Re: [Qemu-devel] [PATCHv2 1/1] Add usb option in machine options to enable/disable usb

[Li Zhang]

On Mon, Jun 18, 2012 at 3:29 PM, Markus Armbruster <address@hidden> wrote:
> Li Zhang <address@hidden> writes:
>
>> On Fri, Jun 15, 2012 at 10:34 PM, Markus Armbruster <address@hidden> wrote:
>>> Li Zhang <address@hidden> writes:
>>>
>>>> On Fri, Jun 15, 2012 at 8:04 PM, Markus Armbruster <address@hidden> wrote:
>>>>> Li Zhang <address@hidden> writes:
>>>>>
>>>>>> For pseries machine, it needs to enable usb to add
>>>>>> keyboard or usb mouse. -usb option won't be used in
>>>>>> the future, and machine options is a better way to
>>>>>> enable usb.
>>>>>>
>>>>>> So this patch is to add usb option to machine options
>>>>>> (-machine type=psereis,usb=on/off)to enable/disable
>>>>>> usb controller.
>>>>>>
>>>>>> In this patch, usb_on is an global option which can
>>>>>> be checked by machines.
>>>>>> For example, on pseries, it will check if usb_on is 1,
>>>>>> if it is 1, it will create one usb ohci controller.
>>>>>> As the following:
>>>>>> if (usb_on == 1) {
>>>>>>      pci_create_simple(bus, -1, "pci-ohci");
>>>>>> }
>>>>>>
>>>>>> In this patch, usb is on by default. So, for -nodefault,
>>>>>> usb should be set off in the command line as the following:
>>>>>>  -machine type=pseries,usb=off.
>>>>>>
>>>>>> Signed-off-by: Li Zhang <address@hidden>
>>>>>>
>>>>>> ---
>>>>>>  hw/spapr.c |    5 +++++
>>>>>>  sysemu.h   |    1 +
>>>>>>  vl.c       |   17 +++++++++++++++++
>>>>>>  3 files changed, 23 insertions(+)
>>>>>>
>>>>>> diff --git a/hw/spapr.c b/hw/spapr.c
>>>>>> index d0bddbc..1feb739 100644
>>>>>> --- a/hw/spapr.c
>>>>>> +++ b/hw/spapr.c
>>>>>> @@ -661,6 +661,11 @@ static void ppc_spapr_init(ram_addr_t ram_size,
>>>>>>          spapr_vscsi_create(spapr->vio_bus);
>>>>>>      }
>>>>>>
>>>>>> +    if (usb_on == 1) {
>>>>>> +        pci_create_simple(QLIST_FIRST(&spapr->phbs)->host_state.bus,
>>>>>> +                          -1, "pci-ohci");
>>>>>> +    }
>>>>>> +
>>>>>>      if (rma_size < (MIN_RMA_SLOF << 20)) {
>>>>>>          fprintf(stderr, "qemu: pSeries SLOF firmware requires >= "
>>>>>>                  "%ldM guest RMA (Real Mode Area memory)\n", 
>>>>>> MIN_RMA_SLOF);
>>>>>> diff --git a/sysemu.h b/sysemu.h
>>>>>> index bc2c788..08134ae 100644
>>>>>> --- a/sysemu.h
>>>>>> +++ b/sysemu.h
>>>>>> @@ -109,6 +109,7 @@ extern int vga_interface_type;
>>>>>>  #define vmsvga_enabled (vga_interface_type == VGA_VMWARE)
>>>>>>  #define qxl_enabled (vga_interface_type == VGA_QXL)
>>>>>>
>>>>>> +extern int usb_on;
>>>>>>  extern int graphic_width;
>>>>>>  extern int graphic_height;
>>>>>>  extern int graphic_depth;
>>>>>> diff --git a/vl.c b/vl.c
>>>>>> index 204d85b..b200203 100644
>>>>>> --- a/vl.c
>>>>>> +++ b/vl.c
>>>>>> @@ -202,6 +202,7 @@ int smp_cpus = 1;
>>>>>>  int max_cpus = 0;
>>>>>>  int smp_cores = 1;
>>>>>>  int smp_threads = 1;
>>>>>> +int usb_on = 0;
>>>>>>  #ifdef CONFIG_VNC
>>>>>>  const char *vnc_display;
>>>>>>  #endif
>>>>>> @@ -758,6 +759,21 @@ static int bt_parse(const char *opt)
>>>>>>      return 1;
>>>>>>  }
>>>>>>
>>>>>> +static int get_usb_opt(QemuOpts *opts)
>>>>>> +{
>>>>>> +    const char *usb_opt = NULL;
>>>>>
>>>>> Useless initializer.
>>>> Thanks. I will remove it.
>>>>>
>>>>>> +    int usb_on = 0;
>>>>>> +
>>>>>> +    if (NULL == qemu_opt_get(opts, "usb"))
>>>>>> +        qemu_opt_set(opts, "usb", "on");
>>>>>
>>>>> Why are you changing opts?
>>>> USB is enabled by default when there is no usb option setting.
>>>> For example,
>>>> using  # qemu-system-ppc64 -machine type=pseries
>>>> There is no usb option, but usb is set on.
>>>
>>> Isn't it off by default for at least some machines now?
>>>
>> OK. This default setting is decided by the machine.
>> In the new version, I put this setting in machine.
>> It can be set off or on.
>> For psereis it sets on.
>
> Makes sense.
>
> Perhaps we really have three kinds of machines, not just two:
>
> 1. Must have USB: main() sets usb_enabled to true.

We only hope to use usb option of machine options, not use -usb or -usbdevice,
which will be removed in the future.
But there are still some machines are using it.
>
> 2. May have USB: usb_enabled = -usb or -usbdevice given
>
For one machine, if it uses usb option in machine, usb_enabled can't
work for this machine.
In fact, even if user use -usb, it still doesn't work.
> 3. Can't have USB: fail if the user tries to enable it.
>
> Code sketch:
>
>    /* init USB devices */
>    if (!machine->has_usb) {
>        if (usb_enabled)
>            [report error; should point to the offending options]
>            exit(1);
>        }
>    } else {
>        if (machine->has_usb > 0) {
>            usb_enabled = 1;
>        }
>        if (usb_enabled) {
>            if (foreach_device_config(DEV_USB, usb_parse) < 0)
>                exit(1);
>        }
>    }
>
>
In fact, I really hope to remove usb_enabled.

>>> Anyway, I don't see why we need to update opts.  Who's using the updated
>>> opts?
>>>
>> psereis will use this opts.
>> usb kbd and mouse will be needed  with vga enabled.
>
> Do they use the updated QemuOpts *opts?  I'd expect them to use usb_on,
> or whatever flag variable governs USB (now: usb_enabled).

Yes, I have modified this for pseries. It will get usb_on from machine options.
But there some other kinds of machines using usb_enabled.

I have send out my latest patches, but it seems that there are some
issues of my network.
I can't see them in the mailing list. I will send out later.
>
> [...]



-- 

Best Regards
-Li