netdev - Re: [PATCH net-next v2 10/12] selftests: ncdevmem: Run selftest when none of the -s or -c has been provided

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]

Message-ID: <Zv7R_dRFZ0VGindy@mini-arch>
Date: Thu, 3 Oct 2024 10:18:53 -0700
From: Stanislav Fomichev <stfomichev@...il.com>
To: Mina Almasry <almasrymina@...gle.com>
Cc: Stanislav Fomichev <sdf@...ichev.me>, netdev@...r.kernel.org,
	davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org,
	pabeni@...hat.com
Subject: Re: [PATCH net-next v2 10/12] selftests: ncdevmem: Run selftest when
 none of the -s or -c has been provided

On 10/03, Mina Almasry wrote:
> On Mon, Sep 30, 2024 at 10:18 AM Stanislav Fomichev <sdf@...ichev.me> wrote:
> >
> > This will be used as a 'probe' mode in the selftest to check whether
> > the device supports the devmem or not.
> 
> It's not really 'probing'. Running ncdevmem with -s or -c does the
> data path tests. Running ncdevmem without does all the control path
> tests in run_devmem_tests(). It's not just probing driver for support,
> it's mean to catch bugs. Maybe rename to 'control path tests' or
> 'config tests' instead of probing.

Re 'probing' vs 'control path tests': I need something I can call
in the python selftest to tell me whether the nic supports devmem
or not (to skip the tests on the unsupported nics), so I'm reusing this
'control path tests' mode for this. I do agree that there might be an
issue where the nic supports devmem, but has some bug which causes
'control path tests' to fail which leads to skipping the data plane tests...

We can try to separate these two in the future. (and I'll keep the word
'probing' for now since it's only in the commit message to describe the
intent)

> > Use hard-coded queue layout
> > (two last queues) and prevent user from passing custom -q and/or -t.
> >
> > Cc: Mina Almasry <almasrymina@...gle.com>
> > Signed-off-by: Stanislav Fomichev <sdf@...ichev.me>
> > ---
> >  tools/testing/selftests/net/ncdevmem.c | 27 +++++++++++++++++++-------
> >  1 file changed, 20 insertions(+), 7 deletions(-)
> >
> > diff --git a/tools/testing/selftests/net/ncdevmem.c b/tools/testing/selftests/net/ncdevmem.c
> > index 900a661a61af..9b0a81b12eac 100644
> > --- a/tools/testing/selftests/net/ncdevmem.c
> > +++ b/tools/testing/selftests/net/ncdevmem.c
> > @@ -688,17 +688,26 @@ int main(int argc, char *argv[])
> >                 }
> >         }
> >
> > -       if (!server_ip)
> > -               error(1, 0, "Missing -s argument\n");
> > -
> > -       if (!port)
> > -               error(1, 0, "Missing -p argument\n");
> > -
> >         if (!ifname)
> >                 error(1, 0, "Missing -f argument\n");
> >
> >         ifindex = if_nametoindex(ifname);
> >
> > +       if (!server_ip && !client_ip) {
> > +               if (start_queue != -1)
> > +                       error(1, 0, "don't support custom start queue for probing\n");
> > +               if (num_queues != 1)
> > +                       error(1, 0, "don't support custom number of queues for probing\n");
> > +
> 
> Remove the start_queue + num_queues check here please. I would like to
> be able to run the control path tests binding dmabuf to all the queues
> or 1 of the queues or some of the queues.

Will remove, but let's continue discussing this in the other thread.