kenel thread

再來就是本文主要談的kthreaddTask,這是在LinuxKernel 2.6所引入的機制,在這之前要產生KernelThread需直接使用函式kernel_thread,而所產生的KernelThread父行程會是當下產生KernelThread的行程(或該父行程結束後,改為initTask(PID=1)).在kthreadd的機制下,UserMode與KernelMode Task的產生方式做了調整,並讓kthreaddTask成為使用kthread_create產生的KernelThread統一的父行程.也因此,在這機制實現下的LinuxKernel,屬於UserMode行程最上層的父行程為initTask (PID=1),而屬於KernelMode行程最上層的父行程為kthreaddTask (PID=2),而這兩個行程共同的父行程就是idle Task (PID=0).

我們選擇直接透過kernel_thread產生KernelThread,跟透過kthreadd機制相比,兩者的差別在於,一個是由當下呼叫的kernel_thread的Task行程所fork出來的,採用kthread_create機制則是由kthreaddTask行程所fork出來的.

kernel_thread

kthread_create

kthread_create_on_node

/**
 * struct platform_hibernation_ops - hibernation platform support
 *
 * The methods in this structure allow a platform to carry out special
 * operations required by it during a hibernation transition.
 *
 * All the methods below, except for @recover(), must be implemented.
 *
 * @begin: Tell the platform driver that we're starting hibernation.
 * Called right after shrinking memory and before freezing devices.
 *
 * @end: Called by the PM core right after resuming devices, to indicate to
 * the platform that the system has returned to the working state.
 *
 * @pre_snapshot: Prepare the platform for creating the hibernation image.
 * Called right after devices have been frozen and before the nonboot
 * CPUs are disabled (runs with IRQs on).
 *
 * @finish: Restore the previous state of the platform after the hibernation
 * image has been created *or* put the platform into the normal operation
 * mode after the hibernation (the same method is executed in both cases).
 * Called right after the nonboot CPUs have been enabled and before
 * thawing devices (runs with IRQs on).
 *
 * @prepare: Prepare the platform for entering the low power state.
 * Called right after the hibernation image has been saved and before
 * devices are prepared for entering the low power state.
 *
 * @enter: Put the system into the low power state after the hibernation image
 * has been saved to disk.
 * Called after the nonboot CPUs have been disabled and all of the low
 * level devices have been shut down (runs with IRQs off).
 *
 * @leave: Perform the first stage of the cleanup after the system sleep state
 * indicated by @set_target() has been left.
 * Called right after the control has been passed from the boot kernel to
 * the image kernel, before the nonboot CPUs are enabled and before devices
 * are resumed.  Executed with interrupts disabled.
 *
 * @pre_restore: Prepare system for the restoration from a hibernation image.
 * Called right after devices have been frozen and before the nonboot
 * CPUs are disabled (runs with IRQs on).
 *
 * @restore_cleanup: Clean up after a failing image restoration.
 * Called right after the nonboot CPUs have been enabled and before
 * thawing devices (runs with IRQs on).
 *
 * @recover: Recover the platform from a failure to suspend devices.
 * Called by the PM core if the suspending of devices during hibernation
 * fails.  This callback is optional and should only be implemented by
 * platforms which require special recovery actions in that situation.
 */
struct platform_hibernation_ops {
int (*begin)(void);
void (*end)(void);
int (*pre_snapshot)(void);
void (*finish)(void);
int (*prepare)(void);
int (*enter)(void);
void (*leave)(void);
int (*pre_restore)(void);
void (*restore_cleanup)(void);
void (*recover)(void);
};

/**
 * struct platform_suspend_ops - Callbacks for managing platform dependent
 * system sleep states.
 *
 * @valid: Callback to determine if given system sleep state is supported by
 * the platform.
 * Valid (ie. supported) states are advertised in /sys/power/state.  Note
 * that it still may be impossible to enter given system sleep state if the
 * conditions aren't right.
 * There is the %suspend_valid_only_mem function available that can be
 * assigned to this if the platform only supports mem sleep.
 *
 * @begin: Initialise a transition to given system sleep state.
 * @begin() is executed right prior to suspending devices.  The information
 * conveyed to the platform code by @begin() should be disregarded by it as
 * soon as @end() is executed.  If @begin() fails (ie. returns nonzero),
 * @prepare(), @enter() and @finish() will not be called by the PM core.
 * This callback is optional.  However, if it is implemented, the argument
 * passed to @enter() is redundant and should be ignored.
 *
 * @prepare: Prepare the platform for entering the system sleep state indicated
 * by @begin().
 * @prepare() is called right after devices have been suspended (ie. the
 * appropriate .suspend() method has been executed for each device) and
 * before device drivers' late suspend callbacks are executed.  It returns
 * 0 on success or a negative error code otherwise, in which case the
 * system cannot enter the desired sleep state (@prepare_late(), @enter(),
 * and @wake() will not be called in that case).
 *
 * @prepare_late: Finish preparing the platform for entering the system sleep
 * state indicated by @begin().
 * @prepare_late is called before disabling nonboot CPUs and after
 * device drivers' late suspend callbacks have been executed.  It returns
 * 0 on success or a negative error code otherwise, in which case the
 * system cannot enter the desired sleep state (@enter() will not be
 * executed).
 *
 * @enter: Enter the system sleep state indicated by @begin() or represented by
 * the argument if @begin() is not implemented.
 * This callback is mandatory.  It returns 0 on success or a negative
 * error code otherwise, in which case the system cannot enter the desired
 * sleep state.
 *
 * @wake: Called when the system has just left a sleep state, right after
 * the nonboot CPUs have been enabled and before device drivers' early
 * resume callbacks are executed.
 * This callback is optional, but should be implemented by the platforms
 * that implement @prepare_late().  If implemented, it is always called
 * after @prepare_late and @enter(), even if one of them fails.
 *
 * @finish: Finish wake-up of the platform.
 * @finish is called right prior to calling device drivers' regular suspend
 * callbacks.
 * This callback is optional, but should be implemented by the platforms
 * that implement @prepare().  If implemented, it is always called after
 * @enter() and @wake(), even if any of them fails.  It is executed after
 * a failing @prepare.
 *
 * @end: Called by the PM core right after resuming devices, to indicate to
 * the platform that the system has returned to the working state or
 * the transition to the sleep state has been aborted.
 * This callback is optional, but should be implemented by the platforms
 * that implement @begin().  Accordingly, platforms implementing @begin()
 * should also provide a @end() which cleans up transitions aborted before
 * @enter().
 *
 * @recover: Recover the platform from a suspend failure.
 * Called by the PM core if the suspending of devices fails.
 * This callback is optional and should only be implemented by platforms
 * which require special recovery actions in that situation.
 */
struct platform_suspend_ops {
int (*valid)(suspend_state_t state);
int (*begin)(suspend_state_t state);
int (*prepare)(void);
int (*prepare_late)(void);
int (*enter)(suspend_state_t state);
void (*wake)(void);
void (*finish)(void);
void (*end)(void);
void (*recover)(void);
};

platform_hibernation_ops   hibernation流程是会执行里面的函数

platform_suspend_ops        suspend流程（on，standby，mem）会执行里面的流程