Don't account for cell dep for MAX_ANCESTORS_COUNT

[chenyukang]

What problem does this PR solve?

We met a scenario that a cell dep may have many transactions as descendants, and we can not submit a new transaction to consume the dep cell.

Problem Summary:

What is changed and how it works?

cell dep should be handled specially when calculating the entry's ancestors_count.
In this PR, we add a direct_ancestors_count to store the directed parent-child relationship between transactions.

This is a short-term solution.

What's Changed:

Related changes

• PR to update owner/repo:
• Need to cherry-pick to the release branch

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No code ci-runs-only: [ quick_checks,linters ]

Side effects

• Performance regression
• Breaking backward compatibility

Release note

Title Only: Include only the PR title in the release note.

[eval-exec]

But I found:
https://github.com/nervosnetwork/ckb/blob/4851b307b018b6bc86d2f79632223c97364f53c6/resource/ckb.toml#L140-L139

the max_ancestors_count in default config is 25, but DEFAULT_MAX_ANCESTORS_COUNT is 125.

[chenyukang]

I think we can change the parameter in ckb.toml to 125 later.

[quake]

I think it's a comment error, we should change the default const value to 25 also.

[eval-exec]

Should we change const DEFAULT_MAX_ANCESTORS_COUNT: usize to 25 too?

[chenyukang]

it should be 125?

ckb/util/app-config/src/legacy/tx_pool.rs

Line 123 in b8ccf92

max_ancestors_count: cmp::max(DEFAULT_MAX_ANCESTORS_COUNT, max_ancestors_count),

25 is originally referred from bitcoin, we use 125.

[eval-exec]

I think it should be min instead of max?

 max_ancestors_count: cmp::min(DEFAULT_MAX_ANCESTORS_COUNT, max_ancestors_count), 

---

Maybe we should only set the max_ancestors_count to DEFAULT_MAX_ANCESTORS_COUNT when the user doesn't specify it?

[chenyukang]

no, it's max.
I think this commit is not complete, since we change from 25 to 125, we'd better also change the value in ckb.toml?
07d4db9

@zhangsoledad

[gpBlockchain]

LGMT