From 72ad40698be9d7dbaa2d5b32f1fb6540d6f9df72 Mon Sep 17 00:00:00 2001
From: Bartosz Taudul <wolf@nereid.pl>
Date: Tue, 16 Aug 2022 13:55:46 +0200
Subject: [PATCH] Move initialization of callstack structs to a thread.

Initializing structures for callstack processing (building memory map of the
process, gathering kernel symbols, etc) takes some time, which in some cases
may be significant.

Callstack queries are now handled on a separate thread. In such setup it no
longer makes sense to block main thread execution with this lengthy init
process.

All the heavy initialization phase has been now moved to this separate
processing thread. Some initial callstack queries may now not produce
responses as promptly as before, but this is only because the main thread is
able to start working earlier.

Some parts of the initialization process may be critical to do in the main
thread, for example because the function responsible for gathering callstacks
must be loaded first. This is done still on the main thread, in a new function
InitCallstackCritical().
---
 public/client/TracyCallstack.cpp | 14 +++++++++++++-
 public/client/TracyCallstack.hpp |  1 +
 public/client/TracyProfiler.cpp  |  6 +++---
 3 files changed, 17 insertions(+), 4 deletions(-)

diff --git a/public/client/TracyCallstack.cpp b/public/client/TracyCallstack.cpp
index 0108deac..6560de33 100644
--- a/public/client/TracyCallstack.cpp
+++ b/public/client/TracyCallstack.cpp
@@ -138,9 +138,13 @@ KernelDriver* s_krnlCache = nullptr;
 size_t s_krnlCacheCnt;
 
 
-void InitCallstack()
+void InitCallstackCritical()
 {
     ___tracy_RtlWalkFrameChain = (___tracy_t_RtlWalkFrameChain)GetProcAddress( GetModuleHandleA( "ntdll.dll" ), "RtlWalkFrameChain" );
+}
+
+void InitCallstack()
+{
     _SymAddrIncludeInlineTrace = (t_SymAddrIncludeInlineTrace)GetProcAddress( GetModuleHandleA( "dbghelp.dll" ), "SymAddrIncludeInlineTrace" );
     _SymQueryInlineTrace = (t_SymQueryInlineTrace)GetProcAddress( GetModuleHandleA( "dbghelp.dll" ), "SymQueryInlineTrace" );
     _SymFromInlineContext = (t_SymFromInlineContext)GetProcAddress( GetModuleHandleA( "dbghelp.dll" ), "SymFromInlineContext" );
@@ -695,6 +699,10 @@ static void InitKernelSymbols()
 }
 #endif
 
+void InitCallstackCritical()
+{
+}
+
 void InitCallstack()
 {
     cb_bts = backtrace_create_state( nullptr, 0, nullptr, nullptr );
@@ -1028,6 +1036,10 @@ CallstackEntryData DecodeCallstackPtr( uint64_t ptr )
 
 #elif TRACY_HAS_CALLSTACK == 5
 
+void InitCallstackCritical()
+{
+}
+
 void InitCallstack()
 {
     ___tracy_init_demangle_buffer();
diff --git a/public/client/TracyCallstack.hpp b/public/client/TracyCallstack.hpp
index 117d8044..bc50561c 100644
--- a/public/client/TracyCallstack.hpp
+++ b/public/client/TracyCallstack.hpp
@@ -55,6 +55,7 @@ CallstackSymbolData DecodeCodeAddress( uint64_t ptr );
 const char* DecodeCallstackPtrFast( uint64_t ptr );
 CallstackEntryData DecodeCallstackPtr( uint64_t ptr );
 void InitCallstack();
+void InitCallstackCritical();
 void EndCallstack();
 const char* GetKernelModulePath( uint64_t addr );
 
diff --git a/public/client/TracyProfiler.cpp b/public/client/TracyProfiler.cpp
index 658f0b39..f48f96e2 100644
--- a/public/client/TracyProfiler.cpp
+++ b/public/client/TracyProfiler.cpp
@@ -1444,7 +1444,7 @@ void Profiler::SpawnWorkerThreads()
 #endif
 
 #ifdef TRACY_HAS_CALLSTACK
-    InitCallstack();
+    InitCallstackCritical();
 #endif
 
     m_timeBegin.store( GetTime(), std::memory_order_relaxed );
@@ -3296,11 +3296,11 @@ void Profiler::SymbolWorker()
 {
     ThreadExitHandler threadExitHandler;
     SetThreadName( "Tracy Symbol Worker" );
-    while( m_timeBegin.load( std::memory_order_relaxed ) == 0 ) std::this_thread::sleep_for( std::chrono::milliseconds( 10 ) );
-
 #ifdef TRACY_USE_RPMALLOC
     rpmalloc_thread_initialize();
 #endif
+    InitCallstack();
+    while( m_timeBegin.load( std::memory_order_relaxed ) == 0 ) std::this_thread::sleep_for( std::chrono::milliseconds( 10 ) );
 
     for(;;)
     {