1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 .. include:: <isonum.txt> 2 .. include:: <isonum.txt>
3 3
4 =================== 4 ===================
5 System Sleep States 5 System Sleep States
6 =================== 6 ===================
7 7
8 :Copyright: |copy| 2017 Intel Corporation 8 :Copyright: |copy| 2017 Intel Corporation
9 9
10 :Author: Rafael J. Wysocki <rafael.j.wysocki@in 10 :Author: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
11 11
12 12
13 Sleep states are global low-power states of th 13 Sleep states are global low-power states of the entire system in which user
14 space code cannot be executed and the overall 14 space code cannot be executed and the overall system activity is significantly
15 reduced. 15 reduced.
16 16
17 17
18 Sleep States That Can Be Supported 18 Sleep States That Can Be Supported
19 ================================== 19 ==================================
20 20
21 Depending on its configuration and the capabil 21 Depending on its configuration and the capabilities of the platform it runs on,
22 the Linux kernel can support up to four system 22 the Linux kernel can support up to four system sleep states, including
23 hibernation and up to three variants of system 23 hibernation and up to three variants of system suspend. The sleep states that
24 can be supported by the kernel are listed belo 24 can be supported by the kernel are listed below.
25 25
26 .. _s2idle: 26 .. _s2idle:
27 27
28 Suspend-to-Idle 28 Suspend-to-Idle
29 --------------- 29 ---------------
30 30
31 This is a generic, pure software, light-weight 31 This is a generic, pure software, light-weight variant of system suspend (also
32 referred to as S2I or S2Idle). It allows more 32 referred to as S2I or S2Idle). It allows more energy to be saved relative to
33 runtime idle by freezing user space, suspendin 33 runtime idle by freezing user space, suspending the timekeeping and putting all
34 I/O devices into low-power states (possibly lo 34 I/O devices into low-power states (possibly lower-power than available in the
35 working state), such that the processors can s 35 working state), such that the processors can spend time in their deepest idle
36 states while the system is suspended. 36 states while the system is suspended.
37 37
38 The system is woken up from this state by in-b 38 The system is woken up from this state by in-band interrupts, so theoretically
39 any devices that can cause interrupts to be ge 39 any devices that can cause interrupts to be generated in the working state can
40 also be set up as wakeup devices for S2Idle. 40 also be set up as wakeup devices for S2Idle.
41 41
42 This state can be used on platforms without su 42 This state can be used on platforms without support for :ref:`standby <standby>`
43 or :ref:`suspend-to-RAM <s2ram>`, or it can be 43 or :ref:`suspend-to-RAM <s2ram>`, or it can be used in addition to any of the
44 deeper system suspend variants to provide redu 44 deeper system suspend variants to provide reduced resume latency. It is always
45 supported if the :c:macro:`CONFIG_SUSPEND` ker 45 supported if the :c:macro:`CONFIG_SUSPEND` kernel configuration option is set.
46 46
47 .. _standby: 47 .. _standby:
48 48
49 Standby 49 Standby
50 ------- 50 -------
51 51
52 This state, if supported, offers moderate, but 52 This state, if supported, offers moderate, but real, energy savings, while
53 providing a relatively straightforward transit 53 providing a relatively straightforward transition back to the working state. No
54 operating state is lost (the system core logic 54 operating state is lost (the system core logic retains power), so the system can
55 go back to where it left off easily enough. 55 go back to where it left off easily enough.
56 56
57 In addition to freezing user space, suspending 57 In addition to freezing user space, suspending the timekeeping and putting all
58 I/O devices into low-power states, which is do 58 I/O devices into low-power states, which is done for :ref:`suspend-to-idle
59 <s2idle>` too, nonboot CPUs are taken offline 59 <s2idle>` too, nonboot CPUs are taken offline and all low-level system functions
60 are suspended during transitions into this sta 60 are suspended during transitions into this state. For this reason, it should
61 allow more energy to be saved relative to :ref 61 allow more energy to be saved relative to :ref:`suspend-to-idle <s2idle>`, but
62 the resume latency will generally be greater t 62 the resume latency will generally be greater than for that state.
63 63
64 The set of devices that can wake up the system 64 The set of devices that can wake up the system from this state usually is
65 reduced relative to :ref:`suspend-to-idle <s2i 65 reduced relative to :ref:`suspend-to-idle <s2idle>` and it may be necessary to
66 rely on the platform for setting up the wakeup 66 rely on the platform for setting up the wakeup functionality as appropriate.
67 67
68 This state is supported if the :c:macro:`CONFI 68 This state is supported if the :c:macro:`CONFIG_SUSPEND` kernel configuration
69 option is set and the support for it is regist 69 option is set and the support for it is registered by the platform with the
70 core system suspend subsystem. On ACPI-based 70 core system suspend subsystem. On ACPI-based systems this state is mapped to
71 the S1 system state defined by ACPI. 71 the S1 system state defined by ACPI.
72 72
73 .. _s2ram: 73 .. _s2ram:
74 74
75 Suspend-to-RAM 75 Suspend-to-RAM
76 -------------- 76 --------------
77 77
78 This state (also referred to as STR or S2RAM), 78 This state (also referred to as STR or S2RAM), if supported, offers significant
79 energy savings as everything in the system is 79 energy savings as everything in the system is put into a low-power state, except
80 for memory, which should be placed into the se 80 for memory, which should be placed into the self-refresh mode to retain its
81 contents. All of the steps carried out when e 81 contents. All of the steps carried out when entering :ref:`standby <standby>`
82 are also carried out during transitions to S2R 82 are also carried out during transitions to S2RAM. Additional operations may
83 take place depending on the platform capabilit 83 take place depending on the platform capabilities. In particular, on ACPI-based
84 systems the kernel passes control to the platf 84 systems the kernel passes control to the platform firmware (BIOS) as the last
85 step during S2RAM transitions and that usually 85 step during S2RAM transitions and that usually results in powering down some
86 more low-level components that are not directl 86 more low-level components that are not directly controlled by the kernel.
87 87
88 The state of devices and CPUs is saved and hel 88 The state of devices and CPUs is saved and held in memory. All devices are
89 suspended and put into low-power states. In m 89 suspended and put into low-power states. In many cases, all peripheral buses
90 lose power when entering S2RAM, so devices mus 90 lose power when entering S2RAM, so devices must be able to handle the transition
91 back to the "on" state. 91 back to the "on" state.
92 92
93 On ACPI-based systems S2RAM requires some mini 93 On ACPI-based systems S2RAM requires some minimal boot-strapping code in the
94 platform firmware to resume the system from it 94 platform firmware to resume the system from it. This may be the case on other
95 platforms too. 95 platforms too.
96 96
97 The set of devices that can wake up the system 97 The set of devices that can wake up the system from S2RAM usually is reduced
98 relative to :ref:`suspend-to-idle <s2idle>` an 98 relative to :ref:`suspend-to-idle <s2idle>` and :ref:`standby <standby>` and it
99 may be necessary to rely on the platform for s 99 may be necessary to rely on the platform for setting up the wakeup functionality
100 as appropriate. 100 as appropriate.
101 101
102 S2RAM is supported if the :c:macro:`CONFIG_SUS 102 S2RAM is supported if the :c:macro:`CONFIG_SUSPEND` kernel configuration option
103 is set and the support for it is registered by 103 is set and the support for it is registered by the platform with the core system
104 suspend subsystem. On ACPI-based systems it i 104 suspend subsystem. On ACPI-based systems it is mapped to the S3 system state
105 defined by ACPI. 105 defined by ACPI.
106 106
107 .. _hibernation: 107 .. _hibernation:
108 108
109 Hibernation 109 Hibernation
110 ----------- 110 -----------
111 111
112 This state (also referred to as Suspend-to-Dis 112 This state (also referred to as Suspend-to-Disk or STD) offers the greatest
113 energy savings and can be used even in the abs 113 energy savings and can be used even in the absence of low-level platform support
114 for system suspend. However, it requires some 114 for system suspend. However, it requires some low-level code for resuming the
115 system to be present for the underlying CPU ar 115 system to be present for the underlying CPU architecture.
116 116
117 Hibernation is significantly different from an 117 Hibernation is significantly different from any of the system suspend variants.
118 It takes three system state changes to put it 118 It takes three system state changes to put it into hibernation and two system
119 state changes to resume it. 119 state changes to resume it.
120 120
121 First, when hibernation is triggered, the kern 121 First, when hibernation is triggered, the kernel stops all system activity and
122 creates a snapshot image of memory to be writt 122 creates a snapshot image of memory to be written into persistent storage. Next,
123 the system goes into a state in which the snap 123 the system goes into a state in which the snapshot image can be saved, the image
124 is written out and finally the system goes int 124 is written out and finally the system goes into the target low-power state in
125 which power is cut from almost all of its hard 125 which power is cut from almost all of its hardware components, including memory,
126 except for a limited set of wakeup devices. 126 except for a limited set of wakeup devices.
127 127
128 Once the snapshot image has been written out, 128 Once the snapshot image has been written out, the system may either enter a
129 special low-power state (like ACPI S4), or it 129 special low-power state (like ACPI S4), or it may simply power down itself.
130 Powering down means minimum power draw and it 130 Powering down means minimum power draw and it allows this mechanism to work on
131 any system. However, entering a special low-p 131 any system. However, entering a special low-power state may allow additional
132 means of system wakeup to be used (e.g. press 132 means of system wakeup to be used (e.g. pressing a key on the keyboard or
133 opening a laptop lid). 133 opening a laptop lid).
134 134
135 After wakeup, control goes to the platform fir 135 After wakeup, control goes to the platform firmware that runs a boot loader
136 which boots a fresh instance of the kernel (co 136 which boots a fresh instance of the kernel (control may also go directly to
137 the boot loader, depending on the system confi 137 the boot loader, depending on the system configuration, but anyway it causes
138 a fresh instance of the kernel to be booted). 138 a fresh instance of the kernel to be booted). That new instance of the kernel
139 (referred to as the ``restore kernel``) looks 139 (referred to as the ``restore kernel``) looks for a hibernation image in
140 persistent storage and if one is found, it is 140 persistent storage and if one is found, it is loaded into memory. Next, all
141 activity in the system is stopped and the rest 141 activity in the system is stopped and the restore kernel overwrites itself with
142 the image contents and jumps into a special tr 142 the image contents and jumps into a special trampoline area in the original
143 kernel stored in the image (referred to as the 143 kernel stored in the image (referred to as the ``image kernel``), which is where
144 the special architecture-specific low-level co 144 the special architecture-specific low-level code is needed. Finally, the
145 image kernel restores the system to the pre-hi 145 image kernel restores the system to the pre-hibernation state and allows user
146 space to run again. 146 space to run again.
147 147
148 Hibernation is supported if the :c:macro:`CONF 148 Hibernation is supported if the :c:macro:`CONFIG_HIBERNATION` kernel
149 configuration option is set. However, this op 149 configuration option is set. However, this option can only be set if support
150 for the given CPU architecture includes the lo 150 for the given CPU architecture includes the low-level code for system resume.
151 151
152 152
153 Basic ``sysfs`` Interfaces for System Suspend 153 Basic ``sysfs`` Interfaces for System Suspend and Hibernation
154 ============================================== 154 =============================================================
155 155
156 The power management subsystem provides usersp 156 The power management subsystem provides userspace with a unified ``sysfs``
157 interface for system sleep regardless of the u 157 interface for system sleep regardless of the underlying system architecture or
158 platform. That interface is located in the :f 158 platform. That interface is located in the :file:`/sys/power/` directory
159 (assuming that ``sysfs`` is mounted at :file:` 159 (assuming that ``sysfs`` is mounted at :file:`/sys`) and it consists of the
160 following attributes (files): 160 following attributes (files):
161 161
162 ``state`` 162 ``state``
163 This file contains a list of strings r 163 This file contains a list of strings representing sleep states supported
164 by the kernel. Writing one of these s 164 by the kernel. Writing one of these strings into it causes the kernel
165 to start a transition of the system in 165 to start a transition of the system into the sleep state represented by
166 that string. 166 that string.
167 167
168 In particular, the "disk", "freeze" an 168 In particular, the "disk", "freeze" and "standby" strings represent the
169 :ref:`hibernation <hibernation>`, :ref 169 :ref:`hibernation <hibernation>`, :ref:`suspend-to-idle <s2idle>` and
170 :ref:`standby <standby>` sleep states, 170 :ref:`standby <standby>` sleep states, respectively. The "mem" string
171 is interpreted in accordance with the 171 is interpreted in accordance with the contents of the ``mem_sleep`` file
172 described below. 172 described below.
173 173
174 If the kernel does not support any sys 174 If the kernel does not support any system sleep states, this file is
175 not present. 175 not present.
176 176
177 ``mem_sleep`` 177 ``mem_sleep``
178 This file contains a list of strings r 178 This file contains a list of strings representing supported system
179 suspend variants and allows user space 179 suspend variants and allows user space to select the variant to be
180 associated with the "mem" string in th 180 associated with the "mem" string in the ``state`` file described above.
181 181
182 The strings that may be present in thi 182 The strings that may be present in this file are "s2idle", "shallow"
183 and "deep". The "s2idle" string alway 183 and "deep". The "s2idle" string always represents :ref:`suspend-to-idle
184 <s2idle>` and, by convention, "shallow 184 <s2idle>` and, by convention, "shallow" and "deep" represent
185 :ref:`standby <standby>` and :ref:`sus 185 :ref:`standby <standby>` and :ref:`suspend-to-RAM <s2ram>`,
186 respectively. 186 respectively.
187 187
188 Writing one of the listed strings into 188 Writing one of the listed strings into this file causes the system
189 suspend variant represented by it to b 189 suspend variant represented by it to be associated with the "mem" string
190 in the ``state`` file. The string rep 190 in the ``state`` file. The string representing the suspend variant
191 currently associated with the "mem" st 191 currently associated with the "mem" string in the ``state`` file is
192 shown in square brackets. 192 shown in square brackets.
193 193
194 If the kernel does not support system 194 If the kernel does not support system suspend, this file is not present.
195 195
196 ``disk`` 196 ``disk``
197 This file controls the operating mode 197 This file controls the operating mode of hibernation (Suspend-to-Disk).
198 Specifically, it tells the kernel what 198 Specifically, it tells the kernel what to do after creating a
199 hibernation image. 199 hibernation image.
200 200
201 Reading from it returns a list of supp 201 Reading from it returns a list of supported options encoded as:
202 202
203 ``platform`` 203 ``platform``
204 Put the system into a special 204 Put the system into a special low-power state (e.g. ACPI S4) to
205 make additional wakeup options 205 make additional wakeup options available and possibly allow the
206 platform firmware to take a si 206 platform firmware to take a simplified initialization path after
207 wakeup. 207 wakeup.
208 208
209 It is only available if the pl 209 It is only available if the platform provides a special
210 mechanism to put the system to 210 mechanism to put the system to sleep after creating a
211 hibernation image (platforms w 211 hibernation image (platforms with ACPI do that as a rule, for
212 example). 212 example).
213 213
214 ``shutdown`` 214 ``shutdown``
215 Power off the system. 215 Power off the system.
216 216
217 ``reboot`` 217 ``reboot``
218 Reboot the system (useful for 218 Reboot the system (useful for diagnostics mostly).
219 219
220 ``suspend`` 220 ``suspend``
221 Hybrid system suspend. Put th 221 Hybrid system suspend. Put the system into the suspend sleep
222 state selected through the ``m 222 state selected through the ``mem_sleep`` file described above.
223 If the system is successfully 223 If the system is successfully woken up from that state, discard
224 the hibernation image and cont 224 the hibernation image and continue. Otherwise, use the image
225 to restore the previous state 225 to restore the previous state of the system.
226 226
227 It is available if system susp 227 It is available if system suspend is supported.
228 228
229 ``test_resume`` 229 ``test_resume``
230 Diagnostic operation. Load th 230 Diagnostic operation. Load the image as though the system had
231 just woken up from hibernation 231 just woken up from hibernation and the currently running kernel
232 instance was a restore kernel 232 instance was a restore kernel and follow up with full system
233 resume. 233 resume.
234 234
235 Writing one of the strings listed abov 235 Writing one of the strings listed above into this file causes the option
236 represented by it to be selected. 236 represented by it to be selected.
237 237
238 The currently selected option is shown 238 The currently selected option is shown in square brackets, which means
239 that the operation represented by it w 239 that the operation represented by it will be carried out after creating
240 and saving the image when hibernation 240 and saving the image when hibernation is triggered by writing ``disk``
241 to :file:`/sys/power/state`. 241 to :file:`/sys/power/state`.
242 242
243 If the kernel does not support hiberna 243 If the kernel does not support hibernation, this file is not present.
244 244
245 ``image_size`` 245 ``image_size``
246 This file controls the size of hiberna 246 This file controls the size of hibernation images.
247 247
248 It can be written a string representin 248 It can be written a string representing a non-negative integer that will
249 be used as a best-effort upper limit o 249 be used as a best-effort upper limit of the image size, in bytes. The
250 hibernation core will do its best to e 250 hibernation core will do its best to ensure that the image size will not
251 exceed that number, but if that turns 251 exceed that number, but if that turns out to be impossible to achieve, a
252 hibernation image will still be create 252 hibernation image will still be created and its size will be as small as
253 possible. In particular, writing '0' 253 possible. In particular, writing '0' to this file causes the size of
254 hibernation images to be minimum. 254 hibernation images to be minimum.
255 255
256 Reading from it returns the current im 256 Reading from it returns the current image size limit, which is set to
257 around 2/5 of the available RAM size b 257 around 2/5 of the available RAM size by default.
258 258
259 ``pm_trace`` 259 ``pm_trace``
260 This file controls the "PM trace" mech 260 This file controls the "PM trace" mechanism saving the last suspend
261 or resume event point in the RTC memor 261 or resume event point in the RTC memory across reboots. It helps to
262 debug hard lockups or reboots due to d 262 debug hard lockups or reboots due to device driver failures that occur
263 during system suspend or resume (which 263 during system suspend or resume (which is more common) more effectively.
264 264
265 If it contains "1", the fingerprint of 265 If it contains "1", the fingerprint of each suspend/resume event point
266 in turn will be stored in the RTC memo 266 in turn will be stored in the RTC memory (overwriting the actual RTC
267 information), so it will survive a sys 267 information), so it will survive a system crash if one occurs right
268 after storing it and it can be used la 268 after storing it and it can be used later to identify the driver that
269 caused the crash to happen. 269 caused the crash to happen.
270 270
271 It contains "0" by default, which may 271 It contains "0" by default, which may be changed to "1" by writing a
272 string representing a nonzero integer 272 string representing a nonzero integer into it.
273 273
274 According to the above, there are two ways to 274 According to the above, there are two ways to make the system go into the
275 :ref:`suspend-to-idle <s2idle>` state. The fi 275 :ref:`suspend-to-idle <s2idle>` state. The first one is to write "freeze"
276 directly to :file:`/sys/power/state`. The sec 276 directly to :file:`/sys/power/state`. The second one is to write "s2idle" to
277 :file:`/sys/power/mem_sleep` and then to write 277 :file:`/sys/power/mem_sleep` and then to write "mem" to
278 :file:`/sys/power/state`. Likewise, there are 278 :file:`/sys/power/state`. Likewise, there are two ways to make the system go
279 into the :ref:`standby <standby>` state (the s 279 into the :ref:`standby <standby>` state (the strings to write to the control
280 files in that case are "standby" or "shallow" 280 files in that case are "standby" or "shallow" and "mem", respectively) if that
281 state is supported by the platform. However, 281 state is supported by the platform. However, there is only one way to make the
282 system go into the :ref:`suspend-to-RAM <s2ram 282 system go into the :ref:`suspend-to-RAM <s2ram>` state (write "deep" into
283 :file:`/sys/power/mem_sleep` and "mem" into :f 283 :file:`/sys/power/mem_sleep` and "mem" into :file:`/sys/power/state`).
284 284
285 The default suspend variant (ie. the one to be 285 The default suspend variant (ie. the one to be used without writing anything
286 into :file:`/sys/power/mem_sleep`) is either " 286 into :file:`/sys/power/mem_sleep`) is either "deep" (on the majority of systems
287 supporting :ref:`suspend-to-RAM <s2ram>`) or " 287 supporting :ref:`suspend-to-RAM <s2ram>`) or "s2idle", but it can be overridden
288 by the value of the ``mem_sleep_default`` para 288 by the value of the ``mem_sleep_default`` parameter in the kernel command line.
289 On some systems with ACPI, depending on the in 289 On some systems with ACPI, depending on the information in the ACPI tables, the
290 default may be "s2idle" even if :ref:`suspend- 290 default may be "s2idle" even if :ref:`suspend-to-RAM <s2ram>` is supported in
291 principle. 291 principle.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.