mirror of
https://github.com/llvm/llvm-project.git
synced 2025-04-24 17:46:32 +00:00

Fix for performance regression introduced by #80061 that slowed down Fuchsia's MallocFree microbenchmark by 3.5 - 8%
199 lines
6.1 KiB
C++
199 lines
6.1 KiB
C++
//===-- tsd_exclusive.h -----------------------------------------*- C++ -*-===//
|
|
//
|
|
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
|
|
// See https://llvm.org/LICENSE.txt for license information.
|
|
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#ifndef SCUDO_TSD_EXCLUSIVE_H_
|
|
#define SCUDO_TSD_EXCLUSIVE_H_
|
|
|
|
#include "tsd.h"
|
|
|
|
#include "string_utils.h"
|
|
|
|
namespace scudo {
|
|
|
|
struct ThreadState {
|
|
bool DisableMemInit : 1;
|
|
enum : unsigned {
|
|
NotInitialized = 0,
|
|
Initialized,
|
|
TornDown,
|
|
} InitState : 2;
|
|
};
|
|
|
|
template <class Allocator> void teardownThread(void *Ptr);
|
|
|
|
template <class Allocator> struct TSDRegistryExT {
|
|
using ThisT = TSDRegistryExT<Allocator>;
|
|
|
|
struct ScopedTSD {
|
|
ALWAYS_INLINE ScopedTSD(ThisT &TSDRegistry) {
|
|
CurrentTSD = TSDRegistry.getTSDAndLock(&UnlockRequired);
|
|
DCHECK_NE(CurrentTSD, nullptr);
|
|
}
|
|
|
|
~ScopedTSD() {
|
|
if (UNLIKELY(UnlockRequired))
|
|
CurrentTSD->unlock();
|
|
}
|
|
|
|
TSD<Allocator> &operator*() { return *CurrentTSD; }
|
|
|
|
TSD<Allocator> *operator->() {
|
|
CurrentTSD->assertLocked(/*BypassCheck=*/!UnlockRequired);
|
|
return CurrentTSD;
|
|
}
|
|
|
|
private:
|
|
TSD<Allocator> *CurrentTSD;
|
|
bool UnlockRequired;
|
|
};
|
|
|
|
void init(Allocator *Instance) REQUIRES(Mutex) {
|
|
DCHECK(!Initialized);
|
|
Instance->init();
|
|
CHECK_EQ(pthread_key_create(&PThreadKey, teardownThread<Allocator>), 0);
|
|
FallbackTSD.init(Instance);
|
|
Initialized = true;
|
|
}
|
|
|
|
void initOnceMaybe(Allocator *Instance) EXCLUDES(Mutex) {
|
|
ScopedLock L(Mutex);
|
|
if (LIKELY(Initialized))
|
|
return;
|
|
init(Instance); // Sets Initialized.
|
|
}
|
|
|
|
void unmapTestOnly(Allocator *Instance) EXCLUDES(Mutex) {
|
|
DCHECK(Instance);
|
|
if (reinterpret_cast<Allocator *>(pthread_getspecific(PThreadKey))) {
|
|
DCHECK_EQ(reinterpret_cast<Allocator *>(pthread_getspecific(PThreadKey)),
|
|
Instance);
|
|
ThreadTSD.commitBack(Instance);
|
|
ThreadTSD = {};
|
|
}
|
|
CHECK_EQ(pthread_key_delete(PThreadKey), 0);
|
|
PThreadKey = {};
|
|
FallbackTSD.commitBack(Instance);
|
|
FallbackTSD = {};
|
|
State = {};
|
|
ScopedLock L(Mutex);
|
|
Initialized = false;
|
|
}
|
|
|
|
void drainCaches(Allocator *Instance) {
|
|
// We don't have a way to iterate all thread local `ThreadTSD`s. Simply
|
|
// drain the `ThreadTSD` of current thread and `FallbackTSD`.
|
|
Instance->drainCache(&ThreadTSD);
|
|
FallbackTSD.lock();
|
|
Instance->drainCache(&FallbackTSD);
|
|
FallbackTSD.unlock();
|
|
}
|
|
|
|
ALWAYS_INLINE void initThreadMaybe(Allocator *Instance, bool MinimalInit) {
|
|
if (LIKELY(State.InitState != ThreadState::NotInitialized))
|
|
return;
|
|
initThread(Instance, MinimalInit);
|
|
}
|
|
|
|
// To disable the exclusive TSD registry, we effectively lock the fallback TSD
|
|
// and force all threads to attempt to use it instead of their local one.
|
|
void disable() NO_THREAD_SAFETY_ANALYSIS {
|
|
Mutex.lock();
|
|
FallbackTSD.lock();
|
|
atomic_store(&Disabled, 1U, memory_order_release);
|
|
}
|
|
|
|
void enable() NO_THREAD_SAFETY_ANALYSIS {
|
|
atomic_store(&Disabled, 0U, memory_order_release);
|
|
FallbackTSD.unlock();
|
|
Mutex.unlock();
|
|
}
|
|
|
|
bool setOption(Option O, sptr Value) {
|
|
if (O == Option::ThreadDisableMemInit)
|
|
State.DisableMemInit = Value;
|
|
if (O == Option::MaxTSDsCount)
|
|
return false;
|
|
return true;
|
|
}
|
|
|
|
bool getDisableMemInit() { return State.DisableMemInit; }
|
|
|
|
void getStats(ScopedString *Str) {
|
|
// We don't have a way to iterate all thread local `ThreadTSD`s. Instead of
|
|
// printing only self `ThreadTSD` which may mislead the usage, we just skip
|
|
// it.
|
|
Str->append("Exclusive TSD don't support iterating each TSD\n");
|
|
}
|
|
|
|
private:
|
|
ALWAYS_INLINE TSD<Allocator> *
|
|
getTSDAndLock(bool *UnlockRequired) NO_THREAD_SAFETY_ANALYSIS {
|
|
if (LIKELY(State.InitState == ThreadState::Initialized &&
|
|
!atomic_load(&Disabled, memory_order_acquire))) {
|
|
*UnlockRequired = false;
|
|
return &ThreadTSD;
|
|
}
|
|
FallbackTSD.lock();
|
|
*UnlockRequired = true;
|
|
return &FallbackTSD;
|
|
}
|
|
|
|
// Using minimal initialization allows for global initialization while keeping
|
|
// the thread specific structure untouched. The fallback structure will be
|
|
// used instead.
|
|
NOINLINE void initThread(Allocator *Instance, bool MinimalInit) {
|
|
initOnceMaybe(Instance);
|
|
if (UNLIKELY(MinimalInit))
|
|
return;
|
|
CHECK_EQ(
|
|
pthread_setspecific(PThreadKey, reinterpret_cast<void *>(Instance)), 0);
|
|
ThreadTSD.init(Instance);
|
|
State.InitState = ThreadState::Initialized;
|
|
Instance->callPostInitCallback();
|
|
}
|
|
|
|
pthread_key_t PThreadKey = {};
|
|
bool Initialized GUARDED_BY(Mutex) = false;
|
|
atomic_u8 Disabled = {};
|
|
TSD<Allocator> FallbackTSD;
|
|
HybridMutex Mutex;
|
|
static thread_local ThreadState State;
|
|
static thread_local TSD<Allocator> ThreadTSD;
|
|
|
|
friend void teardownThread<Allocator>(void *Ptr);
|
|
};
|
|
|
|
template <class Allocator>
|
|
thread_local TSD<Allocator> TSDRegistryExT<Allocator>::ThreadTSD;
|
|
template <class Allocator>
|
|
thread_local ThreadState TSDRegistryExT<Allocator>::State;
|
|
|
|
template <class Allocator>
|
|
void teardownThread(void *Ptr) NO_THREAD_SAFETY_ANALYSIS {
|
|
typedef TSDRegistryExT<Allocator> TSDRegistryT;
|
|
Allocator *Instance = reinterpret_cast<Allocator *>(Ptr);
|
|
// The glibc POSIX thread-local-storage deallocation routine calls user
|
|
// provided destructors in a loop of PTHREAD_DESTRUCTOR_ITERATIONS.
|
|
// We want to be called last since other destructors might call free and the
|
|
// like, so we wait until PTHREAD_DESTRUCTOR_ITERATIONS before draining the
|
|
// quarantine and swallowing the cache.
|
|
if (TSDRegistryT::ThreadTSD.DestructorIterations > 1) {
|
|
TSDRegistryT::ThreadTSD.DestructorIterations--;
|
|
// If pthread_setspecific fails, we will go ahead with the teardown.
|
|
if (LIKELY(pthread_setspecific(Instance->getTSDRegistry()->PThreadKey,
|
|
Ptr) == 0))
|
|
return;
|
|
}
|
|
TSDRegistryT::ThreadTSD.commitBack(Instance);
|
|
TSDRegistryT::State.InitState = ThreadState::TornDown;
|
|
}
|
|
|
|
} // namespace scudo
|
|
|
|
#endif // SCUDO_TSD_EXCLUSIVE_H_
|