vfs: Fix runningspace tuning after maxphys was bumped

The previous maximum value for the upper watermark was based on the old
value of MAXPHYS.  Raise it to allow more parallel I/O on large systems.

This is still a rather flawed mechanism since it's applied without
regard to the number of filesystems or block devices between which this
mechanism sits, but we might as well bump the limits at this point, as
they haven't been revised in quite a long time.

Reviewed by:	imp, kib
MFC after:	2 weeks
Fixes:		cd85379104 ("Make MAXPHYS tunable. Bump MAXPHYS to 1M.")
Differential Revision:	https://reviews.freebsd.org/D47398

sys/kern/vfs_bio.c

@@ -1253,15 +1253,16 @@ bufinit(void)
 	bufspacethresh = lobufspace + (hibufspace - lobufspace) / 2;
 
 	/*
-	 * Note: The 16 MiB upper limit for hirunningspace was chosen
-	 * arbitrarily and may need further tuning. It corresponds to
-	 * 128 outstanding write IO requests (if IO size is 128 KiB),
-	 * which fits with many RAID controllers' tagged queuing limits.
+	 * Note: The upper limit for hirunningspace was chosen arbitrarily and
+	 * may need further tuning. It corresponds to 128 outstanding write IO
+	 * requests, which fits with many RAID controllers' tagged queuing
+	 * limits.
+	 *
 	 * The lower 1 MiB limit is the historical upper limit for
 	 * hirunningspace.
 	 */
 	hirunningspace = lmax(lmin(roundup(hibufspace / 64, maxbcachebuf),
-	    16 * 1024 * 1024), 1024 * 1024);
+	    128 * maxphys), 1024 * 1024);
 	lorunningspace = roundup((hirunningspace * 2) / 3, maxbcachebuf);
 
 	/*