IndexNestedLoopHashJoin send on closed channel

[ichn-hu]

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

the reproduce can be done by #31128

checkout the code and run

make failpoint-enable
cd executor
go test -check.f TestIndexNestedLoopHashJoin > test.result

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

no panic

3. What did you see instead (Required)

tidb/executor/test.result

Line 261 in a5bbee9

[2021/12/29 16:32:57.687 +08:00] [ERROR] [misc.go:94] ["panic in the recoverable goroutine"] [r="\"send on closed channel\""] ["stack trace"="github.com/pingcap/tidb/util.WithRecovery.func1\n\t/home/ichn/Projects/pingcap/tidb/util/misc.go:96\nruntime.gopanic\n\t/home/ichn/.gvm/gos/go1.16/src/runtime/panic.go:965\nruntime.chansend\n\t/home/ichn/.gvm/gos/go1.16/src/runtime/chan.go:204\nruntime.chansend1\n\t/home/ichn/.gvm/gos/go1.16/src/runtime/chan.go:143\ngithub.com/pingcap/tidb/executor.(*indexHashJoinInnerWorker).run\n\t/home/ichn/Projects/pingcap/tidb/executor/index_lookup_hash_join.go:503\ngithub.com/pingcap/tidb/executor.(*IndexNestedLoopHashJoin).startWorkers.func2\n\t/home/ichn/Projects/pingcap/tidb/executor/index_lookup_hash_join.go:196\ngithub.com/pingcap/tidb/util.WithRecovery\n\t/home/ichn/Projects/pingcap/tidb/util/misc.go:99"]

the reason for this panic is because if the result needs to be ordered and it got any error during doJoinInOrder

the resultCh will be closed

however in indexHashJoinInnerWorker

the returned error will be sent to the resultCh which is already closed, causing the panic.

4. What is your TiDB version? (Required)

master

[yudongusa]

@XuHuaiyu assign this to you for now.

[github-actions]

Please check whether the issue should be labeled with 'affects-x.y' or 'fixes-x.y.z', and then remove 'needs-more-info' label.