Hello Alan,
Sorry for the delay. I've sent another version of this patch.
>
> That isn't true. Look again at the code:
>
> > /*
> > * Just before we start our control thread, initialize
> > * the device if it needs initialization
> > */
> > if (us->unusual_dev->initFunction) {
> > p = us->unusual_dev->initFunction(us);
> > if (p)
> > return p;
>
> If p isn't zero then this function uses p as its return value. Thus it
> does more with p than just check whether it is zero.
>
Yeah, you are right. Well what I mean about the "bad thing" is that something wrong but the usb core continue to install this driver, just like what the karam driver is doing.
> > }
> >
> > I will then send patches to make sure all initFunction follow the standard.
> >
> > >
> > > And don't forget the error code for when the rio_karma_send_command()
> > > call fails. In that case the return value should be -EIO.
> > >
> >
> > Okay, will add this in next version of this patch.
> >
> > >
> > > Shouldn't this also be marked for the stable kernels?
> > >
> >
> > Sorry, I didn't get it, do you mean add another tag like "Cc: stable@xxxxxxxxxxxxxxx"?
> > Or I just need to CC the mail to stable mail list?
>
> I mean add another tag. See the description of Option 1 in
> Documentation/process/stable-kernel-rules.rst.
>
Cool, have added this one.
Link: https://marc.info/?l=linux-usb&m=164965695228314&w=2
Thanks
Lin Ma
> Alan Stern
>
> >
> > Regards
> > Lin
