tcp: allow splice() to build full TSO packets

[ This combines upstream commit
  2f53384424251c06038ae612e56231b96ab610ee and the follow-on bug fix
  commit 35f9c09fe9c72eb8ca2b8e89a593e1c151f28fc2 ]

vmsplice()/splice(pipe, socket) call do_tcp_sendpages() one page at a
time, adding at most 4096 bytes to an skb. (assuming PAGE_SIZE=4096)

The call to tcp_push() at the end of do_tcp_sendpages() forces an
immediate xmit when pipe is not already filled, and tso_fragment() try
to split these skb to MSS multiples.

4096 bytes are usually split in a skb with 2 MSS, and a remaining
sub-mss skb (assuming MTU=1500)

This makes slow start suboptimal because many small frames are sent to
qdisc/driver layers instead of big ones (constrained by cwnd and packets
in flight of course)

In fact, applications using sendmsg() (adding an additional memory copy)
instead of vmsplice()/splice()/sendfile() are a bit faster because of
this anomaly, especially if serving small files in environments with
large initial [c]wnd.

Call tcp_push() only if MSG_MORE is not set in the flags parameter.

This bit is automatically provided by splice() internals but for the
last page, or on all pages if user specified SPLICE_F_MORE splice()
flag.

In some workloads, this can reduce number of sent logical packets by an
order of magnitude, making zero-copy TCP actually faster than
one-copy :)

Reported-by: Tom Herbert <therbert@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: H.K. Jerry Chu <hkchu@google.com>
Cc: Maciej Żenczykowski <maze@google.com>
Cc: Mahesh Bandewar <maheshb@google.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>

diff --git a/fs/splice.c b/fs/splice.c
index fa2defa..6d0dfb8 100644 (file)
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -31,6 +31,7 @@
 #include <linux/uio.h>
 #include <linux/security.h>
 #include <linux/gfp.h>
+#include <linux/socket.h>
 
 /*
  * Attempt to steal a page from a pipe buffer. This should perhaps go into
@@ -691,7 +692,9 @@ static int pipe_to_sendpage(struct pipe_inode_info *pipe,
        if (!likely(file->f_op && file->f_op->sendpage))
                return -EINVAL;
 
-       more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len;
+       more = (sd->flags & SPLICE_F_MORE) ? MSG_MORE : 0;
+       if (sd->len < sd->total_len)
+               more |= MSG_SENDPAGE_NOTLAST;
        return file->f_op->sendpage(file, buf->page, buf->offset,
                                    sd->len, &pos, more);
 }

diff --git a/include/linux/socket.h b/include/linux/socket.h
index d0e77f6..ad919e0 100644 (file)
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -265,7 +265,7 @@ struct ucred {
 #define MSG_NOSIGNAL   0x4000  /* Do not generate SIGPIPE */
 #define MSG_MORE       0x8000  /* Sender will send more */
 #define MSG_WAITFORONE 0x10000 /* recvmmsg(): block until 1+ packets avail */
-
+#define MSG_SENDPAGE_NOTLAST 0x20000 /* sendpage() internal : not the last page */
 #define MSG_EOF         MSG_FIN
 
 #define MSG_CMSG_CLOEXEC 0x40000000    /* Set close_on_exit for file

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 34f5db1..36611ab 100644 (file)
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -860,7 +860,7 @@ wait_for_memory:
        }
 
 out:
-       if (copied)
+       if (copied && !(flags & MSG_SENDPAGE_NOTLAST))
                tcp_push(sk, flags, mss_now, tp->nonagle);
        return copied;
 

diff --git a/net/socket.c b/net/socket.c
index 2dce67a..273cbce 100644 (file)
--- a/net/socket.c
+++ b/net/socket.c
@@ -791,9 +791,9 @@ static ssize_t sock_sendpage(struct file *file, struct page *page,
 
        sock = file->private_data;
 
-       flags = !(file->f_flags & O_NONBLOCK) ? 0 : MSG_DONTWAIT;
-       if (more)
-               flags |= MSG_MORE;
+       flags = (file->f_flags & O_NONBLOCK) ? MSG_DONTWAIT : 0;
+       /* more is a combination of MSG_MORE and MSG_SENDPAGE_NOTLAST */
+       flags |= more;
 
        return kernel_sendpage(sock, page, offset, size, flags);
 }