Re: Re: [PATCH 4/5] usb: allow max 8192 bytes for desc

[zhenwei pi]

On 1/4/22 11:22 PM, Philippe Mathieu-Daudé wrote:
> On 27/12/21 15:27, zhenwei pi wrote:
> > A device of USB video class usually uses larger desc structure, so
> > use larger buffer to avoid failure.
> >
> > Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
> > ---
> >   hw/usb/desc.c | 15 ++++++++-------
> >   hw/usb/desc.h |  1 +
> >   2 files changed, 9 insertions(+), 7 deletions(-)
> >
> > diff --git a/hw/usb/desc.c b/hw/usb/desc.c
> > index 8b6eaea407..7f6cc2f99b 100644
> > --- a/hw/usb/desc.c
> > +++ b/hw/usb/desc.c
> > @@ -632,7 +632,8 @@ int usb_desc_get_descriptor(USBDevice *dev, 
> > USBPacket *p,
> >       bool msos = (dev->flags & (1 << USB_DEV_FLAG_MSOS_DESC_IN_USE));
> >       const USBDesc *desc = usb_device_get_usb_desc(dev);
> >       const USBDescDevice *other_dev;
> > -    uint8_t buf[256];
> > +    size_t buflen = USB_DESC_MAX_LEN;
> > +    g_autofree uint8_t *buf = g_malloc(buflen);
>
> Do we want to have a per-device desc_size (in USBDevice, default to
> 256, video devices set it to 8K)?
>
> How "hot" is this path? Could we keep 8K on the stack?
It's an unlikely code path:
1, During guest startup, guest tries to probe device.
2, run 'lsusb' command in guest

Keeping 8K on the stack also seems OK.

> > diff --git a/hw/usb/desc.h b/hw/usb/desc.h
> > index 3ac604ecfa..35babdeff6 100644
> > --- a/hw/usb/desc.h
> > +++ b/hw/usb/desc.h
> > @@ -199,6 +199,7 @@ struct USBDesc {
> >       const USBDescMSOS         *msos;
> >   };
> > +#define USB_DESC_MAX_LEN    8192
> >   #define USB_DESC_FLAG_SUPER (1 << 1)
> >   /* little helpers */

--
zhenwei pi