X-Git-Url: https://git.saurik.com/apple/xnu.git/blobdiff_plain/91447636331957f3d9b5ca5b508f07c526b0074d..743345f9a4b36f7e2f9ba37691e70c50baecb56e:/osfmk/kern/thread.h diff --git a/osfmk/kern/thread.h b/osfmk/kern/thread.h index cde616313..368238042 100644 --- a/osfmk/kern/thread.h +++ b/osfmk/kern/thread.h @@ -1,23 +1,29 @@ /* - * Copyright (c) 2000-2005 Apple Computer, Inc. All rights reserved. + * Copyright (c) 2000-2016 Apple Inc. All rights reserved. * - * @APPLE_LICENSE_HEADER_START@ + * @APPLE_OSREFERENCE_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 file contains Original Code and/or Modifications of Original Code + * as defined in and that are subject to the Apple Public Source License + * Version 2.0 (the 'License'). You may not use this file except in + * compliance with the License. The rights granted to you under the License + * may not be used to create, or enable the creation or redistribution of, + * unlawful or unlicensed copies of an Apple operating system, or to + * circumvent, violate, or enable the circumvention or violation of, any + * terms of an Apple operating system software license agreement. * - * This Original Code and all software distributed under the License are - * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER + * Please obtain a copy of the License at + * http://www.opensource.apple.com/apsl/ and read it before using this file. + * + * The 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. + * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT. + * Please see the License for the specific language governing rights and + * limitations under the License. * - * @APPLE_LICENSE_HEADER_END@ + * @APPLE_OSREFERENCE_LICENSE_HEADER_END@ */ /* * @OSF_FREE_COPYRIGHT@ @@ -93,74 +99,94 @@ #ifdef MACH_KERNEL_PRIVATE -#include - #include -#include -#include #include #include #include #include +#include #include #include -#include +#include #include #include #include +#include #include #include #include #include +#include + +#include #include #include #include + struct thread { + +#if MACH_ASSERT +#define THREAD_MAGIC 0x1234ABCDDCBA4321ULL + /* Ensure nothing uses &thread as a queue entry */ + uint64_t thread_magic; +#endif /* MACH_ASSERT */ + /* * NOTE: The runq field in the thread structure has an unusual - * locking protocol. If its value is RUN_QUEUE_NULL, then it is + * locking protocol. If its value is PROCESSOR_NULL, then it is * locked by the thread_lock, but if its value is something else - * (i.e. a run_queue) then it is locked by that run_queue's lock. + * then it is locked by the associated run queue lock. It is + * set to PROCESSOR_NULL without holding the thread lock, but the + * transition from PROCESSOR_NULL to non-null must be done + * under the thread lock and the run queue lock. * - * When the thread is on a wait queue, these first three fields - * are treated as an unofficial union with a wait_queue_element. - * If you change these, you must change that definition as well - * (kern/wait_queue.h). + * New waitq APIs allow the 'links' and 'runq' fields to be + * anywhere in the thread structure. */ - /* Items examined often, modified infrequently */ - queue_chain_t links; /* run/wait queue links */ - run_queue_t runq; /* run queue thread is on SEE BELOW */ - wait_queue_t wait_queue; /* wait queue we are currently on */ - event64_t wait_event; /* wait queue event */ - integer_t options; /* options set by thread itself */ -#define TH_OPT_INTMASK 0x03 /* interrupt / abort level */ -#define TH_OPT_VMPRIV 0x04 /* may allocate reserved memory */ -#define TH_OPT_DELAYIDLE 0x08 /* performing delayed idle */ -#define TH_OPT_CALLOUT 0x10 /* executing as callout */ + union { + queue_chain_t runq_links; /* run queue links */ + queue_chain_t wait_links; /* wait queue links */ + }; + + processor_t runq; /* run queue assignment */ + + event64_t wait_event; /* wait queue event */ + struct waitq *waitq; /* wait queue this thread is enqueued on */ /* Data updated during assert_wait/thread_wakeup */ +#if __SMP__ decl_simple_lock_data(,sched_lock) /* scheduling lock (thread_lock()) */ - decl_simple_lock_data(,wake_lock) /* covers wake_active (wake_lock())*/ - boolean_t wake_active; /* Someone is waiting for this */ + decl_simple_lock_data(,wake_lock) /* for thread stop / wait (wake_lock()) */ +#endif + integer_t options; /* options set by thread itself */ +#define TH_OPT_INTMASK 0x0003 /* interrupt / abort level */ +#define TH_OPT_VMPRIV 0x0004 /* may allocate reserved memory */ +#define TH_OPT_DTRACE 0x0008 /* executing under dtrace_probe */ +#define TH_OPT_SYSTEM_CRITICAL 0x0010 /* Thread must always be allowed to run - even under heavy load */ +#define TH_OPT_PROC_CPULIMIT 0x0020 /* Thread has a task-wide CPU limit applied to it */ +#define TH_OPT_PRVT_CPULIMIT 0x0040 /* Thread has a thread-private CPU limit applied to it */ +#define TH_OPT_IDLE_THREAD 0x0080 /* Thread is a per-processor idle thread */ +#define TH_OPT_GLOBAL_FORCED_IDLE 0x0100 /* Thread performs forced idle for thermal control */ +#define TH_OPT_SCHED_VM_GROUP 0x0200 /* Thread belongs to special scheduler VM group */ +#define TH_OPT_HONOR_QLIMIT 0x0400 /* Thread will honor qlimit while sending mach_msg, regardless of MACH_SEND_ALWAYS */ +#define TH_OPT_SEND_IMPORTANCE 0x0800 /* Thread will allow importance donation from kernel rpc */ + + boolean_t wake_active; /* wake event on stop */ int at_safe_point; /* thread_abort_safely allowed */ ast_t reason; /* why we blocked */ - wait_result_t wait_result; /* outcome of wait - - * may be examined by this thread - * WITHOUT locking */ + uint32_t quantum_remaining; + wait_result_t wait_result; /* outcome of wait - + * may be examined by this thread + * WITHOUT locking */ thread_continue_t continuation; /* continue here next dispatch */ void *parameter; /* continuation parameter */ /* Data updated/used in thread_invoke */ - struct funnel_lock *funnel_lock; /* Non-reentrancy funnel */ - int funnel_state; -#define TH_FN_OWNED 0x1 /* we own the funnel */ -#define TH_FN_REFUNNEL 0x2 /* re-acquire funnel on dispatch */ - vm_offset_t kernel_stack; /* current kernel stack */ vm_offset_t reserved_stack; /* reserved kernel stack */ @@ -173,94 +199,164 @@ struct thread { #define TH_SUSP 0x02 /* stopped or requested to stop */ #define TH_RUN 0x04 /* running or on runq */ #define TH_UNINT 0x08 /* waiting uninteruptibly */ -#define TH_TERMINATE 0x10 /* halted at termination */ - -#define TH_ABORT 0x20 /* abort interruptible waits */ -#define TH_ABORT_SAFELY 0x40 /* ... but only those at safe point */ +#define TH_TERMINATE 0x10 /* halted at termination */ +#define TH_TERMINATE2 0x20 /* added to termination queue */ -#define TH_IDLE 0x80 /* processor idle thread */ - -#define TH_SCHED_STATE (TH_WAIT|TH_SUSP|TH_RUN|TH_UNINT) +#define TH_IDLE 0x80 /* idling processor */ /* Scheduling information */ - integer_t sched_mode; /* scheduling mode bits */ -#define TH_MODE_REALTIME 0x0001 /* time constraints supplied */ -#define TH_MODE_TIMESHARE 0x0002 /* use timesharing algorithm */ -#define TH_MODE_PREEMPT 0x0004 /* can preempt kernel contexts */ -#define TH_MODE_FAILSAFE 0x0008 /* fail-safe has tripped */ -#define TH_MODE_PROMOTED 0x0010 /* sched pri has been promoted */ -#define TH_MODE_DEPRESS 0x0020 /* normal depress yield */ -#define TH_MODE_POLLDEPRESS 0x0040 /* polled depress yield */ -#define TH_MODE_ISDEPRESSED (TH_MODE_DEPRESS | TH_MODE_POLLDEPRESS) - - integer_t sched_pri; /* scheduled (current) priority */ - integer_t priority; /* base priority */ - integer_t max_priority; /* max base priority */ - integer_t task_priority; /* copy of task base priority */ - - integer_t promotions; /* level of promotion */ - integer_t pending_promoter_index; + sched_mode_t sched_mode; /* scheduling mode */ + sched_mode_t saved_mode; /* saved mode during forced mode demotion */ + + /* This thread's contribution to global sched counters */ + sched_bucket_t th_sched_bucket; + + sfi_class_id_t sfi_class; /* SFI class (XXX Updated on CSW/QE/AST) */ + sfi_class_id_t sfi_wait_class; /* Currently in SFI wait for this class, protected by sfi_lock */ + + + uint32_t sched_flags; /* current flag bits */ +/* TH_SFLAG_FAIRSHARE_TRIPPED (unused) 0x0001 */ +#define TH_SFLAG_FAILSAFE 0x0002 /* fail-safe has tripped */ +#define TH_SFLAG_THROTTLED 0x0004 /* throttled thread forced to timeshare mode (may be applied in addition to failsafe) */ +#define TH_SFLAG_DEMOTED_MASK (TH_SFLAG_THROTTLED | TH_SFLAG_FAILSAFE) /* saved_mode contains previous sched_mode */ + +#define TH_SFLAG_PROMOTED 0x0008 /* sched pri has been promoted */ +#define TH_SFLAG_ABORT 0x0010 /* abort interruptible waits */ +#define TH_SFLAG_ABORTSAFELY 0x0020 /* ... but only those at safe point */ +#define TH_SFLAG_ABORTED_MASK (TH_SFLAG_ABORT | TH_SFLAG_ABORTSAFELY) +#define TH_SFLAG_DEPRESS 0x0040 /* normal depress yield */ +#define TH_SFLAG_POLLDEPRESS 0x0080 /* polled depress yield */ +#define TH_SFLAG_DEPRESSED_MASK (TH_SFLAG_DEPRESS | TH_SFLAG_POLLDEPRESS) +/* unused TH_SFLAG_PRI_UPDATE 0x0100 */ +#define TH_SFLAG_EAGERPREEMPT 0x0200 /* Any preemption of this thread should be treated as if AST_URGENT applied */ +#define TH_SFLAG_RW_PROMOTED 0x0400 /* sched pri has been promoted due to blocking with RW lock held */ +/* unused TH_SFLAG_THROTTLE_DEMOTED 0x0800 */ +#define TH_SFLAG_WAITQ_PROMOTED 0x1000 /* sched pri promoted from waitq wakeup (generally for IPC receive) */ +#define TH_SFLAG_PROMOTED_MASK (TH_SFLAG_PROMOTED | TH_SFLAG_RW_PROMOTED | TH_SFLAG_WAITQ_PROMOTED) + +#define TH_SFLAG_RW_PROMOTED_BIT (10) /* 0x400 */ + + int16_t sched_pri; /* scheduled (current) priority */ + int16_t base_pri; /* base priority */ + int16_t max_priority; /* copy of max base priority */ + int16_t task_priority; /* copy of task base priority */ + +#if defined(CONFIG_SCHED_GRRR) +#if 0 + uint16_t grrr_deficit; /* fixed point (1/1000th quantum) fractional deficit */ +#endif +#endif + + int16_t promotions; /* level of promotion */ + int16_t pending_promoter_index; + uint32_t ref_count; /* number of references to me */ void *pending_promoter[2]; + uint32_t rwlock_count; /* Number of lck_rw_t locks held by thread */ + integer_t importance; /* task-relative importance */ + uint32_t was_promoted_on_wakeup; - /* real-time parameters */ + /* Priority depression expiration */ + integer_t depress_timer_active; + timer_call_data_t depress_timer; + /* real-time parameters */ struct { /* see mach/thread_policy.h */ uint32_t period; uint32_t computation; uint32_t constraint; boolean_t preemptible; - uint64_t deadline; } realtime; - uint32_t current_quantum; /* duration of current quantum */ + uint64_t last_run_time; /* time when thread was switched away from */ + uint64_t last_made_runnable_time; /* time when thread was unblocked or preempted */ + +#if defined(CONFIG_SCHED_MULTIQ) + sched_group_t sched_group; +#endif /* defined(CONFIG_SCHED_MULTIQ) */ /* Data used during setrun/dispatch */ timer_data_t system_timer; /* system mode timer */ - processor_set_t processor_set; /* assigned processor set */ processor_t bound_processor; /* bound to a processor? */ processor_t last_processor; /* processor last dispatched on */ - uint64_t last_switch; /* time of last context switch */ + processor_t chosen_processor; /* Where we want to run this thread */ /* Fail-safe computation since last unblock or qualifying yield */ uint64_t computation_metered; uint64_t computation_epoch; - integer_t safe_mode; /* saved mode during fail-safe */ - natural_t safe_release; /* when to release fail-safe */ - + uint64_t safe_release; /* when to release fail-safe */ + + /* Call out from scheduler */ + void (*sched_call)( + int type, + thread_t thread); +#if defined(CONFIG_SCHED_PROTO) + uint32_t runqueue_generation; /* last time runqueue was drained */ +#endif + /* Statistics and timesharing calculations */ +#if defined(CONFIG_SCHED_TIMESHARE_CORE) natural_t sched_stamp; /* last scheduler tick */ natural_t sched_usage; /* timesharing cpu usage [sched] */ natural_t pri_shift; /* usage -> priority from pset */ natural_t cpu_usage; /* instrumented cpu usage [%cpu] */ natural_t cpu_delta; /* accumulated cpu_usage delta */ +#endif /* CONFIG_SCHED_TIMESHARE_CORE */ + + uint32_t c_switch; /* total context switches */ + uint32_t p_switch; /* total processor switches */ + uint32_t ps_switch; /* total pset switches */ + integer_t mutex_count; /* total count of locks held */ /* Timing data structures */ + int precise_user_kernel_time; /* precise user/kernel enabled for this thread */ timer_data_t user_timer; /* user mode timer */ - uint64_t system_timer_save; /* saved system timer value */ uint64_t user_timer_save; /* saved user timer value */ + uint64_t system_timer_save; /* saved system timer value */ + uint64_t vtimer_user_save; /* saved values for vtimers */ + uint64_t vtimer_prof_save; + uint64_t vtimer_rlim_save; + uint64_t vtimer_qos_save; + +#if CONFIG_SCHED_SFI + /* Timing for wait state */ + uint64_t wait_sfi_begin_time; /* start time for thread waiting in SFI */ +#endif /* Timed wait expiration */ timer_call_data_t wait_timer; integer_t wait_timer_active; boolean_t wait_timer_is_set; - /* Priority depression expiration */ - timer_call_data_t depress_timer; - integer_t depress_timer_active; + + /* + * Processor/cache affinity + * - affinity_threads links task threads with the same affinity set + */ + affinity_set_t affinity_set; + queue_chain_t affinity_threads; /* Various bits of stashed state */ union { struct { mach_msg_return_t state; /* receive state */ + mach_port_seqno_t seqno; /* seqno of recvd message */ ipc_object_t object; /* object received on */ mach_vm_address_t msg_addr; /* receive buffer pointer */ - mach_msg_size_t msize; /* max size for recvd msg */ + mach_msg_size_t rsize; /* max size for recvd msg */ + mach_msg_size_t msize; /* actual size for recvd msg */ mach_msg_option_t option; /* options for receive */ - mach_msg_size_t slist_size; /* scatter list size */ - struct ipc_kmsg *kmsg; /* received message */ - mach_port_seqno_t seqno; /* seqno of recvd message */ + mach_port_name_t receiver_name; /* the receive port name */ + union { + struct ipc_kmsg *kmsg; /* received message */ + struct ipc_mqueue *peekq; /* mqueue to peek at */ + struct { + mach_msg_priority_t qos; /* received message qos */ + mach_msg_priority_t oqos; /* override qos for message */ + } received_qos; + }; mach_msg_continue_t continuation; } receive; struct { @@ -272,98 +368,200 @@ struct thread { } sema; struct { int option; /* switch option */ + boolean_t reenable_workq_callback; /* on entry, callbacks were suspended */ } swtch; int misc; /* catch-all for other state */ } saved; + /* Structure to save information about guard exception */ + struct { + unsigned type; /* EXC_GUARD reason/type */ + mach_exception_data_type_t code; /* Exception code */ + mach_exception_data_type_t subcode; /* Exception sub-code */ + } guard_exc_info; + + /* Kernel holds on this thread */ + int16_t suspend_count; + /* User level suspensions */ + int16_t user_stop_count; + /* IPC data structures */ - struct ipc_kmsg_queue ith_messages; +#if IMPORTANCE_INHERITANCE + natural_t ith_assertions; /* assertions pending drop */ +#endif + struct ipc_kmsg_queue ith_messages; /* messages to reap */ mach_port_t ith_rpc_reply; /* reply port for kernel RPCs */ /* Ast/Halt data structures */ - vm_offset_t recover; /* page fault recover(copyin/out) */ - int ref_count; /* number of references to me */ + vm_offset_t recover; /* page fault recover(copyin/out) */ - /* Processor set info */ - queue_chain_t pset_threads; /* list of all threads in pset */ -#if MACH_HOST - boolean_t may_assign; /* may assignment change? */ - boolean_t assign_active; /* waiting for may_assign */ -#endif /* MACH_HOST */ + queue_chain_t threads; /* global list of all threads */ /* Activation */ queue_chain_t task_threads; - /*** Machine-dependent state ***/ - struct machine_thread machine; - /* Task membership */ struct task *task; vm_map_t map; - decl_mutex_data(,mutex) - - /* Kernel holds on this thread */ - int suspend_count; + decl_lck_mtx_data(,mutex) - /* User level suspensions */ - int user_stop_count; /* Pending thread ast(s) */ ast_t ast; /* Miscellaneous bits guarded by mutex */ uint32_t - /* Indicates that the thread has not been terminated */ - active:1, - - /* Indicates that the thread has been started after creation */ - started:1, - :0; - - /* Return Handers */ - struct ReturnHandler { - struct ReturnHandler *next; - void (*handler)( - struct ReturnHandler *rh, - struct thread *thread); - } *handlers, special_handler; + active:1, /* Thread is active and has not been terminated */ + started:1, /* Thread has been started after creation */ + static_param:1, /* Disallow policy parameter changes */ + inspection:1, /* TRUE when task is being inspected by crash reporter */ + policy_reset:1, /* Disallow policy parameter changes on terminating threads */ + suspend_parked:1, /* thread parked in thread_suspended */ + corpse_dup:1, /* TRUE when thread is an inactive duplicate in a corpse */ + :0; /* Ports associated with this thread */ struct ipc_port *ith_self; /* not a right, doesn't hold ref */ struct ipc_port *ith_sself; /* a send right */ - struct exception_action exc_actions[EXC_TYPES_COUNT]; - - /* Owned ulocks (a lock set element) */ - queue_head_t held_ulocks; - -#if MACH_PROF - /* Profiling */ - boolean_t profiled; - boolean_t profiled_own; - struct prof_data *profil_buffer; -#endif /* MACH_PROF */ + struct exception_action *exc_actions; #ifdef MACH_BSD void *uthread; #endif + +#if CONFIG_DTRACE + uint32_t t_dtrace_flags; /* DTrace thread states */ +#define TH_DTRACE_EXECSUCCESS 0x01 + uint32_t t_dtrace_predcache;/* DTrace per thread predicate value hint */ + int64_t t_dtrace_tracing; /* Thread time under dtrace_probe() */ + int64_t t_dtrace_vtime; +#endif + + clock_sec_t t_page_creation_time; + uint32_t t_page_creation_count; + uint32_t t_page_creation_throttled; +#if (DEVELOPMENT || DEBUG) + uint64_t t_page_creation_throttled_hard; + uint64_t t_page_creation_throttled_soft; +#endif /* DEVELOPMENT || DEBUG */ + +#ifdef KPERF +/* The high 7 bits are the number of frames to sample of a user callstack. */ +#define T_KPERF_CALLSTACK_DEPTH_OFFSET (25) +#define T_KPERF_SET_CALLSTACK_DEPTH(DEPTH) (((uint32_t)(DEPTH)) << T_KPERF_CALLSTACK_DEPTH_OFFSET) +#define T_KPERF_GET_CALLSTACK_DEPTH(FLAGS) ((FLAGS) >> T_KPERF_CALLSTACK_DEPTH_OFFSET) +#endif + +#define T_KPERF_AST_CALLSTACK (1U << 0) /* dump a callstack on thread's next AST */ +#define T_KPERF_AST_DISPATCH (1U << 1) /* dump a name on thread's next AST */ +#define T_KPC_ALLOC (1U << 2) /* thread needs a kpc_buf allocated */ +/* only go up to T_KPERF_CALLSTACK_DEPTH_OFFSET - 1 */ + +#ifdef KPERF + uint32_t kperf_flags; + uint32_t kperf_pet_gen; /* last generation of PET that sampled this thread*/ + uint32_t kperf_c_switch; /* last dispatch detection */ + uint32_t kperf_pet_cnt; /* how many times a thread has been sampled by PET */ +#endif + +#ifdef KPC + /* accumulated performance counters for this thread */ + uint64_t *kpc_buf; +#endif + +#if HYPERVISOR + /* hypervisor virtual CPU object associated with this thread */ + void *hv_thread_target; +#endif /* HYPERVISOR */ + + uint64_t thread_id; /*system wide unique thread-id*/ + + /* Statistics accumulated per-thread and aggregated per-task */ + uint32_t syscalls_unix; + uint32_t syscalls_mach; + ledger_t t_ledger; + ledger_t t_threadledger; /* per thread ledger */ +#ifdef CONFIG_BANK + ledger_t t_bankledger; /* ledger to charge someone */ + uint64_t t_deduct_bank_ledger_time; /* cpu time to be deducted from bank ledger */ +#endif + + /* policy is protected by the thread mutex */ + struct thread_requested_policy requested_policy; + struct thread_effective_policy effective_policy; + + /* usynch override is protected by the task lock, eventually will be thread mutex */ + struct thread_qos_override { + struct thread_qos_override *override_next; + uint32_t override_contended_resource_count; + int16_t override_qos; + int16_t override_resource_type; + user_addr_t override_resource; + } *overrides; + + uint32_t ipc_overrides; + uint32_t user_promotions; + uint16_t user_promotion_basepri; + + int iotier_override; /* atomic operations to set, cleared on ret to user */ + io_stat_info_t thread_io_stats; /* per-thread I/O statistics */ + + + uint32_t thread_callout_interrupt_wakeups; + uint32_t thread_callout_platform_idle_wakeups; + uint32_t thread_timer_wakeups_bin_1; + uint32_t thread_timer_wakeups_bin_2; + uint16_t thread_tag; + uint16_t callout_woken_from_icontext:1, + callout_woken_from_platform_idle:1, + callout_woke_thread:1, + thread_bitfield_unused:13; + + mach_port_name_t ith_voucher_name; + ipc_voucher_t ith_voucher; +#if CONFIG_IOSCHED + void *decmp_upl; +#endif /* CONFIG_IOSCHED */ + + /* work interval ID (if any) associated with the thread. Uses thread mutex */ + uint64_t work_interval_id; + + /*** Machine-dependent state ***/ + struct machine_thread machine; + +#if SCHED_TRACE_THREAD_WAKEUPS + uintptr_t thread_wakeup_bt[64]; +#endif }; -#define ith_state saved.receive.state -#define ith_object saved.receive.object -#define ith_msg_addr saved.receive.msg_addr -#define ith_msize saved.receive.msize -#define ith_option saved.receive.option -#define ith_scatter_list_size saved.receive.slist_size -#define ith_continuation saved.receive.continuation -#define ith_kmsg saved.receive.kmsg -#define ith_seqno saved.receive.seqno - -#define sth_waitsemaphore saved.sema.waitsemaphore -#define sth_signalsemaphore saved.sema.signalsemaphore -#define sth_options saved.sema.options -#define sth_result saved.sema.result -#define sth_continuation saved.sema.continuation +#define ith_state saved.receive.state +#define ith_object saved.receive.object +#define ith_msg_addr saved.receive.msg_addr +#define ith_rsize saved.receive.rsize +#define ith_msize saved.receive.msize +#define ith_option saved.receive.option +#define ith_receiver_name saved.receive.receiver_name +#define ith_continuation saved.receive.continuation +#define ith_kmsg saved.receive.kmsg +#define ith_peekq saved.receive.peekq +#define ith_qos saved.receive.received_qos.qos +#define ith_qos_override saved.receive.received_qos.oqos +#define ith_seqno saved.receive.seqno + +#define sth_waitsemaphore saved.sema.waitsemaphore +#define sth_signalsemaphore saved.sema.signalsemaphore +#define sth_options saved.sema.options +#define sth_result saved.sema.result +#define sth_continuation saved.sema.continuation + +#if MACH_ASSERT +#define assert_thread_magic(thread) assertf((thread)->thread_magic == THREAD_MAGIC, \ + "bad thread magic 0x%llx for thread %p, expected 0x%llx", \ + (thread)->thread_magic, (thread), THREAD_MAGIC) +#else +#define assert_thread_magic(thread) do { (void)(thread); } while (0) +#endif extern void thread_bootstrap(void); @@ -372,28 +570,46 @@ extern void thread_init(void); extern void thread_daemon_init(void); #define thread_reference_internal(thread) \ - hw_atomic_add(&(thread)->ref_count, 1) - -#define thread_deallocate_internal(thread) \ - hw_atomic_sub(&(thread)->ref_count, 1) + (void)hw_atomic_add(&(thread)->ref_count, 1) #define thread_reference(thread) \ MACRO_BEGIN \ if ((thread) != THREAD_NULL) \ - thread_reference_internal(thread); \ + thread_reference_internal(thread); \ MACRO_END extern void thread_deallocate( thread_t thread); +extern void thread_deallocate_safe( + thread_t thread); + extern void thread_terminate_self(void); extern kern_return_t thread_terminate_internal( thread_t thread); +extern void thread_start( + thread_t thread) __attribute__ ((noinline)); + +extern void thread_start_in_assert_wait( + thread_t thread, + event_t event, + wait_interrupt_t interruptible) __attribute__ ((noinline)); + extern void thread_terminate_enqueue( thread_t thread); +extern void thread_exception_enqueue( + task_t task, + thread_t thread); + +extern void thread_copy_resource_info( + thread_t dst_thread, + thread_t src_thread); + +extern void thread_terminate_crashed_threads(void); + extern void thread_stack_enqueue( thread_t thread); @@ -403,26 +619,43 @@ extern void thread_hold( extern void thread_release( thread_t thread); +extern void thread_corpse_continue(void); + +extern boolean_t thread_is_active(thread_t thread); + +/* Locking for scheduler state, always acquired with interrupts disabled (splsched()) */ +#if __SMP__ #define thread_lock_init(th) simple_lock_init(&(th)->sched_lock, 0) #define thread_lock(th) simple_lock(&(th)->sched_lock) #define thread_unlock(th) simple_unlock(&(th)->sched_lock) -#define thread_lock_try(th) simple_lock_try(&(th)->sched_lock) - -#define thread_should_halt_fast(thread) (!(thread)->active) #define wake_lock_init(th) simple_lock_init(&(th)->wake_lock, 0) #define wake_lock(th) simple_lock(&(th)->wake_lock) #define wake_unlock(th) simple_unlock(&(th)->wake_lock) -#define wake_lock_try(th) simple_lock_try(&(th)->wake_lock) +#else +#define thread_lock_init(th) do { (void)th; } while(0) +#define thread_lock(th) do { (void)th; } while(0) +#define thread_unlock(th) do { (void)th; } while(0) + +#define wake_lock_init(th) do { (void)th; } while(0) +#define wake_lock(th) do { (void)th; } while(0) +#define wake_unlock(th) do { (void)th; } while(0) +#endif + +#define thread_should_halt_fast(thread) (!(thread)->active) extern void stack_alloc( thread_t thread); +extern void stack_handoff( + thread_t from, + thread_t to); + extern void stack_free( thread_t thread); -extern void stack_free_stack( - vm_offset_t stack); +extern void stack_free_reserved( + thread_t thread); extern boolean_t stack_alloc_try( thread_t thread); @@ -431,20 +664,6 @@ extern void stack_collect(void); extern void stack_init(void); -extern kern_return_t thread_state_initialize( - thread_t thread); - -extern kern_return_t thread_setstatus( - thread_t thread, - int flavor, - thread_state_t tstate, - mach_msg_type_number_t count); - -extern kern_return_t thread_getstatus( - thread_t thread, - int flavor, - thread_state_t tstate, - mach_msg_type_number_t *count); extern kern_return_t thread_info_internal( thread_t thread, @@ -452,13 +671,7 @@ extern kern_return_t thread_info_internal( thread_info_t thread_info_out, mach_msg_type_number_t *thread_info_count); -extern void thread_task_priority( - thread_t thread, - integer_t priority, - integer_t max_priority); -extern void thread_policy_reset( - thread_t thread); extern kern_return_t kernel_thread_create( thread_continue_t continuation, @@ -515,6 +728,8 @@ extern void machine_thread_init(void); extern kern_return_t machine_thread_create( thread_t thread, task_t task); +extern void machine_thread_switch_addrmode( + thread_t thread); extern void machine_thread_destroy( thread_t thread); @@ -522,41 +737,43 @@ extern void machine_thread_destroy( extern void machine_set_current_thread( thread_t thread); -extern void machine_thread_terminate_self(void); - extern kern_return_t machine_thread_get_kern_state( thread_t thread, thread_flavor_t flavor, thread_state_t tstate, mach_msg_type_number_t *count); +extern kern_return_t machine_thread_inherit_taskwide( + thread_t thread, + task_t parent_task); -/* - * XXX Funnel locks XXX - */ +extern kern_return_t machine_thread_set_tsd_base( + thread_t thread, + mach_vm_offset_t tsd_base); -struct funnel_lock { - int fnl_type; /* funnel type */ - lck_mtx_t *fnl_mutex; /* underlying mutex for the funnel */ - void * fnl_mtxholder; /* thread (last)holdng mutex */ - void * fnl_mtxrelease; /* thread (last)releasing mutex */ - lck_mtx_t *fnl_oldmutex; /* Mutex before collapsing split funnel */ -}; +#define thread_mtx_lock(thread) lck_mtx_lock(&(thread)->mutex) +#define thread_mtx_try(thread) lck_mtx_try_lock(&(thread)->mutex) +#define thread_mtx_unlock(thread) lck_mtx_unlock(&(thread)->mutex) -typedef struct ReturnHandler ReturnHandler; +extern void thread_apc_ast(thread_t thread); -#define thread_mtx_lock(thread) mutex_lock(&(thread)->mutex) -#define thread_mtx_try(thread) mutex_try(&(thread)->mutex) -#define thread_mtx_unlock(thread) mutex_unlock(&(thread)->mutex) +extern void thread_update_qos_cpu_time(thread_t thread); -extern void act_execute_returnhandlers(void); +void act_machine_sv_free(thread_t, int); -extern void install_special_handler( - thread_t thread); +vm_offset_t min_valid_stack_address(void); +vm_offset_t max_valid_stack_address(void); + +static inline uint16_t thread_set_tag_internal(thread_t thread, uint16_t tag) { + return __sync_fetch_and_or(&thread->thread_tag, tag); +} + +static inline uint16_t thread_get_tag_internal(thread_t thread) { + return thread->thread_tag; +} -extern void special_handler( - ReturnHandler *rh, - thread_t thread); + +extern void thread_set_options(uint32_t thopt); #else /* MACH_KERNEL_PRIVATE */ @@ -576,58 +793,109 @@ __END_DECLS #ifdef KERNEL_PRIVATE -typedef struct funnel_lock funnel_t; +__BEGIN_DECLS -#ifdef MACH_KERNEL_PRIVATE +extern uint64_t thread_dispatchqaddr( + thread_t thread); -extern void funnel_lock( - funnel_t *lock); +__END_DECLS -extern void funnel_unlock( - funnel_t *lock); +#endif /* KERNEL_PRIVATE */ -vm_offset_t min_valid_stack_address(void); -vm_offset_t max_valid_stack_address(void); +#ifdef KERNEL +__BEGIN_DECLS -#endif /* MACH_KERNEL_PRIVATE */ +extern uint64_t thread_tid(thread_t thread); + +__END_DECLS + +#endif /* KERNEL */ __BEGIN_DECLS -extern funnel_t *thread_funnel_get(void); +#ifdef XNU_KERNEL_PRIVATE -extern boolean_t thread_funnel_set( - funnel_t *lock, - boolean_t funneled); +/* + * Thread tags; for easy identification. + */ +#define THREAD_TAG_MAINTHREAD 0x1 +#define THREAD_TAG_CALLOUT 0x2 +#define THREAD_TAG_IOWORKLOOP 0x4 -extern thread_t kernel_thread( - task_t task, - void (*start)(void)); +#define THREAD_TAG_PTHREAD 0x10 +#define THREAD_TAG_WORKQUEUE 0x20 -__END_DECLS +uint16_t thread_set_tag(thread_t, uint16_t); +uint16_t thread_get_tag(thread_t); -#endif /* KERNEL_PRIVATE */ +extern kern_return_t thread_state_initialize( + thread_t thread); -__BEGIN_DECLS +extern kern_return_t thread_setstatus( + thread_t thread, + int flavor, + thread_state_t tstate, + mach_msg_type_number_t count); -#ifdef XNU_KERNEL_PRIVATE +extern kern_return_t thread_getstatus( + thread_t thread, + int flavor, + thread_state_t tstate, + mach_msg_type_number_t *count); + +extern kern_return_t thread_create_with_continuation( + task_t task, + thread_t *new_thread, + thread_continue_t continuation); + +extern kern_return_t thread_create_waiting(task_t task, + thread_continue_t continuation, + event_t event, + thread_t *new_thread); + +extern kern_return_t thread_create_workq( + task_t task, + thread_continue_t thread_return, + thread_t *new_thread); + +extern kern_return_t thread_create_workq_waiting( + task_t task, + thread_continue_t thread_return, + event_t event, + thread_t *new_thread); + +extern void thread_yield_internal( + mach_msg_timeout_t interval); + +extern void thread_yield_to_preemption(void); /* - * XXX Funnel locks XXX + * Thread-private CPU limits: apply a private CPU limit to this thread only. Available actions are: + * + * 1) Block. Prevent CPU consumption of the thread from exceeding the limit. + * 2) Exception. Generate a resource consumption exception when the limit is exceeded. + * 3) Disable. Remove any existing CPU limit. */ +#define THREAD_CPULIMIT_BLOCK 0x1 +#define THREAD_CPULIMIT_EXCEPTION 0x2 +#define THREAD_CPULIMIT_DISABLE 0x3 -#define THR_FUNNEL_NULL (funnel_t *)0 +struct _thread_ledger_indices { + int cpu_time; +}; -extern funnel_t *funnel_alloc( - int type); +extern struct _thread_ledger_indices thread_ledgers; -extern void funnel_free( - funnel_t *lock); +extern int thread_get_cpulimit(int *action, uint8_t *percentage, uint64_t *interval_ns); +extern int thread_set_cpulimit(int action, uint8_t percentage, uint64_t interval_ns); extern void thread_read_times( thread_t thread, time_value_t *user_time, time_value_t *system_time); +extern uint64_t thread_get_runtime_self(void); + extern void thread_setuserstack( thread_t thread, mach_vm_offset_t user_stack); @@ -640,34 +908,207 @@ extern void thread_setentrypoint( thread_t thread, mach_vm_offset_t entry); +extern kern_return_t thread_set_tsd_base( + thread_t thread, + mach_vm_offset_t tsd_base); + +extern kern_return_t thread_setsinglestep( + thread_t thread, + int on); + +extern kern_return_t thread_userstack( + thread_t, + int, + thread_state_t, + unsigned int, + mach_vm_offset_t *, + int *, + boolean_t); + +extern kern_return_t thread_entrypoint( + thread_t, + int, + thread_state_t, + unsigned int, + mach_vm_offset_t *); + +extern kern_return_t thread_userstackdefault( + mach_vm_offset_t *, + boolean_t); + extern kern_return_t thread_wire_internal( host_priv_t host_priv, thread_t thread, boolean_t wired, boolean_t *prev_state); -/* JMM - These are only temporary */ -extern boolean_t is_thread_running(thread_t); /* True is TH_RUN */ -extern boolean_t is_thread_idle(thread_t); /* True is TH_IDLE */ extern kern_return_t thread_dup(thread_t); +extern kern_return_t thread_dup2(thread_t, thread_t); + +typedef void (*sched_call_t)( + int type, + thread_t thread); + +#define SCHED_CALL_BLOCK 0x1 +#define SCHED_CALL_UNBLOCK 0x2 + +extern void thread_sched_call( + thread_t thread, + sched_call_t call); + +extern sched_call_t thread_disable_sched_call( + thread_t thread, + sched_call_t call); + +extern void thread_reenable_sched_call( + thread_t thread, + sched_call_t call); + +extern void thread_static_param( + thread_t thread, + boolean_t state); + +extern boolean_t thread_is_static_param( + thread_t thread); + extern task_t get_threadtask(thread_t); +#define thread_is_64bit(thd) \ + task_has_64BitAddr(get_threadtask(thd)) + extern void *get_bsdthread_info(thread_t); extern void set_bsdthread_info(thread_t, void *); -extern void *uthread_alloc(task_t, thread_t); -extern void uthread_free(task_t, void *, void *); +extern void *uthread_alloc(task_t, thread_t, int); +extern void uthread_cleanup_name(void *uthread); +extern void uthread_cleanup(task_t, void *, void *); +extern void uthread_zone_free(void *); +extern void uthread_cred_free(void *); + +#if PROC_REF_DEBUG +extern int uthread_get_proc_refcount(void *); +extern void uthread_reset_proc_refcount(void *); +extern int proc_ref_tracking_disabled; +#endif extern boolean_t thread_should_halt( thread_t thread); +extern boolean_t thread_should_abort( + thread_t); + +extern int is_64signalregset(void); + +extern void act_set_kperf(thread_t); +extern void set_astledger(thread_t thread); +extern void act_set_io_telemetry_ast(thread_t); + +extern uint32_t dtrace_get_thread_predcache(thread_t); +extern int64_t dtrace_get_thread_vtime(thread_t); +extern int64_t dtrace_get_thread_tracing(thread_t); +extern boolean_t dtrace_get_thread_reentering(thread_t); +extern int dtrace_get_thread_last_cpu_id(thread_t); +extern vm_offset_t dtrace_get_kernel_stack(thread_t); +extern void dtrace_set_thread_predcache(thread_t, uint32_t); +extern void dtrace_set_thread_vtime(thread_t, int64_t); +extern void dtrace_set_thread_tracing(thread_t, int64_t); +extern void dtrace_set_thread_reentering(thread_t, boolean_t); +extern vm_offset_t dtrace_set_thread_recover(thread_t, vm_offset_t); +extern void dtrace_thread_bootstrap(void); +extern void dtrace_thread_didexec(thread_t); + +extern int64_t dtrace_calc_thread_recent_vtime(thread_t); + + +extern kern_return_t thread_set_wq_state32( + thread_t thread, + thread_state_t tstate); + +extern kern_return_t thread_set_wq_state64( + thread_t thread, + thread_state_t tstate); + +extern vm_offset_t kernel_stack_mask; +extern vm_offset_t kernel_stack_size; +extern vm_offset_t kernel_stack_depth_max; + +void guard_ast(thread_t thread); +extern void fd_guard_ast(thread_t thread); +extern void mach_port_guard_ast(thread_t thread); +extern void thread_guard_violation(thread_t thread, unsigned type); +extern void thread_update_io_stats(thread_t thread, int size, int io_flags); + +extern kern_return_t thread_set_voucher_name(mach_port_name_t name); +extern kern_return_t thread_get_current_voucher_origin_pid(int32_t *pid); + +extern void set_thread_rwlock_boost(void); +extern void clear_thread_rwlock_boost(void); + +/*! @function thread_has_thread_name + @abstract Checks if a thread has a name. + @discussion This function takes one input, a thread, and returns a boolean value indicating if that thread already has a name associated with it. + @param th The thread to inspect. + @result TRUE if the thread has a name, FALSE otherwise. +*/ +extern boolean_t thread_has_thread_name(thread_t th); + +/*! @function thread_set_thread_name + @abstract Set a thread's name. + @discussion This function takes two input parameters: a thread to name, and the name to apply to the thread. The name will be attached to the thread in order to better identify the thread. + @param th The thread to be named. + @param name The name to apply to the thread. +*/ +extern void thread_set_thread_name(thread_t th, const char* name); + +extern void thread_enable_send_importance(thread_t thread, boolean_t enable); + +/* Get a backtrace for a threads kernel or user stack (user_p), with pc and optionally + * frame pointer (getfp). Returns bytes added to buffer, and kThreadTruncatedBT in + * thread_trace_flags if a user page is not present after kdp_lightweight_fault() is + * called. + */ + +extern int machine_trace_thread( + thread_t thread, + char *tracepos, + char *tracebound, + int nframes, + boolean_t user_p, + boolean_t getfp, + uint32_t *thread_trace_flags); + +extern int machine_trace_thread64(thread_t thread, + char *tracepos, + char *tracebound, + int nframes, + boolean_t user_p, + boolean_t getfp, + uint32_t *thread_trace_flags); + #endif /* XNU_KERNEL_PRIVATE */ + +/*! @function kernel_thread_start + @abstract Create a kernel thread. + @discussion This function takes three input parameters, namely reference to the function that the thread should execute, caller specified data and a reference which is used to return the newly created kernel thread. The function returns KERN_SUCCESS on success or an appropriate kernel code type indicating the error. It may be noted that the caller is responsible for explicitly releasing the reference to the created thread when no longer needed. This should be done by calling thread_deallocate(new_thread). + @param continuation A C-function pointer where the thread will begin execution. + @param parameter Caller specified data to be passed to the new thread. + @param new_thread Reference to the new thread is returned in this parameter. + @result Returns KERN_SUCCESS on success or an appropriate kernel code type. +*/ + extern kern_return_t kernel_thread_start( thread_continue_t continuation, void *parameter, thread_t *new_thread); +#ifdef KERNEL_PRIVATE +void thread_set_eager_preempt(thread_t thread); +void thread_clear_eager_preempt(thread_t thread); +extern ipc_port_t convert_thread_to_port(thread_t); +extern boolean_t is_vm_privileged(void); +extern boolean_t set_vm_privilege(boolean_t); +#endif /* KERNEL_PRIVATE */ __END_DECLS