This also fixes a few memory issues found by valgrind in debug tools/Optick:
Fix memory leak in Optick thread storage on exit while thread is still in scope (e.g. MainThread)
Fix uninitialized variables in ImmediateMode (for debug tools) and Optick
This PR also simplifies / generalizes OPTICK_THREAD() instrumentation for RBDoom3BFG and fixes an Optick thread scope problem in idGameThread::Run(). I previously tested com_smp to enable/disable the thread instrumentation based on smp mode, but stupidly put this inside a conditional scope - something you shouldn't do with Optick. This change eliminates the need for this check and allows Optick to see the full scope for idGameThread::Run() when operating in smp mode, as well as all other <xxx>::Run() threads. Note that In non-smp mode, idGameThread::Run() executes on the main thread and calling OPTICK_THREAD() is not required or desired (e.g. prevents labeling of frame number in main thread).
Fixes #842
This also fixes a few memory issues found by valgrind in debug tools/Optick:
MainThread)This PR also simplifies / generalizes
OPTICK_THREAD()instrumentation for RBDoom3BFG and fixes an Optick thread scope problem inidGameThread::Run(). I previously testedcom_smpto enable/disable the thread instrumentation based on smp mode, but stupidly put this inside a conditional scope - something you shouldn't do with Optick. This change eliminates the need for this check and allows Optick to see the full scope foridGameThread::Run()when operating in smp mode, as well as all other<xxx>::Run()threads. Note that In non-smp mode,idGameThread::Run()executes on the main thread and callingOPTICK_THREAD()is not required or desired (e.g. prevents labeling of frame number in main thread).