Summary: | webauth u2f doesn't seem to work when hw.usb.usbhid.enable=1 is set (no uhid dev entries) | ||
---|---|---|---|
Product: | Base System | Reporter: | Trenton Schulz <trueos> |
Component: | usb | Assignee: | Vladimir Kondratyev <wulf> |
Status: | In Progress --- | ||
Severity: | Affects Some People | CC: | emaste, grahamperrin, hselasky, jakefrancois15, rudolphfroger, sascha.folie, tomek, usb, wulf |
Priority: | --- | ||
Version: | 13.1-RELEASE | ||
Hardware: | Any | ||
OS: | Any |
Description
Trenton Schulz
2022-07-31 06:33:28 UTC
This is a known issue. UGEN had a patch to not do read-ahead, and usbhid.enable=1 needs the same. Adding wulf@ --HPS (In reply to Hans Petter Selasky from comment #2) OK. Thanks for the info. I guess I can do a stable build and see if that works. Unless, was the one commit in comment #2 only part of the fix? This change is already merged to 13/stable, but only for /dev/uhidN access! https://cgit.freebsd.org/src/commit/?id=20d32249196ea6122b481bd61841337a67855337 --HPS (In reply to Hans Petter Selasky from comment #4) OK, Hans Petter, I understand. This fix is in 13-STABLE for uhid, but not for the newer usbhid (hence why there are no *hid entries in /dev when hw.usb.usbhid.enable=1 on 13.1-RELEASE and 13-STABLE) For the time being, running with hw.usb.usbhid.enable=0 (the default) is fine. I was more trying the newer drivers just to see how they worked. I will be happy to test things out when a patch is ready. (In reply to Trenton Schulz from comment #5) Hi Trenton, I have some WIP patches in my local repo to disable read-ahead for YubiKey with usbhid. But they require some additional work. I hope to continue work on them in couple of weeks. Unfortunately usbhid(4) read-ahead can not be disabled with one-line patch like uhid(4) did. +1 I also use YubiKey and could make use of that patch please let me know when its ready for testing I have 13-STABLE :-) (In reply to Tomasz "CeDeROM" CEDRO from comment #7) > let me know See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=263995 It requires some tweaking for STABLE13 I tried out the -stable patch from #263995, and it seems to work well with the webauth demos I reported here. So, for this bug, I would say that it seems to work. Thanks. |