netdev - Re: [PATCH bpf-next] selftests: xsk: changes for setting up NICs to run xsk self-tests

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] [day] [month] [year] [list]

Date:   Wed, 30 Nov 2022 13:13:52 +0100
From:   Magnus Karlsson <magnus.karlsson@...il.com>
To:     Maciej Fijalkowski <maciej.fijalkowski@...el.com>,
        Saeed Mahameed <saeed@...nel.org>
Cc:     Tirthendu Sarkar <tirthendu.sarkar@...el.com>, bjorn@...nel.org,
        magnus.karlsson@...el.com, jonathan.lemon@...il.com,
        ast@...nel.org, daniel@...earbox.net, netdev@...r.kernel.org,
        bpf@...r.kernel.org
Subject: Re: [PATCH bpf-next] selftests: xsk: changes for setting up NICs to
 run xsk self-tests

On Wed, Nov 30, 2022 at 1:09 PM Maciej Fijalkowski
<maciej.fijalkowski@...el.com> wrote:
>
> On Wed, Nov 30, 2022 at 03:11:42PM +0530, Tirthendu Sarkar wrote:
> > ETH devies need to be set up for running xsk self-tests, like enable
> > loopback, set promiscuous mode, MTU etc. This patch adds those settings
> > before running xsk self-tests and reverts them back once done.
> >
> > Signed-off-by: Tirthendu Sarkar <tirthendu.sarkar@...el.com>
> > ---
> >  tools/testing/selftests/bpf/test_xsk.sh | 27 ++++++++++++++++++++++++-
> >  1 file changed, 26 insertions(+), 1 deletion(-)
> >
> > diff --git a/tools/testing/selftests/bpf/test_xsk.sh b/tools/testing/selftests/bpf/test_xsk.sh
> > index d821fd098504..e7a5c5fc4f71 100755
> > --- a/tools/testing/selftests/bpf/test_xsk.sh
> > +++ b/tools/testing/selftests/bpf/test_xsk.sh
> > @@ -106,7 +106,11 @@ MTU=1500
> >  trap ctrl_c INT
> >
> >  function ctrl_c() {
> > -        cleanup_exit ${VETH0} ${VETH1} ${NS1}
> > +     if [ ! -z $ETH ]; then
> > +             cleanup_exit ${VETH0} ${VETH1} ${NS1}
> > +     else
> > +             cleanup_eth
> > +     fi
> >       exit 1
> >  }
> >
> > @@ -138,9 +142,28 @@ setup_vethPairs() {
> >       ip link set ${VETH0} up
> >  }
> >
> > +setup_eth() {
> > +       sudo ethtool -L ${ETH} combined 1
>
> what if particular device has a different way of configuring single
> channel? like
>
> $ sudo ethtool -L ${ETH} tx 1 rx 1
>
> I am not sure if we want to proceed with settings that are specific to
> Intel devices. What if mlx5 will this in a much different way?

Adding Saeed as he will know. How does Mellanox set the number of
channels to 1 and how do you enable loopback mode? At least the rest
should be the same.

> > +       sudo ethtool -K ${ETH} loopback on
> > +       sudo ip link set ${ETH} promisc on
> > +       sudo ip link set ${ETH} mtu ${MTU}
> > +       sudo ip link set ${ETH} up
> > +       IPV6_DISABLE_CMD="sudo sysctl -n net.ipv6.conf.${ETH}.disable_ipv6"
> > +       IPV6_DISABLE=`$IPV6_DISABLE_CMD 2> /dev/null`
> > +       [[ $IPV6_DISABLE == "0" ]] && $IPV6_DISABLE_CMD=1
> > +       sleep 1
> > +}
> > +
> > +cleanup_eth() {
> > +       [[ $IPV6_DISABLE == "0" ]] && $IPV6_DISABLE_CMD=0
> > +       sudo ethtool -K ${ETH} loopback off
> > +       sudo ip link set ${ETH} promisc off
> > +}
> > +
> >  if [ ! -z $ETH ]; then
> >       VETH0=${ETH}
> >       VETH1=${ETH}
> > +     setup_eth
> >       NS1=""
> >  else
> >       validate_root_exec
> > @@ -191,6 +214,8 @@ exec_xskxceiver
> >
> >  if [ -z $ETH ]; then
> >       cleanup_exit ${VETH0} ${VETH1} ${NS1}
> > +else
> > +     cleanup_eth
> >  fi
> >
> >  failures=0
> > --
> > 2.34.1
> >