Thread (17 messages) flat view 17 messages, 2 authors, 2020-12-02

Re: [PATCH v2 net-next 4/4] net: dsa: tag_dsa: Support reception of packets from LAG devices

From: Vladimir Oltean <olteanv@gmail.com>
Date: 2020-12-02 00:34:45

On Tue, Dec 01, 2020 at 11:31:02PM +0100, Tobias Waldekranz wrote:
quoted
quoted
+	if (unlikely(!dsa_slave_dev_check(skb->dev))) {
+		/* Packet is to be injected directly on an upper
+		 * device, e.g. a team/bond, so skip all DSA-port
+		 * specific actions.
+		 */
+		netif_rx(skb);
+		return 0;
netif_rx returns an int code, it seems odd to ignore it.
This is exactly the same treatment that the return code from
gro_cells_receive gets just a few lines down. They return the same set
of codes (NET_RX_{SUCCESS,DROP}).

Looking through the source base, there are a few callers that look at
the return value (the overwhelming majority ignore it). Actions vary
from printing warnings (without rate-limit, yikes), setting variables
that are otherwise unused, or bumping a counter (the only reasonable
thing I have seen).

But looking through enqueue_to_backlog, it seems like there already is a
counter for this that is accessible from /proc/net/softnet_data.
And also, more obviously, in ndo_get_stats64 (ip -s -s link).

Ok, I think you can ignore the return value.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help