| 1 | =====================================================================
|
| 2 | Everything you never wanted to know about kobjects, ksets, and ktypes
|
| 3 | =====================================================================
|
| 4 |
|
| 5 | :Author: Greg Kroah-Hartman <[email protected]>
|
| 6 | :Last updated: December 19, 2007
|
| 7 |
|
| 8 | Based on an original article by Jon Corbet for lwn.net written October 1,
|
| 9 | 2003 and located at https://lwn.net/Articles/51437/
|
| 10 |
|
| 11 | Part of the difficulty in understanding the driver model - and the kobject
|
| 12 | abstraction upon which it is built - is that there is no obvious starting
|
| 13 | place. Dealing with kobjects requires understanding a few different types,
|
| 14 | all of which make reference to each other. In an attempt to make things
|
| 15 | easier, we'll take a multi-pass approach, starting with vague terms and
|
| 16 | adding detail as we go. To that end, here are some quick definitions of
|
| 17 | some terms we will be working with.
|
| 18 |
|
| 19 | - A kobject is an object of type struct kobject. Kobjects have a name
|
| 20 | and a reference count. A kobject also has a parent pointer (allowing
|
| 21 | objects to be arranged into hierarchies), a specific type, and,
|
| 22 | usually, a representation in the sysfs virtual filesystem.
|
| 23 |
|
| 24 | Kobjects are generally not interesting on their own; instead, they are
|
| 25 | usually embedded within some other structure which contains the stuff
|
| 26 | the code is really interested in.
|
| 27 |
|
| 28 | No structure should **EVER** have more than one kobject embedded within it.
|
| 29 | If it does, the reference counting for the object is sure to be messed
|
| 30 | up and incorrect, and your code will be buggy. So do not do this.
|
| 31 |
|
| 32 | - A ktype is the type of object that embeds a kobject. Every structure
|
| 33 | that embeds a kobject needs a corresponding ktype. The ktype controls
|
| 34 | what happens to the kobject when it is created and destroyed.
|
| 35 |
|
| 36 | - A kset is a group of kobjects. These kobjects can be of the same ktype
|
| 37 | or belong to different ktypes. The kset is the basic container type for
|
| 38 | collections of kobjects. Ksets contain their own kobjects, but you can
|
| 39 | safely ignore that implementation detail as the kset core code handles
|
| 40 | this kobject automatically.
|
| 41 |
|
| 42 | When you see a sysfs directory full of other directories, generally each
|
| 43 | of those directories corresponds to a kobject in the same kset.
|
| 44 |
|
| 45 | We'll look at how to create and manipulate all of these types. A bottom-up
|
| 46 | approach will be taken, so we'll go back to kobjects.
|
| 47 |
|
| 48 |
|
| 49 | Embedding kobjects
|
| 50 | ==================
|
| 51 |
|
| 52 | It is rare for kernel code to create a standalone kobject, with one major
|
| 53 | exception explained below. Instead, kobjects are used to control access to
|
| 54 | a larger, domain-specific object. To this end, kobjects will be found
|
| 55 | embedded in other structures. If you are used to thinking of things in
|
| 56 | object-oriented terms, kobjects can be seen as a top-level, abstract class
|
| 57 | from which other classes are derived. A kobject implements a set of
|
| 58 | capabilities which are not particularly useful by themselves, but are
|
| 59 | nice to have in other objects. The C language does not allow for the
|
| 60 | direct expression of inheritance, so other techniques - such as structure
|
| 61 | embedding - must be used.
|
| 62 |
|
| 63 | (As an aside, for those familiar with the kernel linked list implementation,
|
| 64 | this is analogous as to how "list_head" structs are rarely useful on
|
| 65 | their own, but are invariably found embedded in the larger objects of
|
| 66 | interest.)
|
| 67 |
|
| 68 | So, for example, the UIO code in ``drivers/uio/uio.c`` has a structure that
|
| 69 | defines the memory region associated with a uio device::
|
| 70 |
|
| 71 | struct uio_map {
|
| 72 | struct kobject kobj;
|
| 73 | struct uio_mem *mem;
|
| 74 | };
|
| 75 |
|
| 76 | If you have a struct uio_map structure, finding its embedded kobject is
|
| 77 | just a matter of using the kobj member. Code that works with kobjects will
|
| 78 | often have the opposite problem, however: given a struct kobject pointer,
|
| 79 | what is the pointer to the containing structure? You must avoid tricks
|
| 80 | (such as assuming that the kobject is at the beginning of the structure)
|
| 81 | and, instead, use the container_of() macro, found in ``<linux/container_of.h>``::
|
| 82 |
|
| 83 | container_of(ptr, type, member)
|
| 84 |
|
| 85 | where:
|
| 86 |
|
| 87 | * ``ptr`` is the pointer to the embedded kobject,
|
| 88 | * ``type`` is the type of the containing structure, and
|
| 89 | * ``member`` is the name of the structure field to which ``pointer`` points.
|
| 90 |
|
| 91 | The return value from container_of() is a pointer to the corresponding
|
| 92 | container type. So, for example, a pointer ``kp`` to a struct kobject
|
| 93 | embedded **within** a struct uio_map could be converted to a pointer to the
|
| 94 | **containing** uio_map structure with::
|
| 95 |
|
| 96 | struct uio_map *u_map = container_of(kp, struct uio_map, kobj);
|
| 97 |
|
| 98 | For convenience, programmers often define a simple macro for **back-casting**
|
| 99 | kobject pointers to the containing type. Exactly this happens in the
|
| 100 | earlier ``drivers/uio/uio.c``, as you can see here::
|
| 101 |
|
| 102 | struct uio_map {
|
| 103 | struct kobject kobj;
|
| 104 | struct uio_mem *mem;
|
| 105 | };
|
| 106 |
|
| 107 | #define to_map(map) container_of(map, struct uio_map, kobj)
|
| 108 |
|
| 109 | where the macro argument "map" is a pointer to the struct kobject in
|
| 110 | question. That macro is subsequently invoked with::
|
| 111 |
|
| 112 | struct uio_map *map = to_map(kobj);
|
| 113 |
|
| 114 |
|
| 115 | Initialization of kobjects
|
| 116 | ==========================
|
| 117 |
|
| 118 | Code which creates a kobject must, of course, initialize that object. Some
|
| 119 | of the internal fields are setup with a (mandatory) call to kobject_init()::
|
| 120 |
|
| 121 | void kobject_init(struct kobject *kobj, const struct kobj_type *ktype);
|
| 122 |
|
| 123 | The ktype is required for a kobject to be created properly, as every kobject
|
| 124 | must have an associated kobj_type. After calling kobject_init(), to
|
| 125 | register the kobject with sysfs, the function kobject_add() must be called::
|
| 126 |
|
| 127 | int kobject_add(struct kobject *kobj, struct kobject *parent,
|
| 128 | const char *fmt, ...);
|
| 129 |
|
| 130 | This sets up the parent of the kobject and the name for the kobject
|
| 131 | properly. If the kobject is to be associated with a specific kset,
|
| 132 | kobj->kset must be assigned before calling kobject_add(). If a kset is
|
| 133 | associated with a kobject, then the parent for the kobject can be set to
|
| 134 | NULL in the call to kobject_add() and then the kobject's parent will be the
|
| 135 | kset itself.
|
| 136 |
|
| 137 | As the name of the kobject is set when it is added to the kernel, the name
|
| 138 | of the kobject should never be manipulated directly. If you must change
|
| 139 | the name of the kobject, call kobject_rename()::
|
| 140 |
|
| 141 | int kobject_rename(struct kobject *kobj, const char *new_name);
|
| 142 |
|
| 143 | kobject_rename() does not perform any locking or have a solid notion of
|
| 144 | what names are valid so the caller must provide their own sanity checking
|
| 145 | and serialization.
|
| 146 |
|
| 147 | There is a function called kobject_set_name() but that is legacy cruft and
|
| 148 | is being removed. If your code needs to call this function, it is
|
| 149 | incorrect and needs to be fixed.
|
| 150 |
|
| 151 | To properly access the name of the kobject, use the function
|
| 152 | kobject_name()::
|
| 153 |
|
| 154 | const char *kobject_name(const struct kobject * kobj);
|
| 155 |
|
| 156 | There is a helper function to both initialize and add the kobject to the
|
| 157 | kernel at the same time, called surprisingly enough kobject_init_and_add()::
|
| 158 |
|
| 159 | int kobject_init_and_add(struct kobject *kobj, const struct kobj_type *ktype,
|
| 160 | struct kobject *parent, const char *fmt, ...);
|
| 161 |
|
| 162 | The arguments are the same as the individual kobject_init() and
|
| 163 | kobject_add() functions described above.
|
| 164 |
|
| 165 |
|
| 166 | Uevents
|
| 167 | =======
|
| 168 |
|
| 169 | After a kobject has been registered with the kobject core, you need to
|
| 170 | announce to the world that it has been created. This can be done with a
|
| 171 | call to kobject_uevent()::
|
| 172 |
|
| 173 | int kobject_uevent(struct kobject *kobj, enum kobject_action action);
|
| 174 |
|
| 175 | Use the **KOBJ_ADD** action for when the kobject is first added to the kernel.
|
| 176 | This should be done only after any attributes or children of the kobject
|
| 177 | have been initialized properly, as userspace will instantly start to look
|
| 178 | for them when this call happens.
|
| 179 |
|
| 180 | When the kobject is removed from the kernel (details on how to do that are
|
| 181 | below), the uevent for **KOBJ_REMOVE** will be automatically created by the
|
| 182 | kobject core, so the caller does not have to worry about doing that by
|
| 183 | hand.
|
| 184 |
|
| 185 |
|
| 186 | Reference counts
|
| 187 | ================
|
| 188 |
|
| 189 | One of the key functions of a kobject is to serve as a reference counter
|
| 190 | for the object in which it is embedded. As long as references to the object
|
| 191 | exist, the object (and the code which supports it) must continue to exist.
|
| 192 | The low-level functions for manipulating a kobject's reference counts are::
|
| 193 |
|
| 194 | struct kobject *kobject_get(struct kobject *kobj);
|
| 195 | void kobject_put(struct kobject *kobj);
|
| 196 |
|
| 197 | A successful call to kobject_get() will increment the kobject's reference
|
| 198 | counter and return the pointer to the kobject.
|
| 199 |
|
| 200 | When a reference is released, the call to kobject_put() will decrement the
|
| 201 | reference count and, possibly, free the object. Note that kobject_init()
|
| 202 | sets the reference count to one, so the code which sets up the kobject will
|
| 203 | need to do a kobject_put() eventually to release that reference.
|
| 204 |
|
| 205 | Because kobjects are dynamic, they must not be declared statically or on
|
| 206 | the stack, but instead, always allocated dynamically. Future versions of
|
| 207 | the kernel will contain a run-time check for kobjects that are created
|
| 208 | statically and will warn the developer of this improper usage.
|
| 209 |
|
| 210 | If all that you want to use a kobject for is to provide a reference counter
|
| 211 | for your structure, please use the struct kref instead; a kobject would be
|
| 212 | overkill. For more information on how to use struct kref, please see the
|
| 213 | file Documentation/core-api/kref.rst in the Linux kernel source tree.
|
| 214 |
|
| 215 |
|
| 216 | Creating "simple" kobjects
|
| 217 | ==========================
|
| 218 |
|
| 219 | Sometimes all that a developer wants is a way to create a simple directory
|
| 220 | in the sysfs hierarchy, and not have to mess with the whole complication of
|
| 221 | ksets, show and store functions, and other details. This is the one
|
| 222 | exception where a single kobject should be created. To create such an
|
| 223 | entry, use the function::
|
| 224 |
|
| 225 | struct kobject *kobject_create_and_add(const char *name, struct kobject *parent);
|
| 226 |
|
| 227 | This function will create a kobject and place it in sysfs in the location
|
| 228 | underneath the specified parent kobject. To create simple attributes
|
| 229 | associated with this kobject, use::
|
| 230 |
|
| 231 | int sysfs_create_file(struct kobject *kobj, const struct attribute *attr);
|
| 232 |
|
| 233 | or::
|
| 234 |
|
| 235 | int sysfs_create_group(struct kobject *kobj, const struct attribute_group *grp);
|
| 236 |
|
| 237 | Both types of attributes used here, with a kobject that has been created
|
| 238 | with the kobject_create_and_add(), can be of type kobj_attribute, so no
|
| 239 | special custom attribute is needed to be created.
|
| 240 |
|
| 241 | See the example module, ``samples/kobject/kobject-example.c`` for an
|
| 242 | implementation of a simple kobject and attributes.
|
| 243 |
|
| 244 |
|
| 245 |
|
| 246 | ktypes and release methods
|
| 247 | ==========================
|
| 248 |
|
| 249 | One important thing still missing from the discussion is what happens to a
|
| 250 | kobject when its reference count reaches zero. The code which created the
|
| 251 | kobject generally does not know when that will happen; if it did, there
|
| 252 | would be little point in using a kobject in the first place. Even
|
| 253 | predictable object lifecycles become more complicated when sysfs is brought
|
| 254 | in as other portions of the kernel can get a reference on any kobject that
|
| 255 | is registered in the system.
|
| 256 |
|
| 257 | The end result is that a structure protected by a kobject cannot be freed
|
| 258 | before its reference count goes to zero. The reference count is not under
|
| 259 | the direct control of the code which created the kobject. So that code must
|
| 260 | be notified asynchronously whenever the last reference to one of its
|
| 261 | kobjects goes away.
|
| 262 |
|
| 263 | Once you registered your kobject via kobject_add(), you must never use
|
| 264 | kfree() to free it directly. The only safe way is to use kobject_put(). It
|
| 265 | is good practice to always use kobject_put() after kobject_init() to avoid
|
| 266 | errors creeping in.
|
| 267 |
|
| 268 | This notification is done through a kobject's release() method. Usually
|
| 269 | such a method has a form like::
|
| 270 |
|
| 271 | void my_object_release(struct kobject *kobj)
|
| 272 | {
|
| 273 | struct my_object *mine = container_of(kobj, struct my_object, kobj);
|
| 274 |
|
| 275 | /* Perform any additional cleanup on this object, then... */
|
| 276 | kfree(mine);
|
| 277 | }
|
| 278 |
|
| 279 | One important point cannot be overstated: every kobject must have a
|
| 280 | release() method, and the kobject must persist (in a consistent state)
|
| 281 | until that method is called. If these constraints are not met, the code is
|
| 282 | flawed. Note that the kernel will warn you if you forget to provide a
|
| 283 | release() method. Do not try to get rid of this warning by providing an
|
| 284 | "empty" release function.
|
| 285 |
|
| 286 | If all your cleanup function needs to do is call kfree(), then you must
|
| 287 | create a wrapper function which uses container_of() to upcast to the correct
|
| 288 | type (as shown in the example above) and then calls kfree() on the overall
|
| 289 | structure.
|
| 290 |
|
| 291 | Note, the name of the kobject is available in the release function, but it
|
| 292 | must NOT be changed within this callback. Otherwise there will be a memory
|
| 293 | leak in the kobject core, which makes people unhappy.
|
| 294 |
|
| 295 | Interestingly, the release() method is not stored in the kobject itself;
|
| 296 | instead, it is associated with the ktype. So let us introduce struct
|
| 297 | kobj_type::
|
| 298 |
|
| 299 | struct kobj_type {
|
| 300 | void (*release)(struct kobject *kobj);
|
| 301 | const struct sysfs_ops *sysfs_ops;
|
| 302 | const struct attribute_group **default_groups;
|
| 303 | const struct kobj_ns_type_operations *(*child_ns_type)(struct kobject *kobj);
|
| 304 | const void *(*namespace)(struct kobject *kobj);
|
| 305 | void (*get_ownership)(struct kobject *kobj, kuid_t *uid, kgid_t *gid);
|
| 306 | };
|
| 307 |
|
| 308 | This structure is used to describe a particular type of kobject (or, more
|
| 309 | correctly, of containing object). Every kobject needs to have an associated
|
| 310 | kobj_type structure; a pointer to that structure must be specified when you
|
| 311 | call kobject_init() or kobject_init_and_add().
|
| 312 |
|
| 313 | The release field in struct kobj_type is, of course, a pointer to the
|
| 314 | release() method for this type of kobject. The other two fields (sysfs_ops
|
| 315 | and default_groups) control how objects of this type are represented in
|
| 316 | sysfs; they are beyond the scope of this document.
|
| 317 |
|
| 318 | The default_groups pointer is a list of default attributes that will be
|
| 319 | automatically created for any kobject that is registered with this ktype.
|
| 320 |
|
| 321 |
|
| 322 | ksets
|
| 323 | =====
|
| 324 |
|
| 325 | A kset is merely a collection of kobjects that want to be associated with
|
| 326 | each other. There is no restriction that they be of the same ktype, but be
|
| 327 | very careful if they are not.
|
| 328 |
|
| 329 | A kset serves these functions:
|
| 330 |
|
| 331 | - It serves as a bag containing a group of objects. A kset can be used by
|
| 332 | the kernel to track "all block devices" or "all PCI device drivers."
|
| 333 |
|
| 334 | - A kset is also a subdirectory in sysfs, where the associated kobjects
|
| 335 | with the kset can show up. Every kset contains a kobject which can be
|
| 336 | set up to be the parent of other kobjects; the top-level directories of
|
| 337 | the sysfs hierarchy are constructed in this way.
|
| 338 |
|
| 339 | - Ksets can support the "hotplugging" of kobjects and influence how
|
| 340 | uevent events are reported to user space.
|
| 341 |
|
| 342 | In object-oriented terms, "kset" is the top-level container class; ksets
|
| 343 | contain their own kobject, but that kobject is managed by the kset code and
|
| 344 | should not be manipulated by any other user.
|
| 345 |
|
| 346 | A kset keeps its children in a standard kernel linked list. Kobjects point
|
| 347 | back to their containing kset via their kset field. In almost all cases,
|
| 348 | the kobjects belonging to a kset have that kset (or, strictly, its embedded
|
| 349 | kobject) in their parent.
|
| 350 |
|
| 351 | As a kset contains a kobject within it, it should always be dynamically
|
| 352 | created and never declared statically or on the stack. To create a new
|
| 353 | kset use::
|
| 354 |
|
| 355 | struct kset *kset_create_and_add(const char *name,
|
| 356 | const struct kset_uevent_ops *uevent_ops,
|
| 357 | struct kobject *parent_kobj);
|
| 358 |
|
| 359 | When you are finished with the kset, call::
|
| 360 |
|
| 361 | void kset_unregister(struct kset *k);
|
| 362 |
|
| 363 | to destroy it. This removes the kset from sysfs and decrements its reference
|
| 364 | count. When the reference count goes to zero, the kset will be released.
|
| 365 | Because other references to the kset may still exist, the release may happen
|
| 366 | after kset_unregister() returns.
|
| 367 |
|
| 368 | An example of using a kset can be seen in the
|
| 369 | ``samples/kobject/kset-example.c`` file in the kernel tree.
|
| 370 |
|
| 371 | If a kset wishes to control the uevent operations of the kobjects
|
| 372 | associated with it, it can use the struct kset_uevent_ops to handle it::
|
| 373 |
|
| 374 | struct kset_uevent_ops {
|
| 375 | int (* const filter)(struct kobject *kobj);
|
| 376 | const char *(* const name)(struct kobject *kobj);
|
| 377 | int (* const uevent)(struct kobject *kobj, struct kobj_uevent_env *env);
|
| 378 | };
|
| 379 |
|
| 380 |
|
| 381 | The filter function allows a kset to prevent a uevent from being emitted to
|
| 382 | userspace for a specific kobject. If the function returns 0, the uevent
|
| 383 | will not be emitted.
|
| 384 |
|
| 385 | The name function will be called to override the default name of the kset
|
| 386 | that the uevent sends to userspace. By default, the name will be the same
|
| 387 | as the kset itself, but this function, if present, can override that name.
|
| 388 |
|
| 389 | The uevent function will be called when the uevent is about to be sent to
|
| 390 | userspace to allow more environment variables to be added to the uevent.
|
| 391 |
|
| 392 | One might ask how, exactly, a kobject is added to a kset, given that no
|
| 393 | functions which perform that function have been presented. The answer is
|
| 394 | that this task is handled by kobject_add(). When a kobject is passed to
|
| 395 | kobject_add(), its kset member should point to the kset to which the
|
| 396 | kobject will belong. kobject_add() will handle the rest.
|
| 397 |
|
| 398 | If the kobject belonging to a kset has no parent kobject set, it will be
|
| 399 | added to the kset's directory. Not all members of a kset do necessarily
|
| 400 | live in the kset directory. If an explicit parent kobject is assigned
|
| 401 | before the kobject is added, the kobject is registered with the kset, but
|
| 402 | added below the parent kobject.
|
| 403 |
|
| 404 |
|
| 405 | Kobject removal
|
| 406 | ===============
|
| 407 |
|
| 408 | After a kobject has been registered with the kobject core successfully, it
|
| 409 | must be cleaned up when the code is finished with it. To do that, call
|
| 410 | kobject_put(). By doing this, the kobject core will automatically clean up
|
| 411 | all of the memory allocated by this kobject. If a ``KOBJ_ADD`` uevent has been
|
| 412 | sent for the object, a corresponding ``KOBJ_REMOVE`` uevent will be sent, and
|
| 413 | any other sysfs housekeeping will be handled for the caller properly.
|
| 414 |
|
| 415 | If you need to do a two-stage delete of the kobject (say you are not
|
| 416 | allowed to sleep when you need to destroy the object), then call
|
| 417 | kobject_del() which will unregister the kobject from sysfs. This makes the
|
| 418 | kobject "invisible", but it is not cleaned up, and the reference count of
|
| 419 | the object is still the same. At a later time call kobject_put() to finish
|
| 420 | the cleanup of the memory associated with the kobject.
|
| 421 |
|
| 422 | kobject_del() can be used to drop the reference to the parent object, if
|
| 423 | circular references are constructed. It is valid in some cases, that a
|
| 424 | parent objects references a child. Circular references _must_ be broken
|
| 425 | with an explicit call to kobject_del(), so that a release functions will be
|
| 426 | called, and the objects in the former circle release each other.
|
| 427 |
|
| 428 |
|
| 429 | Example code to copy from
|
| 430 | =========================
|
| 431 |
|
| 432 | For a more complete example of using ksets and kobjects properly, see the
|
| 433 | example programs ``samples/kobject/{kobject-example.c,kset-example.c}``,
|
| 434 | which will be built as loadable modules if you select ``CONFIG_SAMPLE_KOBJECT``.
|