netdev - [PATCH v2 net 3/3] qede: Fix single MTU sized packet from firmware GRO flow

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1461135809-9776-4-git-send-email-manish.chopra@qlogic.com>
Date:	Wed, 20 Apr 2016 03:03:29 -0400
From:	Manish Chopra <manish.chopra@...gic.com>
To:	<davem@...emloft.net>
CC:	<netdev@...r.kernel.org>, <Ariel.Elior@...gic.com>,
	<Yuval.Mintz@...gic.com>
Subject: [PATCH v2 net 3/3] qede: Fix single MTU sized packet from firmware GRO flow

In firmware assisted GRO flow there could be a single MTU sized
segment arriving due to firmware aggregation timeout/last segment
in an aggregation flow, which is not expected to be an actual gro
packet. So If a skb has zero frags from the GRO flow then simply
push it in the stack as non gso skb.

Signed-off-by: Manish Chopra <manish.chopra@...gic.com>
Signed-off-by: Yuval Mintz <yuval.mintz@...gic.com>
---
 drivers/net/ethernet/qlogic/qede/qede_main.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/drivers/net/ethernet/qlogic/qede/qede_main.c b/drivers/net/ethernet/qlogic/qede/qede_main.c
index bf0fb99..7869465 100644
--- a/drivers/net/ethernet/qlogic/qede/qede_main.c
+++ b/drivers/net/ethernet/qlogic/qede/qede_main.c
@@ -1069,6 +1069,17 @@ static void qede_gro_receive(struct qede_dev *edev,
 			     struct sk_buff *skb,
 			     u16 vlan_tag)
 {
+	/* FW can send a single MTU sized packet from gro flow
+	 * due to aggregation timeout/last segment etc. which
+	 * is not expected to be a gro packet. If a skb has zero
+	 * frags then simply push it in the stack as non gso skb.
+	 */
+	if (unlikely(!skb->data_len)) {
+		skb_shinfo(skb)->gso_type = 0;
+		skb_shinfo(skb)->gso_size = 0;
+		goto send_skb;
+	}
+
 #ifdef CONFIG_INET
 	if (skb_shinfo(skb)->gso_size) {
 		skb_set_network_header(skb, 0);
@@ -1087,6 +1098,8 @@ static void qede_gro_receive(struct qede_dev *edev,
 		}
 	}
 #endif
+
+send_skb:
 	skb_record_rx_queue(skb, fp->rss_id);
 	qede_skb_receive(edev, fp, skb, vlan_tag);
 }
-- 
2.7.2