diff mbox series

[v2,net-next,08/12] net/sched: mqprio: add an extack message to mqprio_parse_opt()

Message ID 20230219135309.594188-9-vladimir.oltean@nxp.com
State New
Headers
Series Add tc-mqprio and tc-taprio support for preemptible traffic classes |

Commit Message

Vladimir Oltean Feb. 19, 2023, 1:53 p.m. UTC 
  Ferenc reports that a combination of poor iproute2 defaults and obscure
cases where the kernel returns -EINVAL make it difficult to understand
what is wrong with this command:

$ ip link add veth0 numtxqueues 8 numrxqueues 8 type veth peer name veth1
$ tc qdisc add dev veth0 root mqprio num_tc 8 map 0 1 2 3 4 5 6 7 \
        queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7
RTNETLINK answers: Invalid argument

Hopefully with this patch, the cause is clearer:

Error: Device does not support hardware offload.

The kernel was (and still is) rejecting this because iproute2 defaults
to "hw 1" if this command line option is not specified.

Link: https://patchwork.kernel.org/project/netdevbpf/patch/20230204135307.1036988-3-vladimir.oltean@nxp.com/#25215636
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
---
v1->v2: slightly reword last paragraph of commit message

 net/sched/sch_mqprio.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
diff mbox series

Patch

diff --git a/net/sched/sch_mqprio.c b/net/sched/sch_mqprio.c
index 5a9261c38b95..9ee5a9a9b9e9 100644
--- a/net/sched/sch_mqprio.c
+++ b/net/sched/sch_mqprio.c
@@ -133,8 +133,11 @@  static int mqprio_parse_opt(struct net_device *dev, struct tc_mqprio_qopt *qopt,
 	/* If ndo_setup_tc is not present then hardware doesn't support offload
 	 * and we should return an error.
 	 */
-	if (qopt->hw && !dev->netdev_ops->ndo_setup_tc)
+	if (qopt->hw && !dev->netdev_ops->ndo_setup_tc) {
+		NL_SET_ERR_MSG(extack,
+			       "Device does not support hardware offload");
 		return -EINVAL;
+	}
 
 	return 0;
 }