* FEAT_CMOW
@ 2026-02-19 5:21 Manoj Ekbote
2026-02-19 8:52 ` FEAT_CMOW Marc Zyngier
0 siblings, 1 reply; 2+ messages in thread
From: Manoj Ekbote @ 2026-02-19 5:21 UTC (permalink / raw)
To: linux-arm-kernel@lists.infradead.org
Hi,
While looking at the mainline v6.18.6 sources, I noticed that FEAT_CMOW does not appear in the arm64 CPU capability table (arch/arm64/kernel/cpufeature.c). Since CMOW is a mandatory architectural feature starting in Arm v8.8, I expected to see it in the cpufeature framework.
How is the cpu feature list typically updated for architectural extensions like this?
Is it done only when the kernel needs to make use of the feature, or is there a process to actively add all mandatory architecture features?
Thanks
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: FEAT_CMOW
2026-02-19 5:21 FEAT_CMOW Manoj Ekbote
@ 2026-02-19 8:52 ` Marc Zyngier
0 siblings, 0 replies; 2+ messages in thread
From: Marc Zyngier @ 2026-02-19 8:52 UTC (permalink / raw)
To: Manoj Ekbote; +Cc: linux-arm-kernel@lists.infradead.org
On Thu, 19 Feb 2026 05:21:02 +0000,
Manoj Ekbote <mekbote@qti.qualcomm.com> wrote:
>
> Hi,
>
> While looking at the mainline v6.18.6 sources, I noticed that
> FEAT_CMOW does not appear in the arm64 CPU capability table
> (arch/arm64/kernel/cpufeature.c). Since CMOW is a mandatory
> architectural feature starting in Arm v8.8, I expected to see it in
> the cpufeature framework. How is the cpu feature list typically
> updated for architectural extensions like this?
When we have a use for it. FEAT_CMOW has the potential to break the
ABI we already expose to userspace, so I don't immediately see why
we'd care about it.
> Is it done only when the kernel needs to make use of the feature, or
> is there a process to actively add all mandatory architecture
> features?
The Linux kernel is thankfully not a validation tool for the
architecture, and is not aligned to any particular version of the
architecture either. A "mandatory" feature is only a requirement for
implementations, not for SW.
Thanks,
M.
--
Without deviation from the norm, progress is not possible.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-02-19 8:52 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-02-19 5:21 FEAT_CMOW Manoj Ekbote
2026-02-19 8:52 ` FEAT_CMOW Marc Zyngier
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox