Re: [Qemu-devel] [PATCH v3] libxl: usb2 and usb3 controller support for upstream qemu

["Andreas Färber" <afaerber@suse.de>]

Am 12.07.2013 13:06, schrieb George Dunlap:
> On 12/07/13 11:22, Fabio Fantoni wrote:
>> Usage: usbversion=1|2|3 (default=2)
>> Specifies the type of an emulated USB bus in the guest. 1 for usb1,
>> 2 for usb2 and 3 for usb3, it is available only with upstream qemu.
>> Default is 2.
>>
>> Signed-off-by: Fabio Fantoni <fabio.fantoni@m2r.biz>
>> ---
>>   docs/man/xl.cfg.pod.5       |    6 ++++++
>>   tools/libxl/libxl_create.c  |    3 +++
>>   tools/libxl/libxl_dm.c      |   25 ++++++++++++++++++++++++-
>>   tools/libxl/libxl_types.idl |    1 +
>>   tools/libxl/xl_cmdimpl.c    |    2 ++
>>   5 files changed, 36 insertions(+), 1 deletion(-)
>>
>> diff --git a/docs/man/xl.cfg.pod.5 b/docs/man/xl.cfg.pod.5
>> index 069b73f..602d428 100644
>> --- a/docs/man/xl.cfg.pod.5
>> +++ b/docs/man/xl.cfg.pod.5
>> @@ -1154,6 +1154,12 @@ device.
>>     Enables or disables an emulated USB bus in the guest.
>>   +=item B<usbversion=NUMBER>
>> +
>> +Specifies the type of an emulated USB bus in the guest. 1 for usb1,
>> +2 for usb2 and 3 for usb3, it is available only with upstream qemu.
>> +Default is 2.
>> +
>>   =item B<usbdevice=[ "DEVICE", "DEVICE", ...]>
>>     Adds B<DEVICE>s to the emulated USB bus. The USB bus must also be
>> diff --git a/tools/libxl/libxl_create.c b/tools/libxl/libxl_create.c
>> index 0c32d0b..9683740 100644
>> --- a/tools/libxl/libxl_create.c
>> +++ b/tools/libxl/libxl_create.c
>> @@ -229,6 +229,9 @@ int libxl__domain_build_info_setdefault(libxl__gc
>> *gc,
>>               return ERROR_INVAL;
>>           }
>>   +        if (!b_info->u.hvm.usbversion)
>> +            b_info->u.hvm.usbversion = 2;
>> +
>>           if (b_info->u.hvm.timer_mode == LIBXL_TIMER_MODE_DEFAULT)
>>               b_info->u.hvm.timer_mode =
>>                   LIBXL_TIMER_MODE_NO_DELAY_FOR_MISSED_TICKS;
>> diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c
>> index 7e54c02..aa8e131 100644
>> --- a/tools/libxl/libxl_dm.c
>> +++ b/tools/libxl/libxl_dm.c
>> @@ -492,7 +492,30 @@ static char **
>> libxl__build_device_model_args_new(libxl__gc *gc,
>>                       __func__);
>>                   return NULL;
>>               }
>> -            flexarray_append(dm_args, "-usb");
>> +
>> +            switch (b_info->u.hvm.usbversion) {
>> +            case 1:
>> +                flexarray_vappend(dm_args,
>> +                    "-device", "piix3-usb-uhci,id=usb", NULL);
>> +                break;
>> +            case 2:
>> +                flexarray_vappend(dm_args,
>> "-device","ich9-usb-ehci1,id=usb,"
>> +                    "bus=pci.0,addr=0x1d.0x7",
>> "-device","ich9-usb-uhci1,"
>> +                   
>> "masterbus=usb.0,firstport=0,bus=pci.0,multifunction=on,"
>> +                   
>> "addr=0x1d.0x0","-device","ich9-usb-uhci2,masterbus=usb.0,"
>> +                    "firstport=2,bus=pci.0,addr=0x1d.0x1", "-device",
>> +                   
>> "ich9-usb-uhci3,masterbus=usb.0,firstport=4,bus=pci.0,"
>> +                    "addr=0x1d.0x2", NULL);
> 
> I'm just curious, why is this so complicated?  Is this likely to be
> fragile and break in the future?

As pointed out previously, the bus=pci.0 bit will break with different
PCI host bridges, such as the q35 machine existing today (-M q35 uses
pcie.0 instead and it has been discussed to make q35 the default at some
point). I had thus suggested to use a variable for the bus name to
abstract it. For -M pc-i440fx-1.5 etc. pci.0 should continue to work.

Regards,
Andreas

-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg