From owner-cvs-sys  Sat Dec 28 18:45:35 1996
Return-Path: <owner-cvs-sys>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id SAA09856
          for cvs-sys-outgoing; Sat, 28 Dec 1996 18:45:35 -0800 (PST)
Received: (from dyson@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id SAA09849;
          Sat, 28 Dec 1996 18:45:32 -0800 (PST)
Date: Sat, 28 Dec 1996 18:45:32 -0800 (PST)
From: John Dyson <dyson>
Message-Id: <199612290245.SAA09849@freefall.freebsd.org>
To: CVS-committers, cvs-all, cvs-sys
Subject: cvs commit:  src/sys/gnu/ext2fs ext2_readwrite.c src/sys/isofs/cd9660 cd9660_vnops.c src/sys/kern kern_descrip.c vfs_cluster.c vfs_subr.c vfs_vnops.c src/sys/sys buf.h file.h filedesc.h vnode.h src/sys/ufs/ufs ufs_readwrite.c
Sender: owner-cvs-sys@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

dyson       96/12/28 18:45:31

  Modified:    sys/gnu/ext2fs  ext2_readwrite.c
  Log:
  This commit is the embodiment of some VFS read clustering improvements.
  Firstly, now our read-ahead clustering is on a file descriptor basis and not
  on a per-vnode basis.  This will allow multiple processes reading the
  same file to take advantage of read-ahead clustering.  Secondly, there
  previously was a problem with large reads still using the ramp-up
  algorithm.  Of course, that was bogus, and now we read the entire
  "chunk" off of the disk in one operation.   The read-ahead clustering
  algorithm should use less CPU than the previous also (I hope :-)).
  
  Revision  Changes    Path
  1.6       +2 -1      src/sys/gnu/ext2fs/ext2_readwrite.c

  Modified:    sys/isofs/cd9660  cd9660_vnops.c
               sys/kern  kern_descrip.c vfs_cluster.c vfs_subr.c
                        vfs_vnops.c
               sys/sys   buf.h file.h filedesc.h vnode.h
               sys/ufs/ufs  ufs_readwrite.c
  Log:
  This commit is the embodiment of some VFS read clustering improvements.
  Firstly, now our read-ahead clustering is on a file descriptor basis and not
  on a per-vnode basis.  This will allow multiple processes reading the
  same file to take advantage of read-ahead clustering.  Secondly, there
  previously was a problem with large reads still using the ramp-up
  algorithm.  Of course, that was bogus, and now we read the entire
  "chunk" off of the disk in one operation.   The read-ahead clustering
  algorithm should use less CPU than the previous also (I hope :-)).
  
  NOTE:  THAT LKMS MUST BE REBUILT!!!
  
  Revision  Changes    Path
  1.28      +3 -2      src/sys/isofs/cd9660/cd9660_vnops.c
  1.35      +2 -1      src/sys/kern/kern_descrip.c
  1.40      +181 -127  src/sys/kern/vfs_cluster.c
  1.66      +1 -3      src/sys/kern/vfs_subr.c
  1.27      +35 -3     src/sys/kern/vfs_vnops.c
  1.36      +4 -3      src/sys/sys/buf.h
  1.10      +8 -1      src/sys/sys/file.h
  1.9       +3 -1      src/sys/sys/filedesc.h
  1.37      +1 -3      src/sys/sys/vnode.h
  1.24      +10 -8     src/sys/ufs/ufs/ufs_readwrite.c