You are not logged in.
Hello all, I'm mainly a Fedora user but I've been playing with Arch in VirtualBox for the past few days and I've been enjoying it a lot. I decided to try to install it directly on my system this afternoon, however I'm getting the messages:
device descriptor read/64, error -71
Unable to enumerate USB device on port
for anything plugged into a USB 2 port. USB 3.0 ports seem to be working fine. The USB2.0 ports work in the bootloader before Arch boots, I was able to move up and down in the menu with the keyboard plugged into a 2.0 port. Is this something that a complete installation may fix? I'm a bit worried that any issues I have booting to the live system will carry over into the installation.
My motherboard is a GA-990FXA-UD3 R5.
http://www.gigabyte.us/Motherboard/GA-9 … -rev-10#ov
Thanks for reading, and any input you have on the matter.
***********
SOLUTION: The fix was to enable IOMMU on the motherboard and to add "iommu=soft" as a kernel parameter in grub.
Last edited by SonataNo8 (2017-03-13 23:11:57)
Offline
I had the same issue when I installed on my laptop, around 3 years ago. 2 and 3 ports have worked fine ever since.
Offline
Thanks for the response! Did you do anything in particular to fix it or just complete the installation?
Offline
There was nothing to fix really, it was just the odd combination of UEFI firmware, hardware and the ISO.
Offline
I went ahead and completed the installation, no luck after booting into it. Nothing was plugged into the USB2 ports, I'd moved everything into the 3.0 ports.
Any ideas?
Offline
Don't post images of text, paste the actual text: https://wiki.archlinux.org/index.php/Co … s_and_code
Is this with the stock kernel? Paste the dmesg output from a 3 and a 2 port with a knoen good USB drive.
Offline
Unfortunately, I'm not actually able to use the installation to get those text dumps. Those errors in the pic are scrolling by non-stop. This is the default kernel from the live ISO.
I found on the Ubuntu forums that enabling IOMMU in the bios seems to fix it on those installations, I'll try it this evening and update this post if it works.
Last edited by SonataNo8 (2017-03-13 18:24:51)
Offline
If you still habe problems, you can try to set some usbcore parameters in the kernel line like disabling autosuspend or changing the init scheme.
<<usbcore.autosuspend=-1>> or <<usbcore.old_scheme_first=1>>
# https://www.kernel.org/doc/Documentation/admin-guide/kernel-parameters.txt
usbcore.autosuspend=
[USB] The autosuspend time delay (in seconds) used
for newly-detected USB devices (default 2). This
is the time required before an idle device will be
autosuspended. Devices for which the delay is set
to a negative value won't be autosuspended at all.
usbcore.old_scheme_first=
[USB] Start with the old device initialization
scheme (default 0 = off).
Last edited by progandy (2017-03-13 18:34:24)
| alias CUTF='LANG=en_XX.UTF-8@POSIX ' |
Offline
Well, the IOMMU setting worked as far as enabling all of my USB ports, but now I'm getting a few hundred of these errors at startup:
[ 1.453485] AMD-Vi: Event logged [
[ 1.453496] IO_PAGE_FAULT device=02:00.0 domain=0x0011 address=0x00000000be9bf880 flags=0x0000]
I'm trying a few more things, I'll post the results here.
Offline
The fix for *that* was to add "iommu=soft" as a kernel parameter. I just had a clean boot with all USB ports working. I'll go ahead and mark this solved. Thanks, everyone!
Offline
You might want to report this upstream, as a quirk might need to be included in the kernel so that it works without extra parameters.
R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K
Offline
Forgive my ignorance, but where would I report this?
Offline
If the issue is with the kernel you should report a bug at https://bugzilla.kernel.org/
Edit: But search first in the bug tracker to make sure no one has already submitted a bug.
Last edited by circleface (2017-03-14 22:49:11)
Offline