]>
Commit | Line | Data |
---|---|---|
6e1819d6 RW |
1 | Documentation for userland software suspend interface |
2 | (C) 2006 Rafael J. Wysocki <rjw@sisk.pl> | |
3 | ||
4 | First, the warnings at the beginning of swsusp.txt still apply. | |
5 | ||
6 | Second, you should read the FAQ in swsusp.txt _now_ if you have not | |
7 | done it already. | |
8 | ||
9 | Now, to use the userland interface for software suspend you need special | |
10 | utilities that will read/write the system memory snapshot from/to the | |
11 | kernel. Such utilities are available, for example, from | |
12 | <http://www.sisk.pl/kernel/utilities/suspend>. You may want to have | |
13 | a look at them if you are going to develop your own suspend/resume | |
14 | utilities. | |
15 | ||
16 | The interface consists of a character device providing the open(), | |
17 | release(), read(), and write() operations as well as several ioctl() | |
18 | commands defined in kernel/power/power.h. The major and minor | |
19 | numbers of the device are, respectively, 10 and 231, and they can | |
20 | be read from /sys/class/misc/snapshot/dev. | |
21 | ||
22 | The device can be open either for reading or for writing. If open for | |
23 | reading, it is considered to be in the suspend mode. Otherwise it is | |
24 | assumed to be in the resume mode. The device cannot be open for reading | |
25 | and writing. It is also impossible to have the device open more than once | |
26 | at a time. | |
27 | ||
28 | The ioctl() commands recognized by the device are: | |
29 | ||
30 | SNAPSHOT_FREEZE - freeze user space processes (the current process is | |
31 | not frozen); this is required for SNAPSHOT_ATOMIC_SNAPSHOT | |
32 | and SNAPSHOT_ATOMIC_RESTORE to succeed | |
33 | ||
34 | SNAPSHOT_UNFREEZE - thaw user space processes frozen by SNAPSHOT_FREEZE | |
35 | ||
36 | SNAPSHOT_ATOMIC_SNAPSHOT - create a snapshot of the system memory; the | |
37 | last argument of ioctl() should be a pointer to an int variable, | |
38 | the value of which will indicate whether the call returned after | |
39 | creating the snapshot (1) or after restoring the system memory state | |
40 | from it (0) (after resume the system finds itself finishing the | |
41 | SNAPSHOT_ATOMIC_SNAPSHOT ioctl() again); after the snapshot | |
42 | has been created the read() operation can be used to transfer | |
43 | it out of the kernel | |
44 | ||
45 | SNAPSHOT_ATOMIC_RESTORE - restore the system memory state from the | |
46 | uploaded snapshot image; before calling it you should transfer | |
47 | the system memory snapshot back to the kernel using the write() | |
48 | operation; this call will not succeed if the snapshot | |
49 | image is not available to the kernel | |
50 | ||
51 | SNAPSHOT_FREE - free memory allocated for the snapshot image | |
52 | ||
53 | SNAPSHOT_SET_IMAGE_SIZE - set the preferred maximum size of the image | |
54 | (the kernel will do its best to ensure the image size will not exceed | |
55 | this number, but if it turns out to be impossible, the kernel will | |
56 | create the smallest image possible) | |
57 | ||
58 | SNAPSHOT_AVAIL_SWAP - return the amount of available swap in bytes (the last | |
59 | argument should be a pointer to an unsigned int variable that will | |
60 | contain the result if the call is successful). | |
61 | ||
62 | SNAPSHOT_GET_SWAP_PAGE - allocate a swap page from the resume partition | |
63 | (the last argument should be a pointer to a loff_t variable that | |
64 | will contain the swap page offset if the call is successful) | |
65 | ||
66 | SNAPSHOT_FREE_SWAP_PAGES - free all swap pages allocated with | |
67 | SNAPSHOT_GET_SWAP_PAGE | |
68 | ||
69 | SNAPSHOT_SET_SWAP_FILE - set the resume partition (the last ioctl() argument | |
70 | should specify the device's major and minor numbers in the old | |
71 | two-byte format, as returned by the stat() function in the .st_rdev | |
72 | member of the stat structure); it is recommended to always use this | |
73 | call, because the code to set the resume partition could be removed from | |
74 | future kernels | |
75 | ||
76 | The device's read() operation can be used to transfer the snapshot image from | |
77 | the kernel. It has the following limitations: | |
78 | - you cannot read() more than one virtual memory page at a time | |
79 | - read()s accross page boundaries are impossible (ie. if ypu read() 1/2 of | |
80 | a page in the previous call, you will only be able to read() | |
81 | _at_ _most_ 1/2 of the page in the next call) | |
82 | ||
83 | The device's write() operation is used for uploading the system memory snapshot | |
84 | into the kernel. It has the same limitations as the read() operation. | |
85 | ||
86 | The release() operation frees all memory allocated for the snapshot image | |
87 | and all swap pages allocated with SNAPSHOT_GET_SWAP_PAGE (if any). | |
88 | Thus it is not necessary to use either SNAPSHOT_FREE or | |
89 | SNAPSHOT_FREE_SWAP_PAGES before closing the device (in fact it will also | |
90 | unfreeze user space processes frozen by SNAPSHOT_UNFREEZE if they are | |
91 | still frozen when the device is being closed). | |
92 | ||
93 | Currently it is assumed that the userland utilities reading/writing the | |
94 | snapshot image from/to the kernel will use a swap parition, called the resume | |
95 | partition, as storage space. However, this is not really required, as they | |
96 | can use, for example, a special (blank) suspend partition or a file on a partition | |
97 | that is unmounted before SNAPSHOT_ATOMIC_SNAPSHOT and mounted afterwards. | |
98 | ||
99 | These utilities SHOULD NOT make any assumptions regarding the ordering of | |
100 | data within the snapshot image, except for the image header that MAY be | |
101 | assumed to start with an swsusp_info structure, as specified in | |
102 | kernel/power/power.h. This structure MAY be used by the userland utilities | |
103 | to obtain some information about the snapshot image, such as the size | |
104 | of the snapshot image, including the metadata and the header itself, | |
105 | contained in the .size member of swsusp_info. | |
106 | ||
107 | The snapshot image MUST be written to the kernel unaltered (ie. all of the image | |
108 | data, metadata and header MUST be written in _exactly_ the same amount, form | |
109 | and order in which they have been read). Otherwise, the behavior of the | |
110 | resumed system may be totally unpredictable. | |
111 | ||
112 | While executing SNAPSHOT_ATOMIC_RESTORE the kernel checks if the | |
113 | structure of the snapshot image is consistent with the information stored | |
114 | in the image header. If any inconsistencies are detected, | |
115 | SNAPSHOT_ATOMIC_RESTORE will not succeed. Still, this is not a fool-proof | |
116 | mechanism and the userland utilities using the interface SHOULD use additional | |
117 | means, such as checksums, to ensure the integrity of the snapshot image. | |
118 | ||
119 | The suspending and resuming utilities MUST lock themselves in memory, | |
120 | preferrably using mlockall(), before calling SNAPSHOT_FREEZE. | |
121 | ||
122 | The suspending utility MUST check the value stored by SNAPSHOT_ATOMIC_SNAPSHOT | |
123 | in the memory location pointed to by the last argument of ioctl() and proceed | |
124 | in accordance with it: | |
125 | 1. If the value is 1 (ie. the system memory snapshot has just been | |
126 | created and the system is ready for saving it): | |
127 | (a) The suspending utility MUST NOT close the snapshot device | |
128 | _unless_ the whole suspend procedure is to be cancelled, in | |
129 | which case, if the snapshot image has already been saved, the | |
130 | suspending utility SHOULD destroy it, preferrably by zapping | |
131 | its header. If the suspend is not to be cancelled, the | |
132 | system MUST be powered off or rebooted after the snapshot | |
133 | image has been saved. | |
134 | (b) The suspending utility SHOULD NOT attempt to perform any | |
135 | file system operations (including reads) on the file systems | |
136 | that were mounted before SNAPSHOT_ATOMIC_SNAPSHOT has been | |
137 | called. However, it MAY mount a file system that was not | |
138 | mounted at that time and perform some operations on it (eg. | |
139 | use it for saving the image). | |
140 | 2. If the value is 0 (ie. the system state has just been restored from | |
141 | the snapshot image), the suspending utility MUST close the snapshot | |
142 | device. Afterwards it will be treated as a regular userland process, | |
143 | so it need not exit. | |
144 | ||
145 | The resuming utility SHOULD NOT attempt to mount any file systems that could | |
146 | be mounted before suspend and SHOULD NOT attempt to perform any operations | |
147 | involving such file systems. | |
148 | ||
149 | For details, please refer to the source code. |