mass Storage driver...
Aug 25 23:50:11 ub1204 kernel: [   98.133864] scsi6 : usb-storage 1-1:1.0
Aug 25 23:50:11 ub1204 kernel: [   98.133918] usbcore: registered new interface driver usb-storage
Aug 25 23:50:11 ub1204 kernel: [   98.133920] USB Mass Storage support registered.
Aug 25 23:50:33 ub1204 kernel: [  120.072109] usb 1-1: reset high-speed USB device number 2 using ehci_hcd
Aug 25 23:50:43 ub1204 kernel: [  130.392169] usb 1-1: reset high-speed USB device number 2 using ehci_hcd
Aug 25 23:51:00 ub1204 kernel: [  146.752157] usb 1-1: reset high-speed USB device number 2 using ehci_hcd
Aug 25 23:51:00 ub1204 kernel: [  147.084185] usb 1-1: reset high-speed USB device number 2 using ehci_hcd
Aug 25 23:51:10 ub1204 kernel: [  157.444224] usb 1-1: reset high-speed USB device number 2 using ehci_hcd
Aug 25 23:51:10 ub1204 kernel: [  157.669277] scsi 6:0:0:0: Device offlined - not ready after error recovery

On a Solaris 11.1 guest, similar results. Here is the output from /var/adm/messages, and I don't know if this is all related to the USB problem. No obvious error, but the device didn't mount:

Aug 26 00:00:05 s11-1 usba: [ID 912658 kern.info] USB 2.0 device (usb781,5406) operating at hi speed (USB 2.x) on USB 2.0 root hub: storage@1, scsa2usb0 at bus address 2
Aug 26 00:00:05 s11-1 usba: [ID 349649 kern.info] 	SanDisk Corporation U3 Cruzer Micro 0000184CF4736ABE
Aug 26 00:00:05 s11-1 genunix: [ID 936769 kern.info] scsa2usb0 is /pci@0,0/pci8086,265c@1f,5/storage@1
Aug 26 00:00:05 s11-1 genunix: [ID 408114 kern.info] /pci@0,0/pci8086,265c@1f,5/storage@1 (scsa2usb0) online
Aug 26 00:00:17 s11-1 scsi: [ID 583861 kern.info] sd2 at scsa2usb0: target 0 lun 0
Aug 26 00:00:17 s11-1 genunix: [ID 936769 kern.info] sd2 is /pci@0,0/pci8086,265c@1f,5/storage@1/disk@0,0
Aug 26 00:02:11 s11-1 genunix: [ID 408114 kern.info] /pci@0,0/pci8086,265c@1f,5/storage@1/disk@0,0 (sd2) online
Aug 26 00:02:11 s11-1 vga_arbiter: [ID 197929 kern.info] vga_arbiter: device added: PCI:0:0:2.0,decodes=io+mem+IO+MEM,owns=io+mem+IO+MEM,legalocks=none,normlocks=none
Aug 26 00:02:11 s11-1 vga_arbiter: [ID 254375 kern.info] 1 vga device(s) found
Aug 26 00:02:11 s11-1 rootnex: [ID 349649 kern.info] vga_arbiter0 at root
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] vga_arbiter0 is /vga_arbiter
Aug 26 00:02:11 s11-1 pcplusmp: [ID 805372 kern.info] pcplusmp: lp (ecpp) instance 0 irq 0x7 vector 0x40 ioapic 0x2 intin 0x7 is bound to cpu 0
Aug 26 00:02:11 s11-1 isa: [ID 202937 kern.info] ISA-device: ecpp0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] ecpp0 is /pci@0,0/isa@1f/lp@1,378
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: dcpc0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] dcpc0 is /pseudo/dcpc@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: fbt0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] fbt0 is /pseudo/fbt@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: fcp0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] fcp0 is /pseudo/fcp@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: fcsm0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] fcsm0 is /pseudo/fcsm@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: llc10
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] llc10 is /pseudo/llc1@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: lockstat0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] lockstat0 is /pseudo/lockstat@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: lofi0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] lofi0 is /pseudo/lofi@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: profile0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] profile0 is /pseudo/profile@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: ramdisk1024
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] ramdisk1024 is /pseudo/ramdisk@1024
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: sdt0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] sdt0 is /pseudo/sdt@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: systrace0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] systrace0 is /pseudo/systrace@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: ucode0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] ucode0 is /pseudo/ucode@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: bpf0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] bpf0 is /pseudo/bpf@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: fssnap0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] fssnap0 is /pseudo/fssnap@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: nsmb0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] nsmb0 is /pseudo/nsmb@0
Aug 26 00:02:11 s11-1 pseudo: [ID 129642 kern.info] pseudo-device: winlock0
Aug 26 00:02:11 s11-1 genunix: [ID 936769 kern.info] winlock0 is /pseudo/winlock@0

And then when I removed it from the guest:

Aug 26 00:05:57 s11-1 genunix: [ID 408114 kern.info] /pci@0,0/pci8086,265c@1f,5/storage@1/disk@0,0 (sd2) removed
Aug 26 00:05:57 s11-1 genunix: [ID 408114 kern.info] /pci@0,0/pci8086,265c@1f,5/storage@1 (scsa2usb0) removed

And finally on a Windows 8 Pro guest, I couldn't find the right log entry in the Event Viewer. This guest continually froze & unfroze while it was trying to attach the USB device. The other guests paused only once & recovered.

Possibly a good start. I'll try again if you have another. Thanks!

[bq]

Ubuntu 12.10 guest on VB test build 4.2.51.r87810

[bq]

Solaris 11.1 guest on VB test build 4.2.51.r87810

[bq]

Windows 8 Pro guest on VB test build 4.2.51.r87810

[bq]

Tried the 4.3 beta 1, and the results are marginally better. A USB WiFi device connected to a Windows 7 guest is seen by the guest & shows up in the network adapters page, but it can't detect any wireless networks. A storage device doesn't connect, but it tries & doesn't lock up the guest.

For UNIX guests (Solaris won't boot yet on 4.3 beta1), I see the same errors in the guest as most recently posted.

[Srdan]

Tried on Solaris 11.1 "entire 0.5.11-0.175.1.19.0.6.0" with 4.3.12 and it is still the same, USB is not mounted in the guest ( WinXP ). I tested with USB flash disk and printer HP Photosmart. Log is in attachment. I'll be glad to test any latest bits that would help resolve this issue https://www.virtualbox.org/attachment/ticket/11747/VBoxSVC.log

[bq]

VB 4.3.14 still has not resolved this issue, and the vboxusb built by Ram that I've been using is unreliable on Solaris 11.2 Is Oracle going to move these builds to a 64-bit host yet? We're now two years into Solaris 64-bit only versions already!

Ram, might you supply another 64-bit build of the USB drivers for me? I've been using the originals quite successfully until I upgraded to Solaris 11.2 Now they're unreliable.

[aeichner]

Please reopen if still relevant with a recent VirtualBox release.