Re: Re[2]: cpuctl(8) - unable to update amd ucode - no such file

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Sat, 19 Sep 2026 11:54:18 -0000 (UTC)
    From:        mlelstv%serpens.de@localhost (Michael van Elst)
    Message-ID:  <118lt5a$t2o$1%serpens.de@localhost>

  | drixter%e-utp.net@localhost ("Marcin Gondek") writes:
  | >What is wrong in that case:
  | >phoenix# cpuctl ucode
  | >cpuctl: please also check dmesg(8) output for additional error information
  | >cpuctl: IOC_CPU_UCODE_APPLY: No such file or directory
  |
  | There is simply no update for your CPU.

If that was the problem, there should have been a console message
(that the dmesg(8) suggestion should have made known).

  | but even with a file available, you need one that provides a microcode patch
  | for your specific model. Otherwise you get the same error.

We could fix that, make the error in that case be ESRCH or ENXIO
or something, so it is easier to tell the difference, since these
two (no file, no ucode for cpu) are the likely failures.  cpuctl
could be taught to look for these error codes and generate sensible
messages.

In the Intel ucode update, we could possibly also change some of
the zillion different EINVAL errors into something different (EFBIG
would be good for one of them at least).   That, or perhaps (or in
addition) actually add some kernel printfs to explain some of the
validation errors that can occur though they're mostly unlikely.

kre