core/rawdb: use atomic int added in go1.19 (#26935)
This commit is contained in:
parent
8a9a73c99b
commit
905a723fae
@ -43,10 +43,7 @@ const (
|
||||
// The background thread will keep moving ancient chain segments from key-value
|
||||
// database to flat files for saving space on live database.
|
||||
type chainFreezer struct {
|
||||
// WARNING: The `threshold` field is accessed atomically. On 32 bit platforms, only
|
||||
// 64-bit aligned fields can be atomic. The struct is guaranteed to be so aligned,
|
||||
// so take advantage of that (https://golang.org/pkg/sync/atomic/#pkg-note-BUG).
|
||||
threshold uint64 // Number of recent blocks not to freeze (params.FullImmutabilityThreshold apart from tests)
|
||||
threshold atomic.Uint64 // Number of recent blocks not to freeze (params.FullImmutabilityThreshold apart from tests)
|
||||
|
||||
*Freezer
|
||||
quit chan struct{}
|
||||
@ -60,12 +57,13 @@ func newChainFreezer(datadir string, namespace string, readonly bool) (*chainFre
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
return &chainFreezer{
|
||||
Freezer: freezer,
|
||||
threshold: params.FullImmutabilityThreshold,
|
||||
quit: make(chan struct{}),
|
||||
trigger: make(chan chan struct{}),
|
||||
}, nil
|
||||
cf := chainFreezer{
|
||||
Freezer: freezer,
|
||||
quit: make(chan struct{}),
|
||||
trigger: make(chan chan struct{}),
|
||||
}
|
||||
cf.threshold.Store(params.FullImmutabilityThreshold)
|
||||
return &cf, nil
|
||||
}
|
||||
|
||||
// Close closes the chain freezer instance and terminates the background thread.
|
||||
@ -124,8 +122,8 @@ func (f *chainFreezer) freeze(db ethdb.KeyValueStore) {
|
||||
continue
|
||||
}
|
||||
number := ReadHeaderNumber(nfdb, hash)
|
||||
threshold := atomic.LoadUint64(&f.threshold)
|
||||
frozen := atomic.LoadUint64(&f.frozen)
|
||||
threshold := f.threshold.Load()
|
||||
frozen := f.frozen.Load()
|
||||
switch {
|
||||
case number == nil:
|
||||
log.Error("Current full block number unavailable", "hash", hash)
|
||||
@ -186,7 +184,7 @@ func (f *chainFreezer) freeze(db ethdb.KeyValueStore) {
|
||||
|
||||
// Wipe out side chains also and track dangling side chains
|
||||
var dangling []common.Hash
|
||||
frozen = atomic.LoadUint64(&f.frozen) // Needs reload after during freezeRange
|
||||
frozen = f.frozen.Load() // Needs reload after during freezeRange
|
||||
for number := first; number < frozen; number++ {
|
||||
// Always keep the genesis block in active database
|
||||
if number != 0 {
|
||||
|
@ -132,11 +132,12 @@ func iterateTransactions(db ethdb.Database, from uint64, to uint64, reverse bool
|
||||
}
|
||||
}
|
||||
// process runs in parallel
|
||||
nThreadsAlive := int32(threads)
|
||||
var nThreadsAlive atomic.Int32
|
||||
nThreadsAlive.Store(int32(threads))
|
||||
process := func() {
|
||||
defer func() {
|
||||
// Last processor closes the result channel
|
||||
if atomic.AddInt32(&nThreadsAlive, -1) == 0 {
|
||||
if nThreadsAlive.Add(-1) == 0 {
|
||||
close(hashesCh)
|
||||
}
|
||||
}()
|
||||
|
@ -24,7 +24,6 @@ import (
|
||||
"path"
|
||||
"path/filepath"
|
||||
"strings"
|
||||
"sync/atomic"
|
||||
"time"
|
||||
|
||||
"github.com/ethereum/go-ethereum/common"
|
||||
@ -72,9 +71,9 @@ func (frdb *freezerdb) Freeze(threshold uint64) error {
|
||||
}
|
||||
// Set the freezer threshold to a temporary value
|
||||
defer func(old uint64) {
|
||||
atomic.StoreUint64(&frdb.AncientStore.(*chainFreezer).threshold, old)
|
||||
}(atomic.LoadUint64(&frdb.AncientStore.(*chainFreezer).threshold))
|
||||
atomic.StoreUint64(&frdb.AncientStore.(*chainFreezer).threshold, threshold)
|
||||
frdb.AncientStore.(*chainFreezer).threshold.Store(old)
|
||||
}(frdb.AncientStore.(*chainFreezer).threshold.Load())
|
||||
frdb.AncientStore.(*chainFreezer).threshold.Store(threshold)
|
||||
|
||||
// Trigger a freeze cycle and block until it's done
|
||||
trigger := make(chan struct{}, 1)
|
||||
|
@ -62,11 +62,8 @@ const freezerTableSize = 2 * 1000 * 1000 * 1000
|
||||
// reserving it for go-ethereum. This would also reduce the memory requirements
|
||||
// of Geth, and thus also GC overhead.
|
||||
type Freezer struct {
|
||||
// WARNING: The `frozen` and `tail` fields are accessed atomically. On 32 bit platforms, only
|
||||
// 64-bit aligned fields can be atomic. The struct is guaranteed to be so aligned,
|
||||
// so take advantage of that (https://golang.org/pkg/sync/atomic/#pkg-note-BUG).
|
||||
frozen uint64 // Number of blocks already frozen
|
||||
tail uint64 // Number of the first stored item in the freezer
|
||||
frozen atomic.Uint64 // Number of blocks already frozen
|
||||
tail atomic.Uint64 // Number of the first stored item in the freezer
|
||||
|
||||
// This lock synchronizes writers and the truncate operation, as well as
|
||||
// the "atomic" (batched) read operations.
|
||||
@ -212,12 +209,12 @@ func (f *Freezer) AncientRange(kind string, start, count, maxBytes uint64) ([][]
|
||||
|
||||
// Ancients returns the length of the frozen items.
|
||||
func (f *Freezer) Ancients() (uint64, error) {
|
||||
return atomic.LoadUint64(&f.frozen), nil
|
||||
return f.frozen.Load(), nil
|
||||
}
|
||||
|
||||
// Tail returns the number of first stored item in the freezer.
|
||||
func (f *Freezer) Tail() (uint64, error) {
|
||||
return atomic.LoadUint64(&f.tail), nil
|
||||
return f.tail.Load(), nil
|
||||
}
|
||||
|
||||
// AncientSize returns the ancient size of the specified category.
|
||||
@ -251,7 +248,7 @@ func (f *Freezer) ModifyAncients(fn func(ethdb.AncientWriteOp) error) (writeSize
|
||||
defer f.writeLock.Unlock()
|
||||
|
||||
// Roll back all tables to the starting position in case of error.
|
||||
prevItem := atomic.LoadUint64(&f.frozen)
|
||||
prevItem := f.frozen.Load()
|
||||
defer func() {
|
||||
if err != nil {
|
||||
// The write operation has failed. Go back to the previous item position.
|
||||
@ -272,7 +269,7 @@ func (f *Freezer) ModifyAncients(fn func(ethdb.AncientWriteOp) error) (writeSize
|
||||
if err != nil {
|
||||
return 0, err
|
||||
}
|
||||
atomic.StoreUint64(&f.frozen, item)
|
||||
f.frozen.Store(item)
|
||||
return writeSize, nil
|
||||
}
|
||||
|
||||
@ -284,7 +281,7 @@ func (f *Freezer) TruncateHead(items uint64) error {
|
||||
f.writeLock.Lock()
|
||||
defer f.writeLock.Unlock()
|
||||
|
||||
if atomic.LoadUint64(&f.frozen) <= items {
|
||||
if f.frozen.Load() <= items {
|
||||
return nil
|
||||
}
|
||||
for _, table := range f.tables {
|
||||
@ -292,7 +289,7 @@ func (f *Freezer) TruncateHead(items uint64) error {
|
||||
return err
|
||||
}
|
||||
}
|
||||
atomic.StoreUint64(&f.frozen, items)
|
||||
f.frozen.Store(items)
|
||||
return nil
|
||||
}
|
||||
|
||||
@ -304,7 +301,7 @@ func (f *Freezer) TruncateTail(tail uint64) error {
|
||||
f.writeLock.Lock()
|
||||
defer f.writeLock.Unlock()
|
||||
|
||||
if atomic.LoadUint64(&f.tail) >= tail {
|
||||
if f.tail.Load() >= tail {
|
||||
return nil
|
||||
}
|
||||
for _, table := range f.tables {
|
||||
@ -312,7 +309,7 @@ func (f *Freezer) TruncateTail(tail uint64) error {
|
||||
return err
|
||||
}
|
||||
}
|
||||
atomic.StoreUint64(&f.tail, tail)
|
||||
f.tail.Store(tail)
|
||||
return nil
|
||||
}
|
||||
|
||||
@ -343,22 +340,22 @@ func (f *Freezer) validate() error {
|
||||
)
|
||||
// Hack to get boundary of any table
|
||||
for kind, table := range f.tables {
|
||||
head = atomic.LoadUint64(&table.items)
|
||||
tail = atomic.LoadUint64(&table.itemHidden)
|
||||
head = table.items.Load()
|
||||
tail = table.itemHidden.Load()
|
||||
name = kind
|
||||
break
|
||||
}
|
||||
// Now check every table against those boundaries.
|
||||
for kind, table := range f.tables {
|
||||
if head != atomic.LoadUint64(&table.items) {
|
||||
return fmt.Errorf("freezer tables %s and %s have differing head: %d != %d", kind, name, atomic.LoadUint64(&table.items), head)
|
||||
if head != table.items.Load() {
|
||||
return fmt.Errorf("freezer tables %s and %s have differing head: %d != %d", kind, name, table.items.Load(), head)
|
||||
}
|
||||
if tail != atomic.LoadUint64(&table.itemHidden) {
|
||||
return fmt.Errorf("freezer tables %s and %s have differing tail: %d != %d", kind, name, atomic.LoadUint64(&table.itemHidden), tail)
|
||||
if tail != table.itemHidden.Load() {
|
||||
return fmt.Errorf("freezer tables %s and %s have differing tail: %d != %d", kind, name, table.itemHidden.Load(), tail)
|
||||
}
|
||||
}
|
||||
atomic.StoreUint64(&f.frozen, head)
|
||||
atomic.StoreUint64(&f.tail, tail)
|
||||
f.frozen.Store(head)
|
||||
f.tail.Store(tail)
|
||||
return nil
|
||||
}
|
||||
|
||||
@ -369,11 +366,11 @@ func (f *Freezer) repair() error {
|
||||
tail = uint64(0)
|
||||
)
|
||||
for _, table := range f.tables {
|
||||
items := atomic.LoadUint64(&table.items)
|
||||
items := table.items.Load()
|
||||
if head > items {
|
||||
head = items
|
||||
}
|
||||
hidden := atomic.LoadUint64(&table.itemHidden)
|
||||
hidden := table.itemHidden.Load()
|
||||
if hidden > tail {
|
||||
tail = hidden
|
||||
}
|
||||
@ -386,8 +383,8 @@ func (f *Freezer) repair() error {
|
||||
return err
|
||||
}
|
||||
}
|
||||
atomic.StoreUint64(&f.frozen, head)
|
||||
atomic.StoreUint64(&f.tail, tail)
|
||||
f.frozen.Store(head)
|
||||
f.tail.Store(tail)
|
||||
return nil
|
||||
}
|
||||
|
||||
@ -413,7 +410,7 @@ func (f *Freezer) MigrateTable(kind string, convert convertLegacyFn) error {
|
||||
// and that error will be returned.
|
||||
forEach := func(t *freezerTable, offset uint64, fn func(uint64, []byte) error) error {
|
||||
var (
|
||||
items = atomic.LoadUint64(&t.items)
|
||||
items = t.items.Load()
|
||||
batchSize = uint64(1024)
|
||||
maxBytes = uint64(1024 * 1024)
|
||||
)
|
||||
@ -436,7 +433,7 @@ func (f *Freezer) MigrateTable(kind string, convert convertLegacyFn) error {
|
||||
}
|
||||
// TODO(s1na): This is a sanity-check since as of now no process does tail-deletion. But the migration
|
||||
// process assumes no deletion at tail and needs to be modified to account for that.
|
||||
if table.itemOffset > 0 || table.itemHidden > 0 {
|
||||
if table.itemOffset.Load() > 0 || table.itemHidden.Load() > 0 {
|
||||
return fmt.Errorf("migration not supported for tail-deleted freezers")
|
||||
}
|
||||
ancientsPath := filepath.Dir(table.index.Name())
|
||||
@ -452,7 +449,7 @@ func (f *Freezer) MigrateTable(kind string, convert convertLegacyFn) error {
|
||||
out []byte
|
||||
start = time.Now()
|
||||
logged = time.Now()
|
||||
offset = newTable.items
|
||||
offset = newTable.items.Load()
|
||||
)
|
||||
if offset > 0 {
|
||||
log.Info("found previous migration attempt", "migrated", offset)
|
||||
|
@ -18,7 +18,6 @@ package rawdb
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
"sync/atomic"
|
||||
|
||||
"github.com/ethereum/go-ethereum/common/math"
|
||||
"github.com/ethereum/go-ethereum/rlp"
|
||||
@ -107,7 +106,7 @@ func (t *freezerTable) newBatch() *freezerTableBatch {
|
||||
func (batch *freezerTableBatch) reset() {
|
||||
batch.dataBuffer = batch.dataBuffer[:0]
|
||||
batch.indexBuffer = batch.indexBuffer[:0]
|
||||
batch.curItem = atomic.LoadUint64(&batch.t.items)
|
||||
batch.curItem = batch.t.items.Load()
|
||||
batch.totalBytes = 0
|
||||
}
|
||||
|
||||
@ -201,7 +200,7 @@ func (batch *freezerTableBatch) commit() error {
|
||||
|
||||
// Update headBytes of table.
|
||||
batch.t.headBytes += dataSize
|
||||
atomic.StoreUint64(&batch.t.items, batch.curItem)
|
||||
batch.t.items.Store(batch.curItem)
|
||||
|
||||
// Update metrics.
|
||||
batch.t.sizeGauge.Inc(dataSize + indexSize)
|
||||
|
@ -88,18 +88,15 @@ func (i *indexEntry) bounds(end *indexEntry) (startOffset, endOffset, fileId uin
|
||||
// It consists of a data file (snappy encoded arbitrary data blobs) and an indexEntry
|
||||
// file (uncompressed 64 bit indices into the data file).
|
||||
type freezerTable struct {
|
||||
// WARNING: The `items` field is accessed atomically. On 32 bit platforms, only
|
||||
// 64-bit aligned fields can be atomic. The struct is guaranteed to be so aligned,
|
||||
// so take advantage of that (https://golang.org/pkg/sync/atomic/#pkg-note-BUG).
|
||||
items uint64 // Number of items stored in the table (including items removed from tail)
|
||||
itemOffset uint64 // Number of items removed from the table
|
||||
items atomic.Uint64 // Number of items stored in the table (including items removed from tail)
|
||||
itemOffset atomic.Uint64 // Number of items removed from the table
|
||||
|
||||
// itemHidden is the number of items marked as deleted. Tail deletion is
|
||||
// only supported at file level which means the actual deletion will be
|
||||
// delayed until the entire data file is marked as deleted. Before that
|
||||
// these items will be hidden to prevent being visited again. The value
|
||||
// should never be lower than itemOffset.
|
||||
itemHidden uint64
|
||||
itemHidden atomic.Uint64
|
||||
|
||||
noCompression bool // if true, disables snappy compression. Note: does not work retroactively
|
||||
readonly bool
|
||||
@ -241,14 +238,14 @@ func (t *freezerTable) repair() error {
|
||||
// which is not enough in theory but enough in practice.
|
||||
// TODO: use uint64 to represent total removed items.
|
||||
t.tailId = firstIndex.filenum
|
||||
t.itemOffset = uint64(firstIndex.offset)
|
||||
t.itemOffset.Store(uint64(firstIndex.offset))
|
||||
|
||||
// Load metadata from the file
|
||||
meta, err := loadMetadata(t.meta, t.itemOffset)
|
||||
meta, err := loadMetadata(t.meta, t.itemOffset.Load())
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
t.itemHidden = meta.VirtualTail
|
||||
t.itemHidden.Store(meta.VirtualTail)
|
||||
|
||||
// Read the last index, use the default value in case the freezer is empty
|
||||
if offsetsSize == indexEntrySize {
|
||||
@ -331,7 +328,7 @@ func (t *freezerTable) repair() error {
|
||||
}
|
||||
}
|
||||
// Update the item and byte counters and return
|
||||
t.items = t.itemOffset + uint64(offsetsSize/indexEntrySize-1) // last indexEntry points to the end of the data file
|
||||
t.items.Store(t.itemOffset.Load() + uint64(offsetsSize/indexEntrySize-1)) // last indexEntry points to the end of the data file
|
||||
t.headBytes = contentSize
|
||||
t.headId = lastIndex.filenum
|
||||
|
||||
@ -346,9 +343,9 @@ func (t *freezerTable) repair() error {
|
||||
return err
|
||||
}
|
||||
if verbose {
|
||||
t.logger.Info("Chain freezer table opened", "items", t.items, "size", t.headBytes)
|
||||
t.logger.Info("Chain freezer table opened", "items", t.items.Load(), "size", t.headBytes)
|
||||
} else {
|
||||
t.logger.Debug("Chain freezer table opened", "items", t.items, "size", common.StorageSize(t.headBytes))
|
||||
t.logger.Debug("Chain freezer table opened", "items", t.items.Load(), "size", common.StorageSize(t.headBytes))
|
||||
}
|
||||
return nil
|
||||
}
|
||||
@ -382,11 +379,11 @@ func (t *freezerTable) truncateHead(items uint64) error {
|
||||
defer t.lock.Unlock()
|
||||
|
||||
// Ensure the given truncate target falls in the correct range
|
||||
existing := atomic.LoadUint64(&t.items)
|
||||
existing := t.items.Load()
|
||||
if existing <= items {
|
||||
return nil
|
||||
}
|
||||
if items < atomic.LoadUint64(&t.itemHidden) {
|
||||
if items < t.itemHidden.Load() {
|
||||
return errors.New("truncation below tail")
|
||||
}
|
||||
// We need to truncate, save the old size for metrics tracking
|
||||
@ -403,7 +400,7 @@ func (t *freezerTable) truncateHead(items uint64) error {
|
||||
|
||||
// Truncate the index file first, the tail position is also considered
|
||||
// when calculating the new freezer table length.
|
||||
length := items - atomic.LoadUint64(&t.itemOffset)
|
||||
length := items - t.itemOffset.Load()
|
||||
if err := truncateFreezerFile(t.index, int64(length+1)*indexEntrySize); err != nil {
|
||||
return err
|
||||
}
|
||||
@ -438,7 +435,7 @@ func (t *freezerTable) truncateHead(items uint64) error {
|
||||
}
|
||||
// All data files truncated, set internal counters and return
|
||||
t.headBytes = int64(expected.offset)
|
||||
atomic.StoreUint64(&t.items, items)
|
||||
t.items.Store(items)
|
||||
|
||||
// Retrieve the new size and update the total size counter
|
||||
newSize, err := t.sizeNolock()
|
||||
@ -455,10 +452,10 @@ func (t *freezerTable) truncateTail(items uint64) error {
|
||||
defer t.lock.Unlock()
|
||||
|
||||
// Ensure the given truncate target falls in the correct range
|
||||
if atomic.LoadUint64(&t.itemHidden) >= items {
|
||||
if t.itemHidden.Load() >= items {
|
||||
return nil
|
||||
}
|
||||
if atomic.LoadUint64(&t.items) < items {
|
||||
if t.items.Load() < items {
|
||||
return errors.New("truncation above head")
|
||||
}
|
||||
// Load the new tail index by the given new tail position
|
||||
@ -466,10 +463,10 @@ func (t *freezerTable) truncateTail(items uint64) error {
|
||||
newTailId uint32
|
||||
buffer = make([]byte, indexEntrySize)
|
||||
)
|
||||
if atomic.LoadUint64(&t.items) == items {
|
||||
if t.items.Load() == items {
|
||||
newTailId = t.headId
|
||||
} else {
|
||||
offset := items - atomic.LoadUint64(&t.itemOffset)
|
||||
offset := items - t.itemOffset.Load()
|
||||
if _, err := t.index.ReadAt(buffer, int64((offset+1)*indexEntrySize)); err != nil {
|
||||
return err
|
||||
}
|
||||
@ -478,7 +475,7 @@ func (t *freezerTable) truncateTail(items uint64) error {
|
||||
newTailId = newTail.filenum
|
||||
}
|
||||
// Update the virtual tail marker and hidden these entries in table.
|
||||
atomic.StoreUint64(&t.itemHidden, items)
|
||||
t.itemHidden.Store(items)
|
||||
if err := writeMetadata(t.meta, newMetadata(items)); err != nil {
|
||||
return err
|
||||
}
|
||||
@ -501,7 +498,7 @@ func (t *freezerTable) truncateTail(items uint64) error {
|
||||
// Count how many items can be deleted from the file.
|
||||
var (
|
||||
newDeleted = items
|
||||
deleted = atomic.LoadUint64(&t.itemOffset)
|
||||
deleted = t.itemOffset.Load()
|
||||
)
|
||||
for current := items - 1; current >= deleted; current -= 1 {
|
||||
if _, err := t.index.ReadAt(buffer, int64((current-deleted+1)*indexEntrySize)); err != nil {
|
||||
@ -541,7 +538,7 @@ func (t *freezerTable) truncateTail(items uint64) error {
|
||||
}
|
||||
// Release any files before the current tail
|
||||
t.tailId = newTailId
|
||||
atomic.StoreUint64(&t.itemOffset, newDeleted)
|
||||
t.itemOffset.Store(newDeleted)
|
||||
t.releaseFilesBefore(t.tailId, true)
|
||||
|
||||
// Retrieve the new size and update the total size counter
|
||||
@ -654,7 +651,7 @@ func (t *freezerTable) releaseFilesBefore(num uint32, remove bool) {
|
||||
// it will return error.
|
||||
func (t *freezerTable) getIndices(from, count uint64) ([]*indexEntry, error) {
|
||||
// Apply the table-offset
|
||||
from = from - t.itemOffset
|
||||
from = from - t.itemOffset.Load()
|
||||
// For reading N items, we need N+1 indices.
|
||||
buffer := make([]byte, (count+1)*indexEntrySize)
|
||||
if _, err := t.index.ReadAt(buffer, int64(from*indexEntrySize)); err != nil {
|
||||
@ -744,8 +741,8 @@ func (t *freezerTable) retrieveItems(start, count, maxBytes uint64) ([]byte, []i
|
||||
return nil, nil, errClosed
|
||||
}
|
||||
var (
|
||||
items = atomic.LoadUint64(&t.items) // the total items(head + 1)
|
||||
hidden = atomic.LoadUint64(&t.itemHidden) // the number of hidden items
|
||||
items = t.items.Load() // the total items(head + 1)
|
||||
hidden = t.itemHidden.Load() // the number of hidden items
|
||||
)
|
||||
// Ensure the start is written, not deleted from the tail, and that the
|
||||
// caller actually wants something
|
||||
@ -832,7 +829,7 @@ func (t *freezerTable) retrieveItems(start, count, maxBytes uint64) ([]byte, []i
|
||||
// has returns an indicator whether the specified number data is still accessible
|
||||
// in the freezer table.
|
||||
func (t *freezerTable) has(number uint64) bool {
|
||||
return atomic.LoadUint64(&t.items) > number && atomic.LoadUint64(&t.itemHidden) <= number
|
||||
return t.items.Load() > number && t.itemHidden.Load() <= number
|
||||
}
|
||||
|
||||
// size returns the total data size in the freezer table.
|
||||
@ -922,7 +919,7 @@ func (t *freezerTable) dumpIndex(w io.Writer, start, stop int64) {
|
||||
return
|
||||
}
|
||||
fmt.Fprintf(w, "Version %d count %d, deleted %d, hidden %d\n", meta.Version,
|
||||
atomic.LoadUint64(&t.items), atomic.LoadUint64(&t.itemOffset), atomic.LoadUint64(&t.itemHidden))
|
||||
t.items.Load(), t.itemOffset.Load(), t.itemHidden.Load())
|
||||
|
||||
buf := make([]byte, indexEntrySize)
|
||||
|
||||
|
@ -24,7 +24,6 @@ import (
|
||||
"os"
|
||||
"path/filepath"
|
||||
"reflect"
|
||||
"sync/atomic"
|
||||
"testing"
|
||||
"testing/quick"
|
||||
|
||||
@ -191,7 +190,7 @@ func TestFreezerRepairDanglingHeadLarge(t *testing.T) {
|
||||
writeChunks(t, f, 255, 15)
|
||||
|
||||
// The last item should be there
|
||||
if _, err = f.Retrieve(f.items - 1); err != nil {
|
||||
if _, err = f.Retrieve(f.items.Load() - 1); err != nil {
|
||||
t.Fatal(err)
|
||||
}
|
||||
f.Close()
|
||||
@ -317,7 +316,7 @@ func TestFreezerRepairDanglingIndex(t *testing.T) {
|
||||
writeChunks(t, f, 9, 15)
|
||||
|
||||
// The last item should be there
|
||||
if _, err = f.Retrieve(f.items - 1); err != nil {
|
||||
if _, err = f.Retrieve(f.items.Load() - 1); err != nil {
|
||||
f.Close()
|
||||
t.Fatal(err)
|
||||
}
|
||||
@ -350,8 +349,8 @@ func TestFreezerRepairDanglingIndex(t *testing.T) {
|
||||
t.Fatal(err)
|
||||
}
|
||||
defer f.Close()
|
||||
if f.items != 7 {
|
||||
t.Fatalf("expected %d items, got %d", 7, f.items)
|
||||
if f.items.Load() != 7 {
|
||||
t.Fatalf("expected %d items, got %d", 7, f.items.Load())
|
||||
}
|
||||
if err := assertFileSize(fileToCrop, 15); err != nil {
|
||||
t.Fatal(err)
|
||||
@ -374,7 +373,7 @@ func TestFreezerTruncate(t *testing.T) {
|
||||
writeChunks(t, f, 30, 15)
|
||||
|
||||
// The last item should be there
|
||||
if _, err = f.Retrieve(f.items - 1); err != nil {
|
||||
if _, err = f.Retrieve(f.items.Load() - 1); err != nil {
|
||||
t.Fatal(err)
|
||||
}
|
||||
f.Close()
|
||||
@ -388,8 +387,8 @@ func TestFreezerTruncate(t *testing.T) {
|
||||
}
|
||||
defer f.Close()
|
||||
f.truncateHead(10) // 150 bytes
|
||||
if f.items != 10 {
|
||||
t.Fatalf("expected %d items, got %d", 10, f.items)
|
||||
if f.items.Load() != 10 {
|
||||
t.Fatalf("expected %d items, got %d", 10, f.items.Load())
|
||||
}
|
||||
// 45, 45, 45, 15 -- bytes should be 15
|
||||
if f.headBytes != 15 {
|
||||
@ -444,9 +443,9 @@ func TestFreezerRepairFirstFile(t *testing.T) {
|
||||
if err != nil {
|
||||
t.Fatal(err)
|
||||
}
|
||||
if f.items != 1 {
|
||||
if f.items.Load() != 1 {
|
||||
f.Close()
|
||||
t.Fatalf("expected %d items, got %d", 0, f.items)
|
||||
t.Fatalf("expected %d items, got %d", 0, f.items.Load())
|
||||
}
|
||||
|
||||
// Write 40 bytes
|
||||
@ -483,7 +482,7 @@ func TestFreezerReadAndTruncate(t *testing.T) {
|
||||
writeChunks(t, f, 30, 15)
|
||||
|
||||
// The last item should be there
|
||||
if _, err = f.Retrieve(f.items - 1); err != nil {
|
||||
if _, err = f.Retrieve(f.items.Load() - 1); err != nil {
|
||||
t.Fatal(err)
|
||||
}
|
||||
f.Close()
|
||||
@ -495,9 +494,9 @@ func TestFreezerReadAndTruncate(t *testing.T) {
|
||||
if err != nil {
|
||||
t.Fatal(err)
|
||||
}
|
||||
if f.items != 30 {
|
||||
if f.items.Load() != 30 {
|
||||
f.Close()
|
||||
t.Fatalf("expected %d items, got %d", 0, f.items)
|
||||
t.Fatalf("expected %d items, got %d", 0, f.items.Load())
|
||||
}
|
||||
for y := byte(0); y < 30; y++ {
|
||||
f.Retrieve(uint64(y))
|
||||
@ -1210,13 +1209,13 @@ func runRandTest(rt randTest) bool {
|
||||
rt[i].err = fmt.Errorf("failed to reload table %v", err)
|
||||
}
|
||||
case opCheckAll:
|
||||
tail := atomic.LoadUint64(&f.itemHidden)
|
||||
head := atomic.LoadUint64(&f.items)
|
||||
tail := f.itemHidden.Load()
|
||||
head := f.items.Load()
|
||||
|
||||
if tail == head {
|
||||
continue
|
||||
}
|
||||
got, err := f.RetrieveItems(atomic.LoadUint64(&f.itemHidden), head-tail, 100000)
|
||||
got, err := f.RetrieveItems(f.itemHidden.Load(), head-tail, 100000)
|
||||
if err != nil {
|
||||
rt[i].err = err
|
||||
} else {
|
||||
@ -1238,7 +1237,7 @@ func runRandTest(rt randTest) bool {
|
||||
if len(step.items) == 0 {
|
||||
continue
|
||||
}
|
||||
tail := atomic.LoadUint64(&f.itemHidden)
|
||||
tail := f.itemHidden.Load()
|
||||
for i := 0; i < len(step.items); i++ {
|
||||
blobs = append(blobs, values[step.items[i]-tail])
|
||||
}
|
||||
@ -1254,7 +1253,7 @@ func runRandTest(rt randTest) bool {
|
||||
case opTruncateHead:
|
||||
f.truncateHead(step.target)
|
||||
|
||||
length := atomic.LoadUint64(&f.items) - atomic.LoadUint64(&f.itemHidden)
|
||||
length := f.items.Load() - f.itemHidden.Load()
|
||||
values = values[:length]
|
||||
|
||||
case opTruncateHeadAll:
|
||||
@ -1262,10 +1261,10 @@ func runRandTest(rt randTest) bool {
|
||||
values = nil
|
||||
|
||||
case opTruncateTail:
|
||||
prev := atomic.LoadUint64(&f.itemHidden)
|
||||
prev := f.itemHidden.Load()
|
||||
f.truncateTail(step.target)
|
||||
|
||||
truncated := atomic.LoadUint64(&f.itemHidden) - prev
|
||||
truncated := f.itemHidden.Load() - prev
|
||||
values = values[truncated:]
|
||||
|
||||
case opTruncateTailAll:
|
||||
|
@ -267,10 +267,10 @@ func TestFreezerReadonlyValidate(t *testing.T) {
|
||||
bBatch := f.tables["b"].newBatch()
|
||||
require.NoError(t, bBatch.AppendRaw(0, item))
|
||||
require.NoError(t, bBatch.commit())
|
||||
if f.tables["a"].items != 3 {
|
||||
if f.tables["a"].items.Load() != 3 {
|
||||
t.Fatalf("unexpected number of items in table")
|
||||
}
|
||||
if f.tables["b"].items != 1 {
|
||||
if f.tables["b"].items.Load() != 1 {
|
||||
t.Fatalf("unexpected number of items in table")
|
||||
}
|
||||
require.NoError(t, f.Close())
|
||||
|
Loading…
Reference in New Issue
Block a user