Re: git: ff18000da3 - main - HW Relnotes: Add nvd to template

On 11/5/25 18:08, Alexander Ziaee wrote:
> The branch main has been updated by ziaee:
> 
> URL: https://cgit.FreeBSD.org/doc/commit/?id=ff18000da31f4135af522f43beb4d0877b024fea
> 
> commit ff18000da31f4135af522f43beb4d0877b024fea
> Author:     Alexander Ziaee <ziaee@FreeBSD.org>
> AuthorDate: 2025-11-05 23:07:39 +0000
> Commit:     Alexander Ziaee <ziaee@FreeBSD.org>
> CommitDate: 2025-11-05 23:07:39 +0000
> 
>      HW Relnotes: Add nvd to template
> ---
>   website/archetypes/release/hardware.adoc | 2 ++
>   1 file changed, 2 insertions(+)
> 
> diff --git a/website/archetypes/release/hardware.adoc b/website/archetypes/release/hardware.adoc
> index 0a9026c2fd..304641abd1 100644
> --- a/website/archetypes/release/hardware.adoc
> +++ b/website/archetypes/release/hardware.adoc
> @@ -312,6 +312,8 @@ EISA adapters are not supported.
>   
>   &hwlist.nda;
>   
> +&hwlist.nvd;
> +
>   &hwlist.ocs_fc;

Hmm, it's not clear to me that nda/nvd are really hardware devices so much as software
drivers layered on top of a hardware abstraction (NVMe namespaces).  We wouldn't list
vlan(4) support as a distinct hardware feature in a list of NIC devices, and I feel
like listing da(4) or pass(4) or nda(4) or nvd(4) is similar to vlan(4).  In particular,
you can't hold an nda(4) or vlan(4) (or wlan(4)) in your hand the way you can hold an
M.2 NVMe device, or a NIC.  I think for the hardware notes, what people are interested
in is which pieces of hardware will work, not necessarily the software abstractions we
layer on top of them.

Similarly, you wouldn't list mfid(4) for the logical RAID volumes on an mfi(4) controller,
just that we support mfi(4) controllers.  The fact that you can access the RAID volumes
exported by such a controller is implied by supporting the controller.

-- 
John Baldwin