| Age | Commit message (Collapse) | Author |
|---|
|
Bug: 17511334
Change-Id: Ife878bfc039829aef35d048a3d439e6a8c0f2137
|
|
Change-Id: I016607a49207319d92ed1050a802b00b729c11bb
|
|
The update of the sensorhub to use polling was based on the
belief that the workqueue would be frozen before suspend and thawed
afterwards. Unfortunately the system_wq used by schedule_delayed_work,
is not freezable. Update the changes to use the system_freezable_wq.
Change-Id: Ib759664a5bc61dad72207f59d2c8c672cd0f7c9b
Signed-off-by: Jim Wylder <jwylder@motorola.com>
|
|
Bug : 16656953
All sensors were interrupt based, m4 would interrupt omap
when the delay timeout occurs, but to be able mask
non wakable sensor interrupts on suspend and resume, we
r now moving non wakable sensors to be poll based, the
kernel starts a timer for each sensor and reads data from m4
on timer expiry.
Change-Id: I39f1a94563f3c4c65b586b63018e9e1afebd0d28
|
|
Removed IIO settings that caused an unused sysfs file to be created, which causes a kernel panic when read.
Added a gesture count to the gesture driver to differentiate new gestures.
Change-Id: I0e88c88356560c276867388b057ac0a9bb76167e
Signed-off-by: Eric Tashakkor <w36098@motorola.com>
|
|
Added mutex locking in panic restore handlers for all drivers to prevent
losing data when restoring after an M4 panic.
Added a mutex to the mpu9150 driver to prevent concurrent access issues
and loss of data during normal operation.
Change-Id: I9ad5595c632518093ad1aec86afc7a51d3cbbee3
Signed-off-by: Eric Tashakkor <w36098@motorola.com>
|
|
If i2c communication fails while doing a set sample rate for a sensor
retry the same during suspend. It is seems that sometimes omap->M4
i2c bus dies for a brief period and then recovers. But during the failure
period if an app tries to turn off a sensor and this turn off fails,
it could lead to wakeups after omap suspend, and so this fix
retries setting the sample rate on suspend only if it had failed earlier.
Change-Id: I5874738a78ed900b1b039160f852031321dd2e30
|
|
Added orientation using Euler values from M4.
Change-Id: I3dafb73851d30cef5aa98b84a9bda58839bab7c2
Signed-off-by: Eric Tashakkor <w36098@motorola.com>
|
|
Change-Id: I9c090eb29435dea38b43bd367acdf2efa1fde17b
|
|
Change-Id: I53820435227388c323d2beec6e678a94b674fff9
|
|
A few m4 interrupts/events like gesture, rtc alarm that
can occur when kernel is suspended need enough time
to allow the event to reach the higher layer before
kernel suspends again. The absence of this was causing
gestures & rtc alarm to get missed by the higher layers.
Change-Id: Ibf10a1a031bd279ef2a488214498036c648220dc
|
|
Added a separate IIO fusion driver to handle M4 fusion data.
Removed duplicate mutex_destroy() calls.
Deleted unused pressure sensor header.
Change-Id: I2f24a30dbba5b84f590618e3b506d04c92d2a136
Signed-off-by: Eric Tashakkor <w36098@motorola.com>
|
