diff options
| author | Igor Grinberg <grinberg@compulab.co.il> | 2012-04-12 15:43:29 +0300 |
|---|---|---|
| committer | Haojian Zhuang <haojian.zhuang@gmail.com> | 2012-04-27 11:14:06 +0800 |
| commit | ef7c7c693b4b05a293678fd40ed1511bc94d6f61 (patch) | |
| tree | 9ea5270c9139d9dd1047738c5eda6d89ef311162 /sound/pci/ctxfi | |
| parent | a13b878780ea98d8415203c4502a8fc5bc614656 (diff) | |
| download | talos-obmc-linux-ef7c7c693b4b05a293678fd40ed1511bc94d6f61.tar.gz talos-obmc-linux-ef7c7c693b4b05a293678fd40ed1511bc94d6f61.zip | |
ARM: PXA2xx: MFP: fix potential direction bug
Pins configured as input and have MFP_LPM_DRIVE_* flag set, can have a
wrong output value for some period of time (spike) during the suspend
sequence.
This can happen because the direction of the pins (GPDR) is set by
software and the output level is set by hardware (PGSR) at a later
stage.
Fix the above potential bug by setting the output levels first.
Also save the actual levels of the pins before the suspend and restore
them after the resume, but before the direction settings take place, so
the same bug as described above will not happen in the resume sequence.
Reported-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Igor Grinberg <grinberg@compulab.co.il>
Tested-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Haojian Zhuang <haojian.zhuang@gmail.com>
Diffstat (limited to 'sound/pci/ctxfi')
0 files changed, 0 insertions, 0 deletions
