On Mon, Oct 23, 2023 at 05:05:30PM +0530, Siddharth Vadapalli wrote: > On 23/10/23 16:12, Serge Semin wrote: > > ... > > > Siddharth, if it won't be that much bother and you have an access to > > the v3.65-based Keystone PCIe device, could you please have a look > > whether it's possible to implement what Bjorn suggested? > > Unfortunately I don't have any SoC/Device with me that has the v3.65 PCIe > controller, so I will not be able to test Bjorn's suggestion. Huh. 57e1d8206e48 ("MAINTAINERS: move Murali Karicheri to credits") removed the maintainer for pci-keystone.c, so the driver hasn't had a maintainer for over two years. Given the fact that there's no maintainer, I'm more than happy to take a patch to move this code to somewhere in the host_init() callback, even if you don't have the hardware to test it. Bjorn