This is an automated cherry-pick of #19082

What is changed and how it works?

Issue Number: Close #19070

What's Changed:

1. Hibernate condition optimization: leader no need wait for all peers' hibernate votes if any non-voter peer is down and votes reach majority.  
2. If non hibernate vote peers are all unreachable, leader can skip broadcasting hibernate request. This can reduce the number of messages before down peer is detected (10min by default).  

As noted in #19070, when a TiKV peer becomes unavailable, region leaders on other nodes are prevented from entering hibernation due to the current condition logic—which requires the leader to receive votes from all peers before hibernating. This requirement is overly strict. In practice, the leader only needs to confirm that it has received votes from all alive peers(referencing the down_peers field) and that a majority has been reached for raft voters.

Before Modification

After Modification

Related changes

• PR to update pingcap/docs/pingcap/docs-cn:
• 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

Side effects

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

Release note

Hibernate condition optimization: leader no need wait for all peers' hibernate votes if any non-voter peer is down and votes reach majority.