NetBSD/sys/msdosfs/denode.h

284 lines
12 KiB
C
Raw Normal View History

/* $NetBSD: denode.h,v 1.16 1995/09/09 19:38:00 ws Exp $ */
/*-
* Copyright (C) 1994 Wolfgang Solfrank.
* Copyright (C) 1994 TooLs GmbH.
* All rights reserved.
* Original code by Paul Popelka (paulp@uts.amdahl.com) (see below).
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by TooLs GmbH.
* 4. The name of TooLs GmbH may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY TOOLS GMBH ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL TOOLS GMBH BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
* OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
* WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
* ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
/*
* Written by Paul Popelka (paulp@uts.amdahl.com)
*
* You can do anything you want with this software, just don't say you wrote
* it, and don't remove this notice.
*
* This software is provided "as is".
*
* The author supplies this software to be publicly redistributed on the
* understanding that the author is not responsible for the correct
* functioning of this software in any circumstances and is not liable for
* any damages caused by this software.
*
* October 1992
*/
/*
1994-07-17 01:32:06 +04:00
* This is the pc filesystem specific portion of the vnode structure.
*
* To describe a file uniquely the de_dirclust, de_diroffset, and
* de_StartCluster fields are used.
*
* de_dirclust contains the cluster number of the directory cluster
* containing the entry for a file or directory.
* de_diroffset is the index into the cluster for the entry describing
* a file or directory.
* de_StartCluster is the number of the first cluster of the file or directory.
*
* Now to describe the quirks of the pc filesystem.
* - Clusters 0 and 1 are reserved.
* - The first allocatable cluster is 2.
* - The root directory is of fixed size and all blocks that make it up
* are contiguous.
* - Cluster 0 refers to the root directory when it is found in the
* startcluster field of a directory entry that points to another directory.
* - Cluster 0 implies a 0 length file when found in the start cluster field
* of a directory entry that points to a file.
* - You can't use the cluster number 0 to derive the address of the root
* directory.
* - Multiple directory entries can point to a directory. The entry in the
* parent directory points to a child directory. Any directories in the
* child directory contain a ".." entry that points back to the parent.
* The child directory itself contains a "." entry that points to itself.
* - The root directory does not contain a "." or ".." entry.
* - Directory entries for directories are never changed once they are created
* (except when removed). The size stays 0, and the last modification time
* is never changed. This is because so many directory entries can point to
* the physical clusters that make up a directory. It would lead to an
* update nightmare.
* - The length field in a directory entry pointing to a directory contains 0
* (always). The only way to find the end of a directory is to follow the
* cluster chain until the "last cluster" marker is found.
*
* My extensions to make this house of cards work. These apply only to the in
* memory copy of the directory entry.
* - A reference count for each denode will be kept since dos doesn't keep such
* things.
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
*/
/*
* Internal pseudo-offset for (nonexistent) directory entry for the root
* dir in the root dir
*/
#define MSDOSFSROOT_OFS 0x1fffffff
/*
* The fat cache structure. fc_fsrcn is the filesystem relative cluster
* number that corresponds to the file relative cluster number in this
* structure (fc_frcn).
*/
struct fatcache {
u_short fc_frcn; /* file relative cluster number */
u_short fc_fsrcn; /* filesystem relative cluster number */
};
/*
* The fat entry cache as it stands helps make extending files a "quick"
* operation by avoiding having to scan the fat to discover the last
* cluster of the file. The cache also helps sequential reads by
* remembering the last cluster read from the file. This also prevents us
* from having to rescan the fat to find the next cluster to read. This
* cache is probably pretty worthless if a file is opened by multiple
* processes.
*/
#define FC_SIZE 2 /* number of entries in the cache */
#define FC_LASTMAP 0 /* entry the last call to pcbmap() resolved
* to */
#define FC_LASTFC 1 /* entry for the last cluster in the file */
#define FCE_EMPTY 0xffff /* doesn't represent an actual cluster # */
/*
* Set a slot in the fat cache.
*/
#define fc_setcache(dep, slot, frcn, fsrcn) \
(dep)->de_fc[slot].fc_frcn = frcn; \
(dep)->de_fc[slot].fc_fsrcn = fsrcn;
/*
* This is the in memory variant of a dos directory entry. It is usually
* contained within a vnode.
*/
struct denode {
1994-07-17 01:32:06 +04:00
struct denode *de_next; /* Hash chain forward */
struct denode **de_prev; /* Hash chain back */
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
struct vnode *de_vnode; /* addr of vnode we are part of */
struct vnode *de_devvp; /* vnode of blk dev we live on */
u_long de_flag; /* flag bits */
dev_t de_dev; /* device where direntry lives */
u_long de_dirclust; /* cluster of the directory file containing this entry */
u_long de_diroffset; /* ordinal of this entry in the directory */
1994-07-17 01:32:06 +04:00
u_long de_fndclust; /* cluster of found dir entry */
u_long de_fndoffset; /* offset of found dir entry */
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
long de_refcnt; /* reference count */
struct msdosfsmount *de_pmp; /* addr of our mount struct */
struct lockf *de_lockf; /* byte level lock list */
1994-07-17 01:32:06 +04:00
pid_t de_lockholder; /* current lock holder */
pid_t de_lockwaiter; /* lock wanter */
/* the next two fields must be contiguous in memory... */
u_char de_Name[8]; /* name, from directory entry */
u_char de_Extension[3]; /* extension, from directory entry */
u_char de_Attributes; /* attributes, from directory entry */
u_short de_Time; /* creation time */
u_short de_Date; /* creation date */
u_short de_StartCluster; /* starting cluster of file */
u_long de_FileSize; /* size of file in bytes */
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
struct fatcache de_fc[FC_SIZE]; /* fat cache */
};
/*
* Values for the de_flag field of the denode.
*/
1995-01-04 09:32:19 +03:00
#define DE_LOCKED 0x0001 /* Denode lock. */
#define DE_WANTED 0x0002 /* Denode is wanted by a process. */
#define DE_UPDATE 0x0004 /* Modification time update request. */
#define DE_MODIFIED 0x0008 /* Denode has been modified. */
#define DE_RENAME 0x0010 /* Denode is in the process of being renamed */
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
/*
* Transfer directory entries between internal and external form.
* dep is a struct denode * (internal form),
* dp is a struct direntry * (external form).
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
*/
#define DE_INTERNALIZE(dep, dp) \
(bcopy((dp)->deName, (dep)->de_Name, 11), \
(dep)->de_Attributes = (dp)->deAttributes, \
(dep)->de_Time = getushort((dp)->deTime), \
(dep)->de_Date = getushort((dp)->deDate), \
(dep)->de_StartCluster = getushort((dp)->deStartCluster), \
(dep)->de_FileSize = getulong((dp)->deFileSize))
#define DE_EXTERNALIZE(dp, dep) \
(bcopy((dep)->de_Name, (dp)->deName, 11), \
(dp)->deAttributes = (dep)->de_Attributes, \
putushort((dp)->deTime, (dep)->de_Time), \
putushort((dp)->deDate, (dep)->de_Date), \
putushort((dp)->deStartCluster, (dep)->de_StartCluster), \
putulong((dp)->deFileSize, \
((dep)->de_Attributes & ATTR_DIRECTORY) ? 0 : (dep)->de_FileSize))
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
#define de_forw de_chain[0]
#define de_back de_chain[1]
1995-03-30 01:57:43 +04:00
#ifdef _KERNEL
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
#define VTODE(vp) ((struct denode *)(vp)->v_data)
#define DETOV(de) ((de)->de_vnode)
1995-01-04 09:03:11 +03:00
#define DE_TIMES(dep) \
if ((dep)->de_flag & DE_UPDATE) { \
(dep)->de_flag &= ~DE_UPDATE; \
if (((dep)->de_Attributes & ATTR_DIRECTORY) == 0) { \
unix2dostime(NULL, &(dep)->de_Date, &(dep)->de_Time); \
(dep)->de_Attributes |= ATTR_ARCHIVE; \
(dep)->de_flag |= DE_MODIFIED; \
} \
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
}
/*
1995-01-04 09:32:19 +03:00
* This overlays the fid structure (see mount.h)
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
*/
struct defid {
u_short defid_len; /* length of structure */
u_short defid_pad; /* force long alignment */
u_long defid_dirclust; /* cluster this dir entry came from */
u_long defid_dirofs; /* offset of entry within the cluster */
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
/* u_long defid_gen; /* generation number */
};
/*
* Prototypes for MSDOSFS vnode operations
*/
int msdosfs_lookup __P((struct vop_lookup_args *));
int msdosfs_create __P((struct vop_create_args *));
int msdosfs_mknod __P((struct vop_mknod_args *));
int msdosfs_open __P((struct vop_open_args *));
int msdosfs_close __P((struct vop_close_args *));
int msdosfs_access __P((struct vop_access_args *));
int msdosfs_getattr __P((struct vop_getattr_args *));
int msdosfs_setattr __P((struct vop_setattr_args *));
int msdosfs_read __P((struct vop_read_args *));
int msdosfs_write __P((struct vop_write_args *));
#ifdef NFSSERVER
int lease_check __P((struct vop_lease_args *));
#define msdosfs_lease_check lease_check
#else
#define msdosfs_lease_check ((int (*) __P((struct vop_lease_args *)))nullop)
#endif
int msdosfs_ioctl __P((struct vop_ioctl_args *));
int msdosfs_select __P((struct vop_select_args *));
int msdosfs_mmap __P((struct vop_mmap_args *));
int msdosfs_fsync __P((struct vop_fsync_args *));
int msdosfs_seek __P((struct vop_seek_args *));
int msdosfs_remove __P((struct vop_remove_args *));
int msdosfs_link __P((struct vop_link_args *));
int msdosfs_rename __P((struct vop_rename_args *));
int msdosfs_mkdir __P((struct vop_mkdir_args *));
int msdosfs_rmdir __P((struct vop_rmdir_args *));
int msdosfs_symlink __P((struct vop_symlink_args *));
int msdosfs_readdir __P((struct vop_readdir_args *));
int msdosfs_readlink __P((struct vop_readlink_args *));
int msdosfs_abortop __P((struct vop_abortop_args *));
int msdosfs_inactive __P((struct vop_inactive_args *));
int msdosfs_reclaim __P((struct vop_reclaim_args *));
int msdosfs_lock __P((struct vop_lock_args *));
int msdosfs_unlock __P((struct vop_unlock_args *));
int msdosfs_bmap __P((struct vop_bmap_args *));
int msdosfs_strategy __P((struct vop_strategy_args *));
int msdosfs_print __P((struct vop_print_args *));
int msdosfs_islocked __P((struct vop_islocked_args *));
int msdosfs_advlock __P((struct vop_advlock_args *));
int msdosfs_reallocblks __P((struct vop_reallocblks_args *));
brought in fixed/renamed/matching MS-DOS FS code, from Jeff Polk <polk@bsdi.com>. His notes are as follows: ------------------------------------------------------------------------------ July 22, 1993 - Changed name of entire package from PCFS to MSDOSFS - Fixed bugs: root directory size in clusters instead of bytes growing directory didn't update in-core size link, symlink, mknod didn't free locked parent (deadlock) lookup returned real error on create and rename instead of EJUSTRETURN rename changed `.' entry in child instead of name entry in parent rename removed `.' entry in child instead of removing entry in parent when moving a directory from one dir to another createde() left new node locked when write of parent failed (deadlock) removede() decremented refcount even on error (rmdir's which failed due to write errors left in-core cache entries inconsistent) changed validation for filesystem to not check for the boot signature since some disks (e.g., mtools) aren't bootable directories are always show current time as modify time (needed for NFS export since DOS never updates dir mod times -- ctime is true create time). - Added support for cookies changes to the readdir() vnode interface (#ifdef __bsdi__) - Punted on the whole problem of inode generation numbers. This means that there's a chance of using a stale file handle to access a new file, but it doesn't appear to be the common case, and I don't see how to generate reasonable generation numbers without changing something on the disk (which is the way the SVR4 filesystem survival kit guys did it). I don't think it would be very safe to change the on-disk format. Jeff Polk (polk@BSDI.COM) ------------------------------------------------------------------------------
1993-08-13 15:35:13 +04:00
/*
* Internal service routine prototypes.
*/
int createde __P((struct denode *, struct denode *, struct denode **));
int deextend __P((struct denode *, u_long, struct ucred *));
int deget __P((struct msdosfsmount *, u_long, u_long, struct direntry *, struct denode **));
int detrunc __P((struct denode *, u_long, int, struct ucred *, struct proc *));
int deupdat __P((struct denode *, int));
int doscheckpath __P((struct denode *, struct denode *));
int dosdirempty __P((struct denode *));
int readde __P((struct denode *, struct buf **, struct direntry **));
int readep __P((struct msdosfsmount *, u_long, u_long, struct buf **, struct direntry **));
void reinsert __P((struct denode *));
int removede __P((struct denode *, struct denode *));
1995-03-30 01:57:43 +04:00
#endif /* _KERNEL */