Remove Atomic.h

The STL has everything we need nowadays.

I have tried to not alter any behavior or semantics with this
change wherever possible. In particular, WriteLow and WriteHigh
in CommandProcessor retain the ability to accidentally undo
another thread's write to the upper half or lower half
respectively. If that should be fixed, it should be done in a
separate commit for clarity. One thing did change: The places
where we were using += on a volatile variable (not an atomic
operation) are now using fetch_add (actually an atomic operation).

Tested with single core and dual core on x86-64 and AArch64.

Source/Core/VideoCommon/CommandProcessor.cpp

@@ -6,7 +6,6 @@
 #include <cstring>
 
 #include "Common/Assert.h"
-#include "Common/Atomic.h"
 #include "Common/ChunkFile.h"
 #include "Common/CommonTypes.h"
 #include "Common/Flag.h"
@@ -91,21 +90,15 @@ void DoState(PointerWrap& p)
   p.Do(s_interrupt_waiting);
 }
 
-static inline void WriteLow(volatile u32& _reg, u16 lowbits)
+static inline void WriteLow(std::atomic<u32>& reg, u16 lowbits)
 {
-  Common::AtomicStore(_reg, (_reg & 0xFFFF0000) | lowbits);
+  reg.store((reg.load(std::memory_order_relaxed) & 0xFFFF0000) | lowbits,
+            std::memory_order_relaxed);
 }
-static inline void WriteHigh(volatile u32& _reg, u16 highbits)
+static inline void WriteHigh(std::atomic<u32>& reg, u16 highbits)
 {
-  Common::AtomicStore(_reg, (_reg & 0x0000FFFF) | ((u32)highbits << 16));
-}
-static inline u16 ReadLow(u32 _reg)
-{
-  return (u16)(_reg & 0xFFFF);
-}
-static inline u16 ReadHigh(u32 _reg)
-{
-  return (u16)(_reg >> 16);
+  reg.store((reg.load(std::memory_order_relaxed) & 0x0000FFFF) | (static_cast<u32>(highbits) << 16),
+            std::memory_order_relaxed);
 }
 
 void Init()
@@ -259,31 +252,50 @@ void RegisterMMIO(MMIO::Mapping* mmio, u32 base)
   mmio->Register(base | PERF_SELECT, MMIO::InvalidRead<u16>(), MMIO::Nop<u16>());
 
   // Some MMIOs have different handlers for single core vs. dual core mode.
-  mmio->Register(base | FIFO_RW_DISTANCE_LO,
+  mmio->Register(
+      base | FIFO_RW_DISTANCE_LO,
+      IsOnThread() ? MMIO::ComplexRead<u16>([](u32) {
+        if (fifo.CPWritePointer.load(std::memory_order_relaxed) >=
+            fifo.SafeCPReadPointer.load(std::memory_order_relaxed))
+        {
+          return static_cast<u16>(fifo.CPWritePointer.load(std::memory_order_relaxed) -
+                                  fifo.SafeCPReadPointer.load(std::memory_order_relaxed));
+        }
+        else
+        {
+          return static_cast<u16>(fifo.CPEnd.load(std::memory_order_relaxed) -
+                                  fifo.SafeCPReadPointer.load(std::memory_order_relaxed) +
+                                  fifo.CPWritePointer.load(std::memory_order_relaxed) -
+                                  fifo.CPBase.load(std::memory_order_relaxed) + 32);
+        }
+      }) :
+                     MMIO::DirectRead<u16>(MMIO::Utils::LowPart(&fifo.CPReadWriteDistance)),
+      MMIO::DirectWrite<u16>(MMIO::Utils::LowPart(&fifo.CPReadWriteDistance),
+                             WMASK_LO_ALIGN_32BIT));
+  mmio->Register(base | FIFO_RW_DISTANCE_HI,
                  IsOnThread() ?
                      MMIO::ComplexRead<u16>([](u32) {
-                       if (fifo.CPWritePointer >= fifo.SafeCPReadPointer)
-                         return ReadLow(fifo.CPWritePointer - fifo.SafeCPReadPointer);
+                       Fifo::SyncGPUForRegisterAccess();
+                       if (fifo.CPWritePointer.load(std::memory_order_relaxed) >=
+                           fifo.SafeCPReadPointer.load(std::memory_order_relaxed))
+                       {
+                         return (fifo.CPWritePointer.load(std::memory_order_relaxed) -
+                                 fifo.SafeCPReadPointer.load(std::memory_order_relaxed)) >>
+                                16;
+                       }
                        else
-                         return ReadLow(fifo.CPEnd - fifo.SafeCPReadPointer + fifo.CPWritePointer -
-                                        fifo.CPBase + 32);
+                       {
+                         return (fifo.CPEnd.load(std::memory_order_relaxed) -
+                                 fifo.SafeCPReadPointer.load(std::memory_order_relaxed) +
+                                 fifo.CPWritePointer.load(std::memory_order_relaxed) -
+                                 fifo.CPBase.load(std::memory_order_relaxed) + 32) >>
+                                16;
+                       }
                      }) :
-                     MMIO::DirectRead<u16>(MMIO::Utils::LowPart(&fifo.CPReadWriteDistance)),
-                 MMIO::DirectWrite<u16>(MMIO::Utils::LowPart(&fifo.CPReadWriteDistance),
-                                        WMASK_LO_ALIGN_32BIT));
-  mmio->Register(base | FIFO_RW_DISTANCE_HI,
-                 IsOnThread() ? MMIO::ComplexRead<u16>([](u32) {
-                   Fifo::SyncGPUForRegisterAccess();
-                   if (fifo.CPWritePointer >= fifo.SafeCPReadPointer)
-                     return ReadHigh(fifo.CPWritePointer - fifo.SafeCPReadPointer);
-                   else
-                     return ReadHigh(fifo.CPEnd - fifo.SafeCPReadPointer + fifo.CPWritePointer -
-                                     fifo.CPBase + 32);
-                 }) :
-                                MMIO::ComplexRead<u16>([](u32) {
-                                  Fifo::SyncGPUForRegisterAccess();
-                                  return ReadHigh(fifo.CPReadWriteDistance);
-                                }),
+                     MMIO::ComplexRead<u16>([](u32) {
+                       Fifo::SyncGPUForRegisterAccess();
+                       return fifo.CPReadWriteDistance.load(std::memory_order_relaxed) >> 16;
+                     }),
                  MMIO::ComplexWrite<u16>([WMASK_HI_RESTRICT](u32, u16 val) {
                    Fifo::SyncGPUForRegisterAccess();
                    WriteHigh(fifo.CPReadWriteDistance, val & WMASK_HI_RESTRICT);
@@ -297,16 +309,17 @@ void RegisterMMIO(MMIO::Mapping* mmio, u32 base)
   mmio->Register(base | FIFO_READ_POINTER_HI,
                  IsOnThread() ? MMIO::ComplexRead<u16>([](u32) {
                    Fifo::SyncGPUForRegisterAccess();
-                   return ReadHigh(fifo.SafeCPReadPointer);
+                   return fifo.SafeCPReadPointer.load(std::memory_order_relaxed) >> 16;
                  }) :
                                 MMIO::ComplexRead<u16>([](u32) {
                                   Fifo::SyncGPUForRegisterAccess();
-                                  return ReadHigh(fifo.CPReadPointer);
+                                  return fifo.CPReadPointer.load(std::memory_order_relaxed) >> 16;
                                 }),
                  IsOnThread() ? MMIO::ComplexWrite<u16>([WMASK_HI_RESTRICT](u32, u16 val) {
                    Fifo::SyncGPUForRegisterAccess();
                    WriteHigh(fifo.CPReadPointer, val & WMASK_HI_RESTRICT);
-                   fifo.SafeCPReadPointer = fifo.CPReadPointer;
+                   fifo.SafeCPReadPointer.store(fifo.CPReadPointer.load(std::memory_order_relaxed),
+                                                std::memory_order_relaxed);
                  }) :
                                 MMIO::ComplexWrite<u16>([WMASK_HI_RESTRICT](u32, u16 val) {
                                   Fifo::SyncGPUForRegisterAccess();
@@ -325,8 +338,9 @@ void GatherPipeBursted()
     {
       // In multibuffer mode is not allowed write in the same FIFO attached to the GPU.
       // Fix Pokemon XD in DC mode.
-      if ((ProcessorInterface::Fifo_CPUEnd == fifo.CPEnd) &&
-          (ProcessorInterface::Fifo_CPUBase == fifo.CPBase) && fifo.CPReadWriteDistance > 0)
+      if ((ProcessorInterface::Fifo_CPUEnd == fifo.CPEnd.load(std::memory_order_relaxed)) &&
+          (ProcessorInterface::Fifo_CPUBase == fifo.CPBase.load(std::memory_order_relaxed)) &&
+          fifo.CPReadWriteDistance.load(std::memory_order_relaxed) > 0)
       {
         Fifo::FlushGpu();
       }
@@ -336,35 +350,47 @@ void GatherPipeBursted()
   }
 
   // update the fifo pointer
-  if (fifo.CPWritePointer == fifo.CPEnd)
-    fifo.CPWritePointer = fifo.CPBase;
+  if (fifo.CPWritePointer.load(std::memory_order_relaxed) ==
+      fifo.CPEnd.load(std::memory_order_relaxed))
+  {
+    fifo.CPWritePointer.store(fifo.CPBase, std::memory_order_relaxed);
+  }
   else
-    fifo.CPWritePointer += GATHER_PIPE_SIZE;
+  {
+    fifo.CPWritePointer.fetch_add(GATHER_PIPE_SIZE, std::memory_order_relaxed);
+  }
 
   if (m_CPCtrlReg.GPReadEnable && m_CPCtrlReg.GPLinkEnable)
   {
-    ProcessorInterface::Fifo_CPUWritePointer = fifo.CPWritePointer;
-    ProcessorInterface::Fifo_CPUBase = fifo.CPBase;
-    ProcessorInterface::Fifo_CPUEnd = fifo.CPEnd;
+    ProcessorInterface::Fifo_CPUWritePointer = fifo.CPWritePointer.load(std::memory_order_relaxed);
+    ProcessorInterface::Fifo_CPUBase = fifo.CPBase.load(std::memory_order_relaxed);
+    ProcessorInterface::Fifo_CPUEnd = fifo.CPEnd.load(std::memory_order_relaxed);
   }
 
   // If the game is running close to overflowing, make the exception checking more frequent.
   if (fifo.bFF_HiWatermark)
     CoreTiming::ForceExceptionCheck(0);
 
-  Common::AtomicAdd(fifo.CPReadWriteDistance, GATHER_PIPE_SIZE);
+  fifo.CPReadWriteDistance.fetch_add(GATHER_PIPE_SIZE, std::memory_order_seq_cst);
 
   Fifo::RunGpu();
 
-  ASSERT_MSG(COMMANDPROCESSOR, fifo.CPReadWriteDistance <= fifo.CPEnd - fifo.CPBase,
+  ASSERT_MSG(COMMANDPROCESSOR,
+             fifo.CPReadWriteDistance.load(std::memory_order_relaxed) <=
+                 fifo.CPEnd.load(std::memory_order_relaxed) -
+                     fifo.CPBase.load(std::memory_order_relaxed),
              "FIFO is overflowed by GatherPipe !\nCPU thread is too fast!");
 
   // check if we are in sync
-  ASSERT_MSG(COMMANDPROCESSOR, fifo.CPWritePointer == ProcessorInterface::Fifo_CPUWritePointer,
+  ASSERT_MSG(COMMANDPROCESSOR,
+             fifo.CPWritePointer.load(std::memory_order_relaxed) ==
+                 ProcessorInterface::Fifo_CPUWritePointer,
              "FIFOs linked but out of sync");
-  ASSERT_MSG(COMMANDPROCESSOR, fifo.CPBase == ProcessorInterface::Fifo_CPUBase,
+  ASSERT_MSG(COMMANDPROCESSOR,
+             fifo.CPBase.load(std::memory_order_relaxed) == ProcessorInterface::Fifo_CPUBase,
              "FIFOs linked but out of sync");
-  ASSERT_MSG(COMMANDPROCESSOR, fifo.CPEnd == ProcessorInterface::Fifo_CPUEnd,
+  ASSERT_MSG(COMMANDPROCESSOR,
+             fifo.CPEnd.load(std::memory_order_relaxed) == ProcessorInterface::Fifo_CPUEnd,
              "FIFOs linked but out of sync");
 }
 
@@ -403,31 +429,41 @@ void SetCPStatusFromGPU()
   // breakpoint
   if (fifo.bFF_BPEnable)
   {
-    if (fifo.CPBreakpoint == fifo.CPReadPointer)
+    if (fifo.CPBreakpoint.load(std::memory_order_relaxed) ==
+        fifo.CPReadPointer.load(std::memory_order_relaxed))
     {
       if (!fifo.bFF_Breakpoint)
       {
-        DEBUG_LOG_FMT(COMMANDPROCESSOR, "Hit breakpoint at {}", fifo.CPReadPointer);
+        DEBUG_LOG_FMT(COMMANDPROCESSOR, "Hit breakpoint at {}",
+                      fifo.CPReadPointer.load(std::memory_order_relaxed));
         fifo.bFF_Breakpoint = true;
       }
     }
     else
     {
       if (fifo.bFF_Breakpoint)
-        DEBUG_LOG_FMT(COMMANDPROCESSOR, "Cleared breakpoint at {}", fifo.CPReadPointer);
+      {
+        DEBUG_LOG_FMT(COMMANDPROCESSOR, "Cleared breakpoint at {}",
+                      fifo.CPReadPointer.load(std::memory_order_relaxed));
+      }
       fifo.bFF_Breakpoint = false;
     }
   }
   else
   {
     if (fifo.bFF_Breakpoint)
-      DEBUG_LOG_FMT(COMMANDPROCESSOR, "Cleared breakpoint at {}", fifo.CPReadPointer);
+    {
+      DEBUG_LOG_FMT(COMMANDPROCESSOR, "Cleared breakpoint at {}",
+                    fifo.CPReadPointer.load(std::memory_order_relaxed));
+    }
     fifo.bFF_Breakpoint = false;
   }
 
   // overflow & underflow check
-  fifo.bFF_HiWatermark = (fifo.CPReadWriteDistance > fifo.CPHiWatermark);
-  fifo.bFF_LoWatermark = (fifo.CPReadWriteDistance < fifo.CPLoWatermark);
+  fifo.bFF_HiWatermark =
+      (fifo.CPReadWriteDistance.load(std::memory_order_relaxed) > fifo.CPHiWatermark);
+  fifo.bFF_LoWatermark =
+      (fifo.CPReadWriteDistance.load(std::memory_order_relaxed) < fifo.CPLoWatermark);
 
   bool bpInt = fifo.bFF_Breakpoint && fifo.bFF_BPInt;
   bool ovfInt = fifo.bFF_HiWatermark && fifo.bFF_HiWatermarkInt;
@@ -457,8 +493,10 @@ void SetCPStatusFromGPU()
 void SetCPStatusFromCPU()
 {
   // overflow & underflow check
-  fifo.bFF_HiWatermark = (fifo.CPReadWriteDistance > fifo.CPHiWatermark);
-  fifo.bFF_LoWatermark = (fifo.CPReadWriteDistance < fifo.CPLoWatermark);
+  fifo.bFF_HiWatermark =
+      (fifo.CPReadWriteDistance.load(std::memory_order_relaxed) > fifo.CPHiWatermark);
+  fifo.bFF_LoWatermark =
+      (fifo.CPReadWriteDistance.load(std::memory_order_relaxed) < fifo.CPLoWatermark);
 
   bool bpInt = fifo.bFF_Breakpoint && fifo.bFF_BPInt;
   bool ovfInt = fifo.bFF_HiWatermark && fifo.bFF_HiWatermarkInt;
@@ -489,9 +527,11 @@ void SetCpStatusRegister()
 {
   // Here always there is one fifo attached to the GPU
   m_CPStatusReg.Breakpoint = fifo.bFF_Breakpoint;
-  m_CPStatusReg.ReadIdle = !fifo.CPReadWriteDistance || (fifo.CPReadPointer == fifo.CPWritePointer);
-  m_CPStatusReg.CommandIdle =
-      !fifo.CPReadWriteDistance || Fifo::AtBreakpoint() || !fifo.bFF_GPReadEnable;
+  m_CPStatusReg.ReadIdle = !fifo.CPReadWriteDistance.load(std::memory_order_relaxed) ||
+                           (fifo.CPReadPointer.load(std::memory_order_relaxed) ==
+                            fifo.CPWritePointer.load(std::memory_order_relaxed));
+  m_CPStatusReg.CommandIdle = !fifo.CPReadWriteDistance.load(std::memory_order_relaxed) ||
+                              Fifo::AtBreakpoint() || !fifo.bFF_GPReadEnable;
   m_CPStatusReg.UnderflowLoWatermark = fifo.bFF_LoWatermark;
   m_CPStatusReg.OverflowHiWatermark = fifo.bFF_HiWatermark;
 
@@ -548,29 +588,32 @@ void HandleUnknownOpcode(u8 cmd_byte, void* buffer, bool preprocess)
                  cmd_byte, buffer, preprocess ? "preprocess=true" : "preprocess=false");
 
   {
-    PanicAlertFmt("Illegal command {:02x}\n"
-                  "CPBase: {:#010x}\n"
-                  "CPEnd: {:#010x}\n"
-                  "CPHiWatermark: {:#010x}\n"
-                  "CPLoWatermark: {:#010x}\n"
-                  "CPReadWriteDistance: {:#010x}\n"
-                  "CPWritePointer: {:#010x}\n"
-                  "CPReadPointer: {:#010x}\n"
-                  "CPBreakpoint: {:#010x}\n"
-                  "bFF_GPReadEnable: {}\n"
-                  "bFF_BPEnable: {}\n"
-                  "bFF_BPInt: {}\n"
-                  "bFF_Breakpoint: {}\n"
-                  "bFF_GPLinkEnable: {}\n"
-                  "bFF_HiWatermarkInt: {}\n"
-                  "bFF_LoWatermarkInt: {}\n",
-                  cmd_byte, fifo.CPBase, fifo.CPEnd, fifo.CPHiWatermark, fifo.CPLoWatermark,
-                  fifo.CPReadWriteDistance, fifo.CPWritePointer, fifo.CPReadPointer,
-                  fifo.CPBreakpoint, fifo.bFF_GPReadEnable ? "true" : "false",
-                  fifo.bFF_BPEnable ? "true" : "false", fifo.bFF_BPInt ? "true" : "false",
-                  fifo.bFF_Breakpoint ? "true" : "false", fifo.bFF_GPLinkEnable ? "true" : "false",
-                  fifo.bFF_HiWatermarkInt ? "true" : "false",
-                  fifo.bFF_LoWatermarkInt ? "true" : "false");
+    PanicAlertFmt(
+        "Illegal command {:02x}\n"
+        "CPBase: {:#010x}\n"
+        "CPEnd: {:#010x}\n"
+        "CPHiWatermark: {:#010x}\n"
+        "CPLoWatermark: {:#010x}\n"
+        "CPReadWriteDistance: {:#010x}\n"
+        "CPWritePointer: {:#010x}\n"
+        "CPReadPointer: {:#010x}\n"
+        "CPBreakpoint: {:#010x}\n"
+        "bFF_GPReadEnable: {}\n"
+        "bFF_BPEnable: {}\n"
+        "bFF_BPInt: {}\n"
+        "bFF_Breakpoint: {}\n"
+        "bFF_GPLinkEnable: {}\n"
+        "bFF_HiWatermarkInt: {}\n"
+        "bFF_LoWatermarkInt: {}\n",
+        cmd_byte, fifo.CPBase.load(std::memory_order_relaxed),
+        fifo.CPEnd.load(std::memory_order_relaxed), fifo.CPHiWatermark, fifo.CPLoWatermark,
+        fifo.CPReadWriteDistance.load(std::memory_order_relaxed),
+        fifo.CPWritePointer.load(std::memory_order_relaxed),
+        fifo.CPReadPointer.load(std::memory_order_relaxed),
+        fifo.CPBreakpoint.load(std::memory_order_relaxed), fifo.bFF_GPReadEnable ? "true" : "false",
+        fifo.bFF_BPEnable ? "true" : "false", fifo.bFF_BPInt ? "true" : "false",
+        fifo.bFF_Breakpoint ? "true" : "false", fifo.bFF_GPLinkEnable ? "true" : "false",
+        fifo.bFF_HiWatermarkInt ? "true" : "false", fifo.bFF_LoWatermarkInt ? "true" : "false");
   }
 }
 

Source/Core/VideoCommon/CommandProcessor.h

@@ -4,6 +4,8 @@
 
 #pragma once
 
+#include <atomic>
+
 #include "Common/CommonTypes.h"
 
 class PointerWrap;
@@ -17,15 +19,15 @@ namespace CommandProcessor
 struct SCPFifoStruct
 {
   // fifo registers
-  volatile u32 CPBase;
-  volatile u32 CPEnd;
+  std::atomic<u32> CPBase;
+  std::atomic<u32> CPEnd;
   u32 CPHiWatermark;
   u32 CPLoWatermark;
-  volatile u32 CPReadWriteDistance;
-  volatile u32 CPWritePointer;
-  volatile u32 CPReadPointer;
-  volatile u32 CPBreakpoint;
-  volatile u32 SafeCPReadPointer;
+  std::atomic<u32> CPReadWriteDistance;
+  std::atomic<u32> CPWritePointer;
+  std::atomic<u32> CPReadPointer;
+  std::atomic<u32> CPBreakpoint;
+  std::atomic<u32> SafeCPReadPointer;
 
   volatile u32 bFF_GPLinkEnable;
   volatile u32 bFF_GPReadEnable;

Source/Core/VideoCommon/Fifo.cpp

@@ -8,7 +8,6 @@
 #include <cstring>
 
 #include "Common/Assert.h"
-#include "Common/Atomic.h"
 #include "Common/BlockingLoop.h"
 #include "Common/ChunkFile.h"
 #include "Common/Event.h"
@@ -329,33 +328,37 @@ void RunGpuLoop()
 
           // check if we are able to run this buffer
           while (!CommandProcessor::IsInterruptWaiting() && fifo.bFF_GPReadEnable &&
-                 fifo.CPReadWriteDistance && !AtBreakpoint())
+                 fifo.CPReadWriteDistance.load(std::memory_order_relaxed) && !AtBreakpoint())
           {
             if (param.bSyncGPU && s_sync_ticks.load() < param.iSyncGpuMinDistance)
               break;
 
             u32 cyclesExecuted = 0;
-            u32 readPtr = fifo.CPReadPointer;
+            u32 readPtr = fifo.CPReadPointer.load(std::memory_order_relaxed);
             ReadDataFromFifo(readPtr);
 
-            if (readPtr == fifo.CPEnd)
-              readPtr = fifo.CPBase;
+            if (readPtr == fifo.CPEnd.load(std::memory_order_relaxed))
+              readPtr = fifo.CPBase.load(std::memory_order_relaxed);
             else
               readPtr += 32;
 
-            ASSERT_MSG(COMMANDPROCESSOR, (s32)fifo.CPReadWriteDistance - 32 >= 0,
+            ASSERT_MSG(COMMANDPROCESSOR,
+                       (s32)fifo.CPReadWriteDistance.load(std::memory_order_relaxed) - 32 >= 0,
                        "Negative fifo.CPReadWriteDistance = %i in FIFO Loop !\nThat can produce "
                        "instability in the game. Please report it.",
-                       fifo.CPReadWriteDistance - 32);
+                       fifo.CPReadWriteDistance.load(std::memory_order_relaxed) - 32);
 
             u8* write_ptr = s_video_buffer_write_ptr;
             s_video_buffer_read_ptr = OpcodeDecoder::Run(
                 DataReader(s_video_buffer_read_ptr, write_ptr), &cyclesExecuted, false);
 
-            Common::AtomicStore(fifo.CPReadPointer, readPtr);
-            Common::AtomicAdd(fifo.CPReadWriteDistance, static_cast<u32>(-32));
+            fifo.CPReadPointer.store(readPtr, std::memory_order_relaxed);
+            fifo.CPReadWriteDistance.fetch_sub(32, std::memory_order_seq_cst);
             if ((write_ptr - s_video_buffer_read_ptr) == 0)
-              Common::AtomicStore(fifo.SafeCPReadPointer, fifo.CPReadPointer);
+            {
+              fifo.SafeCPReadPointer.store(fifo.CPReadPointer.load(std::memory_order_relaxed),
+                                           std::memory_order_relaxed);
+            }
 
             CommandProcessor::SetCPStatusFromGPU();
 
@@ -412,7 +415,8 @@ void GpuMaySleep()
 bool AtBreakpoint()
 {
   CommandProcessor::SCPFifoStruct& fifo = CommandProcessor::fifo;
-  return fifo.bFF_BPEnable && (fifo.CPReadPointer == fifo.CPBreakpoint);
+  return fifo.bFF_BPEnable && (fifo.CPReadPointer.load(std::memory_order_relaxed) ==
+                               fifo.CPBreakpoint.load(std::memory_order_relaxed));
 }
 
 void RunGpu()
@@ -442,12 +446,12 @@ static int RunGpuOnCpu(int ticks)
   CommandProcessor::SCPFifoStruct& fifo = CommandProcessor::fifo;
   bool reset_simd_state = false;
   int available_ticks = int(ticks * SConfig::GetInstance().fSyncGpuOverclock) + s_sync_ticks.load();
-  while (fifo.bFF_GPReadEnable && fifo.CPReadWriteDistance && !AtBreakpoint() &&
-         available_ticks >= 0)
+  while (fifo.bFF_GPReadEnable && fifo.CPReadWriteDistance.load(std::memory_order_relaxed) &&
+         !AtBreakpoint() && available_ticks >= 0)
   {
     if (s_use_deterministic_gpu_thread)
     {
-      ReadDataFromFifoOnCPU(fifo.CPReadPointer);
+      ReadDataFromFifoOnCPU(fifo.CPReadPointer.load(std::memory_order_relaxed));
       s_gpu_mainloop.Wakeup();
     }
     else
@@ -458,19 +462,25 @@ static int RunGpuOnCpu(int ticks)
         FPURoundMode::LoadDefaultSIMDState();
         reset_simd_state = true;
       }
-      ReadDataFromFifo(fifo.CPReadPointer);
+      ReadDataFromFifo(fifo.CPReadPointer.load(std::memory_order_relaxed));
       u32 cycles = 0;
       s_video_buffer_read_ptr = OpcodeDecoder::Run(
           DataReader(s_video_buffer_read_ptr, s_video_buffer_write_ptr), &cycles, false);
       available_ticks -= cycles;
     }
 
-    if (fifo.CPReadPointer == fifo.CPEnd)
-      fifo.CPReadPointer = fifo.CPBase;
+    if (fifo.CPReadPointer.load(std::memory_order_relaxed) ==
+        fifo.CPEnd.load(std::memory_order_relaxed))
+    {
+      fifo.CPReadPointer.store(fifo.CPBase.load(std::memory_order_relaxed),
+                               std::memory_order_relaxed);
+    }
     else
-      fifo.CPReadPointer += 32;
+    {
+      fifo.CPReadPointer.fetch_add(32, std::memory_order_relaxed);
+    }
 
-    fifo.CPReadWriteDistance -= 32;
+    fifo.CPReadWriteDistance.fetch_sub(32, std::memory_order_relaxed);
   }
 
   CommandProcessor::SetCPStatusFromGPU();

Source/Core/VideoCommon/RenderBase.cpp

@@ -889,7 +889,9 @@ void Renderer::CheckFifoRecording()
     RecordVideoMemory();
   }
 
-  FifoRecorder::GetInstance().EndFrame(CommandProcessor::fifo.CPBase, CommandProcessor::fifo.CPEnd);
+  FifoRecorder::GetInstance().EndFrame(
+      CommandProcessor::fifo.CPBase.load(std::memory_order_relaxed),
+      CommandProcessor::fifo.CPEnd.load(std::memory_order_relaxed));
 }
 
 void Renderer::RecordVideoMemory()