Re: [RFC PATCH 0/8] arm,x86,fs/resctrl: Support asymmetric L2 cache allocation

From: Chen Yu

Date: Tue Nov 03 2026 - 11:58:44 EST


On Tue, Oct 27, 2026 at 11:30:37AM -0700, Reinette Chatre wrote:
> Hi Tony,
>
> > # tree
> > .
> > ├── cpus
> > ├── cpus_list
> > ├── info
> > │   ├── L2
> > │   │   ├── bit_usage
> > │   │   ├── control_mode
> > │   │   ├── min_cbm_bits
> > │   │   ├── num_closids
> > │   │   ├── schemata
>
> This needs an L2 control. As I highlighted in
>
> https://lore.kernel.org/lkml/22ecec13-a703-414e-9f5e-75c89173e845@xxxxxxxxx/
>
> when considering pqos utility user space expects an L2 schemata file entry
> when there is an L2 resource.
>
> Since L2_ASYM is the native interface it could be used to emulate the legacy L2 control
> albeit with poor performance. By default the schemata file only contains the L2 control
> and a user space familiar with new interface can enable L2_ASYM to replace L2 control
> in schemata file.
>

In current patch it seems that we did not create any new controller, and fixes the
per rdt_ctrl_domain's cbm_len, I'm trying to understand why do we need to create a new "native"
controller?

Anyway, if we introduce a new controller, and keep the legacy interfaces, we might want to
encourage the user to switch to the new ones, would the following proposal be doable?

1.When booting up with control_mode = legacy, and the new control mode is supported
(asymmetric, region-aware, etc.), we can print a message saying: new control is available,
please switch the control_mode to "native".

2.Add a mount option to support control_mode.

3.The rdt= option, but only for x86 — not sure why rdt= is not a generic option.

thanks,
Chenyu