netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Xin Long <lucien.xin@gmail.com>
To: network dev <netdev@vger.kernel.org>,
	davem@davemloft.net, kuba@kernel.org,
	Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>,
	linux-sctp@vger.kernel.org
Subject: [PATCH net-next 09/14] sctp: do state transition when receiving an icmp TOOBIG packet
Date: Sun, 20 Jun 2021 21:38:44 -0400	[thread overview]
Message-ID: <079e91bfe9ea9f081df75d04e74253b099ce7a84.1624239422.git.lucien.xin@gmail.com> (raw)
In-Reply-To: <cover.1624239422.git.lucien.xin@gmail.com>

PLPMTUD will short-circuit the old process for icmp TOOBIG packets.
This part is described in rfc8899#section-4.6.2 (PL_PTB_SIZE =
PTB_SIZE - other_headers_len). Note that from rfc8899#section-5.2
State Machine, each case below is for some specific states only:

  a) PL_PTB_SIZE < MIN_PLPMTU || PL_PTB_SIZE >= PROBED_SIZE,
     discard it, for any state

  b) MIN_PLPMTU < PL_PTB_SIZE < BASE_PLPMTU,
     Base -> Error, for Base state

  c) BASE_PLPMTU <= PL_PTB_SIZE < PLPMTU,
     Search -> Base or Complete -> Base, for Search and Complete states.

  d) PLPMTU < PL_PTB_SIZE < PROBED_SIZE,
     set pl.probe_size to PL_PTB_SIZE then verify it, for Search state.

The most important one is case d), which will help find the optimal
fast during searching. Like when pathmtu = 1392 for SCTP over IPv4,
the search will be (20 is iphdr_len):

  1. probe with 1200 - 20
  2. probe with 1232 - 20
  3. probe with 1264 - 20
  ...
  7. probe with 1388 - 20
  8. probe with 1420 - 20

When sending the probe with 1420 - 20, TOOBIG may come with PL_PTB_SIZE =
1392 - 20. Then it matches case d), and saves some rounds to try with the
1392 - 20 probe. But of course, PLPMTUD doesn't trust TOOBIG packets, and
it will go back to the common searching once the probe with the new size
can't be verified.

Signed-off-by: Xin Long <lucien.xin@gmail.com>
---
 net/sctp/input.c     |  4 +++-
 net/sctp/transport.c | 51 +++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 53 insertions(+), 2 deletions(-)

diff --git a/net/sctp/input.c b/net/sctp/input.c
index d508f6f3dd08..9ffdbd6526e9 100644
--- a/net/sctp/input.c
+++ b/net/sctp/input.c
@@ -385,7 +385,9 @@ static int sctp_add_backlog(struct sock *sk, struct sk_buff *skb)
 void sctp_icmp_frag_needed(struct sock *sk, struct sctp_association *asoc,
 			   struct sctp_transport *t, __u32 pmtu)
 {
-	if (!t || (t->pathmtu <= pmtu))
+	if (!t ||
+	    (t->pathmtu <= pmtu &&
+	     t->pl.probe_size + sctp_transport_pl_hlen(t) <= pmtu))
 		return;
 
 	if (sock_owned_by_user(sk)) {
diff --git a/net/sctp/transport.c b/net/sctp/transport.c
index 79ff5ca6b472..5cefb4eab8a0 100644
--- a/net/sctp/transport.c
+++ b/net/sctp/transport.c
@@ -343,10 +343,55 @@ void sctp_transport_pl_recv(struct sctp_transport *t)
 	}
 }
 
+static bool sctp_transport_pl_toobig(struct sctp_transport *t, u32 pmtu)
+{
+	pr_debug("%s: PLPMTUD: transport: %p, state: %d, pmtu: %d, size: %d, ptb: %d\n",
+		 __func__, t, t->pl.state, t->pl.pmtu, t->pl.probe_size, pmtu);
+
+	if (pmtu < SCTP_MIN_PLPMTU || pmtu >= t->pl.probe_size)
+		return false;
+
+	if (t->pl.state == SCTP_PL_BASE) {
+		if (pmtu >= SCTP_MIN_PLPMTU && pmtu < SCTP_BASE_PLPMTU) {
+			t->pl.state = SCTP_PL_ERROR; /* Base -> Error */
+
+			t->pl.pmtu = SCTP_MIN_PLPMTU;
+			t->pathmtu = t->pl.pmtu + sctp_transport_pl_hlen(t);
+		}
+	} else if (t->pl.state == SCTP_PL_SEARCH) {
+		if (pmtu >= SCTP_BASE_PLPMTU && pmtu < t->pl.pmtu) {
+			t->pl.state = SCTP_PL_BASE;  /* Search -> Base */
+			t->pl.probe_size = SCTP_BASE_PLPMTU;
+			t->pl.probe_count = 0;
+
+			t->pl.probe_high = 0;
+			t->pl.pmtu = SCTP_BASE_PLPMTU;
+			t->pathmtu = t->pl.pmtu + sctp_transport_pl_hlen(t);
+		} else if (pmtu > t->pl.pmtu && pmtu < t->pl.probe_size) {
+			t->pl.probe_size = pmtu;
+			t->pl.probe_count = 0;
+
+			return false;
+		}
+	} else if (t->pl.state == SCTP_PL_COMPLETE) {
+		if (pmtu >= SCTP_BASE_PLPMTU && pmtu < t->pl.pmtu) {
+			t->pl.state = SCTP_PL_BASE;  /* Complete -> Base */
+			t->pl.probe_size = SCTP_BASE_PLPMTU;
+			t->pl.probe_count = 0;
+
+			t->pl.probe_high = 0;
+			t->pl.pmtu = SCTP_BASE_PLPMTU;
+			t->pathmtu = t->pl.pmtu + sctp_transport_pl_hlen(t);
+		}
+	}
+
+	return true;
+}
+
 bool sctp_transport_update_pmtu(struct sctp_transport *t, u32 pmtu)
 {
-	struct dst_entry *dst = sctp_transport_dst_check(t);
 	struct sock *sk = t->asoc->base.sk;
+	struct dst_entry *dst;
 	bool change = true;
 
 	if (unlikely(pmtu < SCTP_DEFAULT_MINSEGMENT)) {
@@ -357,6 +402,10 @@ bool sctp_transport_update_pmtu(struct sctp_transport *t, u32 pmtu)
 	}
 	pmtu = SCTP_TRUNC4(pmtu);
 
+	if (sctp_transport_pl_enabled(t))
+		return sctp_transport_pl_toobig(t, pmtu - sctp_transport_pl_hlen(t));
+
+	dst = sctp_transport_dst_check(t);
 	if (dst) {
 		struct sctp_pf *pf = sctp_get_pf_specific(dst->ops->family);
 		union sctp_addr addr;
-- 
2.27.0


  parent reply	other threads:[~2021-06-21  1:39 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-21  1:38 [PATCH net-next 00/14] sctp: implement RFC8899: Packetization Layer Path MTU Discovery for SCTP transport Xin Long
2021-06-21  1:38 ` [PATCH net-next 01/14] sctp: add pad chunk and its make function and event table Xin Long
2021-06-21  1:38 ` [PATCH net-next 02/14] sctp: add probe_interval in sysctl and sock/asoc/transport Xin Long
2021-06-21  1:38 ` [PATCH net-next 03/14] sctp: add SCTP_PLPMTUD_PROBE_INTERVAL sockopt for sock/asoc/transport Xin Long
2021-06-21  1:38 ` [PATCH net-next 04/14] sctp: add the constants/variables and states and some APIs for transport Xin Long
2021-06-21  1:38 ` [PATCH net-next 05/14] sctp: add the probe timer in transport for PLPMTUD Xin Long
2021-06-21  1:38 ` [PATCH net-next 06/14] sctp: do the basic send and recv for PLPMTUD probe Xin Long
2021-06-21  3:49   ` kernel test robot
2021-06-22  1:13     ` Xin Long
2021-06-22 17:02       ` David Miller
2021-06-21  1:38 ` [PATCH net-next 07/14] sctp: do state transition when PROBE_COUNT == MAX_PROBES on HB send path Xin Long
2021-06-21  1:38 ` [PATCH net-next 08/14] sctp: do state transition when a probe succeeds on HB ACK recv path Xin Long
2021-06-21  1:38 ` Xin Long [this message]
2021-06-21  1:38 ` [PATCH net-next 10/14] sctp: enable PLPMTUD when the transport is ready Xin Long
2021-06-21  1:38 ` [PATCH net-next 11/14] sctp: remove the unessessary hold for idev in sctp_v6_err Xin Long
2021-06-21  1:38 ` [PATCH net-next 12/14] sctp: extract sctp_v6_err_handle function from sctp_v6_err Xin Long
2021-06-21  1:38 ` [PATCH net-next 13/14] sctp: extract sctp_v4_err_handle function from sctp_v4_err Xin Long
2021-06-21  1:38 ` [PATCH net-next 14/14] sctp: process sctp over udp icmp err on sctp side Xin Long
2021-06-22  1:30 ` [PATCH net-next 00/14] sctp: implement RFC8899: Packetization Layer Path MTU Discovery for SCTP transport Marcelo Ricardo Leitner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=079e91bfe9ea9f081df75d04e74253b099ce7a84.1624239422.git.lucien.xin@gmail.com \
    --to=lucien.xin@gmail.com \
    --cc=davem@davemloft.net \
    --cc=kuba@kernel.org \
    --cc=linux-sctp@vger.kernel.org \
    --cc=marcelo.leitner@gmail.com \
    --cc=netdev@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).