diff options
author | Felix Blyakher <felixb@sgi.com> | 2009-02-03 11:38:41 -0500 |
---|---|---|
committer | Felix Blyakher <felixb@sgi.com> | 2009-02-03 11:38:41 -0500 |
commit | 6d2160bfe7826aca1c94b4bca77093908a452ae7 (patch) | |
tree | 8153fcd8a7c467e5de136f312e8ef5b27bea9d6b /fs | |
parent | f0e0059b9c18426cffdcc04161062251a8f9741e (diff) | |
parent | b1792e367053968f2ddb48bc911d314143ce6242 (diff) |
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 into for-linus
Diffstat (limited to 'fs')
70 files changed, 1719 insertions, 1927 deletions
diff --git a/fs/9p/Kconfig b/fs/9p/Kconfig new file mode 100644 index 000000000000..74e0723e90bc --- /dev/null +++ b/fs/9p/Kconfig | |||
@@ -0,0 +1,10 @@ | |||
1 | config 9P_FS | ||
2 | tristate "Plan 9 Resource Sharing Support (9P2000) (Experimental)" | ||
3 | depends on INET && NET_9P && EXPERIMENTAL | ||
4 | help | ||
5 | If you say Y here, you will get experimental support for | ||
6 | Plan 9 resource sharing via the 9P2000 protocol. | ||
7 | |||
8 | See <http://v9fs.sf.net> for more information. | ||
9 | |||
10 | If unsure, say N. | ||
diff --git a/fs/Kconfig b/fs/Kconfig index 51307b0fdf0f..93945dd0b1ae 100644 --- a/fs/Kconfig +++ b/fs/Kconfig | |||
@@ -27,141 +27,8 @@ config FS_MBCACHE | |||
27 | default y if EXT4_FS=y && EXT4_FS_XATTR | 27 | default y if EXT4_FS=y && EXT4_FS_XATTR |
28 | default m if EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR | 28 | default m if EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR |
29 | 29 | ||
30 | config REISERFS_FS | 30 | source "fs/reiserfs/Kconfig" |
31 | tristate "Reiserfs support" | 31 | source "fs/jfs/Kconfig" |
32 | help | ||
33 | Stores not just filenames but the files themselves in a balanced | ||
34 | tree. Uses journalling. | ||
35 | |||
36 | Balanced trees are more efficient than traditional file system | ||
37 | architectural foundations. | ||
38 | |||
39 | In general, ReiserFS is as fast as ext2, but is very efficient with | ||
40 | large directories and small files. Additional patches are needed | ||
41 | for NFS and quotas, please see <http://www.namesys.com/> for links. | ||
42 | |||
43 | It is more easily extended to have features currently found in | ||
44 | database and keyword search systems than block allocation based file | ||
45 | systems are. The next version will be so extended, and will support | ||
46 | plugins consistent with our motto ``It takes more than a license to | ||
47 | make source code open.'' | ||
48 | |||
49 | Read <http://www.namesys.com/> to learn more about reiserfs. | ||
50 | |||
51 | Sponsored by Threshold Networks, Emusic.com, and Bigstorage.com. | ||
52 | |||
53 | If you like it, you can pay us to add new features to it that you | ||
54 | need, buy a support contract, or pay us to port it to another OS. | ||
55 | |||
56 | config REISERFS_CHECK | ||
57 | bool "Enable reiserfs debug mode" | ||
58 | depends on REISERFS_FS | ||
59 | help | ||
60 | If you set this to Y, then ReiserFS will perform every check it can | ||
61 | possibly imagine of its internal consistency throughout its | ||
62 | operation. It will also go substantially slower. More than once we | ||
63 | have forgotten that this was on, and then gone despondent over the | ||
64 | latest benchmarks.:-) Use of this option allows our team to go all | ||
65 | out in checking for consistency when debugging without fear of its | ||
66 | effect on end users. If you are on the verge of sending in a bug | ||
67 | report, say Y and you might get a useful error message. Almost | ||
68 | everyone should say N. | ||
69 | |||
70 | config REISERFS_PROC_INFO | ||
71 | bool "Stats in /proc/fs/reiserfs" | ||
72 | depends on REISERFS_FS && PROC_FS | ||
73 | help | ||
74 | Create under /proc/fs/reiserfs a hierarchy of files, displaying | ||
75 | various ReiserFS statistics and internal data at the expense of | ||
76 | making your kernel or module slightly larger (+8 KB). This also | ||
77 | increases the amount of kernel memory required for each mount. | ||
78 | Almost everyone but ReiserFS developers and people fine-tuning | ||
79 | reiserfs or tracing problems should say N. | ||
80 | |||
81 | config REISERFS_FS_XATTR | ||
82 | bool "ReiserFS extended attributes" | ||
83 | depends on REISERFS_FS | ||
84 | help | ||
85 | Extended attributes are name:value pairs associated with inodes by | ||
86 | the kernel or by users (see the attr(5) manual page, or visit | ||
87 | <http://acl.bestbits.at/> for details). | ||
88 | |||
89 | If unsure, say N. | ||
90 | |||
91 | config REISERFS_FS_POSIX_ACL | ||
92 | bool "ReiserFS POSIX Access Control Lists" | ||
93 | depends on REISERFS_FS_XATTR | ||
94 | select FS_POSIX_ACL | ||
95 | help | ||
96 | Posix Access Control Lists (ACLs) support permissions for users and | ||
97 | groups beyond the owner/group/world scheme. | ||
98 | |||
99 | To learn more about Access Control Lists, visit the Posix ACLs for | ||
100 | Linux website <http://acl.bestbits.at/>. | ||
101 | |||
102 | If you don't know what Access Control Lists are, say N | ||
103 | |||
104 | config REISERFS_FS_SECURITY | ||
105 | bool "ReiserFS Security Labels" | ||
106 | depends on REISERFS_FS_XATTR | ||
107 | help | ||
108 | Security labels support alternative access control models | ||
109 | implemented by security modules like SELinux. This option | ||
110 | enables an extended attribute handler for file security | ||
111 | labels in the ReiserFS filesystem. | ||
112 | |||
113 | If you are not using a security module that requires using | ||
114 | extended attributes for file security labels, say N. | ||
115 | |||
116 | config JFS_FS | ||
117 | tristate "JFS filesystem support" | ||
118 | select NLS | ||
119 | help | ||
120 | This is a port of IBM's Journaled Filesystem . More information is | ||
121 | available in the file <file:Documentation/filesystems/jfs.txt>. | ||
122 | |||
123 | If you do not intend to use the JFS filesystem, say N. | ||
124 | |||
125 | config JFS_POSIX_ACL | ||
126 | bool "JFS POSIX Access Control Lists" | ||
127 | depends on JFS_FS | ||
128 | select FS_POSIX_ACL | ||
129 | help | ||
130 | Posix Access Control Lists (ACLs) support permissions for users and | ||
131 | groups beyond the owner/group/world scheme. | ||
132 | |||
133 | To learn more about Access Control Lists, visit the Posix ACLs for | ||
134 | Linux website <http://acl.bestbits.at/>. | ||
135 | |||
136 | If you don't know what Access Control Lists are, say N | ||
137 | |||
138 | config JFS_SECURITY | ||
139 | bool "JFS Security Labels" | ||
140 | depends on JFS_FS | ||
141 | help | ||
142 | Security labels support alternative access control models | ||
143 | implemented by security modules like SELinux. This option | ||
144 | enables an extended attribute handler for file security | ||
145 | labels in the jfs filesystem. | ||
146 | |||
147 | If you are not using a security module that requires using | ||
148 | extended attributes for file security labels, say N. | ||
149 | |||
150 | config JFS_DEBUG | ||
151 | bool "JFS debugging" | ||
152 | depends on JFS_FS | ||
153 | help | ||
154 | If you are experiencing any problems with the JFS filesystem, say | ||
155 | Y here. This will result in additional debugging messages to be | ||
156 | written to the system log. Under normal circumstances, this | ||
157 | results in very little overhead. | ||
158 | |||
159 | config JFS_STATISTICS | ||
160 | bool "JFS statistics" | ||
161 | depends on JFS_FS | ||
162 | help | ||
163 | Enabling this option will cause statistics from the JFS file system | ||
164 | to be made available to the user in the /proc/fs/jfs/ directory. | ||
165 | 32 | ||
166 | config FS_POSIX_ACL | 33 | config FS_POSIX_ACL |
167 | # Posix ACL utility routines (for now, only ext2/ext3/jfs/reiserfs/nfs4) | 34 | # Posix ACL utility routines (for now, only ext2/ext3/jfs/reiserfs/nfs4) |
@@ -182,111 +49,8 @@ config FILE_LOCKING | |||
182 | 49 | ||
183 | source "fs/xfs/Kconfig" | 50 | source "fs/xfs/Kconfig" |
184 | source "fs/gfs2/Kconfig" | 51 | source "fs/gfs2/Kconfig" |
185 | 52 | source "fs/ocfs2/Kconfig" | |
186 | config OCFS2_FS | 53 | source "fs/btrfs/Kconfig" |
187 | tristate "OCFS2 file system support" | ||
188 | depends on NET && SYSFS | ||
189 | select CONFIGFS_FS | ||
190 | select JBD2 | ||
191 | select CRC32 | ||
192 | select QUOTA | ||
193 | select QUOTA_TREE | ||
194 | help | ||
195 | OCFS2 is a general purpose extent based shared disk cluster file | ||
196 | system with many similarities to ext3. It supports 64 bit inode | ||
197 | numbers, and has automatically extending metadata groups which may | ||
198 | also make it attractive for non-clustered use. | ||
199 | |||
200 | You'll want to install the ocfs2-tools package in order to at least | ||
201 | get "mount.ocfs2". | ||
202 | |||
203 | Project web page: http://oss.oracle.com/projects/ocfs2 | ||
204 | Tools web page: http://oss.oracle.com/projects/ocfs2-tools | ||
205 | OCFS2 mailing lists: http://oss.oracle.com/projects/ocfs2/mailman/ | ||
206 | |||
207 | For more information on OCFS2, see the file | ||
208 | <file:Documentation/filesystems/ocfs2.txt>. | ||
209 | |||
210 | config OCFS2_FS_O2CB | ||
211 | tristate "O2CB Kernelspace Clustering" | ||
212 | depends on OCFS2_FS | ||
213 | default y | ||
214 | help | ||
215 | OCFS2 includes a simple kernelspace clustering package, the OCFS2 | ||
216 | Cluster Base. It only requires a very small userspace component | ||
217 | to configure it. This comes with the standard ocfs2-tools package. | ||
218 | O2CB is limited to maintaining a cluster for OCFS2 file systems. | ||
219 | It cannot manage any other cluster applications. | ||
220 | |||
221 | It is always safe to say Y here, as the clustering method is | ||
222 | run-time selectable. | ||
223 | |||
224 | config OCFS2_FS_USERSPACE_CLUSTER | ||
225 | tristate "OCFS2 Userspace Clustering" | ||
226 | depends on OCFS2_FS && DLM | ||
227 | default y | ||
228 | help | ||
229 | This option will allow OCFS2 to use userspace clustering services | ||
230 | in conjunction with the DLM in fs/dlm. If you are using a | ||
231 | userspace cluster manager, say Y here. | ||
232 | |||
233 | It is safe to say Y, as the clustering method is run-time | ||
234 | selectable. | ||
235 | |||
236 | config OCFS2_FS_STATS | ||
237 | bool "OCFS2 statistics" | ||
238 | depends on OCFS2_FS | ||
239 | default y | ||
240 | help | ||
241 | This option allows some fs statistics to be captured. Enabling | ||
242 | this option may increase the memory consumption. | ||
243 | |||
244 | config OCFS2_DEBUG_MASKLOG | ||
245 | bool "OCFS2 logging support" | ||
246 | depends on OCFS2_FS | ||
247 | default y | ||
248 | help | ||
249 | The ocfs2 filesystem has an extensive logging system. The system | ||
250 | allows selection of events to log via files in /sys/o2cb/logmask/. | ||
251 | This option will enlarge your kernel, but it allows debugging of | ||
252 | ocfs2 filesystem issues. | ||
253 | |||
254 | config OCFS2_DEBUG_FS | ||
255 | bool "OCFS2 expensive checks" | ||
256 | depends on OCFS2_FS | ||
257 | default n | ||
258 | help | ||
259 | This option will enable expensive consistency checks. Enable | ||
260 | this option for debugging only as it is likely to decrease | ||
261 | performance of the filesystem. | ||
262 | |||
263 | config OCFS2_FS_POSIX_ACL | ||
264 | bool "OCFS2 POSIX Access Control Lists" | ||
265 | depends on OCFS2_FS | ||
266 | select FS_POSIX_ACL | ||
267 | default n | ||
268 | help | ||
269 | Posix Access Control Lists (ACLs) support permissions for users and | ||
270 | groups beyond the owner/group/world scheme. | ||
271 | |||
272 | config BTRFS_FS | ||
273 | tristate "Btrfs filesystem (EXPERIMENTAL) Unstable disk format" | ||
274 | depends on EXPERIMENTAL | ||
275 | select LIBCRC32C | ||
276 | select ZLIB_INFLATE | ||
277 | select ZLIB_DEFLATE | ||
278 | help | ||
279 | Btrfs is a new filesystem with extents, writable snapshotting, | ||
280 | support for multiple devices and many more features. | ||
281 | |||
282 | Btrfs is highly experimental, and THE DISK FORMAT IS NOT YET | ||
283 | FINALIZED. You should say N here unless you are interested in | ||
284 | testing Btrfs with non-critical data. | ||
285 | |||
286 | To compile this file system support as a module, choose M here. The | ||
287 | module will be called btrfs. | ||
288 | |||
289 | If unsure, say N. | ||
290 | 54 | ||
291 | endif # BLOCK | 55 | endif # BLOCK |
292 | 56 | ||
@@ -348,64 +112,9 @@ config QUOTACTL | |||
348 | depends on XFS_QUOTA || QUOTA | 112 | depends on XFS_QUOTA || QUOTA |
349 | default y | 113 | default y |
350 | 114 | ||
351 | config AUTOFS_FS | 115 | source "fs/autofs/Kconfig" |
352 | tristate "Kernel automounter support" | 116 | source "fs/autofs4/Kconfig" |
353 | help | 117 | source "fs/fuse/Kconfig" |
354 | The automounter is a tool to automatically mount remote file systems | ||
355 | on demand. This implementation is partially kernel-based to reduce | ||
356 | overhead in the already-mounted case; this is unlike the BSD | ||
357 | automounter (amd), which is a pure user space daemon. | ||
358 | |||
359 | To use the automounter you need the user-space tools from the autofs | ||
360 | package; you can find the location in <file:Documentation/Changes>. | ||
361 | You also want to answer Y to "NFS file system support", below. | ||
362 | |||
363 | If you want to use the newer version of the automounter with more | ||
364 | features, say N here and say Y to "Kernel automounter v4 support", | ||
365 | below. | ||
366 | |||
367 | To compile this support as a module, choose M here: the module will be | ||
368 | called autofs. | ||
369 | |||
370 | If you are not a part of a fairly large, distributed network, you | ||
371 | probably do not need an automounter, and can say N here. | ||
372 | |||
373 | config AUTOFS4_FS | ||
374 | tristate "Kernel automounter version 4 support (also supports v3)" | ||
375 | help | ||
376 | The automounter is a tool to automatically mount remote file systems | ||
377 | on demand. This implementation is partially kernel-based to reduce | ||
378 | overhead in the already-mounted case; this is unlike the BSD | ||
379 | automounter (amd), which is a pure user space daemon. | ||
380 | |||
381 | To use the automounter you need the user-space tools from | ||
382 | <ftp://ftp.kernel.org/pub/linux/daemons/autofs/v4/>; you also | ||
383 | want to answer Y to "NFS file system support", below. | ||
384 | |||
385 | To compile this support as a module, choose M here: the module will be | ||
386 | called autofs4. You will need to add "alias autofs autofs4" to your | ||
387 | modules configuration file. | ||
388 | |||
389 | If you are not a part of a fairly large, distributed network or | ||
390 | don't have a laptop which needs to dynamically reconfigure to the | ||
391 | local network, you probably do not need an automounter, and can say | ||
392 | N here. | ||
393 | |||
394 | config FUSE_FS | ||
395 | tristate "FUSE (Filesystem in Userspace) support" | ||
396 | help | ||
397 | With FUSE it is possible to implement a fully functional filesystem | ||
398 | in a userspace program. | ||
399 | |||
400 | There's also companion library: libfuse. This library along with | ||
401 | utilities is available from the FUSE homepage: | ||
402 | <http://fuse.sourceforge.net/> | ||
403 | |||
404 | See <file:Documentation/filesystems/fuse.txt> for more information. | ||
405 | See <file:Documentation/Changes> for needed library/utility version. | ||
406 | |||
407 | If you want to develop a userspace FS, or if you want to use | ||
408 | a filesystem based on FUSE, answer Y or M. | ||
409 | 118 | ||
410 | config GENERIC_ACL | 119 | config GENERIC_ACL |
411 | bool | 120 | bool |
@@ -414,64 +123,8 @@ config GENERIC_ACL | |||
414 | if BLOCK | 123 | if BLOCK |
415 | menu "CD-ROM/DVD Filesystems" | 124 | menu "CD-ROM/DVD Filesystems" |
416 | 125 | ||
417 | config ISO9660_FS | 126 | source "fs/isofs/Kconfig" |
418 | tristate "ISO 9660 CDROM file system support" | 127 | source "fs/udf/Kconfig" |
419 | help | ||
420 | This is the standard file system used on CD-ROMs. It was previously | ||
421 | known as "High Sierra File System" and is called "hsfs" on other | ||
422 | Unix systems. The so-called Rock-Ridge extensions which allow for | ||
423 | long Unix filenames and symbolic links are also supported by this | ||
424 | driver. If you have a CD-ROM drive and want to do more with it than | ||
425 | just listen to audio CDs and watch its LEDs, say Y (and read | ||
426 | <file:Documentation/filesystems/isofs.txt> and the CD-ROM-HOWTO, | ||
427 | available from <http://www.tldp.org/docs.html#howto>), thereby | ||
428 | enlarging your kernel by about 27 KB; otherwise say N. | ||
429 | |||
430 | To compile this file system support as a module, choose M here: the | ||
431 | module will be called isofs. | ||
432 | |||
433 | config JOLIET | ||
434 | bool "Microsoft Joliet CDROM extensions" | ||
435 | depends on ISO9660_FS | ||
436 | select NLS | ||
437 | help | ||
438 | Joliet is a Microsoft extension for the ISO 9660 CD-ROM file system | ||
439 | which allows for long filenames in unicode format (unicode is the | ||
440 | new 16 bit character code, successor to ASCII, which encodes the | ||
441 | characters of almost all languages of the world; see | ||
442 | <http://www.unicode.org/> for more information). Say Y here if you | ||
443 | want to be able to read Joliet CD-ROMs under Linux. | ||
444 | |||
445 | config ZISOFS | ||
446 | bool "Transparent decompression extension" | ||
447 | depends on ISO9660_FS | ||
448 | select ZLIB_INFLATE | ||
449 | help | ||
450 | This is a Linux-specific extension to RockRidge which lets you store | ||
451 | data in compressed form on a CD-ROM and have it transparently | ||
452 | decompressed when the CD-ROM is accessed. See | ||
453 | <http://www.kernel.org/pub/linux/utils/fs/zisofs/> for the tools | ||
454 | necessary to create such a filesystem. Say Y here if you want to be | ||
455 | able to read such compressed CD-ROMs. | ||
456 | |||
457 | config UDF_FS | ||
458 | tristate "UDF file system support" | ||
459 | select CRC_ITU_T | ||
460 | help | ||
461 | This is the new file system used on some CD-ROMs and DVDs. Say Y if | ||
462 | you intend to mount DVD discs or CDRW's written in packet mode, or | ||
463 | if written to by other UDF utilities, such as DirectCD. | ||
464 | Please read <file:Documentation/filesystems/udf.txt>. | ||
465 | |||
466 | To compile this file system support as a module, choose M here: the | ||
467 | module will be called udf. | ||
468 | |||
469 | If unsure, say N. | ||
470 | |||
471 | config UDF_NLS | ||
472 | bool | ||
473 | default y | ||
474 | depends on (UDF_FS=m && NLS) || (UDF_FS=y && NLS=y) | ||
475 | 128 | ||
476 | endmenu | 129 | endmenu |
477 | endif # BLOCK | 130 | endif # BLOCK |
@@ -479,182 +132,8 @@ endif # BLOCK | |||
479 | if BLOCK | 132 | if BLOCK |
480 | menu "DOS/FAT/NT Filesystems" | 133 | menu "DOS/FAT/NT Filesystems" |
481 | 134 | ||
482 | config FAT_FS | 135 | source "fs/fat/Kconfig" |
483 | tristate | 136 | source "fs/ntfs/Kconfig" |
484 | select NLS | ||
485 | help | ||
486 | If you want to use one of the FAT-based file systems (the MS-DOS and | ||
487 | VFAT (Windows 95) file systems), then you must say Y or M here | ||
488 | to include FAT support. You will then be able to mount partitions or | ||
489 | diskettes with FAT-based file systems and transparently access the | ||
490 | files on them, i.e. MSDOS files will look and behave just like all | ||
491 | other Unix files. | ||
492 | |||
493 | This FAT support is not a file system in itself, it only provides | ||
494 | the foundation for the other file systems. You will have to say Y or | ||
495 | M to at least one of "MSDOS fs support" or "VFAT fs support" in | ||
496 | order to make use of it. | ||
497 | |||
498 | Another way to read and write MSDOS floppies and hard drive | ||
499 | partitions from within Linux (but not transparently) is with the | ||
500 | mtools ("man mtools") program suite. You don't need to say Y here in | ||
501 | order to do that. | ||
502 | |||
503 | If you need to move large files on floppies between a DOS and a | ||
504 | Linux box, say Y here, mount the floppy under Linux with an MSDOS | ||
505 | file system and use GNU tar's M option. GNU tar is a program | ||
506 | available for Unix and DOS ("man tar" or "info tar"). | ||
507 | |||
508 | The FAT support will enlarge your kernel by about 37 KB. If unsure, | ||
509 | say Y. | ||
510 | |||
511 | To compile this as a module, choose M here: the module will be called | ||
512 | fat. Note that if you compile the FAT support as a module, you | ||
513 | cannot compile any of the FAT-based file systems into the kernel | ||
514 | -- they will have to be modules as well. | ||
515 | |||
516 | config MSDOS_FS | ||
517 | tristate "MSDOS fs support" | ||
518 | select FAT_FS | ||
519 | help | ||
520 | This allows you to mount MSDOS partitions of your hard drive (unless | ||
521 | they are compressed; to access compressed MSDOS partitions under | ||
522 | Linux, you can either use the DOS emulator DOSEMU, described in the | ||
523 | DOSEMU-HOWTO, available from | ||
524 | <http://www.tldp.org/docs.html#howto>, or try dmsdosfs in | ||
525 | <ftp://ibiblio.org/pub/Linux/system/filesystems/dosfs/>. If you | ||
526 | intend to use dosemu with a non-compressed MSDOS partition, say Y | ||
527 | here) and MSDOS floppies. This means that file access becomes | ||
528 | transparent, i.e. the MSDOS files look and behave just like all | ||
529 | other Unix files. | ||
530 | |||
531 | If you have Windows 95 or Windows NT installed on your MSDOS | ||
532 | partitions, you should use the VFAT file system (say Y to "VFAT fs | ||
533 | support" below), or you will not be able to see the long filenames | ||
534 | generated by Windows 95 / Windows NT. | ||
535 | |||
536 | This option will enlarge your kernel by about 7 KB. If unsure, | ||
537 | answer Y. This will only work if you said Y to "DOS FAT fs support" | ||
538 | as well. To compile this as a module, choose M here: the module will | ||
539 | be called msdos. | ||
540 | |||
541 | config VFAT_FS | ||
542 | tristate "VFAT (Windows-95) fs support" | ||
543 | select FAT_FS | ||
544 | help | ||
545 | This option provides support for normal Windows file systems with | ||
546 | long filenames. That includes non-compressed FAT-based file systems | ||
547 | used by Windows 95, Windows 98, Windows NT 4.0, and the Unix | ||
548 | programs from the mtools package. | ||
549 | |||
550 | The VFAT support enlarges your kernel by about 10 KB and it only | ||
551 | works if you said Y to the "DOS FAT fs support" above. Please read | ||
552 | the file <file:Documentation/filesystems/vfat.txt> for details. If | ||
553 | unsure, say Y. | ||
554 | |||
555 | To compile this as a module, choose M here: the module will be called | ||
556 | vfat. | ||
557 | |||
558 | config FAT_DEFAULT_CODEPAGE | ||
559 | int "Default codepage for FAT" | ||
560 | depends on MSDOS_FS || VFAT_FS | ||
561 | default 437 | ||
562 | help | ||
563 | This option should be set to the codepage of your FAT filesystems. | ||
564 | It can be overridden with the "codepage" mount option. | ||
565 | See <file:Documentation/filesystems/vfat.txt> for more information. | ||
566 | |||
567 | config FAT_DEFAULT_IOCHARSET | ||
568 | string "Default iocharset for FAT" | ||
569 | depends on VFAT_FS | ||
570 | default "iso8859-1" | ||
571 | help | ||
572 | Set this to the default input/output character set you'd | ||
573 | like FAT to use. It should probably match the character set | ||
574 | that most of your FAT filesystems use, and can be overridden | ||
575 | with the "iocharset" mount option for FAT filesystems. | ||
576 | Note that "utf8" is not recommended for FAT filesystems. | ||
577 | If unsure, you shouldn't set "utf8" here. | ||
578 | See <file:Documentation/filesystems/vfat.txt> for more information. | ||
579 | |||
580 | config NTFS_FS | ||
581 | tristate "NTFS file system support" | ||
582 | select NLS | ||
583 | help | ||
584 | NTFS is the file system of Microsoft Windows NT, 2000, XP and 2003. | ||
585 | |||
586 | Saying Y or M here enables read support. There is partial, but | ||
587 | safe, write support available. For write support you must also | ||
588 | say Y to "NTFS write support" below. | ||
589 | |||
590 | There are also a number of user-space tools available, called | ||
591 | ntfsprogs. These include ntfsundelete and ntfsresize, that work | ||
592 | without NTFS support enabled in the kernel. | ||
593 | |||
594 | This is a rewrite from scratch of Linux NTFS support and replaced | ||
595 | the old NTFS code starting with Linux 2.5.11. A backport to | ||
596 | the Linux 2.4 kernel series is separately available as a patch | ||
597 | from the project web site. | ||
598 | |||
599 | For more information see <file:Documentation/filesystems/ntfs.txt> | ||
600 | and <http://www.linux-ntfs.org/>. | ||
601 | |||
602 | To compile this file system support as a module, choose M here: the | ||
603 | module will be called ntfs. | ||
604 | |||
605 | If you are not using Windows NT, 2000, XP or 2003 in addition to | ||
606 | Linux on your computer it is safe to say N. | ||
607 | |||
608 | config NTFS_DEBUG | ||
609 | bool "NTFS debugging support" | ||
610 | depends on NTFS_FS | ||
611 | help | ||
612 | If you are experiencing any problems with the NTFS file system, say | ||
613 | Y here. This will result in additional consistency checks to be | ||
614 | performed by the driver as well as additional debugging messages to | ||
615 | be written to the system log. Note that debugging messages are | ||
616 | disabled by default. To enable them, supply the option debug_msgs=1 | ||
617 | at the kernel command line when booting the kernel or as an option | ||
618 | to insmod when loading the ntfs module. Once the driver is active, | ||
619 | you can enable debugging messages by doing (as root): | ||
620 | echo 1 > /proc/sys/fs/ntfs-debug | ||
621 | Replacing the "1" with "0" would disable debug messages. | ||
622 | |||
623 | If you leave debugging messages disabled, this results in little | ||
624 | overhead, but enabling debug messages results in very significant | ||
625 | slowdown of the system. | ||
626 | |||
627 | When reporting bugs, please try to have available a full dump of | ||
628 | debugging messages while the misbehaviour was occurring. | ||
629 | |||
630 | config NTFS_RW | ||
631 | bool "NTFS write support" | ||
632 | depends on NTFS_FS | ||
633 | help | ||
634 | This enables the partial, but safe, write support in the NTFS driver. | ||
635 | |||
636 | The only supported operation is overwriting existing files, without | ||
637 | changing the file length. No file or directory creation, deletion or | ||
638 | renaming is possible. Note only non-resident files can be written to | ||
639 | so you may find that some very small files (<500 bytes or so) cannot | ||
640 | be written to. | ||
641 | |||
642 | While we cannot guarantee that it will not damage any data, we have | ||
643 | so far not received a single report where the driver would have | ||
644 | damaged someones data so we assume it is perfectly safe to use. | ||
645 | |||
646 | Note: While write support is safe in this version (a rewrite from | ||
647 | scratch of the NTFS support), it should be noted that the old NTFS | ||
648 | write support, included in Linux 2.5.10 and before (since 1997), | ||
649 | is not safe. | ||
650 | |||
651 | This is currently useful with TopologiLinux. TopologiLinux is run | ||
652 | on top of any DOS/Microsoft Windows system without partitioning your | ||
653 | hard disk. Unlike other Linux distributions TopologiLinux does not | ||
654 | need its own partition. For more information see | ||
655 | <http://topologi-linux.sourceforge.net/> | ||
656 | |||
657 | It is perfectly safe to say N here. | ||
658 | 137 | ||
659 | endmenu | 138 | endmenu |
660 | endif # BLOCK | 139 | endif # BLOCK |
@@ -662,30 +141,7 @@ endif # BLOCK | |||
662 | menu "Pseudo filesystems" | 141 | menu "Pseudo filesystems" |
663 | 142 | ||
664 | source "fs/proc/Kconfig" | 143 | source "fs/proc/Kconfig" |
665 | 144 | source "fs/sysfs/Kconfig" | |
666 | config SYSFS | ||
667 | bool "sysfs file system support" if EMBEDDED | ||
668 | default y | ||
669 | help | ||
670 | The sysfs filesystem is a virtual filesystem that the kernel uses to | ||
671 | export internal kernel objects, their attributes, and their | ||
672 | relationships to one another. | ||
673 | |||
674 | Users can use sysfs to ascertain useful information about the running | ||
675 | kernel, such as the devices the kernel has discovered on each bus and | ||
676 | which driver each is bound to. sysfs can also be used to tune devices | ||
677 | and other kernel subsystems. | ||
678 | |||
679 | Some system agents rely on the information in sysfs to operate. | ||
680 | /sbin/hotplug uses device and object attributes in sysfs to assist in | ||
681 | delegating policy decisions, like persistently naming devices. | ||
682 | |||
683 | sysfs is currently used by the block subsystem to mount the root | ||
684 | partition. If sysfs is disabled you must specify the boot device on | ||
685 | the kernel boot command line via its major and minor numbers. For | ||
686 | example, "root=03:01" for /dev/hda1. | ||
687 | |||
688 | Designers of embedded systems may wish to say N here to conserve space. | ||
689 | 145 | ||
690 | config TMPFS | 146 | config TMPFS |
691 | bool "Virtual memory file system support (former shm fs)" | 147 | bool "Virtual memory file system support (former shm fs)" |
@@ -726,17 +182,7 @@ config HUGETLBFS | |||
726 | config HUGETLB_PAGE | 182 | config HUGETLB_PAGE |
727 | def_bool HUGETLBFS | 183 | def_bool HUGETLBFS |
728 | 184 | ||
729 | config CONFIGFS_FS | 185 | source "fs/configfs/Kconfig" |
730 | tristate "Userspace-driven configuration filesystem" | ||
731 | depends on SYSFS | ||
732 | help | ||
733 | configfs is a ram-based filesystem that provides the converse | ||
734 | of sysfs's functionality. Where sysfs is a filesystem-based | ||
735 | view of kernel objects, configfs is a filesystem-based manager | ||
736 | of kernel objects, or config_items. | ||
737 | |||
738 | Both sysfs and configfs can and should exist together on the | ||
739 | same system. One is not a replacement for the other. | ||
740 | 186 | ||
741 | endmenu | 187 | endmenu |
742 | 188 | ||
@@ -755,425 +201,27 @@ menuconfig MISC_FILESYSTEMS | |||
755 | 201 | ||
756 | if MISC_FILESYSTEMS | 202 | if MISC_FILESYSTEMS |
757 | 203 | ||
758 | config ADFS_FS | 204 | source "fs/adfs/Kconfig" |
759 | tristate "ADFS file system support (EXPERIMENTAL)" | 205 | source "fs/affs/Kconfig" |
760 | depends on BLOCK && EXPERIMENTAL | 206 | source "fs/ecryptfs/Kconfig" |
761 | help | 207 | source "fs/hfs/Kconfig" |
762 | The Acorn Disc Filing System is the standard file system of the | 208 | source "fs/hfsplus/Kconfig" |
763 | RiscOS operating system which runs on Acorn's ARM-based Risc PC | 209 | source "fs/befs/Kconfig" |
764 | systems and the Acorn Archimedes range of machines. If you say Y | 210 | source "fs/bfs/Kconfig" |
765 | here, Linux will be able to read from ADFS partitions on hard drives | 211 | source "fs/efs/Kconfig" |
766 | and from ADFS-formatted floppy discs. If you also want to be able to | ||
767 | write to those devices, say Y to "ADFS write support" below. | ||
768 | |||
769 | The ADFS partition should be the first partition (i.e., | ||
770 | /dev/[hs]d?1) on each of your drives. Please read the file | ||
771 | <file:Documentation/filesystems/adfs.txt> for further details. | ||
772 | |||
773 | To compile this code as a module, choose M here: the module will be | ||
774 | called adfs. | ||
775 | |||
776 | If unsure, say N. | ||
777 | |||
778 | config ADFS_FS_RW | ||
779 | bool "ADFS write support (DANGEROUS)" | ||
780 | depends on ADFS_FS | ||
781 | help | ||
782 | If you say Y here, you will be able to write to ADFS partitions on | ||
783 | hard drives and ADFS-formatted floppy disks. This is experimental | ||
784 | codes, so if you're unsure, say N. | ||
785 | |||
786 | config AFFS_FS | ||
787 | tristate "Amiga FFS file system support (EXPERIMENTAL)" | ||
788 | depends on BLOCK && EXPERIMENTAL | ||
789 | help | ||
790 | The Fast File System (FFS) is the common file system used on hard | ||
791 | disks by Amiga(tm) systems since AmigaOS Version 1.3 (34.20). Say Y | ||
792 | if you want to be able to read and write files from and to an Amiga | ||
793 | FFS partition on your hard drive. Amiga floppies however cannot be | ||
794 | read with this driver due to an incompatibility of the floppy | ||
795 | controller used in an Amiga and the standard floppy controller in | ||
796 | PCs and workstations. Read <file:Documentation/filesystems/affs.txt> | ||
797 | and <file:fs/affs/Changes>. | ||
798 | |||
799 | With this driver you can also mount disk files used by Bernd | ||
800 | Schmidt's Un*X Amiga Emulator | ||
801 | (<http://www.freiburg.linux.de/~uae/>). | ||
802 | If you want to do this, you will also need to say Y or M to "Loop | ||
803 | device support", above. | ||
804 | |||
805 | To compile this file system support as a module, choose M here: the | ||
806 | module will be called affs. If unsure, say N. | ||
807 | |||
808 | config ECRYPT_FS | ||
809 | tristate "eCrypt filesystem layer support (EXPERIMENTAL)" | ||
810 | depends on EXPERIMENTAL && KEYS && CRYPTO && NET | ||
811 | help | ||
812 | Encrypted filesystem that operates on the VFS layer. See | ||
813 | <file:Documentation/filesystems/ecryptfs.txt> to learn more about | ||
814 | eCryptfs. Userspace components are required and can be | ||
815 | obtained from <http://ecryptfs.sf.net>. | ||
816 | |||
817 | To compile this file system support as a module, choose M here: the | ||
818 | module will be called ecryptfs. | ||
819 | |||
820 | config HFS_FS | ||
821 | tristate "Apple Macintosh file system support (EXPERIMENTAL)" | ||
822 | depends on BLOCK && EXPERIMENTAL | ||
823 | select NLS | ||
824 | help | ||
825 | If you say Y here, you will be able to mount Macintosh-formatted | ||
826 | floppy disks and hard drive partitions with full read-write access. | ||
827 | Please read <file:Documentation/filesystems/hfs.txt> to learn about | ||
828 | the available mount options. | ||
829 | |||
830 | To compile this file system support as a module, choose M here: the | ||
831 | module will be called hfs. | ||
832 | |||
833 | config HFSPLUS_FS | ||
834 | tristate "Apple Extended HFS file system support" | ||
835 | depends on BLOCK | ||
836 | select NLS | ||
837 | select NLS_UTF8 | ||
838 | help | ||
839 | If you say Y here, you will be able to mount extended format | ||
840 | Macintosh-formatted hard drive partitions with full read-write access. | ||
841 | |||
842 | This file system is often called HFS+ and was introduced with | ||
843 | MacOS 8. It includes all Mac specific filesystem data such as | ||
844 | data forks and creator codes, but it also has several UNIX | ||
845 | style features such as file ownership and permissions. | ||
846 | |||
847 | config BEFS_FS | ||
848 | tristate "BeOS file system (BeFS) support (read only) (EXPERIMENTAL)" | ||
849 | depends on BLOCK && EXPERIMENTAL | ||
850 | select NLS | ||
851 | help | ||
852 | The BeOS File System (BeFS) is the native file system of Be, Inc's | ||
853 | BeOS. Notable features include support for arbitrary attributes | ||
854 | on files and directories, and database-like indices on selected | ||
855 | attributes. (Also note that this driver doesn't make those features | ||
856 | available at this time). It is a 64 bit filesystem, so it supports | ||
857 | extremely large volumes and files. | ||
858 | |||
859 | If you use this filesystem, you should also say Y to at least one | ||
860 | of the NLS (native language support) options below. | ||
861 | |||
862 | If you don't know what this is about, say N. | ||
863 | |||
864 | To compile this as a module, choose M here: the module will be | ||
865 | called befs. | ||
866 | |||
867 | config BEFS_DEBUG | ||
868 | bool "Debug BeFS" | ||
869 | depends on BEFS_FS | ||
870 | help | ||
871 | If you say Y here, you can use the 'debug' mount option to enable | ||
872 | debugging output from the driver. | ||
873 | |||
874 | config BFS_FS | ||
875 | tristate "BFS file system support (EXPERIMENTAL)" | ||
876 | depends on BLOCK && EXPERIMENTAL | ||
877 | help | ||
878 | Boot File System (BFS) is a file system used under SCO UnixWare to | ||
879 | allow the bootloader access to the kernel image and other important | ||
880 | files during the boot process. It is usually mounted under /stand | ||
881 | and corresponds to the slice marked as "STAND" in the UnixWare | ||
882 | partition. You should say Y if you want to read or write the files | ||
883 | on your /stand slice from within Linux. You then also need to say Y | ||
884 | to "UnixWare slices support", below. More information about the BFS | ||
885 | file system is contained in the file | ||
886 | <file:Documentation/filesystems/bfs.txt>. | ||
887 | |||
888 | If you don't know what this is about, say N. | ||
889 | |||
890 | To compile this as a module, choose M here: the module will be called | ||
891 | bfs. Note that the file system of your root partition (the one | ||
892 | containing the directory /) cannot be compiled as a module. | ||
893 | |||
894 | |||
895 | |||
896 | config EFS_FS | ||
897 | tristate "EFS file system support (read only) (EXPERIMENTAL)" | ||
898 | depends on BLOCK && EXPERIMENTAL | ||
899 | help | ||
900 | EFS is an older file system used for non-ISO9660 CD-ROMs and hard | ||
901 | disk partitions by SGI's IRIX operating system (IRIX 6.0 and newer | ||
902 | uses the XFS file system for hard disk partitions however). | ||
903 | |||
904 | This implementation only offers read-only access. If you don't know | ||
905 | what all this is about, it's safe to say N. For more information | ||
906 | about EFS see its home page at <http://aeschi.ch.eu.org/efs/>. | ||
907 | |||
908 | To compile the EFS file system support as a module, choose M here: the | ||
909 | module will be called efs. | ||
910 | |||
911 | source "fs/jffs2/Kconfig" | 212 | source "fs/jffs2/Kconfig" |
912 | # UBIFS File system configuration | 213 | # UBIFS File system configuration |
913 | source "fs/ubifs/Kconfig" | 214 | source "fs/ubifs/Kconfig" |
914 | 215 | source "fs/cramfs/Kconfig" | |
915 | config CRAMFS | 216 | source "fs/squashfs/Kconfig" |
916 | tristate "Compressed ROM file system support (cramfs)" | 217 | source "fs/freevxfs/Kconfig" |
917 | depends on BLOCK | 218 | source "fs/minix/Kconfig" |
918 | select ZLIB_INFLATE | 219 | source "fs/omfs/Kconfig" |
919 | help | 220 | source "fs/hpfs/Kconfig" |
920 | Saying Y here includes support for CramFs (Compressed ROM File | 221 | source "fs/qnx4/Kconfig" |
921 | System). CramFs is designed to be a simple, small, and compressed | 222 | source "fs/romfs/Kconfig" |
922 | file system for ROM based embedded systems. CramFs is read-only, | 223 | source "fs/sysv/Kconfig" |
923 | limited to 256MB file systems (with 16MB files), and doesn't support | 224 | source "fs/ufs/Kconfig" |
924 | 16/32 bits uid/gid, hard links and timestamps. | ||
925 | |||
926 | See <file:Documentation/filesystems/cramfs.txt> and | ||
927 | <file:fs/cramfs/README> for further information. | ||
928 | |||
929 | To compile this as a module, choose M here: the module will be called | ||
930 | cramfs. Note that the root file system (the one containing the | ||
931 | directory /) cannot be compiled as a module. | ||
932 | |||
933 | If unsure, say N. | ||
934 | |||
935 | config SQUASHFS | ||
936 | tristate "SquashFS 4.0 - Squashed file system support" | ||
937 | depends on BLOCK | ||
938 | select ZLIB_INFLATE | ||
939 | help | ||
940 | Saying Y here includes support for SquashFS 4.0 (a Compressed | ||
941 | Read-Only File System). Squashfs is a highly compressed read-only | ||
942 | filesystem for Linux. It uses zlib compression to compress both | ||
943 | files, inodes and directories. Inodes in the system are very small | ||
944 | and all blocks are packed to minimise data overhead. Block sizes | ||
945 | greater than 4K are supported up to a maximum of 1 Mbytes (default | ||
946 | block size 128K). SquashFS 4.0 supports 64 bit filesystems and files | ||
947 | (larger than 4GB), full uid/gid information, hard links and | ||
948 | timestamps. | ||
949 | |||
950 | Squashfs is intended for general read-only filesystem use, for | ||
951 | archival use (i.e. in cases where a .tar.gz file may be used), and in | ||
952 | embedded systems where low overhead is needed. Further information | ||
953 | and tools are available from http://squashfs.sourceforge.net. | ||
954 | |||
955 | If you want to compile this as a module ( = code which can be | ||
956 | inserted in and removed from the running kernel whenever you want), | ||
957 | say M here and read <file:Documentation/modules.txt>. The module | ||
958 | will be called squashfs. Note that the root file system (the one | ||
959 | containing the directory /) cannot be compiled as a module. | ||
960 | |||
961 | If unsure, say N. | ||
962 | |||
963 | config SQUASHFS_EMBEDDED | ||
964 | |||
965 | bool "Additional option for memory-constrained systems" | ||
966 | depends on SQUASHFS | ||
967 | default n | ||
968 | help | ||
969 | Saying Y here allows you to specify cache size. | ||
970 | |||
971 | If unsure, say N. | ||
972 | |||
973 | config SQUASHFS_FRAGMENT_CACHE_SIZE | ||
974 | int "Number of fragments cached" if SQUASHFS_EMBEDDED | ||
975 | depends on SQUASHFS | ||
976 | default "3" | ||
977 | help | ||
978 | By default SquashFS caches the last 3 fragments read from | ||
979 | the filesystem. Increasing this amount may mean SquashFS | ||
980 | has to re-read fragments less often from disk, at the expense | ||
981 | of extra system memory. Decreasing this amount will mean | ||
982 | SquashFS uses less memory at the expense of extra reads from disk. | ||
983 | |||
984 | Note there must be at least one cached fragment. Anything | ||
985 | much more than three will probably not make much difference. | ||
986 | |||
987 | config VXFS_FS | ||
988 | tristate "FreeVxFS file system support (VERITAS VxFS(TM) compatible)" | ||
989 | depends on BLOCK | ||
990 | help | ||
991 | FreeVxFS is a file system driver that support the VERITAS VxFS(TM) | ||
992 | file system format. VERITAS VxFS(TM) is the standard file system | ||
993 | of SCO UnixWare (and possibly others) and optionally available | ||
994 | for Sunsoft Solaris, HP-UX and many other operating systems. | ||
995 | Currently only readonly access is supported. | ||
996 | |||
997 | NOTE: the file system type as used by mount(1), mount(2) and | ||
998 | fstab(5) is 'vxfs' as it describes the file system format, not | ||
999 | the actual driver. | ||
1000 | |||
1001 | To compile this as a module, choose M here: the module will be | ||
1002 | called freevxfs. If unsure, say N. | ||
1003 | |||
1004 | config MINIX_FS | ||
1005 | tristate "Minix file system support" | ||
1006 | depends on BLOCK | ||
1007 | help | ||
1008 | Minix is a simple operating system used in many classes about OS's. | ||
1009 | The minix file system (method to organize files on a hard disk | ||
1010 | partition or a floppy disk) was the original file system for Linux, | ||
1011 | but has been superseded by the second extended file system ext2fs. | ||
1012 | You don't want to use the minix file system on your hard disk | ||
1013 | because of certain built-in restrictions, but it is sometimes found | ||
1014 | on older Linux floppy disks. This option will enlarge your kernel | ||
1015 | by about 28 KB. If unsure, say N. | ||
1016 | |||
1017 | To compile this file system support as a module, choose M here: the | ||
1018 | module will be called minix. Note that the file system of your root | ||
1019 | partition (the one containing the directory /) cannot be compiled as | ||
1020 | a module. | ||
1021 | |||
1022 | config OMFS_FS | ||
1023 | tristate "SonicBlue Optimized MPEG File System support" | ||
1024 | depends on BLOCK | ||
1025 | select CRC_ITU_T | ||
1026 | help | ||
1027 | This is the proprietary file system used by the Rio Karma music | ||
1028 | player and ReplayTV DVR. Despite the name, this filesystem is not | ||
1029 | more efficient than a standard FS for MPEG files, in fact likely | ||
1030 | the opposite is true. Say Y if you have either of these devices | ||
1031 | and wish to mount its disk. | ||
1032 | |||
1033 | To compile this file system support as a module, choose M here: the | ||
1034 | module will be called omfs. If unsure, say N. | ||
1035 | |||
1036 | config HPFS_FS | ||
1037 | tristate "OS/2 HPFS file system support" | ||
1038 | depends on BLOCK | ||
1039 | help | ||
1040 | OS/2 is IBM's operating system for PC's, the same as Warp, and HPFS | ||
1041 | is the file system used for organizing files on OS/2 hard disk | ||
1042 | partitions. Say Y if you want to be able to read files from and | ||
1043 | write files to an OS/2 HPFS partition on your hard drive. OS/2 | ||
1044 | floppies however are in regular MSDOS format, so you don't need this | ||
1045 | option in order to be able to read them. Read | ||
1046 | <file:Documentation/filesystems/hpfs.txt>. | ||
1047 | |||
1048 | To compile this file system support as a module, choose M here: the | ||
1049 | module will be called hpfs. If unsure, say N. | ||
1050 | |||
1051 | |||
1052 | config QNX4FS_FS | ||
1053 | tristate "QNX4 file system support (read only)" | ||
1054 | depends on BLOCK | ||
1055 | help | ||
1056 | This is the file system used by the real-time operating systems | ||
1057 | QNX 4 and QNX 6 (the latter is also called QNX RTP). | ||
1058 | Further information is available at <http://www.qnx.com/>. | ||
1059 | Say Y if you intend to mount QNX hard disks or floppies. | ||
1060 | Unless you say Y to "QNX4FS read-write support" below, you will | ||
1061 | only be able to read these file systems. | ||
1062 | |||
1063 | To compile this file system support as a module, choose M here: the | ||
1064 | module will be called qnx4. | ||
1065 | |||
1066 | If you don't know whether you need it, then you don't need it: | ||
1067 | answer N. | ||
1068 | |||
1069 | config QNX4FS_RW | ||
1070 | bool "QNX4FS write support (DANGEROUS)" | ||
1071 | depends on QNX4FS_FS && EXPERIMENTAL && BROKEN | ||
1072 | help | ||
1073 | Say Y if you want to test write support for QNX4 file systems. | ||
1074 | |||
1075 | It's currently broken, so for now: | ||
1076 | answer N. | ||
1077 | |||
1078 | config ROMFS_FS | ||
1079 | tristate "ROM file system support" | ||
1080 | depends on BLOCK | ||
1081 | ---help--- | ||
1082 | This is a very small read-only file system mainly intended for | ||
1083 | initial ram disks of installation disks, but it could be used for | ||
1084 | other read-only media as well. Read | ||
1085 | <file:Documentation/filesystems/romfs.txt> for details. | ||
1086 | |||
1087 | To compile this file system support as a module, choose M here: the | ||
1088 | module will be called romfs. Note that the file system of your | ||
1089 | root partition (the one containing the directory /) cannot be a | ||
1090 | module. | ||
1091 | |||
1092 | If you don't know whether you need it, then you don't need it: | ||
1093 | answer N. | ||
1094 | |||
1095 | |||
1096 | config SYSV_FS | ||
1097 | tristate "System V/Xenix/V7/Coherent file system support" | ||
1098 | depends on BLOCK | ||
1099 | help | ||
1100 | SCO, Xenix and Coherent are commercial Unix systems for Intel | ||
1101 | machines, and Version 7 was used on the DEC PDP-11. Saying Y | ||
1102 | here would allow you to read from their floppies and hard disk | ||
1103 | partitions. | ||
1104 | |||
1105 | If you have floppies or hard disk partitions like that, it is likely | ||
1106 | that they contain binaries from those other Unix systems; in order | ||
1107 | to run these binaries, you will want to install linux-abi which is | ||
1108 | a set of kernel modules that lets you run SCO, Xenix, Wyse, | ||
1109 | UnixWare, Dell Unix and System V programs under Linux. It is | ||
1110 | available via FTP (user: ftp) from | ||
1111 | <ftp://ftp.openlinux.org/pub/people/hch/linux-abi/>). | ||
1112 | NOTE: that will work only for binaries from Intel-based systems; | ||
1113 | PDP ones will have to wait until somebody ports Linux to -11 ;-) | ||
1114 | |||
1115 | If you only intend to mount files from some other Unix over the | ||
1116 | network using NFS, you don't need the System V file system support | ||
1117 | (but you need NFS file system support obviously). | ||
1118 | |||
1119 | Note that this option is generally not needed for floppies, since a | ||
1120 | good portable way to transport files and directories between unixes | ||
1121 | (and even other operating systems) is given by the tar program ("man | ||
1122 | tar" or preferably "info tar"). Note also that this option has | ||
1123 | nothing whatsoever to do with the option "System V IPC". Read about | ||
1124 | the System V file system in | ||
1125 | <file:Documentation/filesystems/sysv-fs.txt>. | ||
1126 | Saying Y here will enlarge your kernel by about 27 KB. | ||
1127 | |||
1128 | To compile this as a module, choose M here: the module will be called | ||
1129 | sysv. | ||
1130 | |||
1131 | If you haven't heard about all of this before, it's safe to say N. | ||
1132 | |||
1133 | |||
1134 | config UFS_FS | ||
1135 | tristate "UFS file system support (read only)" | ||
1136 | depends on BLOCK | ||
1137 | help | ||
1138 | BSD and derivate versions of Unix (such as SunOS, FreeBSD, NetBSD, | ||
1139 | OpenBSD and NeXTstep) use a file system called UFS. Some System V | ||
1140 | Unixes can create and mount hard disk partitions and diskettes using | ||
1141 | this file system as well. Saying Y here will allow you to read from | ||
1142 | these partitions; if you also want to write to them, say Y to the | ||
1143 | experimental "UFS file system write support", below. Please read the | ||
1144 | file <file:Documentation/filesystems/ufs.txt> for more information. | ||
1145 | |||
1146 | The recently released UFS2 variant (used in FreeBSD 5.x) is | ||
1147 | READ-ONLY supported. | ||
1148 | |||
1149 | Note that this option is generally not needed for floppies, since a | ||
1150 | good portable way to transport files and directories between unixes | ||
1151 | (and even other operating systems) is given by the tar program ("man | ||
1152 | tar" or preferably "info tar"). | ||
1153 | |||
1154 | When accessing NeXTstep files, you may need to convert them from the | ||
1155 | NeXT character set to the Latin1 character set; use the program | ||
1156 | recode ("info recode") for this purpose. | ||
1157 | |||
1158 | To compile the UFS file system support as a module, choose M here: the | ||
1159 | module will be called ufs. | ||
1160 | |||
1161 | If you haven't heard about all of this before, it's safe to say N. | ||
1162 | |||
1163 | config UFS_FS_WRITE | ||
1164 | bool "UFS file system write support (DANGEROUS)" | ||
1165 | depends on UFS_FS && EXPERIMENTAL | ||
1166 | help | ||
1167 | Say Y here if you want to try writing to UFS partitions. This is | ||
1168 | experimental, so you should back up your UFS partitions beforehand. | ||
1169 | |||
1170 | config UFS_DEBUG | ||
1171 | bool "UFS debugging" | ||
1172 | depends on UFS_FS | ||
1173 | help | ||
1174 | If you are experiencing any problems with the UFS filesystem, say | ||
1175 | Y here. This will result in _many_ additional debugging messages to be | ||
1176 | written to the system log. | ||
1177 | 225 | ||
1178 | endif # MISC_FILESYSTEMS | 226 | endif # MISC_FILESYSTEMS |
1179 | 227 | ||
@@ -1193,173 +241,8 @@ menuconfig NETWORK_FILESYSTEMS | |||
1193 | 241 | ||
1194 | if NETWORK_FILESYSTEMS | 242 | if NETWORK_FILESYSTEMS |
1195 | 243 | ||
1196 | config NFS_FS | 244 | source "fs/nfs/Kconfig" |
1197 | tristate "NFS client support" | 245 | source "fs/nfsd/Kconfig" |
1198 | depends on INET | ||
1199 | select LOCKD | ||
1200 | select SUNRPC | ||
1201 | select NFS_ACL_SUPPORT if NFS_V3_ACL | ||
1202 | help | ||
1203 | Choose Y here if you want to access files residing on other | ||
1204 | computers using Sun's Network File System protocol. To compile | ||
1205 | this file system support as a module, choose M here: the module | ||
1206 | will be called nfs. | ||
1207 | |||
1208 | To mount file systems exported by NFS servers, you also need to | ||
1209 | install the user space mount.nfs command which can be found in | ||
1210 | the Linux nfs-utils package, available from http://linux-nfs.org/. | ||
1211 | Information about using the mount command is available in the | ||
1212 | mount(8) man page. More detail about the Linux NFS client | ||
1213 | implementation is available via the nfs(5) man page. | ||
1214 | |||
1215 | Below you can choose which versions of the NFS protocol are | ||
1216 | available in the kernel to mount NFS servers. Support for NFS | ||
1217 | version 2 (RFC 1094) is always available when NFS_FS is selected. | ||
1218 | |||
1219 | To configure a system which mounts its root file system via NFS | ||
1220 | at boot time, say Y here, select "Kernel level IP | ||
1221 | autoconfiguration" in the NETWORK menu, and select "Root file | ||
1222 | system on NFS" below. You cannot compile this file system as a | ||
1223 | module in this case. | ||
1224 | |||
1225 | If unsure, say N. | ||
1226 | |||
1227 | config NFS_V3 | ||
1228 | bool "NFS client support for NFS version 3" | ||
1229 | depends on NFS_FS | ||
1230 | help | ||
1231 | This option enables support for version 3 of the NFS protocol | ||
1232 | (RFC 1813) in the kernel's NFS client. | ||
1233 | |||
1234 | If unsure, say Y. | ||
1235 | |||
1236 | config NFS_V3_ACL | ||
1237 | bool "NFS client support for the NFSv3 ACL protocol extension" | ||
1238 | depends on NFS_V3 | ||
1239 | help | ||
1240 | Some NFS servers support an auxiliary NFSv3 ACL protocol that | ||
1241 | Sun added to Solaris but never became an official part of the | ||
1242 | NFS version 3 protocol. This protocol extension allows | ||
1243 | applications on NFS clients to manipulate POSIX Access Control | ||
1244 | Lists on files residing on NFS servers. NFS servers enforce | ||
1245 | ACLs on local files whether this protocol is available or not. | ||
1246 | |||
1247 | Choose Y here if your NFS server supports the Solaris NFSv3 ACL | ||
1248 | protocol extension and you want your NFS client to allow | ||
1249 | applications to access and modify ACLs on files on the server. | ||
1250 | |||
1251 | Most NFS servers don't support the Solaris NFSv3 ACL protocol | ||
1252 | extension. You can choose N here or specify the "noacl" mount | ||
1253 | option to prevent your NFS client from trying to use the NFSv3 | ||
1254 | ACL protocol. | ||
1255 | |||
1256 | If unsure, say N. | ||
1257 | |||
1258 | config NFS_V4 | ||
1259 | bool "NFS client support for NFS version 4 (EXPERIMENTAL)" | ||
1260 | depends on NFS_FS && EXPERIMENTAL | ||
1261 | select RPCSEC_GSS_KRB5 | ||
1262 | help | ||
1263 | This option enables support for version 4 of the NFS protocol | ||
1264 | (RFC 3530) in the kernel's NFS client. | ||
1265 | |||
1266 | To mount NFS servers using NFSv4, you also need to install user | ||
1267 | space programs which can be found in the Linux nfs-utils package, | ||
1268 | available from http://linux-nfs.org/. | ||
1269 | |||
1270 | If unsure, say N. | ||
1271 | |||
1272 | config ROOT_NFS | ||
1273 | bool "Root file system on NFS" | ||
1274 | depends on NFS_FS=y && IP_PNP | ||
1275 | help | ||
1276 | If you want your system to mount its root file system via NFS, | ||
1277 | choose Y here. This is common practice for managing systems | ||
1278 | without local permanent storage. For details, read | ||
1279 | <file:Documentation/filesystems/nfsroot.txt>. | ||
1280 | |||
1281 | Most people say N here. | ||
1282 | |||
1283 | config NFSD | ||
1284 | tristate "NFS server support" | ||
1285 | depends on INET | ||
1286 | select LOCKD | ||
1287 | select SUNRPC | ||
1288 | select EXPORTFS | ||
1289 | select NFS_ACL_SUPPORT if NFSD_V2_ACL | ||
1290 | help | ||
1291 | Choose Y here if you want to allow other computers to access | ||
1292 | files residing on this system using Sun's Network File System | ||
1293 | protocol. To compile the NFS server support as a module, | ||
1294 | choose M here: the module will be called nfsd. | ||
1295 | |||
1296 | You may choose to use a user-space NFS server instead, in which | ||
1297 | case you can choose N here. | ||
1298 | |||
1299 | To export local file systems using NFS, you also need to install | ||
1300 | user space programs which can be found in the Linux nfs-utils | ||
1301 | package, available from http://linux-nfs.org/. More detail about | ||
1302 | the Linux NFS server implementation is available via the | ||
1303 | exports(5) man page. | ||
1304 | |||
1305 | Below you can choose which versions of the NFS protocol are | ||
1306 | available to clients mounting the NFS server on this system. | ||
1307 | Support for NFS version 2 (RFC 1094) is always available when | ||
1308 | CONFIG_NFSD is selected. | ||
1309 | |||
1310 | If unsure, say N. | ||
1311 | |||
1312 | config NFSD_V2_ACL | ||
1313 | bool | ||
1314 | depends on NFSD | ||
1315 | |||
1316 | config NFSD_V3 | ||
1317 | bool "NFS server support for NFS version 3" | ||
1318 | depends on NFSD | ||
1319 | help | ||
1320 | This option enables support in your system's NFS server for | ||
1321 | version 3 of the NFS protocol (RFC 1813). | ||
1322 | |||
1323 | If unsure, say Y. | ||
1324 | |||
1325 | config NFSD_V3_ACL | ||
1326 | bool "NFS server support for the NFSv3 ACL protocol extension" | ||
1327 | depends on NFSD_V3 | ||
1328 | select NFSD_V2_ACL | ||
1329 | help | ||
1330 | Solaris NFS servers support an auxiliary NFSv3 ACL protocol that | ||
1331 | never became an official part of the NFS version 3 protocol. | ||
1332 | This protocol extension allows applications on NFS clients to | ||
1333 | manipulate POSIX Access Control Lists on files residing on NFS | ||
1334 | servers. NFS servers enforce POSIX ACLs on local files whether | ||
1335 | this protocol is available or not. | ||
1336 | |||
1337 | This option enables support in your system's NFS server for the | ||
1338 | NFSv3 ACL protocol extension allowing NFS clients to manipulate | ||
1339 | POSIX ACLs on files exported by your system's NFS server. NFS | ||
1340 | clients which support the Solaris NFSv3 ACL protocol can then | ||
1341 | access and modify ACLs on your NFS server. | ||
1342 | |||
1343 | To store ACLs on your NFS server, you also need to enable ACL- | ||
1344 | related CONFIG options for your local file systems of choice. | ||
1345 | |||
1346 | If unsure, say N. | ||
1347 | |||
1348 | config NFSD_V4 | ||
1349 | bool "NFS server support for NFS version 4 (EXPERIMENTAL)" | ||
1350 | depends on NFSD && PROC_FS && EXPERIMENTAL | ||
1351 | select NFSD_V3 | ||
1352 | select FS_POSIX_ACL | ||
1353 | select RPCSEC_GSS_KRB5 | ||
1354 | help | ||
1355 | This option enables support in your system's NFS server for | ||
1356 | version 4 of the NFS protocol (RFC 3530). | ||
1357 | |||
1358 | To export files using NFSv4, you need to install additional user | ||
1359 | space programs which can be found in the Linux nfs-utils package, | ||
1360 | available from http://linux-nfs.org/. | ||
1361 | |||
1362 | If unsure, say N. | ||
1363 | 246 | ||
1364 | config LOCKD | 247 | config LOCKD |
1365 | tristate | 248 | tristate |
@@ -1381,221 +264,13 @@ config NFS_COMMON | |||
1381 | depends on NFSD || NFS_FS | 264 | depends on NFSD || NFS_FS |
1382 | default y | 265 | default y |
1383 | 266 | ||
1384 | config SUNRPC | 267 | source "net/sunrpc/Kconfig" |
1385 | tristate | 268 | source "fs/smbfs/Kconfig" |
1386 | |||
1387 | config SUNRPC_GSS | ||
1388 | tristate | ||
1389 | |||
1390 | config SUNRPC_XPRT_RDMA | ||
1391 | tristate | ||
1392 | depends on SUNRPC && INFINIBAND && EXPERIMENTAL | ||
1393 | default SUNRPC && INFINIBAND | ||
1394 | help | ||
1395 | This option enables an RPC client transport capability that | ||
1396 | allows the NFS client to mount servers via an RDMA-enabled | ||
1397 | transport. | ||
1398 | |||
1399 | To compile RPC client RDMA transport support as a module, | ||
1400 | choose M here: the module will be called xprtrdma. | ||
1401 | |||
1402 | If unsure, say N. | ||
1403 | |||
1404 | config SUNRPC_REGISTER_V4 | ||
1405 | bool "Register local RPC services via rpcbind v4 (EXPERIMENTAL)" | ||
1406 | depends on SUNRPC && EXPERIMENTAL | ||
1407 | default n | ||
1408 | help | ||
1409 | Sun added support for registering RPC services at an IPv6 | ||
1410 | address by creating two new versions of the rpcbind protocol | ||
1411 | (RFC 1833). | ||
1412 | |||
1413 | This option enables support in the kernel RPC server for | ||
1414 | registering kernel RPC services via version 4 of the rpcbind | ||
1415 | protocol. If you enable this option, you must run a portmapper | ||
1416 | daemon that supports rpcbind protocol version 4. | ||
1417 | |||
1418 | Serving NFS over IPv6 from knfsd (the kernel's NFS server) | ||
1419 | requires that you enable this option and use a portmapper that | ||
1420 | supports rpcbind version 4. | ||
1421 | |||
1422 | If unsure, say N to get traditional behavior (register kernel | ||
1423 | RPC services using only rpcbind version 2). Distributions | ||
1424 | using the legacy Linux portmapper daemon must say N here. | ||
1425 | |||
1426 | config RPCSEC_GSS_KRB5 | ||
1427 | tristate "Secure RPC: Kerberos V mechanism (EXPERIMENTAL)" | ||
1428 | depends on SUNRPC && EXPERIMENTAL | ||
1429 | select SUNRPC_GSS | ||
1430 | select CRYPTO | ||
1431 | select CRYPTO_MD5 | ||
1432 | select CRYPTO_DES | ||
1433 | select CRYPTO_CBC | ||
1434 | help | ||
1435 | Choose Y here to enable Secure RPC using the Kerberos version 5 | ||
1436 | GSS-API mechanism (RFC 1964). | ||
1437 | |||
1438 | Secure RPC calls with Kerberos require an auxiliary user-space | ||
1439 | daemon which may be found in the Linux nfs-utils package | ||
1440 | available from http://linux-nfs.org/. In addition, user-space | ||
1441 | Kerberos support should be installed. | ||
1442 | |||
1443 | If unsure, say N. | ||
1444 | |||
1445 | config RPCSEC_GSS_SPKM3 | ||
1446 | tristate "Secure RPC: SPKM3 mechanism (EXPERIMENTAL)" | ||
1447 | depends on SUNRPC && EXPERIMENTAL | ||
1448 | select SUNRPC_GSS | ||
1449 | select CRYPTO | ||
1450 | select CRYPTO_MD5 | ||
1451 | select CRYPTO_DES | ||
1452 | select CRYPTO_CAST5 | ||
1453 | select CRYPTO_CBC | ||
1454 | help | ||
1455 | Choose Y here to enable Secure RPC using the SPKM3 public key | ||
1456 | GSS-API mechansim (RFC 2025). | ||
1457 | |||
1458 | Secure RPC calls with SPKM3 require an auxiliary userspace | ||
1459 | daemon which may be found in the Linux nfs-utils package | ||
1460 | available from http://linux-nfs.org/. | ||
1461 | |||
1462 | If unsure, say N. | ||
1463 | |||
1464 | config SMB_FS | ||
1465 | tristate "SMB file system support (OBSOLETE, please use CIFS)" | ||
1466 | depends on INET | ||
1467 | select NLS | ||
1468 | help | ||
1469 | SMB (Server Message Block) is the protocol Windows for Workgroups | ||
1470 | (WfW), Windows 95/98, Windows NT and OS/2 Lan Manager use to share | ||
1471 | files and printers over local networks. Saying Y here allows you to | ||
1472 | mount their file systems (often called "shares" in this context) and | ||
1473 | access them just like any other Unix directory. Currently, this | ||
1474 | works only if the Windows machines use TCP/IP as the underlying | ||
1475 | transport protocol, and not NetBEUI. For details, read | ||
1476 | <file:Documentation/filesystems/smbfs.txt> and the SMB-HOWTO, | ||
1477 | available from <http://www.tldp.org/docs.html#howto>. | ||
1478 | |||
1479 | Note: if you just want your box to act as an SMB *server* and make | ||
1480 | files and printing services available to Windows clients (which need | ||
1481 | to have a TCP/IP stack), you don't need to say Y here; you can use | ||
1482 | the program SAMBA (available from <ftp://ftp.samba.org/pub/samba/>) | ||
1483 | for that. | ||
1484 | |||
1485 | General information about how to connect Linux, Windows machines and | ||
1486 | Macs is on the WWW at <http://www.eats.com/linux_mac_win.html>. | ||
1487 | |||
1488 | To compile the SMB support as a module, choose M here: | ||
1489 | the module will be called smbfs. Most people say N, however. | ||
1490 | |||
1491 | config SMB_NLS_DEFAULT | ||
1492 | bool "Use a default NLS" | ||
1493 | depends on SMB_FS | ||
1494 | help | ||
1495 | Enabling this will make smbfs use nls translations by default. You | ||
1496 | need to specify the local charset (CONFIG_NLS_DEFAULT) in the nls | ||
1497 | settings and you need to give the default nls for the SMB server as | ||
1498 | CONFIG_SMB_NLS_REMOTE. | ||
1499 | |||
1500 | The nls settings can be changed at mount time, if your smbmount | ||
1501 | supports that, using the codepage and iocharset parameters. | ||
1502 | |||
1503 | smbmount from samba 2.2.0 or later supports this. | ||
1504 | |||
1505 | config SMB_NLS_REMOTE | ||
1506 | string "Default Remote NLS Option" | ||
1507 | depends on SMB_NLS_DEFAULT | ||
1508 | default "cp437" | ||
1509 | help | ||
1510 | This setting allows you to specify a default value for which | ||
1511 | codepage the server uses. If this field is left blank no | ||
1512 | translations will be done by default. The local codepage/charset | ||
1513 | default to CONFIG_NLS_DEFAULT. | ||
1514 | |||
1515 | The nls settings can be changed at mount time, if your smbmount | ||
1516 | supports that, using the codepage and iocharset parameters. | ||
1517 | |||
1518 | smbmount from samba 2.2.0 or later supports this. | ||
1519 | |||
1520 | source "fs/cifs/Kconfig" | 269 | source "fs/cifs/Kconfig" |
1521 | |||
1522 | config NCP_FS | ||
1523 | tristate "NCP file system support (to mount NetWare volumes)" | ||
1524 | depends on IPX!=n || INET | ||
1525 | help | ||
1526 | NCP (NetWare Core Protocol) is a protocol that runs over IPX and is | ||
1527 | used by Novell NetWare clients to talk to file servers. It is to | ||
1528 | IPX what NFS is to TCP/IP, if that helps. Saying Y here allows you | ||
1529 | to mount NetWare file server volumes and to access them just like | ||
1530 | any other Unix directory. For details, please read the file | ||
1531 | <file:Documentation/filesystems/ncpfs.txt> in the kernel source and | ||
1532 | the IPX-HOWTO from <http://www.tldp.org/docs.html#howto>. | ||
1533 | |||
1534 | You do not have to say Y here if you want your Linux box to act as a | ||
1535 | file *server* for Novell NetWare clients. | ||
1536 | |||
1537 | General information about how to connect Linux, Windows machines and | ||
1538 | Macs is on the WWW at <http://www.eats.com/linux_mac_win.html>. | ||
1539 | |||
1540 | To compile this as a module, choose M here: the module will be called | ||
1541 | ncpfs. Say N unless you are connected to a Novell network. | ||
1542 | |||
1543 | source "fs/ncpfs/Kconfig" | 270 | source "fs/ncpfs/Kconfig" |
1544 | 271 | source "fs/coda/Kconfig" | |
1545 | config CODA_FS | 272 | source "fs/afs/Kconfig" |
1546 | tristate "Coda file system support (advanced network fs)" | 273 | source "fs/9p/Kconfig" |
1547 | depends on INET | ||
1548 | help | ||
1549 | Coda is an advanced network file system, similar to NFS in that it | ||
1550 | enables you to mount file systems of a remote server and access them | ||
1551 | with regular Unix commands as if they were sitting on your hard | ||
1552 | disk. Coda has several advantages over NFS: support for | ||
1553 | disconnected operation (e.g. for laptops), read/write server | ||
1554 | replication, security model for authentication and encryption, | ||
1555 | persistent client caches and write back caching. | ||
1556 | |||
1557 | If you say Y here, your Linux box will be able to act as a Coda | ||
1558 | *client*. You will need user level code as well, both for the | ||
1559 | client and server. Servers are currently user level, i.e. they need | ||
1560 | no kernel support. Please read | ||
1561 | <file:Documentation/filesystems/coda.txt> and check out the Coda | ||
1562 | home page <http://www.coda.cs.cmu.edu/>. | ||
1563 | |||
1564 | To compile the coda client support as a module, choose M here: the | ||
1565 | module will be called coda. | ||
1566 | |||
1567 | config AFS_FS | ||
1568 | tristate "Andrew File System support (AFS) (EXPERIMENTAL)" | ||
1569 | depends on INET && EXPERIMENTAL | ||
1570 | select AF_RXRPC | ||
1571 | help | ||
1572 | If you say Y here, you will get an experimental Andrew File System | ||
1573 | driver. It currently only supports unsecured read-only AFS access. | ||
1574 | |||
1575 | See <file:Documentation/filesystems/afs.txt> for more information. | ||
1576 | |||
1577 | If unsure, say N. | ||
1578 | |||
1579 | config AFS_DEBUG | ||
1580 | bool "AFS dynamic debugging" | ||
1581 | depends on AFS_FS | ||
1582 | help | ||
1583 | Say Y here to make runtime controllable debugging messages appear. | ||
1584 | |||
1585 | See <file:Documentation/filesystems/afs.txt> for more information. | ||
1586 | |||
1587 | If unsure, say N. | ||
1588 | |||
1589 | config 9P_FS | ||
1590 | tristate "Plan 9 Resource Sharing Support (9P2000) (Experimental)" | ||
1591 | depends on INET && NET_9P && EXPERIMENTAL | ||
1592 | help | ||
1593 | If you say Y here, you will get experimental support for | ||
1594 | Plan 9 resource sharing via the 9P2000 protocol. | ||
1595 | |||
1596 | See <http://v9fs.sf.net> for more information. | ||
1597 | |||
1598 | If unsure, say N. | ||
1599 | 274 | ||
1600 | endif # NETWORK_FILESYSTEMS | 275 | endif # NETWORK_FILESYSTEMS |
1601 | 276 | ||
diff --git a/fs/adfs/Kconfig b/fs/adfs/Kconfig new file mode 100644 index 000000000000..e55182a74605 --- /dev/null +++ b/fs/adfs/Kconfig | |||
@@ -0,0 +1,27 @@ | |||
1 | config ADFS_FS | ||
2 | tristate "ADFS file system support (EXPERIMENTAL)" | ||
3 | depends on BLOCK && EXPERIMENTAL | ||
4 | help | ||
5 | The Acorn Disc Filing System is the standard file system of the | ||
6 | RiscOS operating system which runs on Acorn's ARM-based Risc PC | ||
7 | systems and the Acorn Archimedes range of machines. If you say Y | ||
8 | here, Linux will be able to read from ADFS partitions on hard drives | ||
9 | and from ADFS-formatted floppy discs. If you also want to be able to | ||
10 | write to those devices, say Y to "ADFS write support" below. | ||
11 | |||
12 | The ADFS partition should be the first partition (i.e., | ||
13 | /dev/[hs]d?1) on each of your drives. Please read the file | ||
14 | <file:Documentation/filesystems/adfs.txt> for further details. | ||
15 | |||
16 | To compile this code as a module, choose M here: the module will be | ||
17 | called adfs. | ||
18 | |||
19 | If unsure, say N. | ||
20 | |||
21 | config ADFS_FS_RW | ||
22 | bool "ADFS write support (DANGEROUS)" | ||
23 | depends on ADFS_FS | ||
24 | help | ||
25 | If you say Y here, you will be able to write to ADFS partitions on | ||
26 | hard drives and ADFS-formatted floppy disks. This is experimental | ||
27 | codes, so if you're unsure, say N. | ||
diff --git a/fs/affs/Kconfig b/fs/affs/Kconfig new file mode 100644 index 000000000000..cfad9afb4762 --- /dev/null +++ b/fs/affs/Kconfig | |||
@@ -0,0 +1,21 @@ | |||
1 | config AFFS_FS | ||
2 | tristate "Amiga FFS file system support (EXPERIMENTAL)" | ||
3 | depends on BLOCK && EXPERIMENTAL | ||
4 | help | ||
5 | The Fast File System (FFS) is the common file system used on hard | ||
6 | disks by Amiga(tm) systems since AmigaOS Version 1.3 (34.20). Say Y | ||
7 | if you want to be able to read and write files from and to an Amiga | ||
8 | FFS partition on your hard drive. Amiga floppies however cannot be | ||
9 | read with this driver due to an incompatibility of the floppy | ||
10 | controller used in an Amiga and the standard floppy controller in | ||
11 | PCs and workstations. Read <file:Documentation/filesystems/affs.txt> | ||
12 | and <file:fs/affs/Changes>. | ||
13 | |||
14 | With this driver you can also mount disk files used by Bernd | ||
15 | Schmidt's Un*X Amiga Emulator | ||
16 | (<http://www.freiburg.linux.de/~uae/>). | ||
17 | If you want to do this, you will also need to say Y or M to "Loop | ||
18 | device support", above. | ||
19 | |||
20 | To compile this file system support as a module, choose M here: the | ||
21 | module will be called affs. If unsure, say N. | ||
diff --git a/fs/afs/Kconfig b/fs/afs/Kconfig new file mode 100644 index 000000000000..e7b522fe15e1 --- /dev/null +++ b/fs/afs/Kconfig | |||
@@ -0,0 +1,21 @@ | |||
1 | config AFS_FS | ||
2 | tristate "Andrew File System support (AFS) (EXPERIMENTAL)" | ||
3 | depends on INET && EXPERIMENTAL | ||
4 | select AF_RXRPC | ||
5 | help | ||
6 | If you say Y here, you will get an experimental Andrew File System | ||
7 | driver. It currently only supports unsecured read-only AFS access. | ||
8 | |||
9 | See <file:Documentation/filesystems/afs.txt> for more information. | ||
10 | |||
11 | If unsure, say N. | ||
12 | |||
13 | config AFS_DEBUG | ||
14 | bool "AFS dynamic debugging" | ||
15 | depends on AFS_FS | ||
16 | help | ||
17 | Say Y here to make runtime controllable debugging messages appear. | ||
18 | |||
19 | See <file:Documentation/filesystems/afs.txt> for more information. | ||
20 | |||
21 | If unsure, say N. | ||
diff --git a/fs/autofs/Kconfig b/fs/autofs/Kconfig new file mode 100644 index 000000000000..5f3bea90911e --- /dev/null +++ b/fs/autofs/Kconfig | |||
@@ -0,0 +1,21 @@ | |||
1 | config AUTOFS_FS | ||
2 | tristate "Kernel automounter support" | ||
3 | help | ||
4 | The automounter is a tool to automatically mount remote file systems | ||
5 | on demand. This implementation is partially kernel-based to reduce | ||
6 | overhead in the already-mounted case; this is unlike the BSD | ||
7 | automounter (amd), which is a pure user space daemon. | ||
8 | |||
9 | To use the automounter you need the user-space tools from the autofs | ||
10 | package; you can find the location in <file:Documentation/Changes>. | ||
11 | You also want to answer Y to "NFS file system support", below. | ||
12 | |||
13 | If you want to use the newer version of the automounter with more | ||
14 | features, say N here and say Y to "Kernel automounter v4 support", | ||
15 | below. | ||
16 | |||
17 | To compile this support as a module, choose M here: the module will be | ||
18 | called autofs. | ||
19 | |||
20 | If you are not a part of a fairly large, distributed network, you | ||
21 | probably do not need an automounter, and can say N here. | ||
diff --git a/fs/autofs4/Kconfig b/fs/autofs4/Kconfig new file mode 100644 index 000000000000..1204d6384d39 --- /dev/null +++ b/fs/autofs4/Kconfig | |||
@@ -0,0 +1,20 @@ | |||
1 | config AUTOFS4_FS | ||
2 | tristate "Kernel automounter version 4 support (also supports v3)" | ||
3 | help | ||
4 | The automounter is a tool to automatically mount remote file systems | ||
5 | on demand. This implementation is partially kernel-based to reduce | ||
6 | overhead in the already-mounted case; this is unlike the BSD | ||
7 | automounter (amd), which is a pure user space daemon. | ||
8 | |||
9 | To use the automounter you need the user-space tools from | ||
10 | <ftp://ftp.kernel.org/pub/linux/daemons/autofs/v4/>; you also | ||
11 | want to answer Y to "NFS file system support", below. | ||
12 | |||
13 | To compile this support as a module, choose M here: the module will be | ||
14 | called autofs4. You will need to add "alias autofs autofs4" to your | ||
15 | modules configuration file. | ||
16 | |||
17 | If you are not a part of a fairly large, distributed network or | ||
18 | don't have a laptop which needs to dynamically reconfigure to the | ||
19 | local network, you probably do not need an automounter, and can say | ||
20 | N here. | ||
diff --git a/fs/befs/Kconfig b/fs/befs/Kconfig new file mode 100644 index 000000000000..7835d30f211f --- /dev/null +++ b/fs/befs/Kconfig | |||
@@ -0,0 +1,26 @@ | |||
1 | config BEFS_FS | ||
2 | tristate "BeOS file system (BeFS) support (read only) (EXPERIMENTAL)" | ||
3 | depends on BLOCK && EXPERIMENTAL | ||
4 | select NLS | ||
5 | help | ||
6 | The BeOS File System (BeFS) is the native file system of Be, Inc's | ||
7 | BeOS. Notable features include support for arbitrary attributes | ||
8 | on files and directories, and database-like indices on selected | ||
9 | attributes. (Also note that this driver doesn't make those features | ||
10 | available at this time). It is a 64 bit filesystem, so it supports | ||
11 | extremely large volumes and files. | ||
12 | |||
13 | If you use this filesystem, you should also say Y to at least one | ||
14 | of the NLS (native language support) options below. | ||
15 | |||
16 | If you don't know what this is about, say N. | ||
17 | |||
18 | To compile this as a module, choose M here: the module will be | ||
19 | called befs. | ||
20 | |||
21 | config BEFS_DEBUG | ||
22 | bool "Debug BeFS" | ||
23 | depends on BEFS_FS | ||
24 | help | ||
25 | If you say Y here, you can use the 'debug' mount option to enable | ||
26 | debugging output from the driver. | ||
diff --git a/fs/bfs/Kconfig b/fs/bfs/Kconfig new file mode 100644 index 000000000000..c2336c62024f --- /dev/null +++ b/fs/bfs/Kconfig | |||
@@ -0,0 +1,19 @@ | |||
1 | config BFS_FS | ||
2 | tristate "BFS file system support (EXPERIMENTAL)" | ||
3 | depends on BLOCK && EXPERIMENTAL | ||
4 | help | ||
5 | Boot File System (BFS) is a file system used under SCO UnixWare to | ||
6 | allow the bootloader access to the kernel image and other important | ||
7 | files during the boot process. It is usually mounted under /stand | ||
8 | and corresponds to the slice marked as "STAND" in the UnixWare | ||
9 | partition. You should say Y if you want to read or write the files | ||
10 | on your /stand slice from within Linux. You then also need to say Y | ||
11 | to "UnixWare slices support", below. More information about the BFS | ||
12 | file system is contained in the file | ||
13 | <file:Documentation/filesystems/bfs.txt>. | ||
14 | |||
15 | If you don't know what this is about, say N. | ||
16 | |||
17 | To compile this as a module, choose M here: the module will be called | ||
18 | bfs. Note that the file system of your root partition (the one | ||
19 | containing the directory /) cannot be compiled as a module. | ||
diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index 77ebc3c263d6..549b0144da11 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c | |||
@@ -140,7 +140,6 @@ int bio_integrity_add_page(struct bio *bio, struct page *page, | |||
140 | 140 | ||
141 | iv = bip_vec_idx(bip, bip->bip_vcnt); | 141 | iv = bip_vec_idx(bip, bip->bip_vcnt); |
142 | BUG_ON(iv == NULL); | 142 | BUG_ON(iv == NULL); |
143 | BUG_ON(iv->bv_page != NULL); | ||
144 | 143 | ||
145 | iv->bv_page = page; | 144 | iv->bv_page = page; |
146 | iv->bv_len = len; | 145 | iv->bv_len = len; |
@@ -465,7 +464,7 @@ static int bio_integrity_verify(struct bio *bio) | |||
465 | 464 | ||
466 | if (ret) { | 465 | if (ret) { |
467 | kunmap_atomic(kaddr, KM_USER0); | 466 | kunmap_atomic(kaddr, KM_USER0); |
468 | break; | 467 | return ret; |
469 | } | 468 | } |
470 | 469 | ||
471 | sectors = bv->bv_len / bi->sector_size; | 470 | sectors = bv->bv_len / bi->sector_size; |
@@ -493,18 +492,13 @@ static void bio_integrity_verify_fn(struct work_struct *work) | |||
493 | struct bio_integrity_payload *bip = | 492 | struct bio_integrity_payload *bip = |
494 | container_of(work, struct bio_integrity_payload, bip_work); | 493 | container_of(work, struct bio_integrity_payload, bip_work); |
495 | struct bio *bio = bip->bip_bio; | 494 | struct bio *bio = bip->bip_bio; |
496 | int error = bip->bip_error; | 495 | int error; |
497 | 496 | ||
498 | if (bio_integrity_verify(bio)) { | 497 | error = bio_integrity_verify(bio); |
499 | clear_bit(BIO_UPTODATE, &bio->bi_flags); | ||
500 | error = -EIO; | ||
501 | } | ||
502 | 498 | ||
503 | /* Restore original bio completion handler */ | 499 | /* Restore original bio completion handler */ |
504 | bio->bi_end_io = bip->bip_end_io; | 500 | bio->bi_end_io = bip->bip_end_io; |
505 | 501 | bio_endio(bio, error); | |
506 | if (bio->bi_end_io) | ||
507 | bio->bi_end_io(bio, error); | ||
508 | } | 502 | } |
509 | 503 | ||
510 | /** | 504 | /** |
@@ -525,7 +519,17 @@ void bio_integrity_endio(struct bio *bio, int error) | |||
525 | 519 | ||
526 | BUG_ON(bip->bip_bio != bio); | 520 | BUG_ON(bip->bip_bio != bio); |
527 | 521 | ||
528 | bip->bip_error = error; | 522 | /* In case of an I/O error there is no point in verifying the |
523 | * integrity metadata. Restore original bio end_io handler | ||
524 | * and run it. | ||
525 | */ | ||
526 | if (error) { | ||
527 | bio->bi_end_io = bip->bip_end_io; | ||
528 | bio_endio(bio, error); | ||
529 | |||
530 | return; | ||
531 | } | ||
532 | |||
529 | INIT_WORK(&bip->bip_work, bio_integrity_verify_fn); | 533 | INIT_WORK(&bip->bip_work, bio_integrity_verify_fn); |
530 | queue_work(kintegrityd_wq, &bip->bip_work); | 534 | queue_work(kintegrityd_wq, &bip->bip_work); |
531 | } | 535 | } |
diff --git a/fs/btrfs/Kconfig b/fs/btrfs/Kconfig new file mode 100644 index 000000000000..f8fcf999ea1b --- /dev/null +++ b/fs/btrfs/Kconfig | |||
@@ -0,0 +1,18 @@ | |||
1 | config BTRFS_FS | ||
2 | tristate "Btrfs filesystem (EXPERIMENTAL) Unstable disk format" | ||
3 | depends on EXPERIMENTAL | ||
4 | select LIBCRC32C | ||
5 | select ZLIB_INFLATE | ||
6 | select ZLIB_DEFLATE | ||
7 | help | ||
8 | Btrfs is a new filesystem with extents, writable snapshotting, | ||
9 | support for multiple devices and many more features. | ||
10 | |||
11 | Btrfs is highly experimental, and THE DISK FORMAT IS NOT YET | ||
12 | FINALIZED. You should say N here unless you are interested in | ||
13 | testing Btrfs with non-critical data. | ||
14 | |||
15 | To compile this file system support as a module, choose M here. The | ||
16 | module will be called btrfs. | ||
17 | |||
18 | If unsure, say N. | ||
diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 080703a15f44..73ac7ebd1dfc 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES | |||
@@ -5,7 +5,9 @@ rather than posix (advisory) byte range locks, even though server would | |||
5 | support posix byte range locks. Fix query of root inode when prefixpath | 5 | support posix byte range locks. Fix query of root inode when prefixpath |
6 | specified and user does not have access to query information about the | 6 | specified and user does not have access to query information about the |
7 | top of the share. Fix problem in 2.6.28 resolving DFS paths to | 7 | top of the share. Fix problem in 2.6.28 resolving DFS paths to |
8 | Samba servers (worked to Windows). | 8 | Samba servers (worked to Windows). Fix rmdir so that pending search |
9 | (readdir) requests do not get invalid results which include the now | ||
10 | removed directory. | ||
9 | 11 | ||
10 | Version 1.55 | 12 | Version 1.55 |
11 | ------------ | 13 | ------------ |
diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index d4839cf0cb2c..7c9809523f42 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c | |||
@@ -48,11 +48,11 @@ static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, | |||
48 | if ((cifs_pdu == NULL) || (signature == NULL) || (key == NULL)) | 48 | if ((cifs_pdu == NULL) || (signature == NULL) || (key == NULL)) |
49 | return -EINVAL; | 49 | return -EINVAL; |
50 | 50 | ||
51 | MD5Init(&context); | 51 | cifs_MD5_init(&context); |
52 | MD5Update(&context, (char *)&key->data, key->len); | 52 | cifs_MD5_update(&context, (char *)&key->data, key->len); |
53 | MD5Update(&context, cifs_pdu->Protocol, cifs_pdu->smb_buf_length); | 53 | cifs_MD5_update(&context, cifs_pdu->Protocol, cifs_pdu->smb_buf_length); |
54 | 54 | ||
55 | MD5Final(signature, &context); | 55 | cifs_MD5_final(signature, &context); |
56 | return 0; | 56 | return 0; |
57 | } | 57 | } |
58 | 58 | ||
@@ -96,8 +96,8 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, | |||
96 | if ((iov == NULL) || (signature == NULL) || (key == NULL)) | 96 | if ((iov == NULL) || (signature == NULL) || (key == NULL)) |
97 | return -EINVAL; | 97 | return -EINVAL; |
98 | 98 | ||
99 | MD5Init(&context); | 99 | cifs_MD5_init(&context); |
100 | MD5Update(&context, (char *)&key->data, key->len); | 100 | cifs_MD5_update(&context, (char *)&key->data, key->len); |
101 | for (i = 0; i < n_vec; i++) { | 101 | for (i = 0; i < n_vec; i++) { |
102 | if (iov[i].iov_len == 0) | 102 | if (iov[i].iov_len == 0) |
103 | continue; | 103 | continue; |
@@ -110,13 +110,13 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, | |||
110 | if (i == 0) { | 110 | if (i == 0) { |
111 | if (iov[0].iov_len <= 8) /* cmd field at offset 9 */ | 111 | if (iov[0].iov_len <= 8) /* cmd field at offset 9 */ |
112 | break; /* nothing to sign or corrupt header */ | 112 | break; /* nothing to sign or corrupt header */ |
113 | MD5Update(&context, iov[0].iov_base+4, | 113 | cifs_MD5_update(&context, iov[0].iov_base+4, |
114 | iov[0].iov_len-4); | 114 | iov[0].iov_len-4); |
115 | } else | 115 | } else |
116 | MD5Update(&context, iov[i].iov_base, iov[i].iov_len); | 116 | cifs_MD5_update(&context, iov[i].iov_base, iov[i].iov_len); |
117 | } | 117 | } |
118 | 118 | ||
119 | MD5Final(signature, &context); | 119 | cifs_MD5_final(signature, &context); |
120 | 120 | ||
121 | return 0; | 121 | return 0; |
122 | } | 122 | } |
diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 06f6779988bf..382ba6298809 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h | |||
@@ -35,8 +35,8 @@ extern struct smb_hdr *cifs_buf_get(void); | |||
35 | extern void cifs_buf_release(void *); | 35 | extern void cifs_buf_release(void *); |
36 | extern struct smb_hdr *cifs_small_buf_get(void); | 36 | extern struct smb_hdr *cifs_small_buf_get(void); |
37 | extern void cifs_small_buf_release(void *); | 37 | extern void cifs_small_buf_release(void *); |
38 | extern int smb_send(struct socket *, struct smb_hdr *, | 38 | extern int smb_send(struct TCP_Server_Info *, struct smb_hdr *, |
39 | unsigned int /* length */ , struct sockaddr *, bool); | 39 | unsigned int /* length */); |
40 | extern unsigned int _GetXid(void); | 40 | extern unsigned int _GetXid(void); |
41 | extern void _FreeXid(unsigned int); | 41 | extern void _FreeXid(unsigned int); |
42 | #define GetXid() (int)_GetXid(); cFYI(1,("CIFS VFS: in %s as Xid: %d with uid: %d",__func__, xid,current_fsuid())); | 42 | #define GetXid() (int)_GetXid(); cFYI(1,("CIFS VFS: in %s as Xid: %d with uid: %d",__func__, xid,current_fsuid())); |
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index e9ea394ee075..2209be943051 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c | |||
@@ -1354,7 +1354,7 @@ cifs_parse_mount_options(char *options, const char *devname, | |||
1354 | } | 1354 | } |
1355 | 1355 | ||
1356 | static struct TCP_Server_Info * | 1356 | static struct TCP_Server_Info * |
1357 | cifs_find_tcp_session(struct sockaddr *addr) | 1357 | cifs_find_tcp_session(struct sockaddr_storage *addr) |
1358 | { | 1358 | { |
1359 | struct list_head *tmp; | 1359 | struct list_head *tmp; |
1360 | struct TCP_Server_Info *server; | 1360 | struct TCP_Server_Info *server; |
@@ -1374,11 +1374,11 @@ cifs_find_tcp_session(struct sockaddr *addr) | |||
1374 | if (server->tcpStatus == CifsNew) | 1374 | if (server->tcpStatus == CifsNew) |
1375 | continue; | 1375 | continue; |
1376 | 1376 | ||
1377 | if (addr->sa_family == AF_INET && | 1377 | if (addr->ss_family == AF_INET && |
1378 | (addr4->sin_addr.s_addr != | 1378 | (addr4->sin_addr.s_addr != |
1379 | server->addr.sockAddr.sin_addr.s_addr)) | 1379 | server->addr.sockAddr.sin_addr.s_addr)) |
1380 | continue; | 1380 | continue; |
1381 | else if (addr->sa_family == AF_INET6 && | 1381 | else if (addr->ss_family == AF_INET6 && |
1382 | memcmp(&server->addr.sockAddr6.sin6_addr, | 1382 | memcmp(&server->addr.sockAddr6.sin6_addr, |
1383 | &addr6->sin6_addr, sizeof(addr6->sin6_addr))) | 1383 | &addr6->sin6_addr, sizeof(addr6->sin6_addr))) |
1384 | continue; | 1384 | continue; |
@@ -1419,12 +1419,12 @@ static struct TCP_Server_Info * | |||
1419 | cifs_get_tcp_session(struct smb_vol *volume_info) | 1419 | cifs_get_tcp_session(struct smb_vol *volume_info) |
1420 | { | 1420 | { |
1421 | struct TCP_Server_Info *tcp_ses = NULL; | 1421 | struct TCP_Server_Info *tcp_ses = NULL; |
1422 | struct sockaddr addr; | 1422 | struct sockaddr_storage addr; |
1423 | struct sockaddr_in *sin_server = (struct sockaddr_in *) &addr; | 1423 | struct sockaddr_in *sin_server = (struct sockaddr_in *) &addr; |
1424 | struct sockaddr_in6 *sin_server6 = (struct sockaddr_in6 *) &addr; | 1424 | struct sockaddr_in6 *sin_server6 = (struct sockaddr_in6 *) &addr; |
1425 | int rc; | 1425 | int rc; |
1426 | 1426 | ||
1427 | memset(&addr, 0, sizeof(struct sockaddr)); | 1427 | memset(&addr, 0, sizeof(struct sockaddr_storage)); |
1428 | 1428 | ||
1429 | if (volume_info->UNCip && volume_info->UNC) { | 1429 | if (volume_info->UNCip && volume_info->UNC) { |
1430 | rc = cifs_inet_pton(AF_INET, volume_info->UNCip, | 1430 | rc = cifs_inet_pton(AF_INET, volume_info->UNCip, |
@@ -1435,9 +1435,9 @@ cifs_get_tcp_session(struct smb_vol *volume_info) | |||
1435 | rc = cifs_inet_pton(AF_INET6, volume_info->UNCip, | 1435 | rc = cifs_inet_pton(AF_INET6, volume_info->UNCip, |
1436 | &sin_server6->sin6_addr.in6_u); | 1436 | &sin_server6->sin6_addr.in6_u); |
1437 | if (rc > 0) | 1437 | if (rc > 0) |
1438 | addr.sa_family = AF_INET6; | 1438 | addr.ss_family = AF_INET6; |
1439 | } else { | 1439 | } else { |
1440 | addr.sa_family = AF_INET; | 1440 | addr.ss_family = AF_INET; |
1441 | } | 1441 | } |
1442 | 1442 | ||
1443 | if (rc <= 0) { | 1443 | if (rc <= 0) { |
@@ -1502,7 +1502,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info) | |||
1502 | tcp_ses->tcpStatus = CifsNew; | 1502 | tcp_ses->tcpStatus = CifsNew; |
1503 | ++tcp_ses->srv_count; | 1503 | ++tcp_ses->srv_count; |
1504 | 1504 | ||
1505 | if (addr.sa_family == AF_INET6) { | 1505 | if (addr.ss_family == AF_INET6) { |
1506 | cFYI(1, ("attempting ipv6 connect")); | 1506 | cFYI(1, ("attempting ipv6 connect")); |
1507 | /* BB should we allow ipv6 on port 139? */ | 1507 | /* BB should we allow ipv6 on port 139? */ |
1508 | /* other OS never observed in Wild doing 139 with v6 */ | 1508 | /* other OS never observed in Wild doing 139 with v6 */ |
@@ -1802,7 +1802,7 @@ ipv4_connect(struct TCP_Server_Info *server) | |||
1802 | * user space buffer | 1802 | * user space buffer |
1803 | */ | 1803 | */ |
1804 | socket->sk->sk_rcvtimeo = 7 * HZ; | 1804 | socket->sk->sk_rcvtimeo = 7 * HZ; |
1805 | socket->sk->sk_sndtimeo = 3 * HZ; | 1805 | socket->sk->sk_sndtimeo = 5 * HZ; |
1806 | 1806 | ||
1807 | /* make the bufsizes depend on wsize/rsize and max requests */ | 1807 | /* make the bufsizes depend on wsize/rsize and max requests */ |
1808 | if (server->noautotune) { | 1808 | if (server->noautotune) { |
@@ -1860,9 +1860,7 @@ ipv4_connect(struct TCP_Server_Info *server) | |||
1860 | smb_buf = (struct smb_hdr *)ses_init_buf; | 1860 | smb_buf = (struct smb_hdr *)ses_init_buf; |
1861 | /* sizeof RFC1002_SESSION_REQUEST with no scope */ | 1861 | /* sizeof RFC1002_SESSION_REQUEST with no scope */ |
1862 | smb_buf->smb_buf_length = 0x81000044; | 1862 | smb_buf->smb_buf_length = 0x81000044; |
1863 | rc = smb_send(socket, smb_buf, 0x44, | 1863 | rc = smb_send(server, smb_buf, 0x44); |
1864 | (struct sockaddr *) &server->addr.sockAddr, | ||
1865 | server->noblocksnd); | ||
1866 | kfree(ses_init_buf); | 1864 | kfree(ses_init_buf); |
1867 | msleep(1); /* RFC1001 layer in at least one server | 1865 | msleep(1); /* RFC1001 layer in at least one server |
1868 | requires very short break before negprot | 1866 | requires very short break before negprot |
@@ -1955,7 +1953,7 @@ ipv6_connect(struct TCP_Server_Info *server) | |||
1955 | * user space buffer | 1953 | * user space buffer |
1956 | */ | 1954 | */ |
1957 | socket->sk->sk_rcvtimeo = 7 * HZ; | 1955 | socket->sk->sk_rcvtimeo = 7 * HZ; |
1958 | socket->sk->sk_sndtimeo = 3 * HZ; | 1956 | socket->sk->sk_sndtimeo = 5 * HZ; |
1959 | server->ssocket = socket; | 1957 | server->ssocket = socket; |
1960 | 1958 | ||
1961 | return rc; | 1959 | return rc; |
diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 838d9c720a5c..964aad03c5ad 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c | |||
@@ -129,6 +129,17 @@ cifs_bp_rename_retry: | |||
129 | return full_path; | 129 | return full_path; |
130 | } | 130 | } |
131 | 131 | ||
132 | static void setup_cifs_dentry(struct cifsTconInfo *tcon, | ||
133 | struct dentry *direntry, | ||
134 | struct inode *newinode) | ||
135 | { | ||
136 | if (tcon->nocase) | ||
137 | direntry->d_op = &cifs_ci_dentry_ops; | ||
138 | else | ||
139 | direntry->d_op = &cifs_dentry_ops; | ||
140 | d_instantiate(direntry, newinode); | ||
141 | } | ||
142 | |||
132 | /* Inode operations in similar order to how they appear in Linux file fs.h */ | 143 | /* Inode operations in similar order to how they appear in Linux file fs.h */ |
133 | 144 | ||
134 | int | 145 | int |
@@ -139,14 +150,14 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
139 | int xid; | 150 | int xid; |
140 | int create_options = CREATE_NOT_DIR; | 151 | int create_options = CREATE_NOT_DIR; |
141 | int oplock = 0; | 152 | int oplock = 0; |
153 | /* BB below access is too much for the mknod to request */ | ||
142 | int desiredAccess = GENERIC_READ | GENERIC_WRITE; | 154 | int desiredAccess = GENERIC_READ | GENERIC_WRITE; |
143 | __u16 fileHandle; | 155 | __u16 fileHandle; |
144 | struct cifs_sb_info *cifs_sb; | 156 | struct cifs_sb_info *cifs_sb; |
145 | struct cifsTconInfo *pTcon; | 157 | struct cifsTconInfo *tcon; |
146 | char *full_path = NULL; | 158 | char *full_path = NULL; |
147 | FILE_ALL_INFO *buf = NULL; | 159 | FILE_ALL_INFO *buf = NULL; |
148 | struct inode *newinode = NULL; | 160 | struct inode *newinode = NULL; |
149 | struct cifsFileInfo *pCifsFile = NULL; | ||
150 | struct cifsInodeInfo *pCifsInode; | 161 | struct cifsInodeInfo *pCifsInode; |
151 | int disposition = FILE_OVERWRITE_IF; | 162 | int disposition = FILE_OVERWRITE_IF; |
152 | bool write_only = false; | 163 | bool write_only = false; |
@@ -154,7 +165,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
154 | xid = GetXid(); | 165 | xid = GetXid(); |
155 | 166 | ||
156 | cifs_sb = CIFS_SB(inode->i_sb); | 167 | cifs_sb = CIFS_SB(inode->i_sb); |
157 | pTcon = cifs_sb->tcon; | 168 | tcon = cifs_sb->tcon; |
158 | 169 | ||
159 | full_path = build_path_from_dentry(direntry); | 170 | full_path = build_path_from_dentry(direntry); |
160 | if (full_path == NULL) { | 171 | if (full_path == NULL) { |
@@ -162,6 +173,8 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
162 | return -ENOMEM; | 173 | return -ENOMEM; |
163 | } | 174 | } |
164 | 175 | ||
176 | mode &= ~current->fs->umask; | ||
177 | |||
165 | if (nd && (nd->flags & LOOKUP_OPEN)) { | 178 | if (nd && (nd->flags & LOOKUP_OPEN)) { |
166 | int oflags = nd->intent.open.flags; | 179 | int oflags = nd->intent.open.flags; |
167 | 180 | ||
@@ -196,17 +209,15 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
196 | return -ENOMEM; | 209 | return -ENOMEM; |
197 | } | 210 | } |
198 | 211 | ||
199 | mode &= ~current->fs->umask; | ||
200 | |||
201 | /* | 212 | /* |
202 | * if we're not using unix extensions, see if we need to set | 213 | * if we're not using unix extensions, see if we need to set |
203 | * ATTR_READONLY on the create call | 214 | * ATTR_READONLY on the create call |
204 | */ | 215 | */ |
205 | if (!pTcon->unix_ext && (mode & S_IWUGO) == 0) | 216 | if (!tcon->unix_ext && (mode & S_IWUGO) == 0) |
206 | create_options |= CREATE_OPTION_READONLY; | 217 | create_options |= CREATE_OPTION_READONLY; |
207 | 218 | ||
208 | if (cifs_sb->tcon->ses->capabilities & CAP_NT_SMBS) | 219 | if (cifs_sb->tcon->ses->capabilities & CAP_NT_SMBS) |
209 | rc = CIFSSMBOpen(xid, pTcon, full_path, disposition, | 220 | rc = CIFSSMBOpen(xid, tcon, full_path, disposition, |
210 | desiredAccess, create_options, | 221 | desiredAccess, create_options, |
211 | &fileHandle, &oplock, buf, cifs_sb->local_nls, | 222 | &fileHandle, &oplock, buf, cifs_sb->local_nls, |
212 | cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); | 223 | cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); |
@@ -215,7 +226,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
215 | 226 | ||
216 | if (rc == -EIO) { | 227 | if (rc == -EIO) { |
217 | /* old server, retry the open legacy style */ | 228 | /* old server, retry the open legacy style */ |
218 | rc = SMBLegacyOpen(xid, pTcon, full_path, disposition, | 229 | rc = SMBLegacyOpen(xid, tcon, full_path, disposition, |
219 | desiredAccess, create_options, | 230 | desiredAccess, create_options, |
220 | &fileHandle, &oplock, buf, cifs_sb->local_nls, | 231 | &fileHandle, &oplock, buf, cifs_sb->local_nls, |
221 | cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); | 232 | cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); |
@@ -225,7 +236,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
225 | } else { | 236 | } else { |
226 | /* If Open reported that we actually created a file | 237 | /* If Open reported that we actually created a file |
227 | then we now have to set the mode if possible */ | 238 | then we now have to set the mode if possible */ |
228 | if ((pTcon->unix_ext) && (oplock & CIFS_CREATE_ACTION)) { | 239 | if ((tcon->unix_ext) && (oplock & CIFS_CREATE_ACTION)) { |
229 | struct cifs_unix_set_info_args args = { | 240 | struct cifs_unix_set_info_args args = { |
230 | .mode = mode, | 241 | .mode = mode, |
231 | .ctime = NO_CHANGE_64, | 242 | .ctime = NO_CHANGE_64, |
@@ -244,20 +255,20 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
244 | args.uid = NO_CHANGE_64; | 255 | args.uid = NO_CHANGE_64; |
245 | args.gid = NO_CHANGE_64; | 256 | args.gid = NO_CHANGE_64; |
246 | } | 257 | } |
247 | CIFSSMBUnixSetInfo(xid, pTcon, full_path, &args, | 258 | CIFSSMBUnixSetInfo(xid, tcon, full_path, &args, |
248 | cifs_sb->local_nls, | 259 | cifs_sb->local_nls, |
249 | cifs_sb->mnt_cifs_flags & | 260 | cifs_sb->mnt_cifs_flags & |
250 | CIFS_MOUNT_MAP_SPECIAL_CHR); | 261 | CIFS_MOUNT_MAP_SPECIAL_CHR); |
251 | } else { | 262 | } else { |
252 | /* BB implement mode setting via Windows security | 263 | /* BB implement mode setting via Windows security |
253 | descriptors e.g. */ | 264 | descriptors e.g. */ |
254 | /* CIFSSMBWinSetPerms(xid,pTcon,path,mode,-1,-1,nls);*/ | 265 | /* CIFSSMBWinSetPerms(xid,tcon,path,mode,-1,-1,nls);*/ |
255 | 266 | ||
256 | /* Could set r/o dos attribute if mode & 0222 == 0 */ | 267 | /* Could set r/o dos attribute if mode & 0222 == 0 */ |
257 | } | 268 | } |
258 | 269 | ||
259 | /* server might mask mode so we have to query for it */ | 270 | /* server might mask mode so we have to query for it */ |
260 | if (pTcon->unix_ext) | 271 | if (tcon->unix_ext) |
261 | rc = cifs_get_inode_info_unix(&newinode, full_path, | 272 | rc = cifs_get_inode_info_unix(&newinode, full_path, |
262 | inode->i_sb, xid); | 273 | inode->i_sb, xid); |
263 | else { | 274 | else { |
@@ -283,22 +294,17 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
283 | } | 294 | } |
284 | 295 | ||
285 | if (rc != 0) { | 296 | if (rc != 0) { |
286 | cFYI(1, | 297 | cFYI(1, ("Create worked, get_inode_info failed rc = %d", |
287 | ("Create worked but get_inode_info failed rc = %d", | 298 | rc)); |
288 | rc)); | 299 | } else |
289 | } else { | 300 | setup_cifs_dentry(tcon, direntry, newinode); |
290 | if (pTcon->nocase) | 301 | |
291 | direntry->d_op = &cifs_ci_dentry_ops; | ||
292 | else | ||
293 | direntry->d_op = &cifs_dentry_ops; | ||
294 | d_instantiate(direntry, newinode); | ||
295 | } | ||
296 | if ((nd == NULL /* nfsd case - nfs srv does not set nd */) || | 302 | if ((nd == NULL /* nfsd case - nfs srv does not set nd */) || |
297 | (!(nd->flags & LOOKUP_OPEN))) { | 303 | (!(nd->flags & LOOKUP_OPEN))) { |
298 | /* mknod case - do not leave file open */ | 304 | /* mknod case - do not leave file open */ |
299 | CIFSSMBClose(xid, pTcon, fileHandle); | 305 | CIFSSMBClose(xid, tcon, fileHandle); |
300 | } else if (newinode) { | 306 | } else if (newinode) { |
301 | pCifsFile = | 307 | struct cifsFileInfo *pCifsFile = |
302 | kzalloc(sizeof(struct cifsFileInfo), GFP_KERNEL); | 308 | kzalloc(sizeof(struct cifsFileInfo), GFP_KERNEL); |
303 | 309 | ||
304 | if (pCifsFile == NULL) | 310 | if (pCifsFile == NULL) |
@@ -316,7 +322,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, | |||
316 | /* set the following in open now | 322 | /* set the following in open now |
317 | pCifsFile->pfile = file; */ | 323 | pCifsFile->pfile = file; */ |
318 | write_lock(&GlobalSMBSeslock); | 324 | write_lock(&GlobalSMBSeslock); |
319 | list_add(&pCifsFile->tlist, &pTcon->openFileList); | 325 | list_add(&pCifsFile->tlist, &tcon->openFileList); |
320 | pCifsInode = CIFS_I(newinode); | 326 | pCifsInode = CIFS_I(newinode); |
321 | if (pCifsInode) { | 327 | if (pCifsInode) { |
322 | /* if readable file instance put first in list*/ | 328 | /* if readable file instance put first in list*/ |
diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 5ab9896fdcb2..bcf7b5184664 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c | |||
@@ -1285,6 +1285,11 @@ int cifs_rmdir(struct inode *inode, struct dentry *direntry) | |||
1285 | cifsInode = CIFS_I(direntry->d_inode); | 1285 | cifsInode = CIFS_I(direntry->d_inode); |
1286 | cifsInode->time = 0; /* force revalidate to go get info when | 1286 | cifsInode->time = 0; /* force revalidate to go get info when |
1287 | needed */ | 1287 | needed */ |
1288 | |||
1289 | cifsInode = CIFS_I(inode); | ||
1290 | cifsInode->time = 0; /* force revalidate to get parent dir info | ||
1291 | since cached search results now invalid */ | ||
1292 | |||
1288 | direntry->d_inode->i_ctime = inode->i_ctime = inode->i_mtime = | 1293 | direntry->d_inode->i_ctime = inode->i_ctime = inode->i_mtime = |
1289 | current_fs_time(inode->i_sb); | 1294 | current_fs_time(inode->i_sb); |
1290 | 1295 | ||
diff --git a/fs/cifs/md5.c b/fs/cifs/md5.c index 462bbfefd4b6..98b66a54c319 100644 --- a/fs/cifs/md5.c +++ b/fs/cifs/md5.c | |||
@@ -10,8 +10,8 @@ | |||
10 | * with every copy. | 10 | * with every copy. |
11 | * | 11 | * |
12 | * To compute the message digest of a chunk of bytes, declare an | 12 | * To compute the message digest of a chunk of bytes, declare an |
13 | * MD5Context structure, pass it to MD5Init, call MD5Update as | 13 | * MD5Context structure, pass it to cifs_MD5_init, call cifs_MD5_update as |
14 | * needed on buffers full of bytes, and then call MD5Final, which | 14 | * needed on buffers full of bytes, and then call cifs_MD5_final, which |
15 | * will fill a supplied 16-byte array with the digest. | 15 | * will fill a supplied 16-byte array with the digest. |
16 | */ | 16 | */ |
17 | 17 | ||
@@ -45,7 +45,7 @@ byteReverse(unsigned char *buf, unsigned longs) | |||
45 | * initialization constants. | 45 | * initialization constants. |
46 | */ | 46 | */ |
47 | void | 47 | void |
48 | MD5Init(struct MD5Context *ctx) | 48 | cifs_MD5_init(struct MD5Context *ctx) |
49 | { | 49 | { |
50 | ctx->buf[0] = 0x67452301; | 50 | ctx->buf[0] = 0x67452301; |
51 | ctx->buf[1] = 0xefcdab89; | 51 | ctx->buf[1] = 0xefcdab89; |
@@ -61,7 +61,7 @@ MD5Init(struct MD5Context *ctx) | |||
61 | * of bytes. | 61 | * of bytes. |
62 | */ | 62 | */ |
63 | void | 63 | void |
64 | MD5Update(struct MD5Context *ctx, unsigned char const *buf, unsigned len) | 64 | cifs_MD5_update(struct MD5Context *ctx, unsigned char const *buf, unsigned len) |
65 | { | 65 | { |
66 | register __u32 t; | 66 | register __u32 t; |
67 | 67 | ||
@@ -110,7 +110,7 @@ MD5Update(struct MD5Context *ctx, unsigned char const *buf, unsigned len) | |||
110 | * 1 0* (64-bit count of bits processed, MSB-first) | 110 | * 1 0* (64-bit count of bits processed, MSB-first) |
111 | */ | 111 | */ |
112 | void | 112 | void |
113 | MD5Final(unsigned char digest[16], struct MD5Context *ctx) | 113 | cifs_MD5_final(unsigned char digest[16], struct MD5Context *ctx) |
114 | { | 114 | { |
115 | unsigned int count; | 115 | unsigned int count; |
116 | unsigned char *p; | 116 | unsigned char *p; |
@@ -165,7 +165,7 @@ MD5Final(unsigned char digest[16], struct MD5Context *ctx) | |||
165 | 165 | ||
166 | /* | 166 | /* |
167 | * The core of the MD5 algorithm, this alters an existing MD5 hash to | 167 | * The core of the MD5 algorithm, this alters an existing MD5 hash to |
168 | * reflect the addition of 16 longwords of new data. MD5Update blocks | 168 | * reflect the addition of 16 longwords of new data. cifs_MD5_update blocks |
169 | * the data and converts bytes into longwords for this routine. | 169 | * the data and converts bytes into longwords for this routine. |
170 | */ | 170 | */ |
171 | static void | 171 | static void |
@@ -267,9 +267,9 @@ hmac_md5_init_rfc2104(unsigned char *key, int key_len, | |||
267 | unsigned char tk[16]; | 267 | unsigned char tk[16]; |
268 | struct MD5Context tctx; | 268 | struct MD5Context tctx; |
269 | 269 | ||
270 | MD5Init(&tctx); | 270 | cifs_MD5_init(&tctx); |
271 | MD5Update(&tctx, key, key_len); | 271 | cifs_MD5_update(&tctx, key, key_len); |
272 | MD5Final(tk, &tctx); | 272 | cifs_MD5_final(tk, &tctx); |
273 | 273 | ||
274 | key = tk; | 274 | key = tk; |
275 | key_len = 16; | 275 | key_len = 16; |
@@ -287,8 +287,8 @@ hmac_md5_init_rfc2104(unsigned char *key, int key_len, | |||
287 | ctx->k_opad[i] ^= 0x5c; | 287 | ctx->k_opad[i] ^= 0x5c; |
288 | } | 288 | } |
289 | 289 | ||
290 | MD5Init(&ctx->ctx); | 290 | cifs_MD5_init(&ctx->ctx); |
291 | MD5Update(&ctx->ctx, ctx->k_ipad, 64); | 291 | cifs_MD5_update(&ctx->ctx, ctx->k_ipad, 64); |
292 | } | 292 | } |
293 | #endif | 293 | #endif |
294 | 294 | ||
@@ -317,8 +317,8 @@ hmac_md5_init_limK_to_64(const unsigned char *key, int key_len, | |||
317 | ctx->k_opad[i] ^= 0x5c; | 317 | ctx->k_opad[i] ^= 0x5c; |
318 | } | 318 | } |
319 | 319 | ||
320 | MD5Init(&ctx->ctx); | 320 | cifs_MD5_init(&ctx->ctx); |
321 | MD5Update(&ctx->ctx, ctx->k_ipad, 64); | 321 | cifs_MD5_update(&ctx->ctx, ctx->k_ipad, 64); |
322 | } | 322 | } |
323 | 323 | ||
324 | /*********************************************************************** | 324 | /*********************************************************************** |
@@ -328,7 +328,7 @@ void | |||
328 | hmac_md5_update(const unsigned char *text, int text_len, | 328 | hmac_md5_update(const unsigned char *text, int text_len, |
329 | struct HMACMD5Context *ctx) | 329 | struct HMACMD5Context *ctx) |
330 | { | 330 | { |
331 | MD5Update(&ctx->ctx, text, text_len); /* then text of datagram */ | 331 | cifs_MD5_update(&ctx->ctx, text, text_len); /* then text of datagram */ |
332 | } | 332 | } |
333 | 333 | ||
334 | /*********************************************************************** | 334 | /*********************************************************************** |
@@ -339,12 +339,12 @@ hmac_md5_final(unsigned char *digest, struct HMACMD5Context *ctx) | |||
339 | { | 339 | { |
340 | struct MD5Context ctx_o; | 340 | struct MD5Context ctx_o; |
341 | 341 | ||
342 | MD5Final(digest, &ctx->ctx); | 342 | cifs_MD5_final(digest, &ctx->ctx); |
343 | 343 | ||
344 | MD5Init(&ctx_o); | 344 | cifs_MD5_init(&ctx_o); |
345 | MD5Update(&ctx_o, ctx->k_opad, 64); | 345 | cifs_MD5_update(&ctx_o, ctx->k_opad, 64); |
346 | MD5Update(&ctx_o, digest, 16); | 346 | cifs_MD5_update(&ctx_o, digest, 16); |
347 | MD5Final(digest, &ctx_o); | 347 | cifs_MD5_final(digest, &ctx_o); |
348 | } | 348 | } |
349 | 349 | ||
350 | /*********************************************************** | 350 | /*********************************************************** |
diff --git a/fs/cifs/md5.h b/fs/cifs/md5.h index f7d4f4197bac..6fba8cb402fd 100644 --- a/fs/cifs/md5.h +++ b/fs/cifs/md5.h | |||
@@ -20,10 +20,10 @@ struct HMACMD5Context { | |||
20 | }; | 20 | }; |
21 | #endif /* _HMAC_MD5_H */ | 21 | #endif /* _HMAC_MD5_H */ |
22 | 22 | ||
23 | void MD5Init(struct MD5Context *context); | 23 | void cifs_MD5_init(struct MD5Context *context); |
24 | void MD5Update(struct MD5Context *context, unsigned char const *buf, | 24 | void cifs_MD5_update(struct MD5Context *context, unsigned char const *buf, |
25 | unsigned len); | 25 | unsigned len); |
26 | void MD5Final(unsigned char digest[16], struct MD5Context *context); | 26 | void cifs_MD5_final(unsigned char digest[16], struct MD5Context *context); |
27 | 27 | ||
28 | /* The following definitions come from lib/hmacmd5.c */ | 28 | /* The following definitions come from lib/hmacmd5.c */ |
29 | 29 | ||
diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index 7ebe6599ed3a..0ad3e2d116a6 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c | |||
@@ -154,81 +154,8 @@ void DeleteTconOplockQEntries(struct cifsTconInfo *tcon) | |||
154 | spin_unlock(&GlobalMid_Lock); | 154 | spin_unlock(&GlobalMid_Lock); |
155 | } | 155 | } |
156 | 156 | ||
157 | int | ||
158 | smb_send(struct socket *ssocket, struct smb_hdr *smb_buffer, | ||
159 | unsigned int smb_buf_length, struct sockaddr *sin, bool noblocksnd) | ||
160 | { | ||
161 | int rc = 0; | ||
162 | int i = 0; | ||
163 | struct msghdr smb_msg; | ||
164 | struct kvec iov; | ||
165 | unsigned len = smb_buf_length + 4; | ||
166 | |||
167 | if (ssocket == NULL) | ||
168 | return -ENOTSOCK; /* BB eventually add reconnect code here */ | ||
169 | iov.iov_base = smb_buffer; | ||
170 | iov.iov_len = len; | ||
171 | |||
172 | smb_msg.msg_name = sin; | ||
173 | smb_msg.msg_namelen = sizeof(struct sockaddr); | ||
174 | smb_msg.msg_control = NULL; | ||
175 | smb_msg.msg_controllen = 0; | ||
176 | if (noblocksnd) | ||
177 | smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; | ||
178 | else | ||
179 | smb_msg.msg_flags = MSG_NOSIGNAL; | ||
180 | |||
181 | /* smb header is converted in header_assemble. bcc and rest of SMB word | ||
182 | area, and byte area if necessary, is converted to littleendian in | ||
183 | cifssmb.c and RFC1001 len is converted to bigendian in smb_send | ||
184 | Flags2 is converted in SendReceive */ | ||
185 | |||
186 | smb_buffer->smb_buf_length = cpu_to_be32(smb_buffer->smb_buf_length); | ||
187 | cFYI(1, ("Sending smb of length %d", smb_buf_length)); | ||
188 | dump_smb(smb_buffer, len); | ||
189 | |||
190 | while (len > 0) { | ||
191 | rc = kernel_sendmsg(ssocket, &smb_msg, &iov, 1, len); | ||
192 | if ((rc == -ENOSPC) || (rc == -EAGAIN)) { | ||
193 | i++; | ||
194 | /* smaller timeout here than send2 since smaller size */ | ||
195 | /* Although it may not be required, this also is smaller | ||
196 | oplock break time */ | ||
197 | if (i > 12) { | ||
198 | cERROR(1, | ||
199 | ("sends on sock %p stuck for 7 seconds", | ||
200 | ssocket)); | ||
201 | rc = -EAGAIN; | ||
202 | break; | ||
203 | } | ||
204 | msleep(1 << i); | ||
205 | continue; | ||
206 | } | ||
207 | if (rc < 0) | ||
208 | break; | ||
209 | else | ||
210 | i = 0; /* reset i after each successful send */ | ||
211 | iov.iov_base += rc; | ||
212 | iov.iov_len -= rc; | ||
213 | len -= rc; | ||
214 | } | ||
215 | |||
216 | if (rc < 0) { | ||
217 | cERROR(1, ("Error %d sending data on socket to server", rc)); | ||
218 | } else { | ||
219 | rc = 0; | ||
220 | } | ||
221 | |||
222 | /* Don't want to modify the buffer as a | ||
223 | side effect of this call. */ | ||
224 | smb_buffer->smb_buf_length = smb_buf_length; | ||
225 | |||
226 | return rc; | ||
227 | } | ||
228 | |||
229 | static int | 157 | static int |
230 | smb_send2(struct TCP_Server_Info *server, struct kvec *iov, int n_vec, | 158 | smb_sendv(struct TCP_Server_Info *server, struct kvec *iov, int n_vec) |
231 | struct sockaddr *sin, bool noblocksnd) | ||
232 | { | 159 | { |
233 | int rc = 0; | 160 | int rc = 0; |
234 | int i = 0; | 161 | int i = 0; |
@@ -243,11 +170,11 @@ smb_send2(struct TCP_Server_Info *server, struct kvec *iov, int n_vec, | |||
243 | if (ssocket == NULL) | 170 | if (ssocket == NULL) |
244 | return -ENOTSOCK; /* BB eventually add reconnect code here */ | 171 | return -ENOTSOCK; /* BB eventually add reconnect code here */ |
245 | 172 | ||
246 | smb_msg.msg_name = sin; | 173 | smb_msg.msg_name = (struct sockaddr *) &server->addr.sockAddr; |
247 | smb_msg.msg_namelen = sizeof(struct sockaddr); | 174 | smb_msg.msg_namelen = sizeof(struct sockaddr); |
248 | smb_msg.msg_control = NULL; | 175 | smb_msg.msg_control = NULL; |
249 | smb_msg.msg_controllen = 0; | 176 | smb_msg.msg_controllen = 0; |
250 | if (noblocksnd) | 177 | if (server->noblocksnd) |
251 | smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; | 178 | smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; |
252 | else | 179 | else |
253 | smb_msg.msg_flags = MSG_NOSIGNAL; | 180 | smb_msg.msg_flags = MSG_NOSIGNAL; |
@@ -272,7 +199,25 @@ smb_send2(struct TCP_Server_Info *server, struct kvec *iov, int n_vec, | |||
272 | n_vec - first_vec, total_len); | 199 | n_vec - first_vec, total_len); |
273 | if ((rc == -ENOSPC) || (rc == -EAGAIN)) { | 200 | if ((rc == -ENOSPC) || (rc == -EAGAIN)) { |
274 | i++; | 201 | i++; |
275 | if (i >= 14) { | 202 | /* if blocking send we try 3 times, since each can block |
203 | for 5 seconds. For nonblocking we have to try more | ||
204 | but wait increasing amounts of time allowing time for | ||
205 | socket to clear. The overall time we wait in either | ||
206 | case to send on the socket is about 15 seconds. | ||
207 | Similarly we wait for 15 seconds for | ||
208 | a response from the server in SendReceive[2] | ||
209 | for the server to send a response back for | ||
210 | most types of requests (except SMB Write | ||
211 | past end of file which can be slow, and | ||
212 | blocking lock operations). NFS waits slightly longer | ||
213 | than CIFS, but this can make it take longer for | ||
214 | nonresponsive servers to be detected and 15 seconds | ||
215 | is more than enough time for modern networks to | ||
216 | send a packet. In most cases if we fail to send | ||
217 | after the retries we will kill the socket and | ||
218 | reconnect which may clear the network problem. | ||
219 | */ | ||
220 | if ((i >= 14) || (!server->noblocksnd && (i > 2))) { | ||
276 | cERROR(1, | 221 | cERROR(1, |
277 | ("sends on sock %p stuck for 15 seconds", | 222 | ("sends on sock %p stuck for 15 seconds", |
278 | ssocket)); | 223 | ssocket)); |
@@ -339,6 +284,18 @@ smb_send2(struct TCP_Server_Info *server, struct kvec *iov, int n_vec, | |||
339 | return rc; | 284 | return rc; |
340 | } | 285 | } |
341 | 286 | ||
287 | int | ||
288 | smb_send(struct TCP_Server_Info *server, struct smb_hdr *smb_buffer, | ||
289 | unsigned int smb_buf_length) | ||
290 | { | ||
291 | struct kvec iov; | ||
292 | |||
293 | iov.iov_base = smb_buffer; | ||
294 | iov.iov_len = smb_buf_length + 4; | ||
295 | |||
296 | return smb_sendv(server, &iov, 1); | ||
297 | } | ||
298 | |||
342 | static int wait_for_free_request(struct cifsSesInfo *ses, const int long_op) | 299 | static int wait_for_free_request(struct cifsSesInfo *ses, const int long_op) |
343 | { | 300 | { |
344 | if (long_op == CIFS_ASYNC_OP) { | 301 | if (long_op == CIFS_ASYNC_OP) { |
@@ -540,9 +497,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, | |||
540 | #ifdef CONFIG_CIFS_STATS2 | 497 | #ifdef CONFIG_CIFS_STATS2 |
541 | atomic_inc(&ses->server->inSend); | 498 | atomic_inc(&ses->server->inSend); |
542 | #endif | 499 | #endif |
543 | rc = smb_send2(ses->server, iov, n_vec, | 500 | rc = smb_sendv(ses->server, iov, n_vec); |
544 | (struct sockaddr *) &(ses->server->addr.sockAddr), | ||
545 | ses->server->noblocksnd); | ||
546 | #ifdef CONFIG_CIFS_STATS2 | 501 | #ifdef CONFIG_CIFS_STATS2 |
547 | atomic_dec(&ses->server->inSend); | 502 | atomic_dec(&ses->server->inSend); |
548 | midQ->when_sent = jiffies; | 503 | midQ->when_sent = jiffies; |
@@ -736,9 +691,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, | |||
736 | #ifdef CONFIG_CIFS_STATS2 | 691 | #ifdef CONFIG_CIFS_STATS2 |
737 | atomic_inc(&ses->server->inSend); | 692 | atomic_inc(&ses->server->inSend); |
738 | #endif | 693 | #endif |
739 | rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, | 694 | rc = smb_send(ses->server, in_buf, in_buf->smb_buf_length); |
740 | (struct sockaddr *) &(ses->server->addr.sockAddr), | ||
741 | ses->server->noblocksnd); | ||
742 | #ifdef CONFIG_CIFS_STATS2 | 695 | #ifdef CONFIG_CIFS_STATS2 |
743 | atomic_dec(&ses->server->inSend); | 696 | atomic_dec(&ses->server->inSend); |
744 | midQ->when_sent = jiffies; | 697 | midQ->when_sent = jiffies; |
@@ -879,9 +832,7 @@ send_nt_cancel(struct cifsTconInfo *tcon, struct smb_hdr *in_buf, | |||
879 | mutex_unlock(&ses->server->srv_mutex); | 832 | mutex_unlock(&ses->server->srv_mutex); |
880 | return rc; | 833 | return rc; |
881 | } | 834 | } |
882 | rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, | 835 | rc = smb_send(ses->server, in_buf, in_buf->smb_buf_length); |
883 | (struct sockaddr *) &(ses->server->addr.sockAddr), | ||
884 | ses->server->noblocksnd); | ||
885 | mutex_unlock(&ses->server->srv_mutex); | 836 | mutex_unlock(&ses->server->srv_mutex); |
886 | return rc; | 837 | return rc; |
887 | } | 838 | } |
@@ -973,9 +924,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, | |||
973 | #ifdef CONFIG_CIFS_STATS2 | 924 | #ifdef CONFIG_CIFS_STATS2 |
974 | atomic_inc(&ses->server->inSend); | 925 | atomic_inc(&ses->server->inSend); |
975 | #endif | 926 | #endif |
976 | rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, | 927 | rc = smb_send(ses->server, in_buf, in_buf->smb_buf_length); |
977 | (struct sockaddr *) &(ses->server->addr.sockAddr), | ||
978 | ses->server->noblocksnd); | ||
979 | #ifdef CONFIG_CIFS_STATS2 | 928 | #ifdef CONFIG_CIFS_STATS2 |
980 | atomic_dec(&ses->server->inSend); | 929 | atomic_dec(&ses->server->inSend); |
981 | midQ->when_sent = jiffies; | 930 | midQ->when_sent = jiffies; |
diff --git a/fs/coda/Kconfig b/fs/coda/Kconfig new file mode 100644 index 000000000000..c0e5a7fad06d --- /dev/null +++ b/fs/coda/Kconfig | |||
@@ -0,0 +1,21 @@ | |||
1 | config CODA_FS | ||
2 | tristate "Coda file system support (advanced network fs)" | ||
3 | depends on INET | ||
4 | help | ||
5 | Coda is an advanced network file system, similar to NFS in that it | ||
6 | enables you to mount file systems of a remote server and access them | ||
7 | with regular Unix commands as if they were sitting on your hard | ||
8 | disk. Coda has several advantages over NFS: support for | ||
9 | disconnected operation (e.g. for laptops), read/write server | ||
10 | replication, security model for authentication and encryption, | ||
11 | persistent client caches and write back caching. | ||
12 | |||
13 | If you say Y here, your Linux box will be able to act as a Coda | ||
14 | *client*. You will need user level code as well, both for the | ||
15 | client and server. Servers are currently user level, i.e. they need | ||
16 | no kernel support. Please read | ||
17 | <file:Documentation/filesystems/coda.txt> and check out the Coda | ||
18 | home page <http://www.coda.cs.cmu.edu/>. | ||
19 | |||
20 | To compile the coda client support as a module, choose M here: the | ||
21 | module will be called coda. | ||
diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 5235c67e7594..c8f8d5904f5e 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c | |||
@@ -538,6 +538,7 @@ static int dev_ifsioc(unsigned int fd, unsigned int cmd, unsigned long arg) | |||
538 | * cannot be fixed without breaking all existing apps. | 538 | * cannot be fixed without breaking all existing apps. |
539 | */ | 539 | */ |
540 | case TUNSETIFF: | 540 | case TUNSETIFF: |
541 | case TUNGETIFF: | ||
541 | case SIOCGIFFLAGS: | 542 | case SIOCGIFFLAGS: |
542 | case SIOCGIFMETRIC: | 543 | case SIOCGIFMETRIC: |
543 | case SIOCGIFMTU: | 544 | case SIOCGIFMTU: |
@@ -1982,6 +1983,11 @@ COMPATIBLE_IOCTL(TUNSETNOCSUM) | |||
1982 | COMPATIBLE_IOCTL(TUNSETDEBUG) | 1983 | COMPATIBLE_IOCTL(TUNSETDEBUG) |
1983 | COMPATIBLE_IOCTL(TUNSETPERSIST) | 1984 | COMPATIBLE_IOCTL(TUNSETPERSIST) |
1984 | COMPATIBLE_IOCTL(TUNSETOWNER) | 1985 | COMPATIBLE_IOCTL(TUNSETOWNER) |
1986 | COMPATIBLE_IOCTL(TUNSETLINK) | ||
1987 | COMPATIBLE_IOCTL(TUNSETGROUP) | ||
1988 | COMPATIBLE_IOCTL(TUNGETFEATURES) | ||
1989 | COMPATIBLE_IOCTL(TUNSETOFFLOAD) | ||
1990 | COMPATIBLE_IOCTL(TUNSETTXFILTER) | ||
1985 | /* Big V */ | 1991 | /* Big V */ |
1986 | COMPATIBLE_IOCTL(VT_SETMODE) | 1992 | COMPATIBLE_IOCTL(VT_SETMODE) |
1987 | COMPATIBLE_IOCTL(VT_GETMODE) | 1993 | COMPATIBLE_IOCTL(VT_GETMODE) |
@@ -2573,6 +2579,7 @@ HANDLE_IOCTL(SIOCGIFPFLAGS, dev_ifsioc) | |||
2573 | HANDLE_IOCTL(SIOCGIFTXQLEN, dev_ifsioc) | 2579 | HANDLE_IOCTL(SIOCGIFTXQLEN, dev_ifsioc) |
2574 | HANDLE_IOCTL(SIOCSIFTXQLEN, dev_ifsioc) | 2580 | HANDLE_IOCTL(SIOCSIFTXQLEN, dev_ifsioc) |
2575 | HANDLE_IOCTL(TUNSETIFF, dev_ifsioc) | 2581 | HANDLE_IOCTL(TUNSETIFF, dev_ifsioc) |
2582 | HANDLE_IOCTL(TUNGETIFF, dev_ifsioc) | ||
2576 | HANDLE_IOCTL(SIOCETHTOOL, ethtool_ioctl) | 2583 | HANDLE_IOCTL(SIOCETHTOOL, ethtool_ioctl) |
2577 | HANDLE_IOCTL(SIOCBONDENSLAVE, bond_ioctl) | 2584 | HANDLE_IOCTL(SIOCBONDENSLAVE, bond_ioctl) |
2578 | HANDLE_IOCTL(SIOCBONDRELEASE, bond_ioctl) | 2585 | HANDLE_IOCTL(SIOCBONDRELEASE, bond_ioctl) |
diff --git a/fs/configfs/Kconfig b/fs/configfs/Kconfig new file mode 100644 index 000000000000..13587cc97a0b --- /dev/null +++ b/fs/configfs/Kconfig | |||
@@ -0,0 +1,11 @@ | |||
1 | config CONFIGFS_FS | ||
2 | tristate "Userspace-driven configuration filesystem" | ||
3 | depends on SYSFS | ||
4 | help | ||
5 | configfs is a ram-based filesystem that provides the converse | ||
6 | of sysfs's functionality. Where sysfs is a filesystem-based | ||
7 | view of kernel objects, configfs is a filesystem-based manager | ||
8 | of kernel objects, or config_items. | ||
9 | |||
10 | Both sysfs and configfs can and should exist together on the | ||
11 | same system. One is not a replacement for the other. | ||
diff --git a/fs/cramfs/Kconfig b/fs/cramfs/Kconfig new file mode 100644 index 000000000000..cd06466f365e --- /dev/null +++ b/fs/cramfs/Kconfig | |||
@@ -0,0 +1,19 @@ | |||
1 | config CRAMFS | ||
2 | tristate "Compressed ROM file system support (cramfs)" | ||
3 | depends on BLOCK | ||
4 | select ZLIB_INFLATE | ||
5 | help | ||
6 | Saying Y here includes support for CramFs (Compressed ROM File | ||
7 | System). CramFs is designed to be a simple, small, and compressed | ||
8 | file system for ROM based embedded systems. CramFs is read-only, | ||
9 | limited to 256MB file systems (with 16MB files), and doesn't support | ||
10 | 16/32 bits uid/gid, hard links and timestamps. | ||
11 | |||
12 | See <file:Documentation/filesystems/cramfs.txt> and | ||
13 | <file:fs/cramfs/README> for further information. | ||
14 | |||
15 | To compile this as a module, choose M here: the module will be called | ||
16 | cramfs. Note that the root file system (the one containing the | ||
17 | directory /) cannot be compiled as a module. | ||
18 | |||
19 | If unsure, say N. | ||
diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index eba87ff3177b..894a32d438d5 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c | |||
@@ -168,7 +168,7 @@ static int dlm_plock_callback(struct plock_op *op) | |||
168 | notify = xop->callback; | 168 | notify = xop->callback; |
169 | 169 | ||
170 | if (op->info.rv) { | 170 | if (op->info.rv) { |
171 | notify(flc, NULL, op->info.rv); | 171 | notify(fl, NULL, op->info.rv); |
172 | goto out; | 172 | goto out; |
173 | } | 173 | } |
174 | 174 | ||
@@ -187,7 +187,7 @@ static int dlm_plock_callback(struct plock_op *op) | |||
187 | (unsigned long long)op->info.number, file, fl); | 187 | (unsigned long long)op->info.number, file, fl); |
188 | } | 188 | } |
189 | 189 | ||
190 | rv = notify(flc, NULL, 0); | 190 | rv = notify(fl, NULL, 0); |
191 | if (rv) { | 191 | if (rv) { |
192 | /* XXX: We need to cancel the fs lock here: */ | 192 | /* XXX: We need to cancel the fs lock here: */ |
193 | log_print("dlm_plock_callback: lock granted after lock request " | 193 | log_print("dlm_plock_callback: lock granted after lock request " |
@@ -304,7 +304,9 @@ int dlm_posix_get(dlm_lockspace_t *lockspace, u64 number, struct file *file, | |||
304 | if (rv == -ENOENT) | 304 | if (rv == -ENOENT) |
305 | rv = 0; | 305 | rv = 0; |
306 | else if (rv > 0) { | 306 | else if (rv > 0) { |
307 | locks_init_lock(fl); | ||
307 | fl->fl_type = (op->info.ex) ? F_WRLCK : F_RDLCK; | 308 | fl->fl_type = (op->info.ex) ? F_WRLCK : F_RDLCK; |
309 | fl->fl_flags = FL_POSIX; | ||
308 | fl->fl_pid = op->info.pid; | 310 | fl->fl_pid = op->info.pid; |
309 | fl->fl_start = op->info.start; | 311 | fl->fl_start = op->info.start; |
310 | fl->fl_end = op->info.end; | 312 | fl->fl_end = op->info.end; |
diff --git a/fs/dquot.c b/fs/dquot.c index 48c0571f831d..bca3cac4bee7 100644 --- a/fs/dquot.c +++ b/fs/dquot.c | |||
@@ -87,14 +87,17 @@ | |||
87 | #define __DQUOT_PARANOIA | 87 | #define __DQUOT_PARANOIA |
88 | 88 | ||
89 | /* | 89 | /* |
90 | * There are two quota SMP locks. dq_list_lock protects all lists with quotas | 90 | * There are three quota SMP locks. dq_list_lock protects all lists with quotas |
91 | * and quota formats and also dqstats structure containing statistics about the | 91 | * and quota formats, dqstats structure containing statistics about the lists |
92 | * lists. dq_data_lock protects data from dq_dqb and also mem_dqinfo structures | 92 | * dq_data_lock protects data from dq_dqb and also mem_dqinfo structures and |
93 | * and also guards consistency of dquot->dq_dqb with inode->i_blocks, i_bytes. | 93 | * also guards consistency of dquot->dq_dqb with inode->i_blocks, i_bytes. |
94 | * i_blocks and i_bytes updates itself are guarded by i_lock acquired directly | 94 | * i_blocks and i_bytes updates itself are guarded by i_lock acquired directly |
95 | * in inode_add_bytes() and inode_sub_bytes(). | 95 | * in inode_add_bytes() and inode_sub_bytes(). dq_state_lock protects |
96 | * modifications of quota state (on quotaon and quotaoff) and readers who care | ||
97 | * about latest values take it as well. | ||
96 | * | 98 | * |
97 | * The spinlock ordering is hence: dq_data_lock > dq_list_lock > i_lock | 99 | * The spinlock ordering is hence: dq_data_lock > dq_list_lock > i_lock, |
100 | * dq_list_lock > dq_state_lock | ||
98 | * | 101 | * |
99 | * Note that some things (eg. sb pointer, type, id) doesn't change during | 102 | * Note that some things (eg. sb pointer, type, id) doesn't change during |
100 | * the life of the dquot structure and so needn't to be protected by a lock | 103 | * the life of the dquot structure and so needn't to be protected by a lock |
@@ -103,12 +106,7 @@ | |||
103 | * operation is just reading pointers from inode (or not using them at all) the | 106 | * operation is just reading pointers from inode (or not using them at all) the |
104 | * read lock is enough. If pointers are altered function must hold write lock | 107 | * read lock is enough. If pointers are altered function must hold write lock |
105 | * (these locking rules also apply for S_NOQUOTA flag in the inode - note that | 108 | * (these locking rules also apply for S_NOQUOTA flag in the inode - note that |
106 | * for altering the flag i_mutex is also needed). If operation is holding | 109 | * for altering the flag i_mutex is also needed). |
107 | * reference to dquot in other way (e.g. quotactl ops) it must be guarded by | ||
108 | * dqonoff_mutex. | ||
109 | * This locking assures that: | ||
110 | * a) update/access to dquot pointers in inode is serialized | ||
111 | * b) everyone is guarded against invalidate_dquots() | ||
112 | * | 110 | * |
113 | * Each dquot has its dq_lock mutex. Locked dquots might not be referenced | 111 | * Each dquot has its dq_lock mutex. Locked dquots might not be referenced |
114 | * from inodes (dquot_alloc_space() and such don't check the dq_lock). | 112 | * from inodes (dquot_alloc_space() and such don't check the dq_lock). |
@@ -122,10 +120,17 @@ | |||
122 | * Lock ordering (including related VFS locks) is the following: | 120 | * Lock ordering (including related VFS locks) is the following: |
123 | * i_mutex > dqonoff_sem > journal_lock > dqptr_sem > dquot->dq_lock > | 121 | * i_mutex > dqonoff_sem > journal_lock > dqptr_sem > dquot->dq_lock > |
124 | * dqio_mutex | 122 | * dqio_mutex |
123 | * The lock ordering of dqptr_sem imposed by quota code is only dqonoff_sem > | ||
124 | * dqptr_sem. But filesystem has to count with the fact that functions such as | ||
125 | * dquot_alloc_space() acquire dqptr_sem and they usually have to be called | ||
126 | * from inside a transaction to keep filesystem consistency after a crash. Also | ||
127 | * filesystems usually want to do some IO on dquot from ->mark_dirty which is | ||
128 | * called with dqptr_sem held. | ||
125 | * i_mutex on quota files is special (it's below dqio_mutex) | 129 | * i_mutex on quota files is special (it's below dqio_mutex) |
126 | */ | 130 | */ |
127 | 131 | ||
128 | static DEFINE_SPINLOCK(dq_list_lock); | 132 | static DEFINE_SPINLOCK(dq_list_lock); |
133 | static DEFINE_SPINLOCK(dq_state_lock); | ||
129 | DEFINE_SPINLOCK(dq_data_lock); | 134 | DEFINE_SPINLOCK(dq_data_lock); |
130 | 135 | ||
131 | static char *quotatypes[] = INITQFNAMES; | 136 | static char *quotatypes[] = INITQFNAMES; |
@@ -428,7 +433,7 @@ static inline void do_destroy_dquot(struct dquot *dquot) | |||
428 | * quota is disabled and pointers from inodes removed so there cannot be new | 433 | * quota is disabled and pointers from inodes removed so there cannot be new |
429 | * quota users. There can still be some users of quotas due to inodes being | 434 | * quota users. There can still be some users of quotas due to inodes being |
430 | * just deleted or pruned by prune_icache() (those are not attached to any | 435 | * just deleted or pruned by prune_icache() (those are not attached to any |
431 | * list). We have to wait for such users. | 436 | * list) or parallel quotactl call. We have to wait for such users. |
432 | */ | 437 | */ |
433 | static void invalidate_dquots(struct super_block *sb, int type) | 438 | static void invalidate_dquots(struct super_block *sb, int type) |
434 | { | 439 | { |
@@ -600,7 +605,6 @@ static struct shrinker dqcache_shrinker = { | |||
600 | /* | 605 | /* |
601 | * Put reference to dquot | 606 | * Put reference to dquot |
602 | * NOTE: If you change this function please check whether dqput_blocks() works right... | 607 | * NOTE: If you change this function please check whether dqput_blocks() works right... |
603 | * MUST be called with either dqptr_sem or dqonoff_mutex held | ||
604 | */ | 608 | */ |
605 | void dqput(struct dquot *dquot) | 609 | void dqput(struct dquot *dquot) |
606 | { | 610 | { |
@@ -697,36 +701,30 @@ static struct dquot *get_empty_dquot(struct super_block *sb, int type) | |||
697 | } | 701 | } |
698 | 702 | ||
699 | /* | 703 | /* |
700 | * Check whether dquot is in memory. | ||
701 | * MUST be called with either dqptr_sem or dqonoff_mutex held | ||
702 | */ | ||
703 | int dquot_is_cached(struct super_block *sb, unsigned int id, int type) | ||
704 | { | ||
705 | unsigned int hashent = hashfn(sb, id, type); | ||
706 | int ret = 0; | ||
707 | |||
708 | if (!sb_has_quota_active(sb, type)) | ||
709 | return 0; | ||
710 | spin_lock(&dq_list_lock); | ||
711 | if (find_dquot(hashent, sb, id, type) != NODQUOT) | ||
712 | ret = 1; | ||
713 | spin_unlock(&dq_list_lock); | ||
714 | return ret; | ||
715 | } | ||
716 | |||
717 | /* | ||
718 | * Get reference to dquot | 704 | * Get reference to dquot |
719 | * MUST be called with either dqptr_sem or dqonoff_mutex held | 705 | * |
706 | * Locking is slightly tricky here. We are guarded from parallel quotaoff() | ||
707 | * destroying our dquot by: | ||
708 | * a) checking for quota flags under dq_list_lock and | ||
709 | * b) getting a reference to dquot before we release dq_list_lock | ||
720 | */ | 710 | */ |
721 | struct dquot *dqget(struct super_block *sb, unsigned int id, int type) | 711 | struct dquot *dqget(struct super_block *sb, unsigned int id, int type) |
722 | { | 712 | { |
723 | unsigned int hashent = hashfn(sb, id, type); | 713 | unsigned int hashent = hashfn(sb, id, type); |
724 | struct dquot *dquot, *empty = NODQUOT; | 714 | struct dquot *dquot = NODQUOT, *empty = NODQUOT; |
725 | 715 | ||
726 | if (!sb_has_quota_active(sb, type)) | 716 | if (!sb_has_quota_active(sb, type)) |
727 | return NODQUOT; | 717 | return NODQUOT; |
728 | we_slept: | 718 | we_slept: |
729 | spin_lock(&dq_list_lock); | 719 | spin_lock(&dq_list_lock); |
720 | spin_lock(&dq_state_lock); | ||
721 | if (!sb_has_quota_active(sb, type)) { | ||
722 | spin_unlock(&dq_state_lock); | ||
723 | spin_unlock(&dq_list_lock); | ||
724 | goto out; | ||
725 | } | ||
726 | spin_unlock(&dq_state_lock); | ||
727 | |||
730 | if ((dquot = find_dquot(hashent, sb, id, type)) == NODQUOT) { | 728 | if ((dquot = find_dquot(hashent, sb, id, type)) == NODQUOT) { |
731 | if (empty == NODQUOT) { | 729 | if (empty == NODQUOT) { |
732 | spin_unlock(&dq_list_lock); | 730 | spin_unlock(&dq_list_lock); |
@@ -735,6 +733,7 @@ we_slept: | |||
735 | goto we_slept; | 733 | goto we_slept; |
736 | } | 734 | } |
737 | dquot = empty; | 735 | dquot = empty; |
736 | empty = NODQUOT; | ||
738 | dquot->dq_id = id; | 737 | dquot->dq_id = id; |
739 | /* all dquots go on the inuse_list */ | 738 | /* all dquots go on the inuse_list */ |
740 | put_inuse(dquot); | 739 | put_inuse(dquot); |
@@ -749,8 +748,6 @@ we_slept: | |||
749 | dqstats.cache_hits++; | 748 | dqstats.cache_hits++; |
750 | dqstats.lookups++; | 749 | dqstats.lookups++; |
751 | spin_unlock(&dq_list_lock); | 750 | spin_unlock(&dq_list_lock); |
752 | if (empty) | ||
753 | do_destroy_dquot(empty); | ||
754 | } | 751 | } |
755 | /* Wait for dq_lock - after this we know that either dquot_release() is already | 752 | /* Wait for dq_lock - after this we know that either dquot_release() is already |
756 | * finished or it will be canceled due to dq_count > 1 test */ | 753 | * finished or it will be canceled due to dq_count > 1 test */ |
@@ -758,11 +755,15 @@ we_slept: | |||
758 | /* Read the dquot and instantiate it (everything done only if needed) */ | 755 | /* Read the dquot and instantiate it (everything done only if needed) */ |
759 | if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags) && sb->dq_op->acquire_dquot(dquot) < 0) { | 756 | if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags) && sb->dq_op->acquire_dquot(dquot) < 0) { |
760 | dqput(dquot); | 757 | dqput(dquot); |
761 | return NODQUOT; | 758 | dquot = NODQUOT; |
759 | goto out; | ||
762 | } | 760 | } |
763 | #ifdef __DQUOT_PARANOIA | 761 | #ifdef __DQUOT_PARANOIA |
764 | BUG_ON(!dquot->dq_sb); /* Has somebody invalidated entry under us? */ | 762 | BUG_ON(!dquot->dq_sb); /* Has somebody invalidated entry under us? */ |
765 | #endif | 763 | #endif |
764 | out: | ||
765 | if (empty) | ||
766 | do_destroy_dquot(empty); | ||
766 | 767 | ||
767 | return dquot; | 768 | return dquot; |
768 | } | 769 | } |
@@ -1198,63 +1199,76 @@ static int info_bdq_free(struct dquot *dquot, qsize_t space) | |||
1198 | } | 1199 | } |
1199 | /* | 1200 | /* |
1200 | * Initialize quota pointers in inode | 1201 | * Initialize quota pointers in inode |
1201 | * Transaction must be started at entry | 1202 | * We do things in a bit complicated way but by that we avoid calling |
1203 | * dqget() and thus filesystem callbacks under dqptr_sem. | ||
1202 | */ | 1204 | */ |
1203 | int dquot_initialize(struct inode *inode, int type) | 1205 | int dquot_initialize(struct inode *inode, int type) |
1204 | { | 1206 | { |
1205 | unsigned int id = 0; | 1207 | unsigned int id = 0; |
1206 | int cnt, ret = 0; | 1208 | int cnt, ret = 0; |
1209 | struct dquot *got[MAXQUOTAS] = { NODQUOT, NODQUOT }; | ||
1210 | struct super_block *sb = inode->i_sb; | ||
1207 | 1211 | ||
1208 | /* First test before acquiring mutex - solves deadlocks when we | 1212 | /* First test before acquiring mutex - solves deadlocks when we |
1209 | * re-enter the quota code and are already holding the mutex */ | 1213 | * re-enter the quota code and are already holding the mutex */ |
1210 | if (IS_NOQUOTA(inode)) | 1214 | if (IS_NOQUOTA(inode)) |
1211 | return 0; | 1215 | return 0; |
1212 | down_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | 1216 | |
1217 | /* First get references to structures we might need. */ | ||
1218 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | ||
1219 | if (type != -1 && cnt != type) | ||
1220 | continue; | ||
1221 | switch (cnt) { | ||
1222 | case USRQUOTA: | ||
1223 | id = inode->i_uid; | ||
1224 | break; | ||
1225 | case GRPQUOTA: | ||
1226 | id = inode->i_gid; | ||
1227 | break; | ||
1228 | } | ||
1229 | got[cnt] = dqget(sb, id, cnt); | ||
1230 | } | ||
1231 | |||
1232 | down_write(&sb_dqopt(sb)->dqptr_sem); | ||
1213 | /* Having dqptr_sem we know NOQUOTA flags can't be altered... */ | 1233 | /* Having dqptr_sem we know NOQUOTA flags can't be altered... */ |
1214 | if (IS_NOQUOTA(inode)) | 1234 | if (IS_NOQUOTA(inode)) |
1215 | goto out_err; | 1235 | goto out_err; |
1216 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | 1236 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { |
1217 | if (type != -1 && cnt != type) | 1237 | if (type != -1 && cnt != type) |
1218 | continue; | 1238 | continue; |
1239 | /* Avoid races with quotaoff() */ | ||
1240 | if (!sb_has_quota_active(sb, cnt)) | ||
1241 | continue; | ||
1219 | if (inode->i_dquot[cnt] == NODQUOT) { | 1242 | if (inode->i_dquot[cnt] == NODQUOT) { |
1220 | switch (cnt) { | 1243 | inode->i_dquot[cnt] = got[cnt]; |
1221 | case USRQUOTA: | 1244 | got[cnt] = NODQUOT; |
1222 | id = inode->i_uid; | ||
1223 | break; | ||
1224 | case GRPQUOTA: | ||
1225 | id = inode->i_gid; | ||
1226 | break; | ||
1227 | } | ||
1228 | inode->i_dquot[cnt] = dqget(inode->i_sb, id, cnt); | ||
1229 | } | 1245 | } |
1230 | } | 1246 | } |
1231 | out_err: | 1247 | out_err: |
1232 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | 1248 | up_write(&sb_dqopt(sb)->dqptr_sem); |
1249 | /* Drop unused references */ | ||
1250 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) | ||
1251 | dqput(got[cnt]); | ||
1233 | return ret; | 1252 | return ret; |
1234 | } | 1253 | } |
1235 | 1254 | ||
1236 | /* | 1255 | /* |
1237 | * Release all quotas referenced by inode | 1256 | * Release all quotas referenced by inode |
1238 | * Transaction must be started at an entry | ||
1239 | */ | 1257 | */ |
1240 | int dquot_drop_locked(struct inode *inode) | 1258 | int dquot_drop(struct inode *inode) |
1241 | { | 1259 | { |
1242 | int cnt; | 1260 | int cnt; |
1261 | struct dquot *put[MAXQUOTAS]; | ||
1243 | 1262 | ||
1263 | down_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1244 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | 1264 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { |
1245 | if (inode->i_dquot[cnt] != NODQUOT) { | 1265 | put[cnt] = inode->i_dquot[cnt]; |
1246 | dqput(inode->i_dquot[cnt]); | 1266 | inode->i_dquot[cnt] = NODQUOT; |
1247 | inode->i_dquot[cnt] = NODQUOT; | ||
1248 | } | ||
1249 | } | 1267 | } |
1250 | return 0; | ||
1251 | } | ||
1252 | |||
1253 | int dquot_drop(struct inode *inode) | ||
1254 | { | ||
1255 | down_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1256 | dquot_drop_locked(inode); | ||
1257 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | 1268 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); |
1269 | |||
1270 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) | ||
1271 | dqput(put[cnt]); | ||
1258 | return 0; | 1272 | return 0; |
1259 | } | 1273 | } |
1260 | 1274 | ||
@@ -1470,8 +1484,9 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) | |||
1470 | qsize_t space; | 1484 | qsize_t space; |
1471 | struct dquot *transfer_from[MAXQUOTAS]; | 1485 | struct dquot *transfer_from[MAXQUOTAS]; |
1472 | struct dquot *transfer_to[MAXQUOTAS]; | 1486 | struct dquot *transfer_to[MAXQUOTAS]; |
1473 | int cnt, ret = NO_QUOTA, chuid = (iattr->ia_valid & ATTR_UID) && inode->i_uid != iattr->ia_uid, | 1487 | int cnt, ret = QUOTA_OK; |
1474 | chgid = (iattr->ia_valid & ATTR_GID) && inode->i_gid != iattr->ia_gid; | 1488 | int chuid = iattr->ia_valid & ATTR_UID && inode->i_uid != iattr->ia_uid, |
1489 | chgid = iattr->ia_valid & ATTR_GID && inode->i_gid != iattr->ia_gid; | ||
1475 | char warntype_to[MAXQUOTAS]; | 1490 | char warntype_to[MAXQUOTAS]; |
1476 | char warntype_from_inodes[MAXQUOTAS], warntype_from_space[MAXQUOTAS]; | 1491 | char warntype_from_inodes[MAXQUOTAS], warntype_from_space[MAXQUOTAS]; |
1477 | 1492 | ||
@@ -1479,21 +1494,11 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) | |||
1479 | * re-enter the quota code and are already holding the mutex */ | 1494 | * re-enter the quota code and are already holding the mutex */ |
1480 | if (IS_NOQUOTA(inode)) | 1495 | if (IS_NOQUOTA(inode)) |
1481 | return QUOTA_OK; | 1496 | return QUOTA_OK; |
1482 | /* Clear the arrays */ | 1497 | /* Initialize the arrays */ |
1483 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | 1498 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { |
1484 | transfer_to[cnt] = transfer_from[cnt] = NODQUOT; | 1499 | transfer_from[cnt] = NODQUOT; |
1500 | transfer_to[cnt] = NODQUOT; | ||
1485 | warntype_to[cnt] = QUOTA_NL_NOWARN; | 1501 | warntype_to[cnt] = QUOTA_NL_NOWARN; |
1486 | } | ||
1487 | down_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1488 | /* Now recheck reliably when holding dqptr_sem */ | ||
1489 | if (IS_NOQUOTA(inode)) { /* File without quota accounting? */ | ||
1490 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1491 | return QUOTA_OK; | ||
1492 | } | ||
1493 | /* First build the transfer_to list - here we can block on | ||
1494 | * reading/instantiating of dquots. We know that the transaction for | ||
1495 | * us was already started so we don't violate lock ranking here */ | ||
1496 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | ||
1497 | switch (cnt) { | 1502 | switch (cnt) { |
1498 | case USRQUOTA: | 1503 | case USRQUOTA: |
1499 | if (!chuid) | 1504 | if (!chuid) |
@@ -1507,6 +1512,13 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) | |||
1507 | break; | 1512 | break; |
1508 | } | 1513 | } |
1509 | } | 1514 | } |
1515 | |||
1516 | down_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1517 | /* Now recheck reliably when holding dqptr_sem */ | ||
1518 | if (IS_NOQUOTA(inode)) { /* File without quota accounting? */ | ||
1519 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1520 | goto put_all; | ||
1521 | } | ||
1510 | spin_lock(&dq_data_lock); | 1522 | spin_lock(&dq_data_lock); |
1511 | space = inode_get_bytes(inode); | 1523 | space = inode_get_bytes(inode); |
1512 | /* Build the transfer_from list and check the limits */ | 1524 | /* Build the transfer_from list and check the limits */ |
@@ -1517,7 +1529,7 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) | |||
1517 | if (check_idq(transfer_to[cnt], 1, warntype_to + cnt) == | 1529 | if (check_idq(transfer_to[cnt], 1, warntype_to + cnt) == |
1518 | NO_QUOTA || check_bdq(transfer_to[cnt], space, 0, | 1530 | NO_QUOTA || check_bdq(transfer_to[cnt], space, 0, |
1519 | warntype_to + cnt) == NO_QUOTA) | 1531 | warntype_to + cnt) == NO_QUOTA) |
1520 | goto warn_put_all; | 1532 | goto over_quota; |
1521 | } | 1533 | } |
1522 | 1534 | ||
1523 | /* | 1535 | /* |
@@ -1545,28 +1557,37 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) | |||
1545 | 1557 | ||
1546 | inode->i_dquot[cnt] = transfer_to[cnt]; | 1558 | inode->i_dquot[cnt] = transfer_to[cnt]; |
1547 | } | 1559 | } |
1548 | ret = QUOTA_OK; | ||
1549 | warn_put_all: | ||
1550 | spin_unlock(&dq_data_lock); | 1560 | spin_unlock(&dq_data_lock); |
1561 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1562 | |||
1551 | /* Dirtify all the dquots - this can block when journalling */ | 1563 | /* Dirtify all the dquots - this can block when journalling */ |
1552 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | 1564 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { |
1553 | if (transfer_from[cnt]) | 1565 | if (transfer_from[cnt]) |
1554 | mark_dquot_dirty(transfer_from[cnt]); | 1566 | mark_dquot_dirty(transfer_from[cnt]); |
1555 | if (transfer_to[cnt]) | 1567 | if (transfer_to[cnt]) { |
1556 | mark_dquot_dirty(transfer_to[cnt]); | 1568 | mark_dquot_dirty(transfer_to[cnt]); |
1569 | /* The reference we got is transferred to the inode */ | ||
1570 | transfer_to[cnt] = NODQUOT; | ||
1571 | } | ||
1557 | } | 1572 | } |
1573 | warn_put_all: | ||
1558 | flush_warnings(transfer_to, warntype_to); | 1574 | flush_warnings(transfer_to, warntype_to); |
1559 | flush_warnings(transfer_from, warntype_from_inodes); | 1575 | flush_warnings(transfer_from, warntype_from_inodes); |
1560 | flush_warnings(transfer_from, warntype_from_space); | 1576 | flush_warnings(transfer_from, warntype_from_space); |
1561 | 1577 | put_all: | |
1562 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | 1578 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { |
1563 | if (ret == QUOTA_OK && transfer_from[cnt] != NODQUOT) | 1579 | dqput(transfer_from[cnt]); |
1564 | dqput(transfer_from[cnt]); | 1580 | dqput(transfer_to[cnt]); |
1565 | if (ret == NO_QUOTA && transfer_to[cnt] != NODQUOT) | ||
1566 | dqput(transfer_to[cnt]); | ||
1567 | } | 1581 | } |
1568 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1569 | return ret; | 1582 | return ret; |
1583 | over_quota: | ||
1584 | spin_unlock(&dq_data_lock); | ||
1585 | up_write(&sb_dqopt(inode->i_sb)->dqptr_sem); | ||
1586 | /* Clear dquot pointers we don't want to dqput() */ | ||
1587 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) | ||
1588 | transfer_from[cnt] = NODQUOT; | ||
1589 | ret = NO_QUOTA; | ||
1590 | goto warn_put_all; | ||
1570 | } | 1591 | } |
1571 | 1592 | ||
1572 | /* Wrapper for transferring ownership of an inode */ | 1593 | /* Wrapper for transferring ownership of an inode */ |
@@ -1651,19 +1672,24 @@ int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags) | |||
1651 | continue; | 1672 | continue; |
1652 | 1673 | ||
1653 | if (flags & DQUOT_SUSPENDED) { | 1674 | if (flags & DQUOT_SUSPENDED) { |
1675 | spin_lock(&dq_state_lock); | ||
1654 | dqopt->flags |= | 1676 | dqopt->flags |= |
1655 | dquot_state_flag(DQUOT_SUSPENDED, cnt); | 1677 | dquot_state_flag(DQUOT_SUSPENDED, cnt); |
1678 | spin_unlock(&dq_state_lock); | ||
1656 | } else { | 1679 | } else { |
1680 | spin_lock(&dq_state_lock); | ||
1657 | dqopt->flags &= ~dquot_state_flag(flags, cnt); | 1681 | dqopt->flags &= ~dquot_state_flag(flags, cnt); |
1658 | /* Turning off suspended quotas? */ | 1682 | /* Turning off suspended quotas? */ |
1659 | if (!sb_has_quota_loaded(sb, cnt) && | 1683 | if (!sb_has_quota_loaded(sb, cnt) && |
1660 | sb_has_quota_suspended(sb, cnt)) { | 1684 | sb_has_quota_suspended(sb, cnt)) { |
1661 | dqopt->flags &= ~dquot_state_flag( | 1685 | dqopt->flags &= ~dquot_state_flag( |
1662 | DQUOT_SUSPENDED, cnt); | 1686 | DQUOT_SUSPENDED, cnt); |
1687 | spin_unlock(&dq_state_lock); | ||
1663 | iput(dqopt->files[cnt]); | 1688 | iput(dqopt->files[cnt]); |
1664 | dqopt->files[cnt] = NULL; | 1689 | dqopt->files[cnt] = NULL; |
1665 | continue; | 1690 | continue; |
1666 | } | 1691 | } |
1692 | spin_unlock(&dq_state_lock); | ||
1667 | } | 1693 | } |
1668 | 1694 | ||
1669 | /* We still have to keep quota loaded? */ | 1695 | /* We still have to keep quota loaded? */ |
@@ -1830,7 +1856,9 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id, | |||
1830 | } | 1856 | } |
1831 | mutex_unlock(&dqopt->dqio_mutex); | 1857 | mutex_unlock(&dqopt->dqio_mutex); |
1832 | mutex_unlock(&inode->i_mutex); | 1858 | mutex_unlock(&inode->i_mutex); |
1859 | spin_lock(&dq_state_lock); | ||
1833 | dqopt->flags |= dquot_state_flag(flags, type); | 1860 | dqopt->flags |= dquot_state_flag(flags, type); |
1861 | spin_unlock(&dq_state_lock); | ||
1834 | 1862 | ||
1835 | add_dquot_ref(sb, type); | 1863 | add_dquot_ref(sb, type); |
1836 | mutex_unlock(&dqopt->dqonoff_mutex); | 1864 | mutex_unlock(&dqopt->dqonoff_mutex); |
@@ -1872,9 +1900,11 @@ static int vfs_quota_on_remount(struct super_block *sb, int type) | |||
1872 | } | 1900 | } |
1873 | inode = dqopt->files[type]; | 1901 | inode = dqopt->files[type]; |
1874 | dqopt->files[type] = NULL; | 1902 | dqopt->files[type] = NULL; |
1903 | spin_lock(&dq_state_lock); | ||
1875 | flags = dqopt->flags & dquot_state_flag(DQUOT_USAGE_ENABLED | | 1904 | flags = dqopt->flags & dquot_state_flag(DQUOT_USAGE_ENABLED | |
1876 | DQUOT_LIMITS_ENABLED, type); | 1905 | DQUOT_LIMITS_ENABLED, type); |
1877 | dqopt->flags &= ~dquot_state_flag(DQUOT_STATE_FLAGS, type); | 1906 | dqopt->flags &= ~dquot_state_flag(DQUOT_STATE_FLAGS, type); |
1907 | spin_unlock(&dq_state_lock); | ||
1878 | mutex_unlock(&dqopt->dqonoff_mutex); | 1908 | mutex_unlock(&dqopt->dqonoff_mutex); |
1879 | 1909 | ||
1880 | flags = dquot_generic_flag(flags, type); | 1910 | flags = dquot_generic_flag(flags, type); |
@@ -1952,7 +1982,9 @@ int vfs_quota_enable(struct inode *inode, int type, int format_id, | |||
1952 | ret = -EBUSY; | 1982 | ret = -EBUSY; |
1953 | goto out_lock; | 1983 | goto out_lock; |
1954 | } | 1984 | } |
1985 | spin_lock(&dq_state_lock); | ||
1955 | sb_dqopt(sb)->flags |= dquot_state_flag(flags, type); | 1986 | sb_dqopt(sb)->flags |= dquot_state_flag(flags, type); |
1987 | spin_unlock(&dq_state_lock); | ||
1956 | out_lock: | 1988 | out_lock: |
1957 | mutex_unlock(&dqopt->dqonoff_mutex); | 1989 | mutex_unlock(&dqopt->dqonoff_mutex); |
1958 | return ret; | 1990 | return ret; |
@@ -2039,14 +2071,12 @@ int vfs_get_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *d | |||
2039 | { | 2071 | { |
2040 | struct dquot *dquot; | 2072 | struct dquot *dquot; |
2041 | 2073 | ||
2042 | mutex_lock(&sb_dqopt(sb)->dqonoff_mutex); | 2074 | dquot = dqget(sb, id, type); |
2043 | if (!(dquot = dqget(sb, id, type))) { | 2075 | if (dquot == NODQUOT) |
2044 | mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex); | ||
2045 | return -ESRCH; | 2076 | return -ESRCH; |
2046 | } | ||
2047 | do_get_dqblk(dquot, di); | 2077 | do_get_dqblk(dquot, di); |
2048 | dqput(dquot); | 2078 | dqput(dquot); |
2049 | mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex); | 2079 | |
2050 | return 0; | 2080 | return 0; |
2051 | } | 2081 | } |
2052 | 2082 | ||
@@ -2130,7 +2160,6 @@ int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *d | |||
2130 | struct dquot *dquot; | 2160 | struct dquot *dquot; |
2131 | int rc; | 2161 | int rc; |
2132 | 2162 | ||
2133 | mutex_lock(&sb_dqopt(sb)->dqonoff_mutex); | ||
2134 | dquot = dqget(sb, id, type); | 2163 | dquot = dqget(sb, id, type); |
2135 | if (!dquot) { | 2164 | if (!dquot) { |
2136 | rc = -ESRCH; | 2165 | rc = -ESRCH; |
@@ -2139,7 +2168,6 @@ int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *d | |||
2139 | rc = do_set_dqblk(dquot, di); | 2168 | rc = do_set_dqblk(dquot, di); |
2140 | dqput(dquot); | 2169 | dqput(dquot); |
2141 | out: | 2170 | out: |
2142 | mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex); | ||
2143 | return rc; | 2171 | return rc; |
2144 | } | 2172 | } |
2145 | 2173 | ||
@@ -2370,11 +2398,9 @@ EXPORT_SYMBOL(dquot_release); | |||
2370 | EXPORT_SYMBOL(dquot_mark_dquot_dirty); | 2398 | EXPORT_SYMBOL(dquot_mark_dquot_dirty); |
2371 | EXPORT_SYMBOL(dquot_initialize); | 2399 | EXPORT_SYMBOL(dquot_initialize); |
2372 | EXPORT_SYMBOL(dquot_drop); | 2400 | EXPORT_SYMBOL(dquot_drop); |
2373 | EXPORT_SYMBOL(dquot_drop_locked); | ||
2374 | EXPORT_SYMBOL(vfs_dq_drop); | 2401 | EXPORT_SYMBOL(vfs_dq_drop); |
2375 | EXPORT_SYMBOL(dqget); | 2402 | EXPORT_SYMBOL(dqget); |
2376 | EXPORT_SYMBOL(dqput); | 2403 | EXPORT_SYMBOL(dqput); |
2377 | EXPORT_SYMBOL(dquot_is_cached); | ||
2378 | EXPORT_SYMBOL(dquot_alloc_space); | 2404 | EXPORT_SYMBOL(dquot_alloc_space); |
2379 | EXPORT_SYMBOL(dquot_alloc_inode); | 2405 | EXPORT_SYMBOL(dquot_alloc_inode); |
2380 | EXPORT_SYMBOL(dquot_free_space); | 2406 | EXPORT_SYMBOL(dquot_free_space); |
diff --git a/fs/ecryptfs/Kconfig b/fs/ecryptfs/Kconfig new file mode 100644 index 000000000000..0c754e64232b --- /dev/null +++ b/fs/ecryptfs/Kconfig | |||
@@ -0,0 +1,11 @@ | |||
1 | config ECRYPT_FS | ||
2 | tristate "eCrypt filesystem layer support (EXPERIMENTAL)" | ||
3 | depends on EXPERIMENTAL && KEYS && CRYPTO && NET | ||
4 | help | ||
5 | Encrypted filesystem that operates on the VFS layer. See | ||
6 | <file:Documentation/filesystems/ecryptfs.txt> to learn more about | ||
7 | eCryptfs. Userspace components are required and can be | ||
8 | obtained from <http://ecryptfs.sf.net>. | ||
9 | |||
10 | To compile this file system support as a module, choose M here: the | ||
11 | module will be called ecryptfs. | ||
diff --git a/fs/efs/Kconfig b/fs/efs/Kconfig new file mode 100644 index 000000000000..6ebfc1c207a8 --- /dev/null +++ b/fs/efs/Kconfig | |||
@@ -0,0 +1,14 @@ | |||
1 | config EFS_FS | ||
2 | tristate "EFS file system support (read only) (EXPERIMENTAL)" | ||
3 | depends on BLOCK && EXPERIMENTAL | ||
4 | help | ||
5 | EFS is an older file system used for non-ISO9660 CD-ROMs and hard | ||
6 | disk partitions by SGI's IRIX operating system (IRIX 6.0 and newer | ||
7 | uses the XFS file system for hard disk partitions however). | ||
8 | |||
9 | This implementation only offers read-only access. If you don't know | ||
10 | what all this is about, it's safe to say N. For more information | ||
11 | about EFS see its home page at <http://aeschi.ch.eu.org/efs/>. | ||
12 | |||
13 | To compile the EFS file system support as a module, choose M here: the | ||
14 | module will be called efs. | ||
diff --git a/fs/eventpoll.c b/fs/eventpoll.c index ba2f9ec71192..011b9b8c90c6 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c | |||
@@ -234,8 +234,6 @@ struct ep_pqueue { | |||
234 | /* | 234 | /* |
235 | * Configuration options available inside /proc/sys/fs/epoll/ | 235 | * Configuration options available inside /proc/sys/fs/epoll/ |
236 | */ | 236 | */ |
237 | /* Maximum number of epoll devices, per user */ | ||
238 | static int max_user_instances __read_mostly; | ||
239 | /* Maximum number of epoll watched descriptors, per user */ | 237 | /* Maximum number of epoll watched descriptors, per user */ |
240 | static int max_user_watches __read_mostly; | 238 | static int max_user_watches __read_mostly; |
241 | 239 | ||
@@ -261,14 +259,6 @@ static int zero; | |||
261 | 259 | ||
262 | ctl_table epoll_table[] = { | 260 | ctl_table epoll_table[] = { |
263 | { | 261 | { |
264 | .procname = "max_user_instances", | ||
265 | .data = &max_user_instances, | ||
266 | .maxlen = sizeof(int), | ||
267 | .mode = 0644, | ||
268 | .proc_handler = &proc_dointvec_minmax, | ||
269 | .extra1 = &zero, | ||
270 | }, | ||
271 | { | ||
272 | .procname = "max_user_watches", | 262 | .procname = "max_user_watches", |
273 | .data = &max_user_watches, | 263 | .data = &max_user_watches, |
274 | .maxlen = sizeof(int), | 264 | .maxlen = sizeof(int), |
@@ -491,7 +481,6 @@ static void ep_free(struct eventpoll *ep) | |||
491 | 481 | ||
492 | mutex_unlock(&epmutex); | 482 | mutex_unlock(&epmutex); |
493 | mutex_destroy(&ep->mtx); | 483 | mutex_destroy(&ep->mtx); |
494 | atomic_dec(&ep->user->epoll_devs); | ||
495 | free_uid(ep->user); | 484 | free_uid(ep->user); |
496 | kfree(ep); | 485 | kfree(ep); |
497 | } | 486 | } |
@@ -581,10 +570,6 @@ static int ep_alloc(struct eventpoll **pep) | |||
581 | struct eventpoll *ep; | 570 | struct eventpoll *ep; |
582 | 571 | ||
583 | user = get_current_user(); | 572 | user = get_current_user(); |
584 | error = -EMFILE; | ||
585 | if (unlikely(atomic_read(&user->epoll_devs) >= | ||
586 | max_user_instances)) | ||
587 | goto free_uid; | ||
588 | error = -ENOMEM; | 573 | error = -ENOMEM; |
589 | ep = kzalloc(sizeof(*ep), GFP_KERNEL); | 574 | ep = kzalloc(sizeof(*ep), GFP_KERNEL); |
590 | if (unlikely(!ep)) | 575 | if (unlikely(!ep)) |
@@ -1141,7 +1126,6 @@ SYSCALL_DEFINE1(epoll_create1, int, flags) | |||
1141 | flags & O_CLOEXEC); | 1126 | flags & O_CLOEXEC); |
1142 | if (fd < 0) | 1127 | if (fd < 0) |
1143 | ep_free(ep); | 1128 | ep_free(ep); |
1144 | atomic_inc(&ep->user->epoll_devs); | ||
1145 | 1129 | ||
1146 | error_return: | 1130 | error_return: |
1147 | DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d) = %d\n", | 1131 | DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d) = %d\n", |
@@ -1366,8 +1350,10 @@ static int __init eventpoll_init(void) | |||
1366 | struct sysinfo si; | 1350 | struct sysinfo si; |
1367 | 1351 | ||
1368 | si_meminfo(&si); | 1352 | si_meminfo(&si); |
1369 | max_user_instances = 128; | 1353 | /* |
1370 | max_user_watches = (((si.totalram - si.totalhigh) / 32) << PAGE_SHIFT) / | 1354 | * Allows top 4% of lomem to be allocated for epoll watches (per user). |
1355 | */ | ||
1356 | max_user_watches = (((si.totalram - si.totalhigh) / 25) << PAGE_SHIFT) / | ||
1371 | EP_ITEM_COST; | 1357 | EP_ITEM_COST; |
1372 | 1358 | ||
1373 | /* Initialize the structure used to perform safe poll wait head wake ups */ | 1359 | /* Initialize the structure used to perform safe poll wait head wake ups */ |
diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 69a3d19ca9fd..4db4ffa1edad 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c | |||
@@ -1358,7 +1358,7 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, | |||
1358 | struct fake_dirent *fde; | 1358 | struct fake_dirent *fde; |
1359 | 1359 | ||
1360 | blocksize = dir->i_sb->s_blocksize; | 1360 | blocksize = dir->i_sb->s_blocksize; |
1361 | dxtrace(printk("Creating index\n")); | 1361 | dxtrace(printk(KERN_DEBUG "Creating index: inode %lu\n", dir->i_ino)); |
1362 | retval = ext3_journal_get_write_access(handle, bh); | 1362 | retval = ext3_journal_get_write_access(handle, bh); |
1363 | if (retval) { | 1363 | if (retval) { |
1364 | ext3_std_error(dir->i_sb, retval); | 1364 | ext3_std_error(dir->i_sb, retval); |
@@ -1367,6 +1367,19 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, | |||
1367 | } | 1367 | } |
1368 | root = (struct dx_root *) bh->b_data; | 1368 | root = (struct dx_root *) bh->b_data; |
1369 | 1369 | ||
1370 | /* The 0th block becomes the root, move the dirents out */ | ||
1371 | fde = &root->dotdot; | ||
1372 | de = (struct ext3_dir_entry_2 *)((char *)fde + | ||
1373 | ext3_rec_len_from_disk(fde->rec_len)); | ||
1374 | if ((char *) de >= (((char *) root) + blocksize)) { | ||
1375 | ext3_error(dir->i_sb, __func__, | ||
1376 | "invalid rec_len for '..' in inode %lu", | ||
1377 | dir->i_ino); | ||
1378 | brelse(bh); | ||
1379 | return -EIO; | ||
1380 | } | ||
1381 | len = ((char *) root) + blocksize - (char *) de; | ||
1382 | |||
1370 | bh2 = ext3_append (handle, dir, &block, &retval); | 1383 | bh2 = ext3_append (handle, dir, &block, &retval); |
1371 | if (!(bh2)) { | 1384 | if (!(bh2)) { |
1372 | brelse(bh); | 1385 | brelse(bh); |
@@ -1375,11 +1388,6 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, | |||
1375 | EXT3_I(dir)->i_flags |= EXT3_INDEX_FL; | 1388 | EXT3_I(dir)->i_flags |= EXT3_INDEX_FL; |
1376 | data1 = bh2->b_data; | 1389 | data1 = bh2->b_data; |
1377 | 1390 | ||
1378 | /* The 0th block becomes the root, move the dirents out */ | ||
1379 | fde = &root->dotdot; | ||
1380 | de = (struct ext3_dir_entry_2 *)((char *)fde + | ||
1381 | ext3_rec_len_from_disk(fde->rec_len)); | ||
1382 | len = ((char *) root) + blocksize - (char *) de; | ||
1383 | memcpy (data1, de, len); | 1391 | memcpy (data1, de, len); |
1384 | de = (struct ext3_dir_entry_2 *) data1; | 1392 | de = (struct ext3_dir_entry_2 *) data1; |
1385 | top = data1 + len; | 1393 | top = data1 + len; |
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 6bba06b09dd1..9a50b8052dcf 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c | |||
@@ -684,15 +684,15 @@ ext4_fsblk_t ext4_count_free_blocks(struct super_block *sb) | |||
684 | gdp = ext4_get_group_desc(sb, i, NULL); | 684 | gdp = ext4_get_group_desc(sb, i, NULL); |
685 | if (!gdp) | 685 | if (!gdp) |
686 | continue; | 686 | continue; |
687 | desc_count += le16_to_cpu(gdp->bg_free_blocks_count); | 687 | desc_count += ext4_free_blks_count(sb, gdp); |
688 | brelse(bitmap_bh); | 688 | brelse(bitmap_bh); |
689 | bitmap_bh = ext4_read_block_bitmap(sb, i); | 689 | bitmap_bh = ext4_read_block_bitmap(sb, i); |
690 | if (bitmap_bh == NULL) | 690 | if (bitmap_bh == NULL) |
691 | continue; | 691 | continue; |
692 | 692 | ||
693 | x = ext4_count_free(bitmap_bh, sb->s_blocksize); | 693 | x = ext4_count_free(bitmap_bh, sb->s_blocksize); |
694 | printk(KERN_DEBUG "group %lu: stored = %d, counted = %u\n", | 694 | printk(KERN_DEBUG "group %u: stored = %d, counted = %u\n", |
695 | i, le16_to_cpu(gdp->bg_free_blocks_count), x); | 695 | i, ext4_free_blks_count(sb, gdp), x); |
696 | bitmap_count += x; | 696 | bitmap_count += x; |
697 | } | 697 | } |
698 | brelse(bitmap_bh); | 698 | brelse(bitmap_bh); |
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index c668e4377d76..aafc9eba1c25 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h | |||
@@ -1206,8 +1206,11 @@ static inline void ext4_r_blocks_count_set(struct ext4_super_block *es, | |||
1206 | 1206 | ||
1207 | static inline loff_t ext4_isize(struct ext4_inode *raw_inode) | 1207 | static inline loff_t ext4_isize(struct ext4_inode *raw_inode) |
1208 | { | 1208 | { |
1209 | return ((loff_t)le32_to_cpu(raw_inode->i_size_high) << 32) | | 1209 | if (S_ISREG(le16_to_cpu(raw_inode->i_mode))) |
1210 | le32_to_cpu(raw_inode->i_size_lo); | 1210 | return ((loff_t)le32_to_cpu(raw_inode->i_size_high) << 32) | |
1211 | le32_to_cpu(raw_inode->i_size_lo); | ||
1212 | else | ||
1213 | return (loff_t) le32_to_cpu(raw_inode->i_size_lo); | ||
1211 | } | 1214 | } |
1212 | 1215 | ||
1213 | static inline void ext4_isize_set(struct ext4_inode *raw_inode, loff_t i_size) | 1216 | static inline void ext4_isize_set(struct ext4_inode *raw_inode, loff_t i_size) |
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 54bf0623a9ae..e2eab196875f 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c | |||
@@ -3048,7 +3048,7 @@ retry: | |||
3048 | WARN_ON(ret <= 0); | 3048 | WARN_ON(ret <= 0); |
3049 | printk(KERN_ERR "%s: ext4_ext_get_blocks " | 3049 | printk(KERN_ERR "%s: ext4_ext_get_blocks " |
3050 | "returned error inode#%lu, block=%u, " | 3050 | "returned error inode#%lu, block=%u, " |
3051 | "max_blocks=%lu", __func__, | 3051 | "max_blocks=%u", __func__, |
3052 | inode->i_ino, block, max_blocks); | 3052 | inode->i_ino, block, max_blocks); |
3053 | #endif | 3053 | #endif |
3054 | ext4_mark_inode_dirty(handle, inode); | 3054 | ext4_mark_inode_dirty(handle, inode); |
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index a6444cee0c7e..03ba20be1329 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c | |||
@@ -360,9 +360,9 @@ static int ext4_block_to_path(struct inode *inode, | |||
360 | final = ptrs; | 360 | final = ptrs; |
361 | } else { | 361 | } else { |
362 | ext4_warning(inode->i_sb, "ext4_block_to_path", | 362 | ext4_warning(inode->i_sb, "ext4_block_to_path", |
363 | "block %lu > max", | 363 | "block %lu > max in inode %lu", |
364 | i_block + direct_blocks + | 364 | i_block + direct_blocks + |
365 | indirect_blocks + double_blocks); | 365 | indirect_blocks + double_blocks, inode->i_ino); |
366 | } | 366 | } |
367 | if (boundary) | 367 | if (boundary) |
368 | *boundary = final - 1 - (i_block & (ptrs - 1)); | 368 | *boundary = final - 1 - (i_block & (ptrs - 1)); |
@@ -2821,9 +2821,6 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block) | |||
2821 | filemap_write_and_wait(mapping); | 2821 | filemap_write_and_wait(mapping); |
2822 | } | 2822 | } |
2823 | 2823 | ||
2824 | BUG_ON(!EXT4_JOURNAL(inode) && | ||
2825 | EXT4_I(inode)->i_state & EXT4_STATE_JDATA); | ||
2826 | |||
2827 | if (EXT4_JOURNAL(inode) && EXT4_I(inode)->i_state & EXT4_STATE_JDATA) { | 2824 | if (EXT4_JOURNAL(inode) && EXT4_I(inode)->i_state & EXT4_STATE_JDATA) { |
2828 | /* | 2825 | /* |
2829 | * This is a REALLY heavyweight approach, but the use of | 2826 | * This is a REALLY heavyweight approach, but the use of |
@@ -3622,7 +3619,7 @@ static void ext4_free_data(handle_t *handle, struct inode *inode, | |||
3622 | * block pointed to itself, it would have been detached when | 3619 | * block pointed to itself, it would have been detached when |
3623 | * the block was cleared. Check for this instead of OOPSing. | 3620 | * the block was cleared. Check for this instead of OOPSing. |
3624 | */ | 3621 | */ |
3625 | if (bh2jh(this_bh)) | 3622 | if ((EXT4_JOURNAL(inode) == NULL) || bh2jh(this_bh)) |
3626 | ext4_handle_dirty_metadata(handle, inode, this_bh); | 3623 | ext4_handle_dirty_metadata(handle, inode, this_bh); |
3627 | else | 3624 | else |
3628 | ext4_error(inode->i_sb, __func__, | 3625 | ext4_error(inode->i_sb, __func__, |
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 918aec0c8a11..deba54f6cbed 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c | |||
@@ -3025,7 +3025,7 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, | |||
3025 | goto out_err; | 3025 | goto out_err; |
3026 | 3026 | ||
3027 | ext4_debug("using block group %u(%d)\n", ac->ac_b_ex.fe_group, | 3027 | ext4_debug("using block group %u(%d)\n", ac->ac_b_ex.fe_group, |
3028 | gdp->bg_free_blocks_count); | 3028 | ext4_free_blks_count(sb, gdp)); |
3029 | 3029 | ||
3030 | err = ext4_journal_get_write_access(handle, gdp_bh); | 3030 | err = ext4_journal_get_write_access(handle, gdp_bh); |
3031 | if (err) | 3031 | if (err) |
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index fec0b4c2f5f1..ba702bd7910d 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c | |||
@@ -1368,7 +1368,7 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, | |||
1368 | struct fake_dirent *fde; | 1368 | struct fake_dirent *fde; |
1369 | 1369 | ||
1370 | blocksize = dir->i_sb->s_blocksize; | 1370 | blocksize = dir->i_sb->s_blocksize; |
1371 | dxtrace(printk(KERN_DEBUG "Creating index\n")); | 1371 | dxtrace(printk(KERN_DEBUG "Creating index: inode %lu\n", dir->i_ino)); |
1372 | retval = ext4_journal_get_write_access(handle, bh); | 1372 | retval = ext4_journal_get_write_access(handle, bh); |
1373 | if (retval) { | 1373 | if (retval) { |
1374 | ext4_std_error(dir->i_sb, retval); | 1374 | ext4_std_error(dir->i_sb, retval); |
@@ -1377,6 +1377,20 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, | |||
1377 | } | 1377 | } |
1378 | root = (struct dx_root *) bh->b_data; | 1378 | root = (struct dx_root *) bh->b_data; |
1379 | 1379 | ||
1380 | /* The 0th block becomes the root, move the dirents out */ | ||
1381 | fde = &root->dotdot; | ||
1382 | de = (struct ext4_dir_entry_2 *)((char *)fde + | ||
1383 | ext4_rec_len_from_disk(fde->rec_len)); | ||
1384 | if ((char *) de >= (((char *) root) + blocksize)) { | ||
1385 | ext4_error(dir->i_sb, __func__, | ||
1386 | "invalid rec_len for '..' in inode %lu", | ||
1387 | dir->i_ino); | ||
1388 | brelse(bh); | ||
1389 | return -EIO; | ||
1390 | } | ||
1391 | len = ((char *) root) + blocksize - (char *) de; | ||
1392 | |||
1393 | /* Allocate new block for the 0th block's dirents */ | ||
1380 | bh2 = ext4_append(handle, dir, &block, &retval); | 1394 | bh2 = ext4_append(handle, dir, &block, &retval); |
1381 | if (!(bh2)) { | 1395 | if (!(bh2)) { |
1382 | brelse(bh); | 1396 | brelse(bh); |
@@ -1385,11 +1399,6 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, | |||
1385 | EXT4_I(dir)->i_flags |= EXT4_INDEX_FL; | 1399 | EXT4_I(dir)->i_flags |= EXT4_INDEX_FL; |
1386 | data1 = bh2->b_data; | 1400 | data1 = bh2->b_data; |
1387 | 1401 | ||
1388 | /* The 0th block becomes the root, move the dirents out */ | ||
1389 | fde = &root->dotdot; | ||
1390 | de = (struct ext4_dir_entry_2 *)((char *)fde + | ||
1391 | ext4_rec_len_from_disk(fde->rec_len)); | ||
1392 | len = ((char *) root) + blocksize - (char *) de; | ||
1393 | memcpy (data1, de, len); | 1402 | memcpy (data1, de, len); |
1394 | de = (struct ext4_dir_entry_2 *) data1; | 1403 | de = (struct ext4_dir_entry_2 *) data1; |
1395 | top = data1 + len; | 1404 | top = data1 + len; |
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index c328be5d6885..c06886abd658 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c | |||
@@ -861,12 +861,13 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) | |||
861 | gdp = (struct ext4_group_desc *)((char *)primary->b_data + | 861 | gdp = (struct ext4_group_desc *)((char *)primary->b_data + |
862 | gdb_off * EXT4_DESC_SIZE(sb)); | 862 | gdb_off * EXT4_DESC_SIZE(sb)); |
863 | 863 | ||
864 | memset(gdp, 0, EXT4_DESC_SIZE(sb)); | ||
864 | ext4_block_bitmap_set(sb, gdp, input->block_bitmap); /* LV FIXME */ | 865 | ext4_block_bitmap_set(sb, gdp, input->block_bitmap); /* LV FIXME */ |
865 | ext4_inode_bitmap_set(sb, gdp, input->inode_bitmap); /* LV FIXME */ | 866 | ext4_inode_bitmap_set(sb, gdp, input->inode_bitmap); /* LV FIXME */ |
866 | ext4_inode_table_set(sb, gdp, input->inode_table); /* LV FIXME */ | 867 | ext4_inode_table_set(sb, gdp, input->inode_table); /* LV FIXME */ |
867 | ext4_free_blks_set(sb, gdp, input->free_blocks_count); | 868 | ext4_free_blks_set(sb, gdp, input->free_blocks_count); |
868 | ext4_free_inodes_set(sb, gdp, EXT4_INODES_PER_GROUP(sb)); | 869 | ext4_free_inodes_set(sb, gdp, EXT4_INODES_PER_GROUP(sb)); |
869 | gdp->bg_flags |= cpu_to_le16(EXT4_BG_INODE_ZEROED); | 870 | gdp->bg_flags = cpu_to_le16(EXT4_BG_INODE_ZEROED); |
870 | gdp->bg_checksum = ext4_group_desc_csum(sbi, input->group, gdp); | 871 | gdp->bg_checksum = ext4_group_desc_csum(sbi, input->group, gdp); |
871 | 872 | ||
872 | /* | 873 | /* |
diff --git a/fs/fat/Kconfig b/fs/fat/Kconfig new file mode 100644 index 000000000000..d0a69ff25375 --- /dev/null +++ b/fs/fat/Kconfig | |||
@@ -0,0 +1,97 @@ | |||
1 | config FAT_FS | ||
2 | tristate | ||
3 | select NLS | ||
4 | help | ||
5 | If you want to use one of the FAT-based file systems (the MS-DOS and | ||
6 | VFAT (Windows 95) file systems), then you must say Y or M here | ||
7 | to include FAT support. You will then be able to mount partitions or | ||
8 | diskettes with FAT-based file systems and transparently access the | ||
9 | files on them, i.e. MSDOS files will look and behave just like all | ||
10 | other Unix files. | ||
11 | |||
12 | This FAT support is not a file system in itself, it only provides | ||
13 | the foundation for the other file systems. You will have to say Y or | ||
14 | M to at least one of "MSDOS fs support" or "VFAT fs support" in | ||
15 | order to make use of it. | ||
16 | |||
17 | Another way to read and write MSDOS floppies and hard drive | ||
18 | partitions from within Linux (but not transparently) is with the | ||
19 | mtools ("man mtools") program suite. You don't need to say Y here in | ||
20 | order to do that. | ||
21 | |||
22 | If you need to move large files on floppies between a DOS and a | ||
23 | Linux box, say Y here, mount the floppy under Linux with an MSDOS | ||
24 | file system and use GNU tar's M option. GNU tar is a program | ||
25 | available for Unix and DOS ("man tar" or "info tar"). | ||
26 | |||
27 | The FAT support will enlarge your kernel by about 37 KB. If unsure, | ||
28 | say Y. | ||
29 | |||
30 | To compile this as a module, choose M here: the module will be called | ||
31 | fat. Note that if you compile the FAT support as a module, you | ||
32 | cannot compile any of the FAT-based file systems into the kernel | ||
33 | -- they will have to be modules as well. | ||
34 | |||
35 | config MSDOS_FS | ||
36 | tristate "MSDOS fs support" | ||
37 | select FAT_FS | ||
38 | help | ||
39 | This allows you to mount MSDOS partitions of your hard drive (unless | ||
40 | they are compressed; to access compressed MSDOS partitions under | ||
41 | Linux, you can either use the DOS emulator DOSEMU, described in the | ||
42 | DOSEMU-HOWTO, available from | ||
43 | <http://www.tldp.org/docs.html#howto>, or try dmsdosfs in | ||
44 | <ftp://ibiblio.org/pub/Linux/system/filesystems/dosfs/>. If you | ||
45 | intend to use dosemu with a non-compressed MSDOS partition, say Y | ||
46 | here) and MSDOS floppies. This means that file access becomes | ||
47 | transparent, i.e. the MSDOS files look and behave just like all | ||
48 | other Unix files. | ||
49 | |||
50 | If you have Windows 95 or Windows NT installed on your MSDOS | ||
51 | partitions, you should use the VFAT file system (say Y to "VFAT fs | ||
52 | support" below), or you will not be able to see the long filenames | ||
53 | generated by Windows 95 / Windows NT. | ||
54 | |||
55 | This option will enlarge your kernel by about 7 KB. If unsure, | ||
56 | answer Y. This will only work if you said Y to "DOS FAT fs support" | ||
57 | as well. To compile this as a module, choose M here: the module will | ||
58 | be called msdos. | ||
59 | |||
60 | config VFAT_FS | ||
61 | tristate "VFAT (Windows-95) fs support" | ||
62 | select FAT_FS | ||
63 | help | ||
64 | This option provides support for normal Windows file systems with | ||
65 | long filenames. That includes non-compressed FAT-based file systems | ||
66 | used by Windows 95, Windows 98, Windows NT 4.0, and the Unix | ||
67 | programs from the mtools package. | ||
68 | |||
69 | The VFAT support enlarges your kernel by about 10 KB and it only | ||
70 | works if you said Y to the "DOS FAT fs support" above. Please read | ||
71 | the file <file:Documentation/filesystems/vfat.txt> for details. If | ||
72 | unsure, say Y. | ||
73 | |||
74 | To compile this as a module, choose M here: the module will be called | ||
75 | vfat. | ||
76 | |||
77 | config FAT_DEFAULT_CODEPAGE | ||
78 | int "Default codepage for FAT" | ||
79 | depends on MSDOS_FS || VFAT_FS | ||
80 | default 437 | ||
81 | help | ||
82 | This option should be set to the codepage of your FAT filesystems. | ||
83 | It can be overridden with the "codepage" mount option. | ||
84 | See <file:Documentation/filesystems/vfat.txt> for more information. | ||
85 | |||
86 | config FAT_DEFAULT_IOCHARSET | ||
87 | string "Default iocharset for FAT" | ||
88 | depends on VFAT_FS | ||
89 | default "iso8859-1" | ||
90 | help | ||
91 | Set this to the default input/output character set you'd | ||
92 | like FAT to use. It should probably match the character set | ||
93 | that most of your FAT filesystems use, and can be overridden | ||
94 | with the "iocharset" mount option for FAT filesystems. | ||
95 | Note that "utf8" is not recommended for FAT filesystems. | ||
96 | If unsure, you shouldn't set "utf8" here. | ||
97 | See <file:Documentation/filesystems/vfat.txt> for more information. | ||
diff --git a/fs/freevxfs/Kconfig b/fs/freevxfs/Kconfig new file mode 100644 index 000000000000..8dc1cd5c1efe --- /dev/null +++ b/fs/freevxfs/Kconfig | |||
@@ -0,0 +1,16 @@ | |||
1 | config VXFS_FS | ||
2 | tristate "FreeVxFS file system support (VERITAS VxFS(TM) compatible)" | ||
3 | depends on BLOCK | ||
4 | help | ||
5 | FreeVxFS is a file system driver that support the VERITAS VxFS(TM) | ||
6 | file system format. VERITAS VxFS(TM) is the standard file system | ||
7 | of SCO UnixWare (and possibly others) and optionally available | ||
8 | for Sunsoft Solaris, HP-UX and many other operating systems. | ||
9 | Currently only readonly access is supported. | ||
10 | |||
11 | NOTE: the file system type as used by mount(1), mount(2) and | ||
12 | fstab(5) is 'vxfs' as it describes the file system format, not | ||
13 | the actual driver. | ||
14 | |||
15 | To compile this as a module, choose M here: the module will be | ||
16 | called freevxfs. If unsure, say N. | ||
diff --git a/fs/fuse/Kconfig b/fs/fuse/Kconfig new file mode 100644 index 000000000000..0cf160a94eda --- /dev/null +++ b/fs/fuse/Kconfig | |||
@@ -0,0 +1,15 @@ | |||
1 | config FUSE_FS | ||
2 | tristate "FUSE (Filesystem in Userspace) support" | ||
3 | help | ||
4 | With FUSE it is possible to implement a fully functional filesystem | ||
5 | in a userspace program. | ||
6 | |||
7 | There's also companion library: libfuse. This library along with | ||
8 | utilities is available from the FUSE homepage: | ||
9 | <http://fuse.sourceforge.net/> | ||
10 | |||
11 | See <file:Documentation/filesystems/fuse.txt> for more information. | ||
12 | See <file:Documentation/Changes> for needed library/utility version. | ||
13 | |||
14 | If you want to develop a userspace FS, or if you want to use | ||
15 | a filesystem based on FUSE, answer Y or M. | ||
diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index e0c7ada08a1f..ba76b68c52ff 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c | |||
@@ -281,7 +281,8 @@ __releases(&fc->lock) | |||
281 | fc->blocked = 0; | 281 | fc->blocked = 0; |
282 | wake_up_all(&fc->blocked_waitq); | 282 | wake_up_all(&fc->blocked_waitq); |
283 | } | 283 | } |
284 | if (fc->num_background == FUSE_CONGESTION_THRESHOLD) { | 284 | if (fc->num_background == FUSE_CONGESTION_THRESHOLD && |
285 | fc->connected) { | ||
285 | clear_bdi_congested(&fc->bdi, READ); | 286 | clear_bdi_congested(&fc->bdi, READ); |
286 | clear_bdi_congested(&fc->bdi, WRITE); | 287 | clear_bdi_congested(&fc->bdi, WRITE); |
287 | } | 288 | } |
@@ -825,16 +826,21 @@ static int fuse_notify_poll(struct fuse_conn *fc, unsigned int size, | |||
825 | struct fuse_copy_state *cs) | 826 | struct fuse_copy_state *cs) |
826 | { | 827 | { |
827 | struct fuse_notify_poll_wakeup_out outarg; | 828 | struct fuse_notify_poll_wakeup_out outarg; |
828 | int err; | 829 | int err = -EINVAL; |
829 | 830 | ||
830 | if (size != sizeof(outarg)) | 831 | if (size != sizeof(outarg)) |
831 | return -EINVAL; | 832 | goto err; |
832 | 833 | ||
833 | err = fuse_copy_one(cs, &outarg, sizeof(outarg)); | 834 | err = fuse_copy_one(cs, &outarg, sizeof(outarg)); |
834 | if (err) | 835 | if (err) |
835 | return err; | 836 | goto err; |
836 | 837 | ||
838 | fuse_copy_finish(cs); | ||
837 | return fuse_notify_poll_wakeup(fc, &outarg); | 839 | return fuse_notify_poll_wakeup(fc, &outarg); |
840 | |||
841 | err: | ||
842 | fuse_copy_finish(cs); | ||
843 | return err; | ||
838 | } | 844 | } |
839 | 845 | ||
840 | static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, | 846 | static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, |
@@ -845,6 +851,7 @@ static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, | |||
845 | return fuse_notify_poll(fc, size, cs); | 851 | return fuse_notify_poll(fc, size, cs); |
846 | 852 | ||
847 | default: | 853 | default: |
854 | fuse_copy_finish(cs); | ||
848 | return -EINVAL; | 855 | return -EINVAL; |
849 | } | 856 | } |
850 | } | 857 | } |
@@ -923,7 +930,6 @@ static ssize_t fuse_dev_write(struct kiocb *iocb, const struct iovec *iov, | |||
923 | */ | 930 | */ |
924 | if (!oh.unique) { | 931 | if (!oh.unique) { |
925 | err = fuse_notify(fc, oh.error, nbytes - sizeof(oh), &cs); | 932 | err = fuse_notify(fc, oh.error, nbytes - sizeof(oh), &cs); |
926 | fuse_copy_finish(&cs); | ||
927 | return err ? err : nbytes; | 933 | return err ? err : nbytes; |
928 | } | 934 | } |
929 | 935 | ||
diff --git a/fs/fuse/file.c b/fs/fuse/file.c index e8162646a9b5..d9fdb7cec538 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c | |||
@@ -54,7 +54,7 @@ struct fuse_file *fuse_file_alloc(struct fuse_conn *fc) | |||
54 | ff->reserved_req = fuse_request_alloc(); | 54 | ff->reserved_req = fuse_request_alloc(); |
55 | if (!ff->reserved_req) { | 55 | if (!ff->reserved_req) { |
56 | kfree(ff); | 56 | kfree(ff); |
57 | ff = NULL; | 57 | return NULL; |
58 | } else { | 58 | } else { |
59 | INIT_LIST_HEAD(&ff->write_entry); | 59 | INIT_LIST_HEAD(&ff->write_entry); |
60 | atomic_set(&ff->count, 0); | 60 | atomic_set(&ff->count, 0); |
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 47c96fdca1ac..459b73dd45e1 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c | |||
@@ -292,6 +292,7 @@ static void fuse_put_super(struct super_block *sb) | |||
292 | list_del(&fc->entry); | 292 | list_del(&fc->entry); |
293 | fuse_ctl_remove_conn(fc); | 293 | fuse_ctl_remove_conn(fc); |
294 | mutex_unlock(&fuse_mutex); | 294 | mutex_unlock(&fuse_mutex); |
295 | bdi_destroy(&fc->bdi); | ||
295 | fuse_conn_put(fc); | 296 | fuse_conn_put(fc); |
296 | } | 297 | } |
297 | 298 | ||
@@ -532,7 +533,6 @@ void fuse_conn_put(struct fuse_conn *fc) | |||
532 | if (fc->destroy_req) | 533 | if (fc->destroy_req) |
533 | fuse_request_free(fc->destroy_req); | 534 | fuse_request_free(fc->destroy_req); |
534 | mutex_destroy(&fc->inst_mutex); | 535 | mutex_destroy(&fc->inst_mutex); |
535 | bdi_destroy(&fc->bdi); | ||
536 | fc->release(fc); | 536 | fc->release(fc); |
537 | } | 537 | } |
538 | } | 538 | } |
@@ -805,16 +805,18 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) | |||
805 | int err; | 805 | int err; |
806 | int is_bdev = sb->s_bdev != NULL; | 806 | int is_bdev = sb->s_bdev != NULL; |
807 | 807 | ||
808 | err = -EINVAL; | ||
808 | if (sb->s_flags & MS_MANDLOCK) | 809 | if (sb->s_flags & MS_MANDLOCK) |
809 | return -EINVAL; | 810 | goto err; |
810 | 811 | ||
811 | if (!parse_fuse_opt((char *) data, &d, is_bdev)) | 812 | if (!parse_fuse_opt((char *) data, &d, is_bdev)) |
812 | return -EINVAL; | 813 | goto err; |
813 | 814 | ||
814 | if (is_bdev) { | 815 | if (is_bdev) { |
815 | #ifdef CONFIG_BLOCK | 816 | #ifdef CONFIG_BLOCK |
817 | err = -EINVAL; | ||
816 | if (!sb_set_blocksize(sb, d.blksize)) | 818 | if (!sb_set_blocksize(sb, d.blksize)) |
817 | return -EINVAL; | 819 | goto err; |
818 | #endif | 820 | #endif |
819 | } else { | 821 | } else { |
820 | sb->s_blocksize = PAGE_CACHE_SIZE; | 822 | sb->s_blocksize = PAGE_CACHE_SIZE; |
@@ -826,20 +828,22 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) | |||
826 | sb->s_export_op = &fuse_export_operations; | 828 | sb->s_export_op = &fuse_export_operations; |
827 | 829 | ||
828 | file = fget(d.fd); | 830 | file = fget(d.fd); |
831 | err = -EINVAL; | ||
829 | if (!file) | 832 | if (!file) |
830 | return -EINVAL; | 833 | goto err; |
831 | 834 | ||
832 | if (file->f_op != &fuse_dev_operations) | 835 | if (file->f_op != &fuse_dev_operations) |
833 | return -EINVAL; | 836 | goto err_fput; |
834 | 837 | ||
835 | fc = kmalloc(sizeof(*fc), GFP_KERNEL); | 838 | fc = kmalloc(sizeof(*fc), GFP_KERNEL); |
839 | err = -ENOMEM; | ||
836 | if (!fc) | 840 | if (!fc) |
837 | return -ENOMEM; | 841 | goto err_fput; |
838 | 842 | ||
839 | err = fuse_conn_init(fc, sb); | 843 | err = fuse_conn_init(fc, sb); |
840 | if (err) { | 844 | if (err) { |
841 | kfree(fc); | 845 | kfree(fc); |
842 | return err; | 846 | goto err_fput; |
843 | } | 847 | } |
844 | 848 | ||
845 | fc->release = fuse_free_conn; | 849 | fc->release = fuse_free_conn; |
@@ -854,12 +858,12 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) | |||
854 | err = -ENOMEM; | 858 | err = -ENOMEM; |
855 | root = fuse_get_root_inode(sb, d.rootmode); | 859 | root = fuse_get_root_inode(sb, d.rootmode); |
856 | if (!root) | 860 | if (!root) |
857 | goto err; | 861 | goto err_put_conn; |
858 | 862 | ||
859 | root_dentry = d_alloc_root(root); | 863 | root_dentry = d_alloc_root(root); |
860 | if (!root_dentry) { | 864 | if (!root_dentry) { |
861 | iput(root); | 865 | iput(root); |
862 | goto err; | 866 | goto err_put_conn; |
863 | } | 867 | } |
864 | 868 | ||
865 | init_req = fuse_request_alloc(); | 869 | init_req = fuse_request_alloc(); |
@@ -903,9 +907,11 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) | |||
903 | fuse_request_free(init_req); | 907 | fuse_request_free(init_req); |
904 | err_put_root: | 908 | err_put_root: |
905 | dput(root_dentry); | 909 | dput(root_dentry); |
906 | err: | 910 | err_put_conn: |
907 | fput(file); | ||
908 | fuse_conn_put(fc); | 911 | fuse_conn_put(fc); |
912 | err_fput: | ||
913 | fput(file); | ||
914 | err: | ||
909 | return err; | 915 | return err; |
910 | } | 916 | } |
911 | 917 | ||
diff --git a/fs/hfs/Kconfig b/fs/hfs/Kconfig new file mode 100644 index 000000000000..b77c5bc20f8a --- /dev/null +++ b/fs/hfs/Kconfig | |||
@@ -0,0 +1,12 @@ | |||
1 | config HFS_FS | ||
2 | tristate "Apple Macintosh file system support (EXPERIMENTAL)" | ||
3 | depends on BLOCK && EXPERIMENTAL | ||
4 | select NLS | ||
5 | help | ||
6 | If you say Y here, you will be able to mount Macintosh-formatted | ||
7 | floppy disks and hard drive partitions with full read-write access. | ||
8 | Please read <file:Documentation/filesystems/hfs.txt> to learn about | ||
9 | the available mount options. | ||
10 | |||
11 | To compile this file system support as a module, choose M here: the | ||
12 | module will be called hfs. | ||
diff --git a/fs/hfsplus/Kconfig b/fs/hfsplus/Kconfig new file mode 100644 index 000000000000..a63371815aab --- /dev/null +++ b/fs/hfsplus/Kconfig | |||
@@ -0,0 +1,13 @@ | |||
1 | config HFSPLUS_FS | ||
2 | tristate "Apple Extended HFS file system support" | ||
3 | depends on BLOCK | ||
4 | select NLS | ||
5 | select NLS_UTF8 | ||
6 | help | ||
7 | If you say Y here, you will be able to mount extended format | ||
8 | Macintosh-formatted hard drive partitions with full read-write access. | ||
9 | |||
10 | This file system is often called HFS+ and was introduced with | ||
11 | MacOS 8. It includes all Mac specific filesystem data such as | ||
12 | data forks and creator codes, but it also has several UNIX | ||
13 | style features such as file ownership and permissions. | ||
diff --git a/fs/hpfs/Kconfig b/fs/hpfs/Kconfig new file mode 100644 index 000000000000..56bd15c5bf6c --- /dev/null +++ b/fs/hpfs/Kconfig | |||
@@ -0,0 +1,14 @@ | |||
1 | config HPFS_FS | ||
2 | tristate "OS/2 HPFS file system support" | ||
3 | depends on BLOCK | ||
4 | help | ||
5 | OS/2 is IBM's operating system for PC's, the same as Warp, and HPFS | ||
6 | is the file system used for organizing files on OS/2 hard disk | ||
7 | partitions. Say Y if you want to be able to read files from and | ||
8 | write files to an OS/2 HPFS partition on your hard drive. OS/2 | ||
9 | floppies however are in regular MSDOS format, so you don't need this | ||
10 | option in order to be able to read them. Read | ||
11 | <file:Documentation/filesystems/hpfs.txt>. | ||
12 | |||
13 | To compile this file system support as a module, choose M here: the | ||
14 | module will be called hpfs. If unsure, say N. | ||
diff --git a/fs/isofs/Kconfig b/fs/isofs/Kconfig new file mode 100644 index 000000000000..8ab9878e3671 --- /dev/null +++ b/fs/isofs/Kconfig | |||
@@ -0,0 +1,39 @@ | |||
1 | config ISO9660_FS | ||
2 | tristate "ISO 9660 CDROM file system support" | ||
3 | help | ||
4 | This is the standard file system used on CD-ROMs. It was previously | ||
5 | known as "High Sierra File System" and is called "hsfs" on other | ||
6 | Unix systems. The so-called Rock-Ridge extensions which allow for | ||
7 | long Unix filenames and symbolic links are also supported by this | ||
8 | driver. If you have a CD-ROM drive and want to do more with it than | ||
9 | just listen to audio CDs and watch its LEDs, say Y (and read | ||
10 | <file:Documentation/filesystems/isofs.txt> and the CD-ROM-HOWTO, | ||
11 | available from <http://www.tldp.org/docs.html#howto>), thereby | ||
12 | enlarging your kernel by about 27 KB; otherwise say N. | ||
13 | |||
14 | To compile this file system support as a module, choose M here: the | ||
15 | module will be called isofs. | ||
16 | |||
17 | config JOLIET | ||
18 | bool "Microsoft Joliet CDROM extensions" | ||
19 | depends on ISO9660_FS | ||
20 | select NLS | ||
21 | help | ||
22 | Joliet is a Microsoft extension for the ISO 9660 CD-ROM file system | ||
23 | which allows for long filenames in unicode format (unicode is the | ||
24 | new 16 bit character code, successor to ASCII, which encodes the | ||
25 | characters of almost all languages of the world; see | ||
26 | <http://www.unicode.org/> for more information). Say Y here if you | ||
27 | want to be able to read Joliet CD-ROMs under Linux. | ||
28 | |||
29 | config ZISOFS | ||
30 | bool "Transparent decompression extension" | ||
31 | depends on ISO9660_FS | ||
32 | select ZLIB_INFLATE | ||
33 | help | ||
34 | This is a Linux-specific extension to RockRidge which lets you store | ||
35 | data in compressed form on a CD-ROM and have it transparently | ||
36 | decompressed when the CD-ROM is accessed. See | ||
37 | <http://www.kernel.org/pub/linux/utils/fs/zisofs/> for the tools | ||
38 | necessary to create such a filesystem. Say Y here if you want to be | ||
39 | able to read such compressed CD-ROMs. | ||
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 56675306ed81..eb343008eded 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c | |||
@@ -37,10 +37,10 @@ | |||
37 | #include <linux/proc_fs.h> | 37 | #include <linux/proc_fs.h> |
38 | #include <linux/debugfs.h> | 38 | #include <linux/debugfs.h> |
39 | #include <linux/seq_file.h> | 39 | #include <linux/seq_file.h> |
40 | #include <linux/math64.h> | ||
40 | 41 | ||
41 | #include <asm/uaccess.h> | 42 | #include <asm/uaccess.h> |
42 | #include <asm/page.h> | 43 | #include <asm/page.h> |
43 | #include <asm/div64.h> | ||
44 | 44 | ||
45 | EXPORT_SYMBOL(jbd2_journal_start); | 45 | EXPORT_SYMBOL(jbd2_journal_start); |
46 | EXPORT_SYMBOL(jbd2_journal_restart); | 46 | EXPORT_SYMBOL(jbd2_journal_restart); |
@@ -846,8 +846,8 @@ static int jbd2_seq_info_show(struct seq_file *seq, void *v) | |||
846 | jiffies_to_msecs(s->stats->u.run.rs_flushing / s->stats->ts_tid)); | 846 | jiffies_to_msecs(s->stats->u.run.rs_flushing / s->stats->ts_tid)); |
847 | seq_printf(seq, " %ums logging transaction\n", | 847 | seq_printf(seq, " %ums logging transaction\n", |
848 | jiffies_to_msecs(s->stats->u.run.rs_logging / s->stats->ts_tid)); | 848 | jiffies_to_msecs(s->stats->u.run.rs_logging / s->stats->ts_tid)); |
849 | seq_printf(seq, " %luus average transaction commit time\n", | 849 | seq_printf(seq, " %lluus average transaction commit time\n", |
850 | do_div(s->journal->j_average_commit_time, 1000)); | 850 | div_u64(s->journal->j_average_commit_time, 1000)); |
851 | seq_printf(seq, " %lu handles per transaction\n", | 851 | seq_printf(seq, " %lu handles per transaction\n", |
852 | s->stats->u.run.rs_handle_count / s->stats->ts_tid); | 852 | s->stats->u.run.rs_handle_count / s->stats->ts_tid); |
853 | seq_printf(seq, " %lu blocks per transaction\n", | 853 | seq_printf(seq, " %lu blocks per transaction\n", |
diff --git a/fs/jfs/Kconfig b/fs/jfs/Kconfig new file mode 100644 index 000000000000..9ff619a6f9cc --- /dev/null +++ b/fs/jfs/Kconfig | |||
@@ -0,0 +1,49 @@ | |||
1 | config JFS_FS | ||
2 | tristate "JFS filesystem support" | ||
3 | select NLS | ||
4 | help | ||
5 | This is a port of IBM's Journaled Filesystem . More information is | ||
6 | available in the file <file:Documentation/filesystems/jfs.txt>. | ||
7 | |||
8 | If you do not intend to use the JFS filesystem, say N. | ||
9 | |||
10 | config JFS_POSIX_ACL | ||
11 | bool "JFS POSIX Access Control Lists" | ||
12 | depends on JFS_FS | ||
13 | select FS_POSIX_ACL | ||
14 | help | ||
15 | Posix Access Control Lists (ACLs) support permissions for users and | ||
16 | groups beyond the owner/group/world scheme. | ||
17 | |||
18 | To learn more about Access Control Lists, visit the Posix ACLs for | ||
19 | Linux website <http://acl.bestbits.at/>. | ||
20 | |||
21 | If you don't know what Access Control Lists are, say N | ||
22 | |||
23 | config JFS_SECURITY | ||
24 | bool "JFS Security Labels" | ||
25 | depends on JFS_FS | ||
26 | help | ||
27 | Security labels support alternative access control models | ||
28 | implemented by security modules like SELinux. This option | ||
29 | enables an extended attribute handler for file security | ||
30 | labels in the jfs filesystem. | ||
31 | |||
32 | If you are not using a security module that requires using | ||
33 | extended attributes for file security labels, say N. | ||
34 | |||
35 | config JFS_DEBUG | ||
36 | bool "JFS debugging" | ||
37 | depends on JFS_FS | ||
38 | help | ||
39 | If you are experiencing any problems with the JFS filesystem, say | ||
40 | Y here. This will result in additional debugging messages to be | ||
41 | written to the system log. Under normal circumstances, this | ||
42 | results in very little overhead. | ||
43 | |||
44 | config JFS_STATISTICS | ||
45 | bool "JFS statistics" | ||
46 | depends on JFS_FS | ||
47 | help | ||
48 | Enabling this option will cause statistics from the JFS file system | ||
49 | to be made available to the user in the /proc/fs/jfs/ directory. | ||
diff --git a/fs/minix/Kconfig b/fs/minix/Kconfig new file mode 100644 index 000000000000..0fd7ca994264 --- /dev/null +++ b/fs/minix/Kconfig | |||
@@ -0,0 +1,17 @@ | |||
1 | config MINIX_FS | ||
2 | tristate "Minix file system support" | ||
3 | depends on BLOCK | ||
4 | help | ||
5 | Minix is a simple operating system used in many classes about OS's. | ||
6 | The minix file system (method to organize files on a hard disk | ||
7 | partition or a floppy disk) was the original file system for Linux, | ||
8 | but has been superseded by the second extended file system ext2fs. | ||
9 | You don't want to use the minix file system on your hard disk | ||
10 | because of certain built-in restrictions, but it is sometimes found | ||
11 | on older Linux floppy disks. This option will enlarge your kernel | ||
12 | by about 28 KB. If unsure, say N. | ||
13 | |||
14 | To compile this file system support as a module, choose M here: the | ||
15 | module will be called minix. Note that the file system of your root | ||
16 | partition (the one containing the directory /) cannot be compiled as | ||
17 | a module. | ||
diff --git a/fs/ncpfs/Kconfig b/fs/ncpfs/Kconfig index 142808427b25..c931cf22a1f6 100644 --- a/fs/ncpfs/Kconfig +++ b/fs/ncpfs/Kconfig | |||
@@ -1,6 +1,27 @@ | |||
1 | # | 1 | # |
2 | # NCP Filesystem configuration | 2 | # NCP Filesystem configuration |
3 | # | 3 | # |
4 | config NCP_FS | ||
5 | tristate "NCP file system support (to mount NetWare volumes)" | ||
6 | depends on IPX!=n || INET | ||
7 | help | ||
8 | NCP (NetWare Core Protocol) is a protocol that runs over IPX and is | ||
9 | used by Novell NetWare clients to talk to file servers. It is to | ||
10 | IPX what NFS is to TCP/IP, if that helps. Saying Y here allows you | ||
11 | to mount NetWare file server volumes and to access them just like | ||
12 | any other Unix directory. For details, please read the file | ||
13 | <file:Documentation/filesystems/ncpfs.txt> in the kernel source and | ||
14 | the IPX-HOWTO from <http://www.tldp.org/docs.html#howto>. | ||
15 | |||
16 | You do not have to say Y here if you want your Linux box to act as a | ||
17 | file *server* for Novell NetWare clients. | ||
18 | |||
19 | General information about how to connect Linux, Windows machines and | ||
20 | Macs is on the WWW at <http://www.eats.com/linux_mac_win.html>. | ||
21 | |||
22 | To compile this as a module, choose M here: the module will be called | ||
23 | ncpfs. Say N unless you are connected to a Novell network. | ||
24 | |||
4 | config NCPFS_PACKET_SIGNING | 25 | config NCPFS_PACKET_SIGNING |
5 | bool "Packet signatures" | 26 | bool "Packet signatures" |
6 | depends on NCP_FS | 27 | depends on NCP_FS |
diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig new file mode 100644 index 000000000000..36fe20d6eba2 --- /dev/null +++ b/fs/nfs/Kconfig | |||
@@ -0,0 +1,86 @@ | |||
1 | config NFS_FS | ||
2 | tristate "NFS client support" | ||
3 | depends on INET | ||
4 | select LOCKD | ||
5 | select SUNRPC | ||
6 | select NFS_ACL_SUPPORT if NFS_V3_ACL | ||
7 | help | ||
8 | Choose Y here if you want to access files residing on other | ||
9 | computers using Sun's Network File System protocol. To compile | ||
10 | this file system support as a module, choose M here: the module | ||
11 | will be called nfs. | ||
12 | |||
13 | To mount file systems exported by NFS servers, you also need to | ||
14 | install the user space mount.nfs command which can be found in | ||
15 | the Linux nfs-utils package, available from http://linux-nfs.org/. | ||
16 | Information about using the mount command is available in the | ||
17 | mount(8) man page. More detail about the Linux NFS client | ||
18 | implementation is available via the nfs(5) man page. | ||
19 | |||
20 | Below you can choose which versions of the NFS protocol are | ||
21 | available in the kernel to mount NFS servers. Support for NFS | ||
22 | version 2 (RFC 1094) is always available when NFS_FS is selected. | ||
23 | |||
24 | To configure a system which mounts its root file system via NFS | ||
25 | at boot time, say Y here, select "Kernel level IP | ||
26 | autoconfiguration" in the NETWORK menu, and select "Root file | ||
27 | system on NFS" below. You cannot compile this file system as a | ||
28 | module in this case. | ||
29 | |||
30 | If unsure, say N. | ||
31 | |||
32 | config NFS_V3 | ||
33 | bool "NFS client support for NFS version 3" | ||
34 | depends on NFS_FS | ||
35 | help | ||
36 | This option enables support for version 3 of the NFS protocol | ||
37 | (RFC 1813) in the kernel's NFS client. | ||
38 | |||
39 | If unsure, say Y. | ||
40 | |||
41 | config NFS_V3_ACL | ||
42 | bool "NFS client support for the NFSv3 ACL protocol extension" | ||
43 | depends on NFS_V3 | ||
44 | help | ||
45 | Some NFS servers support an auxiliary NFSv3 ACL protocol that | ||
46 | Sun added to Solaris but never became an official part of the | ||
47 | NFS version 3 protocol. This protocol extension allows | ||
48 | applications on NFS clients to manipulate POSIX Access Control | ||
49 | Lists on files residing on NFS servers. NFS servers enforce | ||
50 | ACLs on local files whether this protocol is available or not. | ||
51 | |||
52 | Choose Y here if your NFS server supports the Solaris NFSv3 ACL | ||
53 | protocol extension and you want your NFS client to allow | ||
54 | applications to access and modify ACLs on files on the server. | ||
55 | |||
56 | Most NFS servers don't support the Solaris NFSv3 ACL protocol | ||
57 | extension. You can choose N here or specify the "noacl" mount | ||
58 | option to prevent your NFS client from trying to use the NFSv3 | ||
59 | ACL protocol. | ||
60 | |||
61 | If unsure, say N. | ||
62 | |||
63 | config NFS_V4 | ||
64 | bool "NFS client support for NFS version 4 (EXPERIMENTAL)" | ||
65 | depends on NFS_FS && EXPERIMENTAL | ||
66 | select RPCSEC_GSS_KRB5 | ||
67 | help | ||
68 | This option enables support for version 4 of the NFS protocol | ||
69 | (RFC 3530) in the kernel's NFS client. | ||
70 | |||
71 | To mount NFS servers using NFSv4, you also need to install user | ||
72 | space programs which can be found in the Linux nfs-utils package, | ||
73 | available from http://linux-nfs.org/. | ||
74 | |||
75 | If unsure, say N. | ||
76 | |||
77 | config ROOT_NFS | ||
78 | bool "Root file system on NFS" | ||
79 | depends on NFS_FS=y && IP_PNP | ||
80 | help | ||
81 | If you want your system to mount its root file system via NFS, | ||
82 | choose Y here. This is common practice for managing systems | ||
83 | without local permanent storage. For details, read | ||
84 | <file:Documentation/filesystems/nfsroot.txt>. | ||
85 | |||
86 | Most people say N here. | ||
diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig new file mode 100644 index 000000000000..44d7d04dab95 --- /dev/null +++ b/fs/nfsd/Kconfig | |||
@@ -0,0 +1,80 @@ | |||
1 | config NFSD | ||
2 | tristate "NFS server support" | ||
3 | depends on INET | ||
4 | select LOCKD | ||
5 | select SUNRPC | ||
6 | select EXPORTFS | ||
7 | select NFS_ACL_SUPPORT if NFSD_V2_ACL | ||
8 | help | ||
9 | Choose Y here if you want to allow other computers to access | ||
10 | files residing on this system using Sun's Network File System | ||
11 | protocol. To compile the NFS server support as a module, | ||
12 | choose M here: the module will be called nfsd. | ||
13 | |||
14 | You may choose to use a user-space NFS server instead, in which | ||
15 | case you can choose N here. | ||
16 | |||
17 | To export local file systems using NFS, you also need to install | ||
18 | user space programs which can be found in the Linux nfs-utils | ||
19 | package, available from http://linux-nfs.org/. More detail about | ||
20 | the Linux NFS server implementation is available via the | ||
21 | exports(5) man page. | ||
22 | |||
23 | Below you can choose which versions of the NFS protocol are | ||
24 | available to clients mounting the NFS server on this system. | ||
25 | Support for NFS version 2 (RFC 1094) is always available when | ||
26 | CONFIG_NFSD is selected. | ||
27 | |||
28 | If unsure, say N. | ||
29 | |||
30 | config NFSD_V2_ACL | ||
31 | bool | ||
32 | depends on NFSD | ||
33 | |||
34 | config NFSD_V3 | ||
35 | bool "NFS server support for NFS version 3" | ||
36 | depends on NFSD | ||
37 | help | ||
38 | This option enables support in your system's NFS server for | ||
39 | version 3 of the NFS protocol (RFC 1813). | ||
40 | |||
41 | If unsure, say Y. | ||
42 | |||
43 | config NFSD_V3_ACL | ||
44 | bool "NFS server support for the NFSv3 ACL protocol extension" | ||
45 | depends on NFSD_V3 | ||
46 | select NFSD_V2_ACL | ||
47 | help | ||
48 | Solaris NFS servers support an auxiliary NFSv3 ACL protocol that | ||
49 | never became an official part of the NFS version 3 protocol. | ||
50 | This protocol extension allows applications on NFS clients to | ||
51 | manipulate POSIX Access Control Lists on files residing on NFS | ||
52 | servers. NFS servers enforce POSIX ACLs on local files whether | ||
53 | this protocol is available or not. | ||
54 | |||
55 | This option enables support in your system's NFS server for the | ||
56 | NFSv3 ACL protocol extension allowing NFS clients to manipulate | ||
57 | POSIX ACLs on files exported by your system's NFS server. NFS | ||
58 | clients which support the Solaris NFSv3 ACL protocol can then | ||
59 | access and modify ACLs on your NFS server. | ||
60 | |||
61 | To store ACLs on your NFS server, you also need to enable ACL- | ||
62 | related CONFIG options for your local file systems of choice. | ||
63 | |||
64 | If unsure, say N. | ||
65 | |||
66 | config NFSD_V4 | ||
67 | bool "NFS server support for NFS version 4 (EXPERIMENTAL)" | ||
68 | depends on NFSD && PROC_FS && EXPERIMENTAL | ||
69 | select NFSD_V3 | ||
70 | select FS_POSIX_ACL | ||
71 | select RPCSEC_GSS_KRB5 | ||
72 | help | ||
73 | This option enables support in your system's NFS server for | ||
74 | version 4 of the NFS protocol (RFC 3530). | ||
75 | |||
76 | To export files using NFSv4, you need to install additional user | ||
77 | space programs which can be found in the Linux nfs-utils package, | ||
78 | available from http://linux-nfs.org/. | ||
79 | |||
80 | If unsure, say N. | ||
diff --git a/fs/nfsd/auth.c b/fs/nfsd/auth.c index c903e04aa217..5573508f707f 100644 --- a/fs/nfsd/auth.c +++ b/fs/nfsd/auth.c | |||
@@ -49,6 +49,8 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) | |||
49 | new->fsuid = exp->ex_anon_uid; | 49 | new->fsuid = exp->ex_anon_uid; |
50 | new->fsgid = exp->ex_anon_gid; | 50 | new->fsgid = exp->ex_anon_gid; |
51 | gi = groups_alloc(0); | 51 | gi = groups_alloc(0); |
52 | if (!gi) | ||
53 | goto oom; | ||
52 | } else if (flags & NFSEXP_ROOTSQUASH) { | 54 | } else if (flags & NFSEXP_ROOTSQUASH) { |
53 | if (!new->fsuid) | 55 | if (!new->fsuid) |
54 | new->fsuid = exp->ex_anon_uid; | 56 | new->fsuid = exp->ex_anon_uid; |
@@ -85,6 +87,7 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) | |||
85 | new->cap_effective = cap_raise_nfsd_set(new->cap_effective, | 87 | new->cap_effective = cap_raise_nfsd_set(new->cap_effective, |
86 | new->cap_permitted); | 88 | new->cap_permitted); |
87 | put_cred(override_creds(new)); | 89 | put_cred(override_creds(new)); |
90 | put_cred(new); | ||
88 | return 0; | 91 | return 0; |
89 | 92 | ||
90 | oom: | 93 | oom: |
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 88db7d3ec120..b6f60f48e94b 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c | |||
@@ -2871,7 +2871,6 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, | |||
2871 | file_lock.fl_owner = (fl_owner_t)lockt->lt_stateowner; | 2871 | file_lock.fl_owner = (fl_owner_t)lockt->lt_stateowner; |
2872 | file_lock.fl_pid = current->tgid; | 2872 | file_lock.fl_pid = current->tgid; |
2873 | file_lock.fl_flags = FL_POSIX; | 2873 | file_lock.fl_flags = FL_POSIX; |
2874 | file_lock.fl_lmops = &nfsd_posix_mng_ops; | ||
2875 | 2874 | ||
2876 | file_lock.fl_start = lockt->lt_offset; | 2875 | file_lock.fl_start = lockt->lt_offset; |
2877 | file_lock.fl_end = last_byte_offset(lockt->lt_offset, lockt->lt_length); | 2876 | file_lock.fl_end = last_byte_offset(lockt->lt_offset, lockt->lt_length); |
diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index d53a1838d6e8..bed766e435b5 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c | |||
@@ -427,10 +427,61 @@ static unsigned int inotify_poll(struct file *file, poll_table *wait) | |||
427 | return ret; | 427 | return ret; |
428 | } | 428 | } |
429 | 429 | ||
430 | /* | ||
431 | * Get an inotify_kernel_event if one exists and is small | ||
432 | * enough to fit in "count". Return an error pointer if | ||
433 | * not large enough. | ||
434 | * | ||
435 | * Called with the device ev_mutex held. | ||
436 | */ | ||
437 | static struct inotify_kernel_event *get_one_event(struct inotify_device *dev, | ||
438 | size_t count) | ||
439 | { | ||
440 | size_t event_size = sizeof(struct inotify_event); | ||
441 | struct inotify_kernel_event *kevent; | ||
442 | |||
443 | if (list_empty(&dev->events)) | ||
444 | return NULL; | ||
445 | |||
446 | kevent = inotify_dev_get_event(dev); | ||
447 | if (kevent->name) | ||
448 | event_size += kevent->event.len; | ||
449 | |||
450 | if (event_size > count) | ||
451 | return ERR_PTR(-EINVAL); | ||
452 | |||
453 | remove_kevent(dev, kevent); | ||
454 | return kevent; | ||
455 | } | ||
456 | |||
457 | /* | ||
458 | * Copy an event to user space, returning how much we copied. | ||
459 | * | ||
460 | * We already checked that the event size is smaller than the | ||
461 | * buffer we had in "get_one_event()" above. | ||
462 | */ | ||
463 | static ssize_t copy_event_to_user(struct inotify_kernel_event *kevent, | ||
464 | char __user *buf) | ||
465 | { | ||
466 | size_t event_size = sizeof(struct inotify_event); | ||
467 | |||
468 | if (copy_to_user(buf, &kevent->event, event_size)) | ||
469 | return -EFAULT; | ||
470 | |||
471 | if (kevent->name) { | ||
472 | buf += event_size; | ||
473 | |||
474 | if (copy_to_user(buf, kevent->name, kevent->event.len)) | ||
475 | return -EFAULT; | ||
476 | |||
477 | event_size += kevent->event.len; | ||
478 | } | ||
479 | return event_size; | ||
480 | } | ||
481 | |||
430 | static ssize_t inotify_read(struct file *file, char __user *buf, | 482 | static ssize_t inotify_read(struct file *file, char __user *buf, |
431 | size_t count, loff_t *pos) | 483 | size_t count, loff_t *pos) |
432 | { | 484 | { |
433 | size_t event_size = sizeof (struct inotify_event); | ||
434 | struct inotify_device *dev; | 485 | struct inotify_device *dev; |
435 | char __user *start; | 486 | char __user *start; |
436 | int ret; | 487 | int ret; |
@@ -440,81 +491,43 @@ static ssize_t inotify_read(struct file *file, char __user *buf, | |||
440 | dev = file->private_data; | 491 | dev = file->private_data; |
441 | 492 | ||
442 | while (1) { | 493 | while (1) { |
494 | struct inotify_kernel_event *kevent; | ||
443 | 495 | ||
444 | prepare_to_wait(&dev->wq, &wait, TASK_INTERRUPTIBLE); | 496 | prepare_to_wait(&dev->wq, &wait, TASK_INTERRUPTIBLE); |
445 | 497 | ||
446 | mutex_lock(&dev->ev_mutex); | 498 | mutex_lock(&dev->ev_mutex); |
447 | if (!list_empty(&dev->events)) { | 499 | kevent = get_one_event(dev, count); |
448 | ret = 0; | ||
449 | break; | ||
450 | } | ||
451 | mutex_unlock(&dev->ev_mutex); | 500 | mutex_unlock(&dev->ev_mutex); |
452 | 501 | ||
453 | if (file->f_flags & O_NONBLOCK) { | 502 | if (kevent) { |
454 | ret = -EAGAIN; | 503 | ret = PTR_ERR(kevent); |
455 | break; | 504 | if (IS_ERR(kevent)) |
456 | } | 505 | break; |
457 | 506 | ret = copy_event_to_user(kevent, buf); | |
458 | if (signal_pending(current)) { | 507 | free_kevent(kevent); |
459 | ret = -EINTR; | 508 | if (ret < 0) |
460 | break; | 509 | break; |
510 | buf += ret; | ||
511 | count -= ret; | ||
512 | continue; | ||
461 | } | 513 | } |
462 | 514 | ||
463 | schedule(); | 515 | ret = -EAGAIN; |
464 | } | 516 | if (file->f_flags & O_NONBLOCK) |
465 | |||
466 | finish_wait(&dev->wq, &wait); | ||
467 | if (ret) | ||
468 | return ret; | ||
469 | |||
470 | while (1) { | ||
471 | struct inotify_kernel_event *kevent; | ||
472 | |||
473 | ret = buf - start; | ||
474 | if (list_empty(&dev->events)) | ||
475 | break; | 517 | break; |
476 | 518 | ret = -EINTR; | |
477 | kevent = inotify_dev_get_event(dev); | 519 | if (signal_pending(current)) |
478 | if (event_size + kevent->event.len > count) { | ||
479 | if (ret == 0 && count > 0) { | ||
480 | /* | ||
481 | * could not get a single event because we | ||
482 | * didn't have enough buffer space. | ||
483 | */ | ||
484 | ret = -EINVAL; | ||
485 | } | ||
486 | break; | 520 | break; |
487 | } | ||
488 | remove_kevent(dev, kevent); | ||
489 | 521 | ||
490 | /* | 522 | if (start != buf) |
491 | * Must perform the copy_to_user outside the mutex in order | ||
492 | * to avoid a lock order reversal with mmap_sem. | ||
493 | */ | ||
494 | mutex_unlock(&dev->ev_mutex); | ||
495 | |||
496 | if (copy_to_user(buf, &kevent->event, event_size)) { | ||
497 | ret = -EFAULT; | ||
498 | break; | 523 | break; |
499 | } | ||
500 | buf += event_size; | ||
501 | count -= event_size; | ||
502 | |||
503 | if (kevent->name) { | ||
504 | if (copy_to_user(buf, kevent->name, kevent->event.len)){ | ||
505 | ret = -EFAULT; | ||
506 | break; | ||
507 | } | ||
508 | buf += kevent->event.len; | ||
509 | count -= kevent->event.len; | ||
510 | } | ||
511 | |||
512 | free_kevent(kevent); | ||
513 | 524 | ||
514 | mutex_lock(&dev->ev_mutex); | 525 | schedule(); |
515 | } | 526 | } |
516 | mutex_unlock(&dev->ev_mutex); | ||
517 | 527 | ||
528 | finish_wait(&dev->wq, &wait); | ||
529 | if (start != buf && ret != -EFAULT) | ||
530 | ret = buf - start; | ||
518 | return ret; | 531 | return ret; |
519 | } | 532 | } |
520 | 533 | ||
diff --git a/fs/ntfs/Kconfig b/fs/ntfs/Kconfig new file mode 100644 index 000000000000..f5a868cc9152 --- /dev/null +++ b/fs/ntfs/Kconfig | |||
@@ -0,0 +1,78 @@ | |||
1 | config NTFS_FS | ||
2 | tristate "NTFS file system support" | ||
3 | select NLS | ||
4 | help | ||
5 | NTFS is the file system of Microsoft Windows NT, 2000, XP and 2003. | ||
6 | |||
7 | Saying Y or M here enables read support. There is partial, but | ||
8 | safe, write support available. For write support you must also | ||
9 | say Y to "NTFS write support" below. | ||
10 | |||
11 | There are also a number of user-space tools available, called | ||
12 | ntfsprogs. These include ntfsundelete and ntfsresize, that work | ||
13 | without NTFS support enabled in the kernel. | ||
14 | |||
15 | This is a rewrite from scratch of Linux NTFS support and replaced | ||
16 | the old NTFS code starting with Linux 2.5.11. A backport to | ||
17 | the Linux 2.4 kernel series is separately available as a patch | ||
18 | from the project web site. | ||
19 | |||
20 | For more information see <file:Documentation/filesystems/ntfs.txt> | ||
21 | and <http://www.linux-ntfs.org/>. | ||
22 | |||
23 | To compile this file system support as a module, choose M here: the | ||
24 | module will be called ntfs. | ||
25 | |||
26 | If you are not using Windows NT, 2000, XP or 2003 in addition to | ||
27 | Linux on your computer it is safe to say N. | ||
28 | |||
29 | config NTFS_DEBUG | ||
30 | bool "NTFS debugging support" | ||
31 | depends on NTFS_FS | ||
32 | help | ||
33 | If you are experiencing any problems with the NTFS file system, say | ||
34 | Y here. This will result in additional consistency checks to be | ||
35 | performed by the driver as well as additional debugging messages to | ||
36 | be written to the system log. Note that debugging messages are | ||
37 | disabled by default. To enable them, supply the option debug_msgs=1 | ||
38 | at the kernel command line when booting the kernel or as an option | ||
39 | to insmod when loading the ntfs module. Once the driver is active, | ||
40 | you can enable debugging messages by doing (as root): | ||
41 | echo 1 > /proc/sys/fs/ntfs-debug | ||
42 | Replacing the "1" with "0" would disable debug messages. | ||
43 | |||
44 | If you leave debugging messages disabled, this results in little | ||
45 | overhead, but enabling debug messages results in very significant | ||
46 | slowdown of the system. | ||
47 | |||
48 | When reporting bugs, please try to have available a full dump of | ||
49 | debugging messages while the misbehaviour was occurring. | ||
50 | |||
51 | config NTFS_RW | ||
52 | bool "NTFS write support" | ||
53 | depends on NTFS_FS | ||
54 | help | ||
55 | This enables the partial, but safe, write support in the NTFS driver. | ||
56 | |||
57 | The only supported operation is overwriting existing files, without | ||
58 | changing the file length. No file or directory creation, deletion or | ||
59 | renaming is possible. Note only non-resident files can be written to | ||
60 | so you may find that some very small files (<500 bytes or so) cannot | ||
61 | be written to. | ||
62 | |||
63 | While we cannot guarantee that it will not damage any data, we have | ||
64 | so far not received a single report where the driver would have | ||
65 | damaged someones data so we assume it is perfectly safe to use. | ||
66 | |||
67 | Note: While write support is safe in this version (a rewrite from | ||
68 | scratch of the NTFS support), it should be noted that the old NTFS | ||
69 | write support, included in Linux 2.5.10 and before (since 1997), | ||
70 | is not safe. | ||
71 | |||
72 | This is currently useful with TopologiLinux. TopologiLinux is run | ||
73 | on top of any DOS/Microsoft Windows system without partitioning your | ||
74 | hard disk. Unlike other Linux distributions TopologiLinux does not | ||
75 | need its own partition. For more information see | ||
76 | <http://topologi-linux.sourceforge.net/> | ||
77 | |||
78 | It is perfectly safe to say N here. | ||
diff --git a/fs/ocfs2/Kconfig b/fs/ocfs2/Kconfig new file mode 100644 index 000000000000..701b7a3a872e --- /dev/null +++ b/fs/ocfs2/Kconfig | |||
@@ -0,0 +1,85 @@ | |||
1 | config OCFS2_FS | ||
2 | tristate "OCFS2 file system support" | ||
3 | depends on NET && SYSFS | ||
4 | select CONFIGFS_FS | ||
5 | select JBD2 | ||
6 | select CRC32 | ||
7 | select QUOTA | ||
8 | select QUOTA_TREE | ||
9 | help | ||
10 | OCFS2 is a general purpose extent based shared disk cluster file | ||
11 | system with many similarities to ext3. It supports 64 bit inode | ||
12 | numbers, and has automatically extending metadata groups which may | ||
13 | also make it attractive for non-clustered use. | ||
14 | |||
15 | You'll want to install the ocfs2-tools package in order to at least | ||
16 | get "mount.ocfs2". | ||
17 | |||
18 | Project web page: http://oss.oracle.com/projects/ocfs2 | ||
19 | Tools web page: http://oss.oracle.com/projects/ocfs2-tools | ||
20 | OCFS2 mailing lists: http://oss.oracle.com/projects/ocfs2/mailman/ | ||
21 | |||
22 | For more information on OCFS2, see the file | ||
23 | <file:Documentation/filesystems/ocfs2.txt>. | ||
24 | |||
25 | config OCFS2_FS_O2CB | ||
26 | tristate "O2CB Kernelspace Clustering" | ||
27 | depends on OCFS2_FS | ||
28 | default y | ||
29 | help | ||
30 | OCFS2 includes a simple kernelspace clustering package, the OCFS2 | ||
31 | Cluster Base. It only requires a very small userspace component | ||
32 | to configure it. This comes with the standard ocfs2-tools package. | ||
33 | O2CB is limited to maintaining a cluster for OCFS2 file systems. | ||
34 | It cannot manage any other cluster applications. | ||
35 | |||
36 | It is always safe to say Y here, as the clustering method is | ||
37 | run-time selectable. | ||
38 | |||
39 | config OCFS2_FS_USERSPACE_CLUSTER | ||
40 | tristate "OCFS2 Userspace Clustering" | ||
41 | depends on OCFS2_FS && DLM | ||
42 | default y | ||
43 | help | ||
44 | This option will allow OCFS2 to use userspace clustering services | ||
45 | in conjunction with the DLM in fs/dlm. If you are using a | ||
46 | userspace cluster manager, say Y here. | ||
47 | |||
48 | It is safe to say Y, as the clustering method is run-time | ||
49 | selectable. | ||
50 | |||
51 | config OCFS2_FS_STATS | ||
52 | bool "OCFS2 statistics" | ||
53 | depends on OCFS2_FS | ||
54 | default y | ||
55 | help | ||
56 | This option allows some fs statistics to be captured. Enabling | ||
57 | this option may increase the memory consumption. | ||
58 | |||
59 | config OCFS2_DEBUG_MASKLOG | ||
60 | bool "OCFS2 logging support" | ||
61 | depends on OCFS2_FS | ||
62 | default y | ||
63 | help | ||
64 | The ocfs2 filesystem has an extensive logging system. The system | ||
65 | allows selection of events to log via files in /sys/o2cb/logmask/. | ||
66 | This option will enlarge your kernel, but it allows debugging of | ||
67 | ocfs2 filesystem issues. | ||
68 | |||
69 | config OCFS2_DEBUG_FS | ||
70 | bool "OCFS2 expensive checks" | ||
71 | depends on OCFS2_FS | ||
72 | default n | ||
73 | help | ||
74 | This option will enable expensive consistency checks. Enable | ||
75 | this option for debugging only as it is likely to decrease | ||
76 | performance of the filesystem. | ||
77 | |||
78 | config OCFS2_FS_POSIX_ACL | ||
79 | bool "OCFS2 POSIX Access Control Lists" | ||
80 | depends on OCFS2_FS | ||
81 | select FS_POSIX_ACL | ||
82 | default n | ||
83 | help | ||
84 | Posix Access Control Lists (ACLs) support permissions for users and | ||
85 | groups beyond the owner/group/world scheme. | ||
diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 6aff8f2d3e49..f4efa89baee5 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c | |||
@@ -810,171 +810,6 @@ out: | |||
810 | return status; | 810 | return status; |
811 | } | 811 | } |
812 | 812 | ||
813 | /* This is difficult. We have to lock quota inode and start transaction | ||
814 | * in this function but we don't want to take the penalty of exlusive | ||
815 | * quota file lock when we are just going to use cached structures. So | ||
816 | * we just take read lock check whether we have dquot cached and if so, | ||
817 | * we don't have to take the write lock... */ | ||
818 | static int ocfs2_dquot_initialize(struct inode *inode, int type) | ||
819 | { | ||
820 | handle_t *handle = NULL; | ||
821 | int status = 0; | ||
822 | struct super_block *sb = inode->i_sb; | ||
823 | struct ocfs2_mem_dqinfo *oinfo; | ||
824 | int exclusive = 0; | ||
825 | int cnt; | ||
826 | qid_t id; | ||
827 | |||
828 | mlog_entry_void(); | ||
829 | |||
830 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | ||
831 | if (type != -1 && cnt != type) | ||
832 | continue; | ||
833 | if (!sb_has_quota_active(sb, cnt)) | ||
834 | continue; | ||
835 | oinfo = sb_dqinfo(sb, cnt)->dqi_priv; | ||
836 | status = ocfs2_lock_global_qf(oinfo, 0); | ||
837 | if (status < 0) | ||
838 | goto out; | ||
839 | /* This is just a performance optimization not a reliable test. | ||
840 | * Since we hold an inode lock, noone can actually release | ||
841 | * the structure until we are finished with initialization. */ | ||
842 | if (inode->i_dquot[cnt] != NODQUOT) { | ||
843 | ocfs2_unlock_global_qf(oinfo, 0); | ||
844 | continue; | ||
845 | } | ||
846 | /* When we have inode lock, we know that no dquot_release() can | ||
847 | * run and thus we can safely check whether we need to | ||
848 | * read+modify global file to get quota information or whether | ||
849 | * our node already has it. */ | ||
850 | if (cnt == USRQUOTA) | ||
851 | id = inode->i_uid; | ||
852 | else if (cnt == GRPQUOTA) | ||
853 | id = inode->i_gid; | ||
854 | else | ||
855 | BUG(); | ||
856 | /* Obtain exclusion from quota off... */ | ||
857 | down_write(&sb_dqopt(sb)->dqptr_sem); | ||
858 | exclusive = !dquot_is_cached(sb, id, cnt); | ||
859 | up_write(&sb_dqopt(sb)->dqptr_sem); | ||
860 | if (exclusive) { | ||
861 | status = ocfs2_lock_global_qf(oinfo, 1); | ||
862 | if (status < 0) { | ||
863 | exclusive = 0; | ||
864 | mlog_errno(status); | ||
865 | goto out_ilock; | ||
866 | } | ||
867 | handle = ocfs2_start_trans(OCFS2_SB(sb), | ||
868 | ocfs2_calc_qinit_credits(sb, cnt)); | ||
869 | if (IS_ERR(handle)) { | ||
870 | status = PTR_ERR(handle); | ||
871 | mlog_errno(status); | ||
872 | goto out_ilock; | ||
873 | } | ||
874 | } | ||
875 | dquot_initialize(inode, cnt); | ||
876 | if (exclusive) { | ||
877 | ocfs2_commit_trans(OCFS2_SB(sb), handle); | ||
878 | ocfs2_unlock_global_qf(oinfo, 1); | ||
879 | } | ||
880 | ocfs2_unlock_global_qf(oinfo, 0); | ||
881 | } | ||
882 | mlog_exit(0); | ||
883 | return 0; | ||
884 | out_ilock: | ||
885 | if (exclusive) | ||
886 | ocfs2_unlock_global_qf(oinfo, 1); | ||
887 | ocfs2_unlock_global_qf(oinfo, 0); | ||
888 | out: | ||
889 | mlog_exit(status); | ||
890 | return status; | ||
891 | } | ||
892 | |||
893 | static int ocfs2_dquot_drop_slow(struct inode *inode) | ||
894 | { | ||
895 | int status = 0; | ||
896 | int cnt; | ||
897 | int got_lock[MAXQUOTAS] = {0, 0}; | ||
898 | handle_t *handle; | ||
899 | struct super_block *sb = inode->i_sb; | ||
900 | struct ocfs2_mem_dqinfo *oinfo; | ||
901 | |||
902 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | ||
903 | if (!sb_has_quota_active(sb, cnt)) | ||
904 | continue; | ||
905 | oinfo = sb_dqinfo(sb, cnt)->dqi_priv; | ||
906 | status = ocfs2_lock_global_qf(oinfo, 1); | ||
907 | if (status < 0) | ||
908 | goto out; | ||
909 | got_lock[cnt] = 1; | ||
910 | } | ||
911 | handle = ocfs2_start_trans(OCFS2_SB(sb), | ||
912 | ocfs2_calc_qinit_credits(sb, USRQUOTA) + | ||
913 | ocfs2_calc_qinit_credits(sb, GRPQUOTA)); | ||
914 | if (IS_ERR(handle)) { | ||
915 | status = PTR_ERR(handle); | ||
916 | mlog_errno(status); | ||
917 | goto out; | ||
918 | } | ||
919 | dquot_drop(inode); | ||
920 | ocfs2_commit_trans(OCFS2_SB(sb), handle); | ||
921 | out: | ||
922 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) | ||
923 | if (got_lock[cnt]) { | ||
924 | oinfo = sb_dqinfo(sb, cnt)->dqi_priv; | ||
925 | ocfs2_unlock_global_qf(oinfo, 1); | ||
926 | } | ||
927 | return status; | ||
928 | } | ||
929 | |||
930 | /* See the comment before ocfs2_dquot_initialize. */ | ||
931 | static int ocfs2_dquot_drop(struct inode *inode) | ||
932 | { | ||
933 | int status = 0; | ||
934 | struct super_block *sb = inode->i_sb; | ||
935 | struct ocfs2_mem_dqinfo *oinfo; | ||
936 | int exclusive = 0; | ||
937 | int cnt; | ||
938 | int got_lock[MAXQUOTAS] = {0, 0}; | ||
939 | |||
940 | mlog_entry_void(); | ||
941 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | ||
942 | if (!sb_has_quota_active(sb, cnt)) | ||
943 | continue; | ||
944 | oinfo = sb_dqinfo(sb, cnt)->dqi_priv; | ||
945 | status = ocfs2_lock_global_qf(oinfo, 0); | ||
946 | if (status < 0) | ||
947 | goto out; | ||
948 | got_lock[cnt] = 1; | ||
949 | } | ||
950 | /* Lock against anyone releasing references so that when when we check | ||
951 | * we know we are not going to be last ones to release dquot */ | ||
952 | down_write(&sb_dqopt(sb)->dqptr_sem); | ||
953 | /* Urgh, this is a terrible hack :( */ | ||
954 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) { | ||
955 | if (inode->i_dquot[cnt] != NODQUOT && | ||
956 | atomic_read(&inode->i_dquot[cnt]->dq_count) > 1) { | ||
957 | exclusive = 1; | ||
958 | break; | ||
959 | } | ||
960 | } | ||
961 | if (!exclusive) | ||
962 | dquot_drop_locked(inode); | ||
963 | up_write(&sb_dqopt(sb)->dqptr_sem); | ||
964 | out: | ||
965 | for (cnt = 0; cnt < MAXQUOTAS; cnt++) | ||
966 | if (got_lock[cnt]) { | ||
967 | oinfo = sb_dqinfo(sb, cnt)->dqi_priv; | ||
968 | ocfs2_unlock_global_qf(oinfo, 0); | ||
969 | } | ||
970 | /* In case we bailed out because we had to do expensive locking | ||
971 | * do it now... */ | ||
972 | if (exclusive) | ||
973 | status = ocfs2_dquot_drop_slow(inode); | ||
974 | mlog_exit(status); | ||
975 | return status; | ||
976 | } | ||
977 | |||
978 | static struct dquot *ocfs2_alloc_dquot(struct super_block *sb, int type) | 813 | static struct dquot *ocfs2_alloc_dquot(struct super_block *sb, int type) |
979 | { | 814 | { |
980 | struct ocfs2_dquot *dquot = | 815 | struct ocfs2_dquot *dquot = |
@@ -991,8 +826,8 @@ static void ocfs2_destroy_dquot(struct dquot *dquot) | |||
991 | } | 826 | } |
992 | 827 | ||
993 | struct dquot_operations ocfs2_quota_operations = { | 828 | struct dquot_operations ocfs2_quota_operations = { |
994 | .initialize = ocfs2_dquot_initialize, | 829 | .initialize = dquot_initialize, |
995 | .drop = ocfs2_dquot_drop, | 830 | .drop = dquot_drop, |
996 | .alloc_space = dquot_alloc_space, | 831 | .alloc_space = dquot_alloc_space, |
997 | .alloc_inode = dquot_alloc_inode, | 832 | .alloc_inode = dquot_alloc_inode, |
998 | .free_space = dquot_free_space, | 833 | .free_space = dquot_free_space, |
diff --git a/fs/omfs/Kconfig b/fs/omfs/Kconfig new file mode 100644 index 000000000000..b1b9a0aba6fd --- /dev/null +++ b/fs/omfs/Kconfig | |||
@@ -0,0 +1,13 @@ | |||
1 | config OMFS_FS | ||
2 | tristate "SonicBlue Optimized MPEG File System support" | ||
3 | depends on BLOCK | ||
4 | select CRC_ITU_T | ||
5 | help | ||
6 | This is the proprietary file system used by the Rio Karma music | ||
7 | player and ReplayTV DVR. Despite the name, this filesystem is not | ||
8 | more efficient than a standard FS for MPEG files, in fact likely | ||
9 | the opposite is true. Say Y if you have either of these devices | ||
10 | and wish to mount its disk. | ||
11 | |||
12 | To compile this file system support as a module, choose M here: the | ||
13 | module will be called omfs. If unsure, say N. | ||
diff --git a/fs/qnx4/Kconfig b/fs/qnx4/Kconfig new file mode 100644 index 000000000000..be8e0e1445b6 --- /dev/null +++ b/fs/qnx4/Kconfig | |||
@@ -0,0 +1,25 @@ | |||
1 | config QNX4FS_FS | ||
2 | tristate "QNX4 file system support (read only)" | ||
3 | depends on BLOCK | ||
4 | help | ||
5 | This is the file system used by the real-time operating systems | ||
6 | QNX 4 and QNX 6 (the latter is also called QNX RTP). | ||
7 | Further information is available at <http://www.qnx.com/>. | ||
8 | Say Y if you intend to mount QNX hard disks or floppies. | ||
9 | Unless you say Y to "QNX4FS read-write support" below, you will | ||
10 | only be able to read these file systems. | ||
11 | |||
12 | To compile this file system support as a module, choose M here: the | ||
13 | module will be called qnx4. | ||
14 | |||
15 | If you don't know whether you need it, then you don't need it: | ||
16 | answer N. | ||
17 | |||
18 | config QNX4FS_RW | ||
19 | bool "QNX4FS write support (DANGEROUS)" | ||
20 | depends on QNX4FS_FS && EXPERIMENTAL && BROKEN | ||
21 | help | ||
22 | Say Y if you want to test write support for QNX4 file systems. | ||
23 | |||
24 | It's currently broken, so for now: | ||
25 | answer N. | ||
diff --git a/fs/reiserfs/Kconfig b/fs/reiserfs/Kconfig new file mode 100644 index 000000000000..949b8c6addc8 --- /dev/null +++ b/fs/reiserfs/Kconfig | |||
@@ -0,0 +1,85 @@ | |||
1 | config REISERFS_FS | ||
2 | tristate "Reiserfs support" | ||
3 | help | ||
4 | Stores not just filenames but the files themselves in a balanced | ||
5 | tree. Uses journalling. | ||
6 | |||
7 | Balanced trees are more efficient than traditional file system | ||
8 | architectural foundations. | ||
9 | |||
10 | In general, ReiserFS is as fast as ext2, but is very efficient with | ||
11 | large directories and small files. Additional patches are needed | ||
12 | for NFS and quotas, please see <http://www.namesys.com/> for links. | ||
13 | |||
14 | It is more easily extended to have features currently found in | ||
15 | database and keyword search systems than block allocation based file | ||
16 | systems are. The next version will be so extended, and will support | ||
17 | plugins consistent with our motto ``It takes more than a license to | ||
18 | make source code open.'' | ||
19 | |||
20 | Read <http://www.namesys.com/> to learn more about reiserfs. | ||
21 | |||
22 | Sponsored by Threshold Networks, Emusic.com, and Bigstorage.com. | ||
23 | |||
24 | If you like it, you can pay us to add new features to it that you | ||
25 | need, buy a support contract, or pay us to port it to another OS. | ||
26 | |||
27 | config REISERFS_CHECK | ||
28 | bool "Enable reiserfs debug mode" | ||
29 | depends on REISERFS_FS | ||
30 | help | ||
31 | If you set this to Y, then ReiserFS will perform every check it can | ||
32 | possibly imagine of its internal consistency throughout its | ||
33 | operation. It will also go substantially slower. More than once we | ||
34 | have forgotten that this was on, and then gone despondent over the | ||
35 | latest benchmarks.:-) Use of this option allows our team to go all | ||
36 | out in checking for consistency when debugging without fear of its | ||
37 | effect on end users. If you are on the verge of sending in a bug | ||
38 | report, say Y and you might get a useful error message. Almost | ||
39 | everyone should say N. | ||
40 | |||
41 | config REISERFS_PROC_INFO | ||
42 | bool "Stats in /proc/fs/reiserfs" | ||
43 | depends on REISERFS_FS && PROC_FS | ||
44 | help | ||
45 | Create under /proc/fs/reiserfs a hierarchy of files, displaying | ||
46 | various ReiserFS statistics and internal data at the expense of | ||
47 | making your kernel or module slightly larger (+8 KB). This also | ||
48 | increases the amount of kernel memory required for each mount. | ||
49 | Almost everyone but ReiserFS developers and people fine-tuning | ||
50 | reiserfs or tracing problems should say N. | ||
51 | |||
52 | config REISERFS_FS_XATTR | ||
53 | bool "ReiserFS extended attributes" | ||
54 | depends on REISERFS_FS | ||
55 | help | ||
56 | Extended attributes are name:value pairs associated with inodes by | ||
57 | the kernel or by users (see the attr(5) manual page, or visit | ||
58 | <http://acl.bestbits.at/> for details). | ||
59 | |||
60 | If unsure, say N. | ||
61 | |||
62 | config REISERFS_FS_POSIX_ACL | ||
63 | bool "ReiserFS POSIX Access Control Lists" | ||
64 | depends on REISERFS_FS_XATTR | ||
65 | select FS_POSIX_ACL | ||
66 | help | ||
67 | Posix Access Control Lists (ACLs) support permissions for users and | ||
68 | groups beyond the owner/group/world scheme. | ||
69 | |||
70 | To learn more about Access Control Lists, visit the Posix ACLs for | ||
71 | Linux website <http://acl.bestbits.at/>. | ||
72 | |||
73 | If you don't know what Access Control Lists are, say N | ||
74 | |||
75 | config REISERFS_FS_SECURITY | ||
76 | bool "ReiserFS Security Labels" | ||
77 | depends on REISERFS_FS_XATTR | ||
78 | help | ||
79 | Security labels support alternative access control models | ||
80 | implemented by security modules like SELinux. This option | ||
81 | enables an extended attribute handler for file security | ||
82 | labels in the ReiserFS filesystem. | ||
83 | |||
84 | If you are not using a security module that requires using | ||
85 | extended attributes for file security labels, say N. | ||
diff --git a/fs/romfs/Kconfig b/fs/romfs/Kconfig new file mode 100644 index 000000000000..1a17020f9faf --- /dev/null +++ b/fs/romfs/Kconfig | |||
@@ -0,0 +1,16 @@ | |||
1 | config ROMFS_FS | ||
2 | tristate "ROM file system support" | ||
3 | depends on BLOCK | ||
4 | ---help--- | ||
5 | This is a very small read-only file system mainly intended for | ||
6 | initial ram disks of installation disks, but it could be used for | ||
7 | other read-only media as well. Read | ||
8 | <file:Documentation/filesystems/romfs.txt> for details. | ||
9 | |||
10 | To compile this file system support as a module, choose M here: the | ||
11 | module will be called romfs. Note that the file system of your | ||
12 | root partition (the one containing the directory /) cannot be a | ||
13 | module. | ||
14 | |||
15 | If you don't know whether you need it, then you don't need it: | ||
16 | answer N. | ||
diff --git a/fs/smbfs/Kconfig b/fs/smbfs/Kconfig new file mode 100644 index 000000000000..e668127c8b2e --- /dev/null +++ b/fs/smbfs/Kconfig | |||
@@ -0,0 +1,55 @@ | |||
1 | config SMB_FS | ||
2 | tristate "SMB file system support (OBSOLETE, please use CIFS)" | ||
3 | depends on INET | ||
4 | select NLS | ||
5 | help | ||
6 | SMB (Server Message Block) is the protocol Windows for Workgroups | ||
7 | (WfW), Windows 95/98, Windows NT and OS/2 Lan Manager use to share | ||
8 | files and printers over local networks. Saying Y here allows you to | ||
9 | mount their file systems (often called "shares" in this context) and | ||
10 | access them just like any other Unix directory. Currently, this | ||
11 | works only if the Windows machines use TCP/IP as the underlying | ||
12 | transport protocol, and not NetBEUI. For details, read | ||
13 | <file:Documentation/filesystems/smbfs.txt> and the SMB-HOWTO, | ||
14 | available from <http://www.tldp.org/docs.html#howto>. | ||
15 | |||
16 | Note: if you just want your box to act as an SMB *server* and make | ||
17 | files and printing services available to Windows clients (which need | ||
18 | to have a TCP/IP stack), you don't need to say Y here; you can use | ||
19 | the program SAMBA (available from <ftp://ftp.samba.org/pub/samba/>) | ||
20 | for that. | ||
21 | |||
22 | General information about how to connect Linux, Windows machines and | ||
23 | Macs is on the WWW at <http://www.eats.com/linux_mac_win.html>. | ||
24 | |||
25 | To compile the SMB support as a module, choose M here: | ||
26 | the module will be called smbfs. Most people say N, however. | ||
27 | |||
28 | config SMB_NLS_DEFAULT | ||
29 | bool "Use a default NLS" | ||
30 | depends on SMB_FS | ||
31 | help | ||
32 | Enabling this will make smbfs use nls translations by default. You | ||
33 | need to specify the local charset (CONFIG_NLS_DEFAULT) in the nls | ||
34 | settings and you need to give the default nls for the SMB server as | ||
35 | CONFIG_SMB_NLS_REMOTE. | ||
36 | |||
37 | The nls settings can be changed at mount time, if your smbmount | ||
38 | supports that, using the codepage and iocharset parameters. | ||
39 | |||
40 | smbmount from samba 2.2.0 or later supports this. | ||
41 | |||
42 | config SMB_NLS_REMOTE | ||
43 | string "Default Remote NLS Option" | ||
44 | depends on SMB_NLS_DEFAULT | ||
45 | default "cp437" | ||
46 | help | ||
47 | This setting allows you to specify a default value for which | ||
48 | codepage the server uses. If this field is left blank no | ||
49 | translations will be done by default. The local codepage/charset | ||
50 | default to CONFIG_NLS_DEFAULT. | ||
51 | |||
52 | The nls settings can be changed at mount time, if your smbmount | ||
53 | supports that, using the codepage and iocharset parameters. | ||
54 | |||
55 | smbmount from samba 2.2.0 or later supports this. | ||
diff --git a/fs/squashfs/Kconfig b/fs/squashfs/Kconfig new file mode 100644 index 000000000000..25a00d19d686 --- /dev/null +++ b/fs/squashfs/Kconfig | |||
@@ -0,0 +1,51 @@ | |||
1 | config SQUASHFS | ||
2 | tristate "SquashFS 4.0 - Squashed file system support" | ||
3 | depends on BLOCK | ||
4 | select ZLIB_INFLATE | ||
5 | help | ||
6 | Saying Y here includes support for SquashFS 4.0 (a Compressed | ||
7 | Read-Only File System). Squashfs is a highly compressed read-only | ||
8 | filesystem for Linux. It uses zlib compression to compress both | ||
9 | files, inodes and directories. Inodes in the system are very small | ||
10 | and all blocks are packed to minimise data overhead. Block sizes | ||
11 | greater than 4K are supported up to a maximum of 1 Mbytes (default | ||
12 | block size 128K). SquashFS 4.0 supports 64 bit filesystems and files | ||
13 | (larger than 4GB), full uid/gid information, hard links and | ||
14 | timestamps. | ||
15 | |||
16 | Squashfs is intended for general read-only filesystem use, for | ||
17 | archival use (i.e. in cases where a .tar.gz file may be used), and in | ||
18 | embedded systems where low overhead is needed. Further information | ||
19 | and tools are available from http://squashfs.sourceforge.net. | ||
20 | |||
21 | If you want to compile this as a module ( = code which can be | ||
22 | inserted in and removed from the running kernel whenever you want), | ||
23 | say M here and read <file:Documentation/modules.txt>. The module | ||
24 | will be called squashfs. Note that the root file system (the one | ||
25 | containing the directory /) cannot be compiled as a module. | ||
26 | |||
27 | If unsure, say N. | ||
28 | |||
29 | config SQUASHFS_EMBEDDED | ||
30 | |||
31 | bool "Additional option for memory-constrained systems" | ||
32 | depends on SQUASHFS | ||
33 | default n | ||
34 | help | ||
35 | Saying Y here allows you to specify cache size. | ||
36 | |||
37 | If unsure, say N. | ||
38 | |||
39 | config SQUASHFS_FRAGMENT_CACHE_SIZE | ||
40 | int "Number of fragments cached" if SQUASHFS_EMBEDDED | ||
41 | depends on SQUASHFS | ||
42 | default "3" | ||
43 | help | ||
44 | By default SquashFS caches the last 3 fragments read from | ||
45 | the filesystem. Increasing this amount may mean SquashFS | ||
46 | has to re-read fragments less often from disk, at the expense | ||
47 | of extra system memory. Decreasing this amount will mean | ||
48 | SquashFS uses less memory at the expense of extra reads from disk. | ||
49 | |||
50 | Note there must be at least one cached fragment. Anything | ||
51 | much more than three will probably not make much difference. | ||
diff --git a/fs/sysfs/Kconfig b/fs/sysfs/Kconfig new file mode 100644 index 000000000000..f4b67588b9d6 --- /dev/null +++ b/fs/sysfs/Kconfig | |||
@@ -0,0 +1,23 @@ | |||
1 | config SYSFS | ||
2 | bool "sysfs file system support" if EMBEDDED | ||
3 | default y | ||
4 | help | ||
5 | The sysfs filesystem is a virtual filesystem that the kernel uses to | ||
6 | export internal kernel objects, their attributes, and their | ||
7 | relationships to one another. | ||
8 | |||
9 | Users can use sysfs to ascertain useful information about the running | ||
10 | kernel, such as the devices the kernel has discovered on each bus and | ||
11 | which driver each is bound to. sysfs can also be used to tune devices | ||
12 | and other kernel subsystems. | ||
13 | |||
14 | Some system agents rely on the information in sysfs to operate. | ||
15 | /sbin/hotplug uses device and object attributes in sysfs to assist in | ||
16 | delegating policy decisions, like persistently naming devices. | ||
17 | |||
18 | sysfs is currently used by the block subsystem to mount the root | ||
19 | partition. If sysfs is disabled you must specify the boot device on | ||
20 | the kernel boot command line via its major and minor numbers. For | ||
21 | example, "root=03:01" for /dev/hda1. | ||
22 | |||
23 | Designers of embedded systems may wish to say N here to conserve space. | ||
diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index 66f6e58a7e4b..f2c478c3424e 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c | |||
@@ -63,6 +63,9 @@ read(struct file *file, char __user *userbuf, size_t bytes, loff_t *off) | |||
63 | int count = min_t(size_t, bytes, PAGE_SIZE); | 63 | int count = min_t(size_t, bytes, PAGE_SIZE); |
64 | char *temp; | 64 | char *temp; |
65 | 65 | ||
66 | if (!bytes) | ||
67 | return 0; | ||
68 | |||
66 | if (size) { | 69 | if (size) { |
67 | if (offs > size) | 70 | if (offs > size) |
68 | return 0; | 71 | return 0; |
@@ -131,6 +134,9 @@ static ssize_t write(struct file *file, const char __user *userbuf, | |||
131 | int count = min_t(size_t, bytes, PAGE_SIZE); | 134 | int count = min_t(size_t, bytes, PAGE_SIZE); |
132 | char *temp; | 135 | char *temp; |
133 | 136 | ||
137 | if (!bytes) | ||
138 | return 0; | ||
139 | |||
134 | if (size) { | 140 | if (size) { |
135 | if (offs > size) | 141 | if (offs > size) |
136 | return 0; | 142 | return 0; |
diff --git a/fs/sysv/Kconfig b/fs/sysv/Kconfig new file mode 100644 index 000000000000..33aeb4b75db1 --- /dev/null +++ b/fs/sysv/Kconfig | |||
@@ -0,0 +1,36 @@ | |||
1 | config SYSV_FS | ||
2 | tristate "System V/Xenix/V7/Coherent file system support" | ||
3 | depends on BLOCK | ||
4 | help | ||
5 | SCO, Xenix and Coherent are commercial Unix systems for Intel | ||
6 | machines, and Version 7 was used on the DEC PDP-11. Saying Y | ||
7 | here would allow you to read from their floppies and hard disk | ||
8 | partitions. | ||
9 | |||
10 | If you have floppies or hard disk partitions like that, it is likely | ||
11 | that they contain binaries from those other Unix systems; in order | ||
12 | to run these binaries, you will want to install linux-abi which is | ||
13 | a set of kernel modules that lets you run SCO, Xenix, Wyse, | ||
14 | UnixWare, Dell Unix and System V programs under Linux. It is | ||
15 | available via FTP (user: ftp) from | ||
16 | <ftp://ftp.openlinux.org/pub/people/hch/linux-abi/>). | ||
17 | NOTE: that will work only for binaries from Intel-based systems; | ||
18 | PDP ones will have to wait until somebody ports Linux to -11 ;-) | ||
19 | |||
20 | If you only intend to mount files from some other Unix over the | ||
21 | network using NFS, you don't need the System V file system support | ||
22 | (but you need NFS file system support obviously). | ||
23 | |||
24 | Note that this option is generally not needed for floppies, since a | ||
25 | good portable way to transport files and directories between unixes | ||
26 | (and even other operating systems) is given by the tar program ("man | ||
27 | tar" or preferably "info tar"). Note also that this option has | ||
28 | nothing whatsoever to do with the option "System V IPC". Read about | ||
29 | the System V file system in | ||
30 | <file:Documentation/filesystems/sysv-fs.txt>. | ||
31 | Saying Y here will enlarge your kernel by about 27 KB. | ||
32 | |||
33 | To compile this as a module, choose M here: the module will be called | ||
34 | sysv. | ||
35 | |||
36 | If you haven't heard about all of this before, it's safe to say N. | ||
diff --git a/fs/udf/Kconfig b/fs/udf/Kconfig new file mode 100644 index 000000000000..0e0e99bd6bce --- /dev/null +++ b/fs/udf/Kconfig | |||
@@ -0,0 +1,18 @@ | |||
1 | config UDF_FS | ||
2 | tristate "UDF file system support" | ||
3 | select CRC_ITU_T | ||
4 | help | ||
5 | This is the new file system used on some CD-ROMs and DVDs. Say Y if | ||
6 | you intend to mount DVD discs or CDRW's written in packet mode, or | ||
7 | if written to by other UDF utilities, such as DirectCD. | ||
8 | Please read <file:Documentation/filesystems/udf.txt>. | ||
9 | |||
10 | To compile this file system support as a module, choose M here: the | ||
11 | module will be called udf. | ||
12 | |||
13 | If unsure, say N. | ||
14 | |||
15 | config UDF_NLS | ||
16 | bool | ||
17 | default y | ||
18 | depends on (UDF_FS=m && NLS) || (UDF_FS=y && NLS=y) | ||
diff --git a/fs/ufs/Kconfig b/fs/ufs/Kconfig new file mode 100644 index 000000000000..e4f10a40768a --- /dev/null +++ b/fs/ufs/Kconfig | |||
@@ -0,0 +1,43 @@ | |||
1 | config UFS_FS | ||
2 | tristate "UFS file system support (read only)" | ||
3 | depends on BLOCK | ||
4 | help | ||
5 | BSD and derivate versions of Unix (such as SunOS, FreeBSD, NetBSD, | ||
6 | OpenBSD and NeXTstep) use a file system called UFS. Some System V | ||
7 | Unixes can create and mount hard disk partitions and diskettes using | ||
8 | this file system as well. Saying Y here will allow you to read from | ||
9 | these partitions; if you also want to write to them, say Y to the | ||
10 | experimental "UFS file system write support", below. Please read the | ||
11 | file <file:Documentation/filesystems/ufs.txt> for more information. | ||
12 | |||
13 | The recently released UFS2 variant (used in FreeBSD 5.x) is | ||
14 | READ-ONLY supported. | ||
15 | |||
16 | Note that this option is generally not needed for floppies, since a | ||
17 | good portable way to transport files and directories between unixes | ||
18 | (and even other operating systems) is given by the tar program ("man | ||
19 | tar" or preferably "info tar"). | ||
20 | |||
21 | When accessing NeXTstep files, you may need to convert them from the | ||
22 | NeXT character set to the Latin1 character set; use the program | ||
23 | recode ("info recode") for this purpose. | ||
24 | |||
25 | To compile the UFS file system support as a module, choose M here: the | ||
26 | module will be called ufs. | ||
27 | |||
28 | If you haven't heard about all of this before, it's safe to say N. | ||
29 | |||
30 | config UFS_FS_WRITE | ||
31 | bool "UFS file system write support (DANGEROUS)" | ||
32 | depends on UFS_FS && EXPERIMENTAL | ||
33 | help | ||
34 | Say Y here if you want to try writing to UFS partitions. This is | ||
35 | experimental, so you should back up your UFS partitions beforehand. | ||
36 | |||
37 | config UFS_DEBUG | ||
38 | bool "UFS debugging" | ||
39 | depends on UFS_FS | ||
40 | help | ||
41 | If you are experiencing any problems with the UFS filesystem, say | ||
42 | Y here. This will result in _many_ additional debugging messages to be | ||
43 | written to the system log. | ||