1Debugging hibernation and suspend
2	(C) 2007 Rafael J. Wysocki <rjw@sisk.pl>, GPL
3
41. Testing hibernation (aka suspend to disk or STD)
5
6To check if hibernation works, you can try to hibernate in the "reboot" mode:
7
8# echo reboot > /sys/power/disk
9# echo disk > /sys/power/state
10
11and the system should create a hibernation image, reboot, resume and get back to
12the command prompt where you have started the transition.  If that happens,
13hibernation is most likely to work correctly.  Still, you need to repeat the
14test at least a couple of times in a row for confidence.  [This is necessary,
15because some problems only show up on a second attempt at suspending and
16resuming the system.]  Moreover, hibernating in the "reboot" and "shutdown"
17modes causes the PM core to skip some platform-related callbacks which on ACPI
18systems might be necessary to make hibernation work.  Thus, if your machine fails
19to hibernate or resume in the "reboot" mode, you should try the "platform" mode:
20
21# echo platform > /sys/power/disk
22# echo disk > /sys/power/state
23
24which is the default and recommended mode of hibernation.
25
26Unfortunately, the "platform" mode of hibernation does not work on some systems
27with broken BIOSes.  In such cases the "shutdown" mode of hibernation might
28work:
29
30# echo shutdown > /sys/power/disk
31# echo disk > /sys/power/state
32
33(it is similar to the "reboot" mode, but it requires you to press the power
34button to make the system resume).
35
36If neither "platform" nor "shutdown" hibernation mode works, you will need to
37identify what goes wrong.
38
39a) Test modes of hibernation
40
41To find out why hibernation fails on your system, you can use a special testing
42facility available if the kernel is compiled with CONFIG_PM_DEBUG set.  Then,
43there is the file /sys/power/pm_test that can be used to make the hibernation
44core run in a test mode.  There are 5 test modes available:
45
46freezer
47- test the freezing of processes
48
49devices
50- test the freezing of processes and suspending of devices
51
52platform
53- test the freezing of processes, suspending of devices and platform
54  global control methods(*)
55
56processors
57- test the freezing of processes, suspending of devices, platform
58  global control methods(*) and the disabling of nonboot CPUs
59
60core
61- test the freezing of processes, suspending of devices, platform global
62  control methods(*), the disabling of nonboot CPUs and suspending of
63  platform/system devices
64
65(*) the platform global control methods are only available on ACPI systems
66    and are only tested if the hibernation mode is set to "platform"
67
68To use one of them it is necessary to write the corresponding string to
69/sys/power/pm_test (eg. "devices" to test the freezing of processes and
70suspending devices) and issue the standard hibernation commands.  For example,
71to use the "devices" test mode along with the "platform" mode of hibernation,
72you should do the following:
73
74# echo devices > /sys/power/pm_test
75# echo platform > /sys/power/disk
76# echo disk > /sys/power/state
77
78Then, the kernel will try to freeze processes, suspend devices, wait 5 seconds,
79resume devices and thaw processes.  If "platform" is written to
80/sys/power/pm_test , then after suspending devices the kernel will additionally
81invoke the global control methods (eg. ACPI global control methods) used to
82prepare the platform firmware for hibernation.  Next, it will wait 5 seconds and
83invoke the platform (eg. ACPI) global methods used to cancel hibernation etc.
84
85Writing "none" to /sys/power/pm_test causes the kernel to switch to the normal
86hibernation/suspend operations.  Also, when open for reading, /sys/power/pm_test
87contains a space-separated list of all available tests (including "none" that
88represents the normal functionality) in which the current test level is
89indicated by square brackets.
90
91Generally, as you can see, each test level is more "invasive" than the previous
92one and the "core" level tests the hardware and drivers as deeply as possible
93without creating a hibernation image.  Obviously, if the "devices" test fails,
94the "platform" test will fail as well and so on.  Thus, as a rule of thumb, you
95should try the test modes starting from "freezer", through "devices", "platform"
96and "processors" up to "core" (repeat the test on each level a couple of times
97to make sure that any random factors are avoided).
98
99If the "freezer" test fails, there is a task that cannot be frozen (in that case
100it usually is possible to identify the offending task by analysing the output of
101dmesg obtained after the failing test).  Failure at this level usually means
102that there is a problem with the tasks freezer subsystem that should be
103reported.
104
105If the "devices" test fails, most likely there is a driver that cannot suspend
106or resume its device (in the latter case the system may hang or become unstable
107after the test, so please take that into consideration).  To find this driver,
108you can carry out a binary search according to the rules:
109- if the test fails, unload a half of the drivers currently loaded and repeat
110(that would probably involve rebooting the system, so always note what drivers
111have been loaded before the test),
112- if the test succeeds, load a half of the drivers you have unloaded most
113recently and repeat.
114
115Once you have found the failing driver (there can be more than just one of
116them), you have to unload it every time before hibernation.  In that case please
117make sure to report the problem with the driver.
118
119It is also possible that the "devices" test will still fail after you have
120unloaded all modules. In that case, you may want to look in your kernel
121configuration for the drivers that can be compiled as modules (and test again
122with these drivers compiled as modules).  You may also try to use some special
123kernel command line options such as "noapic", "noacpi" or even "acpi=off".
124
125If the "platform" test fails, there is a problem with the handling of the
126platform (eg. ACPI) firmware on your system.  In that case the "platform" mode
127of hibernation is not likely to work.  You can try the "shutdown" mode, but that
128is rather a poor man's workaround.
129
130If the "processors" test fails, the disabling/enabling of nonboot CPUs does not
131work (of course, this only may be an issue on SMP systems) and the problem
132should be reported.  In that case you can also try to switch the nonboot CPUs
133off and on using the /sys/devices/system/cpu/cpu*/online sysfs attributes and
134see if that works.
135
136If the "core" test fails, which means that suspending of the system/platform
137devices has failed (these devices are suspended on one CPU with interrupts off),
138the problem is most probably hardware-related and serious, so it should be
139reported.
140
141A failure of any of the "platform", "processors" or "core" tests may cause your
142system to hang or become unstable, so please beware.  Such a failure usually
143indicates a serious problem that very well may be related to the hardware, but
144please report it anyway.
145
146b) Testing minimal configuration
147
148If all of the hibernation test modes work, you can boot the system with the
149"init=/bin/bash" command line parameter and attempt to hibernate in the
150"reboot", "shutdown" and "platform" modes.  If that does not work, there
151probably is a problem with a driver statically compiled into the kernel and you
152can try to compile more drivers as modules, so that they can be tested
153individually.  Otherwise, there is a problem with a modular driver and you can
154find it by loading a half of the modules you normally use and binary searching
155in accordance with the algorithm:
156- if there are n modules loaded and the attempt to suspend and resume fails,
157unload n/2 of the modules and try again (that would probably involve rebooting
158the system),
159- if there are n modules loaded and the attempt to suspend and resume succeeds,
160load n/2 modules more and try again.
161
162Again, if you find the offending module(s), it(they) must be unloaded every time
163before hibernation, and please report the problem with it(them).
164
165c) Advanced debugging
166
167In case that hibernation does not work on your system even in the minimal
168configuration and compiling more drivers as modules is not practical or some
169modules cannot be unloaded, you can use one of the more advanced debugging
170techniques to find the problem.  First, if there is a serial port in your box,
171you can boot the kernel with the 'no_console_suspend' parameter and try to log
172kernel messages using the serial console.  This may provide you with some
173information about the reasons of the suspend (resume) failure.  Alternatively,
174it may be possible to use a FireWire port for debugging with firescope
175(ftp://ftp.firstfloor.org/pub/ak/firescope/).  On x86 it is also possible to
176use the PM_TRACE mechanism documented in Documentation/power/s2ram.txt .
177
1782. Testing suspend to RAM (STR)
179
180To verify that the STR works, it is generally more convenient to use the s2ram
181tool available from http://suspend.sf.net and documented at
182http://en.opensuse.org/SDB:Suspend_to_RAM.
183
184Namely, after writing "freezer", "devices", "platform", "processors", or "core"
185into /sys/power/pm_test (available if the kernel is compiled with
186CONFIG_PM_DEBUG set) the suspend code will work in the test mode corresponding
187to given string.  The STR test modes are defined in the same way as for
188hibernation, so please refer to Section 1 for more information about them.  In
189particular, the "core" test allows you to test everything except for the actual
190invocation of the platform firmware in order to put the system into the sleep
191state.
192
193Among other things, the testing with the help of /sys/power/pm_test may allow
194you to identify drivers that fail to suspend or resume their devices.  They
195should be unloaded every time before an STR transition.
196
197Next, you can follow the instructions at http://en.opensuse.org/s2ram to test
198the system, but if it does not work "out of the box", you may need to boot it
199with "init=/bin/bash" and test s2ram in the minimal configuration.  In that
200case, you may be able to search for failing drivers by following the procedure
201analogous to the one described in section 1.  If you find some failing drivers,
202you will have to unload them every time before an STR transition (ie. before
203you run s2ram), and please report the problems with them.
204
205There is a debugfs entry which shows the suspend to RAM statistics. Here is an
206example of its output.
207	# mount -t debugfs none /sys/kernel/debug
208	# cat /sys/kernel/debug/suspend_stats
209	success: 20
210	fail: 5
211	failed_freeze: 0
212	failed_prepare: 0
213	failed_suspend: 5
214	failed_suspend_noirq: 0
215	failed_resume: 0
216	failed_resume_noirq: 0
217	failures:
218	  last_failed_dev:	alarm
219				adc
220	  last_failed_errno:	-16
221				-16
222	  last_failed_step:	suspend
223				suspend
224Field success means the success number of suspend to RAM, and field fail means
225the failure number. Others are the failure number of different steps of suspend
226to RAM. suspend_stats just lists the last 2 failed devices, error number and
227failed step of suspend.
228