Based on kernel version 3.9. Page generated on 2013-05-02 23:03 EST.
1 <?xml version="1.0" encoding="UTF-8"?> 2 <!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN" 3 "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" []> 4 5 <book id="Linux-filesystems-API"> 6 <bookinfo> 7 <title>Linux Filesystems API</title> 8 9 <legalnotice> 10 <para> 11 This documentation is free software; you can redistribute 12 it and/or modify it under the terms of the GNU General Public 13 License as published by the Free Software Foundation; either 14 version 2 of the License, or (at your option) any later 15 version. 16 </para> 17 18 <para> 19 This program is distributed in the hope that it will be 20 useful, but WITHOUT ANY WARRANTY; without even the implied 21 warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. 22 See the GNU General Public License for more details. 23 </para> 24 25 <para> 26 You should have received a copy of the GNU General Public 27 License along with this program; if not, write to the Free 28 Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, 29 MA 02111-1307 USA 30 </para> 31 32 <para> 33 For more details see the file COPYING in the source 34 distribution of Linux. 35 </para> 36 </legalnotice> 37 </bookinfo> 38 39 <toc></toc> 40 41 <chapter id="vfs"> 42 <title>The Linux VFS</title> 43 <sect1 id="the_filesystem_types"><title>The Filesystem types</title> 44 !Iinclude/linux/fs.h 45 </sect1> 46 <sect1 id="the_directory_cache"><title>The Directory Cache</title> 47 !Efs/dcache.c 48 !Iinclude/linux/dcache.h 49 </sect1> 50 <sect1 id="inode_handling"><title>Inode Handling</title> 51 !Efs/inode.c 52 !Efs/bad_inode.c 53 </sect1> 54 <sect1 id="registration_and_superblocks"><title>Registration and Superblocks</title> 55 !Efs/super.c 56 </sect1> 57 <sect1 id="file_locks"><title>File Locks</title> 58 !Efs/locks.c 59 !Ifs/locks.c 60 </sect1> 61 <sect1 id="other_functions"><title>Other Functions</title> 62 !Efs/mpage.c 63 !Efs/namei.c 64 !Efs/buffer.c 65 !Efs/bio.c 66 !Efs/seq_file.c 67 !Efs/filesystems.c 68 !Efs/fs-writeback.c 69 !Efs/block_dev.c 70 </sect1> 71 </chapter> 72 73 <chapter id="proc"> 74 <title>The proc filesystem</title> 75 76 <sect1 id="sysctl_interface"><title>sysctl interface</title> 77 !Ekernel/sysctl.c 78 </sect1> 79 80 <sect1 id="proc_filesystem_interface"><title>proc filesystem interface</title> 81 !Ifs/proc/base.c 82 </sect1> 83 </chapter> 84 85 <chapter id="fs_events"> 86 <title>Events based on file descriptors</title> 87 !Efs/eventfd.c 88 </chapter> 89 90 <chapter id="sysfs"> 91 <title>The Filesystem for Exporting Kernel Objects</title> 92 !Efs/sysfs/file.c 93 !Efs/sysfs/symlink.c 94 !Efs/sysfs/bin.c 95 </chapter> 96 97 <chapter id="debugfs"> 98 <title>The debugfs filesystem</title> 99 100 <sect1 id="debugfs_interface"><title>debugfs interface</title> 101 !Efs/debugfs/inode.c 102 !Efs/debugfs/file.c 103 </sect1> 104 </chapter> 105 106 <chapter id="LinuxJDBAPI"> 107 <chapterinfo> 108 <title>The Linux Journalling API</title> 109 110 <authorgroup> 111 <author> 112 <firstname>Roger</firstname> 113 <surname>Gammans</surname> 114 <affiliation> 115 <address> 116 <email>rgammans@computer-surgery.co.uk</email> 117 </address> 118 </affiliation> 119 </author> 120 </authorgroup> 121 122 <authorgroup> 123 <author> 124 <firstname>Stephen</firstname> 125 <surname>Tweedie</surname> 126 <affiliation> 127 <address> 128 <email>sct@redhat.com</email> 129 </address> 130 </affiliation> 131 </author> 132 </authorgroup> 133 134 <copyright> 135 <year>2002</year> 136 <holder>Roger Gammans</holder> 137 </copyright> 138 </chapterinfo> 139 140 <title>The Linux Journalling API</title> 141 142 <sect1 id="journaling_overview"> 143 <title>Overview</title> 144 <sect2 id="journaling_details"> 145 <title>Details</title> 146 <para> 147 The journalling layer is easy to use. You need to 148 first of all create a journal_t data structure. There are 149 two calls to do this dependent on how you decide to allocate the physical 150 media on which the journal resides. The journal_init_inode() call 151 is for journals stored in filesystem inodes, or the journal_init_dev() 152 call can be use for journal stored on a raw device (in a continuous range 153 of blocks). A journal_t is a typedef for a struct pointer, so when 154 you are finally finished make sure you call journal_destroy() on it 155 to free up any used kernel memory. 156 </para> 157 158 <para> 159 Once you have got your journal_t object you need to 'mount' or load the journal 160 file, unless of course you haven't initialised it yet - in which case you 161 need to call journal_create(). 162 </para> 163 164 <para> 165 Most of the time however your journal file will already have been created, but 166 before you load it you must call journal_wipe() to empty the journal file. 167 Hang on, you say , what if the filesystem wasn't cleanly umount()'d . Well, it is the 168 job of the client file system to detect this and skip the call to journal_wipe(). 169 </para> 170 171 <para> 172 In either case the next call should be to journal_load() which prepares the 173 journal file for use. Note that journal_wipe(..,0) calls journal_skip_recovery() 174 for you if it detects any outstanding transactions in the journal and similarly 175 journal_load() will call journal_recover() if necessary. 176 I would advise reading fs/ext3/super.c for examples on this stage. 177 [RGG: Why is the journal_wipe() call necessary - doesn't this needlessly 178 complicate the API. Or isn't a good idea for the journal layer to hide 179 dirty mounts from the client fs] 180 </para> 181 182 <para> 183 Now you can go ahead and start modifying the underlying 184 filesystem. Almost. 185 </para> 186 187 <para> 188 189 You still need to actually journal your filesystem changes, this 190 is done by wrapping them into transactions. Additionally you 191 also need to wrap the modification of each of the buffers 192 with calls to the journal layer, so it knows what the modifications 193 you are actually making are. To do this use journal_start() which 194 returns a transaction handle. 195 </para> 196 197 <para> 198 journal_start() 199 and its counterpart journal_stop(), which indicates the end of a transaction 200 are nestable calls, so you can reenter a transaction if necessary, 201 but remember you must call journal_stop() the same number of times as 202 journal_start() before the transaction is completed (or more accurately 203 leaves the update phase). Ext3/VFS makes use of this feature to simplify 204 quota support. 205 </para> 206 207 <para> 208 Inside each transaction you need to wrap the modifications to the 209 individual buffers (blocks). Before you start to modify a buffer you 210 need to call journal_get_{create,write,undo}_access() as appropriate, 211 this allows the journalling layer to copy the unmodified data if it 212 needs to. After all the buffer may be part of a previously uncommitted 213 transaction. 214 At this point you are at last ready to modify a buffer, and once 215 you are have done so you need to call journal_dirty_{meta,}data(). 216 Or if you've asked for access to a buffer you now know is now longer 217 required to be pushed back on the device you can call journal_forget() 218 in much the same way as you might have used bforget() in the past. 219 </para> 220 221 <para> 222 A journal_flush() may be called at any time to commit and checkpoint 223 all your transactions. 224 </para> 225 226 <para> 227 Then at umount time , in your put_super() you can then call journal_destroy() 228 to clean up your in-core journal object. 229 </para> 230 231 <para> 232 Unfortunately there a couple of ways the journal layer can cause a deadlock. 233 The first thing to note is that each task can only have 234 a single outstanding transaction at any one time, remember nothing 235 commits until the outermost journal_stop(). This means 236 you must complete the transaction at the end of each file/inode/address 237 etc. operation you perform, so that the journalling system isn't re-entered 238 on another journal. Since transactions can't be nested/batched 239 across differing journals, and another filesystem other than 240 yours (say ext3) may be modified in a later syscall. 241 </para> 242 243 <para> 244 The second case to bear in mind is that journal_start() can 245 block if there isn't enough space in the journal for your transaction 246 (based on the passed nblocks param) - when it blocks it merely(!) needs to 247 wait for transactions to complete and be committed from other tasks, 248 so essentially we are waiting for journal_stop(). So to avoid 249 deadlocks you must treat journal_start/stop() as if they 250 were semaphores and include them in your semaphore ordering rules to prevent 251 deadlocks. Note that journal_extend() has similar blocking behaviour to 252 journal_start() so you can deadlock here just as easily as on journal_start(). 253 </para> 254 255 <para> 256 Try to reserve the right number of blocks the first time. ;-). This will 257 be the maximum number of blocks you are going to touch in this transaction. 258 I advise having a look at at least ext3_jbd.h to see the basis on which 259 ext3 uses to make these decisions. 260 </para> 261 262 <para> 263 Another wriggle to watch out for is your on-disk block allocation strategy. 264 why? Because, if you undo a delete, you need to ensure you haven't reused any 265 of the freed blocks in a later transaction. One simple way of doing this 266 is make sure any blocks you allocate only have checkpointed transactions 267 listed against them. Ext3 does this in ext3_test_allocatable(). 268 </para> 269 270 <para> 271 Lock is also providing through journal_{un,}lock_updates(), 272 ext3 uses this when it wants a window with a clean and stable fs for a moment. 273 eg. 274 </para> 275 276 <programlisting> 277 278 journal_lock_updates() //stop new stuff happening.. 279 journal_flush() // checkpoint everything. 280 ..do stuff on stable fs 281 journal_unlock_updates() // carry on with filesystem use. 282 </programlisting> 283 284 <para> 285 The opportunities for abuse and DOS attacks with this should be obvious, 286 if you allow unprivileged userspace to trigger codepaths containing these 287 calls. 288 </para> 289 290 <para> 291 A new feature of jbd since 2.5.25 is commit callbacks with the new 292 journal_callback_set() function you can now ask the journalling layer 293 to call you back when the transaction is finally committed to disk, so that 294 you can do some of your own management. The key to this is the journal_callback 295 struct, this maintains the internal callback information but you can 296 extend it like this:- 297 </para> 298 <programlisting> 299 struct myfs_callback_s { 300 //Data structure element required by jbd.. 301 struct journal_callback for_jbd; 302 // Stuff for myfs allocated together. 303 myfs_inode* i_commited; 304 305 } 306 </programlisting> 307 308 <para> 309 this would be useful if you needed to know when data was committed to a 310 particular inode. 311 </para> 312 313 </sect2> 314 315 <sect2 id="jbd_summary"> 316 <title>Summary</title> 317 <para> 318 Using the journal is a matter of wrapping the different context changes, 319 being each mount, each modification (transaction) and each changed buffer 320 to tell the journalling layer about them. 321 </para> 322 323 <para> 324 Here is a some pseudo code to give you an idea of how it works, as 325 an example. 326 </para> 327 328 <programlisting> 329 journal_t* my_jnrl = journal_create(); 330 journal_init_{dev,inode}(jnrl,...) 331 if (clean) journal_wipe(); 332 journal_load(); 333 334 foreach(transaction) { /*transactions must be 335 completed before 336 a syscall returns to 337 userspace*/ 338 339 handle_t * xct=journal_start(my_jnrl); 340 foreach(bh) { 341 journal_get_{create,write,undo}_access(xact,bh); 342 if ( myfs_modify(bh) ) { /* returns true 343 if makes changes */ 344 journal_dirty_{meta,}data(xact,bh); 345 } else { 346 journal_forget(bh); 347 } 348 } 349 journal_stop(xct); 350 } 351 journal_destroy(my_jrnl); 352 </programlisting> 353 </sect2> 354 355 </sect1> 356 357 <sect1 id="data_types"> 358 <title>Data Types</title> 359 <para> 360 The journalling layer uses typedefs to 'hide' the concrete definitions 361 of the structures used. As a client of the JBD layer you can 362 just rely on the using the pointer as a magic cookie of some sort. 363 364 Obviously the hiding is not enforced as this is 'C'. 365 </para> 366 <sect2 id="structures"><title>Structures</title> 367 !Iinclude/linux/jbd.h 368 </sect2> 369 </sect1> 370 371 <sect1 id="functions"> 372 <title>Functions</title> 373 <para> 374 The functions here are split into two groups those that 375 affect a journal as a whole, and those which are used to 376 manage transactions 377 </para> 378 <sect2 id="journal_level"><title>Journal Level</title> 379 !Efs/jbd/journal.c 380 !Ifs/jbd/recovery.c 381 </sect2> 382 <sect2 id="transaction_level"><title>Transasction Level</title> 383 !Efs/jbd/transaction.c 384 </sect2> 385 </sect1> 386 <sect1 id="see_also"> 387 <title>See also</title> 388 <para> 389 <citation> 390 <ulink url="http://kernel.org/pub/linux/kernel/people/sct/ext3/journal-design.ps.gz"> 391 Journaling the Linux ext2fs Filesystem, LinuxExpo 98, Stephen Tweedie 392 </ulink> 393 </citation> 394 </para> 395 <para> 396 <citation> 397 <ulink url="http://olstrans.sourceforge.net/release/OLS2000-ext3/OLS2000-ext3.html"> 398 Ext3 Journalling FileSystem, OLS 2000, Dr. Stephen Tweedie 399 </ulink> 400 </citation> 401 </para> 402 </sect1> 403 404 </chapter> 405 406 <chapter id="splice"> 407 <title>splice API</title> 408 <para> 409 splice is a method for moving blocks of data around inside the 410 kernel, without continually transferring them between the kernel 411 and user space. 412 </para> 413 !Ffs/splice.c 414 </chapter> 415 416 <chapter id="pipes"> 417 <title>pipes API</title> 418 <para> 419 Pipe interfaces are all for in-kernel (builtin image) use. 420 They are not exported for use by modules. 421 </para> 422 !Iinclude/linux/pipe_fs_i.h 423 !Ffs/pipe.c 424 </chapter> 425 426 </book>