summaryrefslogtreecommitdiff
path: root/zen/globals.h
diff options
context:
space:
mode:
authorB Stack <bgstack15@gmail.com>2020-09-01 00:24:17 +0000
committerB Stack <bgstack15@gmail.com>2020-09-01 00:24:17 +0000
commit5a3f52b016581a6a0cb4513614b6c620d365dde2 (patch)
treeacfdfb3e1046db87040477033fda0df76d92916a /zen/globals.h
parentMerge branch '11.0' into 'master' (diff)
parentadd upstream 11.1 (diff)
downloadFreeFileSync-11.1.tar.gz
FreeFileSync-11.1.tar.bz2
FreeFileSync-11.1.zip
Merge branch '11.1' into 'master'11.1
add upstream 11.1 See merge request opensource-tracking/FreeFileSync!25
Diffstat (limited to 'zen/globals.h')
-rw-r--r--zen/globals.h85
1 files changed, 45 insertions, 40 deletions
diff --git a/zen/globals.h b/zen/globals.h
index 876d2598..635909f7 100644
--- a/zen/globals.h
+++ b/zen/globals.h
@@ -10,19 +10,18 @@
#include <atomic>
#include <memory>
#include "scope_guard.h"
+#include "legacy_compiler.h"
namespace zen
{
-/*
-Solve static destruction order fiasco by providing shared ownership and serialized access to global variables
+/* Solve static destruction order fiasco by providing shared ownership and serialized access to global variables
-=> there may be accesses to "Global<T>::get()" during process shutdown e.g. _("") used by message in debug_minidump.cpp or by some detached thread assembling an error message!
-=> use trivially-destructible POD only!!!
-
-ATTENTION: function-static globals have the compiler generate "magic statics" == compiler-genenerated locking code which will crash or leak memory when accessed after global is "dead"
- => "solved" by FunStatGlobal, but we can't have "too many" of these... */
+ => there may be accesses to "Global<T>::get()" during process shutdown e.g. _("") used by message in debug_minidump.cpp or by some detached thread assembling an error message!
+ => use trivially-destructible POD only!!!
+ ATTENTION: function-static globals have the compiler generate "magic statics" == compiler-genenerated locking code which will crash or leak memory when accessed after global is "dead"
+ => "solved" by FunStatGlobal, but we can't have "too many" of these... */
class PodSpinMutex
{
public:
@@ -32,7 +31,7 @@ public:
bool isLocked();
private:
- std::atomic_flag flag_; /* => avoid potential contention with worker thread during Global<> construction!
+ std::atomic_flag flag_{}; /* => avoid potential contention with worker thread during Global<> construction!
- "For an atomic_flag with static storage duration, this guarantees static initialization:" => just what the doctor ordered!
- "[default initialization] initializes std::atomic_flag to clear state" - since C++20 =>
- "std::atomic_flag is [...] guaranteed to be lock-free"
@@ -40,21 +39,25 @@ private:
};
+#define GLOBAL_RUN_ONCE(X) \
+ struct ZEN_CONCAT(GlobalInitializer, __LINE__) \
+ { \
+ ZEN_CONCAT(GlobalInitializer, __LINE__)() { X; } \
+ } ZEN_CONCAT(globalInitializer, __LINE__)
+
+
template <class T>
class Global //don't use for function-scope statics!
{
public:
- Global()
+ consteval2 Global() {}; //demand static zero-initialization!
+
+ ~Global()
{
static_assert(std::is_trivially_destructible_v<Pod>, "this memory needs to live forever");
- assert(!pod_.spinLock.isLocked()); //we depend on static zero-initialization!
- assert(!pod_.inst); //
+ set(nullptr);
}
- explicit Global(std::unique_ptr<T>&& newInst) { set(std::move(newInst)); }
-
- ~Global() { set(nullptr); }
-
std::shared_ptr<T> get() //=> return std::shared_ptr to let instance life time be handled by caller (MT usage!)
{
pod_.spinLock.lock();
@@ -83,8 +86,8 @@ private:
struct Pod
{
PodSpinMutex spinLock; //rely entirely on static zero-initialization! => avoid potential contention with worker thread during Global<> construction!
- //serialize access; can't use std::mutex: has non-trival destructor
- std::shared_ptr<T>* inst; //= nullptr;
+ //serialize access: can't use std::mutex: has non-trival destructor
+ std::shared_ptr<T>* inst = nullptr;
} pod_;
};
@@ -94,9 +97,9 @@ private:
struct CleanUpEntry
{
using CleanUpFunction = void (*)(void* callbackData);
- CleanUpFunction cleanUpFun;
- void* callbackData;
- CleanUpEntry* prev;
+ CleanUpFunction cleanUpFun = nullptr;
+ void* callbackData = nullptr;
+ CleanUpEntry* prev = nullptr;
};
void registerGlobalForDestruction(CleanUpEntry& entry);
@@ -105,15 +108,31 @@ template <class T>
class FunStatGlobal
{
public:
- //No FunStatGlobal() or ~FunStatGlobal()!
+ consteval2 FunStatGlobal() {}; //demand static zero-initialization!
- std::shared_ptr<T> get()
+ //No ~FunStatGlobal()!
+
+ void initOnce(std::unique_ptr<T> (*getInitialValue)())
{
static_assert(std::is_trivially_destructible_v<FunStatGlobal>, "this class must not generate code for magic statics!");
pod_.spinLock.lock();
ZEN_ON_SCOPE_EXIT(pod_.spinLock.unlock());
+ if (!pod_.cleanUpEntry.cleanUpFun)
+ {
+ assert(!pod_.inst);
+ if (std::unique_ptr<T> newInst = (*getInitialValue)())
+ pod_.inst = new std::shared_ptr<T>(std::move(newInst));
+ registerDestruction();
+ }
+ }
+
+ std::shared_ptr<T> get()
+ {
+ pod_.spinLock.lock();
+ ZEN_ON_SCOPE_EXIT(pod_.spinLock.unlock());
+
if (pod_.inst)
return *pod_.inst;
return nullptr;
@@ -134,20 +153,6 @@ public:
delete tmpInst;
}
- void initOnce(std::unique_ptr<T> (*getInitialValue)())
- {
- pod_.spinLock.lock();
- ZEN_ON_SCOPE_EXIT(pod_.spinLock.unlock());
-
- if (!pod_.cleanUpEntry.cleanUpFun)
- {
- assert(!pod_.inst);
- if (std::unique_ptr<T> newInst = (*getInitialValue)())
- pod_.inst = new std::shared_ptr<T>(std::move(newInst));
- registerDestruction();
- }
- }
-
private:
//call while holding pod_.spinLock
void registerDestruction()
@@ -171,7 +176,7 @@ private:
{
PodSpinMutex spinLock; //rely entirely on static zero-initialization! => avoid potential contention with worker thread during Global<> construction!
//serialize access; can't use std::mutex: has non-trival destructor
- std::shared_ptr<T>* inst; // = nullptr;
+ std::shared_ptr<T>* inst = nullptr;
CleanUpEntry cleanUpEntry;
} pod_;
};
@@ -206,7 +211,7 @@ void registerGlobalForDestruction(CleanUpEntry& entry)
}
//------------------------------------------------------------------------------------------
-#if __cpp_lib_atomic_wait
+#ifdef __cpp_lib_atomic_wait
#error implement + rewiew improvements
#endif
@@ -222,7 +227,7 @@ inline
void PodSpinMutex::lock()
{
while (!tryLock())
-#if __cpp_lib_atomic_wait
+#ifdef __cpp_lib_atomic_wait
flag_.wait(true, std::memory_order_relaxed);
#else
;
@@ -234,7 +239,7 @@ inline
void PodSpinMutex::unlock()
{
flag_.clear(std::memory_order_release);
-#if __cpp_lib_atomic_wait
+#ifdef __cpp_lib_atomic_wait
flag_.notify_one();
#endif
}
bgstack15