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: <20180425185601.55511-4-willemdebruijn.kernel@gmail.com>
Date:   Wed, 25 Apr 2018 14:55:54 -0400
From:   Willem de Bruijn <willemdebruijn.kernel@...il.com>
To:     netdev@...r.kernel.org
Cc:     davem@...emloft.net, alexander.duyck@...il.com,
        dmichail@...gle.com, Willem de Bruijn <willemb@...gle.com>
Subject: [PATCH net-next 03/10] udp: better wmem accounting on gso

From: Willem de Bruijn <willemb@...gle.com>

skb_segment by default transfers allocated wmem from the gso skb
to the tail of the segment list. This underreports real truesize
of the list, especially if the tail might be dropped.

Similar to tcp_gso_segment, update wmem_alloc with the aggregate
list truesize and make each segment responsible for its own
share by setting skb->destructor.

Clear gso_skb->destructor prior to calling skb_segment to skip
the default assignment to tail.

Signed-off-by: Willem de Bruijn <willemb@...gle.com>
---
 net/ipv4/udp_offload.c | 20 ++++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)

diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c
index a3244768e45f..acfef71459f8 100644
--- a/net/ipv4/udp_offload.c
+++ b/net/ipv4/udp_offload.c
@@ -192,7 +192,9 @@ struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
 				  unsigned int mss, __sum16 check)
 {
 	struct udphdr *uh = udp_hdr(gso_skb);
-	struct sk_buff *segs;
+	struct sock *sk = gso_skb->sk;
+	struct sk_buff *segs, *seg;
+	unsigned int sum_truesize = 0;
 	unsigned int hdrlen;
 
 	if (gso_skb->len <= sizeof(*uh) + mss)
@@ -203,9 +205,23 @@ struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
 	skb_pull(gso_skb, sizeof(*uh));
 	hdrlen = gso_skb->data - skb_mac_header(gso_skb);
 
+	/* clear destructor to avoid skb_segment assigning it to tail */
+	WARN_ON_ONCE(gso_skb->destructor != sock_wfree);
+	gso_skb->destructor = NULL;
+
 	segs = skb_segment(gso_skb, features);
-	if (unlikely(IS_ERR_OR_NULL(segs)))
+	if (unlikely(IS_ERR_OR_NULL(segs))) {
+		gso_skb->destructor = sock_wfree;
 		return segs;
+	}
+
+	for (seg = segs; seg; seg = seg->next) {
+		seg->destructor = sock_wfree;
+		seg->sk = sk;
+		sum_truesize += seg->truesize;
+	}
+
+	refcount_add(sum_truesize - gso_skb->truesize, &sk->sk_wmem_alloc);
 
 	/* If last packet is not full, fix up its header */
 	if (segs->prev->len != hdrlen + mss) {
-- 
2.17.0.441.gb46fe60e1d-goog

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ