| msdosfs_vnops.c (295370) | msdosfs_vnops.c (298806) |
|---|---|
| 1/* $FreeBSD: head/sys/fs/msdosfs/msdosfs_vnops.c 295370 2016-02-07 15:36:16Z pfg $ */ | 1/* $FreeBSD: head/sys/fs/msdosfs/msdosfs_vnops.c 298806 2016-04-29 20:51:24Z pfg $ */ |
| 2/* $NetBSD: msdosfs_vnops.c,v 1.68 1998/02/10 14:10:04 mrg Exp $ */ 3 4/*- 5 * Copyright (C) 1994, 1995, 1997 Wolfgang Solfrank. 6 * Copyright (C) 1994, 1995, 1997 TooLs GmbH. 7 * All rights reserved. 8 * Original code by Paul Popelka (paulp@uts.amdahl.com) (see below). 9 * --- 101 unchanged lines hidden (view full) --- 111 * (including directories when they are read/written as files). This 112 * presents problems for the dos filesystem because data that should be in 113 * an inode (if dos had them) resides in the directory itself. Since we 114 * must update directory entries without the benefit of having the vnode 115 * for the directory we must use the vnode for the filesystem. This means 116 * that when a directory is actually read/written (via read, write, or 117 * readdir, or seek) we must use the vnode for the filesystem instead of 118 * the vnode for the directory as would happen in ufs. This is to insure we | 2/* $NetBSD: msdosfs_vnops.c,v 1.68 1998/02/10 14:10:04 mrg Exp $ */ 3 4/*- 5 * Copyright (C) 1994, 1995, 1997 Wolfgang Solfrank. 6 * Copyright (C) 1994, 1995, 1997 TooLs GmbH. 7 * All rights reserved. 8 * Original code by Paul Popelka (paulp@uts.amdahl.com) (see below). 9 * --- 101 unchanged lines hidden (view full) --- 111 * (including directories when they are read/written as files). This 112 * presents problems for the dos filesystem because data that should be in 113 * an inode (if dos had them) resides in the directory itself. Since we 114 * must update directory entries without the benefit of having the vnode 115 * for the directory we must use the vnode for the filesystem. This means 116 * that when a directory is actually read/written (via read, write, or 117 * readdir, or seek) we must use the vnode for the filesystem instead of 118 * the vnode for the directory as would happen in ufs. This is to insure we |
| 119 * retreive the correct block from the buffer cache since the hash value is | 119 * retrieve the correct block from the buffer cache since the hash value is |
| 120 * based upon the vnode address and the desired block number. 121 */ 122 123/* 124 * Create a regular file. On entry the directory to contain the file being 125 * created is locked. We must release before we return. We must also free 126 * the pathname buffer pointed at by cnp->cn_pnbuf, always on error, or 127 * only if the SAVESTART bit in cn_flags is clear on success. --- 1792 unchanged lines hidden --- | 120 * based upon the vnode address and the desired block number. 121 */ 122 123/* 124 * Create a regular file. On entry the directory to contain the file being 125 * created is locked. We must release before we return. We must also free 126 * the pathname buffer pointed at by cnp->cn_pnbuf, always on error, or 127 * only if the SAVESTART bit in cn_flags is clear on success. --- 1792 unchanged lines hidden --- |