gcc/libgo/runtime/mprof.goc
2013-11-06 19:49:01 +00:00

561 lines
12 KiB
Plaintext

// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Malloc profiling.
// Patterned after tcmalloc's algorithms; shorter code.
package runtime
#include "runtime.h"
#include "arch.h"
#include "malloc.h"
#include "defs.h"
#include "go-type.h"
#include "go-string.h"
// NOTE(rsc): Everything here could use cas if contention became an issue.
static Lock proflock;
// All memory allocations are local and do not escape outside of the profiler.
// The profiler is forbidden from referring to garbage-collected memory.
enum { MProf, BProf }; // profile types
// Per-call-stack profiling information.
// Lookup by hashing call stack into a linked-list hash table.
typedef struct Bucket Bucket;
struct Bucket
{
Bucket *next; // next in hash list
Bucket *allnext; // next in list of all mbuckets/bbuckets
int32 typ;
// Generally unions can break precise GC,
// this one is fine because it does not contain pointers.
union
{
struct // typ == MProf
{
uintptr allocs;
uintptr frees;
uintptr alloc_bytes;
uintptr free_bytes;
uintptr recent_allocs; // since last gc
uintptr recent_frees;
uintptr recent_alloc_bytes;
uintptr recent_free_bytes;
};
struct // typ == BProf
{
int64 count;
int64 cycles;
};
};
uintptr hash;
uintptr nstk;
Location stk[1];
};
enum {
BuckHashSize = 179999,
};
static Bucket **buckhash;
static Bucket *mbuckets; // memory profile buckets
static Bucket *bbuckets; // blocking profile buckets
static uintptr bucketmem;
// Return the bucket for stk[0:nstk], allocating new bucket if needed.
static Bucket*
stkbucket(int32 typ, Location *stk, int32 nstk, bool alloc)
{
int32 i, j;
uintptr h;
Bucket *b;
if(buckhash == nil) {
buckhash = runtime_SysAlloc(BuckHashSize*sizeof buckhash[0], &mstats.buckhash_sys);
if(buckhash == nil)
runtime_throw("runtime: cannot allocate memory");
}
// Hash stack.
h = 0;
for(i=0; i<nstk; i++) {
h += stk[i].pc;
h += h<<10;
h ^= h>>6;
}
h += h<<3;
h ^= h>>11;
i = h%BuckHashSize;
for(b = buckhash[i]; b; b=b->next) {
if(b->typ == typ && b->hash == h && b->nstk == (uintptr)nstk) {
for(j = 0; j < nstk; j++) {
if(b->stk[j].pc != stk[j].pc ||
b->stk[j].lineno != stk[j].lineno ||
!__go_strings_equal(b->stk[j].filename, stk[j].filename))
break;
}
if (j == nstk)
return b;
}
}
if(!alloc)
return nil;
b = runtime_persistentalloc(sizeof *b + nstk*sizeof stk[0], 0, &mstats.buckhash_sys);
bucketmem += sizeof *b + nstk*sizeof stk[0];
runtime_memmove(b->stk, stk, nstk*sizeof stk[0]);
b->typ = typ;
b->hash = h;
b->nstk = nstk;
b->next = buckhash[i];
buckhash[i] = b;
if(typ == MProf) {
b->allnext = mbuckets;
mbuckets = b;
} else {
b->allnext = bbuckets;
bbuckets = b;
}
return b;
}
static void
MProf_GC(void)
{
Bucket *b;
for(b=mbuckets; b; b=b->allnext) {
b->allocs += b->recent_allocs;
b->frees += b->recent_frees;
b->alloc_bytes += b->recent_alloc_bytes;
b->free_bytes += b->recent_free_bytes;
b->recent_allocs = 0;
b->recent_frees = 0;
b->recent_alloc_bytes = 0;
b->recent_free_bytes = 0;
}
}
// Record that a gc just happened: all the 'recent' statistics are now real.
void
runtime_MProf_GC(void)
{
runtime_lock(&proflock);
MProf_GC();
runtime_unlock(&proflock);
}
// Map from pointer to Bucket* that allocated it.
// Three levels:
// Linked-list hash table for top N-AddrHashShift bits.
// Array index for next AddrDenseBits bits.
// Linked list for next AddrHashShift-AddrDenseBits bits.
// This is more efficient than using a general map,
// because of the typical clustering of the pointer keys.
typedef struct AddrHash AddrHash;
typedef struct AddrEntry AddrEntry;
enum {
AddrHashBits = 12, // good for 4GB of used address space
AddrHashShift = 20, // each AddrHash knows about 1MB of address space
AddrDenseBits = 8, // good for a profiling rate of 4096 bytes
};
struct AddrHash
{
AddrHash *next; // next in top-level hash table linked list
uintptr addr; // addr>>20
AddrEntry *dense[1<<AddrDenseBits];
};
struct AddrEntry
{
AddrEntry *next; // next in bottom-level linked list
uint32 addr;
Bucket *b;
};
static AddrHash **addrhash; // points to (AddrHash*)[1<<AddrHashBits]
static AddrEntry *addrfree;
static uintptr addrmem;
// Multiplicative hash function:
// hashMultiplier is the bottom 32 bits of int((sqrt(5)-1)/2 * (1<<32)).
// This is a good multiplier as suggested in CLR, Knuth. The hash
// value is taken to be the top AddrHashBits bits of the bottom 32 bits
// of the multiplied value.
enum {
HashMultiplier = 2654435769U
};
// Set the bucket associated with addr to b.
static void
setaddrbucket(uintptr addr, Bucket *b)
{
int32 i;
uint32 h;
AddrHash *ah;
AddrEntry *e;
h = (uint32)((addr>>AddrHashShift)*HashMultiplier) >> (32-AddrHashBits);
for(ah=addrhash[h]; ah; ah=ah->next)
if(ah->addr == (addr>>AddrHashShift))
goto found;
ah = runtime_persistentalloc(sizeof *ah, 0, &mstats.buckhash_sys);
addrmem += sizeof *ah;
ah->next = addrhash[h];
ah->addr = addr>>AddrHashShift;
addrhash[h] = ah;
found:
if((e = addrfree) == nil) {
e = runtime_persistentalloc(64*sizeof *e, 0, &mstats.buckhash_sys);
addrmem += 64*sizeof *e;
for(i=0; i+1<64; i++)
e[i].next = &e[i+1];
e[63].next = nil;
}
addrfree = e->next;
e->addr = (uint32)~(addr & ((1<<AddrHashShift)-1));
e->b = b;
h = (addr>>(AddrHashShift-AddrDenseBits))&(nelem(ah->dense)-1); // entry in dense is top 8 bits of low 20.
e->next = ah->dense[h];
ah->dense[h] = e;
}
// Get the bucket associated with addr and clear the association.
static Bucket*
getaddrbucket(uintptr addr)
{
uint32 h;
AddrHash *ah;
AddrEntry *e, **l;
Bucket *b;
h = (uint32)((addr>>AddrHashShift)*HashMultiplier) >> (32-AddrHashBits);
for(ah=addrhash[h]; ah; ah=ah->next)
if(ah->addr == (addr>>AddrHashShift))
goto found;
return nil;
found:
h = (addr>>(AddrHashShift-AddrDenseBits))&(nelem(ah->dense)-1); // entry in dense is top 8 bits of low 20.
for(l=&ah->dense[h]; (e=*l) != nil; l=&e->next) {
if(e->addr == (uint32)~(addr & ((1<<AddrHashShift)-1))) {
*l = e->next;
b = e->b;
e->next = addrfree;
addrfree = e;
return b;
}
}
return nil;
}
// Called by malloc to record a profiled block.
void
runtime_MProf_Malloc(void *p, uintptr size)
{
int32 nstk;
Location stk[32];
Bucket *b;
nstk = runtime_callers(1, stk, 32);
runtime_lock(&proflock);
b = stkbucket(MProf, stk, nstk, true);
b->recent_allocs++;
b->recent_alloc_bytes += size;
setaddrbucket((uintptr)p, b);
runtime_unlock(&proflock);
}
// Called when freeing a profiled block.
void
runtime_MProf_Free(void *p, uintptr size)
{
Bucket *b;
runtime_lock(&proflock);
b = getaddrbucket((uintptr)p);
if(b != nil) {
b->recent_frees++;
b->recent_free_bytes += size;
}
runtime_unlock(&proflock);
}
int64 runtime_blockprofilerate; // in CPU ticks
void runtime_SetBlockProfileRate(intgo) __asm__ (GOSYM_PREFIX "runtime.SetBlockProfileRate");
void
runtime_SetBlockProfileRate(intgo rate)
{
int64 r;
if(rate <= 0)
r = 0; // disable profiling
else {
// convert ns to cycles, use float64 to prevent overflow during multiplication
r = (float64)rate*runtime_tickspersecond()/(1000*1000*1000);
if(r == 0)
r = 1;
}
runtime_atomicstore64((uint64*)&runtime_blockprofilerate, r);
}
void
runtime_blockevent(int64 cycles, int32 skip)
{
int32 nstk;
int64 rate;
Location stk[32];
Bucket *b;
if(cycles <= 0)
return;
rate = runtime_atomicload64((uint64*)&runtime_blockprofilerate);
if(rate <= 0 || (rate > cycles && runtime_fastrand1()%rate > cycles))
return;
nstk = runtime_callers(skip, stk, 32);
runtime_lock(&proflock);
b = stkbucket(BProf, stk, nstk, true);
b->count++;
b->cycles += cycles;
runtime_unlock(&proflock);
}
// Go interface to profile data. (Declared in debug.go)
// Must match MemProfileRecord in debug.go.
typedef struct Record Record;
struct Record {
int64 alloc_bytes, free_bytes;
int64 alloc_objects, free_objects;
uintptr stk[32];
};
// Write b's data to r.
static void
record(Record *r, Bucket *b)
{
uint32 i;
r->alloc_bytes = b->alloc_bytes;
r->free_bytes = b->free_bytes;
r->alloc_objects = b->allocs;
r->free_objects = b->frees;
for(i=0; i<b->nstk && i<nelem(r->stk); i++)
r->stk[i] = b->stk[i].pc;
for(; i<nelem(r->stk); i++)
r->stk[i] = 0;
}
func MemProfile(p Slice, include_inuse_zero bool) (n int, ok bool) {
Bucket *b;
Record *r;
bool clear;
runtime_lock(&proflock);
n = 0;
clear = true;
for(b=mbuckets; b; b=b->allnext) {
if(include_inuse_zero || b->alloc_bytes != b->free_bytes)
n++;
if(b->allocs != 0 || b->frees != 0)
clear = false;
}
if(clear) {
// Absolutely no data, suggesting that a garbage collection
// has not yet happened. In order to allow profiling when
// garbage collection is disabled from the beginning of execution,
// accumulate stats as if a GC just happened, and recount buckets.
MProf_GC();
n = 0;
for(b=mbuckets; b; b=b->allnext)
if(include_inuse_zero || b->alloc_bytes != b->free_bytes)
n++;
}
ok = false;
if(n <= p.__count) {
ok = true;
r = (Record*)p.__values;
for(b=mbuckets; b; b=b->allnext)
if(include_inuse_zero || b->alloc_bytes != b->free_bytes)
record(r++, b);
}
runtime_unlock(&proflock);
}
void
runtime_MProf_Mark(void (*addroot)(Obj))
{
// buckhash is not allocated via mallocgc.
addroot((Obj){(byte*)&mbuckets, sizeof mbuckets, 0});
addroot((Obj){(byte*)&bbuckets, sizeof bbuckets, 0});
addroot((Obj){(byte*)&addrhash, sizeof addrhash, 0});
addroot((Obj){(byte*)&addrfree, sizeof addrfree, 0});
}
// Must match BlockProfileRecord in debug.go.
typedef struct BRecord BRecord;
struct BRecord {
int64 count;
int64 cycles;
uintptr stk[32];
};
func BlockProfile(p Slice) (n int, ok bool) {
Bucket *b;
BRecord *r;
int32 i;
runtime_lock(&proflock);
n = 0;
for(b=bbuckets; b; b=b->allnext)
n++;
ok = false;
if(n <= p.__count) {
ok = true;
r = (BRecord*)p.__values;
for(b=bbuckets; b; b=b->allnext, r++) {
r->count = b->count;
r->cycles = b->cycles;
for(i=0; (uintptr)i<b->nstk && (uintptr)i<nelem(r->stk); i++)
r->stk[i] = b->stk[i].pc;
for(; (uintptr)i<nelem(r->stk); i++)
r->stk[i] = 0;
}
}
runtime_unlock(&proflock);
}
// Must match StackRecord in debug.go.
typedef struct TRecord TRecord;
struct TRecord {
uintptr stk[32];
};
func ThreadCreateProfile(p Slice) (n int, ok bool) {
TRecord *r;
M *first, *mp;
int32 i;
first = runtime_atomicloadp(&runtime_allm);
n = 0;
for(mp=first; mp; mp=mp->alllink)
n++;
ok = false;
if(n <= p.__count) {
ok = true;
r = (TRecord*)p.__values;
for(mp=first; mp; mp=mp->alllink) {
for(i = 0; (uintptr)i < nelem(r->stk); i++) {
r->stk[i] = mp->createstack[i].pc;
}
r++;
}
}
}
func Stack(b Slice, all bool) (n int) {
byte *pc, *sp;
bool enablegc;
sp = runtime_getcallersp(&b);
pc = (byte*)(uintptr)runtime_getcallerpc(&b);
if(all) {
runtime_semacquire(&runtime_worldsema, false);
runtime_m()->gcing = 1;
runtime_stoptheworld();
enablegc = mstats.enablegc;
mstats.enablegc = false;
}
if(b.__count == 0)
n = 0;
else{
G* g = runtime_g();
g->writebuf = (byte*)b.__values;
g->writenbuf = b.__count;
USED(pc);
USED(sp);
runtime_goroutineheader(g);
runtime_traceback();
runtime_printcreatedby(g);
if(all)
runtime_tracebackothers(g);
n = b.__count - g->writenbuf;
g->writebuf = nil;
g->writenbuf = 0;
}
if(all) {
runtime_m()->gcing = 0;
mstats.enablegc = enablegc;
runtime_semrelease(&runtime_worldsema);
runtime_starttheworld();
}
}
static void
saveg(G *gp, TRecord *r)
{
int32 n, i;
Location locstk[nelem(r->stk)];
if(gp == runtime_g()) {
n = runtime_callers(0, locstk, nelem(r->stk));
for(i = 0; i < n; i++)
r->stk[i] = locstk[i].pc;
}
else {
// FIXME: Not implemented.
n = 0;
}
if((size_t)n < nelem(r->stk))
r->stk[n] = 0;
}
func GoroutineProfile(b Slice) (n int, ok bool) {
TRecord *r;
G *gp;
ok = false;
n = runtime_gcount();
if(n <= b.__count) {
runtime_semacquire(&runtime_worldsema, false);
runtime_m()->gcing = 1;
runtime_stoptheworld();
n = runtime_gcount();
if(n <= b.__count) {
G* g = runtime_g();
ok = true;
r = (TRecord*)b.__values;
saveg(g, r++);
for(gp = runtime_allg; gp != nil; gp = gp->alllink) {
if(gp == g || gp->status == Gdead)
continue;
saveg(gp, r++);
}
}
runtime_m()->gcing = 0;
runtime_semrelease(&runtime_worldsema);
runtime_starttheworld();
}
}
void
runtime_mprofinit(void)
{
addrhash = runtime_persistentalloc((1<<AddrHashBits)*sizeof *addrhash, 0, &mstats.buckhash_sys);
}