netdev - Re: [PATCH RFC v2 virtio 4/7] pds_vdpa: add vdpa config client commands

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 17 Mar 2023 11:36:28 +0800
From:   Jason Wang <jasowang@...hat.com>
To:     Shannon Nelson <shannon.nelson@....com>
Cc:     mst@...hat.com, virtualization@...ts.linux-foundation.org,
        brett.creeley@....com, davem@...emloft.net, netdev@...r.kernel.org,
        kuba@...nel.org, drivers@...sando.io
Subject: Re: [PATCH RFC v2 virtio 4/7] pds_vdpa: add vdpa config client commands

On Thu, Mar 16, 2023 at 11:25 AM Shannon Nelson <shannon.nelson@....com> wrote:
>
> On 3/15/23 12:05 AM, Jason Wang wrote:
> > On Thu, Mar 9, 2023 at 9:31 AM Shannon Nelson <shannon.nelson@....com> wrote:
> >>
> >> These are the adminq commands that will be needed for
> >> setting up and using the vDPA device.
> >
> > It's better to explain under which case the driver should use adminq,
> > I see some functions overlap with common configuration capability.
> > More below.
>
> Yes, I agree this needs to be more clearly stated.  The overlap is
> because the original FW didn't have the virtio device as well modeled
> and we had to go through adminq calls to get things done.

Does this mean the device could be actually probed by a virtio-pci driver?

>  Now that we
> have a reasonable virtio emulation and can use the virtio_net_config, we
> have a lot less need for the adminq calls.

Please add those in the changelog. Btw, adminq should be more flexible
since it's easier to extend for new features. If there's no plan to
model a virtio-pci driver we can even avoid mapping PCI capabilities
which may simplify the codes.

Thanks