> -----Original Message-----
> From: Jason Gunthorpe <j...@ziepe.ca>
> Sent: Monday, June 29, 2020 16:00
> To: Mark Brown <broo...@kernel.org>
> Cc: Greg KH <gre...@linuxfoundation.org>; Takashi Iwai <ti...@suse.de>;
> Pierre-Louis Bossart <pierre-louis.boss...@linux.intel.com>; Ranjani Sridharan
> <ranjani.sridha...@linux.intel.com>; Kirsher, Jeffrey T
> <jeffrey.t.kirs...@intel.com>; da...@davemloft.net; netdev@vger.kernel.org;
> linux-r...@vger.kernel.org; nhor...@redhat.com; sassm...@redhat.com;
> Fred Oh <fred...@linux.intel.com>
> Subject: Re: [net-next v4 10/12] ASoC: SOF: Introduce descriptors for SOF
> client
>
> On Mon, Jun 29, 2020 at 09:33:17PM +0100, Mark Brown wrote:
> > On Wed, May 27, 2020 at 09:17:33AM +0200, Greg KH wrote:
> >
> > > Ok, that's good to hear. But platform devices should never be
> > > showing up as a child of a PCI device. In the "near future" when we
> > > get the virtual bus code merged, we can convert any existing users
> > > like this to the new code.
> >
> > What are we supposed to do with things like PCI attached FPGAs and
> > ASICs in that case? They can have host visible devices with physical
> > resources like MMIO ranges and interrupts without those being split up
> > neatly as PCI subfunctions - the original use case for MFD was such
> > ASICs, there's a few PCI drivers in there now.
>
> Greg has been pretty clear that MFD shouldn't have been used on top of PCI
> drivers.
>
> In a sense virtual bus is pretty much MFD v2.
With the big distinction that MFD uses Platform bus/devices, which is why we
could not
use MFD as a solution, and virtbus does not.
