Android does not use standby power state by default.
So we can take this state as eink suspend. switch to mem
state when request for suspend before earlysuspend.

Quick suspend request for this case will confuse the earlysuspend store routing.
So record the last suspend mode. If entering another suspend mode,
and early resume is not called, let resume work go by flushing the workqueue,
then request for another suspend mode.

Signed-off-by: Xinyu Chen <xinyu.chen@freescale.com>

Add earlysuspend status to current kernel power subsystem.
Kernel will first enter earlysuspend, later if no wakelock,
then it goes into the normal suspend state.
There's earlysuspend api exported to be used by drivers who want to
register itself's pm callback into earlysuspend state.

Add console and fb earlysuspend driver to do display earlysuspend.

Add kernel level wakelock api.
Drivers can hold/release the wakelock to prevent system going to
the suspend mode.

Add user level wakelock interface based on wakelock api.
Application can hold/release the wakelock by /sys/power/wake_lock/unlock.

Signed-off-by: Arve Hjønnevåg <arve@android.com>

the default state of suspend driver is default off,
if you want to debug the suspend time, echo time in
microsecond(u Second) to /sys/powe/device_suspend_time_threshold

eg: I want to know which driver suspend & resume takes
more that 0.5 ms (500 us), you can just :

ehco 500 > /sys/power/device_suspend_time_threshold

Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>

There was some driver is slow on suspend/resume,
but some embeded system like eReader,Cellphone
are time sensitive,this commit will report the slow
driver on suspend/resume, the default value is 500us(0.5ms)

Also, the threshold can be change by modify
'/sys/power/device_suspend_time_threshold' to change the threshold,
it is in microsecond.

The output is like:

PM: device platform:soc-audio.2 suspend too slow, takes          606.696 msecs
PM: device platform:mxc_sdc_fb.1 suspend too slow, takes         7.708 msecs

Signed-off-by: Zhang Jiejing <jiejing.zhang@freescale.com>

Saving platform non-volatile state may be required for suspend to RAM as
well as hibernation. Move it to more generic code.

Signed-off-by: Matthew Garrett <mjg@redhat.com>
Acked-by: Rafael J. Wysocki <rjw@sisk.pl>
Tested-by: Maxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: Len Brown <len.brown@intel.com>

Fix printk format warning in block_io.c:

kernel/power/block_io.c:41: warning: format '%ld' expects type 'long int', but argument 2 has type 'sector_t'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

Move all the swap processing into one function. It will make swap
calls from a non-swap code easier.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

The first sector knowledge is swap-only specific. Move it into the
swap handle. This will be needed for later non-swap specific code
moving into snapshot.c.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: "Rafael J. Wysocki" <rjw@sisk.pl>

Move block I/O operations to a separate file. It is because it will
be used later not only by the swap writer.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>

Remove support of reads with offset. This means snapshot_read/write_next
now does not accept count parameter. It allows to clean up the functions
and snapshot handle which no longer needs to care about offsets.

/dev/snapshot handler is converted to simple_{read_from,write_to}_buffer
which take care of offsets.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>