tcp: allow for bigger reordering level

While testing upcoming Yaogong patch (converting out of order queue
into an RB tree), I hit the max reordering level of linux TCP stack.

Reordering level was limited to 127 for no good reason, and some
network setups [1] can easily reach this limit and get limited
throughput.

Allow a new max limit of 300, and add a sysctl to allow admins to even
allow bigger (or lower) values if needed.

[1] Aggregation of links, per packet load balancing, fabrics not doing
 deep packet inspections, alternative TCP congestion modules...

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yaogong Wang <wygivan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

6 files changed, 23 insertions, 12 deletions

diff --git a/Documentation/networking/bonding.txt b/Documentation/networking/bonding.txt
index eeb5b2e97bed..83bf4986baea 100644
--- a/Documentation/networking/bonding.txt
+++ b/Documentation/networking/bonding.txt
@@ -2230,11 +2230,8 @@ balance-rr: This mode is the only mode that will permit a single
 
         It is possible to adjust TCP/IP's congestion limits by
         altering the net.ipv4.tcp_reordering sysctl parameter.  The
-        usual default value is 3, and the maximum useful value is 127.
-        For a four interface balance-rr bond, expect that a single
-        TCP/IP stream will utilize no more than approximately 2.3
-        interface's worth of throughput, even after adjusting
-        tcp_reordering.
+        usual default value is 3. But keep in mind TCP stack is able
+        to automatically increase this when it detects reorders.
 
         Note that the fraction of packets that will be delivered out of
         order is highly variable, and is unlikely to be zero.  The level
diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index 0307e2875f21..9028b879a97b 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -376,9 +376,17 @@ tcp_orphan_retries - INTEGER
         may consume significant resources. Cf. tcp_max_orphans.
 
 tcp_reordering - INTEGER
-        Maximal reordering of packets in a TCP stream.
+        Initial reordering level of packets in a TCP stream.
+        TCP stack can then dynamically adjust flow reordering level
+        between this initial value and tcp_max_reordering
         Default: 3
 
+tcp_max_reordering - INTEGER
+        Maximal reordering level of packets in a TCP stream.
+        300 is a fairly conservative value, but you might increase it
+        if paths are using per packet load balancing (like bonding rr mode)
+        Default: 300
+
 tcp_retrans_collapse - BOOLEAN
         Bug-to-bug compatibility with some broken printers.
         On retransmit try to send bigger packets to work around bugs in
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index c2dee7deefa8..f566b8567892 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -204,10 +204,10 @@ struct tcp_sock {
 
         u16     urg_data;       /* Saved octet of OOB data and control flags */
         u8      ecn_flags;      /* ECN status bits.                     */
-        u8      reordering;     /* Packet reordering metric.            */
+        u8      keepalive_probes; /* num of allowed keep alive probes   */
+        u32     reordering;     /* Packet reordering metric.            */
         u32     snd_up;         /* Urgent pointer               */
 
-        u8      keepalive_probes; /* num of allowed keep alive probes   */
 /*
  *      Options received (usually on last packet, some only on SYN packets).
  */
diff --git a/include/net/tcp.h b/include/net/tcp.h
index c73fc145ee45..3a35b1500359 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -70,9 +70,6 @@ void tcp_time_wait(struct sock *sk, int state, int timeo);
 /* After receiving this amount of duplicate ACKs fast retransmit starts. */
 #define TCP_FASTRETRANS_THRESH 3
 
-/* Maximal reordering. */
-#define TCP_MAX_REORDERING      127
-
 /* Maximal number of ACKs sent quickly to accelerate slow-start. */
 #define TCP_MAX_QUICKACKS       16U
 
@@ -252,6 +249,7 @@ extern int sysctl_tcp_abort_on_overflow;
 extern int sysctl_tcp_max_orphans;
 extern int sysctl_tcp_fack;
 extern int sysctl_tcp_reordering;
+extern int sysctl_tcp_max_reordering;
 extern int sysctl_tcp_dsack;
 extern long sysctl_tcp_mem[3];
 extern int sysctl_tcp_wmem[3];
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index b3c53c8b331e..e0ee384a448f 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -496,6 +496,13 @@ static struct ctl_table ipv4_table[] = {
                 .proc_handler   = proc_dointvec
         },
         {
+                .procname       = "tcp_max_reordering",
+                .data           = &sysctl_tcp_max_reordering,
+                .maxlen         = sizeof(int),
+                .mode           = 0644,
+                .proc_handler   = proc_dointvec
+        },
+        {
                 .procname       = "tcp_dsack",
                 .data           = &sysctl_tcp_dsack,
                 .maxlen         = sizeof(int),
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index a12b455928e5..9a18cdd633f3 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -81,6 +81,7 @@ int sysctl_tcp_window_scaling __read_mostly = 1;
 int sysctl_tcp_sack __read_mostly = 1;
 int sysctl_tcp_fack __read_mostly = 1;
 int sysctl_tcp_reordering __read_mostly = TCP_FASTRETRANS_THRESH;
+int sysctl_tcp_max_reordering __read_mostly = 300;
 EXPORT_SYMBOL(sysctl_tcp_reordering);
 int sysctl_tcp_dsack __read_mostly = 1;
 int sysctl_tcp_app_win __read_mostly = 31;
@@ -833,7 +834,7 @@ static void tcp_update_reordering(struct sock *sk, const int metric,
         if (metric > tp->reordering) {
                 int mib_idx;
 
-                tp->reordering = min(TCP_MAX_REORDERING, metric);
+                tp->reordering = min(sysctl_tcp_max_reordering, metric);
 
                 /* This exciting event is worth to be remembered. 8) */
                 if (ts)