Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Make sysctl variable lfs_clean_vnhead do what it was supposed to do, 

namely, toggle whether vnodes loaded only for cleaning (as opposed to
normal filesystem use) are freed to the *head* of the vnode free list,
rather than the tail.  This should avoid a possible cache flushing
effect, if the cleaner cleans a segment containing a large number of
live inodes.
This commit is contained in:
perseant 1999-03-25 22:38:28 +00:00
parent cffa290130
commit fbc5e63cd6
1 changed files with 9 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
sys/ufs/lfs/lfs_segment.c
View File

@ -1,4 +1,4 @@
/* $NetBSD: lfs_segment.c,v 1.19 1999/03/25 22:26:52 perseant Exp $ */
/* $NetBSD: lfs_segment.c,v 1.20 1999/03/25 22:38:28 perseant Exp $ */
/*-
* Copyright (c) 1999 The NetBSD Foundation, Inc.
@ -135,6 +135,7 @@ int lfs_writevnodes __P((struct lfs *fs, struct mount *mp,
int lfs_allclean_wakeup; /* Cleaner wakeup address. */
int lfs_writeindir = 1; /* whether to flush indir on non-ckp */
int lfs_clean_vnhead = 1; /* Allow freeing to head of vn list */
/* Statistics Counters */
int lfs_dostats = 1;
@ -280,7 +281,7 @@ lfs_writevnodes(fs, mp, sp, op)
{
struct inode *ip;
struct vnode *vp;
int inodes_written=0;
int inodes_written=0, only_cleaning;
#ifndef LFS_NO_BACKVP_HACK
/* BEGIN HACK */
@ -350,6 +351,8 @@ lfs_writevnodes(fs, mp, sp, op)
(IN_ACCESS | IN_CHANGE | IN_MODIFIED | IN_UPDATE | IN_CLEANING) ||
vp->v_dirtyblkhd.lh_first != NULL))
{
only_cleaning = ((ip->i_flag & (IN_ACCESS|IN_CHANGE|IN_MODIFIED|IN_UPDATE|IN_CLEANING))==IN_CLEANING);
if(ip->i_number != LFS_IFILE_INUM
&& vp->v_dirtyblkhd.lh_first != NULL)
{
@ -379,20 +382,14 @@ lfs_writevnodes(fs, mp, sp, op)
lfs_vunref(vp);
}
lfs_vunref(vp);
if(lfs_clean_vnhead && only_cleaning)
lfs_vunref_head(vp);
else
lfs_vunref(vp);
}
return inodes_written;
}
/*
* There is a distinct difference in the interpretation of SEGM_CLEAN,
* depending on whether it is passed *directly* to lfs_segwrite (i.e., we
* were called from lfs_markv), or whether it was just in the segment flags
* (we were called indirectly through getnewvnode/lfs_vflush). In the former
* case, we only want to write vnodes where cleaning is in progress; but
* in the latter case, we might want to write all empty vnodes, or possibly
* all vnodes.
*/
int
lfs_segwrite(mp, flags)
struct mount *mp;