| Message ID | 20240910-iio-bmi323-remove-config_pm-guards-v1-1-0552249207af@kernel.org (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
| Series | iio: bmi323: Drop CONFIG_PM guards around runtime functions | expand |
On Tue, 10 Sep 2024 10:22:54 -0700 Nathan Chancellor <nathan@kernel.org> wrote: > When building with clang and CONFIG_PM disabled (such as with s390), it > warns: > > drivers/iio/imu/bmi323/bmi323_core.c:121:27: warning: variable 'bmi323_reg_savestate' is not needed and will not be emitted [-Wunneeded-internal-declaration] > 121 | static const unsigned int bmi323_reg_savestate[] = { > | ^~~~~~~~~~~~~~~~~~~~ > drivers/iio/imu/bmi323/bmi323_core.c:133:27: warning: variable 'bmi323_ext_reg_savestate' is not needed and will not be emitted [-Wunneeded-internal-declaration] > 133 | static const unsigned int bmi323_ext_reg_savestate[] = { > | ^~~~~~~~~~~~~~~~~~~~~~~~ > > These arrays have no references outside of sizeof(), which will be > evaluated at compile time. To avoid these warnings, remove the CONFIG_PM > ifdef guard and use the RUNTIME_PM_OPS macro to ensure these functions > always appear used to the compiler, which allows the references to the > arrays to be visible as well. This results in no difference in runtime > behavior because bmi323_core_pm_ops is only used when CONFIG_PM is set > with the pm_ptr() macro. > > Fixes: b09999ee1e86 ("iio: bmi323: suspend and resume triggering on relevant pm operations") > Signed-off-by: Nathan Chancellor <nathan@kernel.org> Applied to the fixes togreg branch of iio.git. I'll rebase that tree on rc1 once available then send a pull request with this in. Thanks, Jonathan > --- > NOTE: The second warning will still be visible without this change as > well: > > https://lore.kernel.org/20240909-iio-bmi323-fix-array-ref-v1-1-51c220f22229@kernel.org/ > > Technically, the blamed change for that fix is somewhat responsible for > this issue as well but I believe the one I assigned was the real culprit > because this diff should have been in that change. > --- > drivers/iio/imu/bmi323/bmi323_core.c | 7 ++----- > 1 file changed, 2 insertions(+), 5 deletions(-) > > diff --git a/drivers/iio/imu/bmi323/bmi323_core.c b/drivers/iio/imu/bmi323/bmi323_core.c > index 671401ce80dcf947b7b64ea3af112d2a42ca5501..1d1405d37c5adb2717fc27d6b22540da50dea6d4 100644 > --- a/drivers/iio/imu/bmi323/bmi323_core.c > +++ b/drivers/iio/imu/bmi323/bmi323_core.c > @@ -2172,7 +2172,6 @@ int bmi323_core_probe(struct device *dev) > } > EXPORT_SYMBOL_NS_GPL(bmi323_core_probe, IIO_BMI323); > > -#if defined(CONFIG_PM) > static int bmi323_core_runtime_suspend(struct device *dev) > { > struct iio_dev *indio_dev = dev_get_drvdata(dev); > @@ -2293,11 +2292,9 @@ static int bmi323_core_runtime_resume(struct device *dev) > return iio_device_resume_triggering(indio_dev); > } > > -#endif > - > const struct dev_pm_ops bmi323_core_pm_ops = { > - SET_RUNTIME_PM_OPS(bmi323_core_runtime_suspend, > - bmi323_core_runtime_resume, NULL) > + RUNTIME_PM_OPS(bmi323_core_runtime_suspend, > + bmi323_core_runtime_resume, NULL) > }; > EXPORT_SYMBOL_NS_GPL(bmi323_core_pm_ops, IIO_BMI323); > > > --- > base-commit: 5ba0cb92584ba5e107c97001e09013c1da0772a8 > change-id: 20240910-iio-bmi323-remove-config_pm-guards-5707958e0ac1 > > Best regards,
diff --git a/drivers/iio/imu/bmi323/bmi323_core.c b/drivers/iio/imu/bmi323/bmi323_core.c index 671401ce80dcf947b7b64ea3af112d2a42ca5501..1d1405d37c5adb2717fc27d6b22540da50dea6d4 100644 --- a/drivers/iio/imu/bmi323/bmi323_core.c +++ b/drivers/iio/imu/bmi323/bmi323_core.c @@ -2172,7 +2172,6 @@ int bmi323_core_probe(struct device *dev) } EXPORT_SYMBOL_NS_GPL(bmi323_core_probe, IIO_BMI323); -#if defined(CONFIG_PM) static int bmi323_core_runtime_suspend(struct device *dev) { struct iio_dev *indio_dev = dev_get_drvdata(dev); @@ -2293,11 +2292,9 @@ static int bmi323_core_runtime_resume(struct device *dev) return iio_device_resume_triggering(indio_dev); } -#endif - const struct dev_pm_ops bmi323_core_pm_ops = { - SET_RUNTIME_PM_OPS(bmi323_core_runtime_suspend, - bmi323_core_runtime_resume, NULL) + RUNTIME_PM_OPS(bmi323_core_runtime_suspend, + bmi323_core_runtime_resume, NULL) }; EXPORT_SYMBOL_NS_GPL(bmi323_core_pm_ops, IIO_BMI323);
When building with clang and CONFIG_PM disabled (such as with s390), it warns: drivers/iio/imu/bmi323/bmi323_core.c:121:27: warning: variable 'bmi323_reg_savestate' is not needed and will not be emitted [-Wunneeded-internal-declaration] 121 | static const unsigned int bmi323_reg_savestate[] = { | ^~~~~~~~~~~~~~~~~~~~ drivers/iio/imu/bmi323/bmi323_core.c:133:27: warning: variable 'bmi323_ext_reg_savestate' is not needed and will not be emitted [-Wunneeded-internal-declaration] 133 | static const unsigned int bmi323_ext_reg_savestate[] = { | ^~~~~~~~~~~~~~~~~~~~~~~~ These arrays have no references outside of sizeof(), which will be evaluated at compile time. To avoid these warnings, remove the CONFIG_PM ifdef guard and use the RUNTIME_PM_OPS macro to ensure these functions always appear used to the compiler, which allows the references to the arrays to be visible as well. This results in no difference in runtime behavior because bmi323_core_pm_ops is only used when CONFIG_PM is set with the pm_ptr() macro. Fixes: b09999ee1e86 ("iio: bmi323: suspend and resume triggering on relevant pm operations") Signed-off-by: Nathan Chancellor <nathan@kernel.org> --- NOTE: The second warning will still be visible without this change as well: https://lore.kernel.org/20240909-iio-bmi323-fix-array-ref-v1-1-51c220f22229@kernel.org/ Technically, the blamed change for that fix is somewhat responsible for this issue as well but I believe the one I assigned was the real culprit because this diff should have been in that change. --- drivers/iio/imu/bmi323/bmi323_core.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) --- base-commit: 5ba0cb92584ba5e107c97001e09013c1da0772a8 change-id: 20240910-iio-bmi323-remove-config_pm-guards-5707958e0ac1 Best regards,