From: Ido Schimmel <idosch@idosch.org>
To: Sriram Yagnaraman <sriram.yagnaraman@est.tech>
Cc: netdev@vger.kernel.org, linux-kselftest@vger.kernel.org,
"David S . Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
David Ahern <dsahern@kernel.org>,
Ido Schimmel <idosch@nvidia.com>, Shuah Khan <shuah@kernel.org>,
Petr Machata <petrm@nvidia.com>
Subject: Re: [PATCH net v4 3/3] selftests: fib_tests: Add multipath list receive tests
Date: Wed, 30 Aug 2023 18:48:46 +0300 [thread overview]
Message-ID: <ZO9k3kSVHpxO1owX@shredder> (raw)
In-Reply-To: <20230830091341.17273-4-sriram.yagnaraman@est.tech>
On Wed, Aug 30, 2023 at 11:13:41AM +0200, Sriram Yagnaraman wrote:
> +link_stats_get()
> +{
> + local ns=$1; shift
> + local dev=$1; shift
> + local dir=$1; shift
> + local stat=$1; shift
Indentation is off...
> +
> + ip -n $ns -j -s link show dev $dev \
> + | jq '.[]["stats64"]["'$dir'"]["'$stat'"]'
> +}
> +
> +list_rcv_eval()
> +{
> + local name=$1; shift
> + local file=$1; shift
> + local expected=$1; shift
> + local exp=$1; shift
Nit: You can drop the first and last arguments since they don't change
between both invocations.
> +
> +
Unnecessary blank line
> + local count=$(tail -n 1 $file | jq '.["counter-value"] | tonumber | floor')
> + local ratio=$(echo "scale=2; $count / $expected" | bc -l)
> + local res=$(echo "$ratio $exp" | bc)
> + [[ $res -eq 1 ]]
> + log_test $? 0 "$name route hit ratio ($ratio)"
> +}
> +
> +ipv4_mpath_list_test()
> +{
> + echo
> + echo "IPv4 multipath list receive tests"
> +
> + mpath_dep_check || return 1
> +
> + route_setup
> +
> + set -e
> + run_cmd "ip netns exec ns1 ethtool -K veth1 tcp-segmentation-offload off"
> +
> + run_cmd "ip netns exec ns2 bash -c \"echo 20000 > /sys/class/net/veth2/gro_flush_timeout\""
> + run_cmd "ip netns exec ns2 bash -c \"echo 1 > /sys/class/net/veth2/napi_defer_hard_irqs\""
> + run_cmd "ip netns exec ns2 ethtool -K veth2 generic-receive-offload on"
> + run_cmd "ip -n ns2 link add name nh1 up type dummy"
> + run_cmd "ip -n ns2 link add name nh2 up type dummy"
> + run_cmd "ip -n ns2 address add 172.16.201.1/24 dev nh1"
> + run_cmd "ip -n ns2 address add 172.16.202.1/24 dev nh2"
> + run_cmd "ip -n ns2 neigh add 172.16.201.2 lladdr 00:11:22:33:44:55 nud perm dev nh1"
> + run_cmd "ip -n ns2 neigh add 172.16.202.2 lladdr 00:aa:bb:cc:dd:ee nud perm dev nh2"
> + run_cmd "ip -n ns2 route add 203.0.113.0/24
> + nexthop via 172.16.201.2 nexthop via 172.16.202.2"
> + run_cmd "ip netns exec ns2 sysctl -qw net.ipv4.fib_multipath_hash_policy=1"
> + set +e
> +
> + local dmac=$(ip -n ns2 -j link show dev veth2 | jq -r '.[]["address"]')
> + local tmp_file=$(mktemp)
> + local cmd="ip netns exec ns1 mausezahn veth1 -a own -b $dmac
> + -A 172.16.101.1 -B 203.0.113.1 -t udp 'sp=12345,dp=0-65535' -q"
> +
> + # Packets forwarded in a list using a multipath route must not reuse a
> + # cached result so that a flow always hits the same nexthop. In other
> + # words, the FIB lookup tracepoint needs to be triggered for every
> + # packet.
> + local t0_rx_pkts=$(link_stats_get ns2 veth2 rx packets)
> + run_cmd "perf stat -e fib:fib_table_lookup --filter 'err == 0' -j -o $tmp_file -- $cmd"
> + local t1_rx_pkts=$(link_stats_get ns2 veth2 rx packets)
> + local diff=$(echo $t1_rx_pkts - $t0_rx_pkts | bc -l)
> + list_rcv_eval "Multipath" $tmp_file $diff ">= 0.95"
> +
> + rm $tmp_file
> + route_cleanup
> +}
> +
> +ipv6_mpath_list_test()
> +{
> + echo
> + echo "IPv6 multipath list receive tests"
> +
> + mpath_dep_check || return 1
> +
> + route_setup
> +
> + set -e
> + run_cmd "ip netns exec ns1 ethtool -K veth1 tcp-segmentation-offload off"
> +
> + run_cmd "ip netns exec ns2 bash -c \"echo 20000 > /sys/class/net/veth2/gro_flush_timeout\""
> + run_cmd "ip netns exec ns2 bash -c \"echo 1 > /sys/class/net/veth2/napi_defer_hard_irqs\""
> + run_cmd "ip netns exec ns2 ethtool -K veth2 generic-receive-offload on"
> + run_cmd "ip -n ns2 link add name nh1 up type dummy"
> + run_cmd "ip -n ns2 link add name nh2 up type dummy"
> + run_cmd "ip -n ns2 -6 address add 2001:db8:201::1/64 dev nh1"
> + run_cmd "ip -n ns2 -6 address add 2001:db8:202::1/64 dev nh2"
> + run_cmd "ip -n ns2 -6 neigh add 2001:db8:201::2 lladdr 00:11:22:33:44:55 nud perm dev nh1"
> + run_cmd "ip -n ns2 -6 neigh add 2001:db8:202::2 lladdr 00:aa:bb:cc:dd:ee nud perm dev nh2"
> + run_cmd "ip -n ns2 -6 route add 2001:db8:301::/64
> + nexthop via 2001:db8:201::2 nexthop via 2001:db8:202::2"
> + run_cmd "ip netns exec ns2 sysctl -qw net.ipv6.fib_multipath_hash_policy=1"
> + set +e
> +
> + local dmac=$(ip -n ns2 -j link show dev veth2 | jq -r '.[]["address"]')
> + local tmp_file=$(mktemp)
> + local cmd="ip netns exec ns1 mausezahn -6 veth1 -a own -b $dmac
> + -A 2001:db8:101::1 -B 2001:db8:301::1 -t udp 'sp=12345,dp=0-65535' -q"
> +
> + # Packets forwarded in a list using a multipath route must not reuse a
> + # cached result so that a flow always hits the same nexthop. In other
> + # words, the FIB lookup tracepoint needs to be triggered for every
> + # packet.
> + local t0_rx_pkts=$(link_stats_get ns2 veth2 rx packets)
> + run_cmd "perf stat -e fib6:fib6_table_lookup --filter 'err == 0' -j -o $tmp_file -- $cmd"
> + local t1_rx_pkts=$(link_stats_get ns2 veth2 rx packets)
> + local diff=$(echo $t1_rx_pkts - $t0_rx_pkts | bc -l)
> + list_rcv_eval "Multipath" $tmp_file $diff ">= 0.95"
> +
> + rm $tmp_file
> + route_cleanup
> +}
> +
> ################################################################################
> # usage
>
> @@ -2433,6 +2587,8 @@ do
> ipv6_mangle) ipv6_mangle_test;;
> ipv4_bcast_neigh) ipv4_bcast_neigh_test;;
> fib6_gc_test|ipv6_gc) fib6_gc_test;;
> + ipv4_mpath_list) ipv4_mpath_list_test;;
> + ipv6_mpath_list) ipv6_mpath_list_test;;
>
> help) echo "Test names: $TESTS"; exit 0;;
> esac
> --
> 2.34.1
>
>
prev parent reply other threads:[~2023-08-30 15:48 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-30 9:13 [PATCH net v4 0/3] Avoid TCP resets when using ECMP for load-balancing between multiple servers Sriram Yagnaraman
2023-08-30 9:13 ` [PATCH net v4 1/3] ipv4: ignore dst hint for multipath routes Sriram Yagnaraman
2023-08-30 9:13 ` [PATCH net v4 2/3] ipv6: " Sriram Yagnaraman
2023-08-30 9:13 ` [PATCH net v4 3/3] selftests: fib_tests: Add multipath list receive tests Sriram Yagnaraman
2023-08-30 15:48 ` Ido Schimmel [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZO9k3kSVHpxO1owX@shredder \
--to=idosch@idosch.org \
--cc=davem@davemloft.net \
--cc=dsahern@kernel.org \
--cc=edumazet@google.com \
--cc=idosch@nvidia.com \
--cc=kuba@kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=petrm@nvidia.com \
--cc=shuah@kernel.org \
--cc=sriram.yagnaraman@est.tech \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).