*: update client-go (tidb-6.1)

[sticnarf]

What problem does this PR solve?

Issue Number: close #37141

Problem Summary:

Our region cache B-Tree uses the start key of the region as the entry key. So, insertRegionToCache only replaces the region with the same start key:
https://github.com/tikv/client-go/blob/516cfcdecce865d90ec05dd1cc1de060c35438dc/internal/locate/region_cache.go#L1287

Considering there's a table with lots of regions, the user removes all data from the table. So, PD will finally merge all regions into one. In our region cache implementation, only the first region (containing the table head) is updated, while all other regions still exist in the cache.

When we locate a key near the end of the table, DescendLessOrEqual will iterate over all regions that are left in the cache. This will waste a lot of CPU and make region search really slow, and it is also possible to cause lots of unnecessary PD loadRegion RPCs.

To solve the problem, when we do insertRegionToCache, we should iterate the B-Tree for all regions that intersect with the inserted region and remove them.

What is changed and how it works?

It includes tikv/client-go#566, so when new region info is inserted to the cache, old invalid region cache that will be cleared.

This also includes tikv/client-go#585, which covers more cases like #26832 to reduce the risk of data race of SelectLock executor under other parallel executors.

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No code

Side effects

• Performance regression: Consumes more CPU
• Performance regression: Consumes more Memory
• Breaking backward compatibility

Documentation

• Affects user behaviors
• Contains syntax changes
• Contains variable changes
• Contains experimental features
• Changes MySQL compatibility

Release note

Please refer to Release Notes Language Style Guide to write a quality release note.

- Optimize the performance of the region cache after merging many regions.

- Reduce the risk of data race of SelectLock under parallel executors 

[ti-chi-bot]

[REVIEW NOTIFICATION]

This pull request has been approved by:

• cfzjywxk
• jackysp

To complete the pull request process, please ask the reviewers in the list to review by filling /cc @reviewer in the comment.
After your PR has acquired the required number of LGTMs, you can assign this pull request to the committer in the list by filling /assign @committer in the comment to help you merge this pull request.

The full list of commands accepted by this bot can be found here.

Reviewer can indicate their review by submitting an approval review.
Reviewer can cancel approval by submitting a request changes review.

[jackysp]

LGTM

[jackysp]

ah, maybe need go mod tidy?

[sticnarf]

ah, maybe need go mod tidy?

Oops. Fixed.

[sticnarf]

/merge

[ti-chi-bot]

This pull request has been accepted and is ready to merge.

Commit hash: 685d796631edea87e658a6695d4ad07280c801f1

[sticnarf]

/run-all-tests

[sticnarf]

/hold

[sticnarf]

/merge

[ti-chi-bot]

This pull request has been accepted and is ready to merge.

Commit hash: 8556719

[sticnarf]

/unhold