Re: Path forward for Virtualized Swap?

[Nhat Pham]

On Sun, Sep 20, 2026 at 5:13 PM Chris Li <chrisl@xxxxxxxxxx> wrote:
>
> That is not what I have in mind. The example I have in mind is that
> zswap is charged to swap counters. Changing that behavior will break
> our current deployment and that of many other. Kairui give some
> example as well. That behavior should not change.

We already have a knob to select vswap. Users can turn it off
initially, and/or update their software to handle the transition
based on this knob.

Otherwise, you'd never be able to introduce any behavior into the
kernel. Hugetlb accounting/charging is a new behavior, and there were
softwares written before that assumed hugetlb usage is not charged. If
it's not a good argument to block the implementation of hugetlb
accounting then, it's not a good argument now.