1/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
2#ifndef _UAPI_LINUX_FS_H
3#define _UAPI_LINUX_FS_H
4
5/*
6 * This file has definitions for some important file table structures
7 * and constants and structures used by various generic file system
8 * ioctl's.  Please do not make any changes in this file before
9 * sending patches for review to linux-fsdevel@vger.kernel.org and
10 * linux-api@vger.kernel.org.
11 */
12
13#include <linux/limits.h>
14#include <linux/ioctl.h>
15#include <linux/types.h>
16#ifndef __KERNEL__
17#include <linux/fscrypt.h>
18#endif
19
20/* Use of MS_* flags within the kernel is restricted to core mount(2) code. */
21#if !defined(__KERNEL__)
22#include <linux/mount.h>
23#endif
24
25/*
26 * It's silly to have NR_OPEN bigger than NR_FILE, but you can change
27 * the file limit at runtime and only root can increase the per-process
28 * nr_file rlimit, so it's safe to set up a ridiculously high absolute
29 * upper limit on files-per-process.
30 *
31 * Some programs (notably those using select()) may have to be
32 * recompiled to take full advantage of the new limits..
33 */
34
35/* Fixed constants first: */
36#undef NR_OPEN
37#define INR_OPEN_CUR 1024	/* Initial setting for nfile rlimits */
38#define INR_OPEN_MAX 4096	/* Hard limit for nfile rlimits */
39
40#define BLOCK_SIZE_BITS 10
41#define BLOCK_SIZE (1<<BLOCK_SIZE_BITS)
42
43#define SEEK_SET	0	/* seek relative to beginning of file */
44#define SEEK_CUR	1	/* seek relative to current file position */
45#define SEEK_END	2	/* seek relative to end of file */
46#define SEEK_DATA	3	/* seek to the next data */
47#define SEEK_HOLE	4	/* seek to the next hole */
48#define SEEK_MAX	SEEK_HOLE
49
50#define RENAME_NOREPLACE	(1 << 0)	/* Don't overwrite target */
51#define RENAME_EXCHANGE		(1 << 1)	/* Exchange source and dest */
52#define RENAME_WHITEOUT		(1 << 2)	/* Whiteout source */
53
54struct file_clone_range {
55	__s64 src_fd;
56	__u64 src_offset;
57	__u64 src_length;
58	__u64 dest_offset;
59};
60
61struct fstrim_range {
62	__u64 start;
63	__u64 len;
64	__u64 minlen;
65};
66
67/* extent-same (dedupe) ioctls; these MUST match the btrfs ioctl definitions */
68#define FILE_DEDUPE_RANGE_SAME		0
69#define FILE_DEDUPE_RANGE_DIFFERS	1
70
71/* from struct btrfs_ioctl_file_extent_same_info */
72struct file_dedupe_range_info {
73	__s64 dest_fd;		/* in - destination file */
74	__u64 dest_offset;	/* in - start of extent in destination */
75	__u64 bytes_deduped;	/* out - total # of bytes we were able
76				 * to dedupe from this file. */
77	/* status of this dedupe operation:
78	 * < 0 for error
79	 * == FILE_DEDUPE_RANGE_SAME if dedupe succeeds
80	 * == FILE_DEDUPE_RANGE_DIFFERS if data differs
81	 */
82	__s32 status;		/* out - see above description */
83	__u32 reserved;		/* must be zero */
84};
85
86/* from struct btrfs_ioctl_file_extent_same_args */
87struct file_dedupe_range {
88	__u64 src_offset;	/* in - start of extent in source */
89	__u64 src_length;	/* in - length of extent */
90	__u16 dest_count;	/* in - total elements in info array */
91	__u16 reserved1;	/* must be zero */
92	__u32 reserved2;	/* must be zero */
93	struct file_dedupe_range_info info[];
94};
95
96/* And dynamically-tunable limits and defaults: */
97struct files_stat_struct {
98	unsigned long nr_files;		/* read only */
99	unsigned long nr_free_files;	/* read only */
100	unsigned long max_files;		/* tunable */
101};
102
103struct inodes_stat_t {
104	long nr_inodes;
105	long nr_unused;
106	long dummy[5];		/* padding for sysctl ABI compatibility */
107};
108
109
110#define NR_FILE  8192	/* this can well be larger on a larger system */
111
112/*
113 * Structure for FS_IOC_FSGETXATTR[A] and FS_IOC_FSSETXATTR.
114 */
115struct fsxattr {
116	__u32		fsx_xflags;	/* xflags field value (get/set) */
117	__u32		fsx_extsize;	/* extsize field value (get/set)*/
118	__u32		fsx_nextents;	/* nextents field value (get)	*/
119	__u32		fsx_projid;	/* project identifier (get/set) */
120	__u32		fsx_cowextsize;	/* CoW extsize field value (get/set)*/
121	unsigned char	fsx_pad[8];
122};
123
124/*
125 * Flags for the fsx_xflags field
126 */
127#define FS_XFLAG_REALTIME	0x00000001	/* data in realtime volume */
128#define FS_XFLAG_PREALLOC	0x00000002	/* preallocated file extents */
129#define FS_XFLAG_IMMUTABLE	0x00000008	/* file cannot be modified */
130#define FS_XFLAG_APPEND		0x00000010	/* all writes append */
131#define FS_XFLAG_SYNC		0x00000020	/* all writes synchronous */
132#define FS_XFLAG_NOATIME	0x00000040	/* do not update access time */
133#define FS_XFLAG_NODUMP		0x00000080	/* do not include in backups */
134#define FS_XFLAG_RTINHERIT	0x00000100	/* create with rt bit set */
135#define FS_XFLAG_PROJINHERIT	0x00000200	/* create with parents projid */
136#define FS_XFLAG_NOSYMLINKS	0x00000400	/* disallow symlink creation */
137#define FS_XFLAG_EXTSIZE	0x00000800	/* extent size allocator hint */
138#define FS_XFLAG_EXTSZINHERIT	0x00001000	/* inherit inode extent size */
139#define FS_XFLAG_NODEFRAG	0x00002000	/* do not defragment */
140#define FS_XFLAG_FILESTREAM	0x00004000	/* use filestream allocator */
141#define FS_XFLAG_DAX		0x00008000	/* use DAX for IO */
142#define FS_XFLAG_COWEXTSIZE	0x00010000	/* CoW extent size allocator hint */
143#define FS_XFLAG_HASATTR	0x80000000	/* no DIFLAG for this	*/
144
145/* the read-only stuff doesn't really belong here, but any other place is
146   probably as bad and I don't want to create yet another include file. */
147
148#define BLKROSET   _IO(0x12,93)	/* set device read-only (0 = read-write) */
149#define BLKROGET   _IO(0x12,94)	/* get read-only status (0 = read_write) */
150#define BLKRRPART  _IO(0x12,95)	/* re-read partition table */
151#define BLKGETSIZE _IO(0x12,96)	/* return device size /512 (long *arg) */
152#define BLKFLSBUF  _IO(0x12,97)	/* flush buffer cache */
153#define BLKRASET   _IO(0x12,98)	/* set read ahead for block device */
154#define BLKRAGET   _IO(0x12,99)	/* get current read ahead setting */
155#define BLKFRASET  _IO(0x12,100)/* set filesystem (mm/filemap.c) read-ahead */
156#define BLKFRAGET  _IO(0x12,101)/* get filesystem (mm/filemap.c) read-ahead */
157#define BLKSECTSET _IO(0x12,102)/* set max sectors per request (ll_rw_blk.c) */
158#define BLKSECTGET _IO(0x12,103)/* get max sectors per request (ll_rw_blk.c) */
159#define BLKSSZGET  _IO(0x12,104)/* get block device sector size */
160#if 0
161#define BLKPG      _IO(0x12,105)/* See blkpg.h */
162
163/* Some people are morons.  Do not use sizeof! */
164
165#define BLKELVGET  _IOR(0x12,106,size_t)/* elevator get */
166#define BLKELVSET  _IOW(0x12,107,size_t)/* elevator set */
167/* This was here just to show that the number is taken -
168   probably all these _IO(0x12,*) ioctls should be moved to blkpg.h. */
169#endif
170/* A jump here: 108-111 have been used for various private purposes. */
171#define BLKBSZGET  _IOR(0x12,112,size_t)
172#define BLKBSZSET  _IOW(0x12,113,size_t)
173#define BLKGETSIZE64 _IOR(0x12,114,size_t)	/* return device size in bytes (u64 *arg) */
174#define BLKTRACESETUP _IOWR(0x12,115,struct blk_user_trace_setup)
175#define BLKTRACESTART _IO(0x12,116)
176#define BLKTRACESTOP _IO(0x12,117)
177#define BLKTRACETEARDOWN _IO(0x12,118)
178#define BLKDISCARD _IO(0x12,119)
179#define BLKIOMIN _IO(0x12,120)
180#define BLKIOOPT _IO(0x12,121)
181#define BLKALIGNOFF _IO(0x12,122)
182#define BLKPBSZGET _IO(0x12,123)
183#define BLKDISCARDZEROES _IO(0x12,124)
184#define BLKSECDISCARD _IO(0x12,125)
185#define BLKROTATIONAL _IO(0x12,126)
186#define BLKZEROOUT _IO(0x12,127)
187#define BLKGETDISKSEQ _IOR(0x12,128,__u64)
188/*
189 * A jump here: 130-136 are reserved for zoned block devices
190 * (see uapi/linux/blkzoned.h)
191 */
192
193#define BMAP_IOCTL 1		/* obsolete - kept for compatibility */
194#define FIBMAP	   _IO(0x00,1)	/* bmap access */
195#define FIGETBSZ   _IO(0x00,2)	/* get the block size used for bmap */
196#define FIFREEZE	_IOWR('X', 119, int)	/* Freeze */
197#define FITHAW		_IOWR('X', 120, int)	/* Thaw */
198#define FITRIM		_IOWR('X', 121, struct fstrim_range)	/* Trim */
199#define FICLONE		_IOW(0x94, 9, int)
200#define FICLONERANGE	_IOW(0x94, 13, struct file_clone_range)
201#define FIDEDUPERANGE	_IOWR(0x94, 54, struct file_dedupe_range)
202
203#define FSLABEL_MAX 256	/* Max chars for the interface; each fs may differ */
204
205#define	FS_IOC_GETFLAGS			_IOR('f', 1, long)
206#define	FS_IOC_SETFLAGS			_IOW('f', 2, long)
207#define	FS_IOC_GETVERSION		_IOR('v', 1, long)
208#define	FS_IOC_SETVERSION		_IOW('v', 2, long)
209#define FS_IOC_FIEMAP			_IOWR('f', 11, struct fiemap)
210#define FS_IOC32_GETFLAGS		_IOR('f', 1, int)
211#define FS_IOC32_SETFLAGS		_IOW('f', 2, int)
212#define FS_IOC32_GETVERSION		_IOR('v', 1, int)
213#define FS_IOC32_SETVERSION		_IOW('v', 2, int)
214#define FS_IOC_FSGETXATTR		_IOR('X', 31, struct fsxattr)
215#define FS_IOC_FSSETXATTR		_IOW('X', 32, struct fsxattr)
216#define FS_IOC_GETFSLABEL		_IOR(0x94, 49, char[FSLABEL_MAX])
217#define FS_IOC_SETFSLABEL		_IOW(0x94, 50, char[FSLABEL_MAX])
218
219/*
220 * Inode flags (FS_IOC_GETFLAGS / FS_IOC_SETFLAGS)
221 *
222 * Note: for historical reasons, these flags were originally used and
223 * defined for use by ext2/ext3, and then other file systems started
224 * using these flags so they wouldn't need to write their own version
225 * of chattr/lsattr (which was shipped as part of e2fsprogs).  You
226 * should think twice before trying to use these flags in new
227 * contexts, or trying to assign these flags, since they are used both
228 * as the UAPI and the on-disk encoding for ext2/3/4.  Also, we are
229 * almost out of 32-bit flags.  :-)
230 *
231 * We have recently hoisted FS_IOC_FSGETXATTR / FS_IOC_FSSETXATTR from
232 * XFS to the generic FS level interface.  This uses a structure that
233 * has padding and hence has more room to grow, so it may be more
234 * appropriate for many new use cases.
235 *
236 * Please do not change these flags or interfaces before checking with
237 * linux-fsdevel@vger.kernel.org and linux-api@vger.kernel.org.
238 */
239#define	FS_SECRM_FL			0x00000001 /* Secure deletion */
240#define	FS_UNRM_FL			0x00000002 /* Undelete */
241#define	FS_COMPR_FL			0x00000004 /* Compress file */
242#define FS_SYNC_FL			0x00000008 /* Synchronous updates */
243#define FS_IMMUTABLE_FL			0x00000010 /* Immutable file */
244#define FS_APPEND_FL			0x00000020 /* writes to file may only append */
245#define FS_NODUMP_FL			0x00000040 /* do not dump file */
246#define FS_NOATIME_FL			0x00000080 /* do not update atime */
247/* Reserved for compression usage... */
248#define FS_DIRTY_FL			0x00000100
249#define FS_COMPRBLK_FL			0x00000200 /* One or more compressed clusters */
250#define FS_NOCOMP_FL			0x00000400 /* Don't compress */
251/* End compression flags --- maybe not all used */
252#define FS_ENCRYPT_FL			0x00000800 /* Encrypted file */
253#define FS_BTREE_FL			0x00001000 /* btree format dir */
254#define FS_INDEX_FL			0x00001000 /* hash-indexed directory */
255#define FS_IMAGIC_FL			0x00002000 /* AFS directory */
256#define FS_JOURNAL_DATA_FL		0x00004000 /* Reserved for ext3 */
257#define FS_NOTAIL_FL			0x00008000 /* file tail should not be merged */
258#define FS_DIRSYNC_FL			0x00010000 /* dirsync behaviour (directories only) */
259#define FS_TOPDIR_FL			0x00020000 /* Top of directory hierarchies*/
260#define FS_HUGE_FILE_FL			0x00040000 /* Reserved for ext4 */
261#define FS_EXTENT_FL			0x00080000 /* Extents */
262#define FS_VERITY_FL			0x00100000 /* Verity protected inode */
263#define FS_EA_INODE_FL			0x00200000 /* Inode used for large EA */
264#define FS_EOFBLOCKS_FL			0x00400000 /* Reserved for ext4 */
265#define FS_NOCOW_FL			0x00800000 /* Do not cow file */
266#define FS_DAX_FL			0x02000000 /* Inode is DAX */
267#define FS_INLINE_DATA_FL		0x10000000 /* Reserved for ext4 */
268#define FS_PROJINHERIT_FL		0x20000000 /* Create with parents projid */
269#define FS_CASEFOLD_FL			0x40000000 /* Folder is case insensitive */
270#define FS_RESERVED_FL			0x80000000 /* reserved for ext2 lib */
271
272#define FS_FL_USER_VISIBLE		0x0003DFFF /* User visible flags */
273#define FS_FL_USER_MODIFIABLE		0x000380FF /* User modifiable flags */
274
275
276#define SYNC_FILE_RANGE_WAIT_BEFORE	1
277#define SYNC_FILE_RANGE_WRITE		2
278#define SYNC_FILE_RANGE_WAIT_AFTER	4
279#define SYNC_FILE_RANGE_WRITE_AND_WAIT	(SYNC_FILE_RANGE_WRITE | \
280					 SYNC_FILE_RANGE_WAIT_BEFORE | \
281					 SYNC_FILE_RANGE_WAIT_AFTER)
282
283/*
284 * Flags for preadv2/pwritev2:
285 */
286
287typedef int __bitwise __kernel_rwf_t;
288
289/* high priority request, poll if possible */
290#define RWF_HIPRI	((__force __kernel_rwf_t)0x00000001)
291
292/* per-IO O_DSYNC */
293#define RWF_DSYNC	((__force __kernel_rwf_t)0x00000002)
294
295/* per-IO O_SYNC */
296#define RWF_SYNC	((__force __kernel_rwf_t)0x00000004)
297
298/* per-IO, return -EAGAIN if operation would block */
299#define RWF_NOWAIT	((__force __kernel_rwf_t)0x00000008)
300
301/* per-IO O_APPEND */
302#define RWF_APPEND	((__force __kernel_rwf_t)0x00000010)
303
304/* mask of flags supported by the kernel */
305#define RWF_SUPPORTED	(RWF_HIPRI | RWF_DSYNC | RWF_SYNC | RWF_NOWAIT |\
306			 RWF_APPEND)
307
308/* Pagemap ioctl */
309#define PAGEMAP_SCAN	_IOWR('f', 16, struct pm_scan_arg)
310
311/* Bitmasks provided in pm_scan_args masks and reported in page_region.categories. */
312#define PAGE_IS_WPALLOWED	(1 << 0)
313#define PAGE_IS_WRITTEN		(1 << 1)
314#define PAGE_IS_FILE		(1 << 2)
315#define PAGE_IS_PRESENT		(1 << 3)
316#define PAGE_IS_SWAPPED		(1 << 4)
317#define PAGE_IS_PFNZERO		(1 << 5)
318#define PAGE_IS_HUGE		(1 << 6)
319#define PAGE_IS_SOFT_DIRTY	(1 << 7)
320
321/*
322 * struct page_region - Page region with flags
323 * @start:	Start of the region
324 * @end:	End of the region (exclusive)
325 * @categories:	PAGE_IS_* category bitmask for the region
326 */
327struct page_region {
328	__u64 start;
329	__u64 end;
330	__u64 categories;
331};
332
333/* Flags for PAGEMAP_SCAN ioctl */
334#define PM_SCAN_WP_MATCHING	(1 << 0)	/* Write protect the pages matched. */
335#define PM_SCAN_CHECK_WPASYNC	(1 << 1)	/* Abort the scan when a non-WP-enabled page is found. */
336
337/*
338 * struct pm_scan_arg - Pagemap ioctl argument
339 * @size:		Size of the structure
340 * @flags:		Flags for the IOCTL
341 * @start:		Starting address of the region
342 * @end:		Ending address of the region
343 * @walk_end		Address where the scan stopped (written by kernel).
344 *			walk_end == end (address tags cleared) informs that the scan completed on entire range.
345 * @vec:		Address of page_region struct array for output
346 * @vec_len:		Length of the page_region struct array
347 * @max_pages:		Optional limit for number of returned pages (0 = disabled)
348 * @category_inverted:	PAGE_IS_* categories which values match if 0 instead of 1
349 * @category_mask:	Skip pages for which any category doesn't match
350 * @category_anyof_mask: Skip pages for which no category matches
351 * @return_mask:	PAGE_IS_* categories that are to be reported in `page_region`s returned
352 */
353struct pm_scan_arg {
354	__u64 size;
355	__u64 flags;
356	__u64 start;
357	__u64 end;
358	__u64 walk_end;
359	__u64 vec;
360	__u64 vec_len;
361	__u64 max_pages;
362	__u64 category_inverted;
363	__u64 category_mask;
364	__u64 category_anyof_mask;
365	__u64 return_mask;
366};
367
368#endif /* _UAPI_LINUX_FS_H */
369