[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: 60-serial.rules, broken

On 6/14/23 12:01, Sascha Silbe wrote:

Hello,

gene heskett <gheskett@shentel.net> writes:


That's Debian#1035094 [1]. A fix was uploaded three weeks ago and is
available in bullseye-proposed-updates [2] but not the regular Bullseye
repository. If I understand the status page [3] correctly it's only
scheduled for the next point release. Given that there already was a
point release recently (11.7 on 2023-05-01) it will probably another
couple of months until this bug gets fixed in bullseye. :-/

[...]

That is most unfortunate for 3d printer users, our support group on
discord is blowing up with the problems this is cause for klipper users,
can its priority be raised?  We use it because that gives us a
bulletproof method using the devices ChipID, which is totally unique,
like MAC addresses.


Tell me about it. As an embedded developer this bug is a huge PITA. I'm
surprised the fix for this regression hasn't already landed in
stable. But then I don't know what else changed and what the risks are
for other bugs to show up and potentially affect even more users. After
all systemd is a rather central component of most Debian systems.

You may want to try asking on the bug report (#1035094 [1]) about
it. I'm not a Debian Developer myself and don't have much experience
trying to get a fix landed in stable but I expect there is a way for the
maintainers of the package to request an exception and get the fix in
the repository before the next point release.

As a first step however we should try out the version from
bullseye-proposed-updates to make sure it's safe to use. I've already
installed it on one of my hosts (where it fixed the bug) and will
monitor it for regressions; would be great if others did the same.

Sascha

[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1035094
A fixit script was posted on discord/linux/klipper/configuration, and pinned. It worked ok on several of my bananapi-m5's and my i5 version of bullseye had a /dev/serial/by-path, but no by-id, so I ran that same script here, but it pulled in 39 other updates at the same time and blew away my passwd. no root pw had ever been set as I use sudo for that, so I was locked out. 5 installs later I'm finally back in business. But my setup is a bit unusual, /home is a raid10 of 1.7 TB on 4 1Tb samsung ssd's, with 60Gb of swap on the rest of that raid10. And I still haven't recreated my local /sshnet network of all my cnc'd machines and 3d printers but that's just details I haven't addressed yet.
That script does an apt update && apt upgrade -y before it pulls in this new 60-persistent-serial into /etc/udev/rules.d. Then recommended strongly that I reboot.
It has not actually bothered this machine. so given the nearly 3 days of hell it gave me to recover from it, I'll wait for it to make 64 bit intel bullseye.
As for the networking, I've found sshfs to be many times more dependable than nfs any version, it Just Works. That forces me to go to that machines keyboard for root stuffs, its only disadvantage. But I'll go check that bug since to see if its the same fix. 

Take care & stay well.

Cheers, Gene Heskett.
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page <http://geneslinuxbox.net:6309/>
Reply to: