potential data race on add index when other routine exit due to error but adjust worker size routine keeps running

[D3Hunter]

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

the task sender routine A uses error group context

tidb/pkg/ddl/backfilling.go

Lines 1039 to 1040 in 889bf45

	// generate task goroutine
	eg.Go(func() error {

but the adjust worker size routine B uses parent ctx

tidb/pkg/ddl/backfilling.go

Lines 1074 to 1081 in 889bf45

	// update the worker cnt goroutine
	go func() {
	ticker := time.NewTicker(UpdateDDLJobReorgCfgInterval)
	defer ticker.Stop()
	for {
	select {
	case <-ctx.Done():
	return

so if A failed with error, cause error group context cancelled, B keeps running, so we might met potential data race in

tidb/pkg/ddl/backfilling_scheduler.go

Line 342 in 889bf45

closeBackfillWorkers(b.workers)

and

tidb/pkg/ddl/backfilling_scheduler.go

Line 322 in 889bf45

b.workers = append(b.workers, runner)

we should use the same context, and run together under the same group. introduced in #57468

2. What did you expect to see? (Required)

3. What did you see instead (Required)

4. What is your TiDB version? (Required)

master