3 USB device persistence during system suspend
4 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
6 :Author: Alan Stern <stern@rowland.harvard.edu>
7 :Date: September 2, 2006 (Updated February 25, 2008)
13 According to the USB specification, when a USB bus is suspended the
14 bus must continue to supply suspend current (around 1-5 mA). This
15 is so that devices can maintain their internal state and hubs can
16 detect connect-change events (devices being plugged in or unplugged).
17 The technical term is "power session".
19 If a USB device's power session is interrupted then the system is
20 required to behave as though the device has been unplugged. It's a
21 conservative approach; in the absence of suspend current the computer
22 has no way to know what has actually happened. Perhaps the same
23 device is still attached or perhaps it was removed and a different
24 device plugged into the port. The system must assume the worst.
26 By default, Linux behaves according to the spec. If a USB host
27 controller loses power during a system suspend, then when the system
28 wakes up all the devices attached to that controller are treated as
29 though they had disconnected. This is always safe and it is the
30 "officially correct" thing to do.
32 For many sorts of devices this behavior doesn't matter in the least.
33 If the kernel wants to believe that your USB keyboard was unplugged
34 while the system was asleep and a new keyboard was plugged in when the
35 system woke up, who cares? It'll still work the same when you type on
38 Unfortunately problems _can_ arise, particularly with mass-storage
39 devices. The effect is exactly the same as if the device really had
40 been unplugged while the system was suspended. If you had a mounted
41 filesystem on the device, you're out of luck -- everything in that
42 filesystem is now inaccessible. This is especially annoying if your
43 root filesystem was located on the device, since your system will
46 Loss of power isn't the only mechanism to worry about. Anything that
47 interrupts a power session will have the same effect. For example,
48 even though suspend current may have been maintained while the system
49 was asleep, on many systems during the initial stages of wakeup the
50 firmware (i.e., the BIOS) resets the motherboard's USB host
51 controllers. Result: all the power sessions are destroyed and again
52 it's as though you had unplugged all the USB devices. Yes, it's
53 entirely the BIOS's fault, but that doesn't do _you_ any good unless
54 you can convince the BIOS supplier to fix the problem (lots of luck!).
56 On many systems the USB host controllers will get reset after a
57 suspend-to-RAM. On almost all systems, no suspend current is
58 available during hibernation (also known as swsusp or suspend-to-disk).
59 You can check the kernel log after resuming to see if either of these
60 has happened; look for lines saying "root hub lost power or was reset".
62 In practice, people are forced to unmount any filesystems on a USB
63 device before suspending. If the root filesystem is on a USB device,
64 the system can't be suspended at all. (All right, it _can_ be
65 suspended -- but it will crash as soon as it wakes up, which isn't
72 The kernel includes a feature called USB-persist. It tries to work
73 around these issues by allowing the core USB device data structures to
74 persist across a power-session disruption.
76 It works like this. If the kernel sees that a USB host controller is
77 not in the expected state during resume (i.e., if the controller was
78 reset or otherwise had lost power) then it applies a persistence check
79 to each of the USB devices below that controller for which the
80 "persist" attribute is set. It doesn't try to resume the device; that
81 can't work once the power session is gone. Instead it issues a USB
82 port reset and then re-enumerates the device. (This is exactly the
83 same thing that happens whenever a USB device is reset.) If the
84 re-enumeration shows that the device now attached to that port has the
85 same descriptors as before, including the Vendor and Product IDs, then
86 the kernel continues to use the same device structure. In effect, the
87 kernel treats the device as though it had merely been reset instead of
90 The same thing happens if the host controller is in the expected state
91 but a USB device was unplugged and then replugged, or if a USB device
92 fails to carry out a normal resume.
94 If no device is now attached to the port, or if the descriptors are
95 different from what the kernel remembers, then the treatment is what
96 you would expect. The kernel destroys the old device structure and
97 behaves as though the old device had been unplugged and a new device
100 The end result is that the USB device remains available and usable.
101 Filesystem mounts and memory mappings are unaffected, and the world is
102 now a good and happy place.
104 Note that the "USB-persist" feature will be applied only to those
105 devices for which it is enabled. You can enable the feature by doing
108 echo 1 >/sys/bus/usb/devices/.../power/persist
110 where the "..." should be filled in the with the device's ID. Disable
111 the feature by writing 0 instead of 1. For hubs the feature is
112 automatically and permanently enabled and the power/persist file
113 doesn't even exist, so you only have to worry about setting it for
114 devices where it really matters.
117 Is this the best solution?
118 ==========================
120 Perhaps not. Arguably, keeping track of mounted filesystems and
121 memory mappings across device disconnects should be handled by a
122 centralized Logical Volume Manager. Such a solution would allow you
123 to plug in a USB flash device, create a persistent volume associated
124 with it, unplug the flash device, plug it back in later, and still
125 have the same persistent volume associated with the device. As such
126 it would be more far-reaching than USB-persist.
128 On the other hand, writing a persistent volume manager would be a big
129 job and using it would require significant input from the user. This
130 solution is much quicker and easier -- and it exists now, a giant
133 Furthermore, the USB-persist feature applies to _all_ USB devices, not
134 just mass-storage devices. It might turn out to be equally useful for
135 other device types, such as network interfaces.
138 WARNING: USB-persist can be dangerous!!
139 =======================================
141 When recovering an interrupted power session the kernel does its best
142 to make sure the USB device hasn't been changed; that is, the same
143 device is still plugged into the port as before. But the checks
144 aren't guaranteed to be 100% accurate.
146 If you replace one USB device with another of the same type (same
147 manufacturer, same IDs, and so on) there's an excellent chance the
148 kernel won't detect the change. The serial number string and other
149 descriptors are compared with the kernel's stored values, but this
150 might not help since manufacturers frequently omit serial numbers
151 entirely in their devices.
153 Furthermore it's quite possible to leave a USB device exactly the same
154 while changing its media. If you replace the flash memory card in a
155 USB card reader while the system is asleep, the kernel will have no
156 way to know you did it. The kernel will assume that nothing has
157 happened and will continue to use the partition tables, inodes, and
158 memory mappings for the old card.
160 If the kernel gets fooled in this way, it's almost certain to cause
161 data corruption and to crash your system. You'll have no one to blame
164 For those devices with avoid_reset_quirk attribute being set, persist
165 maybe fail because they may morph after reset.
167 YOU HAVE BEEN WARNED! USE AT YOUR OWN RISK!
169 That having been said, most of the time there shouldn't be any trouble
170 at all. The USB-persist feature can be extremely useful. Make the