1 .. SPDX-License-Identifier: GPL-2.0
2 .. include:: <isonum.txt>
4 ===================================================
5 ACPI Device Tree - Representation of ACPI Namespace
6 ===================================================
8 :Copyright: |copy| 2013, Intel Corporation
10 :Author: Lv Zheng <lv.zheng@intel.com>
12 :Credit: Thanks for the help from Zhang Rui <rui.zhang@intel.com> and
13 Rafael J.Wysocki <rafael.j.wysocki@intel.com>.
17 The Linux ACPI subsystem converts ACPI namespace objects into a Linux
18 device tree under the /sys/devices/LNXSYSTEM:00 and updates it upon
19 receiving ACPI hotplug notification events. For each device object
20 in this hierarchy there is a corresponding symbolic link in the
21 /sys/bus/acpi/devices.
23 This document illustrates the structure of the ACPI device tree.
25 ACPI Definition Blocks
26 ======================
28 The ACPI firmware sets up RSDP (Root System Description Pointer) in the
29 system memory address space pointing to the XSDT (Extended System
30 Description Table). The XSDT always points to the FADT (Fixed ACPI
31 Description Table) using its first entry, the data within the FADT
32 includes various fixed-length entries that describe fixed ACPI features
33 of the hardware. The FADT contains a pointer to the DSDT
34 (Differentiated System Description Table). The XSDT also contains
35 entries pointing to possibly multiple SSDTs (Secondary System
38 The DSDT and SSDT data is organized in data structures called definition
39 blocks that contain definitions of various objects, including ACPI
40 control methods, encoded in AML (ACPI Machine Language). The data block
41 of the DSDT along with the contents of SSDTs represents a hierarchical
42 data structure called the ACPI namespace whose topology reflects the
43 structure of the underlying hardware platform.
45 The relationships between ACPI System Definition Tables described above
46 are illustrated in the following diagram::
48 +---------+ +-------+ +--------+ +------------------------+
49 | RSDP | +->| XSDT | +->| FADT | | +-------------------+ |
50 +---------+ | +-------+ | +--------+ +-|->| DSDT | |
51 | Pointer | | | Entry |-+ | ...... | | | +-------------------+ |
52 +---------+ | +-------+ | X_DSDT |--+ | | Definition Blocks | |
53 | Pointer |-+ | ..... | | ...... | | +-------------------+ |
54 +---------+ +-------+ +--------+ | +-------------------+ |
55 | Entry |------------------|->| SSDT | |
56 +- - - -+ | +-------------------| |
57 | Entry | - - - - - - - -+ | | Definition Blocks | |
58 +- - - -+ | | +-------------------+ |
59 | | +- - - - - - - - - -+ |
61 | +-------------------+ |
62 | | Definition Blocks | |
63 | +- - - - - - - - - -+ |
64 +------------------------+
72 Figure 1. ACPI Definition Blocks
74 .. note:: RSDP can also contain a pointer to the RSDT (Root System
75 Description Table). Platforms provide RSDT to enable
76 compatibility with ACPI 1.0 operating systems. The OS is expected
77 to use XSDT, if present.
80 Example ACPI Namespace
81 ======================
83 All definition blocks are loaded into a single namespace. The namespace
84 is a hierarchy of objects identified by names and paths.
85 The following naming conventions apply to object names in the ACPI
88 1. All names are 32 bits long.
89 2. The first byte of a name must be one of 'A' - 'Z', '_'.
90 3. Each of the remaining bytes of a name must be one of 'A' - 'Z', '0'
92 4. Names starting with '_' are reserved by the ACPI specification.
93 5. The '\' symbol represents the root of the namespace (i.e. names
94 prepended with '\' are relative to the namespace root).
95 6. The '^' symbol represents the parent of the current namespace node
96 (i.e. names prepended with '^' are relative to the parent of the
97 current namespace node).
99 The figure below shows an example ACPI namespace::
106 +-| _PR | Scope(_PR): the processor namespace
110 | +-| CPU0 | Processor(CPU0): the first processor
114 +-| _SB | Scope(_SB): the system bus namespace
118 | +-| LID0 | Device(LID0); the lid device
122 | | +-| _HID | Name(_HID, "PNP0C0D"): the hardware ID
126 | | +-| _STA | Method(_STA): the status control method
130 | +-| PCI0 | Device(PCI0); the PCI root bridge
134 | +-| _HID | Name(_HID, "PNP0A08"): the hardware ID
138 | +-| _CID | Name(_CID, "PNP0A03"): the compatible ID
142 | +-| RP03 | Scope(RP03): the PCI0 power scope
146 | | +-| PXP3 | PowerResource(PXP3): the PCI0 power resource
150 | +-| GFX0 | Device(GFX0): the graphics adapter
154 | +-| _ADR | Name(_ADR, 0x00020000): the PCI bus address
158 | +-| DD01 | Device(DD01): the LCD output device
162 | +-| _BCL | Method(_BCL): the backlight control method
166 +-| _TZ | Scope(_TZ): the thermal zone namespace
170 | +-| FN00 | PowerResource(FN00): the FAN0 power resource
174 | +-| FAN0 | Device(FAN0): the FAN0 cooling device
178 | | +-| _HID | Name(_HID, "PNP0A0B"): the hardware ID
182 | +-| TZ00 | ThermalZone(TZ00); the FAN thermal zone
186 +-| _GPE | Scope(_GPE): the GPE namespace
189 Figure 2. Example ACPI Namespace
192 Linux ACPI Device Objects
193 =========================
195 The Linux kernel's core ACPI subsystem creates struct acpi_device
196 objects for ACPI namespace objects representing devices, power resources
197 processors, thermal zones. Those objects are exported to user space via
198 sysfs as directories in the subtree under /sys/devices/LNXSYSTM:00. The
199 format of their names is <bus_id:instance>, where 'bus_id' refers to the
200 ACPI namespace representation of the given object and 'instance' is used
201 for distinguishing different object of the same 'bus_id' (it is
202 two-digit decimal representation of an unsigned integer).
204 The value of 'bus_id' depends on the type of the object whose name it is
205 part of as listed in the table below::
207 +---+-----------------+-------+----------+
208 | | Object/Feature | Table | bus_id |
209 +---+-----------------+-------+----------+
210 | N | Root | xSDT | LNXSYSTM |
211 +---+-----------------+-------+----------+
212 | N | Device | xSDT | _HID |
213 +---+-----------------+-------+----------+
214 | N | Processor | xSDT | LNXCPU |
215 +---+-----------------+-------+----------+
216 | N | ThermalZone | xSDT | LNXTHERM |
217 +---+-----------------+-------+----------+
218 | N | PowerResource | xSDT | LNXPOWER |
219 +---+-----------------+-------+----------+
220 | N | Other Devices | xSDT | device |
221 +---+-----------------+-------+----------+
222 | F | PWR_BUTTON | FADT | LNXPWRBN |
223 +---+-----------------+-------+----------+
224 | F | SLP_BUTTON | FADT | LNXSLPBN |
225 +---+-----------------+-------+----------+
226 | M | Video Extension | xSDT | LNXVIDEO |
227 +---+-----------------+-------+----------+
228 | M | ATA Controller | xSDT | LNXIOBAY |
229 +---+-----------------+-------+----------+
230 | M | Docking Station | xSDT | LNXDOCK |
231 +---+-----------------+-------+----------+
233 Table 1. ACPI Namespace Objects Mapping
235 The following rules apply when creating struct acpi_device objects on
236 the basis of the contents of ACPI System Description Tables (as
237 indicated by the letter in the first column and the notation in the
238 second column of the table above):
241 The object's source is an ACPI namespace node (as indicated by the
242 named object's type in the second column). In that case the object's
243 directory in sysfs will contain the 'path' attribute whose value is
244 the full path to the node from the namespace root.
246 The struct acpi_device object is created for a fixed hardware
247 feature (as indicated by the fixed feature flag's name in the second
248 column), so its sysfs directory will not contain the 'path'
251 The struct acpi_device object is created for an ACPI namespace node
252 with specific control methods (as indicated by the ACPI defined
253 device's type in the second column). The 'path' attribute containing
254 its namespace path will be present in its sysfs directory. For
255 example, if the _BCL method is present for an ACPI namespace node, a
256 struct acpi_device object with LNXVIDEO 'bus_id' will be created for
259 The third column of the above table indicates which ACPI System
260 Description Tables contain information used for the creation of the
261 struct acpi_device objects represented by the given row (xSDT means DSDT
264 The fourth column of the above table indicates the 'bus_id' generation
265 rule of the struct acpi_device object:
268 _HID in the last column of the table means that the object's bus_id
269 is derived from the _HID/_CID identification objects present under
270 the corresponding ACPI namespace node. The object's sysfs directory
271 will then contain the 'hid' and 'modalias' attributes that can be
272 used to retrieve the _HID and _CIDs of that object.
274 The 'modalias' attribute is also present for struct acpi_device
275 objects having bus_id of the "LNXxxxxx" form (pseudo devices), in
276 which cases it contains the bus_id string itself.
278 'device' in the last column of the table indicates that the object's
279 bus_id cannot be determined from _HID/_CID of the corresponding
280 ACPI namespace node, although that object represents a device (for
281 example, it may be a PCI device with _ADR defined and without _HID
282 or _CID). In that case the string 'device' will be used as the
286 Linux ACPI Physical Device Glue
287 ===============================
289 ACPI device (i.e. struct acpi_device) objects may be linked to other
290 objects in the Linux' device hierarchy that represent "physical" devices
291 (for example, devices on the PCI bus). If that happens, it means that
292 the ACPI device object is a "companion" of a device otherwise
293 represented in a different way and is used (1) to provide configuration
294 information on that device which cannot be obtained by other means and
295 (2) to do specific things to the device with the help of its ACPI
296 control methods. One ACPI device object may be linked this way to
297 multiple "physical" devices.
299 If an ACPI device object is linked to a "physical" device, its sysfs
300 directory contains the "physical_node" symbolic link to the sysfs
301 directory of the target device object. In turn, the target device's
302 sysfs directory will then contain the "firmware_node" symbolic link to
303 the sysfs directory of the companion ACPI device object.
304 The linking mechanism relies on device identification provided by the
305 ACPI namespace. For example, if there's an ACPI namespace object
306 representing a PCI device (i.e. a device object under an ACPI namespace
307 object representing a PCI bridge) whose _ADR returns 0x00020000 and the
308 bus number of the parent PCI bridge is 0, the sysfs directory
309 representing the struct acpi_device object created for that ACPI
310 namespace object will contain the 'physical_node' symbolic link to the
311 /sys/devices/pci0000:00/0000:00:02:0/ sysfs directory of the
312 corresponding PCI device.
314 The linking mechanism is generally bus-specific. The core of its
315 implementation is located in the drivers/acpi/glue.c file, but there are
316 complementary parts depending on the bus types in question located
317 elsewhere. For example, the PCI-specific part of it is located in
318 drivers/pci/pci-acpi.c.
321 Example Linux ACPI Device Tree
322 =================================
324 The sysfs hierarchy of struct acpi_device objects corresponding to the
325 example ACPI namespace illustrated in Figure 2 with the addition of
326 fixed PWR_BUTTON/SLP_BUTTON devices is shown below::
328 +--------------+---+-----------------+
329 | LNXSYSTEM:00 | \ | acpi:LNXSYSTEM: |
330 +--------------+---+-----------------+
332 | +-------------+-----+----------------+
333 +-| LNXPWRBN:00 | N/A | acpi:LNXPWRBN: |
334 | +-------------+-----+----------------+
336 | +-------------+-----+----------------+
337 +-| LNXSLPBN:00 | N/A | acpi:LNXSLPBN: |
338 | +-------------+-----+----------------+
340 | +-----------+------------+--------------+
341 +-| LNXCPU:00 | \_PR_.CPU0 | acpi:LNXCPU: |
342 | +-----------+------------+--------------+
344 | +-------------+-------+----------------+
345 +-| LNXSYBUS:00 | \_SB_ | acpi:LNXSYBUS: |
346 | +-------------+-------+----------------+
348 | | +- - - - - - - +- - - - - - +- - - - - - - -+
349 | +-| PNP0C0D:00 | \_SB_.LID0 | acpi:PNP0C0D: |
350 | | +- - - - - - - +- - - - - - +- - - - - - - -+
352 | | +------------+------------+-----------------------+
353 | +-| PNP0A08:00 | \_SB_.PCI0 | acpi:PNP0A08:PNP0A03: |
354 | +------------+------------+-----------------------+
356 | | +-----------+-----------------+-----+
357 | +-| device:00 | \_SB_.PCI0.RP03 | N/A |
358 | | +-----------+-----------------+-----+
360 | | | +-------------+----------------------+----------------+
361 | | +-| LNXPOWER:00 | \_SB_.PCI0.RP03.PXP3 | acpi:LNXPOWER: |
362 | | +-------------+----------------------+----------------+
364 | | +-------------+-----------------+----------------+
365 | +-| LNXVIDEO:00 | \_SB_.PCI0.GFX0 | acpi:LNXVIDEO: |
366 | +-------------+-----------------+----------------+
368 | | +-----------+-----------------+-----+
369 | +-| device:01 | \_SB_.PCI0.DD01 | N/A |
370 | +-----------+-----------------+-----+
372 | +-------------+-------+----------------+
373 +-| LNXSYBUS:01 | \_TZ_ | acpi:LNXSYBUS: |
374 +-------------+-------+----------------+
376 | +-------------+------------+----------------+
377 +-| LNXPOWER:0a | \_TZ_.FN00 | acpi:LNXPOWER: |
378 | +-------------+------------+----------------+
380 | +------------+------------+---------------+
381 +-| PNP0C0B:00 | \_TZ_.FAN0 | acpi:PNP0C0B: |
382 | +------------+------------+---------------+
384 | +-------------+------------+----------------+
385 +-| LNXTHERM:00 | \_TZ_.TZ00 | acpi:LNXTHERM: |
386 +-------------+------------+----------------+
388 Figure 3. Example Linux ACPI Device Tree
390 .. note:: Each node is represented as "object/path/modalias", where:
392 1. 'object' is the name of the object's directory in sysfs.
393 2. 'path' is the ACPI namespace path of the corresponding
394 ACPI namespace object, as returned by the object's 'path'
396 3. 'modalias' is the value of the object's 'modalias' sysfs
397 attribute (as described earlier in this document).
399 .. note:: N/A indicates the device object does not have the 'path' or the
400 'modalias' attribute.