1 // Copyright 2009 The Go Authors. All rights reserved.
2 // Use of this source code is governed by a BSD-style
3 // license that can be found in the LICENSE file.
9 "runtime/internal/atomic"
17 // Beyond indicating the general state of a G, the G status
18 // acts like a lock on the goroutine's stack (and hence its
19 // ability to execute user code).
21 // If you add to this list, add to the list
22 // of "okay during garbage collection" status
25 // TODO(austin): The _Gscan bit could be much lighter-weight.
26 // For example, we could choose not to run _Gscanrunnable
27 // goroutines found in the run queue, rather than CAS-looping
28 // until they become _Grunnable. And transitions like
29 // _Gscanwaiting -> _Gscanrunnable are actually okay because
30 // they don't affect stack ownership.
32 // _Gidle means this goroutine was just allocated and has not
33 // yet been initialized.
36 // _Grunnable means this goroutine is on a run queue. It is
37 // not currently executing user code. The stack is not owned.
40 // _Grunning means this goroutine may execute user code. The
41 // stack is owned by this goroutine. It is not on a run queue.
42 // It is assigned an M and a P (g.m and g.m.p are valid).
45 // _Gsyscall means this goroutine is executing a system call.
46 // It is not executing user code. The stack is owned by this
47 // goroutine. It is not on a run queue. It is assigned an M.
50 // _Gwaiting means this goroutine is blocked in the runtime.
51 // It is not executing user code. It is not on a run queue,
52 // but should be recorded somewhere (e.g., a channel wait
53 // queue) so it can be ready()d when necessary. The stack is
54 // not owned *except* that a channel operation may read or
55 // write parts of the stack under the appropriate channel
56 // lock. Otherwise, it is not safe to access the stack after a
57 // goroutine enters _Gwaiting (e.g., it may get moved).
60 // _Gmoribund_unused is currently unused, but hardcoded in gdb
62 _Gmoribund_unused // 5
64 // _Gdead means this goroutine is currently unused. It may be
65 // just exited, on a free list, or just being initialized. It
66 // is not executing user code. It may or may not have a stack
67 // allocated. The G and its stack (if any) are owned by the M
68 // that is exiting the G or that obtained the G from the free
72 // _Genqueue_unused is currently unused.
75 // _Gcopystack means this goroutine's stack is being moved. It
76 // is not executing user code and is not on a run queue. The
77 // stack is owned by the goroutine that put it in _Gcopystack.
80 // _Gpreempted means this goroutine stopped itself for a
81 // suspendG preemption. It is like _Gwaiting, but nothing is
82 // yet responsible for ready()ing it. Some suspendG must CAS
83 // the status to _Gwaiting to take responsibility for
87 // _Gscan combined with one of the above states other than
88 // _Grunning indicates that GC is scanning the stack. The
89 // goroutine is not executing user code and the stack is owned
90 // by the goroutine that set the _Gscan bit.
92 // _Gscanrunning is different: it is used to briefly block
93 // state transitions while GC signals the G to scan its own
94 // stack. This is otherwise like _Grunning.
96 // atomicstatus&~Gscan gives the state the goroutine will
97 // return to when the scan completes.
99 _Gscanrunnable = _Gscan + _Grunnable // 0x1001
100 _Gscanrunning = _Gscan + _Grunning // 0x1002
101 _Gscansyscall = _Gscan + _Gsyscall // 0x1003
102 _Gscanwaiting = _Gscan + _Gwaiting // 0x1004
103 _Gscanpreempted = _Gscan + _Gpreempted // 0x1009
109 // _Pidle means a P is not being used to run user code or the
110 // scheduler. Typically, it's on the idle P list and available
111 // to the scheduler, but it may just be transitioning between
114 // The P is owned by the idle list or by whatever is
115 // transitioning its state. Its run queue is empty.
118 // _Prunning means a P is owned by an M and is being used to
119 // run user code or the scheduler. Only the M that owns this P
120 // is allowed to change the P's status from _Prunning. The M
121 // may transition the P to _Pidle (if it has no more work to
122 // do), _Psyscall (when entering a syscall), or _Pgcstop (to
123 // halt for the GC). The M may also hand ownership of the P
124 // off directly to another M (e.g., to schedule a locked G).
127 // _Psyscall means a P is not running user code. It has
128 // affinity to an M in a syscall but is not owned by it and
129 // may be stolen by another M. This is similar to _Pidle but
130 // uses lightweight transitions and maintains M affinity.
132 // Leaving _Psyscall must be done with a CAS, either to steal
133 // or retake the P. Note that there's an ABA hazard: even if
134 // an M successfully CASes its original P back to _Prunning
135 // after a syscall, it must understand the P may have been
136 // used by another M in the interim.
139 // _Pgcstop means a P is halted for STW and owned by the M
140 // that stopped the world. The M that stopped the world
141 // continues to use its P, even in _Pgcstop. Transitioning
142 // from _Prunning to _Pgcstop causes an M to release its P and
145 // The P retains its run queue and startTheWorld will restart
146 // the scheduler on Ps with non-empty run queues.
149 // _Pdead means a P is no longer used (GOMAXPROCS shrank). We
150 // reuse Ps if GOMAXPROCS increases. A dead P is mostly
151 // stripped of its resources, though a few things remain
152 // (e.g., trace buffers).
156 // Mutual exclusion locks. In the uncontended case,
157 // as fast as spin locks (just a few user-level instructions),
158 // but on the contention path they sleep in the kernel.
159 // A zeroed Mutex is unlocked (no need to initialize each lock).
160 // Initialization is helpful for static lock ranking, but not required.
162 // Empty struct if lock ranking is disabled, otherwise includes the lock rank
164 // Futex-based impl treats it as uint32 key,
165 // while sema-based impl as M* waitm.
166 // Used to be a union, but unions break precise GC.
170 // sleep and wakeup on one-time events.
171 // before any calls to notesleep or notewakeup,
172 // must call noteclear to initialize the Note.
173 // then, exactly one thread can call notesleep
174 // and exactly one thread can call notewakeup (once).
175 // once notewakeup has been called, the notesleep
176 // will return. future notesleep will return immediately.
177 // subsequent noteclear must be called only after
178 // previous notesleep has returned, e.g. it's disallowed
179 // to call noteclear straight after notewakeup.
181 // notetsleep is like notesleep but wakes up after
182 // a given number of nanoseconds even if the event
183 // has not yet happened. if a goroutine uses notetsleep to
184 // wake up early, it must wait to call noteclear until it
185 // can be sure that no other goroutine is calling
188 // notesleep/notetsleep are generally called on g0,
189 // notetsleepg is similar to notetsleep but is called on user g.
191 // Futex-based impl treats it as uint32 key,
192 // while sema-based impl as M* waitm.
193 // Used to be a union, but unions break precise GC.
197 type funcval struct {
199 // variable-size, fn-specific data here
212 func efaceOf(ep *any) *eface {
213 return (*eface)(unsafe.Pointer(ep))
216 // The guintptr, muintptr, and puintptr are all used to bypass write barriers.
217 // It is particularly important to avoid write barriers when the current P has
218 // been released, because the GC thinks the world is stopped, and an
219 // unexpected write barrier would not be synchronized with the GC,
220 // which can lead to a half-executed write barrier that has marked the object
221 // but not queued it. If the GC skips the object and completes before the
222 // queuing can occur, it will incorrectly free the object.
224 // We tried using special assignment functions invoked only when not
225 // holding a running P, but then some updates to a particular memory
226 // word went through write barriers and some did not. This breaks the
227 // write barrier shadow checking mode, and it is also scary: better to have
228 // a word that is completely ignored by the GC than to have one for which
229 // only a few updates are ignored.
231 // Gs and Ps are always reachable via true pointers in the
232 // allgs and allp lists or (during allocation before they reach those lists)
233 // from stack variables.
235 // Ms are always reachable via true pointers either from allm or
236 // freem. Unlike Gs and Ps we do free Ms, so it's important that
237 // nothing ever hold an muintptr across a safe point.
239 // A guintptr holds a goroutine pointer, but typed as a uintptr
240 // to bypass write barriers. It is used in the Gobuf goroutine state
241 // and in scheduling lists that are manipulated without a P.
243 // The Gobuf.g goroutine pointer is almost always updated by assembly code.
244 // In one of the few places it is updated by Go code - func save - it must be
245 // treated as a uintptr to avoid a write barrier being emitted at a bad time.
246 // Instead of figuring out how to emit the write barriers missing in the
247 // assembly manipulation, we change the type of the field to uintptr,
248 // so that it does not require write barriers at all.
250 // Goroutine structs are published in the allg list and never freed.
251 // That will keep the goroutine structs from being collected.
252 // There is never a time that Gobuf.g's contain the only references
253 // to a goroutine: the publishing of the goroutine in allg comes first.
254 // Goroutine pointers are also kept in non-GC-visible places like TLS,
255 // so I can't see them ever moving. If we did want to start moving data
256 // in the GC, we'd need to allocate the goroutine structs from an
257 // alternate arena. Using guintptr doesn't make that problem any worse.
258 // Note that pollDesc.rg, pollDesc.wg also store g in uintptr form,
259 // so they would need to be updated too if g's start moving.
260 type guintptr uintptr
263 func (gp guintptr) ptr() *g { return (*g)(unsafe.Pointer(gp)) }
266 func (gp *guintptr) set(g *g) { *gp = guintptr(unsafe.Pointer(g)) }
269 func (gp *guintptr) cas(old, new guintptr) bool {
270 return atomic.Casuintptr((*uintptr)(unsafe.Pointer(gp)), uintptr(old), uintptr(new))
273 // setGNoWB performs *gp = new without a write barrier.
274 // For times when it's impractical to use a guintptr.
278 func setGNoWB(gp **g, new *g) {
279 (*guintptr)(unsafe.Pointer(gp)).set(new)
282 type puintptr uintptr
285 func (pp puintptr) ptr() *p { return (*p)(unsafe.Pointer(pp)) }
288 func (pp *puintptr) set(p *p) { *pp = puintptr(unsafe.Pointer(p)) }
290 // muintptr is a *m that is not tracked by the garbage collector.
292 // Because we do free Ms, there are some additional constrains on
295 // 1. Never hold an muintptr locally across a safe point.
297 // 2. Any muintptr in the heap must be owned by the M itself so it can
298 // ensure it is not in use when the last true *m is released.
299 type muintptr uintptr
302 func (mp muintptr) ptr() *m { return (*m)(unsafe.Pointer(mp)) }
305 func (mp *muintptr) set(m *m) { *mp = muintptr(unsafe.Pointer(m)) }
307 // setMNoWB performs *mp = new without a write barrier.
308 // For times when it's impractical to use an muintptr.
312 func setMNoWB(mp **m, new *m) {
313 (*muintptr)(unsafe.Pointer(mp)).set(new)
317 // The offsets of sp, pc, and g are known to (hard-coded in) libmach.
319 // ctxt is unusual with respect to GC: it may be a
320 // heap-allocated funcval, so GC needs to track it, but it
321 // needs to be set and cleared from assembly, where it's
322 // difficult to have write barriers. However, ctxt is really a
323 // saved, live register, and we only ever exchange it between
324 // the real register and the gobuf. Hence, we treat it as a
325 // root during stack scanning, which means assembly that saves
326 // and restores it doesn't need write barriers. It's still
327 // typed as a pointer so that any other writes from Go get
335 bp uintptr // for framepointer-enabled architectures
338 // sudog represents a g in a wait list, such as for sending/receiving
341 // sudog is necessary because the g ↔ synchronization object relation
342 // is many-to-many. A g can be on many wait lists, so there may be
343 // many sudogs for one g; and many gs may be waiting on the same
344 // synchronization object, so there may be many sudogs for one object.
346 // sudogs are allocated from a special pool. Use acquireSudog and
347 // releaseSudog to allocate and free them.
349 // The following fields are protected by the hchan.lock of the
350 // channel this sudog is blocking on. shrinkstack depends on
351 // this for sudogs involved in channel ops.
357 elem unsafe.Pointer // data element (may point to stack)
359 // The following fields are never accessed concurrently.
360 // For channels, waitlink is only accessed by g.
361 // For semaphores, all fields (including the ones above)
362 // are only accessed when holding a semaRoot lock.
368 // isSelect indicates g is participating in a select, so
369 // g.selectDone must be CAS'd to win the wake-up race.
372 // success indicates whether communication over channel c
373 // succeeded. It is true if the goroutine was awoken because a
374 // value was delivered over channel c, and false if awoken
375 // because c was closed.
378 parent *sudog // semaRoot binary tree
379 waitlink *sudog // g.waiting list or semaRoot
380 waittail *sudog // semaRoot
384 type libcall struct {
386 n uintptr // number of parameters
387 args uintptr // parameters
388 r1 uintptr // return values
390 err uintptr // error number
393 // Stack describes a Go execution stack.
394 // The bounds of the stack are exactly [lo, hi),
395 // with no implicit data structures on either side.
401 // heldLockInfo gives info on a held lock and the rank of that lock
402 type heldLockInfo struct {
409 // stack describes the actual stack memory: [stack.lo, stack.hi).
410 // stackguard0 is the stack pointer compared in the Go stack growth prologue.
411 // It is stack.lo+StackGuard normally, but can be StackPreempt to trigger a preemption.
412 // stackguard1 is the stack pointer compared in the C stack growth prologue.
413 // It is stack.lo+StackGuard on g0 and gsignal stacks.
414 // It is ~0 on other goroutine stacks, to trigger a call to morestackc (and crash).
415 stack stack // offset known to runtime/cgo
416 stackguard0 uintptr // offset known to liblink
417 stackguard1 uintptr // offset known to liblink
419 _panic *_panic // innermost panic - offset known to liblink
420 _defer *_defer // innermost defer
421 m *m // current m; offset known to arm liblink
423 syscallsp uintptr // if status==Gsyscall, syscallsp = sched.sp to use during gc
424 syscallpc uintptr // if status==Gsyscall, syscallpc = sched.pc to use during gc
425 stktopsp uintptr // expected sp at top of stack, to check in traceback
426 // param is a generic pointer parameter field used to pass
427 // values in particular contexts where other storage for the
428 // parameter would be difficult to find. It is currently used
430 // 1. When a channel operation wakes up a blocked goroutine, it sets param to
431 // point to the sudog of the completed blocking operation.
432 // 2. By gcAssistAlloc1 to signal back to its caller that the goroutine completed
433 // the GC cycle. It is unsafe to do so in any other way, because the goroutine's
434 // stack may have moved in the meantime.
435 // 3. By debugCallWrap to pass parameters to a new goroutine because allocating a
436 // closure in the runtime is forbidden.
438 atomicstatus atomic.Uint32
439 stackLock uint32 // sigprof/scang lock; TODO: fold in to atomicstatus
442 waitsince int64 // approx time when the g become blocked
443 waitreason waitReason // if status==Gwaiting
445 preempt bool // preemption signal, duplicates stackguard0 = stackpreempt
446 preemptStop bool // transition to _Gpreempted on preemption; otherwise, just deschedule
447 preemptShrink bool // shrink stack at synchronous safe point
449 // asyncSafePoint is set if g is stopped at an asynchronous
450 // safe point. This means there are frames on the stack
451 // without precise pointer information.
454 paniconfault bool // panic (instead of crash) on unexpected fault address
455 gcscandone bool // g has scanned stack; protected by _Gscan bit in status
456 throwsplit bool // must not split stack
457 // activeStackChans indicates that there are unlocked channels
458 // pointing into this goroutine's stack. If true, stack
459 // copying needs to acquire channel locks to protect these
460 // areas of the stack.
461 activeStackChans bool
462 // parkingOnChan indicates that the goroutine is about to
463 // park on a chansend or chanrecv. Used to signal an unsafe point
464 // for stack shrinking.
465 parkingOnChan atomic.Bool
467 raceignore int8 // ignore race detection events
468 sysblocktraced bool // StartTrace has emitted EvGoInSyscall about this goroutine
469 tracking bool // whether we're tracking this G for sched latency statistics
470 trackingSeq uint8 // used to decide whether to track this G
471 trackingStamp int64 // timestamp of when the G last started being tracked
472 runnableTime int64 // the amount of time spent runnable, cleared when running, only used when tracking
473 sysexitticks int64 // cputicks when syscall has returned (for tracing)
474 traceseq uint64 // trace event sequencer
475 tracelastp puintptr // last P emitted an event for this goroutine
482 gopc uintptr // pc of go statement that created this goroutine
483 ancestors *[]ancestorInfo // ancestor information goroutine(s) that created this goroutine (only used if debug.tracebackancestors)
484 startpc uintptr // pc of goroutine function
486 waiting *sudog // sudog structures this g is waiting on (that have a valid elem ptr); in lock order
487 cgoCtxt []uintptr // cgo traceback context
488 labels unsafe.Pointer // profiler labels
489 timer *timer // cached timer for time.Sleep
490 selectDone atomic.Uint32 // are we participating in a select and did someone win the race?
492 // goroutineProfiled indicates the status of this goroutine's stack for the
493 // current in-progress goroutine profile
494 goroutineProfiled goroutineProfileStateHolder
498 // gcAssistBytes is this G's GC assist credit in terms of
499 // bytes allocated. If this is positive, then the G has credit
500 // to allocate gcAssistBytes bytes without assisting. If this
501 // is negative, then the G must correct this by performing
502 // scan work. We track this in bytes to make it fast to update
503 // and check for debt in the malloc hot path. The assist ratio
504 // determines how this corresponds to scan work debt.
508 // gTrackingPeriod is the number of transitions out of _Grunning between
509 // latency tracking runs.
510 const gTrackingPeriod = 8
513 // tlsSlots is the number of pointer-sized slots reserved for TLS on some platforms,
516 tlsSize = tlsSlots * goarch.PtrSize
520 g0 *g // goroutine with scheduling stack
521 morebuf gobuf // gobuf arg to morestack
522 divmod uint32 // div/mod denominator for arm - known to liblink
523 _ uint32 // align next field to 8 bytes
525 // Fields not known to debuggers.
526 procid uint64 // for debuggers, but offset not hard-coded
527 gsignal *g // signal-handling g
528 goSigStack gsignalStack // Go-allocated signal handling stack
529 sigmask sigset // storage for saved signal mask
530 tls [tlsSlots]uintptr // thread-local storage (for x86 extern register)
532 curg *g // current running goroutine
533 caughtsig guintptr // goroutine running during fatal signal
534 p puintptr // attached p for executing go code (nil if not executing go code)
536 oldp puintptr // the p that was attached before executing a syscall
540 preemptoff string // if != "", keep curg running on this m
544 spinning bool // m is out of work and is actively looking for work
545 blocked bool // m is blocked on a note
546 newSigstack bool // minit on C thread called sigaltstack
548 incgo bool // m is executing a cgo call
549 isextra bool // m is an extra m
550 freeWait uint32 // if == 0, safe to free g0 and delete m (atomic)
554 ncgocall uint64 // number of cgo calls in total
555 ncgo int32 // number of cgo calls currently in progress
556 cgoCallersUse atomic.Uint32 // if non-zero, cgoCallers in use temporarily
557 cgoCallers *cgoCallers // cgo traceback if crashing in cgo call
559 alllink *m // on allm
562 createstack [32]uintptr // stack that created this thread.
563 lockedExt uint32 // tracking for external LockOSThread
564 lockedInt uint32 // tracking for internal lockOSThread
565 nextwaitm muintptr // next m waiting for lock
566 waitunlockf func(*g, unsafe.Pointer) bool
567 waitlock unsafe.Pointer
572 freelink *m // on sched.freem
574 // these are here because they are too large to be on the stack
575 // of low-level NOSPLIT functions.
577 libcallpc uintptr // for cpu profiler
580 syscall libcall // stores syscall parameters on windows
582 vdsoSP uintptr // SP for traceback while in VDSO call (0 if not in call)
583 vdsoPC uintptr // PC for traceback while in VDSO call
585 // preemptGen counts the number of completed preemption
586 // signals. This is used to detect when a preemption is
587 // requested, but fails.
588 preemptGen atomic.Uint32
590 // Whether this is a pending preemption signal on this M.
591 signalPending atomic.Uint32
597 // Up to 10 locks held by this m, maintained by the lock ranking code.
599 locksHeld [10]heldLockInfo
604 status uint32 // one of pidle/prunning/...
606 schedtick uint32 // incremented on every scheduler call
607 syscalltick uint32 // incremented on every system call
608 sysmontick sysmontick // last tick observed by sysmon
609 m muintptr // back-link to associated m (nil if idle)
614 deferpool []*_defer // pool of available defer structs (see panic.go)
615 deferpoolbuf [32]*_defer
617 // Cache of goroutine ids, amortizes accesses to runtime·sched.goidgen.
621 // Queue of runnable goroutines. Accessed without lock.
625 // runnext, if non-nil, is a runnable G that was ready'd by
626 // the current G and should be run next instead of what's in
627 // runq if there's time remaining in the running G's time
628 // slice. It will inherit the time left in the current time
629 // slice. If a set of goroutines is locked in a
630 // communicate-and-wait pattern, this schedules that set as a
631 // unit and eliminates the (potentially large) scheduling
632 // latency that otherwise arises from adding the ready'd
633 // goroutines to the end of the run queue.
635 // Note that while other P's may atomically CAS this to zero,
636 // only the owner P can CAS it to a valid G.
639 // Available G's (status == Gdead)
648 // Cache of mspan objects from the heap.
650 // We need an explicit length here because this field is used
651 // in allocation codepaths where write barriers are not allowed,
652 // and eliminating the write barrier/keeping it eliminated from
653 // slice updates is tricky, moreso than just managing the length
661 // traceSweep indicates the sweep events should be traced.
662 // This is used to defer the sweep start event until a span
663 // has actually been swept.
665 // traceSwept and traceReclaimed track the number of bytes
666 // swept and reclaimed by sweeping in the current sweep loop.
667 traceSwept, traceReclaimed uintptr
669 palloc persistentAlloc // per-P to avoid mutex
671 // The when field of the first entry on the timer heap.
672 // This is 0 if the timer heap is empty.
673 timer0When atomic.Int64
675 // The earliest known nextwhen field of a timer with
676 // timerModifiedEarlier status. Because the timer may have been
677 // modified again, there need not be any timer with this value.
678 // This is 0 if there are no timerModifiedEarlier timers.
679 timerModifiedEarliest atomic.Int64
682 gcAssistTime int64 // Nanoseconds in assistAlloc
683 gcFractionalMarkTime int64 // Nanoseconds in fractional mark worker (atomic)
685 // limiterEvent tracks events for the GC CPU limiter.
686 limiterEvent limiterEvent
688 // gcMarkWorkerMode is the mode for the next mark worker to run in.
689 // That is, this is used to communicate with the worker goroutine
690 // selected for immediate execution by
691 // gcController.findRunnableGCWorker. When scheduling other goroutines,
692 // this field must be set to gcMarkWorkerNotWorker.
693 gcMarkWorkerMode gcMarkWorkerMode
694 // gcMarkWorkerStartTime is the nanotime() at which the most recent
695 // mark worker started.
696 gcMarkWorkerStartTime int64
698 // gcw is this P's GC work buffer cache. The work buffer is
699 // filled by write barriers, drained by mutator assists, and
700 // disposed on certain GC state transitions.
703 // wbBuf is this P's GC write barrier buffer.
705 // TODO: Consider caching this in the running G.
708 runSafePointFn uint32 // if 1, run sched.safePointFn at next safe point
710 // statsSeq is a counter indicating whether this P is currently
711 // writing any stats. Its value is even when not, odd when it is.
712 statsSeq atomic.Uint32
714 // Lock for timers. We normally access the timers while running
715 // on this P, but the scheduler can also do it from a different P.
718 // Actions to take at some time. This is used to implement the
719 // standard library's time package.
720 // Must hold timersLock to access.
723 // Number of timers in P's heap.
724 numTimers atomic.Uint32
726 // Number of timerDeleted timers in P's heap.
727 deletedTimers atomic.Uint32
729 // Race context used while executing timer functions.
732 // maxStackScanDelta accumulates the amount of stack space held by
733 // live goroutines (i.e. those eligible for stack scanning).
734 // Flushed to gcController.maxStackScan once maxStackScanSlack
735 // or -maxStackScanSlack is reached.
736 maxStackScanDelta int64
738 // gc-time statistics about current goroutines
739 // Note that this differs from maxStackScan in that this
740 // accumulates the actual stack observed to be used at GC time (hi - sp),
741 // not an instantaneous measure of the total stack size that might need
742 // to be scanned (hi - lo).
743 scannedStackSize uint64 // stack size of goroutines scanned by this P
744 scannedStacks uint64 // number of goroutines scanned by this P
746 // preempt is set to indicate that this P should be enter the
747 // scheduler ASAP (regardless of what G is running on it).
750 // Padding is no longer needed. False sharing is now not a worry because p is large enough
751 // that its size class is an integer multiple of the cache line size (for any of our architectures).
755 goidgen atomic.Uint64
756 lastpoll atomic.Int64 // time of last network poll, 0 if currently polling
757 pollUntil atomic.Int64 // time to which current poll is sleeping
761 // When increasing nmidle, nmidlelocked, nmsys, or nmfreed, be
762 // sure to call checkdead().
764 midle muintptr // idle m's waiting for work
765 nmidle int32 // number of idle m's waiting for work
766 nmidlelocked int32 // number of locked m's waiting for work
767 mnext int64 // number of m's that have been created and next M ID
768 maxmcount int32 // maximum number of m's allowed (or die)
769 nmsys int32 // number of system m's not counted for deadlock
770 nmfreed int64 // cumulative number of freed m's
772 ngsys atomic.Int32 // number of system goroutines
774 pidle puintptr // idle p's
776 nmspinning atomic.Int32 // See "Worker thread parking/unparking" comment in proc.go.
777 needspinning atomic.Uint32 // See "Delicate dance" comment in proc.go. Boolean. Must hold sched.lock to set to 1.
779 // Global runnable queue.
783 // disable controls selective disabling of the scheduler.
785 // Use schedEnableUser to control this.
787 // disable is protected by sched.lock.
789 // user disables scheduling of user goroutines.
791 runnable gQueue // pending runnable Gs
792 n int32 // length of runnable
795 // Global cache of dead G's.
798 stack gList // Gs with stacks
799 noStack gList // Gs without stacks
803 // Central cache of sudog structs.
807 // Central pool of available defer structs.
811 // freem is the list of m's waiting to be freed when their
812 // m.exited is set. Linked through m.freelink.
815 gcwaiting atomic.Bool // gc is waiting to run
818 sysmonwait atomic.Bool
821 // safepointFn should be called on each P at the next GC
822 // safepoint if p.runSafePointFn is set.
827 profilehz int32 // cpu profiling rate
829 procresizetime int64 // nanotime() of last change to gomaxprocs
830 totaltime int64 // ∫gomaxprocs dt up to procresizetime
832 // sysmonlock protects sysmon's actions on the runtime.
834 // Acquire and hold this mutex to block sysmon from interacting
835 // with the rest of the runtime.
838 // timeToRun is a distribution of scheduling latencies, defined
839 // as the sum of time a G spends in the _Grunnable state before
840 // it transitions to _Grunning.
841 timeToRun timeHistogram
843 // idleTime is the total CPU time Ps have "spent" idle.
845 // Reset on each GC cycle.
846 idleTime atomic.Int64
848 // totalMutexWaitTime is the sum of time goroutines have spent in _Gwaiting
849 // with a waitreason of the form waitReasonSync{RW,}Mutex{R,}Lock.
850 totalMutexWaitTime atomic.Int64
853 // Values for the flags field of a sigTabT.
855 _SigNotify = 1 << iota // let signal.Notify have signal, even if from kernel
856 _SigKill // if signal.Notify doesn't take it, exit quietly
857 _SigThrow // if signal.Notify doesn't take it, exit loudly
858 _SigPanic // if the signal is from the kernel, panic
859 _SigDefault // if the signal isn't explicitly requested, don't monitor it
860 _SigGoExit // cause all runtime procs to exit (only used on Plan 9).
861 _SigSetStack // Don't explicitly install handler, but add SA_ONSTACK to existing libc handler
862 _SigUnblock // always unblock; see blockableSig
863 _SigIgn // _SIG_DFL action is to ignore the signal
866 // Layout of in-memory per-function information prepared by linker
867 // See https://golang.org/s/go12symtab.
868 // Keep in sync with linker (../cmd/link/internal/ld/pcln.go:/pclntab)
869 // and with package debug/gosym and with symtab.go in package runtime.
871 entryOff uint32 // start pc, as offset from moduledata.text/pcHeader.textStart
872 nameOff int32 // function name, as index into moduledata.funcnametab.
874 args int32 // in/out args size
875 deferreturn uint32 // offset of start of a deferreturn call instruction from entry, if any.
881 cuOffset uint32 // runtime.cutab offset of this function's CU
882 funcID funcID // set for certain special runtime functions
885 nfuncdata uint8 // must be last, must end on a uint32-aligned boundary
887 // The end of the struct is followed immediately by two variable-length
888 // arrays that reference the pcdata and funcdata locations for this
891 // pcdata contains the offset into moduledata.pctab for the start of
892 // that index's table. e.g.,
893 // &moduledata.pctab[_func.pcdata[_PCDATA_UnsafePoint]] is the start of
894 // the unsafe point table.
896 // An offset of 0 indicates that there is no table.
898 // pcdata [npcdata]uint32
900 // funcdata contains the offset past moduledata.gofunc which contains a
901 // pointer to that index's funcdata. e.g.,
902 // *(moduledata.gofunc + _func.funcdata[_FUNCDATA_ArgsPointerMaps]) is
903 // the argument pointer map.
905 // An offset of ^uint32(0) indicates that there is no entry.
907 // funcdata [nfuncdata]uint32
910 // Pseudo-Func that is returned for PCs that occur in inlined code.
911 // A *Func can be either a *_func or a *funcinl, and they are distinguished
912 // by the first uintptr.
913 type funcinl struct {
914 ones uint32 // set to ^0 to distinguish from _func
915 entry uintptr // entry of the real (the "outermost") frame
921 // layout of Itab known to compilers
922 // allocated in non-garbage-collected memory
923 // Needs to be in sync with
924 // ../cmd/compile/internal/reflectdata/reflect.go:/^func.WriteTabs.
928 hash uint32 // copy of _type.hash. Used for type switches.
930 fun [1]uintptr // variable sized. fun[0]==0 means _type does not implement inter.
933 // Lock-free stack node.
934 // Also known to export_test.go.
940 type forcegcstate struct {
946 // extendRandom extends the random numbers in r[:n] to the whole slice r.
947 // Treats n<0 as n==0.
948 func extendRandom(r []byte, n int) {
953 // Extend random bits using hash function & time seed
958 h := memhash(unsafe.Pointer(&r[n-w]), uintptr(nanotime()), uintptr(w))
959 for i := 0; i < goarch.PtrSize && n < len(r); i++ {
967 // A _defer holds an entry on the list of deferred calls.
968 // If you add a field here, add code to clear it in deferProcStack.
969 // This struct must match the code in cmd/compile/internal/ssagen/ssa.go:deferstruct
970 // and cmd/compile/internal/ssagen/ssa.go:(*state).call.
971 // Some defers will be allocated on the stack and some on the heap.
972 // All defers are logically part of the stack, so write barriers to
973 // initialize them are not required. All defers must be manually scanned,
974 // and for heap defers, marked.
978 // openDefer indicates that this _defer is for a frame with open-coded
979 // defers. We have only one defer record for the entire frame (which may
980 // currently have 0, 1, or more defers active).
982 sp uintptr // sp at time of defer
983 pc uintptr // pc at time of defer
984 fn func() // can be nil for open-coded defers
985 _panic *_panic // panic that is running defer
986 link *_defer // next defer on G; can point to either heap or stack!
988 // If openDefer is true, the fields below record values about the stack
989 // frame and associated function that has the open-coded defer(s). sp
990 // above will be the sp for the frame, and pc will be address of the
991 // deferreturn call in the function.
992 fd unsafe.Pointer // funcdata for the function associated with the frame
993 varp uintptr // value of varp for the stack frame
994 // framepc is the current pc associated with the stack frame. Together,
995 // with sp above (which is the sp associated with the stack frame),
996 // framepc/sp can be used as pc/sp pair to continue a stack trace via
1001 // A _panic holds information about an active panic.
1003 // A _panic value must only ever live on the stack.
1005 // The argp and link fields are stack pointers, but don't need special
1006 // handling during stack growth: because they are pointer-typed and
1007 // _panic values only live on the stack, regular stack pointer
1008 // adjustment takes care of them.
1009 type _panic struct {
1010 argp unsafe.Pointer // pointer to arguments of deferred call run during panic; cannot move - known to liblink
1011 arg any // argument to panic
1012 link *_panic // link to earlier panic
1013 pc uintptr // where to return to in runtime if this panic is bypassed
1014 sp unsafe.Pointer // where to return to in runtime if this panic is bypassed
1015 recovered bool // whether this panic is over
1016 aborted bool // the panic was aborted
1020 // ancestorInfo records details of where a goroutine was started.
1021 type ancestorInfo struct {
1022 pcs []uintptr // pcs from the stack of this goroutine
1023 goid uint64 // goroutine id of this goroutine; original goroutine possibly dead
1024 gopc uintptr // pc of go statement that created this goroutine
1028 _TraceRuntimeFrames = 1 << iota // include frames for internal runtime functions.
1029 _TraceTrap // the initial PC, SP are from a trap, not a return PC from a call
1030 _TraceJumpStack // if traceback is on a systemstack, resume trace at g that called into it
1033 // The maximum number of frames we print for a traceback
1034 const _TracebackMaxFrames = 100
1036 // A waitReason explains why a goroutine has been stopped.
1037 // See gopark. Do not re-use waitReasons, add new ones.
1038 type waitReason uint8
1041 waitReasonZero waitReason = iota // ""
1042 waitReasonGCAssistMarking // "GC assist marking"
1043 waitReasonIOWait // "IO wait"
1044 waitReasonChanReceiveNilChan // "chan receive (nil chan)"
1045 waitReasonChanSendNilChan // "chan send (nil chan)"
1046 waitReasonDumpingHeap // "dumping heap"
1047 waitReasonGarbageCollection // "garbage collection"
1048 waitReasonGarbageCollectionScan // "garbage collection scan"
1049 waitReasonPanicWait // "panicwait"
1050 waitReasonSelect // "select"
1051 waitReasonSelectNoCases // "select (no cases)"
1052 waitReasonGCAssistWait // "GC assist wait"
1053 waitReasonGCSweepWait // "GC sweep wait"
1054 waitReasonGCScavengeWait // "GC scavenge wait"
1055 waitReasonChanReceive // "chan receive"
1056 waitReasonChanSend // "chan send"
1057 waitReasonFinalizerWait // "finalizer wait"
1058 waitReasonForceGCIdle // "force gc (idle)"
1059 waitReasonSemacquire // "semacquire"
1060 waitReasonSleep // "sleep"
1061 waitReasonSyncCondWait // "sync.Cond.Wait"
1062 waitReasonSyncMutexLock // "sync.Mutex.Lock"
1063 waitReasonSyncRWMutexRLock // "sync.RWMutex.RLock"
1064 waitReasonSyncRWMutexLock // "sync.RWMutex.Lock"
1065 waitReasonTraceReaderBlocked // "trace reader (blocked)"
1066 waitReasonWaitForGCCycle // "wait for GC cycle"
1067 waitReasonGCWorkerIdle // "GC worker (idle)"
1068 waitReasonGCWorkerActive // "GC worker (active)"
1069 waitReasonPreempted // "preempted"
1070 waitReasonDebugCall // "debug call"
1071 waitReasonGCMarkTermination // "GC mark termination"
1072 waitReasonStoppingTheWorld // "stopping the world"
1075 var waitReasonStrings = [...]string{
1077 waitReasonGCAssistMarking: "GC assist marking",
1078 waitReasonIOWait: "IO wait",
1079 waitReasonChanReceiveNilChan: "chan receive (nil chan)",
1080 waitReasonChanSendNilChan: "chan send (nil chan)",
1081 waitReasonDumpingHeap: "dumping heap",
1082 waitReasonGarbageCollection: "garbage collection",
1083 waitReasonGarbageCollectionScan: "garbage collection scan",
1084 waitReasonPanicWait: "panicwait",
1085 waitReasonSelect: "select",
1086 waitReasonSelectNoCases: "select (no cases)",
1087 waitReasonGCAssistWait: "GC assist wait",
1088 waitReasonGCSweepWait: "GC sweep wait",
1089 waitReasonGCScavengeWait: "GC scavenge wait",
1090 waitReasonChanReceive: "chan receive",
1091 waitReasonChanSend: "chan send",
1092 waitReasonFinalizerWait: "finalizer wait",
1093 waitReasonForceGCIdle: "force gc (idle)",
1094 waitReasonSemacquire: "semacquire",
1095 waitReasonSleep: "sleep",
1096 waitReasonSyncCondWait: "sync.Cond.Wait",
1097 waitReasonSyncMutexLock: "sync.Mutex.Lock",
1098 waitReasonSyncRWMutexRLock: "sync.RWMutex.RLock",
1099 waitReasonSyncRWMutexLock: "sync.RWMutex.Lock",
1100 waitReasonTraceReaderBlocked: "trace reader (blocked)",
1101 waitReasonWaitForGCCycle: "wait for GC cycle",
1102 waitReasonGCWorkerIdle: "GC worker (idle)",
1103 waitReasonGCWorkerActive: "GC worker (active)",
1104 waitReasonPreempted: "preempted",
1105 waitReasonDebugCall: "debug call",
1106 waitReasonGCMarkTermination: "GC mark termination",
1107 waitReasonStoppingTheWorld: "stopping the world",
1110 func (w waitReason) String() string {
1111 if w < 0 || w >= waitReason(len(waitReasonStrings)) {
1112 return "unknown wait reason"
1114 return waitReasonStrings[w]
1117 func (w waitReason) isMutexWait() bool {
1118 return w == waitReasonSyncMutexLock ||
1119 w == waitReasonSyncRWMutexRLock ||
1120 w == waitReasonSyncRWMutexLock
1127 forcegc forcegcstate
1131 // allpLock protects P-less reads and size changes of allp, idlepMask,
1132 // and timerpMask, and all writes to allp.
1134 // len(allp) == gomaxprocs; may change at safe points, otherwise
1137 // Bitmask of Ps in _Pidle list, one bit per P. Reads and writes must
1138 // be atomic. Length may change at safe points.
1140 // Each P must update only its own bit. In order to maintain
1141 // consistency, a P going idle must the idle mask simultaneously with
1142 // updates to the idle P list under the sched.lock, otherwise a racing
1143 // pidleget may clear the mask before pidleput sets the mask,
1144 // corrupting the bitmap.
1146 // N.B., procresize takes ownership of all Ps in stopTheWorldWithSema.
1148 // Bitmask of Ps that may have a timer, one bit per P. Reads and writes
1149 // must be atomic. Length may change at safe points.
1152 // Pool of GC parked background workers. Entries are type
1153 // *gcBgMarkWorkerNode.
1154 gcBgMarkWorkerPool lfstack
1156 // Total number of gcBgMarkWorker goroutines. Protected by worldsema.
1157 gcBgMarkWorkerCount int32
1159 // Information about what cpu features are available.
1160 // Packages outside the runtime should not use these
1161 // as they are not an external api.
1162 // Set on startup in asm_{386,amd64}.s
1163 processorVersionInfo uint32
1166 goarm uint8 // set by cmd/link on arm systems
1169 // Set by the linker so the runtime can determine the buildmode.
1171 islibrary bool // -buildmode=c-shared
1172 isarchive bool // -buildmode=c-archive
1175 // Must agree with internal/buildcfg.FramePointerEnabled.
1176 const framepointer_enabled = GOARCH == "amd64" || GOARCH == "arm64"