linux/net/sunrpc/sched.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 * linux/net/sunrpc/sched.c
 *
 * Scheduling for synchronous and asynchronous RPC requests.
 *
 * Copyright (C) 1996 Olaf Kirch, <[email protected]>
 *
 * TCP NFS related read + write fixes
 * (C) 1999 Dave Airlie, University of Limerick, Ireland <[email protected]>
 */

#include <linux/module.h>

#include <linux/sched.h>
#include <linux/interrupt.h>
#include <linux/slab.h>
#include <linux/mempool.h>
#include <linux/smp.h>
#include <linux/spinlock.h>
#include <linux/mutex.h>
#include <linux/freezer.h>
#include <linux/sched/mm.h>

#include <linux/sunrpc/clnt.h>
#include <linux/sunrpc/metrics.h>

#include "sunrpc.h"

#define CREATE_TRACE_POINTS
#include <trace/events/sunrpc.h>

/*
 * RPC slabs and memory pools
 */
#define RPC_BUFFER_MAXSIZE …
#define RPC_BUFFER_POOLSIZE …
#define RPC_TASK_POOLSIZE …
static struct kmem_cache	*rpc_task_slabp __read_mostly;
static struct kmem_cache	*rpc_buffer_slabp __read_mostly;
static mempool_t	*rpc_task_mempool __read_mostly;
static mempool_t	*rpc_buffer_mempool __read_mostly;

static void			rpc_async_schedule(struct work_struct *);
static void			 rpc_release_task(struct rpc_task *task);
static void __rpc_queue_timer_fn(struct work_struct *);

/*
 * RPC tasks sit here while waiting for conditions to improve.
 */
static struct rpc_wait_queue delay_queue;

/*
 * rpciod-related stuff
 */
struct workqueue_struct *rpciod_workqueue __read_mostly;
struct workqueue_struct *xprtiod_workqueue __read_mostly;
EXPORT_SYMBOL_GPL(…);

gfp_t rpc_task_gfp_mask(void)
{ … }
EXPORT_SYMBOL_GPL(…);

bool rpc_task_set_rpc_status(struct rpc_task *task, int rpc_status)
{ … }

unsigned long
rpc_task_timeout(const struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

/*
 * Disable the timer for a given RPC task. Should be called with
 * queue->lock and bh_disabled in order to avoid races within
 * rpc_run_timer().
 */
static void
__rpc_disable_timer(struct rpc_wait_queue *queue, struct rpc_task *task)
{ … }

static void
rpc_set_queue_timer(struct rpc_wait_queue *queue, unsigned long expires)
{ … }

/*
 * Set up a timer for the current task.
 */
static void
__rpc_add_timer(struct rpc_wait_queue *queue, struct rpc_task *task,
		unsigned long timeout)
{ … }

static void rpc_set_waitqueue_priority(struct rpc_wait_queue *queue, int priority)
{ … }

static void rpc_reset_waitqueue_priority(struct rpc_wait_queue *queue)
{ … }

/*
 * Add a request to a queue list
 */
static void
__rpc_list_enqueue_task(struct list_head *q, struct rpc_task *task)
{ … }

/*
 * Remove request from a queue list
 */
static void
__rpc_list_dequeue_task(struct rpc_task *task)
{ … }

/*
 * Add new request to a priority queue.
 */
static void __rpc_add_wait_queue_priority(struct rpc_wait_queue *queue,
		struct rpc_task *task,
		unsigned char queue_priority)
{ … }

/*
 * Add new request to wait queue.
 */
static void __rpc_add_wait_queue(struct rpc_wait_queue *queue,
		struct rpc_task *task,
		unsigned char queue_priority)
{ … }

/*
 * Remove request from a priority queue.
 */
static void __rpc_remove_wait_queue_priority(struct rpc_task *task)
{ … }

/*
 * Remove request from queue.
 * Note: must be called with spin lock held.
 */
static void __rpc_remove_wait_queue(struct rpc_wait_queue *queue, struct rpc_task *task)
{ … }

static void __rpc_init_priority_wait_queue(struct rpc_wait_queue *queue, const char *qname, unsigned char nr_queues)
{ … }

void rpc_init_priority_wait_queue(struct rpc_wait_queue *queue, const char *qname)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_init_wait_queue(struct rpc_wait_queue *queue, const char *qname)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_destroy_wait_queue(struct rpc_wait_queue *queue)
{ … }
EXPORT_SYMBOL_GPL(…);

static int rpc_wait_bit_killable(struct wait_bit_key *key, int mode)
{ … }

#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) || IS_ENABLED(CONFIG_TRACEPOINTS)
static void rpc_task_set_debuginfo(struct rpc_task *task)
{ … }
#else
static inline void rpc_task_set_debuginfo(struct rpc_task *task)
{
}
#endif

static void rpc_set_active(struct rpc_task *task)
{ … }

/*
 * Mark an RPC call as having completed by clearing the 'active' bit
 * and then waking up all tasks that were sleeping.
 */
static int rpc_complete_task(struct rpc_task *task)
{ … }

/*
 * Allow callers to wait for completion of an RPC call
 *
 * Note the use of out_of_line_wait_on_bit() rather than wait_on_bit()
 * to enforce taking of the wq->lock and hence avoid races with
 * rpc_complete_task().
 */
int rpc_wait_for_completion_task(struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

/*
 * Make an RPC task runnable.
 *
 * Note: If the task is ASYNC, and is being made runnable after sitting on an
 * rpc_wait_queue, this must be called with the queue spinlock held to protect
 * the wait queue operation.
 * Note the ordering of rpc_test_and_set_running() and rpc_clear_queued(),
 * which is needed to ensure that __rpc_execute() doesn't loop (due to the
 * lockless RPC_IS_QUEUED() test) before we've had a chance to test
 * the RPC_TASK_RUNNING flag.
 */
static void rpc_make_runnable(struct workqueue_struct *wq,
		struct rpc_task *task)
{ … }

/*
 * Prepare for sleeping on a wait queue.
 * By always appending tasks to the list we ensure FIFO behavior.
 * NB: An RPC task will only receive interrupt-driven events as long
 * as it's on a wait queue.
 */
static void __rpc_do_sleep_on_priority(struct rpc_wait_queue *q,
		struct rpc_task *task,
		unsigned char queue_priority)
{ … }

static void __rpc_sleep_on_priority(struct rpc_wait_queue *q,
		struct rpc_task *task,
		unsigned char queue_priority)
{ … }

static void __rpc_sleep_on_priority_timeout(struct rpc_wait_queue *q,
		struct rpc_task *task, unsigned long timeout,
		unsigned char queue_priority)
{ … }

static void rpc_set_tk_callback(struct rpc_task *task, rpc_action action)
{ … }

static bool rpc_sleep_check_activated(struct rpc_task *task)
{ … }

void rpc_sleep_on_timeout(struct rpc_wait_queue *q, struct rpc_task *task,
				rpc_action action, unsigned long timeout)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_sleep_on(struct rpc_wait_queue *q, struct rpc_task *task,
				rpc_action action)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_sleep_on_priority_timeout(struct rpc_wait_queue *q,
		struct rpc_task *task, unsigned long timeout, int priority)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_sleep_on_priority(struct rpc_wait_queue *q, struct rpc_task *task,
		int priority)
{ … }
EXPORT_SYMBOL_GPL(…);

/**
 * __rpc_do_wake_up_task_on_wq - wake up a single rpc_task
 * @wq: workqueue on which to run task
 * @queue: wait queue
 * @task: task to be woken up
 *
 * Caller must hold queue->lock, and have cleared the task queued flag.
 */
static void __rpc_do_wake_up_task_on_wq(struct workqueue_struct *wq,
		struct rpc_wait_queue *queue,
		struct rpc_task *task)
{ … }

/*
 * Wake up a queued task while the queue lock is being held
 */
static struct rpc_task *
rpc_wake_up_task_on_wq_queue_action_locked(struct workqueue_struct *wq,
		struct rpc_wait_queue *queue, struct rpc_task *task,
		bool (*action)(struct rpc_task *, void *), void *data)
{ … }

/*
 * Wake up a queued task while the queue lock is being held
 */
static void rpc_wake_up_task_queue_locked(struct rpc_wait_queue *queue,
					  struct rpc_task *task)
{ … }

/*
 * Wake up a task on a specific queue
 */
void rpc_wake_up_queued_task(struct rpc_wait_queue *queue, struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

static bool rpc_task_action_set_status(struct rpc_task *task, void *status)
{ … }

static void
rpc_wake_up_task_queue_set_status_locked(struct rpc_wait_queue *queue,
		struct rpc_task *task, int status)
{ … }

/**
 * rpc_wake_up_queued_task_set_status - wake up a task and set task->tk_status
 * @queue: pointer to rpc_wait_queue
 * @task: pointer to rpc_task
 * @status: integer error value
 *
 * If @task is queued on @queue, then it is woken up, and @task->tk_status is
 * set to the value of @status.
 */
void
rpc_wake_up_queued_task_set_status(struct rpc_wait_queue *queue,
		struct rpc_task *task, int status)
{ … }

/*
 * Wake up the next task on a priority queue.
 */
static struct rpc_task *__rpc_find_next_queued_priority(struct rpc_wait_queue *queue)
{ … }

static struct rpc_task *__rpc_find_next_queued(struct rpc_wait_queue *queue)
{ … }

/*
 * Wake up the first task on the wait queue.
 */
struct rpc_task *rpc_wake_up_first_on_wq(struct workqueue_struct *wq,
		struct rpc_wait_queue *queue,
		bool (*func)(struct rpc_task *, void *), void *data)
{ … }

/*
 * Wake up the first task on the wait queue.
 */
struct rpc_task *rpc_wake_up_first(struct rpc_wait_queue *queue,
		bool (*func)(struct rpc_task *, void *), void *data)
{ … }
EXPORT_SYMBOL_GPL(…);

static bool rpc_wake_up_next_func(struct rpc_task *task, void *data)
{ … }

/*
 * Wake up the next task on the wait queue.
*/
struct rpc_task *rpc_wake_up_next(struct rpc_wait_queue *queue)
{ … }
EXPORT_SYMBOL_GPL(…);

/**
 * rpc_wake_up_locked - wake up all rpc_tasks
 * @queue: rpc_wait_queue on which the tasks are sleeping
 *
 */
static void rpc_wake_up_locked(struct rpc_wait_queue *queue)
{ … }

/**
 * rpc_wake_up - wake up all rpc_tasks
 * @queue: rpc_wait_queue on which the tasks are sleeping
 *
 * Grabs queue->lock
 */
void rpc_wake_up(struct rpc_wait_queue *queue)
{ … }
EXPORT_SYMBOL_GPL(…);

/**
 * rpc_wake_up_status_locked - wake up all rpc_tasks and set their status value.
 * @queue: rpc_wait_queue on which the tasks are sleeping
 * @status: status value to set
 */
static void rpc_wake_up_status_locked(struct rpc_wait_queue *queue, int status)
{ … }

/**
 * rpc_wake_up_status - wake up all rpc_tasks and set their status value.
 * @queue: rpc_wait_queue on which the tasks are sleeping
 * @status: status value to set
 *
 * Grabs queue->lock
 */
void rpc_wake_up_status(struct rpc_wait_queue *queue, int status)
{ … }
EXPORT_SYMBOL_GPL(…);

static void __rpc_queue_timer_fn(struct work_struct *work)
{ … }

static void __rpc_atrun(struct rpc_task *task)
{ … }

/*
 * Run a task at a later time
 */
void rpc_delay(struct rpc_task *task, unsigned long delay)
{ … }
EXPORT_SYMBOL_GPL(…);

/*
 * Helper to call task->tk_ops->rpc_call_prepare
 */
void rpc_prepare_task(struct rpc_task *task)
{ … }

static void
rpc_init_task_statistics(struct rpc_task *task)
{ … }

static void
rpc_reset_task_statistics(struct rpc_task *task)
{ … }

/*
 * Helper that calls task->tk_ops->rpc_call_done if it exists
 */
void rpc_exit_task(struct rpc_task *task)
{ … }

void rpc_signal_task(struct rpc_task *task)
{ … }

void rpc_task_try_cancel(struct rpc_task *task, int error)
{ … }

void rpc_exit(struct rpc_task *task, int status)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_release_calldata(const struct rpc_call_ops *ops, void *calldata)
{ … }

static bool xprt_needs_memalloc(struct rpc_xprt *xprt, struct rpc_task *tk)
{ … }

/*
 * This is the RPC `scheduler' (or rather, the finite state machine).
 */
static void __rpc_execute(struct rpc_task *task)
{ … }

/*
 * User-visible entry point to the scheduler.
 *
 * This may be called recursively if e.g. an async NFS task updates
 * the attributes and finds that dirty pages must be flushed.
 * NOTE: Upon exit of this function the task is guaranteed to be
 *	 released. In particular note that tk_release() will have
 *	 been called, so your task memory may have been freed.
 */
void rpc_execute(struct rpc_task *task)
{ … }

static void rpc_async_schedule(struct work_struct *work)
{ … }

/**
 * rpc_malloc - allocate RPC buffer resources
 * @task: RPC task
 *
 * A single memory region is allocated, which is split between the
 * RPC call and RPC reply that this task is being used for. When
 * this RPC is retired, the memory is released by calling rpc_free.
 *
 * To prevent rpciod from hanging, this allocator never sleeps,
 * returning -ENOMEM and suppressing warning if the request cannot
 * be serviced immediately. The caller can arrange to sleep in a
 * way that is safe for rpciod.
 *
 * Most requests are 'small' (under 2KiB) and can be serviced from a
 * mempool, ensuring that NFS reads and writes can always proceed,
 * and that there is good locality of reference for these buffers.
 */
int rpc_malloc(struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

/**
 * rpc_free - free RPC buffer resources allocated via rpc_malloc
 * @task: RPC task
 *
 */
void rpc_free(struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

/*
 * Creation and deletion of RPC task structures
 */
static void rpc_init_task(struct rpc_task *task, const struct rpc_task_setup *task_setup_data)
{ … }

static struct rpc_task *rpc_alloc_task(void)
{ … }

/*
 * Create a new task for the specified client.
 */
struct rpc_task *rpc_new_task(const struct rpc_task_setup *setup_data)
{ … }

/*
 * rpc_free_task - release rpc task and perform cleanups
 *
 * Note that we free up the rpc_task _after_ rpc_release_calldata()
 * in order to work around a workqueue dependency issue.
 *
 * Tejun Heo states:
 * "Workqueue currently considers two work items to be the same if they're
 * on the same address and won't execute them concurrently - ie. it
 * makes a work item which is queued again while being executed wait
 * for the previous execution to complete.
 *
 * If a work function frees the work item, and then waits for an event
 * which should be performed by another work item and *that* work item
 * recycles the freed work item, it can create a false dependency loop.
 * There really is no reliable way to detect this short of verifying
 * every memory free."
 *
 */
static void rpc_free_task(struct rpc_task *task)
{ … }

static void rpc_async_release(struct work_struct *work)
{ … }

static void rpc_release_resources_task(struct rpc_task *task)
{ … }

static void rpc_final_put_task(struct rpc_task *task,
		struct workqueue_struct *q)
{ … }

static void rpc_do_put_task(struct rpc_task *task, struct workqueue_struct *q)
{ … }

void rpc_put_task(struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

void rpc_put_task_async(struct rpc_task *task)
{ … }
EXPORT_SYMBOL_GPL(…);

static void rpc_release_task(struct rpc_task *task)
{ … }

int rpciod_up(void)
{ … }

void rpciod_down(void)
{ … }

/*
 * Start up the rpciod workqueue.
 */
static int rpciod_start(void)
{ … }

static void rpciod_stop(void)
{ … }

void
rpc_destroy_mempool(void)
{ … }

int
rpc_init_mempool(void)
{ … }