Page MenuHomeFreeBSD
Differential D59440

14.5/errata.adoc: xhci(4) data toggle issue
Needs ReviewPublic
Actions

Authored by vladlen on Sat, Sep 5, 12:15 PM.
Tags
Referenced Files
F170962449: D59440.id185913.diff
Mon, Sep 7, 9:32 PM
F170909435: D59440.diff
Mon, Sep 7, 12:27 PM
F170888263: D59440.id185913.diff
Mon, Sep 7, 8:16 AM
F170874607: D59440.diff
Mon, Sep 7, 5:11 AM
Unknown Object (File)
Sun, Sep 6, 2:06 AM
Unknown Object (File)
Sun, Sep 6, 1:40 AM
Subscribers
None

Details

Reviewers
aokblast
ziaee
cperciva
Group Reviewers
releng
Summary

Description:
This change adds an Open Issues entry to errata.adoc for
FreeBSD 14.5-RELEASE regarding a regression in the xHCI driver
that affects some USB 2.0 devices -
listed in https://wiki.freebsd.org/action/diff/Releng/14.5ISSUES.

The problem was fixed in stable/14 by commit f656bde0c2ef
MFC of 0f59df83869d). However, the fix landed after
the 14.5-RELEASE branch point, so the issue remains
in the release image.

Impact:
Users with certain USB 2.0 umass devices may
experience communication errors.

Status: Open

Please review and confirm the wording is appropriate for the errata.adoc.

Diff Detail

Repository
R9 FreeBSD doc repository
Lint
Lint Skipped
Unit
Tests Skipped

Event Timeline

vladlen requested review of this revision.Sat, Sep 5, 12:15 PM
vladlen created this revision.
ziaee added inline comments.Sat, Sep 5, 1:34 PM
website/content/en/releases/14.5R/errata.adoc
66–68

Please simplify this. The mechanical details are not appropriate here, only explaining what is going on to the operators perspective.

What type of devices? In the review it says yubikeys, the existing text reads like its internet adapters. I don't expect 100% of the target audience to know USB is transfered in packets.

Also, there is *always* a workaround.
My suggested text here is just a suggestion.

ziaee added a comment.Sat, Sep 5, 1:44 PM

Also thank you for doing this!

vladlen added inline comments.Sat, Sep 5, 8:12 PM
website/content/en/releases/14.5R/errata.adoc
66–68

Thank you for the feedback. Based on the bug report and discussion, I understand that affected devices may stop responding and require reconnection. If the authors of the fix could confirm this behavior, that would be very helpful.

If that's accurate, how about this wording:

Some USB 2.0 devices (e.g., Yubikeys) may stop responding and require a manual reconnection to resume operation. +
Workaround: If you are affected, unplug and reinsert the device, or use man:usbconfig[8] to reset the device.

ziaee added inline comments.Sat, Sep 5, 8:34 PM
website/content/en/releases/14.5R/errata.adoc
66–68

That sounds perfect to me.

aokblast added inline comments.Sat, Sep 5, 8:52 PM
website/content/en/releases/14.5R/errata.adoc
66–68

Hi. Thanks for doing this! But it is not the correct problem that we aim to solve. Commit rG28d85db46b484589e2ee74cf4b270db066821de1 fix the EZ-USB2 and yubikey problem. However, it introduce the regression on some umass devices, which is fixed in rGf656bde0c2efd016f1c5d1f22460d657e73a7eb5 but not in 14.5 now.

vladlen added inline comments.Sun, Sep 6, 12:17 PM
website/content/en/releases/14.5R/errata.adoc
66–68

Hi. Thanks for doing this! But it is not the correct problem that we aim to solve. Commit rG28d85db46b484589e2ee74cf4b270db066821de1 fix the EZ-USB2 and yubikey problem. However, it introduce the regression on some umass devices, which is fixed in rGf656bde0c2efd016f1c5d1f22460d657e73a7eb5 but not in 14.5 now.

Thanks for the clarification. So the remaining issue in 14.5-RELEASE is a regression affecting some umass devices (USB mass storage). Could you clarify:

  • What does the user actually experience? Do devices become completely inaccessible, or just intermittent errors? I need to describe this from the user's perspective.
  • Is there a workaround? For example, does unplugging and replugging the device help, or is a reboot required?

I'm concerned that if this affects many USB storage devices, users may have a very poor experience until the fix arrives. The release docs should clearly warn them.

vladlen edited the summary of this revision. (Show Details)Sun, Sep 6, 12:18 PM
ziaee added a comment.Sun, Sep 6, 1:08 PM

I'm concerned that if this affects many USB storage devices, users may have a very poor experience until the fix arrives. The release docs should clearly warn them.

+1

Revision Contents
Changeset List

PathSize
website/
content/
en/
releases/
14.5R/
4 lines

Diff 185913

View Options

website/content/en/releases/14.5R/errata.adoc

Loading...