From eee827ad6ffb1eb6ea4684478cfcfbb6f1884cbe Mon Sep 17 00:00:00 2001
From: Matthew Dillon <dillon@dragonflybsd.org>
Date: Tue, 10 Jun 2008 05:02:09 +0000
Subject: [PATCH] Change bwillwrite() to smooth out performance under heavy
 loads.  Blocking based on strict hystersis was being used to try to gang
 flushes together but filesystems can still blow out the buffer cache and
 cause processes to block for long periods of time waiting for the dirty count
 to drop significantly.

Instead, as the number of dirty buffers exceeds the desired maximum
bwillwrite() imposes a dynamic delay which increases as the number of
dirty buffers increase.  This improves the stall behavior under heavy loads
and keeps the system responsive.

TODO: The algorithm needs to have a per-LWP heuristic to penalize heavy
writers more then light ones.
---
 sys/kern/vfs_bio.c | 53 ++++++++++++++++++++++++++++++++++++++++++-----------
 1 file changed, 42 insertions(+), 11 deletions(-)

diff --git a/sys/kern/vfs_bio.c b/sys/kern/vfs_bio.c
index 1c3f0c7658..38c45ff4ab 100644
--- a/sys/kern/vfs_bio.c
+++ b/sys/kern/vfs_bio.c
@@ -12,7 +12,7 @@
  *		John S. Dyson.
  *
  * $FreeBSD: src/sys/kern/vfs_bio.c,v 1.242.2.20 2003/05/28 18:38:10 alc Exp $
- * $DragonFly: src/sys/kern/vfs_bio.c,v 1.102 2008/05/09 07:24:45 dillon Exp $
+ * $DragonFly: src/sys/kern/vfs_bio.c,v 1.103 2008/06/10 05:02:09 dillon Exp $
  */
 
 /*
@@ -968,19 +968,50 @@ bowrite(struct buf *bp)
 void
 bwillwrite(void)
 {
-	if (runningbufcount + numdirtybuffers >= hidirtybuffers / 2) {
+	int mid1 = hidirtybuffers / 2;
+	int mid2 = mid1 + hidirtybuffers / 4;
+	int delay;
+	int count;
+	int priority = 0;
+
+	count = runningbufcount + numdirtybuffers;
+
+	/*
+	 * Nothing to do if nothing is stressed.
+	 */
+	if (count < mid1)
+		return;
+
+	/*
+	 * Get the buffer daemon heated up
+	 */
+	bd_wakeup(1);
+
+	while (count >= mid2) {
+		/*
+		 * Start slowing down writes, down to 1 per second.
+		 */
+		if (count < hidirtybuffers) {
+			delay = (count - mid2) * hz / (hidirtybuffers - mid2);
+			delay = delay * 10 / (10 + priority);
+			if (delay == 0)
+				delay = 1;
+			tsleep(&count, 0, "flstik", delay);
+			return;
+		}
+
+		/*
+		 * Now we are really in trouble.
+		 */
 		bd_wakeup(1);
-		while (runningbufcount + numdirtybuffers >= hidirtybuffers) {
-			bd_wakeup(1);
-			spin_lock_wr(&needsbuffer_spin);
-			if (runningbufcount + numdirtybuffers >= 
-			    hidirtybuffers) {
-				needsbuffer |= VFS_BIO_NEED_DIRTYFLUSH;
-				msleep(&needsbuffer, &needsbuffer_spin, 0,
-				       "flswai", 0);
-			}
+		spin_lock_wr(&needsbuffer_spin);
+		count = runningbufcount + numdirtybuffers;
+		if (count >= hidirtybuffers) {
+			needsbuffer |= VFS_BIO_NEED_DIRTYFLUSH;
+			msleep(&needsbuffer, &needsbuffer_spin, 0, "flswai", 0);
 			spin_unlock_wr(&needsbuffer_spin);
 		}
+		count = runningbufcount + numdirtybuffers;
 	} 
 #if 0
 	/* FUTURE - maybe */
-- 
2.11.4.GIT