From: Edward Cree <ecree@solarflare.com>
Date: Fri, 22 Nov 2019 17:57:19 +0000
Subject: sfc: suppress MCDI errors from ARFS
Patch-mainline: v5.5-rc1
Git-commit: 0aa6608daec5d19275809d30df6a02529eb66c7d
References: jsc#SLE-8314

In high connection count usage, the NIC's filter table may be filled with
 sufficiently many ARFS filters that further insertions fail.  As this
 does not represent a correctness issue, do not log the resulting MCDI
 errors.  Add a debug-level message under the (by default disabled)
 rx_status category instead; and take the opportunity to do a little extra
 expiry work.

Since there are now multiple workitems able to call __efx_filter_rfs_expire
 on a given channel, it is possible for them to race and thus pass quotas
 which, combined, exceed rfs_filter_count.  Thus, don't WARN_ON if we loop
 all the way around the table with quota left over.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Tested-by: David Ahern <dahern@digitalocean.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Acked-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
---
 drivers/net/ethernet/sfc/ef10.c |    8 ++++++--
 drivers/net/ethernet/sfc/rx.c   |   28 ++++++++++++++++++++++++----
 2 files changed, 30 insertions(+), 6 deletions(-)

--- a/drivers/net/ethernet/sfc/ef10.c
+++ b/drivers/net/ethernet/sfc/ef10.c
@@ -4202,11 +4202,15 @@ static int efx_ef10_filter_push(struct e
 {
 	MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN);
 	MCDI_DECLARE_BUF(outbuf, MC_CMD_FILTER_OP_EXT_OUT_LEN);
+	size_t outlen;
 	int rc;
 
 	efx_ef10_filter_push_prep(efx, spec, inbuf, *handle, ctx, replacing);
-	rc = efx_mcdi_rpc(efx, MC_CMD_FILTER_OP, inbuf, sizeof(inbuf),
-			  outbuf, sizeof(outbuf), NULL);
+	rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, inbuf, sizeof(inbuf),
+				outbuf, sizeof(outbuf), &outlen);
+	if (rc && spec->priority != EFX_FILTER_PRI_HINT)
+		efx_mcdi_display_error(efx, MC_CMD_FILTER_OP, sizeof(inbuf),
+				       outbuf, outlen, rc);
 	if (rc == 0)
 		*handle = MCDI_QWORD(outbuf, FILTER_OP_OUT_HANDLE);
 	if (rc == -ENOSPC)
--- a/drivers/net/ethernet/sfc/rx.c
+++ b/drivers/net/ethernet/sfc/rx.c
@@ -1032,6 +1032,26 @@ static void efx_filter_rfs_work(struct w
 				   req->spec.rem_host, ntohs(req->spec.rem_port),
 				   req->spec.loc_host, ntohs(req->spec.loc_port),
 				   req->rxq_index, req->flow_id, rc, arfs_id);
+	} else {
+		if (req->spec.ether_type == htons(ETH_P_IP))
+			netif_dbg(efx, rx_status, efx->net_dev,
+				  "failed to steer %s %pI4:%u:%pI4:%u to queue %u [flow %u rc %d id %u]\n",
+				  (req->spec.ip_proto == IPPROTO_TCP) ? "TCP" : "UDP",
+				  req->spec.rem_host, ntohs(req->spec.rem_port),
+				  req->spec.loc_host, ntohs(req->spec.loc_port),
+				  req->rxq_index, req->flow_id, rc, arfs_id);
+		else
+			netif_dbg(efx, rx_status, efx->net_dev,
+				  "failed to steer %s [%pI6]:%u:[%pI6]:%u to queue %u [flow %u rc %d id %u]\n",
+				  (req->spec.ip_proto == IPPROTO_TCP) ? "TCP" : "UDP",
+				  req->spec.rem_host, ntohs(req->spec.rem_port),
+				  req->spec.loc_host, ntohs(req->spec.loc_port),
+				  req->rxq_index, req->flow_id, rc, arfs_id);
+		/* We're overloading the NIC's filter tables, so let's do a
+		 * chunk of extra expiry work.
+		 */
+		__efx_filter_rfs_expire(channel, min(channel->rfs_filter_count,
+						     100u));
 	}
 
 	/* Release references */
@@ -1170,11 +1190,11 @@ bool __efx_filter_rfs_expire(struct efx_
 		if (++index == size)
 			index = 0;
 		/* If we were called with a quota that exceeds the total number
-		 * of filters in the table (which should never happen), ensure
-		 * that we don't loop forever - stop when we've examined every
-		 * row of the table.
+		 * of filters in the table (which shouldn't happen, but could
+		 * if two callers race), ensure that we don't loop forever -
+		 * stop when we've examined every row of the table.
 		 */
-		if (WARN_ON(index == start && quota))
+		if (index == start)
 			break;
 	}