4.3.14 Now segfaults when unplugging USB devices

[DigiAngel]

Topic says it:

Jul 22 05:49:47 kernel: [ 6071.233090] EMT-0[7742]: segfault at 0 ip 00007fa5c35ed7f1 sp 00007fa5f3245bd0 error 6 in VBoxDD.so[7fa5c3541000+254000]

I can pretty much reproduce this. Thoughts on this?

[socratis]

Unfortunately for you, we can't reproduce it (unless we have more details). See: Minimum information needed for assistance.

[RayC]

Same thing here. Started with 4.3.14. 4.3.12 works fine.

Guest - Windows 8.1 64-bit
Host - Linux Mint 17 Quiana

VirtualBox log added, although there is nothing notable around the time of failure in the log.

[1383007.564892] EMT-0[24455]: segfault at 0 ip 00007fbdfd68a7f1 sp 00007fbe2d221b90 error 6 in VBoxDD.so[7fbdfd5de000+254000]

To reproduce:
1) I am attaching an iPhone 5c. I have a USB filter setup for vendor ID 05ac so it automatically is attached to VirtualBox once I plug in the device.
2) Start VM
3) Connect USB device
4) Unplug USB device - segfault occurs and VirtualBox goes bye-bye.

This does not happen to me if I don't have USB 2.0 support enabled...but then again my device won't work without USB 2.0 support enabled either (identify fails if I don't have 2.0 enabled).

[manolo]

Similar problem here.
Ubuntu 14.04 host wuth 4G ram. Windows Xp Service pack 3 guest. This appeared in lastest VB update. I have two serial devices shared with the client. They exchange data with a POS terminal to download apps and monitor logs. Whenever the POS closes the port, the guest vm crashes. A workaround is to detach the device from the menu "Devices->USB Devices" before the POS closes the port.

[ 5424.876067] usb 2-2: new high-speed USB device number 5 using ehci-pci
[ 5425.008736] usb 2-2: New USB device found, idVendor=123a, idProduct=0001
[ 5425.008740] usb 2-2: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[ 5425.008743] usb 2-2: Product: SpectraUSBSerial
[ 5425.008745] usb 2-2: Manufacturer: SPECTRA
[ 5425.398882] cdc_acm 2-2:1.0: This device cannot do calls on its own. It is not a modem.
[ 5425.398996] cdc_acm 2-2:1.0: ttyACM0: USB ACM device
[ 5425.399833] usbcore: registered new interface driver cdc_acm
[ 5425.399837] cdc_acm: USB Abstract Control Model driver for USB modems and ISDN adapters
[ 5487.048059] usb 2-2: USB disconnect, device number 5
[ 5487.182220] EMT[9781]: segfault at 0 ip a1a1942e sp a4064f80 error 6 in VBoxDD.so[a194f000+27f000]
[ 5743.692686] usb 6-1: usbfs: process 30698 (USBDevIo-2) did not claim interface 0 before use
[ 6048.132450] usb 6-1: reset full-speed USB device number 2 using uhci_hcd
[ 6048.276801] pl2303 6-1:1.0: pl2303 converter detected
[ 6048.288090] usb 6-1: pl2303 converter now attached to ttyUSB0

[socratis]

@RayC, @manoio
This seems to be a reproducible case. Similar segfaults when detaching USB devices. I'm not sure if the post by 'DigiAngel' was due to the same conditions, since s/he didn't give any details on how to trigger the segfault.

These kinds of problems that are reproducible should be reported to the bugtracker and either search for an appropriate bug where you'll attach your logs and other vital information, or open a new bug report. It would also help if you can post the bug report number here so other can see the progress, add information, or see the final resolution.

[RayC]

Ticket 13239 has been created.

[mauromol]

I also have this exact same problem on Linux Mint 16 KDE host, Windows XP SP3 guest. I added my comment to ticket 13239.

[lyallp]

I am also experiencing the same issue.

It does not matter what the USB device is, unplugging it causes immedate failure of my Windows 7 64 bit Guest.

No diagnostics appear to be written to the VM Log.

Code: Select all

# VBoxManage list --long hostinfo
Host Information:

Host time: 2014-08-22T05:07:05.158000000Z
Processor online count: 4
Processor count: 4
Processor online core count: 2
Processor core count: 2
Processor#0 speed: 2667 MHz
Processor#0 description: Intel(R) Core(TM) i7 CPU       M 620  @ 2.67GHz
Processor#1 speed: 2667 MHz
Processor#1 description: Intel(R) Core(TM) i7 CPU       M 620  @ 2.67GHz
Processor#2 speed: 2667 MHz
Processor#2 description: Intel(R) Core(TM) i7 CPU       M 620  @ 2.67GHz
Processor#3 speed: 2667 MHz
Processor#3 description: Intel(R) Core(TM) i7 CPU       M 620  @ 2.67GHz
Memory size: 7840 MByte
Memory available: 3316 MByte
Operating system: Linux
Operating system version: 3.14.14-gentoo

Code: Select all

$ VBoxManage  list --long vms
Name:            HP COE Win7 x64
Groups:          /
Guest OS:        Windows 7 (64 bit)
UUID:            3cdbc5d3-c526-493f-9671-4d6d98b866ee
Config file:     /home/pearcely/vms/VM_Settings/Machines/HP COE Win7 x64/HP COE Win7 x64.vbox
Snapshot folder: /home/pearcely/vms/VM_Settings/Machines/HP COE Win7 x64/Snapshots
Log folder:      /home/pearcely/vms/VM_Settings/Machines/HP COE Win7 x64/Logs
Hardware UUID:   3cdbc5d3-c526-493f-9671-4d6d98b866ee
Memory size:     3500MB
Page Fusion:     off
VRAM size:       64MB
CPU exec cap:    90%
HPET:            off
Chipset:         piix3
Firmware:        BIOS
Number of CPUs:  2
PAE:             on
Long Mode:       on
Synthetic CPU:   off
CPUID overrides: None
Boot menu mode:  message and menu
Boot Device (1): DVD
Boot Device (2): HardDisk
Boot Device (3): Not Assigned
Boot Device (4): Not Assigned
ACPI:            on
IOAPIC:          on
Time offset:     0ms
RTC:             UTC
Hardw. virt.ext: on
Nested Paging:   on
Large Pages:     off
VT-x VPID:       on
VT-x unr. exec.: on
State:           running (since 2014-08-22T04:51:05.584000000)
Monitor count:   2
3D Acceleration: off
2D Video Acceleration: off
Teleporter Enabled: off
Teleporter Port: 0
Teleporter Address: 
Teleporter Password: 
Tracing Enabled: off
Allow Tracing to Access VM: off
Tracing Configuration: 
Autostart Enabled: off
Autostart Delay: 0
Default Frontend: 
Storage Controller Name (0):            IDE Controller
Storage Controller Type (0):            PIIX4
Storage Controller Instance Number (0): 0
Storage Controller Max Port Count (0):  2
Storage Controller Port Count (0):      2
Storage Controller Bootable (0):        on
Storage Controller Name (1):            SATA Controller
Storage Controller Type (1):            IntelAhci

[socratis]

@lyallp
You seem to be using the Gentoo version of VirtualBox, not the plain vanilla that is "supported" in these forums.

[lyallp]

@lyallp
You seem to be using the Gentoo version of VirtualBox, not the plain vanilla that is "supported" in these forums.

Regardless, I am affected in the same way others are, and have provided more info than most.

[Perryg]

@lyallp
You seem to be using the Gentoo version of VirtualBox, not the plain vanilla that is "supported" in these forums.

Regardless, I am affected in the same way others are, and have provided more info than most.

At this point while the information was in fact something that we normally have to beg for and you did provide it ( thank you very much ) it would be better to go ahead and continue this at
https://www.virtualbox.org/ticket/13239 Be sure to provide information there as well as they do not like to bounce all over the place to find it.

[socratis]

lyallp, I simply wanted to make sure that you have the same issue with the official release, and not just a distro-fork. It doesn't mean it's not the same problem, it means the devs can't know for sure unless they take that parameter away. As Perryg said, thank you and yes, that's exactly the info the devs are looking for, but (and that is a big but), I know I wouldn't take the reports at face value if they were coming from a forked version, which I had no control over. That's why the "supported" had quotes. I've learned from the experts around here to dismiss pretty much everything that is not from the official sources.

[lyallp]

I thought I would clarify what I consider to be a misunderstanding on what gentoo does.
Gentoo has not forked virtualbox.
In general, Gentoo does not fork, what it does do, on occasion, is to supply a patch or two to the base source, maybe to fix a known problem, more generally to enable the build to operate successfully in the gentoo build environment.
Gentoo builds virtualbox from the OSE source, downloaded, in my case from a mirror at http://mirror.internode.on.net/pub/gent ... 12.tar.bz2
As such, gentoo should be the ideal system to obtain bug reports, as I can turn on full debug (-gdb) as a part of the build. Curiously enough, I only just disabled that on my home system as debug symbols took in excess of 20G for the entire system (everything from glibc upwards libraries upward had debug symbols).

I don't know why I could not find that bug in the tracker, as I created ticket #13309. I suspect they should be merged.