*: optimize MemTracker object allocation in point get

[tiancaiamao]

What problem does this PR solve?

Optimize memory tracker object allocation.

After TxnInfo, TSSet, RuntimeStatsColl and so on, memory tracker allocation would be the next.

Problem Summary:

Improve point get performance for #25573

What's Changed:

How it Works:

Avoid memory tracker object allocation in the point get query.

Check List

Tests

• Manual test (add detailed scripts or steps below)

cd session
go test -test.run XXX -bench BenchmarkPreparedPointGet  -test.benchmem -test.benchtime 30s  -test.memprofile mem.out

Before vs After:

BenchmarkPreparedPointGet-16    	 3750292	      9442 ns/op	    7830 B/op	     101 allocs/op
BenchmarkPreparedPointGet-16    	 3915133	      9125 ns/op	    7926 B/op	      97 allocs/op

Release note

• No release note

[ti-chi-bot]

[REVIEW NOTIFICATION]

This pull request has been approved by:

• guo-shaoge
• tangenta

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.

[tiancaiamao]

/merge

[ti-chi-bot]

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

Commit hash: 0ae8a6e