xref: /linux/Documentation/ABI/testing/sysfs-firmware-acpi (revision ec8a42e7343234802b9054874fe01810880289ce)
1What:		/sys/firmware/acpi/bgrt/
2Date:		January 2012
3Contact:	Matthew Garrett <mjg@redhat.com>
4Description:
5		The BGRT is an ACPI 5.0 feature that allows the OS
6		to obtain a copy of the firmware boot splash and
7		some associated metadata. This is intended to be used
8		by boot splash applications in order to interact with
9		the firmware boot splash in order to avoid jarring
10		transitions.
11
12		image: The image bitmap. Currently a 32-bit BMP.
13		status: 1 if the image is valid, 0 if firmware invalidated it.
14		type: 0 indicates image is in BMP format.
15
16		======== ===================================================
17		version: The version of the BGRT. Currently 1.
18		xoffset: The number of pixels between the left of the screen
19			 and the left edge of the image.
20		yoffset: The number of pixels between the top of the screen
21			 and the top edge of the image.
22		======== ===================================================
23
24What:		/sys/firmware/acpi/hotplug/
25Date:		February 2013
26Contact:	Rafael J. Wysocki <rafael.j.wysocki@intel.com>
27Description:
28		There are separate hotplug profiles for different classes of
29		devices supported by ACPI, such as containers, memory modules,
30		processors, PCI root bridges etc.  A hotplug profile for a given
31		class of devices is a collection of settings defining the way
32		that class of devices will be handled by the ACPI core hotplug
33		code.  Those profiles are represented in sysfs as subdirectories
34		of /sys/firmware/acpi/hotplug/.
35
36		The following setting is available to user space for each
37		hotplug profile:
38
39		======== =======================================================
40		enabled: If set, the ACPI core will handle notifications of
41			 hotplug events associated with the given class of
42			 devices and will allow those devices to be ejected with
43			 the help of the _EJ0 control method.  Unsetting it
44			 effectively disables hotplug for the correspoinding
45			 class of devices.
46		======== =======================================================
47
48		The value of the above attribute is an integer number: 1 (set)
49		or 0 (unset).  Attempts to write any other values to it will
50		cause -EINVAL to be returned.
51
52What:		/sys/firmware/acpi/interrupts/
53Date:		February 2008
54Contact:	Len Brown <lenb@kernel.org>
55Description:
56		All ACPI interrupts are handled via a single IRQ,
57		the System Control Interrupt (SCI), which appears
58		as "acpi" in /proc/interrupts.
59
60		However, one of the main functions of ACPI is to make
61		the platform understand random hardware without
62		special driver support.  So while the SCI handles a few
63		well known (fixed feature) interrupts sources, such
64		as the power button, it can also handle a variable
65		number of a "General Purpose Events" (GPE).
66
67		A GPE vectors to a specified handler in AML, which
68		can do a anything the BIOS writer wants from
69		OS context.  GPE 0x12, for example, would vector
70		to a level or edge handler called _L12 or _E12.
71		The handler may do its business and return.
72		Or the handler may send send a Notify event
73		to a Linux device driver registered on an ACPI device,
74		such as a battery, or a processor.
75
76		To figure out where all the SCI's are coming from,
77		/sys/firmware/acpi/interrupts contains a file listing
78		every possible source, and the count of how many
79		times it has triggered::
80
81		  $ cd /sys/firmware/acpi/interrupts
82		  $ grep . *
83		  error:	     0
84		  ff_gbl_lock:	     0   enable
85		  ff_pmtimer:	     0  invalid
86		  ff_pwr_btn:	     0   enable
87		  ff_rt_clk:	     2  disable
88		  ff_slp_btn:	     0  invalid
89		  gpe00:	     0	invalid
90		  gpe01:	     0	 enable
91		  gpe02:	   108	 enable
92		  gpe03:	     0	invalid
93		  gpe04:	     0	invalid
94		  gpe05:	     0	invalid
95		  gpe06:	     0	 enable
96		  gpe07:	     0	 enable
97		  gpe08:	     0	invalid
98		  gpe09:	     0	invalid
99		  gpe0A:	     0	invalid
100		  gpe0B:	     0	invalid
101		  gpe0C:	     0	invalid
102		  gpe0D:	     0	invalid
103		  gpe0E:	     0	invalid
104		  gpe0F:	     0	invalid
105		  gpe10:	     0	invalid
106		  gpe11:	     0	invalid
107		  gpe12:	     0	invalid
108		  gpe13:	     0	invalid
109		  gpe14:	     0	invalid
110		  gpe15:	     0	invalid
111		  gpe16:	     0	invalid
112		  gpe17:	  1084	 enable
113		  gpe18:	     0	 enable
114		  gpe19:	     0	invalid
115		  gpe1A:	     0	invalid
116		  gpe1B:	     0	invalid
117		  gpe1C:	     0	invalid
118		  gpe1D:	     0	invalid
119		  gpe1E:	     0	invalid
120		  gpe1F:	     0	invalid
121		  gpe_all:	  1192
122		  sci:		  1194
123		  sci_not:	     0
124
125		===========  ==================================================
126		sci	     The number of times the ACPI SCI
127			     has been called and claimed an interrupt.
128
129		sci_not	     The number of times the ACPI SCI
130			     has been called and NOT claimed an interrupt.
131
132		gpe_all	     count of SCI caused by GPEs.
133
134		gpeXX	     count for individual GPE source
135
136		ff_gbl_lock  Global Lock
137
138		ff_pmtimer   PM Timer
139
140		ff_pwr_btn   Power Button
141
142		ff_rt_clk    Real Time Clock
143
144		ff_slp_btn   Sleep Button
145
146		error	     an interrupt that can't be accounted for above.
147
148		invalid      it's either a GPE or a Fixed Event that
149			     doesn't have an event handler.
150
151		disable	     the GPE/Fixed Event is valid but disabled.
152
153		enable       the GPE/Fixed Event is valid and enabled.
154		===========  ==================================================
155
156		Root has permission to clear any of these counters.  Eg.::
157
158		  # echo 0 > gpe11
159
160		All counters can be cleared by clearing the total "sci"::
161
162		  # echo 0 > sci
163
164		None of these counters has an effect on the function
165		of the system, they are simply statistics.
166
167		Besides this, user can also write specific strings to these files
168		to enable/disable/clear ACPI interrupts in user space, which can be
169		used to debug some ACPI interrupt storm issues.
170
171		Note that only writing to VALID GPE/Fixed Event is allowed,
172		i.e. user can only change the status of runtime GPE and
173		Fixed Event with event handler installed.
174
175		Let's take power button fixed event for example, please kill acpid
176		and other user space applications so that the machine won't shutdown
177		when pressing the power button::
178
179		  # cat ff_pwr_btn
180		  0	enabled
181		  # press the power button for 3 times;
182		  # cat ff_pwr_btn
183		  3	enabled
184		  # echo disable > ff_pwr_btn
185		  # cat ff_pwr_btn
186		  3	disabled
187		  # press the power button for 3 times;
188		  # cat ff_pwr_btn
189		  3	disabled
190		  # echo enable > ff_pwr_btn
191		  # cat ff_pwr_btn
192		  4	enabled
193		  /*
194		   * this is because the status bit is set even if the enable
195		   * bit is cleared, and it triggers an ACPI fixed event when
196		   * the enable bit is set again
197		   */
198		  # press the power button for 3 times;
199		  # cat ff_pwr_btn
200		  7	enabled
201		  # echo disable > ff_pwr_btn
202		  # press the power button for 3 times;
203		  # echo clear > ff_pwr_btn	/* clear the status bit */
204		  # echo disable > ff_pwr_btn
205		  # cat ff_pwr_btn
206		  7	enabled
207
208