[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH 5/5] hw/nvme: flexible data placement emulation
From: |
Jesper Devantier |
Subject: |
Re: [PATCH 5/5] hw/nvme: flexible data placement emulation |
Date: |
Thu, 16 Feb 2023 19:36:55 +0100 |
Correct. We deliberately allow enabling fdp as a startup parameter to side-step
the need for NS mgmt.
You are also correct on the second point - the NS placement handle list can at
most hold 128 elements.
> On 16 Feb 2023, at 19.26, Keith Busch <kbusch@kernel.org> wrote:
>
> This mostly looks fine. I need to clarify some spec decisions, though.
>
> By default, FDP feature is disabled: "The default value of this Feature shall
> be 0h." You can't change the value as long as namespaces exist within the
> group, so FDP requires NS Mgmt be supported if you're going to enable it. The
> spec, however, doesn't seem to explicitly say that, and NS Mgmt isn't
> currently
> supported in this controller.
>
> We can look past that for this implementation to allow static settings even if
> that doesn't perfectly align with this feature (NS Mgmt is kind of difficult
> here). In order to match what the spec says is possible though, we can't have
> a
> namespace with more than 128 placement handles since that's the most you can
> provide when you create the namespace.
>
> Does that make sense, or am I totally misunderstanding the details?