Source to osfmk/kern/thread.c
/*
* Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
*
* @APPLE_LICENSE_HEADER_START@
*
* The contents of this file constitute Original Code as defined in and
* are subject to the Apple Public Source License Version 1.1 (the
* "License"). You may not use this file except in compliance with the
* License. Please obtain a copy of the License at
* http://www.apple.com/publicsource and read it before using this file.
*
* This Original Code and all software distributed under the License are
* distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT. Please see the
* License for the specific language governing rights and limitations
* under the License.
*
* @APPLE_LICENSE_HEADER_END@
*/
/*
* @OSF_FREE_COPYRIGHT@
*/
/*
* Mach Operating System
* Copyright (c) 1991,1990,1989,1988,1987 Carnegie Mellon University
* All Rights Reserved.
*
* Permission to use, copy, modify and distribute this software and its
* documentation is hereby granted, provided that both the copyright
* notice and this permission notice appear in all copies of the
* software, derivative works or modified versions, and any portions
* thereof, and that both notices appear in supporting documentation.
*
* CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS "AS IS"
* CONDITION. CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND FOR
* ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.
*
* Carnegie Mellon requests users of this software to return to
*
* Software Distribution Coordinator or [email protected]
* School of Computer Science
* Carnegie Mellon University
* Pittsburgh PA 15213-3890
*
* any improvements or extensions that they make and grant Carnegie Mellon
* the rights to redistribute these changes.
*/
/*
*/
/*
* File: kern/thread.c
* Author: Avadis Tevanian, Jr., Michael Wayne Young, David Golub
* Date: 1986
*
* Thread/thread_shuttle management primitives implementation.
*/
/*
* Copyright (c) 1993 The University of Utah and
* the Computer Systems Laboratory (CSL). All rights reserved.
*
* Permission to use, copy, modify and distribute this software and its
* documentation is hereby granted, provided that both the copyright
* notice and this permission notice appear in all copies of the
* software, derivative works or modified versions, and any portions
* thereof, and that both notices appear in supporting documentation.
*
* THE UNIVERSITY OF UTAH AND CSL ALLOW FREE USE OF THIS SOFTWARE IN ITS "AS
* IS" CONDITION. THE UNIVERSITY OF UTAH AND CSL DISCLAIM ANY LIABILITY OF
* ANY KIND FOR ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.
*
* CSL requests users of this software to return to [email protected] any
* improvements that they make and grant CSL redistribution rights.
*
*/
#include <cpus.h>
#include <mach_host.h>
#include <simple_clock.h>
#include <mach_debug.h>
#include <mach_prof.h>
#include <dipc.h>
#include <stack_usage.h>
#include <mach/boolean.h>
#include <mach/policy.h>
#include <mach/thread_info.h>
#include <mach/thread_special_ports.h>
#include <mach/thread_status.h>
#include <mach/time_value.h>
#include <mach/vm_param.h>
#include <kern/ast.h>
#include <kern/cpu_data.h>
#include <kern/counters.h>
#include <kern/etap_macros.h>
#include <kern/ipc_mig.h>
#include <kern/ipc_tt.h>
#include <kern/mach_param.h>
#include <kern/machine.h>
#include <kern/misc_protos.h>
#include <kern/processor.h>
#include <kern/queue.h>
#include <kern/sched.h>
#include <kern/sched_prim.h>
#include <kern/sf.h>
#include <kern/mk_sp.h> /*** ??? fix so this can be removed ***/
#include <kern/task.h>
#include <kern/thread.h>
#include <kern/thread_act.h>
#include <kern/thread_swap.h>
#include <kern/host.h>
#include <kern/zalloc.h>
#include <vm/vm_kern.h>
#include <ipc/ipc_kmsg.h>
#include <ipc/ipc_port.h>
#include <machine/thread.h> /* for MACHINE_STACK */
#include <kern/profile.h>
#include <kern/assert.h>
#include <sys/kdebug.h>
/*
* Exported interfaces
*/
#include <mach/thread_act_server.h>
#include <mach/mach_host_server.h>
/*
* Per-Cpu stashed global state
*/
vm_offset_t active_stacks[NCPUS]; /* per-cpu active stacks */
vm_offset_t kernel_stack[NCPUS]; /* top of active stacks */
thread_act_t active_kloaded[NCPUS]; /* + act if kernel loaded */
decl_mutex_data(, funnel_lock)
struct zone *thread_shuttle_zone;
queue_head_t reaper_queue;
decl_simple_lock_data(,reaper_lock)
thread_call_t thread_reaper_call;
extern int tick;
extern void pcb_module_init(void);
/* private */
static struct thread_shuttle thr_sh_template;
#if MACH_DEBUG
#if STACK_USAGE
static void stack_init(vm_offset_t stack, unsigned int bytes);
void stack_finalize(vm_offset_t stack);
vm_size_t stack_usage(vm_offset_t stack);
#else /*STACK_USAGE*/
#define stack_init(stack, size)
#define stack_finalize(stack)
#define stack_usage(stack) (vm_size_t)0
#endif /*STACK_USAGE*/
#ifdef MACHINE_STACK
extern
#endif
void stack_statistics(
unsigned int *totalp,
vm_size_t *maxusagep);
#define STACK_MARKER 0xdeadbeef
#if STACK_USAGE
boolean_t stack_check_usage = TRUE;
#else /* STACK_USAGE */
boolean_t stack_check_usage = FALSE;
#endif /* STACK_USAGE */
decl_simple_lock_data(,stack_usage_lock)
vm_size_t stack_max_usage = 0;
vm_size_t stack_max_use = KERNEL_STACK_SIZE - 64;
#endif /* MACH_DEBUG */
/* Forwards */
void thread_collect_scan(void);
kern_return_t thread_create_shuttle(
thread_act_t thr_act,
sp_attributes_t policy_attributes,
void (*start_at)(void),
thread_t *new_thread);
extern void Load_context(
thread_t thread);
/*
* Machine-dependent code must define:
* thread_machine_init
* thread_machine_terminate
* thread_machine_collect
*
* The thread->pcb field is reserved for machine-dependent code.
*/
#ifdef MACHINE_STACK
/*
* Machine-dependent code must define:
* stack_alloc_try
* stack_alloc
* stack_free
* stack_collect
* and if MACH_DEBUG:
* stack_statistics
*/
#else /* MACHINE_STACK */
/*
* We allocate stacks from generic kernel VM.
* Machine-dependent code must define:
* machine_kernel_stack_init
*
* The stack_free_list can only be accessed at splsched,
* because stack_alloc_try/thread_invoke operate at splsched.
*/
decl_simple_lock_data(,stack_lock_data) /* splsched only */
#define stack_lock() simple_lock(&stack_lock_data)
#define stack_unlock() simple_unlock(&stack_lock_data)
vm_offset_t stack_free_list; /* splsched only */
unsigned int stack_free_count = 0; /* splsched only */
unsigned int stack_free_limit = 1; /* patchable */
unsigned int stack_alloc_hits = 0; /* debugging */
unsigned int stack_alloc_misses = 0; /* debugging */
unsigned int stack_alloc_max = 0; /* debugging */
unsigned int stack_alloc_total = 0;
unsigned int stack_alloc_hiwater = 0;
/*
* The next field is at the base of the stack,
* so the low end is left unsullied.
*/
#define stack_next(stack) (*((vm_offset_t *)((stack) + KERNEL_STACK_SIZE) - 1))
/*
* stack_alloc:
*
* Allocate a kernel stack for an activation.
* May block.
*/
vm_offset_t
stack_alloc(
thread_t thread,
void (*continuation)(void))
{
vm_offset_t stack;
spl_t s;
/*
* We first try the free list. It is probably empty,
* or stack_alloc_try would have succeeded, but possibly
* a stack was freed before the swapin thread got to us.
*/
s = splsched();
stack_lock();
stack = stack_free_list;
if (stack != 0) {
stack_free_list = stack_next(stack);
stack_free_count--;
}
stack_unlock();
splx(s);
if (stack == 0) {
/*
* Kernel stacks should be naturally aligned,
* so that it is easy to find the starting/ending
* addresses of a stack given an address in the middle.
*/
if (kmem_alloc_aligned(kernel_map, &stack,
round_page(KERNEL_STACK_SIZE)) != KERN_SUCCESS)
panic("stack_alloc");
stack_alloc_total++;
if (stack_alloc_total > stack_alloc_hiwater)
stack_alloc_hiwater = stack_alloc_total;
#if MACH_DEBUG
stack_init(stack, round_page(KERNEL_STACK_SIZE));
#endif /* MACH_DEBUG */
/*
* If using fractional pages, free the remainder(s)
*/
if (KERNEL_STACK_SIZE < round_page(KERNEL_STACK_SIZE)) {
vm_offset_t ptr = stack + KERNEL_STACK_SIZE;
vm_offset_t endp = stack + round_page(KERNEL_STACK_SIZE);
while (ptr < endp) {
#if MACH_DEBUG
/*
* We need to initialize just the end of the
* region.
*/
stack_init(ptr, (unsigned int) (endp - ptr));
#endif
stack_lock();
stack_next(stack) = stack_free_list;
stack_free_list = stack;
if (++stack_free_count > stack_alloc_max)
stack_alloc_max = stack_free_count;
stack_unlock();
ptr += KERNEL_STACK_SIZE;
}
}
}
stack_attach(thread, stack, continuation);
return (stack);
}
/*
* stack_free:
*
* Free a kernel stack.
* Called at splsched.
*/
void
stack_free(
thread_t thread)
{
vm_offset_t stack = stack_detach(thread);
assert(stack);
if (stack != thread->stack_privilege) {
stack_lock();
stack_next(stack) = stack_free_list;
stack_free_list = stack;
if (++stack_free_count > stack_alloc_max)
stack_alloc_max = stack_free_count;
stack_unlock();
}
}
/*
* stack_collect:
*
* Free excess kernel stacks.
* May block.
*/
void
stack_collect(void)
{
register vm_offset_t stack;
spl_t s;
/* If using fractional pages, Cannot just call kmem_free(),
* and we're too lazy to coalesce small chunks.
*/
if (KERNEL_STACK_SIZE < round_page(KERNEL_STACK_SIZE))
return;
s = splsched();
stack_lock();
while (stack_free_count > stack_free_limit) {
stack = stack_free_list;
stack_free_list = stack_next(stack);
stack_free_count--;
stack_unlock();
splx(s);
#if MACH_DEBUG
stack_finalize(stack);
#endif /* MACH_DEBUG */
kmem_free(kernel_map, stack, KERNEL_STACK_SIZE);
s = splsched();
stack_alloc_total--;
stack_lock();
}
stack_unlock();
splx(s);
}
#if MACH_DEBUG
/*
* stack_statistics:
*
* Return statistics on cached kernel stacks.
* *maxusagep must be initialized by the caller.
*/
void
stack_statistics(
unsigned int *totalp,
vm_size_t *maxusagep)
{
spl_t s;
s = splsched();
stack_lock();
#if STACK_USAGE
if (stack_check_usage) {
vm_offset_t stack;
/*
* This is pretty expensive to do at splsched,
* but it only happens when someone makes
* a debugging call, so it should be OK.
*/
for (stack = stack_free_list; stack != 0;
stack = stack_next(stack)) {
vm_size_t usage = stack_usage(stack);
if (usage > *maxusagep)
*maxusagep = usage;
}
}
#endif /* STACK_USAGE */
*totalp = stack_free_count;
stack_unlock();
splx(s);
}
#endif /* MACH_DEBUG */
#endif /* MACHINE_STACK */
/*
* stack_privilege:
*
* stack_alloc_try on this thread must always succeed.
*/
void
stack_privilege(
register thread_t thread)
{
/*
* This implementation only works for the current thread.
*/
if (thread != current_thread())
panic("stack_privilege");
if (thread->stack_privilege == 0)
thread->stack_privilege = current_stack();
}
/*
* stack_alloc_try:
*
* Non-blocking attempt to allocate a kernel stack.
* Called at splsched with the thread locked.
*/
boolean_t stack_alloc_try(
thread_t thread,
void (*resume)(void))
{
register vm_offset_t stack;
if ((stack = thread->stack_privilege) == (vm_offset_t)0) {
stack_lock();
stack = stack_free_list;
if (stack != (vm_offset_t)0) {
stack_free_list = stack_next(stack);
stack_free_count--;
}
stack_unlock();
}
if (stack != 0) {
stack_attach(thread, stack, resume);
stack_alloc_hits++;
return TRUE;
} else {
stack_alloc_misses++;
return FALSE;
}
}
void
thread_init(void)
{
thread_shuttle_zone = zinit(
sizeof(struct thread_shuttle),
THREAD_MAX * sizeof(struct thread_shuttle),
THREAD_CHUNK * sizeof(struct thread_shuttle),
"threads");
/*
* Fill in a template thread_shuttle for fast initialization.
* [Fields that must be (or are typically) reset at
* time of creation are so noted.]
*/
/* thr_sh_template.links (none) */
thr_sh_template.runq = RUN_QUEUE_NULL;
/* thr_sh_template.task (later) */
/* thr_sh_template.thread_list (later) */
/* thr_sh_template.pset_threads (later) */
/* one ref for being alive, one to return to the creator */
thr_sh_template.ref_count = 2;
thr_sh_template.wait_event = NO_EVENT;
thr_sh_template.wait_result = KERN_SUCCESS;
thr_sh_template.wait_queue = WAIT_QUEUE_NULL;
thr_sh_template.wake_active = FALSE;
/*thr_sh_template.state = TH_SUSP;*/
thr_sh_template.state = 0;
thr_sh_template.continuation = (void (*)(void))thread_bootstrap_return;
thr_sh_template.top_act = THR_ACT_NULL;
/* thr_sh_template.priority (later) */
/***??? thr_sh_template.max_priority = BASEPRI_USER; ***/
/* thr_sh_template.sched_pri (later - compute_priority) */
/***??? thr_sh_template.sched_data = 0; ***/
thr_sh_template.policy = POLICY_TIMESHARE;
/***??? thr_sh_template.depress_priority = -1; ***/
/***??? thr_sh_template.cpu_usage = 0; ***/
/***??? thr_sh_template.sched_usage = 0; ***/
/* thr_sh_template.sched_stamp (later) */
/***??? thr_sh_template.sched_change_stamp = 1; ***/
thr_sh_template.vm_privilege = FALSE;
/* thr_sh_template.<IPC structures> (later) */
timer_init(&(thr_sh_template.user_timer));
timer_init(&(thr_sh_template.system_timer));
thr_sh_template.user_timer_save.low = 0;
thr_sh_template.user_timer_save.high = 0;
thr_sh_template.system_timer_save.low = 0;
thr_sh_template.system_timer_save.high = 0;
thr_sh_template.cpu_delta = 0;
thr_sh_template.sched_delta = 0;
thr_sh_template.active = FALSE; /* reset */
/* thr_sh_template.processor_set (later) */
#if NCPUS > 1
thr_sh_template.bound_processor = PROCESSOR_NULL;
#endif /*NCPUS > 1*/
#if MACH_HOST
thr_sh_template.may_assign = TRUE;
thr_sh_template.assign_active = FALSE;
#endif /* MACH_HOST */
thr_sh_template.funnel_state = 0;
#if NCPUS > 1
/* thr_sh_template.last_processor (later) */
#endif /* NCPUS > 1 */
/*
* Initialize other data structures used in
* this module.
*/
queue_init(&reaper_queue);
simple_lock_init(&reaper_lock, ETAP_THREAD_REAPER);
mutex_init(&funnel_lock,0);
#ifndef MACHINE_STACK
simple_lock_init(&stack_lock_data, ETAP_THREAD_STACK);
#endif /* MACHINE_STACK */
#if MACH_DEBUG
simple_lock_init(&stack_usage_lock, ETAP_THREAD_STACK_USAGE);
#endif /* MACH_DEBUG */
#if MACH_LDEBUG
thr_sh_template.kthread = FALSE;
thr_sh_template.mutex_count = 0;
#endif /* MACH_LDEBUG */
/*
* Initialize any machine-dependent
* per-thread structures necessary.
*/
thread_machine_init();
}
void
thread_reaper_enqueue(
thread_t thread)
{
/*
* thread lock is already held, splsched()
* not necessary here.
*/
simple_lock(&reaper_lock);
enqueue_tail(&reaper_queue, (queue_entry_t)thread);
#if 0 /* CHECKME! */
/*
* Since thread has been put in the reaper_queue, it must no longer
* be preempted (otherwise, it could be put back in a run queue).
*/
thread->preempt = TH_NOT_PREEMPTABLE;
#endif
simple_unlock(&reaper_lock);
thread_call_enter(thread_reaper_call);
}
void
thread_terminate_self(void)
{
register thread_t thread = current_thread();
thread_act_t thr_act, prev_act;
task_t task;
sched_policy_t *policy;
spl_t s;
/*
* Check for rpc chain. If so, switch to the previous
* activation, set error code, switch stacks and jump
* to mach_rpc_return_error.
*/
thr_act = thread->top_act;
thread = thr_act->thread;
task = thr_act->task;
if (task) {
time_value_t user_time, system_time;
void *proc = NULL;
/*
* Accumulate times for dead threads into task.
*/
thread_read_times(thread, &user_time, &system_time);
task_lock(task);
time_value_add(&task->total_user_time, &user_time);
time_value_add(&task->total_system_time, &system_time);
if (task->thr_act_count == 1)
proc = task->bsd_info;
task_unlock(task);
if (proc)
proc_exit(proc);
}
thread = act_lock_thread(thr_act);
/* Unlink the thr_act from the task's thr_act list,
* so it doesn't appear in calls to task_threads and such.
* The thr_act still keeps its ref on the task, however.
*/
task_lock(task);
mutex_lock(&task->act_list_lock);
queue_remove(&task->thr_acts, thr_act, thread_act_t, thr_acts);
/*
* Decrement the act count for this task.
*/
task->thr_act_count--;
#if THREAD_SWAPPER
/*
* Thread is supposed to be unswappable by now...
*/
assert(thr_act->swap_state == TH_SW_UNSWAPPABLE ||
!(thread_swap_unwire_stack ||
thread_swap_unwire_user_stack));
#endif /* THREAD_SWAPPER */
task->res_act_count--;
thr_act->thr_acts.next = NULL;
mutex_unlock(&task->act_list_lock);
task_unlock(task);
#ifdef CALLOUT_RPC_MODEL
if (thr_act->lower) {
/*
* JMM - RPC will not be using a callout/stack manipulation
* mechanism. instead we will let it return normally as if
* from a continuation. Accordingly, these need to be cleaned
* up a bit.
*/
act_unlock(thr_act);
act_switch_swapcheck(thread, (ipc_port_t)0);
act_lock(thr_act); /* hierarchy violation XXX */
(void) switch_act(THR_ACT_NULL);
assert(thr_act->ref_count == 1); /* XXX */
/* act_deallocate(thr_act); XXX */
prev_act = thread->top_act;
/* disable preemption to protect kernel stack changes */
disable_preemption();
MACH_RPC_RET(prev_act) = KERN_RPC_SERVER_TERMINATED;
* machine_kernel_stack_init(thread,
* (void (*)(void)) mach_rpc_return_error);
* Load_context(thread);
*/
/* NOTREACHED */
}
#else /* !CALLOUT_RPC_MODEL */
assert(!thr_act->lower);
#endif /* CALLOUT_RPC_MODEL */
act_unlock_thread(thr_act);
s = splsched();
thread_lock(thread);
policy = &sched_policy[thread->policy];
thr_act = thread->top_act;
thread->active = FALSE;
thread_unlock(thread);
splx(s);
policy->sp_ops.sp_thread_depress_abort(policy, thread);
thread_cancel_timer();
/* flush any lazy HW state while in own context */
thread_machine_flush(thr_act);
/* Reap times from dying threads */
ipc_thr_act_disable(thr_act);
/*
* the test for task_active seems unnecessary because
* the thread holds a reference to the task (so it
* can't be deleted out from under it).
*/
if( task && task->active) {
#if THREAD_SWAPPER
thread_swap_disable(thr_act);
#endif /* THREAD_SWAPPER */
task_lock(task);
/* Make act inactive iff it was born as a base activation */
act_lock_thread(thr_act);
if( thr_act->active && (thr_act->pool_port == IP_NULL))
act_disable_task_locked( thr_act );
act_unlock_thread(thr_act);
task_unlock( task );
}
thread_deallocate(thread); /* take caller's ref; 1 left for reaper */
ipc_thread_terminate(thread);
s = splsched();
thread_lock(thread);
thread->state |= (TH_HALTED|TH_TERMINATE);
assert((thread->state & TH_UNINT) == 0);
#if 0 /* CHECKME! */
/*
* Since thread has been put in the reaper_queue, it must no longer
* be preempted (otherwise, it could be put back in a run queue).
*/
thread->preempt = TH_NOT_PREEMPTABLE;
#endif
thread_mark_wait_locked(thread);
thread_unlock(thread);
/* splx(s); */
ETAP_SET_REASON(thread, BLOCKED_ON_TERMINATION);
thread_block((void (*)(void)) 0);
panic("the zombie walks!");
/*NOTREACHED*/
}
/*
* Create a new thread in the specified activation (i.e. "populate" the
* activation). The activation can be either user or kernel, but it must
* be brand-new: no thread, no pool_port, nobody else knows about it.
* Doesn't start the thread running; use thread_setrun to start it.
*/
kern_return_t
thread_create_shuttle(
thread_act_t thr_act,
sp_attributes_t attributes,
void (*start_at)(void),
thread_t *new_thread)
{
thread_t new_shuttle;
task_t parent_task = thr_act->task;
processor_set_t pset;
kern_return_t result;
sched_policy_t *policy;
sf_return_t sfr;
int suspcnt;
/*
* Allocate a thread and initialize static fields
*/
new_shuttle = (thread_t)zalloc(thread_shuttle_zone);
if (new_shuttle == THREAD_NULL)
return (KERN_RESOURCE_SHORTAGE);
*new_shuttle = thr_sh_template;
/* Allocate space for scheduling information and attributes */
/*** Think about integrating with shuttle structure someday ***/
new_shuttle->sp_info = (sp_info_t)kalloc(max_sched_info_size);
if (new_shuttle->sp_info == SP_INFO_NULL) {
zfree(thread_shuttle_zone, (vm_offset_t)new_shuttle);
return (KERN_RESOURCE_SHORTAGE);
}
new_shuttle->pending_sched_attr =
(sp_attributes_t)kalloc(max_sched_attributes_size);
if (new_shuttle->pending_sched_attr == SP_ATTRIBUTES_NULL) {
kfree((vm_offset_t)new_shuttle->sp_info,
(vm_size_t)max_sched_info_size);
zfree(thread_shuttle_zone, (vm_offset_t)new_shuttle);
return (KERN_RESOURCE_SHORTAGE);
}
thread_lock_init(new_shuttle);
rpc_lock_init(new_shuttle);
wake_lock_init(new_shuttle);
new_shuttle->sleep_stamp = sched_tick;
/*
* No need to lock thr_act, since it can't be known to anyone --
* we set its suspend_count to one more than the task suspend_count
* by calling thread_hold.
*/
thr_act->user_stop_count = 1;
for (suspcnt = thr_act->task->suspend_count + 1; suspcnt; --suspcnt)
thread_hold(thr_act);
simple_lock_init(&new_shuttle->lock, ETAP_THREAD_NEW);
/*
* Initialize system-dependent part.
*/
result = thread_machine_create(new_shuttle, thr_act, start_at);
if (result != KERN_SUCCESS) {
kfree((vm_offset_t)new_shuttle->pending_sched_attr,
(vm_size_t)max_sched_attributes_size);
kfree((vm_offset_t)new_shuttle->sp_info,
(vm_size_t)max_sched_info_size);
zfree(thread_shuttle_zone, (vm_offset_t)new_shuttle);
return (result);
}
/* Attach the thread to the activation. */
assert(!thr_act->thread);
assert(!thr_act->pool_port);
/* Synchronize with act_lock_thread() et al. */
act_lock(thr_act);
/* Thread holds a ref to the thr_act */
act_locked_act_reference(thr_act);
act_attach(thr_act, new_shuttle, 0);
act_unlock(thr_act);
/*
* Initialize runtime-dependent fields
*/
thread_timer_setup(new_shuttle);
machine_kernel_stack_init(new_shuttle, (void (*)(void))thread_continue);
ipc_thread_init(new_shuttle);
thread_start(new_shuttle, start_at);
pset = parent_task->processor_set;
if (!pset->active) {
pset = &default_pset;
}
pset_lock(pset);
task_lock(parent_task);
if (attributes == SP_ATTRIBUTES_NULL)
attributes = parent_task->sp_attributes;
/* Associate the thread with that scheduling policy */
new_shuttle->policy = attributes->policy_id;
policy = &sched_policy[new_shuttle->policy];
sfr = policy->sp_ops.sp_thread_attach(policy, new_shuttle);
if (sfr != SF_SUCCESS)
panic("thread_create_shuttle: sp_thread_attach");
/* Indicate that no change in scheduling policy is pending */
new_shuttle->pending_policy = POLICY_NULL;
/* Associate the thread with the processor set */
sfr = policy->sp_ops.sp_thread_processor_set(policy, new_shuttle, pset);
if (sfr != SF_SUCCESS)
panic("thread_create_shuttle: sp_thread_proceessor_set");
/* Set the thread's scheduling parameters */
sfr = policy->sp_ops.sp_thread_set(policy, new_shuttle, attributes);
if (sfr != SF_SUCCESS)
panic("thread_create_shuttle: sp_thread_set");
/***
*** ??? Have to do something here. I want the call above to
*** send the parameters to the policy and have the policy do
*** its thing. Unfortunately, I can't see the "artful" way
*** to use the current MK SP routines to do this without
*** alteration.
***
*** I must return to this.
***
*** Perhaps a state bit should be associated with each thread
*** under the MK SP indicating whether that thread is runnable.
*** If it is not, `compute_priority()' or one of its siblings
*** is used to adjust scheduling parameter values; if it is
*** runnable, then the scheduling parameter adjustments can be
*** followed by code that tries to place the thread on the
*** appropriate run queue, or tries to run it immediately.
***/
#if ETAP_EVENT_MONITOR
new_thread->etap_reason = 0;
new_thread->etap_trace = FALSE;
#endif /* ETAP_EVENT_MONITOR */
new_shuttle->active = TRUE;
/*
* Don't need to initialize because the context switch
* code will set it before it can be used.
*/
if (!parent_task->active) {
task_unlock(parent_task);
pset_unlock(pset);
thread_deallocate(new_shuttle);
/* Drop ref we'd have given caller */
thread_deallocate(new_shuttle);
return (KERN_FAILURE);
}
task_unlock(parent_task);
pset_unlock(pset);
*new_thread = new_shuttle;
{
long dbg_arg1, dbg_arg2, dbg_arg3, dbg_arg4;
KERNEL_DEBUG_CONSTANT((TRACEDBG_CODE(DBG_TRACE_DATA, 1)) | DBG_FUNC_NONE,
new_shuttle, 0,0,0,0);
kdbg_trace_string(parent_task->bsd_info, &dbg_arg1, &dbg_arg2, &dbg_arg3,
&dbg_arg4);
KERNEL_DEBUG_CONSTANT((TRACEDBG_CODE(DBG_TRACE_STRING, 1)) | DBG_FUNC_NONE,
dbg_arg1, dbg_arg2, dbg_arg3, dbg_arg4, 0);
}
return (KERN_SUCCESS);
}
kern_return_t
thread_create(
task_t task,
thread_act_t *new_act)
{
thread_act_t thr_act;
thread_t thread;
kern_return_t result;
sched_policy_t *policy;
sf_return_t sfr;
spl_t s;
extern void thread_bootstrap_return(void);
result = act_create(task, NULL_PARAMS, &thr_act);
if (result != KERN_SUCCESS)
return (result);
result = thread_create_shuttle(thr_act, SP_ATTRIBUTES_NULL,
thread_bootstrap_return, &thread);
if (result != KERN_SUCCESS) {
thread_terminate(thr_act);
act_deallocate(thr_act);
return (result);
}
if (task->kernel_loaded)
thread_user_to_kernel(thread);
/* Start the thread running (it will immediately suspend itself). */
s = splsched();
thread_ast_set(thr_act, AST_APC);
thread_lock(thread);
thread->state |= TH_RUN; /*** ??? I think this is okay ***/
/* Allow the thread to execute */
policy = &sched_policy[thread->policy];
sfr = policy->sp_ops.sp_thread_dispatch(policy, thread);
if (sfr != SF_SUCCESS)
panic("thread_create: sp_thread_dispatch");
thread_unlock(thread);
splx(s);
/*****
act_lock_thread(thr_act);
thread_dowait( thr_act, FALSE);
act_unlock_thread(thr_act);
*****/
*new_act = thr_act;
return (KERN_SUCCESS);
}
/*
* Update thread that belongs to a task created via kernel_task_create().
*/
void
thread_user_to_kernel(
thread_t thread)
{
/*
* Used to set special swap_func here...
*/
}
kern_return_t
thread_create_running(
register task_t parent_task,
int flavor,
thread_state_t new_state,
mach_msg_type_number_t new_state_count,
thread_act_t *child_act) /* OUT */
{
register kern_return_t result;
result = thread_create(parent_task, child_act);
if (result != KERN_SUCCESS)
return (result);
result = act_set_state(*child_act, flavor, new_state, new_state_count);
if (result != KERN_SUCCESS) {
(void) thread_terminate(*child_act);
return (result);
}
result = thread_resume(*child_act);
if (result != KERN_SUCCESS) {
(void) thread_terminate(*child_act);
return (result);
}
return (result);
}
/*
* kernel_thread:
*
* Create and kernel thread in the specified task, and
* optionally start it running.
*/
thread_t
kernel_thread_with_attributes(
task_t task,
sp_attributes_t attributes,
void (*start_at)(void),
boolean_t start_running)
{
kern_return_t result;
thread_t thread;
thread_act_t thr_act;
sched_policy_t *policy;
sf_return_t sfr;
spl_t s;
result = act_create(task, NULL_PARAMS, &thr_act);
if (result != KERN_SUCCESS) {
printf("kernel_thread act_create %x\n", result);
panic("act_create failure");
}
result = thread_create_shuttle(thr_act, attributes, start_at, &thread);
if (result != KERN_SUCCESS) {
printf("kernel_thread create_shuttle %x\n", result);
panic("create_shuttle failure");
}
thread_swappable(thr_act, FALSE);
s = splsched();
thread_lock(thread);
thr_act = thread->top_act;
#if MACH_LDEBUG
thread->kthread = TRUE;
#endif /* MACH_LDEBUG */
if (start_running) {
policy = &sched_policy[thread->policy];
thread->state |= TH_RUN;
sfr = policy->sp_ops.sp_thread_unblock(policy, thread);
if (sfr != SF_SUCCESS)
panic("kernel_thread: sp_thread_unblock");
}
thread_unlock(thread);
splx(s);
act_deallocate(thr_act);
if (start_running)
thread_resume(thr_act);
return (thread);
}
thread_t
kernel_thread(
task_t task,
void (*start_at)(void))
{
return kernel_thread_with_attributes(
task, SP_ATTRIBUTES_NULL, start_at, TRUE);
}
unsigned int c_weird_pset_ref_exit = 0; /* pset code raced us */
void
thread_deallocate(
thread_t thread)
{
task_t task;
processor_set_t pset;
sched_policy_t *policy;
sf_return_t sfr;
spl_t s;
if (thread == THREAD_NULL)
return;
/*
* First, check for new count > 0 (the common case).
* Only the thread needs to be locked.
*/
s = splsched();
thread_lock(thread);
if (--thread->ref_count > 0) {
thread_unlock(thread);
splx(s);
return;
}
/*
* Count is zero. However, the processor set's
* thread list has an implicit reference to
* the thread, and may make new ones. Its lock also
* dominate the thread lock. To check for this, we
* temporarily restore the one thread reference, unlock
* the thread, and then lock the pset in the proper order.
*/
assert(thread->ref_count == 0); /* Else this is an extra dealloc! */
thread->ref_count++;
thread_unlock(thread);
splx(s);
#if MACH_HOST
thread_freeze(thread);
#endif /* MACH_HOST */
pset = thread->processor_set;
pset_lock(pset);
s = splsched();
thread_lock(thread);
if (--thread->ref_count > 0) {
#if MACH_HOST
boolean_t need_wakeup = FALSE;
/*
* processor_set made extra reference.
*/
/* Inline the unfreeze */
thread->may_assign = TRUE;
if (thread->assign_active) {
need_wakeup = TRUE;
thread->assign_active = FALSE;
}
#endif /* MACH_HOST */
thread_unlock(thread);
splx(s);
pset_unlock(pset);
#if MACH_HOST
if (need_wakeup)
thread_wakeup((event_t)&thread->assign_active);
#endif /* MACH_HOST */
c_weird_pset_ref_exit++;
return;
}
#if MACH_HOST
assert(thread->assign_active == FALSE);
#endif /* MACH_HOST */
/*
* Thread has no references - we can remove it.
*/
/*
* A quick sanity check
*/
if (thread == current_thread())
panic("thread deallocating itself");
/* Detach thread (shuttle) from its sched policy */
policy = &sched_policy[thread->policy];
sfr = policy->sp_ops.sp_thread_detach(policy, thread);
if (sfr != SF_SUCCESS)
panic("thread_deallocate: sp_thread_detach");
/* Release storage used for scheduling info and attributes */
assert(thread->pending_sched_attr != SP_ATTRIBUTES_NULL);
kfree((vm_offset_t) thread->pending_sched_attr,
(vm_size_t) max_sched_attributes_size);
assert(thread->sp_info != SP_INFO_NULL);
kfree((vm_offset_t) thread->sp_info,
(vm_size_t) max_sched_info_size);
pset_remove_thread(pset, thread);
thread_unlock(thread); /* no more references - safe */
splx(s);
pset_unlock(pset);
pset_deallocate(thread->processor_set);
/* frees kernel stack & other MD resources */
thread->stack_privilege = 0;
thread_machine_destroy(thread);
zfree(thread_shuttle_zone, (vm_offset_t) thread);
}
void
thread_reference(
thread_t thread)
{
spl_t s;
if (thread == THREAD_NULL)
return;
s = splsched();
thread_lock(thread);
thread->ref_count++;
thread_unlock(thread);
splx(s);
}
/*
* Called with "appropriate" thread-related locks held on
* thread and its top_act for synchrony with RPC (see
* act_lock_thread()).
*/
kern_return_t
thread_info_shuttle(
register thread_act_t thr_act,
thread_flavor_t flavor,
thread_info_t thread_info_out, /* ptr to OUT array */
mach_msg_type_number_t *thread_info_count) /*IN/OUT*/
{
register thread_t thread = thr_act->thread;
int state, flags;
spl_t s;
if (thread == THREAD_NULL)
return (KERN_INVALID_ARGUMENT);
if (flavor == THREAD_BASIC_INFO) {
register thread_basic_info_t basic_info;
if (*thread_info_count < THREAD_BASIC_INFO_COUNT)
return (KERN_INVALID_ARGUMENT);
basic_info = (thread_basic_info_t) thread_info_out;
s = splsched();
thread_lock(thread);
/* fill in info */
thread_read_times(thread, &basic_info->user_time,
&basic_info->system_time);
/*** ??? fix me ***/
if (thread->policy & (POLICY_TIMESHARE|POLICY_RR|POLICY_FIFO)) {
mk_sp_info_t sp_info = (mk_sp_info_t)thread->sp_info;
/*
* Update lazy-evaluated scheduler info because someone wants it.
*/
assert(sp_info != SP_INFO_NULL);
if (sp_info->sched_stamp != sched_tick)
update_priority(thread);
basic_info->sleep_time = 0;
/*
* To calculate cpu_usage, first correct for timer rate,
* then for 5/8 ageing. The correction factor [3/5] is
* (1/(5/8) - 1).
*/
basic_info->cpu_usage = sp_info->cpu_usage /
(TIMER_RATE / TH_USAGE_SCALE);
basic_info->cpu_usage = (basic_info->cpu_usage * 3) / 5;
#if SIMPLE_CLOCK
/*
* Clock drift compensation.
*/
basic_info->cpu_usage =
(basic_info->cpu_usage * 1000000) / sched_usec;
#endif /* SIMPLE_CLOCK */
}
else
basic_info->sleep_time = basic_info->cpu_usage = 0;
basic_info->policy = thread->policy;
flags = 0;
if (thread->state & TH_SWAPPED_OUT)
flags = TH_FLAGS_SWAPPED;
else
if (thread->state & TH_IDLE)
flags = TH_FLAGS_IDLE;
state = 0;
if (thread->state & TH_HALTED)
state = TH_STATE_HALTED;
else
if (thread->state & TH_RUN)
state = TH_STATE_RUNNING;
else
if (thread->state & TH_UNINT)
state = TH_STATE_UNINTERRUPTIBLE;
else
if (thread->state & TH_SUSP)
state = TH_STATE_STOPPED;
else
if (thread->state & TH_WAIT)
state = TH_STATE_WAITING;
basic_info->run_state = state;
basic_info->flags = flags;
basic_info->suspend_count = thr_act->user_stop_count;
thread_unlock(thread);
splx(s);
*thread_info_count = THREAD_BASIC_INFO_COUNT;
return (KERN_SUCCESS);
}
else
if (flavor == THREAD_SCHED_TIMESHARE_INFO) {
policy_timeshare_info_t ts_info;
mk_sp_info_t sp_info = (mk_sp_info_t)thread->sp_info;
if (*thread_info_count < POLICY_TIMESHARE_INFO_COUNT)
return (KERN_INVALID_ARGUMENT);
ts_info = (policy_timeshare_info_t)thread_info_out;
s = splsched();
thread_lock(thread);
if (thread->policy != POLICY_TIMESHARE) {
thread_unlock(thread);
splx(s);
return (KERN_INVALID_POLICY);
}
/*** ??? fix me ***/
assert(sp_info != SP_INFO_NULL);
ts_info->base_priority = sp_info->priority;
ts_info->max_priority = sp_info->max_priority;
ts_info->cur_priority = thread->sched_pri;
ts_info->depressed = (sp_info->depress_priority >= 0);
ts_info->depress_priority = sp_info->depress_priority;
thread_unlock(thread);
splx(s);
*thread_info_count = POLICY_TIMESHARE_INFO_COUNT;
return (KERN_SUCCESS);
}
else
if (flavor == THREAD_SCHED_FIFO_INFO) {
policy_fifo_info_t fifo_info;
mk_sp_info_t sp_info = (mk_sp_info_t)thread->sp_info;
if (*thread_info_count < POLICY_FIFO_INFO_COUNT)
return (KERN_INVALID_ARGUMENT);
fifo_info = (policy_fifo_info_t)thread_info_out;
s = splsched();
thread_lock(thread);
if (thread->policy != POLICY_FIFO) {
thread_unlock(thread);
splx(s);
return (KERN_INVALID_POLICY);
}
/*** ??? fix me ***/
assert(sp_info != SP_INFO_NULL);
fifo_info->base_priority = sp_info->priority;
fifo_info->max_priority = sp_info->max_priority;
fifo_info->depressed = (sp_info->depress_priority >= 0);
fifo_info->depress_priority = sp_info->depress_priority;
thread_unlock(thread);
splx(s);
*thread_info_count = POLICY_FIFO_INFO_COUNT;
return (KERN_SUCCESS);
}
else
if (flavor == THREAD_SCHED_RR_INFO) {
policy_rr_info_t rr_info;
mk_sp_info_t sp_info = (mk_sp_info_t)thread->sp_info;
if (*thread_info_count < POLICY_RR_INFO_COUNT)
return (KERN_INVALID_ARGUMENT);
rr_info = (policy_rr_info_t) thread_info_out;
s = splsched();
thread_lock(thread);
if (thread->policy != POLICY_RR) {
thread_unlock(thread);
splx(s);
return (KERN_INVALID_POLICY);
}
/*** ??? fix me ***/
assert(sp_info != SP_INFO_NULL);
rr_info->base_priority = sp_info->priority;
rr_info->max_priority = sp_info->max_priority;
rr_info->quantum = (sp_info->sched_data * tick) / 1000;
rr_info->depressed = (sp_info->depress_priority >= 0);
rr_info->depress_priority = sp_info->depress_priority;
thread_unlock(thread);
splx(s);
*thread_info_count = POLICY_RR_INFO_COUNT;
return (KERN_SUCCESS);
}
return (KERN_INVALID_ARGUMENT);
}
void
thread_doreap(
register thread_t thread)
{
thread_act_t thr_act;
struct ipc_port *pool_port;
thr_act = thread_lock_act(thread);
assert(thr_act && thr_act->thread == thread);
act_locked_act_reference(thr_act);
pool_port = thr_act->pool_port;
/*
* Replace `act_unlock_thread()' with individual
* calls. (`act_detach()' can change fields used
* to determine which locks are held, confusing
* `act_unlock_thread()'.)
*/
rpc_unlock(thread);
if (pool_port != IP_NULL)
ip_unlock(pool_port);
act_unlock(thr_act);
/* Remove the reference held by a rooted thread */
if (pool_port == IP_NULL)
act_deallocate(thr_act);
/* Remove the reference held by the thread: */
act_deallocate(thr_act);
}
static thread_call_data_t thread_reaper_call_data;
/*
* reaper_thread:
*
* This kernel thread runs forever looking for threads to destroy
* (when they request that they be destroyed, of course).
*
* The reaper thread will disappear in the next revision of thread
* control when it's function will be moved into thread_dispatch.
*/
void
thread_reaper(void)
{
register thread_t thread;
spl_t s;
s = splsched();
simple_lock(&reaper_lock);
if (thread_reaper_call == NULL) {
thread_call_setup(&thread_reaper_call_data, thread_reaper, NULL);
thread_reaper_call = &thread_reaper_call_data;
}
while ((thread = (thread_t) dequeue_head(&reaper_queue)) != THREAD_NULL) {
simple_unlock(&reaper_lock);
/*
* wait for run bit to clear
*/
thread_lock(thread);
if (thread->state & TH_RUN)
panic("thread reaper: TH_RUN");
thread_unlock(thread);
splx(s);
thread_doreap(thread);
s = splsched();
simple_lock(&reaper_lock);
}
simple_unlock(&reaper_lock);
splx(s);
}
#if MACH_HOST
/*
* thread_assign:
*
* Change processor set assignment.
* Caller must hold an extra reference to the thread (if this is
* called directly from the ipc interface, this is an operation
* in progress reference). Caller must hold no locks -- this may block.
*/
kern_return_t
thread_assign(
thread_act_t thr_act,
processor_set_t new_pset)
{
thread_t thread;
if (thr_act == THR_ACT_NULL || new_pset == PROCESSOR_SET_NULL)
return(KERN_INVALID_ARGUMENT);
thread = act_lock_thread(thr_act);
if (thread == THREAD_NULL) {
act_unlock_thread(thr_act);
return(KERN_INVALID_ARGUMENT);
}
thread_freeze(thread);
thread_doassign(thread, new_pset, TRUE);
act_unlock_thread(thr_act);
return(KERN_SUCCESS);
}
/*
* thread_freeze:
*
* Freeze thread's assignment. Prelude to assigning thread.
* Only one freeze may be held per thread.
*/
void
thread_freeze(
thread_t thread)
{
spl_t s;
/*
* Freeze the assignment, deferring to a prior freeze.
*/
s = splsched();
thread_lock(thread);
while (thread->may_assign == FALSE) {
thread->assign_active = TRUE;
thread_sleep_simple_lock((event_t) &thread->assign_active,
simple_lock_addr(thread->lock), TRUE);
thread_lock(thread);
}
thread->may_assign = FALSE;
thread_unlock(thread);
splx(s);
}
/*
* thread_unfreeze: release freeze on thread's assignment.
*/
void
thread_unfreeze(
thread_t thread)
{
spl_t s;
s = splsched();
thread_lock(thread);
thread->may_assign = TRUE;
if (thread->assign_active) {
thread->assign_active = FALSE;
thread_unlock(thread);
splx(s);
thread_wakeup((event_t)&thread->assign_active);
return;
}
thread_unlock(thread);
splx(s);
}
/*
* thread_doassign:
*
* Actually do thread assignment. thread_will_assign must have been
* called on the thread. release_freeze argument indicates whether
* to release freeze on thread.
*
* Called with "appropriate" thread-related locks held on thread (see
* act_lock_thread()). Returns with thread unlocked.
*/
void
thread_doassign(
register thread_t thread,
register processor_set_t new_pset,
boolean_t release_freeze)
{
register boolean_t old_empty, new_empty;
register processor_set_t pset;
boolean_t recompute_pri = FALSE;
int max_priority;
spl_t s;
thread_act_t thr_act = thread->top_act;
/*
* Check for silly no-op.
*/
pset = thread->processor_set;
if (pset == new_pset) {
if (release_freeze)
thread_unfreeze(thread);
return;
}
/*
* Suspend the thread and stop it if it's not the current thread.
*/
thread_hold(thr_act);
act_locked_act_reference(thr_act);
act_unlock_thread(thr_act);
if (thread != current_thread()) {
if (thread_stop_wait(thread) == FALSE ){
(void)act_lock_thread(thr_act);
thread_release(thr_act);
act_locked_act_deallocate(thr_act);
act_unlock_thread(thr_act);
if (release_freeze )
thread_unfreeze(thread);
return;
}
}
/*
* Had to release thread-related locks before acquiring pset
* locks.
*/
/*
* Lock both psets now, use ordering to avoid deadlocks.
*/
Restart:
if (pset < new_pset) {
pset_lock(pset);
pset_lock(new_pset);
} else {
pset_lock(new_pset);
pset_lock(pset);
}
/*
* Check if new_pset is ok to assign to. If not, reassign
* to default_pset.
*/
if (!new_pset->active) {
pset_unlock(pset);
pset_unlock(new_pset);
new_pset = &default_pset;
goto Restart;
}
/*
* Grab the thread lock and move the thread.
* Then drop the lock on the old pset and the thread's
* reference to it.
*/
s = splsched();
thread_lock(thread);
thread_change_psets(thread, pset, new_pset);
old_empty = pset->empty;
new_empty = new_pset->empty;
pset_unlock(pset);
pset_deallocate(pset);
/*
* Reset policy and priorities if needed.
*
* There are three rules for threads under assignment:
*
* (1) If the new pset has the old policy enabled, keep the
* old policy. Otherwise, use the default policy for the pset.
* (2) The new limits will be the pset limits for the new policy.
* (3) The new base will be the same as the old base unless either
* (a) the new policy changed to the pset default policy;
* in this case, the new base is the default policy
* base,
* or
* (b) the new limits are different from the old limits;
* in this case, the new base is the new limits.
*/
/*** ??? fix me to fit into MK Scheduling Framework ***/
max_priority = pset_max_priority(new_pset, thread->policy);
if ((thread->policy & new_pset->policies) == 0) {
thread->policy = new_pset->policy_default;
thread->sched_data =
pset_sched_data(new_pset, thread->policy);
thread->unconsumed_quantum = thread->sched_data;
thread->priority =
pset_base_priority(new_pset, thread->policy);
max_priority = pset_max_priority(new_pset, thread->policy);
recompute_pri = TRUE;
}
else if (thread->max_priority != max_priority) {
thread->priority = max_priority;
recompute_pri = TRUE;
}
thread->max_priority = max_priority;
if ((thread->depress_priority >= 0) &&
(thread->depress_priority > thread->max_priority)) {
thread->depress_priority = thread->max_priority;
}
pset_unlock(new_pset);
if (recompute_pri)
compute_priority(thread, TRUE);
if (release_freeze) {
boolean_t need_wakeup = FALSE;
thread->may_assign = TRUE;
if (thread->assign_active) {
thread->assign_active = FALSE;
need_wakeup = TRUE;
}
thread_unlock(thread);
splx(s);
if (need_wakeup)
thread_wakeup((event_t)&thread->assign_active);
} else {
thread_unlock(thread);
splx(s);
}
if (thread != current_thread())
thread_unstop(thread);
/*
* Figure out hold status of thread. Threads assigned to empty
* psets must be held. Therefore:
* If old pset was empty release its hold.
* Release our hold from above unless new pset is empty.
*/
(void)act_lock_thread(thr_act);
if (old_empty)
thread_release(thr_act);
if (!new_empty)
thread_release(thr_act);
act_locked_act_deallocate(thr_act);
act_unlock_thread(thr_act);
/*
* If current_thread is assigned, context switch to force
* assignment to happen. This also causes hold to take
* effect if the new pset is empty.
*/
if (thread == current_thread()) {
s = splsched();
mp_disable_preemption();
ast_on(AST_BLOCK);
mp_enable_preemption();
splx(s);
}
}
#else /* MACH_HOST */
kern_return_t
thread_assign(
thread_act_t thr_act,
processor_set_t new_pset)
{
#ifdef lint
thread++; new_pset++;
#endif /* lint */
return(KERN_FAILURE);
}
#endif /* MACH_HOST */
/*
* thread_assign_default:
*
* Special version of thread_assign for assigning threads to default
* processor set.
*/
kern_return_t
thread_assign_default(
thread_act_t thr_act)
{
return (thread_assign(thr_act, &default_pset));
}
/*
* thread_get_assignment
*
* Return current assignment for this thread.
*/
kern_return_t
thread_get_assignment(
thread_act_t thr_act,
processor_set_t *pset)
{
thread_t thread;
if (thr_act == THR_ACT_NULL)
return(KERN_INVALID_ARGUMENT);
thread = act_lock_thread(thr_act);
if (thread == THREAD_NULL) {
act_unlock_thread(thr_act);
return(KERN_INVALID_ARGUMENT);
}
*pset = thread->processor_set;
act_unlock_thread(thr_act);
pset_reference(*pset);
return(KERN_SUCCESS);
}
/*
* thread_wire:
*
* Specify that the target thread must always be able
* to run and to allocate memory.
*/
kern_return_t
thread_wire(
host_t host,
thread_act_t thr_act,
boolean_t wired)
{
spl_t s;
thread_t thread;
extern void vm_page_free_reserve(int pages);
if (thr_act == THR_ACT_NULL || host == HOST_NULL)
return (KERN_INVALID_ARGUMENT);
thread = act_lock_thread(thr_act);
if (thread ==THREAD_NULL) {
act_unlock_thread(thr_act);
return(KERN_INVALID_ARGUMENT);
}
/*
* This implementation only works for the current thread.
* See stack_privilege.
*/
if (thr_act != current_act())
return KERN_INVALID_ARGUMENT;
s = splsched();
thread_lock(thread);
if (wired) {
if (thread->vm_privilege == FALSE)
vm_page_free_reserve(1); /* XXX */
thread->vm_privilege = TRUE;
} else {
if (thread->vm_privilege == TRUE)
vm_page_free_reserve(-1); /* XXX */
thread->vm_privilege = FALSE;
}
thread_unlock(thread);
splx(s);
act_unlock_thread(thr_act);
/*
* Make the thread unswappable.
*/
thread_swappable(thr_act, FALSE);
return KERN_SUCCESS;
}
/*
* thread_collect_scan:
*
* Attempt to free resources owned by threads.
*/
void
thread_collect_scan(void)
{
/* This code runs very quickly! */
}
boolean_t thread_collect_allowed = TRUE;
unsigned thread_collect_last_tick = 0;
unsigned thread_collect_max_rate = 0; /* in ticks */
/*
* consider_thread_collect:
*
* Called by the pageout daemon when the system needs more free pages.
*/
void
consider_thread_collect(void)
{
/*
* By default, don't attempt thread collection more frequently
* than once a second (one scheduler tick).
*/
if (thread_collect_max_rate == 0)
thread_collect_max_rate = 2; /* sched_tick is a 1 second resolution 2 here insures at least 1 second interval */
if (thread_collect_allowed &&
(sched_tick >
(thread_collect_last_tick + thread_collect_max_rate))) {
thread_collect_last_tick = sched_tick;
thread_collect_scan();
}
}
#if MACH_DEBUG
#if STACK_USAGE
vm_size_t
stack_usage(
register vm_offset_t stack)
{
int i;
for (i = 0; i < KERNEL_STACK_SIZE/sizeof(unsigned int); i++)
if (((unsigned int *)stack)[i] != STACK_MARKER)
break;
return KERNEL_STACK_SIZE - i * sizeof(unsigned int);
}
/*
* Machine-dependent code should call stack_init
* before doing its own initialization of the stack.
*/
static void
stack_init(
register vm_offset_t stack,
unsigned int bytes)
{
if (stack_check_usage) {
int i;
for (i = 0; i < bytes / sizeof(unsigned int); i++)
((unsigned int *)stack)[i] = STACK_MARKER;
}
}
/*
* Machine-dependent code should call stack_finalize
* before releasing the stack memory.
*/
void
stack_finalize(
register vm_offset_t stack)
{
if (stack_check_usage) {
vm_size_t used = stack_usage(stack);
simple_lock(&stack_usage_lock);
if (used > stack_max_usage)
stack_max_usage = used;
simple_unlock(&stack_usage_lock);
if (used > stack_max_use) {
printf("stack usage = %x\n", used);
panic("stack overflow");
}
}
}
#endif /*STACK_USAGE*/
#endif /* MACH_DEBUG */
kern_return_t
host_stack_usage(
host_t host,
vm_size_t *reservedp,
unsigned int *totalp,
vm_size_t *spacep,
vm_size_t *residentp,
vm_size_t *maxusagep,
vm_offset_t *maxstackp)
{
#if !MACH_DEBUG
return KERN_NOT_SUPPORTED;
#else
unsigned int total;
vm_size_t maxusage;
if (host == HOST_NULL)
return KERN_INVALID_HOST;
simple_lock(&stack_usage_lock);
maxusage = stack_max_usage;
simple_unlock(&stack_usage_lock);
stack_statistics(&total, &maxusage);
*reservedp = 0;
*totalp = total;
*spacep = *residentp = total * round_page(KERNEL_STACK_SIZE);
*maxusagep = maxusage;
*maxstackp = 0;
return KERN_SUCCESS;
#endif /* MACH_DEBUG */
}
/*
* Return info on stack usage for threads in a specific processor set
*/
kern_return_t
processor_set_stack_usage(
processor_set_t pset,
unsigned int *totalp,
vm_size_t *spacep,
vm_size_t *residentp,
vm_size_t *maxusagep,
vm_offset_t *maxstackp)
{
#if !MACH_DEBUG
return KERN_NOT_SUPPORTED;
#else
unsigned int total;
vm_size_t maxusage;
vm_offset_t maxstack;
register thread_t *threads;
register thread_t thread;
unsigned int actual; /* this many things */
unsigned int i;
vm_size_t size, size_needed;
vm_offset_t addr;
if (pset == PROCESSOR_SET_NULL)
return KERN_INVALID_ARGUMENT;
size = 0; addr = 0;
for (;;) {
pset_lock(pset);
if (!pset->active) {
pset_unlock(pset);
return KERN_INVALID_ARGUMENT;
}
actual = pset->thread_count;
/* do we have the memory we need? */
size_needed = actual * sizeof(thread_t);
if (size_needed <= size)
break;
/* unlock the pset and allocate more memory */
pset_unlock(pset);
if (size != 0)
kfree(addr, size);
assert(size_needed > 0);
size = size_needed;
addr = kalloc(size);
if (addr == 0)
return KERN_RESOURCE_SHORTAGE;
}
/* OK, have memory and the processor_set is locked & active */
threads = (thread_t *) addr;
for (i = 0, thread = (thread_t) queue_first(&pset->threads);
i < actual;
i++,
thread = (thread_t) queue_next(&thread->pset_threads)) {
thread_reference(thread);
threads[i] = thread;
}
assert(queue_end(&pset->threads, (queue_entry_t) thread));
/* can unlock processor set now that we have the thread refs */
pset_unlock(pset);
/* calculate maxusage and free thread references */
total = 0;
maxusage = 0;
maxstack = 0;
for (i = 0; i < actual; i++) {
int cpu;
thread_t thread = threads[i];
vm_offset_t stack = 0;
/*
* thread->kernel_stack is only accurate if the
* thread isn't swapped and is not executing.
*
* Of course, we don't have the appropriate locks
* for these shenanigans.
*/
stack = thread->kernel_stack;
for (cpu = 0; cpu < NCPUS; cpu++)
if (cpu_data[cpu].active_thread == thread) {
stack = active_stacks[cpu];
break;
}
if (stack != 0) {
total++;
if (stack_check_usage) {
vm_size_t usage = stack_usage(stack);
if (usage > maxusage) {
maxusage = usage;
maxstack = (vm_offset_t) thread;
}
}
}
thread_deallocate(thread);
}
if (size != 0)
kfree(addr, size);
*totalp = total;
*residentp = *spacep = total * round_page(KERNEL_STACK_SIZE);
*maxusagep = maxusage;
*maxstackp = maxstack;
return KERN_SUCCESS;
#endif /* MACH_DEBUG */
}
/*
* We consider a thread not preemptable if it is marked as either
* suspended or waiting.
*/
boolean_t thread_not_preemptable(
thread_t thread)
{
/* XXX - when scheduling framework and such is done, the
thread state check can be eliminated */
if ((thread->state & (TH_WAIT|TH_SUSP)) || thread->preempt)
return (TRUE);
else
return (FALSE);
}
/*
* thread_set_sched
*
* Set scheduling policy and parameters for the given thread.
* Policy must be a policy which is enabled for the
* processor set.
* (This should replace `thread_set_policy()' with the addition
* of the MK Scheduling Framework to the kernel.)
*/
kern_return_t
thread_set_sched(
thread_act_t thr_act,
policy_t policy_id,
sched_attr_t sched_attr,
mach_msg_type_number_t sched_attrCnt)
{
thread_t thread;
processor_set_t pset;
kern_return_t result = KERN_SUCCESS;
sched_policy_t *policy;
sf_return_t sfr;
boolean_t do_dispatch = FALSE;
spl_t s;
if (thr_act == THR_ACT_NULL)
return (KERN_INVALID_ARGUMENT);
thread = act_lock_thread(thr_act);
if ( thread == THREAD_NULL ||
(sched_attrCnt * sizeof(int)) !=
sched_policy[policy_id].sched_attributes_size ) {
act_unlock_thread(thr_act);
return(KERN_INVALID_ARGUMENT);
}
if (invalid_policy(policy_id)) {
act_unlock_thread(thr_act);
return(KERN_INVALID_POLICY);
}
/* coordinate changes to thread scheduling state with others */
s = splsched();
thread_lock(thread);
/* see if thread's policy is to be changed */
if (thread->policy != policy_id) {
/* it is; see if the target is executing */
if (thread != current_thread()) { /*** fix for SMP ***/
/* it is not; detach from old policy */
policy = &sched_policy[thread->policy];
sfr = policy->sp_ops.sp_thread_detach(policy, thread);
if (sfr != SF_SUCCESS)
panic("thread_set_sched: sp_thread_detach");
/* attach to this policy */
policy = &sched_policy[policy_id];
sfr = policy->sp_ops.sp_thread_attach(policy, thread);
if (sfr != SF_SUCCESS) {
thread_unlock(thread);
splx(s);
act_unlock_thread(thr_act);
return(KERN_FAILURE);
}
/* remember to dispatch thread after setting parms */
do_dispatch = TRUE;
}
else {
/* target is currently executing */
/* defer policy change until thread leaves processor */
thread->pending_policy = policy_id;
bcopy((char *)sched_attr,
(char *)thread->pending_sched_attr,
(sched_attrCnt * sizeof(int)));
/* try to get off processor to effect change */
thread_unlock(thread);
splx(s);
act_unlock_thread(thr_act);
thread_block((void (*)(void)) 0);
/* by now, new attributes have been installed */
act_lock_thread(thr_act);
/*** check to see it's same thread? ***/
s = splsched();
thread_lock(thread);
result = (thread->change_sfr == SF_SUCCESS)?
KERN_SUCCESS : KERN_FAILURE;
thread_unlock(thread);
splx(s);
act_unlock_thread(thr_act);
return(result);
}
}
/* call policy-specific routine to install new scheduling parameters */
policy = &sched_policy[policy_id];
sfr = policy->sp_ops.sp_thread_set(
policy, thread, (sp_attributes_t)sched_attr);
/* dispatch thread under new policy, if appropriate */
if (do_dispatch == TRUE && sfr == SF_SUCCESS) {
sfr = policy->sp_ops.sp_thread_dispatch(policy, thread);
}
thread_unlock(thread);
splx(s);
if (sfr != SF_SUCCESS)
result = KERN_FAILURE;
act_unlock_thread(thr_act);
return(result);
}
/*
* thread_get_sched
*
* Get scheduling policy and parameters for the given thread.
* (This was added as part of the MK Scheduling Framework.)
*/
kern_return_t
thread_get_sched(
thread_act_t thr_act,
policy_t *policy_id_out,
sched_attr_t sched_attr,
mach_msg_type_number_t *sched_attrCnt)
{
thread_t thread;
kern_return_t result = KERN_SUCCESS;
policy_t policy_id;
sched_policy_t *policy;
sf_return_t sfr;
int size;
spl_t s;
if (thr_act == THR_ACT_NULL)
return (KERN_INVALID_ARGUMENT);
thread = act_lock_thread(thr_act);
if (thread == THREAD_NULL) {
act_unlock_thread(thr_act);
*policy_id_out = POLICY_NULL;
return(KERN_INVALID_ARGUMENT);
}
size = *sched_attrCnt * sizeof(int);
/* coordinate changes to thread scheduling state with others */
s = splsched();
thread_lock(thread);
policy_id = thread->policy;
if (size < sched_policy[policy_id].sched_attributes_size) {
thread_unlock(thread);
splx(s);
act_unlock_thread(thr_act);
*policy_id_out = policy_id;
return(KERN_INVALID_ARGUMENT);
}
/* note policy for caller */
*policy_id_out = policy_id;
/* call policy-specific routine */
policy = &sched_policy[policy_id];
sfr = policy->sp_ops.sp_thread_get(
policy, thread, (sp_attributes_t)sched_attr, size);
if (sfr != SF_SUCCESS)
result = KERN_FAILURE;
*sched_attrCnt = policy->sched_attributes_size / sizeof(int);
thread_unlock(thread);
splx(s);
act_unlock_thread(thr_act);
return(result);
}
boolean_t
thread_get_funneled(
void)
{
return((current_thread()->funnel_state & TH_FN_OWNED) == TH_FN_OWNED);
}
boolean_t
thread_set_funneled(
boolean_t funneled)
{
thread_t cur_thread;
boolean_t funnel_state_prev;
cur_thread = current_thread();
funnel_state_prev = ((cur_thread->funnel_state & TH_FN_OWNED) == TH_FN_OWNED);
if (funnel_state_prev != funneled) {
if (funneled == TRUE) {
mutex_lock(&funnel_lock);
cur_thread->funnel_state |= TH_FN_OWNED;
} else {
cur_thread->funnel_state &= ~TH_FN_OWNED;
mutex_unlock(&funnel_lock);
}
}
return(funnel_state_prev);
}
void
thread_set_cont_arg(int arg)
{
thread_t th = current_thread();
th->cont_arg = arg;
}
int
thread_get_cont_arg(void)
{
thread_t th = current_thread();
return(th->cont_arg);
}
/*
* Export routines to other components for things that are done as macros
* within the osfmk component.
*/
#undef thread_should_halt
boolean_t
thread_should_halt(
thread_shuttle_t th)
{
return(thread_should_halt_fast(th));
}