Apologies for not being clear - yes, that resolved it. Also, to clarify
by happening again, the kernel changed to 6.14.0-37-generic to trigger
it this time.

Thanks for adding more background. TBH, I had to look up what HWE
kernels are.

So I realize there is some trade-offs on keeping things up to date, but my 
assumptions (may be wrong, please correct :)) are:
1. A new (HWE) kernel is available.
2. It is installed, including the nvidia drivers for it.
3. During that installation something goes wrong (in this case, not having that 
latest version of a couple packages, fine, it happens)
4. UEFI/Bootloader is still updated (or automatically uses the latest) - so 
even though there was a partial installation failure, we switch to that new 
kernel next time giving a quite poor user experience (i.e. one low-resolution 
monitor)

So my suggestion is a step 3, _if_ something noticeable goes wrong, then
the kernel version is not updated - and potentially something is told to
the user or reported centrally.

Alternatively, if that is somehow not possible, it would seem nice that
when booting into a fairly broken kernel that it could tell the user
"something critical did not load, do you want the previous version".

And I'll re-iterate, I don't know any of the details - just some
musings. :)

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2131020

Title:
  New kernel version failed to prepare nvidia driver

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+bug/2131020/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to