mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-12-02 08:16:41 +07:00
728ce22b69
The cpuidle sysfs code is designed to have a single instance of per CPU cpuidle directory. It is not possible to remove the sysfs entry and create it again. This is not a problem with the current code but future changes will add CPU hotplug support to enable/disable the device, so it will need to remove the sysfs entry like other subsystems do. That won't be possible without this change, because the kobj is a static object which can't be reused for kobj_init_and_add(). Add cpuidle_device_kobj to be allocated dynamically when adding/removing a sysfs entry which is consistent with the other cpuidle's sysfs entries. An added benefit is that the sysfs code is now more self-contained and the includes needed for sysfs can be moved from cpuidle.h directly into sysfs.c so as to reduce the total number of headers dragged along with cpuidle.h. [rjw: Changelog] Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> |
||
|---|---|---|
| .. | ||
| governors | ||
| coupled.c | ||
| cpuidle-calxeda.c | ||
| cpuidle-kirkwood.c | ||
| cpuidle-zynq.c | ||
| cpuidle.c | ||
| cpuidle.h | ||
| driver.c | ||
| governor.c | ||
| Kconfig | ||
| Makefile | ||
| sysfs.c | ||