Fix period backup crash due to the dead lock of std::lock_guard (2nd Solution)

[donho]

The crash occurs because the thread terminates the task prematurely due to PostMessage's nature, so FileManager::backupCurrentBuffer() is run always by the main thread - it makes the line std::lock_guard<std::mutex> lock(backup_mutex); has the dead lock on the 2nd main thread's entry and cause the crash:
https://en.cppreference.com/w/cpp/thread/mutex/lock

Using SendMessage instead of PostMessage to ensure that thread keeps the mutex until entire operation is terminated.

This solution seems better than #14914 because for the following reasons:

1. The old solution (Fix period backup crash due to the dead lock of std::lock_guard #14914) has no lock at all: though it doesn't crash while the main thread enters the 2nd time into FileManager::backupCurrentBuffer() (with write method blocked), the goal of the lock is to prevent the race condition. Whereas this PR prevents the 2nd thread from entering into FileManager::backupCurrentBuffer(), until the 1st thread finish its task.
2. Keep using standard C++11 function.

Fix #14906

[donho]

@xomx
Please check this PR.

[donho]

@xomx
Since this PR has fixed the pontential crash #14906, I have to merge this PR for the long waiting new release.
Please feel free for any reviewing/suggestion of this PR for the future release.

[xomx]

@donho

Sorry for my unresponsiveness - many IRL stuff + I am ill.

Seems this is not the right solution, I am still able to crash it at the N++ exit from the same reason:

I cannot do much more now, but I will try ASAP.

[donho]

@xomx
Damn! Hope you'll get recovered soon!

If it's possible, could you just provide me the instructions to reproduce the crash please?
Otherwise, no problem - it may not fix the crash, it's not getting worse anyway.

Bon rétablissement !