xref: /illumos-gate/usr/src/uts/common/sys/fs/pc_dir.h (revision 2d6eb4a5)

/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
 * Use is subject to license terms.
 */

#ifndef	_SYS_FS_PC_DIR_H
#define	_SYS_FS_PC_DIR_H

#include <sys/dirent.h>

#ifdef	__cplusplus
extern "C" {
#endif

#define	PCFNAMESIZE	8
#define	PCFEXTSIZE	3
#define	PCMAXNAMLEN	255
#define	PCMAXNAM_UTF16	(256 * sizeof (uint16_t))	/* for UTF-16 */
#define	PCLFNCHUNKSIZE	13

struct pctime {
	ushort_t pct_time;		/* hh:mm:ss (little endian) */
	ushort_t pct_date;		/* yy:mm:dd (little endian) */
};

/*
 * Shifts and masks for time and date fields, in host byte order.
 */
#define	SECSHIFT	0
#define	SECMASK		0x1F
#define	MINSHIFT	5
#define	MINMASK		0x3F
#define	HOURSHIFT	11
#define	HOURMASK	0x1F

#define	DAYSHIFT	0
#define	DAYMASK		0x1F
#define	MONSHIFT	5
#define	MONMASK		0x0F
#define	YEARSHIFT	9
#define	YEARMASK	0x7F

struct pcdir {
	char pcd_filename[PCFNAMESIZE];	/* file name */
	char pcd_ext[PCFEXTSIZE];	/* file extension */
	uchar_t	pcd_attr;		/* file attributes */
	uchar_t	pcd_ntattr;		/* reserved for NT attributes */
	uchar_t	pcd_crtime_msec;	/* milliseconds after the minute */
	struct pctime pcd_crtime;	/* creation time/date */
	ushort_t pcd_ladate;		/* last-access date */
	union {
		uint16_t pcd_eattr;	/* OS/2 extended attribute */
		pc_cluster16_t pcd_scluster_hi;
	} un;
	struct pctime pcd_mtime;	/* last modified time/date */
	pc_cluster16_t pcd_scluster_lo;	/* starting cluster (little endian) */
	uint_t	pcd_size;		/* file size (little endian) */
};

#ifdef	__cplusplus
}
#endif

#include <sys/fs/pc_node.h>

#ifdef	__cplusplus
extern "C" {
#endif

/*
 * Long filename support (introduced by Windows 95) is an interesting
 * exercise in compatibility. Basically, it is now no longer the case
 * that an entry in a directory (as described by the 'pcdir' structure above)
 * contains the entire name of a file. Now, a directory entry can consist of
 * a long filename component (a series of 'pcdir'-like structures) followed
 * by a short filename (the old form). Each long filename component is
 * identified by having it's Read-Only, Hidden, System, and Volume Label
 * attributes set.  Each can store 13 Unicode characters (16-bits, of
 * which we only look at the lower 8 for now), broken into (gak) three
 * sections of the entry (since that's the way the available bits fall out).
 * In addition, each long filename entry has a sequence number (starting
 * from 1). The first entry has bit 7 (0x40) set in the sequence number, and
 * has the maximum value in the sequence. This may seem a bit backwards, and
 * it gets worse: the first entry stores the last component of
 * the name. So the directory entry for a file named
 * "This is a very long filename indeed" might look like this:
 *
 * Offset  Sequence      Component         Attributes    Cluster    Size
 *    0      0x43       "me indeed"          RSHV           0         0
 *   32      0x02       "y long filena"      RSHV           0         0
 *   64      0x01       "This is a ver"      RSHV           0         0
 *   96      ----       "THISIS~1.TXT"       <whatever>  2122       110
 *
 * The last entry is for the short filename, which stores actual information
 * about the file (like type, permissions, cluster, and size). The short name
 * is also used by non-long-filename aware applications (like Windows 3.X and
 * DOS). This name is generated by Windows 95 (and now Solaris) from the
 * long name, and must (of course) be unique within the directory.
 * Solaris continues to this entry to actually identify the file and its
 * attributes (filenames only really matter when names are used, like at
 * lookup/readdir/remove/create/rename time - for general access to the file
 * they aren't used).
 *
 * Long filenames can also be broken by applications that don't
 * understand them (for example, a Solaris 2.5.1 user could rename
 * "THISIS~1.TXT" to "test.exe"). This can be detected because each long
 * filename component has a checksum which is based on the short filename.
 * After reading the long filename entry, if the checksum doesn't match the
 * short name that follows, we simply ignore it and use the short name.
 *
 * One subtle thing - though long file names are case-sensitive,
 * searches for them are not.
 *
 * Another _very_ subtle thing. The number of characters in the
 * last long filename chunk (the first entry, with the 0x40 bit set) is
 * either all the characters (if there is no null, '\0'), or all the
 * characters up to the null. _However_, if the remaining characters are
 * null, Norton Disk Doctor and Microsoft ScanDisk will claim
 * that the filename entry is damaged. The remaining bytes must actually
 * contain 0xff (discovered with Disk Doctor).
 *
 * Some information about long filename support can be found in the
 * book "Inside Windows 95" by Adrian King.
 */

/*
 * The number of bytes in each section of the name in a long filename
 * entry. This is _bytes_, not characters: each character is actually
 * 16 bits.
 */
#define	PCLF_FIRSTNAMESIZE	10
#define	PCLF_SECONDNAMESIZE	12
#define	PCLF_THIRDNAMESIZE	4

/*
 * A long filename entry. It must match the 'pcdir' structure in size,
 * and pcdl_attr must overlap pcd_attr.
 */
struct pcdir_lfn {
	uchar_t pcdl_ordinal;	/* lfn order. First is 01, next 02, */
				/* last has bit 7 (0x40) set */
	uchar_t pcdl_firstfilename[PCLF_FIRSTNAMESIZE];
	uchar_t pcdl_attr;
	uchar_t pcdl_type;	/* type - always contains 0 for an LFN entry */
	uchar_t pcdl_checksum;	/* checksum to validate the LFN entry - */
				/* based on the short name */
	uchar_t pcdl_secondfilename[PCLF_SECONDNAMESIZE];
	pc_cluster16_t pcd_scluster;	/* (not used, always 0) */
	uchar_t pcdl_thirdfilename[PCLF_THIRDNAMESIZE];
};

/*
 * FAT LFN entries are consecutively numbered downwards, and the last
 * entry of a LFN chain will have the 0x40 'termination' marker logically
 * or'ed in. The entry immediately preceeding the short name has number 1,
 * consecutively increasing. Since the filename length limit on FAT is
 * 255 unicode characters and every LFN entry contributes 13 characters,
 * the maximum sequence number is 255/13 + 1 == 20.
 */
#define	PCDL_IS_LAST_LFN(x) ((x->pcdl_ordinal) & 0x40)
#define	PCDL_LFN_BITS (PCA_RDONLY | PCA_HIDDEN | PCA_SYSTEM | PCA_LABEL)
#define	PCDL_LFN_MASK (PCDL_LFN_BITS | PCA_DIR | PCA_ARCH)
#define	PCDL_LFN_VALID_ORD(x)						\
	(((((struct pcdir_lfn *)(x))->pcdl_ordinal & ~0x40) > 0) &&	\
	((((struct pcdir_lfn *)(x))->pcdl_ordinal & ~0x40) <= 20))
#define	PCDL_IS_LFN(x)							\
	(enable_long_filenames &&					\
	(((x)->pcd_attr & PCDL_LFN_MASK) == PCDL_LFN_BITS) &&		\
	PCDL_LFN_VALID_ORD((x)))

/*
 * The first char of the file name has special meaning as follows:
 */
#define	PCD_UNUSED	((char)0x00)	/* entry has never been used */
#define	PCD_ERASED	((char)0xE5)	/* entry was erased */

/*
 * File attributes.
 */
#define	PCA_RDONLY	0x01	/* file is read only */
#define	PCA_HIDDEN	0x02	/* file is hidden */
#define	PCA_SYSTEM	0x04	/* system file */
#define	PCA_LABEL	0x08	/* entry contains the volume label */
#define	PCA_DIR		0x10	/* subdirectory */
#define	PCA_ARCH	0x20	/* file has been modified since last backup */

/*
 * Avoid hidden files unless the private variable is set.
 * Always avoid the label.
 */
#define	PCA_IS_HIDDEN(fsp, attr) \
	((((attr) & PCA_LABEL) == PCA_LABEL) || \
	((((fsp)->pcfs_flags & PCFS_HIDDEN) == 0) && \
	    ((attr) & (PCA_HIDDEN | PCA_SYSTEM))))

#define	PC_NAME_IS_DOT(namep) \
	(((namep)[0] == '.') && ((namep)[1] == '\0'))
#define	PC_NAME_IS_DOTDOT(namep) \
	(((namep)[0] == '.') && ((namep)[1] == '.') && ((namep)[2] == '\0'))
#define	PC_SHORTNAME_IS_DOT(namep) \
	(((namep)[0] == '.') && ((namep)[1] == ' '))
#define	PC_SHORTNAME_IS_DOTDOT(namep) \
	(((namep)[0] == '.') && ((namep)[1] == '.') && ((namep)[2] == ' '))
/*
 * slot structure is used by the directory search routine to return
 * the results of the search.  If the search is successful sl_blkno and
 * sl_offset reflect the disk address of the entry and sl_ep points to
 * the actual entry data in buffer sl_bp. sl_flags is set to whether the
 * entry is dot or dotdot. If the search is unsuccessful sl_blkno and
 * sl_offset points to an empty directory slot if there are any. Otherwise
 * it is set to -1.
 */
struct pcslot {
	enum {SL_NONE, SL_FOUND, SL_EXTEND} sl_status;	/* slot status */
	daddr_t		sl_blkno;	/* disk block number which has entry */
	int		sl_offset;	/* offset of entry within block */
	struct buf	*sl_bp;		/* buffer containing entry data */
	struct pcdir	*sl_ep;		/* pointer to entry data */
	int		sl_flags;	/* flags (see below) */
};
#define	SL_DOT		1	/* entry point to self */
#define	SL_DOTDOT	2	/* entry points to parent */

/*
 * A pcfs directory entry. Directory entries are actually variable
 * length, but this is the maximum size.
 *
 * This _must_ match a dirent64 structure in format.
 * d_name is 512 bytes long to accomodate 256 UTF-16 characters.
 */
struct pc_dirent {
	ino64_t		d_ino;		/* "inode number" of entry */
	off64_t		d_off;		/* offset of disk directory entry */
	unsigned short	d_reclen;	/* length of this record */
	char		d_name[PCMAXNAM_UTF16];
};

/*
 * Check FAT 8.3 filename characters for validity.
 * Lacking a kernel iconv, codepage support for short filenames
 * is not provided.
 * Short names must be uppercase ASCII (no support for MSDOS
 * codepages right now, sorry) and may not contain any of
 * *+=|\[];:",<>.?/ which are explicitly forbidden by the
 * FAT specifications.
 */
#define	pc_invalchar(c)						\
	(((c) >= 'a' && (c) <= 'z') ||				\
	(c) == '"' || (c) == '*' || (c) == '+' || (c) == ',' || \
	(c) == '.' || (c) == '/' || (c) == ':' || (c) == ';' || \
	(c) == '<' || (c) == '=' || (c) == '>' || (c) == '?' || \
	(c) == '[' || (c) == '|' || (c) == ']' || (c) == '\\')

#define	pc_validchar(c)	(((c) >= ' ' && !((c) & ~0177)) && !pc_invalchar(c))


#ifdef _KERNEL

/*
 * macros for converting ASCII to/from upper or lower case.
 * users may give and get names in lower case, but they are stored on the
 * disk in upper case to be PCDOS compatible.
 * These would better come from some shared source in <sys/...> but
 * there is no such place yet.
 */
#define	toupper(C)	(((C) >= 'a' && (C) <= 'z') ? (C) - 'a' + 'A' : (C))
#define	tolower(C)	(((C) >= 'A' && (C) <= 'Z') ? (C) - 'A' + 'a' : (C))

extern int pc_tvtopct(timestruc_t *, struct pctime *);	/* timeval to pctime */
extern void pc_pcttotv(struct pctime *, int64_t *);	/* pctime to timeval */
extern int pc_valid_lfn_char(char);		/* valid long filename ch */

extern int pc_read_long_fn(struct vnode *, struct uio *,
    struct pc_dirent *, struct pcdir **, offset_t *, struct buf **);
extern int pc_read_short_fn(struct vnode *, struct uio *,
    struct pc_dirent *, struct pcdir **, offset_t *, struct buf **);
extern int pc_match_long_fn(struct pcnode *, char *, struct pcdir **,
    struct pcslot *, offset_t *);
extern int pc_match_short_fn(struct pcnode *, char *,
    struct pcdir **, struct pcslot *, offset_t *);
extern uchar_t pc_checksum_long_fn(char *, char *);
extern void set_long_fn_chunk(struct pcdir_lfn *, char *, int);
extern int pc_valid_long_fn(char *, int);
extern int pc_extract_long_fn(struct pcnode *, char *,
    struct pcdir **, offset_t *offset, struct buf **);
extern int pc_fname_ext_to_name(char *, char *, char *, int);

extern pc_cluster32_t pc_getstartcluster(struct pcfs *, struct pcdir *);
extern void pc_setstartcluster(struct pcfs *, struct pcdir *, pc_cluster32_t);

/*
 * Private tunables
 */

/*
 * Use long filenames (Windows 95). Disabling this causes pcfs
 * to not recognize long filenames at all, which may cause it to
 * break associations between the short and long names. This is likely
 * to leave unused long filename entries  in directories (which may make
 * apparently empty directories unremovable), and would require a fsck_pcfs
 * to find and fix (or a Windows utility like Norton Disk Doctor or
 * Microsoft ScanDisk).
 */
extern int enable_long_filenames;	/* default: on */

#endif

#ifdef	__cplusplus
}
#endif

#endif	/* _SYS_FS_PC_DIR_H */