On 4 Feb 2026, at 23:42, Ilya Maximets wrote:
> On 2/2/26 11:29 AM, Eelco Chaudron wrote:
>> Previously, dpif_offload_set_config() was skipped if the dpif
>> provider did not implement the corresponding class function.
>> This was incorrect: the offload configuration function should be
>> called regardless of whether the dpif-provider-specific function
>> exists.
>>
>> Fixes: bd6543466dc6 ("dpif-offload: Add offload provider set_config API.")
>> Signed-off-by: Eelco Chaudron <[email protected]>
>> ---
>>
>> v2: Added a self-test.
>> ---
>> lib/dpif.c | 2 +-
>> tests/system-offloads-traffic.at | 35 ++++++++++++++++++++++++++++++++
>> 2 files changed, 36 insertions(+), 1 deletion(-)
>>
>> diff --git a/lib/dpif.c b/lib/dpif.c
>> index 88b5edfad5..cab5884254 100644
>> --- a/lib/dpif.c
>> +++ b/lib/dpif.c
>> @@ -1609,8 +1609,8 @@ dpif_set_config(struct dpif *dpif, const struct smap
>> *cfg)
>> if (error) {
>> log_operation(dpif, "set_config", error);
>> }
>> - dpif_offload_set_config(dpif, cfg);
>> }
>> + dpif_offload_set_config(dpif, cfg);
>>
>> return error;
>> }
>> diff --git a/tests/system-offloads-traffic.at
>> b/tests/system-offloads-traffic.at
>> index ba1f9fe5e8..6e71f672f9 100644
>> --- a/tests/system-offloads-traffic.at
>> +++ b/tests/system-offloads-traffic.at
>> @@ -93,6 +93,41 @@ AT_CHECK([ovs-appctl upcall/show | grep -E "offloaded
>> flows : [[1-9]]"], [0], [i
>> OVS_TRAFFIC_VSWITCHD_STOP
>> AT_CLEANUP
>>
>> +AT_SETUP([offloads - tc-policy configuration])
>> +OVS_TRAFFIC_VSWITCHD_START([], [],
>> + [-- set Open_vSwitch . other_config:hw-offload=true \
>> + -- set Open_vSwitch . other_config:tc-policy=skip_hw])
>> +
>> +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"])
>> +
>> +ADD_NAMESPACES(at_ns0, at_ns1)
>> +
>> +ADD_VETH(p0, at_ns0, br0, "10.1.1.1/24")
>> +ADD_VETH(p1, at_ns1, br0, "10.1.1.2/24")
>> +AT_CHECK([ovs-appctl dpctl/dump-flows], [0], [ignore])
>> +
>> +NS_CHECK_EXEC([at_ns0], [ping -q -c 10 -i 0.1 -W 2 10.1.1.2 | FORMAT_PING],
>> [0], [dnl
>> +10 packets transmitted, 10 received, 0% packet loss, time 0ms
>> +])
>> +
>> +AT_CHECK([ovs-appctl dpctl/dump-flows | grep "eth_type(0x0800)" |
>> DUMP_CLEAN_SORTED], [0], [dnl
>> +in_port(2),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756,
>> used:0.001s, actions:output
>> +in_port(3),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756,
>> used:0.001s, actions:output
>> +])
>> +
>> +AT_CHECK([ovs-appctl dpctl/dump-flows type=ovs | grep "eth_type(0x0800)" |
>> DUMP_CLEAN_SORTED], [0], [])
>> +
>> +AT_CHECK([ovs-appctl dpctl/dump-flows type=tc,offloaded | grep
>> "eth_type(0x0800)" | DUMP_CLEAN_SORTED], [0], [dnl
>> +in_port(2),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756,
>> used:0.001s, actions:output
>> +in_port(3),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756,
>> used:0.001s, actions:output
>> +])
>> +
>> +AT_CHECK([ovs-appctl upcall/show | grep -E "offloaded flows : [[1-9]]"],
>> [0], [ignore])
>> +AT_CHECK([tc -d filter show dev ovs-p0 ingress | grep -q "skip_hw"], [0])
>
> nit: The [0] at the end is redundant, it's the default.
>
> Otherwise, LGTM!
>
> Acked-by: Ilya Maximets <[email protected]>
Thanks Ilya for the review. With nit fixed, applied to main and branch-3.7.
//Eelco
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
