Re: [PATCH RFC] x86/microcode/intel: Refresh the revisions that determine old_microcode to 20250812 (Aug 2025)

[Borislav Petkov <bp@alien8.de>]

On October 5, 2025 6:58:05 AM GMT+03:00, Dave Hansen <dave.hansen@intel.com> wrote:
>On 10/4/25 19:42, Borislav Petkov wrote:
>> This is turning into exactly what I was afraid and I warned it would turn onto:
>> 
>> 1. https://git.kernel.org/tip/952df63ef426b21d6da14bb48748f12b0ae2fe36 - we *just* updated it and there's already a new one
>> 
>> 2. Random people - not Intel - are going to be sending updates too.
>> 
>> This is a mess waiting to happen. ;-/
>
>In the end, we've got an lightly documented patch that doesn't make a
>lot of sense to apply. Not exactly a unique situation. ;)
>
>Xose, do you have any specific, practical reasons this should be
>applied? Honestly, I don't hold the "HIGH" Intel SA rating in super high
>regard.

That's the problem. It'll be needless and endless discussions of the sort: oh, why do you need to update those revisions? But but, I need this and that. Oh but you don't. Blablabla... Upstream has these revisions, why aren't they backported...?

Versus: people should always update to the latest microcode and we're not tracking this in the kernel because there's really no need to.

And none of those issues exist anymore.

This is one of the reasons why I'm making sure the loader always works and is very easy to update microcode. You always update and that's it.

So you don't need any of that unnecessary work.

Thx.


Small device. Typos and formatting crap