diff options
| author | Yuval Peress <peress@chromium.org> | 2019-06-21 15:04:17 -0600 |
|---|---|---|
| committer | Commit Bot <commit-bot@chromium.org> | 2019-06-22 02:11:37 +0000 |
| commit | 3f80f258a4aad6a7ec048683586e0ecb6c122faf (patch) | |
| tree | 866147938a7601ff6b48048078815ba30242bd7e /driver/mag_lis2mdl.c | |
| parent | e7024509621ddf9cd64b55bcf2898bf2cb3e3cdb (diff) | |
| download | chrome-ec-3f80f258a4aad6a7ec048683586e0ecb6c122faf.tar.gz | |
driver:lis2mdl: Remove todo
After some digging around I've found that the issue wasn't in the
call to init_mag_cal. Instead it was the fact that we weren't
resetting the chip using the LIS2MDL_FLAG_SW_RESET flag every time
we reset the calibration. Instead of resetting every time, it
actually improves performance to ignore no-op calls here. So the
comment can be resolved.
BUG=b:130417518
TEST=Added reset flag to set ODR command and remove early bail.
Everything works.
BRANCH=None
Change-Id: Ida4afefc4aad0630bb60a4e25371aa050f0886d0
Signed-off-by: Yuval Peress <peress@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/ec/+/1672207
Reviewed-by: Jack Rosenthal <jrosenth@chromium.org>
Commit-Queue: Jack Rosenthal <jrosenth@chromium.org>
Diffstat (limited to 'driver/mag_lis2mdl.c')
| -rw-r--r-- | driver/mag_lis2mdl.c | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/driver/mag_lis2mdl.c b/driver/mag_lis2mdl.c index 1dedb23882..40725c180d 100644 --- a/driver/mag_lis2mdl.c +++ b/driver/mag_lis2mdl.c @@ -342,7 +342,10 @@ int lis2mdl_set_data_rate(const struct motion_sensor_t *s, int rate, int rnd) : reg_val == LIS2MDL_ODR_20HZ ? 20000 : 50000; - /* b/130417518 - verify why skipping init_mag_cal is needed here. */ + /* + * If no change is needed just bail. Not doing so will require a reset + * of the chip which only leads to re-calibration and lost samples. + */ if (normalized_rate == data->base.odr) return ret; |