Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

executor: fix hashjoin goleak #39023

Merged
merged 3 commits into from
Nov 10, 2022
Merged
Show file tree
Hide file tree
Changes from 2 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
48 changes: 22 additions & 26 deletions executor/join.go
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,6 @@ import (
"fmt"
"runtime/trace"
"strconv"
"sync"
"sync/atomic"
"time"

Expand Down Expand Up @@ -70,6 +69,8 @@ type HashJoinExec struct {

// closeCh add a lock for closing executor.
closeCh chan struct{}
worker util.WaitGroupWrapper
waiter util.WaitGroupWrapper
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

waitJoinWorkersAndCloseResultChan will call worker.Wait(), but it is running in another goroutine. So waiter is to prevent leak of waitJoinWorkersAndCloseResultChan.

joinType plannercore.JoinType
requiredRows int64

Expand All @@ -92,9 +93,7 @@ type HashJoinExec struct {
prepared bool
isOuterJoin bool

// joinWorkerWaitGroup is for sync multiple join workers.
joinWorkerWaitGroup sync.WaitGroup
finished atomic.Value
finished atomic.Bool

stats *hashJoinRuntimeStats

Expand Down Expand Up @@ -154,6 +153,7 @@ func (e *HashJoinExec) Close() error {
e.probeChkResourceCh = nil
e.joinChkResourceCh = nil
terror.Call(e.rowContainer.Close)
e.waiter.Wait()
}
e.outerMatchedStatus = e.outerMatchedStatus[:0]
e.buildSideRows = nil
Expand Down Expand Up @@ -181,9 +181,10 @@ func (e *HashJoinExec) Open(ctx context.Context) error {
e.diskTracker = disk.NewTracker(e.id, -1)
e.diskTracker.AttachTo(e.ctx.GetSessionVars().StmtCtx.DiskTracker)

e.worker = util.WaitGroupWrapper{}
e.waiter = util.WaitGroupWrapper{}
e.closeCh = make(chan struct{})
e.finished.Store(false)
e.joinWorkerWaitGroup = sync.WaitGroup{}

if e.probeTypes == nil {
e.probeTypes = retTypes(e.probeSideExec)
Expand All @@ -205,7 +206,7 @@ func (e *HashJoinExec) Open(ctx context.Context) error {
func (e *HashJoinExec) fetchProbeSideChunks(ctx context.Context) {
hasWaitedForBuild := false
for {
if e.finished.Load().(bool) {
if e.finished.Load() {
return
}

Expand Down Expand Up @@ -283,24 +284,24 @@ func (e *HashJoinExec) wait4BuildSide() (emptyBuild bool, err error) {

// fetchBuildSideRows fetches all rows from build side executor, and append them
// to e.buildSideResult.
func (e *HashJoinExec) fetchBuildSideRows(ctx context.Context, chkCh chan<- *chunk.Chunk, doneCh <-chan struct{}) {
func (e *HashJoinExec) fetchBuildSideRows(ctx context.Context, chkCh chan<- *chunk.Chunk, errCh chan<- error, doneCh <-chan struct{}) {
defer close(chkCh)
var err error
failpoint.Inject("issue30289", func(val failpoint.Value) {
if val.(bool) {
err = errors.Errorf("issue30289 build return error")
e.buildFinished <- errors.Trace(err)
errCh <- errors.Trace(err)
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

fetchBuildSideRows and fetchAndBuildHashTable are running in parallel. So fetchAndBuildHashTable may close buildFinished before fetchBuildSideRows is finised.

return
}
})
for {
if e.finished.Load().(bool) {
if e.finished.Load() {
return
}
chk := e.ctx.GetSessionVars().GetNewChunk(e.buildSideExec.base().retFieldTypes, e.ctx.GetSessionVars().MaxChunkSize)
err = Next(ctx, e.buildSideExec, chk)
if err != nil {
e.buildFinished <- errors.Trace(err)
errCh <- errors.Trace(err)
return
}
failpoint.Inject("errorFetchBuildSideRowsMockOOMPanic", nil)
Expand Down Expand Up @@ -357,8 +358,7 @@ func (e *HashJoinExec) initializeForProbe() {

func (e *HashJoinExec) fetchAndProbeHashTable(ctx context.Context) {
e.initializeForProbe()
e.joinWorkerWaitGroup.Add(1)
go util.WithRecovery(func() {
e.worker.RunWithRecover(func() {
defer trace.StartRegion(ctx, "HashJoinProbeSideFetcher").End()
e.fetchProbeSideChunks(ctx)
}, e.handleProbeSideFetcherPanic)
Expand All @@ -373,14 +373,13 @@ func (e *HashJoinExec) fetchAndProbeHashTable(ctx context.Context) {
}

for i := uint(0); i < e.concurrency; i++ {
e.joinWorkerWaitGroup.Add(1)
workID := i
go util.WithRecovery(func() {
e.worker.RunWithRecover(func() {
defer trace.StartRegion(ctx, "HashJoinWorker").End()
e.runJoinWorker(workID, probeKeyColIdx, probeNAKeColIdx)
}, e.handleJoinWorkerPanic)
}
go util.WithRecovery(e.waitJoinWorkersAndCloseResultChan, nil)
e.waiter.RunWithRecover(e.waitJoinWorkersAndCloseResultChan, nil)
}

func (e *HashJoinExec) handleProbeSideFetcherPanic(r interface{}) {
Expand All @@ -390,14 +389,12 @@ func (e *HashJoinExec) handleProbeSideFetcherPanic(r interface{}) {
if r != nil {
e.joinResultCh <- &hashjoinWorkerResult{err: errors.Errorf("%v", r)}
}
e.joinWorkerWaitGroup.Done()
}

func (e *HashJoinExec) handleJoinWorkerPanic(r interface{}) {
if r != nil {
e.joinResultCh <- &hashjoinWorkerResult{err: errors.Errorf("%v", r)}
}
e.joinWorkerWaitGroup.Done()
}

// Concurrently handling unmatched rows from the hash table
Expand Down Expand Up @@ -437,15 +434,14 @@ func (e *HashJoinExec) handleUnmatchedRowsFromHashTable(workerID uint) {
}

func (e *HashJoinExec) waitJoinWorkersAndCloseResultChan() {
e.joinWorkerWaitGroup.Wait()
e.worker.Wait()
if e.useOuterToBuild {
// Concurrently handling unmatched rows from the hash table at the tail
for i := uint(0); i < e.concurrency; i++ {
var workerID = i
e.joinWorkerWaitGroup.Add(1)
go util.WithRecovery(func() { e.handleUnmatchedRowsFromHashTable(workerID) }, e.handleJoinWorkerPanic)
e.worker.RunWithRecover(func() { e.handleUnmatchedRowsFromHashTable(workerID) }, e.handleJoinWorkerPanic)
}
e.joinWorkerWaitGroup.Wait()
e.worker.Wait()
}
close(e.joinResultCh)
}
Expand Down Expand Up @@ -481,7 +477,7 @@ func (e *HashJoinExec) runJoinWorker(workerID uint, probeKeyColIdx, probeNAKeyCo
naKeyColIdx: probeNAKeyColIdx,
}
for ok := true; ok; {
if e.finished.Load().(bool) {
if e.finished.Load() {
break
}
select {
Expand Down Expand Up @@ -1121,7 +1117,7 @@ func (e *HashJoinExec) Next(ctx context.Context, req *chunk.Chunk) (err error) {
for i := uint(0); i < e.concurrency; i++ {
e.rowIters = append(e.rowIters, chunk.NewIterator4Slice([]chunk.Row{}).(*chunk.Iterator4Slice))
}
go util.WithRecovery(func() {
e.worker.RunWithRecover(func() {
defer trace.StartRegion(ctx, "HashJoinHashTableBuilder").End()
e.fetchAndBuildHashTable(ctx)
}, e.handleFetchAndBuildHashTablePanic)
Expand Down Expand Up @@ -1164,10 +1160,10 @@ func (e *HashJoinExec) fetchAndBuildHashTable(ctx context.Context) {
buildSideResultCh := make(chan *chunk.Chunk, 1)
doneCh := make(chan struct{})
fetchBuildSideRowsOk := make(chan error, 1)
go util.WithRecovery(
e.worker.RunWithRecover(
func() {
defer trace.StartRegion(ctx, "HashJoinBuildSideFetcher").End()
e.fetchBuildSideRows(ctx, buildSideResultCh, doneCh)
e.fetchBuildSideRows(ctx, buildSideResultCh, fetchBuildSideRowsOk, doneCh)
},
func(r interface{}) {
if r != nil {
Expand Down Expand Up @@ -1214,7 +1210,7 @@ func (e *HashJoinExec) buildHashTableForList(buildSideResultCh <-chan *chunk.Chu
e.ctx.GetSessionVars().MemTracker.FallbackOldAndSetNewAction(actionSpill)
}
for chk := range buildSideResultCh {
if e.finished.Load().(bool) {
if e.finished.Load() {
return nil
}
if !e.useOuterToBuild {
Expand Down
2 changes: 1 addition & 1 deletion util/wait_group_wrapper.go
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ func (w *WaitGroupWrapper) RunWithRecover(exec func(), recoverFn func(r interfac
go func() {
defer func() {
r := recover()
if r != nil && recoverFn != nil {
if recoverFn != nil {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

util.WithRecovery will always call recoveryFn, align with it and hashjoin relies on this behavior.

recoverFn(r)
}
w.Done()
Expand Down