On Fri, Jun 11, 2021 at 04:56:52PM +0530, Vinod Koul wrote:
On 09-06-21, 12:10, Jason Gunthorpe wrote:
On Wed, Jun 09, 2021 at 09:44:08AM -0500, Pierre-Louis Bossart wrote:
The consensus for the auxiliary_device model was hard to reach, and the agreement was to align on a minimal model. If you disagree with the directions, you will have to convince Nvidia/Mellanox and Intel networking folks who contributed the solution to do something different.
The purpose of the aux devices was primarily to bind a *software* interface between two parts of the kernel.
Then I dont think this example is valid... This example has a PCI device, which represents a DSP, HDA controller, DMICs, Soundwire links... So at least here it is hardware.
Yes, and that's fine, and exactly what aux devices were created for.
You divide up a single logically addressable device into differently handled portions.
Not all hardware specs were as "smart" as USB was in allowing multiple drivers to bind to the same physical USB device and talk to it at the same time :)
Luckily the USB spec authors learned from the mistakes of PCI...
thanks,
greg k-h