Re: [PATCH 1/2] x86/resctrl, Documentation: Keep mbm_assign_mode "default" on boot

[Reinette Chatre]

Hi Boris,

On 7/24/26 4:12 PM, Borislav Petkov wrote:
> On Fri, Jul 24, 2026 at 03:28:18PM -0700, Reinette Chatre wrote:
>> I understand that if users only use a fraction of the supported monitoring
>> groups then the counters will indeed be valid but (a) this is a small number
>> compared to the supported monitoring groups, and (b) hardware does not
>> expose this number.
> 
> It takes only one user to go to Linus and say, my tool doesn't work anymore
> the way I used to use it. Then stuff gets reverted immediately. Because we
> don't break userspace. Ever. 

ack.

One clarification here is that as I understand there has not yet been
an actual user complaint. At least not that the pqos utility mentioned in
this thread is aware of. Instead I struggle with the speculation about possible
user complaints with different interpretations on how this change could be
perceived by users. 
> It really is that simple. And I see your point about wrong usage and so on.
> But as long as users do use it this way, we don't break that usage.

ok. 

When considering a change like this, should it be a stable candidate to
help reduce the number of kernels with different behaviors?

The patch notes mention that this change is planned to be reverted at some
point in the future and I have not heard of any changes to this plan.
Considering the stated intention from everybody to maintain user interface
I assume this plan is off the table?

> 
> But let's see what other x86 folks think.
Thank you very much for your guidance.

Reinette