@ryannathans@captainkangaroo I’m going to make the wild assumption that the kernel will have a table of the current microcode versions at the time of it’s release, but I doubt that
will get updated except by kernel upgrades.
@DaPorkchop_ Oddly, if you build your own kernel and remove the system provided one, the package gets automatically removed as well which is weird, because it is really still needed regardless.
@ryannathans Why bloat the kernel with the microcode for every intel processor that might need it (and there is a similar thing for AMD) when you don’t have that specific processor? It does make more sense for it to be a separate, especially on memory constrained systems. I mean if you’ve got 256GB of RAM probably not a big deal but if you’ve got 256MB a big deal.
How does it know if the microcode is outdated?
@ryannathans @captainkangaroo I’m going to make the wild assumption that the kernel will have a table of the current microcode versions at the time of it’s release, but I doubt that
will get updated except by kernel upgrades.
There’s probably an efivar that reads the current microcode version.
Debian-based distros (and probably most othera as well) actually have a package called “intel-microcode” which gets updated fairly regularly.
@DaPorkchop_ Oddly, if you build your own kernel and remove the system provided one, the package gets automatically removed as well which is weird, because it is really still needed regardless.
If that’s the case, why wouldn’t they put the microcode in the kernel?
@ryannathans Why bloat the kernel with the microcode for every intel processor that might need it (and there is a similar thing for AMD) when you don’t have that specific processor? It does make more sense for it to be a separate, especially on memory constrained systems. I mean if you’ve got 256GB of RAM probably not a big deal but if you’ve got 256MB a big deal.
The kernel compilation is already configurable between megabytes and gigabyte+