[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20241002071602.793d3e2d@kernel.org>
Date: Wed, 2 Oct 2024 07:16:02 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Aleksandr Mikhalitsyn <aleksandr.mikhalitsyn@...onical.com>
Cc: Stefano Garzarella <sgarzare@...hat.com>, stefanha@...hat.com, "Michael
S. Tsirkin" <mst@...hat.com>, Jason Wang <jasowang@...hat.com>, Eugenio
PĂ©rez <eperezma@...hat.com>, kvm@...r.kernel.org,
virtualization@...ts.linux.dev, netdev@...r.kernel.org,
linux-kernel@...r.kernel.org, mcgrof@...nel.org
Subject: Re: [PATCH v2] vhost/vsock: specify module version
On Mon, 30 Sep 2024 19:03:52 +0200 Aleksandr Mikhalitsyn wrote:
> > At this point my question is, should we solve the problem higher and
> > show all the modules in /sys/modules, either way?
>
> Probably, yes. We can ask Luis Chamberlain's opinion on this one.
>
> +cc Luis Chamberlain <mcgrof@...nel.org>
>
> >
> > Your use case makes sense to me, so that we could try something like
> > that, but obviously it requires more work I think.
>
> I personally am pretty happy to do more work on the generic side if
> it's really valuable
> for other use cases and folks support the idea.
IMHO a generic solution would be much better. I can't help but feel
like exposing an arbitrary version to get the module to show up in
sysfs is a hack.
IIUC the list of built in modules is available in
/lib/modules/*/modules.builtin, the user space can't read that?
Powered by blists - more mailing lists