linux-kernel - Re: [PATCH v1 1/3] sound: usb: Add vendor's hooking interface

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Yj1md+TOz2nCnRWz@kroah.com>
Date:   Fri, 25 Mar 2022 07:51:35 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Oh Eomji <eomji.oh@...sung.com>
Cc:     Jaroslav Kysela <perex@...ex.cz>, Takashi Iwai <tiwai@...e.com>,
        open list <linux-kernel@...r.kernel.org>,
        alsa-devel@...a-project.org, Leon Romanovsky <leon@...nel.org>,
        Pavel Skripkin <paskripkin@...il.com>,
        JaeHun Jung <jh0801.jung@...sung.com>
Subject: Re: [PATCH v1 1/3] sound: usb: Add vendor's hooking interface

On Fri, Mar 25, 2022 at 03:44:13PM +0900, Oh Eomji wrote:
> On Thu, Mar 24, 2022 at 09:33:40AM +0100, Greg Kroah-Hartman wrote:
> > On Thu, Mar 24, 2022 at 05:10:42PM +0900, Oh Eomji wrote:
> > > In mobile, a co-processor can be used with USB audio to improve power
> > > consumption.  To support this type of hardware, hooks need to be added
> > > to the USB audio subsystem to be able to call into the hardware when
> > > needed.
> > > 
> > > The main operation of the call-backs are:
> > >   - Initialize the co-processor by transmitting data when initializing.
> > >   - Change the co-processor setting value through the interface
> > >     function.
> > >   - Configure sampling rate
> > >   - pcm open/close
> > >   - other housekeeping
> > > 
> > > Known issues:
> > >   - This only supports one set of callback hooks, meaning that this only
> > >     works if there is one type of USB controller in the system.  This
> > >     should be changed to be a per-host-controller interface instead of
> > >     one global set of callbacks.
> > > 
> > > Signed-off-by: JaeHun Jung <jh0801.jung@...sung.com>
> > > Signed-off-by: Oh Eomji <eomji.oh@...sung.com>
> > > ---
> > >  sound/usb/card.c     | 119 +++++++++++++++++++++++++++++++++++++++++++++++++++
> > >  sound/usb/card.h     |  20 +++++++++
> > >  sound/usb/usbaudio.h |  45 +++++++++++++++++++
> > >  3 files changed, 184 insertions(+)
> > > 
> > > diff --git a/sound/usb/card.c b/sound/usb/card.c
> > > index 3769622..bd59311 100644
> > > --- a/sound/usb/card.c
> > > +++ b/sound/usb/card.c
> > > @@ -117,6 +117,117 @@ MODULE_PARM_DESC(skip_validation, "Skip unit descriptor validation (default: no)
> > >  static DEFINE_MUTEX(register_mutex);
> > >  static struct snd_usb_audio *usb_chip[SNDRV_CARDS];
> > >  static struct usb_driver usb_audio_driver;
> > > +static struct snd_usb_audio_vendor_ops *usb_vendor_ops;
> > > +
> > > +int snd_vendor_set_ops(struct snd_usb_audio_vendor_ops *ops)
> > > +{
> > > +	if ((!ops->connect) ||
> > > +	    (!ops->disconnect) ||
> > > +	    (!ops->set_interface) ||
> > > +	    (!ops->set_rate) ||
> > > +	    (!ops->set_pcm_buf) ||
> > > +	    (!ops->set_pcm_intf) ||
> > > +	    (!ops->set_pcm_connection) ||
> > > +	    (!ops->set_pcm_binterval) ||
> > > +	    (!ops->usb_add_ctls))
> > > +		return -EINVAL;
> > > +
> > > +	usb_vendor_ops = ops;
> > > +	return 0;
> > > +}
> > > +EXPORT_SYMBOL_GPL(snd_vendor_set_ops);
> > > +
> > > +struct snd_usb_audio_vendor_ops *snd_vendor_get_ops(void)
> > > +{
> > > +	return usb_vendor_ops;
> > > +}
> > 
> > This is the function you need to fix up, and add proper reference
> > counting to, in order to solve your "this breaks with multiple USB
> > controllers" problem.  So this really should not be all that difficult
> > of a task.  Why has it taken years to do so?
> > 
> > thanks,
> > 
> > greg k-h
> >
> Hi,
> 
> Is that mean the scenario when two or more usb audio devices are
> connected through the hub?

I have no idea, it's your hardware and code to test, not mine :)

greg k-h