From mboxrd@z Thu Jan  1 00:00:00 1970
From: Stefano Brivio <sbrivio@redhat.com>
To: passt-dev@passt.top
Subject: Re: More mbuto pain
Date: Tue, 28 Jun 2022 09:17:22 +0200
Message-ID: <20220628091722.39572330@elisabeth>
In-Reply-To: <YrqSKzP600gzFjB4@yekko>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============6800582103277221209=="

--===============6800582103277221209==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit

[Adding Sevinj... if you have some time to have a look]

On Tue, 28 Jun 2022 15:31:23 +1000
David Gibson <david(a)gibson.dropbear.id.au> wrote:

> Now that the mbuto profile change is around, I tried getting the passt
> tests working.  I fixed a couple of minor problems, before hitting
> another mbuto issue.
> 
> On my host, mbuto isn't managing to find and include the virtio_net
> module.  I think this is because on current Fedora the modules in
> /lib/modules are compressed.  So it needs to include:
>     /lib/modules/5.18.6-200.fc36.x86_64/kernel/drivers/net/virtio_net.ko.xz
> but I think it's only looking for a plain '.ko' file.

Correct. I guess all it takes is to accept .ko, .ko.gz, .ko.xz, perhaps
something else in these expressions:

	__t="$(${BASENAME} -- "${__t%*.ko}")"

	__src="$(${FIND} ${__find_path}	-name "${__t}.ko")"

in kmod_add(), but I haven't tried yet.

> I had a look at the mbuto code to see if I could fix it, but I'm
> afraid mbuto is a long way past the complexity level at which I find
> shell remotely comprehensible.

-- 
Stefano


--===============6800582103277221209==--