Previous commit to vfs_cluster accidentally disabled read-ahead. Problem

fixed by initializing "alreadyincore" to 0 in the case of
sequential reads.
This commit is contained in:
John Dyson 1996-01-20 23:24:16 +00:00
parent 9f29a57754
commit f69170e601
Notes: svn2git 2020-12-20 02:59:44 +00:00
svn path=/head/; revision=13524

View File

@ -33,7 +33,7 @@
* SUCH DAMAGE.
*
* @(#)vfs_cluster.c 8.7 (Berkeley) 2/13/94
* $Id: vfs_cluster.c,v 1.31 1995/12/22 16:06:46 bde Exp $
* $Id: vfs_cluster.c,v 1.32 1996/01/19 03:58:14 dyson Exp $
*/
#include <sys/param.h>
@ -179,6 +179,7 @@ cluster_read(vp, filesize, lblkno, size, cred, bpp)
* if we have been doing sequential I/O, then do some read-ahead
*/
if (seq) {
alreadyincore = 0;
/*
* bump ralen a bit...