Re: [Qemu-devel] Revert commit 5af35d7feccaa7d26b72c6c3d14116421d736b36 - "usb-host-libusb: Fix reset handling"

[Hans de Goede]

Hi,

On 10-02-15 22:02, Dennis Ostermann wrote:
> Hello Hans,
>
> thanks for taking care.
>
>
> 09-02-15 09:09, Hans de Goede wrote:
> > Hi,
> >
> > On 09-02-15 22:09, Dennis Ostermann wrote:
> > > Hi there,
> > >
> > > please revert commit 5af35d7feccaa7d26b72c6c3d14116421d736b36 - "usb-host-libusb: 
> > > Fix reset handling"
> > >
> > > This breaks usb pass through of FTDI based usb devices:
> > >
> > > On the host:
> > >
> > > lsusb | grep FT2232
> > > Bus 003 Device 008: ID 0403:6010 Future Technology Devices International, Ltd 
> > > FT2232C Dual USB-UART/FIFO IC
> > >
> > >
> > > ~/qemu-install/bin$ sudo ./qemu-system-x86_64 -monitor 
> > > telnet:127.0.0.1:1234,server,nowait -hda /dev/sdd2 -redir tcp:20022::22 
> > > --enable-kvm -cpu host -smp 4 -vga vmware --vnc :0 -m 8192  -usb -device 
> > > usb-ehci,id=ehci -device usb-host,bus=ehci.0,vendorid=0x0403,productid=0x6010
> > > WARNING: Image format was not specified for '/dev/sdd2' and probing guessed raw.
> > >            Automatically detecting the format is dangerous for raw images, 
> > > write operations on block 0 will be restricted.
> > >            Specify the 'raw' format explicitly to remove the restrictions.
> > > libusbx: error [_get_usbfs_fd] libusbx couldn't open USB device 
> > > /dev/bus/usb/003/005: No such file or directory
> > > libusbx: error [_get_usbfs_fd] libusbx couldn't open USB device 
> > > /dev/bus/usb/003/006: No such file or directory
> > > libusbx: error [_get_usbfs_fd] libusbx couldn't open USB device 
> > > /dev/bus/usb/003/007: No such device
> > >
> > > The device gets reset again and again and is re-enumerated every time and 
> > > finally not passed through.
> > This looks like the device drops of the bus when it is reset, that is not 
> > normal behavior,
> > there seems to be something unique to your setup causing this. Have you tried 
> > this on
> > multiple machines / different usb ports on your pc ? This may be something 
> > weird with the usb
> > controller in your machine.
> The machine uses Intel H97 chipset, so not that unique. I tried every port, 
> every USB BIOS, with and without hub, it doesn't make any difference. But 
> you're right, I also tried it on a T61 with Intel 900 series chipset and it 
> worked alright. But this is an 'old' USB 2.0 chipset. I'll try on another box 
> with USB 3.0 chipset tomorrow, if I can get one.

Ok, lets wait and see what results another model usb-3 capable pc / laptop gets 
us, I've
the feeling that this is a machine specific issue. Normally a usb-reset should 
not cause
a device disconnect as you're seeing here.

Regards,

Hans