[core] Fixed bug: tsbpd might miss m_bClosing flag setting, flag not always properly set.

[ethouris]

Fixed:

• Properly set the flag before calling releaseSync
• TSBPD: check the flag immediately after re-locking m_RecvLock

Added some sanity checks

[maxsharabayko]

The scope of this variable can be reduced. It is used both in the if and else statements, but independently. Thus can be defined inside those.

[ethouris]

You are right, this was likely tied to some earlier version that provided some log instruction after this if-branch, hence the scope was bigger. Now it has its own log in every branch, so this isn't necessary.

[codecov]

Codecov Report

Attention: Patch coverage is 66.66667% with 4 lines in your changes missing coverage. Please review.

Project coverage is 68.63%. Comparing base (5f16494) to head (83fcba8).
Report is 6 commits behind head on master.

Files with missing lines	Patch %	Lines
srtcore/core.cpp	66.66%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1709      +/-   ##
==========================================
+ Coverage   65.75%   68.63%   +2.87%     
==========================================
  Files         101      101              
  Lines       17966    18327     +361     
==========================================
+ Hits        11814    12579     +765     
+ Misses       6152     5748     -404     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[maxsharabayko]

Tested on a Raspberry Pi3 alike platform.
Hard to tell for sure from a single run, but might be a small performance degradation of the TSBPD thread.

This PR

SRT v1.5.4 master 2024-12-12 f109fb1

[maxsharabayko]

Second run.

This PR.

Master 2024-12-12 f109fb1

[maxsharabayko]

Third run

This PR.

Master 2024-12-12 f109fb1

[ethouris]

Would you be able to check also a similar version, just in place where there's a wait on tsbpd_cc just replace the timely waiting with forever waiting, while assigning true to the boolean variable?

Hard to believe that this is the point of performance degradation, but there are about 3 candidates among the changes that could cause it. I also hope that you made the measurements on C++11 version.