Revise the direct or optimized case to use uiomove_fromphys() by the reader

instead of ephemeral mappings using pmap_qenter() by the writer.  The
writer is still, however, responsible for wiring the pages, just not
mapping them.  Consequently, the allocation of KVA for the direct case is
unnecessary.  Remove it and the sysctls limiting it, i.e.,
kern.ipc.maxpipekvawired and kern.ipc.amountpipekvawired.  The number
of temporarily wired pages is still, however, limited by
kern.ipc.maxpipekva.

Note: On platforms lacking a direct virtual-to-physical mapping,
uiomove_fromphys() uses sf_bufs to cache ephemeral mappings.  Thus,
the number of available sf_bufs can influence the performance of pipes
on platforms such i386.  Surprisingly, I saw the greatest gain from this
change on such a machine: lmbench's pipe bandwidth result increased from
~1050MB/s to ~1850MB/s on my 2.4GHz, 400MHz FSB P4 Xeon.

1 files changed, 1 insertions, 6 deletions

diff --git a/sys/kern/subr_param.c b/sys/kern/subr_param.c
index 9eca45f..e7421b7 100644
--- a/sys/kern/subr_param.c
+++ b/sys/kern/subr_param.c
@@ -82,7 +82,6 @@ int	nswbuf;
 int	maxswzone;			/* max swmeta KVA storage */
 int	maxbcache;			/* max buffer cache KVA storage */
 int	maxpipekva;			/* Limit on pipe KVA */
-int	maxpipekvawired;		/* Limit on wired pipe KVA */
 u_quad_t	maxtsiz;			/* max text size */
 u_quad_t	dfldsiz;			/* initial data size limit */
 u_quad_t	maxdsiz;			/* max data size */
@@ -184,17 +183,13 @@ init_param3(long kmempages)
 {
 	/*
 	 * Limit pageable pipe memory usage to 5% of the kernel map
-	 * (via pipe_map), and nonpageable pipe memory usage to 2.5%
-	 * of the same.  Ensure that all have reasonable floors.
+	 * (via pipe_map).  Ensure that all have reasonable floors.
 	 * (See sys_pipe.c for more info.)
 	 */
 	maxpipekva = (kmempages / 20) * PAGE_SIZE;
-	maxpipekvawired = (kmempages / 40) * PAGE_SIZE;
 
 	if (maxpipekva < 512 * 1024)
 		maxpipekva = 512 * 1024;
-	if (maxpipekvawired < 512 * 1024)
-		maxpipekvawired = 512 * 1024;
 
 	TUNABLE_INT_FETCH("kern.ipc.maxpipekva", &maxpipekva);
 }