Re: [PATCH v5 0/3] x86,fs/resctrl,arm_mpam: Factor MBA parse-time conversion to be per-arch

[Borislav Petkov]

On Fri, Sep 04, 2026 at 02:15:08PM -0700, Reinette Chatre wrote:
> Sashiko's second comment is about resctrl's existing (not introduced by this patch)
> Intel specific enumeration being resilient against a hypervisor misconfiguration.
> Thus far resctrl assumed the enumeration would return what the actual hardware
> is known to support.
> 
> Sashiko had similar "hypervisor misconfiguration" comments in response to another
> resctrl series that already falls under the "cleanup" umbrella. It is not clear to
> me how far resctrl needs to go to be resilient against hypervisor misconfiguration
> but I can include needed validation addressing this report as part of [1]. 

I would always ask myself how realistic is the use case. Here, I'm thinking
not that realistic unless you guys wanna support resctrl in a guest. Which
would be weird, IMHO, because a guest can fake all kinds of hw so controlling
resources of a fake hw is an exercise in futility.

So might as well answer that Sashiko review comment with "thanks, but no
thanks."

:-)

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette