Update sendfile.2 to include a TUNING section that documents the

various tunables that are applicable to sendfile(2).  Update tuning.7
to mention a reference to sendfile.2.

Approved by:	keramida

1 files changed, 38 insertions, 0 deletions

diff --git a/lib/libc/sys/sendfile.2 b/lib/libc/sys/sendfile.2
index a0d8239..fac2577 100644
--- a/lib/libc/sys/sendfile.2
+++ b/lib/libc/sys/sendfile.2
@@ -106,6 +106,43 @@ The
 implementation of
 .Fn sendfile
 is "zero-copy", meaning that it has been optimized so that copying of the file data is avoided.
+.Sh TUNING
+Internally this system call uses a special
+.Xr sendfile 2
+buffer
+.Pq Fa "struct sf_buf"
+to handle sending file data to the client.
+If the sending socket is
+blocking and there are not enough sendfile buffers available,
+.Xr sendfile 2
+will block and report a state of
+.Dq sfbufa .
+If the sending socket is non-blocking and there are not enough
+sendfile buffers available, the call will block and wait for the
+necessary buffers to become available before finishing the call.
+.Pp
+The number of
+.Fa sf_buf Ns 's
+allocated should be proportional to the number of nmbclusters used to
+send data to a client via
+.Xr sendfile 2 .
+Tune accordingly to avoid blocking!
+Busy installations that make extensive use of
+.Xr sendfile 2
+may want to increase these values to be inline with their
+.Ar kern.ipc.nmbclusters
+.Pq see Xr tuning 7 for details .
+.Pp
+The number of
+.Xr sendfile 2
+buffers in use is determined at boot time by either the
+.Em kern.ipc.nsfbufs
+.Xr loader.conf 5
+variable or the
+.Em NSFBUFS
+kernel configuration tunable.
+The number of sendfile buffers scales with
+.Em kern.maxusers .
 .Sh RETURN VALUES
 .Rv -std sendfile
 .Sh ERRORS
@@ -161,6 +198,7 @@ If specified, the number of bytes successfully sent will be returned in
 .Xr open 2 ,
 .Xr send 2 ,
 .Xr socket 2 ,
+.Xr tuning 7 ,
 .Xr writev 2
 .Sh HISTORY
 The