On Sat, Apr 30, 2016 at 01:01:36AM +0200, Arnd Bergmann wrote: > As mentioned in another thread, I have tried to come up with > a way to make the PCI host driver registration more flexible > and simpler. > > We have actually discussed this multiple times in the past, > but always ended up elsewhere, so this is a proof of concept > work, leaving all the existing interfaces in place, and > adding a way to allocate a pci_host_bridge structure from > a driver and register that after filling out all the interesting > fields. > > This is not tested at all, and certainly not meant for > inclusion until the concept has been discussed better. > > Please have a look. Hi Arnd, I like this idea very much and welcome the timing. After the longer than expected XUSB detour I was going to look at PCI next. Making the driver work on both 32-bit and 64-bit ARM is the first thing that needs solving but once that's in place it should be fairly trivial to add support for Tegra X1 on top. From a quick glance I think this series is a really good starting point. I had a couple of questions, but I'll ask them as replies to each patch for context. Thierry
Attachment:
signature.asc
Description: PGP signature
