There is no Beaglebone release for 12.2. Discussion thread: https://lists.freebsd.org/pipermail/freebsd-arm/2020-October/022469.html Forum thread: https://forums.freebsd.org/threads/no-freebsd-12-2-release-for-beaglebone.77499/#post-482529 As per the release errata: https://www.freebsd.org/releases/12.2R/errata.html#open-issues I have a Beaglebone white and am happy to help test if given some pointers on where to start.
Please see the third bullet point here: https://www.freebsd.org/releases/12.2R/errata.html
I don't understand why this was closed out. I didn't see any bug reports mentioning what the bug was or if there was a fix. The other two errata points (which I already linked to) had corresponding FreeBSD bug references. It sounds like there is now an image available: https://download.freebsd.org/ftp/snapshots/ISO-IMAGES/12.2/FreeBSD-12.2-STABLE-arm-armv7-BEAGLEBONE-20201029-r367116.img.xz I can definitely test this.
It was closed because there will not be a BEAGLEBONE image for 12.2-RELEASE. If you can test the 12.2-STABLE image you linked to, that would be great. As I understand it, the excessively-long I/O issue is not present in 13-CURRENT, but it would definitely be nice to have this tracked down before 12.3-RELEASE when the time comes.
Right, I understand 12.2-RELEASE won't have a Beaglebone release. This is more about the slow I/O issue. It's definitely still happening, I pulled down the 12.2-STABLE image and booting probably took two hours.
The Right People(tm) are already involved in this. :) https://lists.freebsd.org/pipermail/freebsd-arm/2020-September/022423.html
Re-open, adding mmel@ to CC.
There's something messed up with the clock, I don't think it has anything to do with the SD infrastructure. If I run `date` and wait ten seconds to run `date` again, the time is the same. dd reading from the SD card thinks it's instantaneous, even though it takes a good minute or two to read a gig and write it to /dev/null. I think the clock issue is messing with the scheduler. Now I presume there's no RTC on these, but whatever "clock" interface we are using seems to be the issue.
My bad, I think that's what you all already concluded. I'll leave you to it but let me know if there's anything I can test.
Hi everyone. Any update on this? Thank you!
We do not retroactively build releases. Provided the issue(s) have been resolved, there will be BeagleBone images for 12.3.
12.3-BETA1 has the same problem.
https://download.freebsd.org/ftp/releases/arm/armv7/ISO-IMAGES/12.3/FreeBSD-12.3-BETA1-arm-armv7-BEAGLEBONE.img.xz I'll double-check the website to make sure there is a URL to it.
Added the URL to the BEAGLEBONE image on where.html. Thank you for spotting this.
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/doc/commit/?id=826e4a81a6e8ee2a02a2486befcd28f575a62329 commit 826e4a81a6e8ee2a02a2486befcd28f575a62329 Author: Glen Barber <gjb@FreeBSD.org> AuthorDate: 2021-10-25 17:50:38 +0000 Commit: Glen Barber <gjb@FreeBSD.org> CommitDate: 2021-10-25 17:50:38 +0000 where.html: add link to 12.3-BETA1 BEAGLEBONE PR: 250709 Submitted by: karels Sponsored by: Rubicon Communications, LLC ("Netgate") website/content/en/where.adoc | 1 + 1 file changed, 1 insertion(+)
I should have been more explicit. 12.3-BETA1 still has the clock problem that caused Beaglebone to be removed from 12.2. Should we have a separate bug about the clock problem?
Sorry, I misunderstood the problem. Re-opening the ticket with (somewhat low) hopes that it can be resolved for 12.3-RELEASE.
Unfortunately, this does not seem to be resolvable before 12.3 final release builds. Closing this as "overcome by events", and pulling the BeagleBone images from the 12.3 builds for the remainder of the release cycle.
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/doc/commit/?id=b91feb85b8652215a5bd20eaf9040a6b910757ac commit b91feb85b8652215a5bd20eaf9040a6b910757ac Author: Glen Barber <gjb@FreeBSD.org> AuthorDate: 2021-11-12 17:50:11 +0000 Commit: Glen Barber <gjb@FreeBSD.org> CommitDate: 2021-11-12 17:50:11 +0000 12.3: pull BEAGLEBONE from the list of arm images PR: 250709 (related) Sponsored by: Rubicon Communications, LLC ("Netgate") website/content/en/where.adoc | 1 - 1 file changed, 1 deletion(-)