Bugzilla – Bug 1174017
Laptop: one of two usb3 hubs gone with 5.7 - works with 5.3.6 (linux live system)
Last modified: 2020-12-23 11:23:07 UTC
Today I've discovered that with 5.7.7 and back to 5.7.2 one of my laptop's internal usb3 hub is just gone (one type c and one type a port). I've tried removing the battery, bios reset etc. It's just GONE. Booting a live system with 5.3.6 works every time - no issues. Same hardware as in bug report 1167146 I will do some testing with other kernel versions and report back. Please let me know what info is required to get my usb ports back.
I've checked with kernel 5.6.12 (partedmagic) - no issues whatsoever I can boot from the ports that TW doesn't like - no issues I can boot from those ports a Leap 15.2 (5.3.18-lp152.20.7) live system - no issues
Disabling LPM like so seems to be beneficial: https://askubuntu.com/questions/802712/how-to-disable-usb3-lpm-hardware-controlled-power-management
It seems that running this script also unfreezes frozen ports. No reboot needed. I've added it as a one-shot on system startup now. Is there some sort of "quirks" file for permanently adding this to the kernel? I haven't tested it yet, but I suspect that booting a recent TW installer from the affected ports on that hub would fail as well.
Thinking about this a bit more, this recent effect on my laptop might actually be an escalation of the issues I've had and reported in 1167146, as these issues go away as well, as far as I've checked.
Just booted openSUSE-Tumbleweed-KDE-Live-x86_64-Snapshot20200708-Media.iso successfully from the "bad" ports. No issues. Is LPM disabled on Live ISOs ?
Hi Robert, sorry for the late reply (In reply to robert spitzenpfeil from comment #3) > It seems that running this script also unfreezes frozen ports. No reboot > needed. I think you missed attaching the script here. Otherwise, any behavior changes with the latest TW kernels?
I used to run this as a one-shot after booting: --- snip--- for i in /sys/bus/usb/devices/*/power/autosuspend_delay_ms; do echo -1 > $i; done for foo in /sys/bus/usb/devices/*/power/control; do echo on > $foo; done --- snip --- For some time now it is disabled and everything seems to run just fine without it. The real acid test would be a fresh installation, but I'm reluctant to disturb my working machine. As described in https://bugzilla.opensuse.org/show_bug.cgi?id=1167146, I now suspect the same cause for this and that issue. As far as I can tell, both are gone now. I'm testing it sporadically when I need my scanner.
Let's close then. Thanks Robert.