1 ==========================
2 Hard disk shock protection
3 ==========================
5 Author: Elias Oltmanns <eo@nebensachen.de>
7 Last modified: 2008-10-03
21 ATA/ATAPI-7 specifies the IDLE IMMEDIATE command with unload feature.
22 Issuing this command should cause the drive to switch to idle mode and
23 unload disk heads. This feature is being used in modern laptops in
24 conjunction with accelerometers and appropriate software to implement
25 a shock protection facility. The idea is to stop all I/O operations on
26 the internal hard drive and park its heads on the ramp when critical
27 situations are anticipated. The desire to have such a feature
28 available on GNU/Linux systems has been the original motivation to
29 implement a generic disk head parking interface in the Linux kernel.
30 Please note, however, that other components have to be set up on your
31 system in order to get disk shock protection working (see
32 section 3. References below for pointers to more information about
39 For each ATA device, the kernel exports the file
40 `block/*/device/unload_heads` in sysfs (here assumed to be mounted under
41 /sys). Access to `/sys/block/*/device/unload_heads` is denied with
42 -EOPNOTSUPP if the device does not support the unload feature.
43 Otherwise, writing an integer value to this file will take the heads
44 of the respective drive off the platter and block all I/O operations
45 for the specified number of milliseconds. When the timeout expires and
46 no further disk head park request has been issued in the meantime,
47 normal operation will be resumed. The maximal value accepted for a
48 timeout is 30000 milliseconds. Exceeding this limit will return
49 -EOVERFLOW, but heads will be parked anyway and the timeout will be
50 set to 30 seconds. However, you can always change a timeout to any
51 value between 0 and 30000 by issuing a subsequent head park request
52 before the timeout of the previous one has expired. In particular, the
53 total timeout can exceed 30 seconds and, more importantly, you can
54 cancel a previously set timeout and resume normal operation
55 immediately by specifying a timeout of 0. Values below -2 are rejected
56 with -EINVAL (see below for the special meaning of -1 and -2). If the
57 timeout specified for a recent head park request has not yet expired,
58 reading from `/sys/block/*/device/unload_heads` will report the number
59 of milliseconds remaining until normal operation will be resumed;
60 otherwise, reading the unload_heads attribute will return 0.
62 For example, do the following in order to park the heads of drive
63 /dev/sda and stop all I/O operations for five seconds::
65 # echo 5000 > /sys/block/sda/device/unload_heads
69 # cat /sys/block/sda/device/unload_heads
71 will show you how many milliseconds are left before normal operation
74 A word of caution: The fact that the interface operates on a basis of
75 milliseconds may raise expectations that cannot be satisfied in
76 reality. In fact, the ATA specs clearly state that the time for an
77 unload operation to complete is vendor specific. The hint in ATA-7
78 that this will typically be within 500 milliseconds apparently has
79 been dropped in ATA-8.
81 There is a technical detail of this implementation that may cause some
82 confusion and should be discussed here. When a head park request has
83 been issued to a device successfully, all I/O operations on the
84 controller port this device is attached to will be deferred. That is
85 to say, any other device that may be connected to the same port will
86 be affected too. The only exception is that a subsequent head unload
87 request to that other device will be executed immediately. Further
88 operations on that port will be deferred until the timeout specified
89 for either device on the port has expired. As far as PATA (old style
90 IDE) configurations are concerned, there can only be two devices
91 attached to any single port. In SATA world we have port multipliers
92 which means that a user-issued head parking request to one device may
93 actually result in stopping I/O to a whole bunch of devices. However,
94 since this feature is supposed to be used on laptops and does not seem
95 to be very useful in any other environment, there will be mostly one
96 device per port. Even if the CD/DVD writer happens to be connected to
97 the same port as the hard drive, it generally *should* recover just
98 fine from the occasional buffer under-run incurred by a head park
99 request to the HD. Actually, when you are using an ide driver rather
100 than its libata counterpart (i.e. your disk is called /dev/hda
101 instead of /dev/sda), then parking the heads of one drive (drive X)
102 will generally not affect the mode of operation of another drive
103 (drive Y) on the same port as described above. It is only when a port
104 reset is required to recover from an exception on drive Y that further
105 I/O operations on that drive (and the reset itself) will be delayed
106 until drive X is no longer in the parked state.
108 Finally, there are some hard drives that only comply with an earlier
109 version of the ATA standard than ATA-7, but do support the unload
110 feature nonetheless. Unfortunately, there is no safe way Linux can
111 detect these devices, so you won't be able to write to the
112 unload_heads attribute. If you know that your device really does
113 support the unload feature (for instance, because the vendor of your
114 laptop or the hard drive itself told you so), then you can tell the
115 kernel to enable the usage of this feature for that drive by writing
116 the special value -1 to the unload_heads attribute::
118 # echo -1 > /sys/block/sda/device/unload_heads
120 will enable the feature for /dev/sda, and giving -2 instead of -1 will
127 There are several laptops from different vendors featuring shock
128 protection capabilities. As manufacturers have refused to support open
129 source development of the required software components so far, Linux
130 support for shock protection varies considerably between different
131 hardware implementations. Ideally, this section should contain a list
132 of pointers at different projects aiming at an implementation of shock
133 protection on different systems. Unfortunately, I only know of a
134 single project which, although still considered experimental, is fit
135 for use. Please feel free to add projects that have been the victims
138 - https://www.thinkwiki.org/wiki/HDAPS
140 See this page for information about Linux support of the hard disk
141 active protection system as implemented in IBM/Lenovo Thinkpads.
147 This implementation of disk head parking has been inspired by a patch
148 originally published by Jon Escombe <lists@dresco.co.uk>. My efforts
149 to develop an implementation of this feature that is fit to be merged
150 into mainline have been aided by various kernel developers, in
151 particular by Tejun Heo and Bartlomiej Zolnierkiewicz.