On Fri, Nov 5, 2021 at 8:47 PM Qian Cai <quic_qiancai@xxxxxxxxxxx> wrote: > On 11/1/21 7:51 PM, Andy Shevchenko wrote: > > No, it’s not so easy. What you are doing is a papering over the real issue > > which is the limitation of the firmware nodes to two. What we need is to > > drop the link from struct fwnode_handle, move it to upper layer and modify > > all fwnode ops to be used over the list of fwnode:s. > > > > XHCI driver and DWC3 are sharing the primary fwnode, but at the same time > > they wanted to have _different_ secondary ones when role is switched. This > > can’t be done in the current design. And here is the symptom what you got. > > Andy, thanks for the pointers so far. I was able to trace > set_primary_fwnode() and set_secondary_fwnode(). Can you share the trace you have got? > Anyway, what's the "upper layer"? Is that "struct device" or "struct > swnode"? I suppose you meant: struct device here. > - Remove "secondary" field from "struct fwnode_handle". > - Replace "fwnode" from "upper layer" with > "struct list_head fwnode_head;". > - Modify all functions in "software_node_ops" to use "fwnode_head". > > Is that correct? Yes. It might be a bit complicated taking into account how much fwnode is spreaded in the kernel... Basically, you need to fix all direct accesses to the dev->fwnode first. Besides that you need to check that fwnode, which is used out of the device scope, like in IRQ domains, doesn't use secondary pointer(s). This nevertheless adds a lot of flexibility and we may add whatever type of fwnodes and mix them together. -- With Best Regards, Andy Shevchenko
