Control: tag -1 moreinfo On Sun, 2014-11-16 at 15:47 -0500, Zack Weinberg wrote: > Package: initramfs-tools > Version: 0.118 > Severity: normal > > If I set break=premount to drop into the initramfs embedded shell > (I have not tested other values of break=) I see a bunch of error messages > about various USB-related modules not being in modules.dep, and then the > (USB-based) keyboard is completely nonfunctional and the computer has > to be reset using the hardware reset button. Before starting a shell, the initramfs init script currently tries to load many modules for different USB controller drivers, and not all of them are actually built for any particular architecture. So it is expected that some of them will fail to load, but we ought to do something to avoid those error messages. > I played around with adding the missing modules to /etc/initramfs-tools/modules > but was not able to find a combination that worked. The closest I got > was with xhci_hcd + usbhid; then I see plausible-looking kernel probe > messages for the USB hub and the keyboard, but after that the keyboard > is still completely nonfunctional. > > The keyboard does work just fine in single-user mode, so I presume this is > indeed a problem with necessary modules not getting loaded. > > I am using MODULES=most, so this is not the same as #701656. [...] I am very surprised that the keyboard is not functional. Based on the lsmod output, your keyboard driver is hid-microsoft and that will be included in the initramfs. If this system has some sockets that don't support USB 3, could you test whether the keyboard works when attached to them? (The plastic inside a USB socket should be blue if it supports USB 3, and some other colour if it doesn't.) Ben. -- Ben Hutchings Any smoothly functioning technology is indistinguishable from a rigged demo.
Attachment:
signature.asc
Description: This is a digitally signed message part