Get rid of dual mutex in blockchain.go

[hqjang-pepper]

Proposed changes

• This PR removes chainmu in blockchain.go.
• The mu supposed to protect internal state from getting corrupted whilst chaimu from multiple modifications to the chain.
• I found that functions such as Rollback or SetHead may cause race condition problems with chain insertion, since they use different mutex but can be called at the same time.

Related Ethereum PR : https://github.com/ethereum/go-ethereum/pull/18436/files

Types of changes

Please put an x in the boxes related to your change.

• Bugfix
• New feature or enhancement
• Others

Checklist

Put an x in the boxes that apply. You can also fill these out after creating the PR. If you're unsure about any of them, don't hesitate to ask. We're here to help! This is simply a reminder of what we are going to look for before merging your code.

• I have read the CONTRIBUTING GUIDELINES doc
• I have signed the CLA
• Lint and unit tests pass locally with my changes ($ make test)
• I have added tests that prove my fix is effective or that my feature works
• I have added necessary documentation (if appropriate)
• Any dependent changes have been merged and published in downstream modules

Related issues

• Please leave the issue numbers or links related to this PR here.

Further comments

If this is a relatively large or complex change, kick off the discussion by explaining why you chose the solution you did and what alternatives you considered, etc...

[jeongkyun-oh]

@hqjang-pepper Could you please check why the test cases are failed.

[hqjang-pepper]

@hqjang-pepper Could you please check why the test cases are failed.

Following check-fails are not directly from this PR changes.
Since I had no problems in local tests, so it's hard to reason with..
Gotta need some time to figure it out to deal with.

[ehnuje]

LGTM, however, have you found the root cause of timeout issue on CircleCI test?

[hqjang-pepper]

LGTM, however, have you found the root cause of timeout issue on CircleCI test?

I missed removing mutex at WriteBlockWithState.
Without removing, it causes race condition when InsertChain is called.
Changes were applied at this following PR. e93aae4

[ehnuje]

LGTM, however, have you found the root cause of timeout issue on CircleCI test?

I missed removing mutex at WriteBlockWithState. Without removing, it causes race condition when InsertChain is called. Changes were applied at this following PR. e93aae4

@hqjang-pepper If so, why didn't it make any problem on the local test? Due to timing issue?

[hqjang-pepper]

LGTM, however, have you found the root cause of timeout issue on CircleCI test?

I missed removing mutex at WriteBlockWithState. Without removing, it causes race condition when InsertChain is called. Changes were applied at this following PR. e93aae4

@hqjang-pepper If so, why didn't it make any problem on the local test? Due to timing issue?

@ehnuje Sorry for late response.
There was no timeout issue when I tested through make test, which was different with results from CircleCi. I'm not sure that the reason no timeout failure was timing issue. But after I checked again, timeout issue also occurred in my local environment, and I found the cause by debugging.

[kjhman21]

@hqjang-pepper It requires heavy testing especially changing calling locations of locks. How did you conduct testing?

[hqjang-pepper]

I conducted load test on two version, before/after mutex integration.
As a result of sending enough requests(more than 1m), the TPS difference between the two versions was 1~10.
It's not that big difference, so we decided to merge this PR.