On Wed, Feb 26, 2025 at 01:41:14AM +0800, Mingcong Bai wrote: > Hi Alan, > > 在 2025/2/9 18:22, Mingcong Bai 写道: > > Hi again, > > > > Oops. I missed the dmesg. > > > > <snip> > > Gentle ping as it has been almost a month since our last correspondence. Can > you please advise if you would need any further information and, since the > fix is probably incorrect, if you have any suggestions as to how we could > move forward with a better fix or platform-specific quirk? I was waiting to see if your last email (Feb 9) to Huacai Chen got a reply, since you suggested the problem was Loongson-specific. As for a platform-specific solution... I wonder if it might not be better to fix this in userspace, by making Powertop (or whatever tool is responsible for managing wakeup settings) automatically enable root-hub wakeups on Loongson platforms. Another possibility is to ask somebody at Intel if Loongson has some special way to handle UHCI wakeups that we don't currently support. The driver does contain support for the USBRES_INTEL register (see uhci-hcd.h and uhci-pci.c). Is this different for Loongson? Alan Stern
