ppsspp/Core/HLE/sceKernelThread.cpp
2013-04-07 11:47:00 -07:00

3216 lines
82 KiB
C++

// Copyright (c) 2012- PPSSPP Project.
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, version 2.0 or later versions.
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License 2.0 for more details.
// A copy of the GPL 2.0 should have been included with the program.
// If not, see http://www.gnu.org/licenses/
// Official git repository and contact information can be found at
// https://github.com/hrydgard/ppsspp and http://www.ppsspp.org/.
#include <set>
#include <map>
#include <queue>
#include <algorithm>
#include "Common/LogManager.h"
#include "HLE.h"
#include "HLETables.h"
#include "../MIPS/MIPSInt.h"
#include "../MIPS/MIPSCodeUtils.h"
#include "../MIPS/MIPS.h"
#include "Core/CoreTiming.h"
#include "Core/MemMap.h"
#include "Core/Reporting.h"
#include "ChunkFile.h"
#include "sceAudio.h"
#include "sceKernel.h"
#include "sceKernelMemory.h"
#include "sceKernelThread.h"
#include "sceKernelModule.h"
#include "sceKernelInterrupt.h"
enum {
ERROR_KERNEL_THREAD_ALREADY_DORMANT = 0x800201a2,
ERROR_KERNEL_THREAD_ALREADY_SUSPEND = 0x800201a3,
ERROR_KERNEL_THREAD_IS_NOT_DORMANT = 0x800201a4,
ERROR_KERNEL_THREAD_IS_NOT_SUSPEND = 0x800201a5,
ERROR_KERNEL_THREAD_IS_NOT_WAIT = 0x800201a6,
};
enum
{
PSP_THREAD_ATTR_USER = 0x80000000,
PSP_THREAD_ATTR_USBWLAN = 0xa0000000,
PSP_THREAD_ATTR_VSH = 0xc0000000,
PSP_THREAD_ATTR_KERNEL = 0x00001000,
PSP_THREAD_ATTR_VFPU = 0x00004000, // TODO: Should not bother saving VFPU context except when switching between two thread that has this attribute
PSP_THREAD_ATTR_SCRATCH_SRAM = 0x00008000, // Save/restore scratch as part of context???
PSP_THREAD_ATTR_NO_FILLSTACK = 0x00100000, // TODO: No filling of 0xff
PSP_THREAD_ATTR_CLEAR_STACK = 0x00200000, // TODO: Clear thread stack when deleted
};
struct NativeCallback
{
SceUInt size;
char name[32];
SceUID threadId;
u32 entrypoint;
u32 commonArgument;
int notifyCount;
int notifyArg;
};
class Callback : public KernelObject
{
public:
const char *GetName() {return nc.name;}
const char *GetTypeName() {return "CallBack";}
void GetQuickInfo(char *ptr, int size)
{
sprintf(ptr, "thread=%i, argument= %08x",
//hackAddress,
nc.threadId,
nc.commonArgument);
}
~Callback()
{
}
static u32 GetMissingErrorCode() { return SCE_KERNEL_ERROR_UNKNOWN_CBID; }
int GetIDType() const { return SCE_KERNEL_TMID_Callback; }
virtual void DoState(PointerWrap &p)
{
p.Do(nc);
p.Do(savedPC);
p.Do(savedRA);
p.Do(savedV0);
p.Do(savedV1);
p.Do(savedIdRegister);
p.Do(forceDelete);
p.DoMarker("Callback");
}
NativeCallback nc;
u32 savedPC;
u32 savedRA;
u32 savedV0;
u32 savedV1;
u32 savedIdRegister;
/*
SceUInt attr;
SceUInt initPattern;
SceUInt currentPattern;
int numWaitThreads;
*/
bool forceDelete;
};
// Real PSP struct, don't change the fields
struct NativeThread
{
u32 nativeSize;
char name[KERNELOBJECT_MAX_NAME_LENGTH+1];
// Threading stuff
u32 attr;
u32 status;
u32 entrypoint;
u32 initialStack;
u32 stackSize;
u32 gpreg;
int initialPriority;
int currentPriority;
WaitType waitType;
SceUID waitID;
int wakeupCount;
int exitStatus;
SceKernelSysClock runForClocks;
int numInterruptPreempts;
int numThreadPreempts;
int numReleases;
};
struct ThreadWaitInfo {
u32 waitValue;
u32 timeoutPtr;
};
// Owns outstanding MIPS calls and provides a way to get them by ID.
class MipsCallManager {
public:
MipsCallManager() : idGen_(0) {}
u32 add(MipsCall *call) {
u32 id = genId();
calls_.insert(std::pair<int, MipsCall *>(id, call));
return id;
}
MipsCall *get(u32 id) {
auto iter = calls_.find(id);
if (iter == calls_.end())
return NULL;
return iter->second;
}
MipsCall *pop(u32 id) {
MipsCall *temp = calls_[id];
calls_.erase(id);
return temp;
}
void clear() {
for (auto it = calls_.begin(), end = calls_.end(); it != end; ++it) {
delete it->second;
}
calls_.clear();
idGen_ = 0;
}
int registerActionType(ActionCreator creator) {
types_.push_back(creator);
return (int) types_.size() - 1;
}
void restoreActionType(int actionType, ActionCreator creator) {
if (actionType >= (int) types_.size())
types_.resize(actionType + 1, NULL);
types_[actionType] = creator;
}
Action *createActionByType(int actionType) {
if (actionType < (int) types_.size() && types_[actionType] != NULL) {
Action *a = types_[actionType]();
a->actionTypeID = actionType;
return a;
}
return NULL;
}
void DoState(PointerWrap &p) {
p.Do(calls_);
p.Do(idGen_);
p.DoMarker("MipsCallManager");
}
private:
u32 genId() { return ++idGen_; }
std::map<u32, MipsCall *> calls_;
std::vector<ActionCreator> types_;
u32 idGen_;
};
class ActionAfterMipsCall : public Action
{
ActionAfterMipsCall()
{
chainedAction = NULL;
}
public:
virtual void run(MipsCall &call);
static Action *Create()
{
return new ActionAfterMipsCall();
}
virtual void DoState(PointerWrap &p)
{
p.Do(threadID);
p.Do(status);
p.Do(waitType);
p.Do(waitID);
p.Do(waitInfo);
p.Do(isProcessingCallbacks);
p.Do(currentCallbackId);
p.DoMarker("ActionAfterMipsCall");
int chainedActionType = 0;
if (chainedAction != NULL)
chainedActionType = chainedAction->actionTypeID;
p.Do(chainedActionType);
if (chainedActionType != 0)
{
if (p.mode == p.MODE_READ)
chainedAction = __KernelCreateAction(chainedActionType);
chainedAction->DoState(p);
}
}
SceUID threadID;
// Saved thread state
int status;
WaitType waitType;
int waitID;
ThreadWaitInfo waitInfo;
bool isProcessingCallbacks;
SceUID currentCallbackId;
Action *chainedAction;
};
class ActionAfterCallback : public Action
{
public:
ActionAfterCallback() {}
virtual void run(MipsCall &call);
static Action *Create()
{
return new ActionAfterCallback;
}
void setCallback(SceUID cbId_)
{
cbId = cbId_;
}
void DoState(PointerWrap &p)
{
p.Do(cbId);
p.DoMarker("ActionAfterCallback");
}
SceUID cbId;
};
class Thread : public KernelObject
{
public:
const char *GetName() {return nt.name;}
const char *GetTypeName() {return "Thread";}
void GetQuickInfo(char *ptr, int size)
{
sprintf(ptr, "pc= %08x sp= %08x %s %s %s %s %s %s (wt=%i wid=%i wv= %08x )",
context.pc, context.r[MIPS_REG_SP],
(nt.status & THREADSTATUS_RUNNING) ? "RUN" : "",
(nt.status & THREADSTATUS_READY) ? "READY" : "",
(nt.status & THREADSTATUS_WAIT) ? "WAIT" : "",
(nt.status & THREADSTATUS_SUSPEND) ? "SUSPEND" : "",
(nt.status & THREADSTATUS_DORMANT) ? "DORMANT" : "",
(nt.status & THREADSTATUS_DEAD) ? "DEAD" : "",
nt.waitType,
nt.waitID,
waitInfo.waitValue);
}
static u32 GetMissingErrorCode() { return SCE_KERNEL_ERROR_UNKNOWN_THID; }
int GetIDType() const { return SCE_KERNEL_TMID_Thread; }
bool AllocateStack(u32 &stackSize)
{
FreeStack();
if (nt.attr & PSP_THREAD_ATTR_KERNEL)
{
// Allocate stacks for kernel threads (idle) in kernel RAM
stackBlock = kernelMemory.Alloc(stackSize, true, (std::string("stack/") + nt.name).c_str());
}
else
{
stackBlock = userMemory.Alloc(stackSize, true, (std::string("stack/") + nt.name).c_str());
}
if (stackBlock == (u32)-1)
{
stackBlock = 0;
ERROR_LOG(HLE, "Failed to allocate stack for thread");
return false;
}
// Fill the stack.
Memory::Memset(stackBlock, 0xFF, stackSize);
context.r[MIPS_REG_SP] = stackBlock + stackSize;
stackEnd = context.r[MIPS_REG_SP];
nt.initialStack = stackBlock;
nt.stackSize = stackSize;
// What's this 512?
context.r[MIPS_REG_K0] = context.r[MIPS_REG_SP] - 256;
context.r[MIPS_REG_SP] -= 512;
u32 k0 = context.r[MIPS_REG_K0];
Memory::Memset(k0, 0, 0x100);
Memory::Write_U32(nt.initialStack, k0 + 0xc0);
Memory::Write_U32(GetUID(), k0 + 0xca);
Memory::Write_U32(0xffffffff, k0 + 0xf8);
Memory::Write_U32(0xffffffff, k0 + 0xfc);
Memory::Write_U32(GetUID(), nt.initialStack);
return true;
}
void FreeStack() {
if (stackBlock != 0) {
DEBUG_LOG(HLE, "Freeing thread stack %s", nt.name);
if (nt.attr & PSP_THREAD_ATTR_KERNEL) {
kernelMemory.Free(stackBlock);
} else {
userMemory.Free(stackBlock);
}
stackBlock = 0;
}
}
Thread() : stackBlock(0)
{
}
~Thread()
{
FreeStack();
}
ActionAfterMipsCall *getRunningCallbackAction();
void setReturnValue(u32 retval);
void setReturnValue(u64 retval);
void resumeFromWait();
bool isWaitingFor(WaitType type, int id);
int getWaitID(WaitType type);
ThreadWaitInfo getWaitInfo();
// Utils
inline bool isRunning() const { return (nt.status & THREADSTATUS_RUNNING) != 0; }
inline bool isStopped() const { return (nt.status & THREADSTATUS_DORMANT) != 0; }
inline bool isReady() const { return (nt.status & THREADSTATUS_READY) != 0; }
inline bool isWaiting() const { return (nt.status & THREADSTATUS_WAIT) != 0; }
inline bool isSuspended() const { return (nt.status & THREADSTATUS_SUSPEND) != 0; }
virtual void DoState(PointerWrap &p)
{
p.Do(nt);
p.Do(waitInfo);
p.Do(moduleId);
p.Do(isProcessingCallbacks);
p.Do(currentMipscallId);
p.Do(currentCallbackId);
p.Do(context);
u32 numCallbacks = THREAD_CALLBACK_NUM_TYPES;
p.Do(numCallbacks);
if (numCallbacks != THREAD_CALLBACK_NUM_TYPES)
ERROR_LOG(HLE, "Unable to load state: different kernel object storage.");
for (size_t i = 0; i < THREAD_CALLBACK_NUM_TYPES; ++i)
{
p.Do(registeredCallbacks[i]);
p.Do(readyCallbacks[i]);
}
p.Do(pendingMipsCalls);
p.Do(stackBlock);
p.DoMarker("Thread");
}
NativeThread nt;
ThreadWaitInfo waitInfo;
SceUID moduleId;
bool isProcessingCallbacks;
u32 currentMipscallId;
SceUID currentCallbackId;
ThreadContext context;
std::set<SceUID> registeredCallbacks[THREAD_CALLBACK_NUM_TYPES];
std::list<SceUID> readyCallbacks[THREAD_CALLBACK_NUM_TYPES];
std::list<u32> pendingMipsCalls;
u32 stackBlock;
u32 stackEnd;
};
struct ThreadQueueList
{
// Number of queues (number of priority levels starting at 0.)
static const int NUM_QUEUES = 128;
// Initial number of threads a single queue can handle.
static const int INITIAL_CAPACITY = 32;
struct Queue
{
// Next ever-been-used queue (worse priority.)
Queue *next;
// First valid item in data.
int first;
// One after last valid item in data.
int end;
// A too-large array with room on the front and end.
SceUID *data;
// Size of data array.
int capacity;
};
ThreadQueueList()
{
memset(queues, 0, sizeof(queues));
first = invalid();
}
~ThreadQueueList()
{
for (int i = 0; i < NUM_QUEUES; ++i)
{
if (queues[i].data != NULL)
free(queues[i].data);
}
}
inline SceUID pop_first()
{
Queue *cur = first;
while (cur != invalid())
{
if (cur->end - cur->first > 0)
return cur->data[cur->first++];
cur = cur->next;
}
_dbg_assert_msg_(HLE, false, "ThreadQueueList should not be empty.");
return 0;
}
inline SceUID pop_first_better(u32 priority)
{
Queue *cur = first;
Queue *stop = &queues[priority];
while (cur < stop)
{
if (cur->end - cur->first > 0)
return cur->data[cur->first++];
cur = cur->next;
}
return 0;
}
inline void push_front(u32 priority, const SceUID threadID)
{
Queue *cur = &queues[priority];
cur->data[--cur->first] = threadID;
if (cur->first == 0)
rebalance(priority);
}
inline void push_back(u32 priority, const SceUID threadID)
{
Queue *cur = &queues[priority];
cur->data[cur->end++] = threadID;
if (cur->end == cur->capacity)
rebalance(priority);
}
inline void remove(u32 priority, const SceUID threadID)
{
Queue *cur = &queues[priority];
_dbg_assert_msg_(HLE, cur->next != NULL, "ThreadQueueList::Queue should already be linked up.");
for (int i = cur->first; i < cur->end; ++i)
{
if (cur->data[i] == threadID)
{
int remaining = --cur->end - i;
if (remaining > 0)
memmove(&cur->data[i], &cur->data[i + 1], remaining * sizeof(SceUID));
return;
}
}
// Wasn't there.
}
inline void rotate(u32 priority)
{
Queue *cur = &queues[priority];
_dbg_assert_msg_(HLE, cur->next != NULL, "ThreadQueueList::Queue should already be linked up.");
if (cur->end - cur->first > 1)
{
cur->data[cur->end++] = cur->data[cur->first++];
if (cur->end == cur->capacity)
rebalance(priority);
}
}
inline void clear()
{
for (int i = 0; i < NUM_QUEUES; ++i)
{
if (queues[i].data != NULL)
free(queues[i].data);
}
memset(queues, 0, sizeof(queues));
first = invalid();
}
inline bool empty(u32 priority) const
{
const Queue *cur = &queues[priority];
return cur->first == cur->end;
}
inline void prepare(u32 priority)
{
Queue *cur = &queues[priority];
if (cur->next == NULL)
link(priority, INITIAL_CAPACITY);
}
void DoState(PointerWrap &p)
{
int numQueues = NUM_QUEUES;
p.Do(numQueues);
if (numQueues != NUM_QUEUES)
{
ERROR_LOG(HLE, "Savestate loading error: invalid data");
return;
}
if (p.mode == p.MODE_READ)
clear();
for (int i = 0; i < NUM_QUEUES; ++i)
{
Queue *cur = &queues[i];
int size = cur->end - cur->first;
p.Do(size);
int capacity = cur->capacity;
p.Do(capacity);
if (capacity == 0)
continue;
if (p.mode == p.MODE_READ)
{
link(i, capacity);
cur->first = (cur->capacity - size) / 2;
cur->end = cur->first + size;
}
if (size != 0)
p.DoArray(&cur->data[cur->first], size);
}
p.DoMarker("ThreadQueueList");
}
private:
Queue *invalid() const
{
return (Queue *) -1;
}
void link(u32 priority, int size)
{
_dbg_assert_msg_(HLE, queues[priority].data == NULL, "ThreadQueueList::Queue should only be initialized once.");
if (size <= INITIAL_CAPACITY)
size = INITIAL_CAPACITY;
else
{
int goal = size;
size = INITIAL_CAPACITY;
while (size < goal)
size *= 2;
}
Queue *cur = &queues[priority];
cur->data = (SceUID *) malloc(sizeof(SceUID) * size);
cur->capacity = size;
cur->first = size / 2;
cur->end = size / 2;
for (int i = (int) priority - 1; i >= 0; --i)
{
if (queues[i].next != NULL)
{
cur->next = queues[i].next;
queues[i].next = cur;
return;
}
}
cur->next = first;
first = cur;
}
void rebalance(u32 priority)
{
Queue *cur = &queues[priority];
int size = cur->end - cur->first;
if (size >= cur->capacity - 2)
{
cur->capacity *= 2;
cur->data = (SceUID *)realloc(cur->data, cur->capacity * sizeof(SceUID));
}
int newFirst = (cur->capacity - size) / 2;
if (newFirst != cur->first)
{
memmove(&cur->data[newFirst], &cur->data[cur->first], size * sizeof(SceUID));
cur->first = newFirst;
cur->end = newFirst + size;
}
}
// The first queue that's ever been used.
Queue *first;
// The priority level queues of thread ids.
Queue queues[NUM_QUEUES];
};
struct WaitTypeFuncs
{
WaitBeginCallbackFunc beginFunc;
WaitEndCallbackFunc endFunc;
};
void __KernelExecuteMipsCallOnCurrentThread(u32 callId, bool reschedAfter);
Thread *__KernelCreateThread(SceUID &id, SceUID moduleID, const char *name, u32 entryPoint, u32 priority, int stacksize, u32 attr);
void __KernelResetThread(Thread *t);
void __KernelCancelWakeup(SceUID threadID);
void __KernelCancelThreadEndTimeout(SceUID threadID);
bool __KernelCheckThreadCallbacks(Thread *thread, bool force);
//////////////////////////////////////////////////////////////////////////
//STATE BEGIN
//////////////////////////////////////////////////////////////////////////
int g_inCbCount = 0;
// Normally, the same as currentThread. In an interrupt, remembers the callback's thread id.
SceUID currentCallbackThreadID = 0;
int readyCallbacksCount = 0;
SceUID currentThread;
u32 idleThreadHackAddr;
u32 threadReturnHackAddr;
u32 cbReturnHackAddr;
u32 intReturnHackAddr;
std::vector<ThreadCallback> threadEndListeners;
// Lists all thread ids that aren't deleted/etc.
std::vector<SceUID> threadqueue;
// Lists only ready thread ids.
ThreadQueueList threadReadyQueue;
SceUID threadIdleID[2];
int eventScheduledWakeup;
int eventThreadEndTimeout;
bool dispatchEnabled = true;
MipsCallManager mipsCalls;
int actionAfterCallback;
int actionAfterMipsCall;
// This seems nasty
SceUID curModule;
// Doesn't need state saving.
WaitTypeFuncs waitTypeFuncs[NUM_WAITTYPES];
//////////////////////////////////////////////////////////////////////////
//STATE END
//////////////////////////////////////////////////////////////////////////
int __KernelRegisterActionType(ActionCreator creator)
{
return mipsCalls.registerActionType(creator);
}
void __KernelRestoreActionType(int actionType, ActionCreator creator)
{
mipsCalls.restoreActionType(actionType, creator);
}
Action *__KernelCreateAction(int actionType)
{
return mipsCalls.createActionByType(actionType);
}
void MipsCall::DoState(PointerWrap &p)
{
p.Do(entryPoint);
p.Do(cbId);
p.DoArray(args, ARRAY_SIZE(args));
p.Do(numArgs);
p.Do(savedIdRegister);
p.Do(savedRa);
p.Do(savedPc);
p.Do(savedV0);
p.Do(savedV1);
p.Do(tag);
p.Do(savedId);
p.Do(reschedAfter);
p.DoMarker("MipsCall");
int actionTypeID = 0;
if (doAfter != NULL)
actionTypeID = doAfter->actionTypeID;
p.Do(actionTypeID);
if (actionTypeID != 0)
{
if (p.mode == p.MODE_READ)
doAfter = __KernelCreateAction(actionTypeID);
doAfter->DoState(p);
}
}
void MipsCall::setReturnValue(u32 value)
{
savedV0 = value;
}
void MipsCall::setReturnValue(u64 value)
{
savedV0 = value & 0xFFFFFFFF;
savedV1 = (value >> 32) & 0xFFFFFFFF;
}
Thread *__GetCurrentThread() {
if (currentThread != 0)
return kernelObjects.GetFast<Thread>(currentThread);
else
return NULL;
}
u32 __KernelMipsCallReturnAddress()
{
return cbReturnHackAddr;
}
u32 __KernelInterruptReturnAddress()
{
return intReturnHackAddr;
}
void hleScheduledWakeup(u64 userdata, int cyclesLate);
void hleThreadEndTimeout(u64 userdata, int cyclesLate);
void __KernelThreadingInit()
{
u32 blockSize = 4 * 4 + 4 * 2 * 3; // One 16-byte thread plus 3 8-byte "hacks"
dispatchEnabled = true;
memset(waitTypeFuncs, 0, sizeof(waitTypeFuncs));
currentThread = 0;
g_inCbCount = 0;
currentCallbackThreadID = 0;
readyCallbacksCount = 0;
idleThreadHackAddr = kernelMemory.Alloc(blockSize, false, "threadrethack");
// Make sure it got allocated where we expect it... at the very start of kernel RAM
//CHECK_EQ(idleThreadHackAddr & 0x3FFFFFFF, 0x08000000);
// Yeah, this is straight out of JPCSP, I should be ashamed.
Memory::Write_U32(MIPS_MAKE_ADDIU(MIPS_REG_A0, MIPS_REG_ZERO, 0), idleThreadHackAddr);
Memory::Write_U32(MIPS_MAKE_LUI(MIPS_REG_RA, 0x0800), idleThreadHackAddr + 4);
Memory::Write_U32(MIPS_MAKE_JR_RA(), idleThreadHackAddr + 8);
//Memory::Write_U32(MIPS_MAKE_SYSCALL("ThreadManForUser", "sceKernelDelayThread"), idleThreadHackAddr + 12);
Memory::Write_U32(MIPS_MAKE_SYSCALL("FakeSysCalls", "_sceKernelIdle"), idleThreadHackAddr + 12);
Memory::Write_U32(MIPS_MAKE_BREAK(), idleThreadHackAddr + 16);
threadReturnHackAddr = idleThreadHackAddr + 20;
WriteSyscall("FakeSysCalls", NID_THREADRETURN, threadReturnHackAddr);
cbReturnHackAddr = threadReturnHackAddr + 8;
WriteSyscall("FakeSysCalls", NID_CALLBACKRETURN, cbReturnHackAddr);
intReturnHackAddr = cbReturnHackAddr + 8;
WriteSyscall("FakeSysCalls", NID_INTERRUPTRETURN, intReturnHackAddr);
eventScheduledWakeup = CoreTiming::RegisterEvent("ScheduledWakeup", &hleScheduledWakeup);
eventThreadEndTimeout = CoreTiming::RegisterEvent("ThreadEndTimeout", &hleThreadEndTimeout);
actionAfterMipsCall = __KernelRegisterActionType(ActionAfterMipsCall::Create);
actionAfterCallback = __KernelRegisterActionType(ActionAfterCallback::Create);
// Create the two idle threads, as well. With the absolute minimal possible priority.
// 4096 stack size - don't know what the right value is. Hm, if callbacks are ever to run on these threads...
__KernelResetThread(__KernelCreateThread(threadIdleID[0], 0, "idle0", idleThreadHackAddr, 0x7f, 4096, PSP_THREAD_ATTR_KERNEL));
__KernelResetThread(__KernelCreateThread(threadIdleID[1], 0, "idle1", idleThreadHackAddr, 0x7f, 4096, PSP_THREAD_ATTR_KERNEL));
// These idle threads are later started in LoadExec, which calls __KernelStartIdleThreads below.
__KernelListenThreadEnd(__KernelCancelWakeup);
__KernelListenThreadEnd(__KernelCancelThreadEndTimeout);
}
void __KernelThreadingDoState(PointerWrap &p)
{
p.Do(g_inCbCount);
p.Do(currentCallbackThreadID);
p.Do(readyCallbacksCount);
p.Do(idleThreadHackAddr);
p.Do(threadReturnHackAddr);
p.Do(cbReturnHackAddr);
p.Do(intReturnHackAddr);
p.Do(currentThread);
SceUID dv = 0;
p.Do(threadqueue, dv);
p.DoArray(threadIdleID, ARRAY_SIZE(threadIdleID));
p.Do(dispatchEnabled);
p.Do(curModule);
p.Do(threadReadyQueue);
p.Do(eventScheduledWakeup);
CoreTiming::RestoreRegisterEvent(eventScheduledWakeup, "ScheduledWakeup", &hleScheduledWakeup);
p.Do(eventThreadEndTimeout);
CoreTiming::RestoreRegisterEvent(eventThreadEndTimeout, "ThreadEndTimeout", &hleThreadEndTimeout);
p.Do(actionAfterMipsCall);
__KernelRestoreActionType(actionAfterMipsCall, ActionAfterMipsCall::Create);
p.Do(actionAfterCallback);
__KernelRestoreActionType(actionAfterCallback, ActionAfterCallback::Create);
hleCurrentThreadName = __KernelGetThreadName(currentThread);
p.DoMarker("sceKernelThread");
}
void __KernelThreadingDoStateLate(PointerWrap &p)
{
// We do this late to give modules time to register actions.
mipsCalls.DoState(p);
p.DoMarker("sceKernelThread Late");
}
KernelObject *__KernelThreadObject()
{
return new Thread;
}
KernelObject *__KernelCallbackObject()
{
return new Callback;
}
void __KernelListenThreadEnd(ThreadCallback callback)
{
threadEndListeners.push_back(callback);
}
void __KernelFireThreadEnd(SceUID threadID)
{
for (auto iter = threadEndListeners.begin(), end = threadEndListeners.end(); iter != end; ++iter)
{
ThreadCallback cb = *iter;
cb(threadID);
}
}
// TODO: Use __KernelChangeThreadState instead? It has other affects...
void __KernelChangeReadyState(Thread *thread, SceUID threadID, bool ready)
{
int prio = thread->nt.currentPriority;
if (thread->isReady())
{
if (!ready)
threadReadyQueue.remove(prio, threadID);
}
else if (ready)
{
if (thread->isRunning())
threadReadyQueue.push_front(prio, threadID);
else
threadReadyQueue.push_back(prio, threadID);
thread->nt.status = THREADSTATUS_READY;
}
}
void __KernelChangeReadyState(SceUID threadID, bool ready)
{
u32 error;
Thread *thread = kernelObjects.Get<Thread>(threadID, error);
if (thread)
__KernelChangeReadyState(thread, threadID, ready);
else
WARN_LOG(HLE, "Trying to change the ready state of an unknown thread?");
}
void __KernelStartIdleThreads()
{
for (int i = 0; i < 2; i++)
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadIdleID[i], error);
t->nt.gpreg = __KernelGetModuleGP(curModule);
t->context.r[MIPS_REG_GP] = t->nt.gpreg;
//t->context.pc += 4; // ADJUSTPC
threadReadyQueue.prepare(t->nt.currentPriority);
__KernelChangeReadyState(t, threadIdleID[i], true);
}
}
bool __KernelSwitchOffThread(const char *reason)
{
if (!reason)
reason = "switch off thread";
SceUID threadID = currentThread;
if (threadID != threadIdleID[0] && threadID != threadIdleID[1])
{
Thread *current = __GetCurrentThread();
if (current && current->isRunning())
__KernelChangeReadyState(current, threadID, true);
// Idle 0 chosen entirely arbitrarily.
Thread *t = kernelObjects.GetFast<Thread>(threadIdleID[0]);
if (t)
{
__KernelSwitchContext(t, reason);
return true;
}
else
ERROR_LOG(HLE, "Unable to switch to idle thread.");
}
return false;
}
bool __KernelSwitchToThread(SceUID threadID, const char *reason)
{
if (!reason)
reason = "switch to thread";
if (currentThread != threadIdleID[0] && currentThread != threadIdleID[1])
{
ERROR_LOG(HLE, "__KernelSwitchToThread used when already on a thread.");
return false;
}
if (currentThread == threadID)
return false;
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (!t)
ERROR_LOG(HLE, "__KernelSwitchToThread: %x doesn't exist", threadID)
else
{
Thread *current = __GetCurrentThread();
if (current && current->isRunning())
__KernelChangeReadyState(current, threadID, true);
__KernelSwitchContext(t, reason);
return true;
}
return false;
}
void __KernelIdle()
{
CoreTiming::Idle();
// Advance must happen between Idle and Reschedule, so that threads that were waiting for something
// that was triggered at the end of the Idle period must get a chance to be scheduled.
CoreTiming::AdvanceQuick();
// We must've exited a callback?
if (__KernelInCallback())
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(currentCallbackThreadID, error);
if (t)
{
__KernelChangeReadyState(t, currentCallbackThreadID, false);
t->nt.status = (t->nt.status | THREADSTATUS_RUNNING) & ~THREADSTATUS_READY;
__KernelSwitchContext(t, "idle");
}
else
{
WARN_LOG_REPORT(HLE, "UNTESTED - Callback thread deleted during interrupt?");
g_inCbCount = 0;
currentCallbackThreadID = 0;
}
}
// In Advance, we might trigger an interrupt such as vblank.
// If we end up in an interrupt, we don't want to reschedule.
// However, we have to reschedule... damn.
__KernelReSchedule("idle");
}
void __KernelThreadingShutdown()
{
kernelMemory.Free(threadReturnHackAddr);
threadqueue.clear();
threadReadyQueue.clear();
threadEndListeners.clear();
mipsCalls.clear();
threadReturnHackAddr = 0;
cbReturnHackAddr = 0;
currentThread = 0;
intReturnHackAddr = 0;
curModule = 0;
hleCurrentThreadName = NULL;
}
const char *__KernelGetThreadName(SceUID threadID)
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
return t->nt.name;
return "ERROR";
}
u32 __KernelGetWaitValue(SceUID threadID, u32 &error)
{
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
return t->getWaitInfo().waitValue;
}
else
{
ERROR_LOG(HLE, "__KernelGetWaitValue ERROR: thread %i", threadID);
return 0;
}
}
u32 __KernelGetWaitTimeoutPtr(SceUID threadID, u32 &error)
{
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
return t->getWaitInfo().timeoutPtr;
}
else
{
ERROR_LOG(HLE, "__KernelGetWaitTimeoutPtr ERROR: thread %i", threadID);
return 0;
}
}
SceUID __KernelGetWaitID(SceUID threadID, WaitType type, u32 &error)
{
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
return t->getWaitID(type);
}
else
{
ERROR_LOG(HLE, "__KernelGetWaitID ERROR: thread %i", threadID);
return 0;
}
}
SceUID __KernelGetCurrentCallbackID(SceUID threadID, u32 &error)
{
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
return t->currentCallbackId;
else
{
ERROR_LOG(HLE, "__KernelGetCurrentCallbackID ERROR: thread %i", threadID);
return 0;
}
}
u32 sceKernelReferThreadStatus(u32 threadID, u32 statusPtr)
{
if (threadID == 0)
threadID = __KernelGetCurThread();
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (!t)
{
ERROR_LOG(HLE,"sceKernelReferThreadStatus Error %08x", error);
return error;
}
DEBUG_LOG(HLE,"sceKernelReferThreadStatus(%i, %08x)", threadID, statusPtr);
u32 wantedSize = Memory::Read_U32(PARAM(1));
u32 sz = sizeof(NativeThread);
if (wantedSize) {
t->nt.nativeSize = sz = std::min(sz, wantedSize);
}
Memory::Memcpy(statusPtr, &(t->nt), sz);
return 0;
}
// Thanks JPCSP
u32 sceKernelReferThreadRunStatus(u32 threadID, u32 statusPtr)
{
if (threadID == 0)
threadID = __KernelGetCurThread();
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (!t)
{
ERROR_LOG(HLE,"sceKernelReferThreadRunStatus Error %08x", error);
return error;
}
DEBUG_LOG(HLE,"sceKernelReferThreadRunStatus(%i, %08x)", threadID, statusPtr);
if (!Memory::IsValidAddress(statusPtr))
return -1;
Memory::Write_U32(t->nt.status, statusPtr);
Memory::Write_U32(t->nt.currentPriority, statusPtr + 4);
Memory::Write_U32(t->nt.waitType, statusPtr + 8);
Memory::Write_U32(t->nt.waitID, statusPtr + 12);
Memory::Write_U32(t->nt.wakeupCount, statusPtr + 16);
Memory::Write_U32(t->nt.runForClocks.lo, statusPtr + 20);
Memory::Write_U32(t->nt.runForClocks.hi, statusPtr + 24);
Memory::Write_U32(t->nt.numInterruptPreempts, statusPtr + 28);
Memory::Write_U32(t->nt.numThreadPreempts, statusPtr + 32);
Memory::Write_U32(t->nt.numReleases, statusPtr + 36);
return 0;
}
void sceKernelGetThreadExitStatus()
{
SceUID threadID = PARAM(0);
if (threadID == 0)
threadID = __KernelGetCurThread();
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
if (t->nt.status == THREADSTATUS_DORMANT) // TODO: can be dormant before starting, too, need to avoid that
{
DEBUG_LOG(HLE,"sceKernelGetThreadExitStatus(%i)", threadID);
RETURN(t->nt.exitStatus);
}
else
{
RETURN(SCE_KERNEL_ERROR_NOT_DORMANT);
}
}
else
{
ERROR_LOG(HLE,"sceKernelGetThreadExitStatus Error %08x", error);
RETURN(SCE_KERNEL_ERROR_UNKNOWN_THID);
}
}
u32 sceKernelGetThreadmanIdType(u32 uid) {
int type;
if (kernelObjects.GetIDType(uid, &type)) {
DEBUG_LOG(HLE, "%i=sceKernelGetThreadmanIdType(%i)", type, uid);
return type;
} else {
ERROR_LOG(HLE, "sceKernelGetThreadmanIdType(%i) - FAILED", uid);
return SCE_KERNEL_ERROR_ILLEGAL_ARGUMENT;
}
}
u32 sceKernelGetThreadmanIdList(u32 type, u32 readBufPtr, u32 readBufSize, u32 idCountPtr)
{
DEBUG_LOG(HLE, "sceKernelGetThreadmanIdList(%i, %08x, %i, %08x)",
type, readBufPtr, readBufSize, idCountPtr);
if (!Memory::IsValidAddress(readBufPtr))
return SCE_KERNEL_ERROR_ILLEGAL_ARGUMENT;
if (type != SCE_KERNEL_TMID_Thread) {
ERROR_LOG(HLE, "sceKernelGetThreadmanIdList only implemented for threads");
return SCE_KERNEL_ERROR_ILLEGAL_ARGUMENT;
}
for (size_t i = 0; i < std::min((size_t)readBufSize, threadqueue.size()); i++)
{
Memory::Write_U32(threadqueue[i], readBufPtr + (u32)i * 4);
}
Memory::Write_U32((u32)threadqueue.size(), idCountPtr);
return 0;
}
// Saves the current CPU context
void __KernelSaveContext(ThreadContext *ctx, bool vfpuEnabled)
{
memcpy(ctx->r, currentMIPS->r, sizeof(ctx->r));
memcpy(ctx->f, currentMIPS->f, sizeof(ctx->f));
if (vfpuEnabled)
{
memcpy(ctx->v, currentMIPS->v, sizeof(ctx->v));
memcpy(ctx->vfpuCtrl, currentMIPS->vfpuCtrl, sizeof(ctx->vfpuCtrl));
}
ctx->pc = currentMIPS->pc;
ctx->hi = currentMIPS->hi;
ctx->lo = currentMIPS->lo;
ctx->fcr0 = currentMIPS->fcr0;
ctx->fcr31 = currentMIPS->fcr31;
ctx->fpcond = currentMIPS->fpcond;
}
// Loads a CPU context
void __KernelLoadContext(ThreadContext *ctx, bool vfpuEnabled)
{
memcpy(currentMIPS->r, ctx->r, sizeof(ctx->r));
memcpy(currentMIPS->f, ctx->f, sizeof(ctx->f));
if (vfpuEnabled)
{
memcpy(currentMIPS->v, ctx->v, sizeof(ctx->v));
memcpy(currentMIPS->vfpuCtrl, ctx->vfpuCtrl, sizeof(ctx->vfpuCtrl));
}
currentMIPS->pc = ctx->pc;
currentMIPS->hi = ctx->hi;
currentMIPS->lo = ctx->lo;
currentMIPS->fcr0 = ctx->fcr0;
currentMIPS->fcr31 = ctx->fcr31;
currentMIPS->fpcond = ctx->fpcond;
// Reset the llBit, the other thread may have touched memory.
currentMIPS->llBit = 0;
}
u32 __KernelResumeThreadFromWait(SceUID threadID)
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
t->resumeFromWait();
return 0;
}
else
{
ERROR_LOG(HLE, "__KernelResumeThreadFromWait(%d): bad thread: %08x", threadID, error);
return error;
}
}
u32 __KernelResumeThreadFromWait(SceUID threadID, u32 retval)
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
t->resumeFromWait();
t->setReturnValue(retval);
return 0;
}
else
{
ERROR_LOG(HLE, "__KernelResumeThreadFromWait(%d): bad thread: %08x", threadID, error);
return error;
}
}
u32 __KernelResumeThreadFromWait(SceUID threadID, u64 retval)
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
t->resumeFromWait();
t->setReturnValue(retval);
return 0;
}
else
{
ERROR_LOG(HLE, "__KernelResumeThreadFromWait(%d): bad thread: %08x", threadID, error);
return error;
}
}
// Only run when you can safely accept a context switch
// Triggers a waitable event, that is, it wakes up all threads that waits for it
// If any changes were made, it will context switch after the syscall
bool __KernelTriggerWait(WaitType type, int id, bool useRetVal, int retVal, const char *reason, bool dontSwitch)
{
bool doneAnything = false;
u32 error;
for (std::vector<SceUID>::iterator iter = threadqueue.begin(); iter != threadqueue.end(); iter++)
{
Thread *t = kernelObjects.Get<Thread>(*iter, error);
if (t && t->isWaitingFor(type, id))
{
// This thread was waiting for the triggered object.
t->resumeFromWait();
if (useRetVal)
t->setReturnValue((u32)retVal);
doneAnything = true;
if (type == WAITTYPE_THREADEND)
__KernelCancelThreadEndTimeout(*iter);
}
}
// if (doneAnything) // lumines?
{
if (!dontSwitch)
{
// TODO: time waster
hleReSchedule(reason);
}
}
return doneAnything;
}
bool __KernelTriggerWait(WaitType type, int id, const char *reason, bool dontSwitch)
{
return __KernelTriggerWait(type, id, false, 0, reason, dontSwitch);
}
bool __KernelTriggerWait(WaitType type, int id, int retVal, const char *reason, bool dontSwitch)
{
return __KernelTriggerWait(type, id, true, retVal, reason, dontSwitch);
}
// makes the current thread wait for an event
void __KernelWaitCurThread(WaitType type, SceUID waitID, u32 waitValue, u32 timeoutPtr, bool processCallbacks, const char *reason)
{
if (!dispatchEnabled)
{
WARN_LOG_REPORT(HLE, "Ignoring wait, dispatching disabled... right thing to do?");
return;
}
// TODO: Need to defer if in callback?
if (g_inCbCount > 0)
WARN_LOG_REPORT(HLE, "UNTESTED - waiting within a callback, probably bad mojo.");
Thread *thread = __GetCurrentThread();
thread->nt.waitID = waitID;
thread->nt.waitType = type;
__KernelChangeThreadState(thread, THREADSTATUS_WAIT);
thread->nt.numReleases++;
thread->waitInfo.waitValue = waitValue;
thread->waitInfo.timeoutPtr = timeoutPtr;
// TODO: Remove this once all callers are cleaned up.
RETURN(0); //pretend all went OK
// TODO: time waster
if (!reason)
reason = "started wait";
hleReSchedule(processCallbacks, reason);
// TODO: Remove thread from Ready queue?
}
void __KernelWaitCallbacksCurThread(WaitType type, SceUID waitID, u32 waitValue, u32 timeoutPtr)
{
if (!dispatchEnabled)
{
WARN_LOG_REPORT(HLE, "Ignoring wait, dispatching disabled... right thing to do?");
return;
}
Thread *thread = __GetCurrentThread();
thread->nt.waitID = waitID;
thread->nt.waitType = type;
__KernelChangeThreadState(thread, THREADSTATUS_WAIT);
// TODO: Probably not...?
thread->nt.numReleases++;
thread->waitInfo.waitValue = waitValue;
thread->waitInfo.timeoutPtr = timeoutPtr;
__KernelForceCallbacks();
}
void hleScheduledWakeup(u64 userdata, int cyclesLate)
{
SceUID threadID = (SceUID)userdata;
u32 error;
if (__KernelGetWaitID(threadID, WAITTYPE_DELAY, error) == threadID)
__KernelResumeThreadFromWait(threadID);
}
void __KernelScheduleWakeup(SceUID threadID, s64 usFromNow)
{
s64 cycles = usToCycles(usFromNow);
CoreTiming::ScheduleEvent(cycles, eventScheduledWakeup, threadID);
}
void __KernelCancelWakeup(SceUID threadID)
{
CoreTiming::UnscheduleEvent(eventScheduledWakeup, threadID);
}
void hleThreadEndTimeout(u64 userdata, int cyclesLate)
{
SceUID threadID = (SceUID) userdata;
u32 error;
// Just in case it was woken on its own.
if (__KernelGetWaitID(threadID, WAITTYPE_THREADEND, error) != 0)
{
u32 timeoutPtr = __KernelGetWaitTimeoutPtr(threadID, error);
if (Memory::IsValidAddress(timeoutPtr))
Memory::Write_U32(0, timeoutPtr);
__KernelResumeThreadFromWait(threadID, SCE_KERNEL_ERROR_WAIT_TIMEOUT);
}
}
void __KernelScheduleThreadEndTimeout(SceUID threadID, SceUID waitForID, s64 usFromNow)
{
s64 cycles = usToCycles(usFromNow);
CoreTiming::ScheduleEvent(cycles, eventThreadEndTimeout, threadID);
}
void __KernelCancelThreadEndTimeout(SceUID threadID)
{
CoreTiming::UnscheduleEvent(eventThreadEndTimeout, threadID);
}
void __KernelRemoveFromThreadQueue(SceUID threadID)
{
int prio = __KernelGetThreadPrio(threadID);
if (prio != 0)
threadReadyQueue.remove(prio, threadID);
threadqueue.erase(std::remove(threadqueue.begin(), threadqueue.end(), threadID), threadqueue.end());
}
u32 __KernelDeleteThread(SceUID threadID, int exitStatus, const char *reason, bool dontSwitch)
{
__KernelFireThreadEnd(threadID);
__KernelRemoveFromThreadQueue(threadID);
__KernelTriggerWait(WAITTYPE_THREADEND, threadID, exitStatus, reason, dontSwitch);
if (currentThread == threadID)
{
currentThread = 0;
hleCurrentThreadName = NULL;
}
if (currentCallbackThreadID == threadID)
{
currentCallbackThreadID = 0;
g_inCbCount = 0;
}
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
// TODO: Unless they should be run before deletion?
for (int i = 0; i < THREAD_CALLBACK_NUM_TYPES; i++)
readyCallbacksCount -= (int)t->readyCallbacks[i].size();
}
return kernelObjects.Destroy<Thread>(threadID);
}
// Returns NULL if the current thread is fine.
Thread *__KernelNextThread() {
SceUID bestThread;
// If the current thread is running, it's a valid candidate.
Thread *cur = __GetCurrentThread();
if (cur && cur->isRunning())
{
bestThread = threadReadyQueue.pop_first_better(cur->nt.currentPriority);
if (bestThread != 0)
__KernelChangeReadyState(cur, currentThread, true);
}
else
bestThread = threadReadyQueue.pop_first();
// Assume threadReadyQueue has not become corrupt.
if (bestThread != 0)
return kernelObjects.GetFast<Thread>(bestThread);
else
return 0;
}
void __KernelReSchedule(const char *reason)
{
// cancel rescheduling when in interrupt or callback, otherwise everything will be fucked up
if (__IsInInterrupt() || __KernelInCallback() || !__KernelIsDispatchEnabled())
{
reason = "In Interrupt Or Callback";
return;
}
// This may get us running a callback, don't reschedule out of it.
if (__KernelCheckCallbacks())
{
reason = "Began interrupt or callback.";
return;
}
// Execute any pending events while we're doing scheduling.
CoreTiming::AdvanceQuick();
if (__IsInInterrupt() || __KernelInCallback() || !__KernelIsDispatchEnabled())
{
reason = "In Interrupt Or Callback";
return;
}
Thread *nextThread = __KernelNextThread();
if (nextThread)
__KernelSwitchContext(nextThread, reason);
// Otherwise, no need to switch.
}
void __KernelReSchedule(bool doCallbacks, const char *reason)
{
Thread *thread = __GetCurrentThread();
if (doCallbacks)
{
if (thread)
thread->isProcessingCallbacks = doCallbacks;
}
__KernelReSchedule(reason);
if (doCallbacks && thread != NULL && thread->GetUID() == currentThread) {
if (thread->isRunning()) {
thread->isProcessingCallbacks = false;
}
}
}
//////////////////////////////////////////////////////////////////////////
// Thread Management
//////////////////////////////////////////////////////////////////////////
int sceKernelCheckThreadStack()
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(__KernelGetCurThread(), error);
if (t) {
u32 diff = labs((long)((s64)t->stackEnd - (s64)currentMIPS->r[MIPS_REG_SP]));
WARN_LOG(HLE, "%i=sceKernelCheckThreadStack()", diff);
return diff;
} else {
// WTF?
ERROR_LOG(HLE, "sceKernelCheckThreadStack() - not on thread");
return -1;
}
}
void ThreadContext::reset()
{
for (int i = 0; i<32; i++)
{
r[i] = 0;
f[i] = 0.0f;
}
for (int i = 0; i<128; i++)
{
v[i] = 0.0f;
}
for (int i = 0; i<15; i++)
{
vfpuCtrl[i] = 0x00000000;
}
vfpuCtrl[VFPU_CTRL_SPREFIX] = 0xe4; // neutral
vfpuCtrl[VFPU_CTRL_TPREFIX] = 0xe4; // neutral
vfpuCtrl[VFPU_CTRL_DPREFIX] = 0x0; // neutral
vfpuCtrl[VFPU_CTRL_CC] = 0x3f;
vfpuCtrl[VFPU_CTRL_INF4] = 0;
vfpuCtrl[VFPU_CTRL_RCX0] = 0x3f800001;
vfpuCtrl[VFPU_CTRL_RCX1] = 0x3f800002;
vfpuCtrl[VFPU_CTRL_RCX2] = 0x3f800004;
vfpuCtrl[VFPU_CTRL_RCX3] = 0x3f800008;
vfpuCtrl[VFPU_CTRL_RCX4] = 0x3f800000;
vfpuCtrl[VFPU_CTRL_RCX5] = 0x3f800000;
vfpuCtrl[VFPU_CTRL_RCX6] = 0x3f800000;
vfpuCtrl[VFPU_CTRL_RCX7] = 0x3f800000;
fpcond = 0;
fcr0 = 0;
fcr31 = 0;
hi = 0;
lo = 0;
}
void __KernelResetThread(Thread *t)
{
t->context.reset();
t->context.hi = 0;
t->context.lo = 0;
t->context.pc = t->nt.entrypoint;
// TODO: Reset the priority?
t->nt.waitType = WAITTYPE_NONE;
t->nt.waitID = 0;
memset(&t->waitInfo, 0, sizeof(t->waitInfo));
t->nt.exitStatus = SCE_KERNEL_ERROR_NOT_DORMANT;
t->isProcessingCallbacks = false;
t->currentCallbackId = 0;
t->currentMipscallId = 0;
t->pendingMipsCalls.clear();
t->context.r[MIPS_REG_RA] = threadReturnHackAddr; //hack! TODO fix
t->AllocateStack(t->nt.stackSize); // can change the stacksize!
}
Thread *__KernelCreateThread(SceUID &id, SceUID moduleId, const char *name, u32 entryPoint, u32 priority, int stacksize, u32 attr)
{
Thread *t = new Thread;
id = kernelObjects.Create(t);
threadqueue.push_back(id);
threadReadyQueue.prepare(priority);
memset(&t->nt, 0xCD, sizeof(t->nt));
t->nt.entrypoint = entryPoint;
t->nt.nativeSize = sizeof(t->nt);
t->nt.attr = attr;
t->nt.initialPriority = t->nt.currentPriority = priority;
t->nt.stackSize = stacksize;
t->nt.status = THREADSTATUS_DORMANT;
t->nt.numInterruptPreempts = 0;
t->nt.numReleases = 0;
t->nt.numThreadPreempts = 0;
t->nt.runForClocks.lo = 0;
t->nt.runForClocks.hi = 0;
t->nt.wakeupCount = 0;
t->nt.initialStack = 0;
t->nt.waitID = 0;
t->nt.exitStatus = SCE_KERNEL_ERROR_DORMANT;
t->nt.waitType = WAITTYPE_NONE;
if (moduleId)
t->nt.gpreg = __KernelGetModuleGP(moduleId);
else
t->nt.gpreg = 0; // sceKernelStartThread will take care of this.
t->moduleId = moduleId;
strncpy(t->nt.name, name, KERNELOBJECT_MAX_NAME_LENGTH);
t->nt.name[KERNELOBJECT_MAX_NAME_LENGTH] = '\0';
return t;
}
void __KernelSetupRootThread(SceUID moduleID, int args, const char *argp, int prio, int stacksize, int attr)
{
curModule = moduleID;
//grab mips regs
SceUID id;
Thread *thread = __KernelCreateThread(id, moduleID, "root", currentMIPS->pc, prio, stacksize, attr);
__KernelResetThread(thread);
Thread *prevThread = __GetCurrentThread();
if (prevThread && prevThread->isRunning())
__KernelChangeReadyState(currentThread, true);
currentThread = id;
hleCurrentThreadName = "root";
thread->nt.status = THREADSTATUS_RUNNING; // do not schedule
strcpy(thread->nt.name, "root");
__KernelLoadContext(&thread->context, (attr & PSP_THREAD_ATTR_VFPU) != 0);
mipsr4k.r[MIPS_REG_A0] = args;
mipsr4k.r[MIPS_REG_SP] -= 256;
u32 location = mipsr4k.r[MIPS_REG_SP];
mipsr4k.r[MIPS_REG_A1] = location;
for (int i = 0; i < args; i++)
Memory::Write_U8(argp[i], location + i);
}
int __KernelCreateThread(const char *threadName, SceUID moduleID, u32 entry, u32 prio, int stacksize, u32 attr, u32 optionAddr)
{
if (threadName == NULL)
{
ERROR_LOG(HLE, "SCE_KERNEL_ERROR_ERROR = sceKernelCreateThread(): NULL name");
return SCE_KERNEL_ERROR_ERROR;
}
// TODO: PSP actually fails for many of these cases, but trying for compat.
if (stacksize < 0x200 || stacksize >= 0x20000000)
{
WARN_LOG_REPORT(HLE, "sceKernelCreateThread(name=\"%s\"): bogus stack size %08x, using 0x4000", threadName, stacksize);
stacksize = 0x4000;
}
if (prio < 0x08 || prio > 0x77)
WARN_LOG_REPORT(HLE, "sceKernelCreateThread(name=\"%s\"): bogus priority %08x", threadName, prio);
if (!Memory::IsValidAddress(entry))
WARN_LOG_REPORT(HLE, "sceKernelCreateThread(name=\"%s\"): invalid entry %08x", threadName, entry);
// We're assuming all threads created are user threads.
if ((attr & PSP_THREAD_ATTR_KERNEL) == 0)
attr |= PSP_THREAD_ATTR_USER;
SceUID id;
__KernelCreateThread(id, moduleID, threadName, entry, prio, stacksize, attr);
INFO_LOG(HLE, "%i = sceKernelCreateThread(name=\"%s\", entry=%08x, prio=%x, stacksize=%i)", id, threadName, entry, prio, stacksize);
if (optionAddr != 0)
WARN_LOG_REPORT(HLE, "sceKernelCreateThread(name=\"%s\"): unsupported options parameter %08x", threadName, optionAddr);
return id;
}
int sceKernelCreateThread(const char *threadName, u32 entry, u32 prio, int stacksize, u32 attr, u32 optionAddr)
{
SceUID moduleId = curModule;
if (__GetCurrentThread())
moduleId = __GetCurrentThread()->moduleId;
return __KernelCreateThread(threadName, moduleId, entry, prio, stacksize, attr, optionAddr);
}
// int sceKernelStartThread(SceUID threadToStartID, SceSize argSize, void *argBlock)
int sceKernelStartThread(SceUID threadToStartID, u32 argSize, u32 argBlockPtr)
{
if (threadToStartID != currentThread)
{
u32 error;
Thread *startThread = kernelObjects.Get<Thread>(threadToStartID, error);
if (startThread == 0)
{
ERROR_LOG(HLE,"%08x=sceKernelStartThread(thread=%i, argSize=%i, argPtr= %08x): thread does not exist!",
error,threadToStartID,argSize,argBlockPtr)
return error;
}
if (startThread->nt.status != THREADSTATUS_DORMANT)
{
//Not dormant, WTF?
return ERROR_KERNEL_THREAD_IS_NOT_DORMANT;
}
INFO_LOG(HLE,"sceKernelStartThread(thread=%i, argSize=%i, argPtr= %08x )",
threadToStartID,argSize,argBlockPtr);
__KernelResetThread(startThread);
u32 sp = startThread->context.r[MIPS_REG_SP];
if (argBlockPtr && argSize > 0)
{
startThread->context.r[MIPS_REG_A0] = argSize;
startThread->context.r[MIPS_REG_A1] = sp;
}
else
{
startThread->context.r[MIPS_REG_A0] = 0;
startThread->context.r[MIPS_REG_A1] = 0;
}
startThread->context.r[MIPS_REG_GP] = startThread->nt.gpreg;
//now copy argument to stack
for (int i = 0; i < (int)argSize; i++)
Memory::Write_U8(argBlockPtr ? Memory::Read_U8(argBlockPtr + i) : 0, sp + i);
if (!argBlockPtr && argSize > 0) {
WARN_LOG(HLE,"sceKernelStartThread : had NULL arg");
}
Thread *cur = __GetCurrentThread();
// Smaller is better for priority. Only switch if the new thread is better.
if (cur && cur->nt.currentPriority > startThread->nt.currentPriority)
{
// Starting a thread automatically resumes the dispatch thread.
// TODO: Maybe this happens even for worse-priority started threads?
dispatchEnabled = true;
if (cur && cur->isRunning())
cur->nt.status &= ~THREADSTATUS_RUNNING;
__KernelChangeReadyState(cur, currentThread, true);
hleReSchedule("thread started");
}
else if (!dispatchEnabled)
WARN_LOG_REPORT(HLE, "UNTESTED Dispatch disabled while starting worse-priority thread");
__KernelChangeReadyState(startThread, threadToStartID, true);
return 0;
}
else
{
ERROR_LOG(HLE,"thread %i trying to start itself", threadToStartID);
return -1;
}
}
void sceKernelGetThreadStackFreeSize()
{
SceUID threadID = PARAM(0);
Thread *thread;
INFO_LOG(HLE,"sceKernelGetThreadStackFreeSize(%i)", threadID);
if (threadID == 0)
thread = __GetCurrentThread();
else
{
u32 error;
thread = kernelObjects.Get<Thread>(threadID, error);
if (thread == 0)
{
ERROR_LOG(HLE,"sceKernelGetThreadStackFreeSize: invalid thread id %i", threadID);
RETURN(error);
return;
}
}
// Scan the stack for 0xFF
int sz = 0;
for (u32 addr = thread->stackBlock; addr < thread->stackBlock + thread->nt.stackSize; addr++)
{
if (Memory::Read_U8(addr) != 0xFF)
break;
sz++;
}
RETURN(sz & ~3);
}
// Internal function
void __KernelReturnFromThread()
{
Thread *thread = __GetCurrentThread();
_dbg_assert_msg_(HLE, thread != NULL, "Returned from a NULL thread.");
INFO_LOG(HLE,"__KernelReturnFromThread : %s", thread->GetName());
// TEMPORARY HACK: kill the stack of the root thread early:
if (!strcmp(thread->GetName(), "root")) {
thread->FreeStack();
}
thread->nt.exitStatus = currentMIPS->r[2];
__KernelChangeReadyState(thread, currentThread, false);
thread->nt.status = THREADSTATUS_DORMANT;
__KernelFireThreadEnd(thread->GetUID());
__KernelTriggerWait(WAITTYPE_THREADEND, __KernelGetCurThread(), thread->nt.exitStatus, "thread returned", true);
hleReSchedule("thread returned");
// The stack will be deallocated when the thread is deleted.
}
void sceKernelExitThread()
{
Thread *thread = __GetCurrentThread();
_dbg_assert_msg_(HLE, thread != NULL, "Exited from a NULL thread.");
ERROR_LOG(HLE,"sceKernelExitThread FAKED");
__KernelChangeReadyState(thread, currentThread, false);
thread->nt.status = THREADSTATUS_DORMANT;
thread->nt.exitStatus = PARAM(0);
__KernelFireThreadEnd(thread->GetUID());
__KernelTriggerWait(WAITTYPE_THREADEND, __KernelGetCurThread(), thread->nt.exitStatus, "thread exited", true);
hleReSchedule("thread exited");
// The stack will be deallocated when the thread is deleted.
}
void _sceKernelExitThread()
{
Thread *thread = __GetCurrentThread();
_dbg_assert_msg_(HLE, thread != NULL, "_Exited from a NULL thread.");
ERROR_LOG(HLE,"_sceKernelExitThread FAKED");
thread->nt.status = THREADSTATUS_DORMANT;
thread->nt.exitStatus = PARAM(0);
__KernelFireThreadEnd(thread->GetUID());
__KernelTriggerWait(WAITTYPE_THREADEND, __KernelGetCurThread(), thread->nt.exitStatus, "thread _exited", true);
hleReSchedule("thread _exited");
// The stack will be deallocated when the thread is deleted.
}
void sceKernelExitDeleteThread()
{
int threadHandle = __KernelGetCurThread();
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadHandle, error);
if (t)
{
INFO_LOG(HLE,"sceKernelExitDeleteThread()");
__KernelChangeReadyState(t, threadHandle, false);
t->nt.status = THREADSTATUS_DORMANT;
t->nt.exitStatus = PARAM(0);
error = __KernelDeleteThread(threadHandle, PARAM(0), "thread exited with delete", true);
hleReSchedule("thread exited with delete");
RETURN(error);
}
else
{
ERROR_LOG(HLE,"sceKernelExitDeleteThread() ERROR - could not find myself!");
RETURN(error);
}
}
u32 sceKernelSuspendDispatchThread()
{
if (!__InterruptsEnabled())
return SCE_KERNEL_ERROR_CPUDI;
u32 oldDispatchEnabled = dispatchEnabled;
dispatchEnabled = false;
DEBUG_LOG(HLE, "%i=sceKernelSuspendDispatchThread()", oldDispatchEnabled);
return oldDispatchEnabled;
}
u32 sceKernelResumeDispatchThread(u32 enabled)
{
if (!__InterruptsEnabled())
return SCE_KERNEL_ERROR_CPUDI;
u32 oldDispatchEnabled = dispatchEnabled;
dispatchEnabled = enabled != 0;
DEBUG_LOG(HLE, "sceKernelResumeDispatchThread(%i) - from %i", enabled, oldDispatchEnabled);
hleReSchedule("dispatch resumed");
return 0;
}
bool __KernelIsDispatchEnabled()
{
// Dispatch can never be enabled when interrupts are disabled.
return dispatchEnabled && __InterruptsEnabled();
}
int sceKernelRotateThreadReadyQueue(int priority)
{
DEBUG_LOG(HLE, "sceKernelRotateThreadReadyQueue(%x)", priority);
Thread *cur = __GetCurrentThread();
// 0 is special, it means "my current priority."
if (priority == 0)
priority = cur->nt.currentPriority;
if (priority <= 0x07 || priority > 0x77)
return SCE_KERNEL_ERROR_ILLEGAL_PRIORITY;
if (!threadReadyQueue.empty(priority))
{
// In other words, yield to everyone else.
if (cur->nt.currentPriority == priority)
{
threadReadyQueue.push_back(priority, currentThread);
cur->nt.status = THREADSTATUS_READY;
}
// Yield the next thread of this priority to all other threads of same priority.
else
threadReadyQueue.rotate(priority);
hleReSchedule("rotatethreadreadyqueue");
}
return 0;
}
int sceKernelDeleteThread(int threadHandle)
{
if (threadHandle != currentThread)
{
DEBUG_LOG(HLE,"sceKernelDeleteThread(%i)",threadHandle);
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadHandle, error);
if (t)
{
// TODO: Should this reschedule ever? Probably no?
return __KernelDeleteThread(threadHandle, SCE_KERNEL_ERROR_THREAD_TERMINATED, "thread deleted", true);
}
// TODO: Error when doesn't exist?
return 0;
}
else
{
ERROR_LOG(HLE, "Thread \"%s\" tries to delete itself! :(", __GetCurrentThread() ? __GetCurrentThread()->GetName() : "NULL");
return -1;
}
}
int sceKernelTerminateDeleteThread(int threadno)
{
if (threadno != currentThread)
{
INFO_LOG(HLE, "sceKernelTerminateDeleteThread(%i)", threadno);
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadno, error);
if (t)
{
//TODO: should we really reschedule here?
error = __KernelDeleteThread(threadno, SCE_KERNEL_ERROR_THREAD_TERMINATED, "thread terminated with delete", false);
hleReSchedule("thread terminated with delete");
return error;
}
// TODO: Error when doesn't exist?
return 0;
}
else
{
ERROR_LOG(HLE, "Thread \"%s\" trying to delete itself! :(", __GetCurrentThread() ? __GetCurrentThread()->GetName() : "NULL");
return -1;
}
}
int sceKernelTerminateThread(SceUID threadID)
{
if (threadID != currentThread)
{
INFO_LOG(HLE, "sceKernelTerminateThread(%i)", threadID);
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
t->nt.exitStatus = SCE_KERNEL_ERROR_THREAD_TERMINATED;
__KernelChangeReadyState(t, threadID, false);
t->nt.status = THREADSTATUS_DORMANT;
__KernelFireThreadEnd(threadID);
// TODO: Should this really reschedule?
__KernelTriggerWait(WAITTYPE_THREADEND, threadID, t->nt.exitStatus, "thread terminated", true);
}
// TODO: Return an error if it doesn't exist?
return 0;
}
else
{
ERROR_LOG(HLE, "Thread \"%s\" trying to delete itself! :(", __GetCurrentThread() ? __GetCurrentThread()->GetName() : "NULL");
return -1;
}
}
SceUID __KernelGetCurThread()
{
return currentThread;
}
SceUID __KernelGetCurThreadModuleId()
{
Thread *t = __GetCurrentThread();
if (t)
return t->moduleId;
return 0;
}
u32 __KernelGetCurThreadStack()
{
Thread *t = __GetCurrentThread();
if (t)
return t->stackEnd;
return 0;
}
void sceKernelGetThreadId()
{
u32 retVal = currentThread;
// DEBUG_LOG(HLE,"%i = sceKernelGetThreadId()", retVal);
RETURN(retVal);
}
void sceKernelGetThreadCurrentPriority()
{
u32 retVal = __GetCurrentThread()->nt.currentPriority;
DEBUG_LOG(HLE,"%i = sceKernelGetThreadCurrentPriority()", retVal);
RETURN(retVal);
}
void sceKernelChangeCurrentThreadAttr()
{
int clearAttr = PARAM(0);
int setAttr = PARAM(1);
DEBUG_LOG(HLE,"0 = sceKernelChangeCurrentThreadAttr(clear = %08x, set = %08x", clearAttr, setAttr);
Thread *t = __GetCurrentThread();
if (t)
t->nt.attr = (t->nt.attr & ~clearAttr) | setAttr;
else
ERROR_LOG(HLE, "%s(): No current thread?", __FUNCTION__);
RETURN(0);
}
void sceKernelChangeThreadPriority()
{
int id = PARAM(0);
if (id == 0) id = currentThread; //special
u32 error;
Thread *thread = kernelObjects.Get<Thread>(id, error);
if (thread)
{
DEBUG_LOG(HLE,"sceKernelChangeThreadPriority(%i, %i)", id, PARAM(1));
int prio = thread->nt.currentPriority;
threadReadyQueue.remove(prio, id);
thread->nt.currentPriority = PARAM(1);
threadReadyQueue.prepare(thread->nt.currentPriority);
if (thread->isReady())
threadReadyQueue.push_back(thread->nt.currentPriority, id);
RETURN(0);
}
else
{
ERROR_LOG(HLE,"%08x=sceKernelChangeThreadPriority(%i, %i) failed - no such thread", error, id, PARAM(1));
RETURN(error);
}
}
int sceKernelDelayThreadCB(u32 usec)
{
if (usec < 200) usec = 200;
DEBUG_LOG(HLE,"sceKernelDelayThreadCB(%i usec)",usec);
SceUID curThread = __KernelGetCurThread();
__KernelScheduleWakeup(curThread, usec);
__KernelWaitCurThread(WAITTYPE_DELAY, curThread, 0, 0, true, "thread delayed");
return 0;
}
int sceKernelDelayThread(u32 usec)
{
if (usec < 200) usec = 200;
DEBUG_LOG(HLE,"sceKernelDelayThread(%i usec)",usec);
SceUID curThread = __KernelGetCurThread();
__KernelScheduleWakeup(curThread, usec);
__KernelWaitCurThread(WAITTYPE_DELAY, curThread, 0, 0, false, "thread delayed");
return 0;
}
void sceKernelDelaySysClockThreadCB()
{
u32 sysclockAddr = PARAM(0);
if (!Memory::IsValidAddress(sysclockAddr)) {
ERROR_LOG(HLE, "sceKernelDelaySysClockThread(%08x) - bad pointer", sysclockAddr);
RETURN(-1);
return;
}
SceKernelSysClock sysclock;
Memory::ReadStruct(sysclockAddr, &sysclock);
// TODO: Which unit?
u64 usec = sysclock.lo | ((u64)sysclock.hi << 32);
if (usec < 200) usec = 200;
DEBUG_LOG(HLE, "sceKernelDelaySysClockThread(%08x (%llu))", sysclockAddr, usec);
SceUID curThread = __KernelGetCurThread();
__KernelScheduleWakeup(curThread, usec);
__KernelWaitCurThread(WAITTYPE_DELAY, curThread, 0, 0, true, "thread delayed");
}
void sceKernelDelaySysClockThread()
{
u32 sysclockAddr = PARAM(0);
if (!Memory::IsValidAddress(sysclockAddr)) {
ERROR_LOG(HLE, "sceKernelDelaySysClockThread(%08x) - bad pointer", sysclockAddr);
RETURN(-1);
return;
}
SceKernelSysClock sysclock;
Memory::ReadStruct(sysclockAddr, &sysclock);
// TODO: Which unit?
u64 usec = sysclock.lo | ((u64)sysclock.hi << 32);
if (usec < 200) usec = 200;
DEBUG_LOG(HLE, "sceKernelDelaySysClockThread(%08x (%llu))", sysclockAddr, usec);
SceUID curThread = __KernelGetCurThread();
__KernelScheduleWakeup(curThread, usec);
__KernelWaitCurThread(WAITTYPE_DELAY, curThread, 0, 0, false, "thread delayed");
}
u32 __KernelGetThreadPrio(SceUID id)
{
u32 error;
Thread *thread = kernelObjects.Get<Thread>(id, error);
if (thread)
return thread->nt.currentPriority;
return 0;
}
bool __KernelThreadSortPriority(SceUID thread1, SceUID thread2)
{
return __KernelGetThreadPrio(thread1) < __KernelGetThreadPrio(thread2);
}
//////////////////////////////////////////////////////////////////////////
// WAIT/SLEEP ETC
//////////////////////////////////////////////////////////////////////////
void sceKernelWakeupThread()
{
SceUID uid = PARAM(0);
u32 error;
Thread *t = kernelObjects.Get<Thread>(uid, error);
if (t)
{
if (t->nt.waitType != WAITTYPE_SLEEP) {
t->nt.wakeupCount++;
DEBUG_LOG(HLE,"sceKernelWakeupThread(%i) - wakeupCount incremented to %i", uid, t->nt.wakeupCount);
RETURN(0);
} else {
__KernelResumeThreadFromWait(uid);
}
}
else {
ERROR_LOG(HLE,"sceKernelWakeupThread(%i) - bad thread id", uid);
RETURN(error);
}
}
void sceKernelCancelWakeupThread()
{
SceUID uid = PARAM(0);
u32 error;
if (uid == 0) uid = __KernelGetCurThread();
Thread *t = kernelObjects.Get<Thread>(uid, error);
if (t)
{
int wCount = t->nt.wakeupCount;
t->nt.wakeupCount = 0;
DEBUG_LOG(HLE,"sceKernelCancelWakeupThread(%i) - wakeupCount reset from %i", uid, wCount);
RETURN(wCount);
}
else {
ERROR_LOG(HLE,"sceKernelCancelWakeupThread(%i) - bad thread id", uid);
RETURN(error);
}
}
static void __KernelSleepThread(bool doCallbacks) {
Thread *thread = __GetCurrentThread();
if (!thread)
{
ERROR_LOG(HLE, "sceKernelSleepThread*(): bad current thread");
return;
}
if (thread->nt.wakeupCount > 0) {
thread->nt.wakeupCount--;
DEBUG_LOG(HLE, "sceKernelSleepThread() - wakeupCount decremented to %i", thread->nt.wakeupCount);
RETURN(0);
} else {
VERBOSE_LOG(HLE, "sceKernelSleepThread()");
RETURN(0);
__KernelWaitCurThread(WAITTYPE_SLEEP, 0, 0, 0, doCallbacks, "thread slept");
}
}
void sceKernelSleepThread()
{
__KernelSleepThread(false);
}
//the homebrew PollCallbacks
void sceKernelSleepThreadCB()
{
VERBOSE_LOG(HLE, "sceKernelSleepThreadCB()");
__KernelSleepThread(true);
__KernelCheckCallbacks();
}
int sceKernelWaitThreadEnd(SceUID threadID, u32 timeoutPtr)
{
DEBUG_LOG(HLE, "sceKernelWaitThreadEnd(%i, %08x)", threadID, timeoutPtr);
if (threadID == 0 || threadID == currentThread)
return SCE_KERNEL_ERROR_ILLEGAL_THID;
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
if (t->nt.status != THREADSTATUS_DORMANT)
{
if (Memory::IsValidAddress(timeoutPtr))
__KernelScheduleThreadEndTimeout(currentThread, threadID, Memory::Read_U32(timeoutPtr));
__KernelWaitCurThread(WAITTYPE_THREADEND, threadID, 0, timeoutPtr, false, "thread wait end");
}
return t->nt.exitStatus;
}
else
{
ERROR_LOG(HLE, "sceKernelWaitThreadEnd - bad thread %i", threadID);
return error;
}
}
int sceKernelWaitThreadEndCB(SceUID threadID, u32 timeoutPtr)
{
DEBUG_LOG(HLE, "sceKernelWaitThreadEndCB(%i, 0x%X)", threadID, timeoutPtr);
if (threadID == 0 || threadID == currentThread)
return SCE_KERNEL_ERROR_ILLEGAL_THID;
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
hleCheckCurrentCallbacks();
if (t->nt.status != THREADSTATUS_DORMANT)
{
if (Memory::IsValidAddress(timeoutPtr))
__KernelScheduleThreadEndTimeout(currentThread, threadID, Memory::Read_U32(timeoutPtr));
__KernelWaitCurThread(WAITTYPE_THREADEND, threadID, 0, timeoutPtr, true, "thread wait end");
}
return t->nt.exitStatus;
}
else
{
ERROR_LOG(HLE, "sceKernelWaitThreadEndCB - bad thread %i", threadID);
return error;
}
}
int sceKernelReleaseWaitThread(SceUID threadID)
{
DEBUG_LOG(HLE, "sceKernelReleaseWaitThread(%i)", threadID);
if (__KernelInCallback())
WARN_LOG_REPORT(HLE, "UNTESTED sceKernelReleaseWaitThread() might not do the right thing in a callback");
if (threadID == 0 || threadID == currentThread)
return SCE_KERNEL_ERROR_ILLEGAL_THID;
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadID, error);
if (t)
{
if (!t->isWaiting())
return SCE_KERNEL_ERROR_NOT_WAIT;
__KernelResumeThreadFromWait(threadID, SCE_KERNEL_ERROR_RELEASE_WAIT);
hleReSchedule("thread released from wait");
return 0;
}
else
{
ERROR_LOG(HLE, "sceKernelReleaseWaitThread - bad thread %i", threadID);
return error;
}
}
void sceKernelSuspendThread()
{
DEBUG_LOG(HLE,"UNIMPL sceKernelSuspendThread");
RETURN(0);
}
void sceKernelResumeThread()
{
DEBUG_LOG(HLE,"UNIMPL sceKernelResumeThread");
RETURN(0);
}
//////////////////////////////////////////////////////////////////////////
// CALLBACKS
//////////////////////////////////////////////////////////////////////////
// Internal API
u32 __KernelCreateCallback(const char *name, u32 entrypoint, u32 commonArg)
{
Callback *cb = new Callback;
SceUID id = kernelObjects.Create(cb);
cb->nc.size = sizeof(NativeCallback);
strncpy(cb->nc.name, name, 32);
cb->nc.entrypoint = entrypoint;
cb->nc.threadId = __KernelGetCurThread();
cb->nc.commonArgument = commonArg;
cb->nc.notifyCount = 0;
cb->nc.notifyArg = 0;
cb->forceDelete = false;
return id;
}
void sceKernelCreateCallback()
{
u32 entrypoint = PARAM(1);
u32 callbackArg = PARAM(2);
const char *name = Memory::GetCharPointer(PARAM(0));
u32 id = __KernelCreateCallback(name, entrypoint, callbackArg);
DEBUG_LOG(HLE,"%i=sceKernelCreateCallback(name=%s,entry= %08x, callbackArg = %08x)", id, name, entrypoint, callbackArg);
RETURN(id);
}
void sceKernelDeleteCallback()
{
SceUID id = PARAM(0);
DEBUG_LOG(HLE,"sceKernelDeleteCallback(%i)", id);
// TODO: Make sure it's gone from all threads first!
RETURN(kernelObjects.Destroy<Callback>(id));
}
// Rarely used
void sceKernelNotifyCallback()
{
SceUID cbId = PARAM(0);
u32 arg = PARAM(1);
DEBUG_LOG(HLE,"sceKernelNotifyCallback(%i, %i)", cbId, arg);
__KernelNotifyCallback(THREAD_CALLBACK_USER_DEFINED, cbId, arg);
RETURN(0);
}
void sceKernelCancelCallback()
{
SceUID cbId = PARAM(0);
ERROR_LOG(HLE,"sceKernelCancelCallback(%i) - BAD", cbId);
u32 error;
Callback *cb = kernelObjects.Get<Callback>(cbId, error);
if (cb) {
// This is what JPCSP does. Huh?
cb->nc.notifyArg = 0;
RETURN(0);
} else {
ERROR_LOG(HLE,"sceKernelCancelCallback(%i) - bad cbId", cbId);
RETURN(error);
}
RETURN(0);
}
void sceKernelGetCallbackCount()
{
SceUID cbId = PARAM(0);
u32 error;
Callback *cb = kernelObjects.Get<Callback>(cbId, error);
if (cb) {
RETURN(cb->nc.notifyCount);
} else {
ERROR_LOG(HLE,"sceKernelGetCallbackCount(%i) - bad cbId", cbId);
RETURN(error);
}
}
u32 sceKernelExtendThreadStack(u32 cpu, u32 size, u32 entryAddr, u32 entryParameter)
{
ERROR_LOG(HLE,"UNIMPL sceKernelExtendThreadStack(%08x, %08x, %08x, %08x)", cpu, size, entryAddr, entryParameter);
return 0;
}
void sceKernelReferCallbackStatus()
{
SceUID cbId = PARAM(0);
u32 statusAddr = PARAM(1);
u32 error;
Callback *c = kernelObjects.Get<Callback>(cbId, error);
if (c) {
DEBUG_LOG(HLE,"sceKernelReferCallbackStatus(%i, %08x)", cbId, statusAddr);
if (Memory::IsValidAddress(statusAddr)) {
Memory::WriteStruct(statusAddr, &c->nc);
} // else TODO
RETURN(0);
} else {
ERROR_LOG(HLE,"sceKernelReferCallbackStatus(%i, %08x) - bad cbId", cbId, statusAddr);
RETURN(error);
}
}
void ActionAfterMipsCall::run(MipsCall &call) {
u32 error;
Thread *thread = kernelObjects.Get<Thread>(threadID, error);
if (thread) {
__KernelChangeReadyState(thread, threadID, (status & THREADSTATUS_READY) != 0);
thread->nt.status = status;
thread->nt.waitType = waitType;
thread->nt.waitID = waitID;
thread->waitInfo = waitInfo;
thread->isProcessingCallbacks = isProcessingCallbacks;
thread->currentCallbackId = currentCallbackId;
}
if (chainedAction) {
chainedAction->run(call);
delete chainedAction;
}
}
ActionAfterMipsCall *Thread::getRunningCallbackAction()
{
if (this->GetUID() == currentThread && g_inCbCount > 0)
{
MipsCall *call = mipsCalls.get(this->currentMipscallId);
ActionAfterMipsCall *action = 0;
if (call)
action = dynamic_cast<ActionAfterMipsCall *>(call->doAfter);
if (!call || !action)
{
ERROR_LOG(HLE, "Failed to access deferred info for thread: %s", this->nt.name);
return NULL;
}
return action;
}
return NULL;
}
void Thread::setReturnValue(u32 retval)
{
if (this->GetUID() == currentThread) {
if (g_inCbCount) {
u32 callId = this->currentMipscallId;
MipsCall *call = mipsCalls.get(callId);
if (call) {
call->setReturnValue(retval);
} else {
ERROR_LOG(HLE, "Failed to inject return value %08x in thread", retval);
}
} else {
currentMIPS->r[2] = retval;
}
} else {
context.r[2] = retval;
}
}
void Thread::setReturnValue(u64 retval)
{
if (this->GetUID() == currentThread) {
if (g_inCbCount) {
u32 callId = this->currentMipscallId;
MipsCall *call = mipsCalls.get(callId);
if (call) {
call->setReturnValue(retval);
} else {
ERROR_LOG(HLE, "Failed to inject return value %08llx in thread", retval);
}
} else {
currentMIPS->r[2] = retval & 0xFFFFFFFF;
currentMIPS->r[3] = (retval >> 32) & 0xFFFFFFFF;
}
} else {
context.r[2] = retval & 0xFFFFFFFF;
context.r[3] = (retval >> 32) & 0xFFFFFFFF;
}
}
void Thread::resumeFromWait()
{
// Do we need to "inject" it?
ActionAfterMipsCall *action = getRunningCallbackAction();
if (action)
{
action->status &= ~THREADSTATUS_WAIT;
// TODO: What if DORMANT or DEAD?
if (!(action->status & THREADSTATUS_WAITSUSPEND))
action->status = THREADSTATUS_READY;
// Non-waiting threads do not process callbacks.
action->isProcessingCallbacks = false;
}
else
{
this->nt.status &= ~THREADSTATUS_WAIT;
// TODO: What if DORMANT or DEAD?
if (!(this->nt.status & THREADSTATUS_WAITSUSPEND))
__KernelChangeReadyState(this, this->GetUID(), true);
// Non-waiting threads do not process callbacks.
this->isProcessingCallbacks = false;
}
}
bool Thread::isWaitingFor(WaitType type, int id)
{
// Thread might be in a callback right now.
ActionAfterMipsCall *action = getRunningCallbackAction();
if (action)
{
if (action->status & THREADSTATUS_WAIT)
return action->waitType == type && action->waitID == id;
return false;
}
if (this->nt.status & THREADSTATUS_WAIT)
return this->nt.waitType == type && this->nt.waitID == id;
return false;
}
int Thread::getWaitID(WaitType type)
{
// Thread might be in a callback right now.
ActionAfterMipsCall *action = getRunningCallbackAction();
if (action)
{
if (action->waitType == type)
return action->waitID;
return 0;
}
if (this->nt.waitType == type)
return this->nt.waitID;
return 0;
}
ThreadWaitInfo Thread::getWaitInfo()
{
// Thread might be in a callback right now.
ActionAfterMipsCall *action = getRunningCallbackAction();
if (action)
return action->waitInfo;
return this->waitInfo;
}
void __KernelSwitchContext(Thread *target, const char *reason)
{
u32 oldPC = 0;
SceUID oldUID = 0;
const char *oldName = "(none)";
Thread *cur = __GetCurrentThread();
if (cur) // It might just have been deleted.
{
__KernelSaveContext(&cur->context, (cur->nt.attr & PSP_THREAD_ATTR_VFPU) != 0);
oldPC = currentMIPS->pc;
oldUID = cur->GetUID();
// Profile on Windows shows this takes time, skip it.
if (DEBUG_LEVEL <= MAX_LOGLEVEL)
oldName = cur->GetName();
// Normally this is taken care of in __KernelNextThread().
if (cur->isRunning())
__KernelChangeReadyState(cur, oldUID, true);
}
if (target)
{
currentThread = target->GetUID();
hleCurrentThreadName = target->nt.name;
__KernelChangeReadyState(target, currentThread, false);
target->nt.status = (target->nt.status | THREADSTATUS_RUNNING) & ~THREADSTATUS_READY;
}
else
{
currentThread = 0;
hleCurrentThreadName = NULL;
}
__KernelLoadContext(&target->context, (target->nt.attr & PSP_THREAD_ATTR_VFPU) != 0);
bool fromIdle = oldUID == threadIdleID[0] || oldUID == threadIdleID[1];
bool toIdle = currentThread == threadIdleID[0] || currentThread == threadIdleID[1];
if (!(fromIdle && toIdle))
{
DEBUG_LOG(HLE,"Context switched: %s -> %s (%s) (%i - pc: %08x -> %i - pc: %08x)",
oldName, hleCurrentThreadName,
reason,
oldUID, oldPC, currentThread, currentMIPS->pc);
}
// No longer waiting.
target->nt.waitType = WAITTYPE_NONE;
target->nt.waitID = 0;
__KernelExecutePendingMipsCalls(target, true);
}
void __KernelChangeThreadState(Thread *thread, ThreadStatus newStatus) {
if (!thread || thread->nt.status == newStatus)
return;
if (!dispatchEnabled && thread == __GetCurrentThread() && newStatus != THREADSTATUS_RUNNING) {
ERROR_LOG(HLE, "Dispatching suspended, not changing thread state");
return;
}
// TODO: JPSCP has many conditions here, like removing wait timeout actions etc.
// if (thread->nt.status == THREADSTATUS_WAIT && newStatus != THREADSTATUS_WAITSUSPEND) {
__KernelChangeReadyState(thread, thread->GetUID(), (newStatus & THREADSTATUS_READY) != 0);
thread->nt.status = newStatus;
if (newStatus == THREADSTATUS_WAIT) {
if (thread->nt.waitType == WAITTYPE_NONE) {
ERROR_LOG(HLE, "Waittype none not allowed here");
}
// Schedule deletion of stopped threads here. if (thread->isStopped())
}
}
bool __CanExecuteCallbackNow(Thread *thread) {
return g_inCbCount == 0;
}
void __KernelCallAddress(Thread *thread, u32 entryPoint, Action *afterAction, const u32 args[], int numargs, bool reschedAfter, SceUID cbId)
{
_dbg_assert_msg_(HLE, numargs <= 6, "MipsCalls can only take 6 args.");
if (thread) {
ActionAfterMipsCall *after = (ActionAfterMipsCall *) __KernelCreateAction(actionAfterMipsCall);
after->chainedAction = afterAction;
after->threadID = thread->GetUID();
after->status = thread->nt.status;
after->waitType = thread->nt.waitType;
after->waitID = thread->nt.waitID;
after->waitInfo = thread->waitInfo;
after->isProcessingCallbacks = thread->isProcessingCallbacks;
after->currentCallbackId = thread->currentCallbackId;
afterAction = after;
if (thread->nt.waitType != WAITTYPE_NONE) {
// If it's a callback, tell the wait to stop.
if (waitTypeFuncs[thread->nt.waitType].beginFunc != NULL && cbId > 0) {
waitTypeFuncs[thread->nt.waitType].beginFunc(after->threadID, thread->currentCallbackId);
}
// Release thread from waiting
thread->nt.waitType = WAITTYPE_NONE;
}
__KernelChangeThreadState(thread, THREADSTATUS_READY);
}
MipsCall *call = new MipsCall();
call->entryPoint = entryPoint;
for (int i = 0; i < numargs; i++) {
call->args[i] = args[i];
}
call->numArgs = (int) numargs;
call->doAfter = afterAction;
call->tag = "callAddress";
call->cbId = cbId;
u32 callId = mipsCalls.add(call);
bool called = false;
if (!thread || thread == __GetCurrentThread()) {
if (__CanExecuteCallbackNow(thread)) {
thread = __GetCurrentThread();
__KernelChangeThreadState(thread, THREADSTATUS_RUNNING);
__KernelExecuteMipsCallOnCurrentThread(callId, reschedAfter);
called = true;
}
}
if (!called) {
if (thread) {
DEBUG_LOG(HLE, "Making mipscall pending on thread");
thread->pendingMipsCalls.push_back(callId);
} else {
WARN_LOG(HLE, "Ignoring mispcall on NULL/deleted thread");
}
}
}
void __KernelDirectMipsCall(u32 entryPoint, Action *afterAction, u32 args[], int numargs, bool reschedAfter)
{
__KernelCallAddress(__GetCurrentThread(), entryPoint, afterAction, args, numargs, reschedAfter, 0);
}
void __KernelExecuteMipsCallOnCurrentThread(u32 callId, bool reschedAfter)
{
Thread *cur = __GetCurrentThread();
if (cur == NULL)
{
ERROR_LOG(HLE, "__KernelExecuteMipsCallOnCurrentThread(): Bad current thread");
return;
}
if (g_inCbCount > 0) {
WARN_LOG_REPORT(HLE, "__KernelExecuteMipsCallOnCurrentThread(): Already in a callback!");
}
DEBUG_LOG(HLE, "Executing mipscall %i", callId);
MipsCall *call = mipsCalls.get(callId);
// Save the few regs that need saving
call->savedPc = currentMIPS->pc;
call->savedRa = currentMIPS->r[MIPS_REG_RA];
call->savedV0 = currentMIPS->r[MIPS_REG_V0];
call->savedV1 = currentMIPS->r[MIPS_REG_V1];
call->savedIdRegister = currentMIPS->r[MIPS_REG_CALL_ID];
call->savedId = cur->currentMipscallId;
call->reschedAfter = reschedAfter;
// Set up the new state
currentMIPS->pc = call->entryPoint;
currentMIPS->r[MIPS_REG_RA] = __KernelMipsCallReturnAddress();
// We put this two places in case the game overwrites it.
// We may want it later to "inject" return values.
currentMIPS->r[MIPS_REG_CALL_ID] = callId;
cur->currentMipscallId = callId;
for (int i = 0; i < call->numArgs; i++) {
currentMIPS->r[MIPS_REG_A0 + i] = call->args[i];
}
if (call->cbId != 0)
g_inCbCount++;
currentCallbackThreadID = currentThread;
}
void __KernelReturnFromMipsCall()
{
Thread *cur = __GetCurrentThread();
if (cur == NULL)
{
ERROR_LOG(HLE, "__KernelReturnFromMipsCall(): Bad current thread");
return;
}
u32 callId = cur->currentMipscallId;
if (currentMIPS->r[MIPS_REG_CALL_ID] != callId)
WARN_LOG_REPORT(HLE, "__KernelReturnFromMipsCall(): s0 is %08x != %08x", currentMIPS->r[MIPS_REG_CALL_ID], callId);
MipsCall *call = mipsCalls.pop(callId);
// Value returned by the callback function
u32 retVal = currentMIPS->r[MIPS_REG_V0];
DEBUG_LOG(HLE,"__KernelReturnFromMipsCall(), returned %08x", retVal);
// Should also save/restore wait state here.
if (call->doAfter)
{
call->doAfter->run(*call);
delete call->doAfter;
}
currentMIPS->pc = call->savedPc;
currentMIPS->r[MIPS_REG_RA] = call->savedRa;
currentMIPS->r[MIPS_REG_V0] = call->savedV0;
currentMIPS->r[MIPS_REG_V1] = call->savedV1;
currentMIPS->r[MIPS_REG_CALL_ID] = call->savedIdRegister;
cur->currentMipscallId = call->savedId;
if (call->cbId != 0)
g_inCbCount--;
currentCallbackThreadID = 0;
if (cur->nt.waitType != WAITTYPE_NONE)
{
if (waitTypeFuncs[cur->nt.waitType].endFunc != NULL && call->cbId > 0)
waitTypeFuncs[cur->nt.waitType].endFunc(cur->GetUID(), cur->currentCallbackId, currentMIPS->r[MIPS_REG_V0]);
}
// yeah! back in the real world, let's keep going. Should we process more callbacks?
if (!__KernelExecutePendingMipsCalls(cur, call->reschedAfter))
{
// Sometimes, we want to stay on the thread.
int threadReady = cur->nt.status & (THREADSTATUS_READY | THREADSTATUS_RUNNING);
if (call->reschedAfter || threadReady == 0)
__KernelReSchedule("return from callback");
}
delete call;
}
// First arg must be current thread, passed to avoid perf cost of a lookup.
bool __KernelExecutePendingMipsCalls(Thread *thread, bool reschedAfter)
{
_dbg_assert_msg_(HLE, thread->GetUID() == __KernelGetCurThread(), "__KernelExecutePendingMipsCalls() should be called only with the current thread.");
if (thread->pendingMipsCalls.empty()) {
// Nothing to do
return false;
}
if (__CanExecuteCallbackNow(thread))
{
// Pop off the first pending mips call
u32 callId = thread->pendingMipsCalls.front();
thread->pendingMipsCalls.pop_front();
__KernelExecuteMipsCallOnCurrentThread(callId, reschedAfter);
return true;
}
return false;
}
// Executes the callback, when it next is context switched to.
void __KernelRunCallbackOnThread(SceUID cbId, Thread *thread, bool reschedAfter)
{
u32 error;
Callback *cb = kernelObjects.Get<Callback>(cbId, error);
if (!cb) {
ERROR_LOG(HLE, "__KernelRunCallbackOnThread: Bad cbId %i", cbId);
return;
}
DEBUG_LOG(HLE, "__KernelRunCallbackOnThread: Turning callback %i into pending mipscall", cbId);
// Alright, we're on the right thread
// Should save/restore wait state?
const u32 args[] = {(u32) cb->nc.notifyCount, (u32) cb->nc.notifyArg, cb->nc.commonArgument};
// Clear the notify count / arg
cb->nc.notifyCount = 0;
cb->nc.notifyArg = 0;
ActionAfterCallback *action = (ActionAfterCallback *) __KernelCreateAction(actionAfterCallback);
if (action != NULL)
action->setCallback(cbId);
else
ERROR_LOG(HLE, "Something went wrong creating a restore action for a callback.");
__KernelCallAddress(thread, cb->nc.entrypoint, action, args, 3, reschedAfter, cbId);
}
void ActionAfterCallback::run(MipsCall &call) {
if (cbId != -1) {
u32 error;
Callback *cb = kernelObjects.Get<Callback>(cbId, error);
if (cb)
{
Thread *t = kernelObjects.Get<Thread>(cb->nc.threadId, error);
if (t)
{
// Check for other callbacks to run (including ones this callback scheduled.)
__KernelCheckThreadCallbacks(t, true);
}
DEBUG_LOG(HLE, "Left callback %i - %s", cbId, cb->nc.name);
// Callbacks that don't return 0 are deleted. But should this be done here?
if (currentMIPS->r[MIPS_REG_V0] != 0 || cb->forceDelete)
{
DEBUG_LOG(HLE, "ActionAfterCallback::run(): Callback returned non-zero, gets deleted!");
kernelObjects.Destroy<Callback>(cbId);
}
}
}
}
bool __KernelCurHasReadyCallbacks() {
if (readyCallbacksCount == 0)
return false;
Thread *thread = __GetCurrentThread();
for (int i = 0; i < THREAD_CALLBACK_NUM_TYPES; i++) {
if (thread->readyCallbacks[i].size()) {
return true;
}
}
return false;
}
// Check callbacks on the current thread only.
// Returns true if any callbacks were processed on the current thread.
bool __KernelCheckThreadCallbacks(Thread *thread, bool force)
{
if (!thread || (!thread->isProcessingCallbacks && !force))
return false;
for (int i = 0; i < THREAD_CALLBACK_NUM_TYPES; i++) {
if (thread->readyCallbacks[i].size()) {
SceUID readyCallback = thread->readyCallbacks[i].front();
thread->readyCallbacks[i].pop_front();
readyCallbacksCount--;
// If the callback was deleted, we're good. Just skip it.
if (kernelObjects.IsValid(readyCallback))
{
__KernelRunCallbackOnThread(readyCallback, thread, !force); // makes pending
return true;
}
else
{
WARN_LOG(HLE, "Ignoring deleted callback %08x", readyCallback);
}
}
}
return false;
}
// Checks for callbacks on all threads
bool __KernelCheckCallbacks() {
// Let's not check every thread all the time, callbacks are fairly uncommon.
if (readyCallbacksCount == 0) {
return false;
}
if (readyCallbacksCount < 0) {
ERROR_LOG(HLE, "readyCallbacksCount became negative: %i", readyCallbacksCount);
}
// SceUID currentThread = __KernelGetCurThread();
// __GetCurrentThread()->isProcessingCallbacks = true;
// do {
bool processed = false;
u32 error;
for (std::vector<SceUID>::iterator iter = threadqueue.begin(); iter != threadqueue.end(); iter++) {
Thread *thread = kernelObjects.Get<Thread>(*iter, error);
if (thread && __KernelCheckThreadCallbacks(thread, false)) {
processed = true;
}
}
// } while (processed && currentThread == __KernelGetCurThread());
if (processed)
return __KernelExecutePendingMipsCalls(__GetCurrentThread(), true);
return processed;
}
bool __KernelForceCallbacks()
{
// Let's not check every thread all the time, callbacks are fairly uncommon.
if (readyCallbacksCount == 0) {
return false;
}
if (readyCallbacksCount < 0) {
ERROR_LOG(HLE, "readyCallbacksCount became negative: %i", readyCallbacksCount);
}
Thread *curThread = __GetCurrentThread();
bool callbacksProcessed = __KernelCheckThreadCallbacks(curThread, true);
if (callbacksProcessed)
__KernelExecutePendingMipsCalls(curThread, false);
return callbacksProcessed;
}
void sceKernelCheckCallback()
{
// Start with yes.
RETURN(1);
bool callbacksProcessed = __KernelForceCallbacks();
if (callbacksProcessed) {
DEBUG_LOG(HLE,"sceKernelCheckCallback() - processed a callback.");
} else {
RETURN(0);
}
}
bool __KernelInCallback()
{
return (g_inCbCount != 0);
}
u32 __KernelRegisterCallback(RegisteredCallbackType type, SceUID cbId)
{
Thread *t = __GetCurrentThread();
if (cbId > 0 && t->registeredCallbacks[type].find(cbId) == t->registeredCallbacks[type].end()) {
t->registeredCallbacks[type].insert(cbId);
return 0;
} else {
return SCE_KERNEL_ERROR_INVAL;
}
}
u32 __KernelUnregisterCallback(RegisteredCallbackType type, SceUID cbId)
{
Thread *t = __GetCurrentThread();
if (t->registeredCallbacks[type].find(cbId) != t->registeredCallbacks[type].end()) {
t->registeredCallbacks[type].erase(cbId);
return 0;
} else {
return 0x80010016;
}
}
void __KernelNotifyCallback(RegisteredCallbackType type, SceUID cbId, int notifyArg)
{
u32 error;
Callback *cb = kernelObjects.Get<Callback>(cbId, error);
if (!cb) {
// Yeah, we're screwed, this shouldn't happen.
ERROR_LOG(HLE, "__KernelNotifyCallback - invalid callback %08x", cbId);
return;
}
cb->nc.notifyCount++;
cb->nc.notifyArg = notifyArg;
Thread *t = kernelObjects.Get<Thread>(cb->nc.threadId, error);
std::list<SceUID> &readyCallbacks = t->readyCallbacks[type];
auto iter = std::find(readyCallbacks.begin(), readyCallbacks.end(), cbId);
if (iter == readyCallbacks.end())
{
t->readyCallbacks[type].push_back(cbId);
readyCallbacksCount++;
}
}
// TODO: If cbId == -1, notify the callback ID on all threads that have it.
u32 __KernelNotifyCallbackType(RegisteredCallbackType type, SceUID cbId, int notifyArg)
{
u32 error;
for (std::vector<SceUID>::iterator iter = threadqueue.begin(); iter != threadqueue.end(); iter++) {
Thread *t = kernelObjects.Get<Thread>(*iter, error);
if (!t)
continue;
for (std::set<SceUID>::iterator citer = t->registeredCallbacks[type].begin(); citer != t->registeredCallbacks[type].end(); citer++) {
if (cbId == -1 || cbId == *citer) {
__KernelNotifyCallback(type, *citer, notifyArg);
}
}
}
// checkCallbacks on other threads?
return 0;
}
void __KernelRegisterWaitTypeFuncs(WaitType type, WaitBeginCallbackFunc beginFunc, WaitEndCallbackFunc endFunc)
{
waitTypeFuncs[type].beginFunc = beginFunc;
waitTypeFuncs[type].endFunc = endFunc;
}
std::vector<DebugThreadInfo> GetThreadsInfo()
{
std::vector<DebugThreadInfo> threadList;
u32 error;
for (std::vector<SceUID>::iterator iter = threadqueue.begin(); iter != threadqueue.end(); iter++)
{
Thread *t = kernelObjects.Get<Thread>(*iter, error);
if (!t)
continue;
DebugThreadInfo info;
info.id = *iter;
strncpy(info.name,t->GetName(),KERNELOBJECT_MAX_NAME_LENGTH);
info.name[KERNELOBJECT_MAX_NAME_LENGTH] = 0;
info.status = t->nt.status;
info.entrypoint = t->nt.entrypoint;
if(*iter == currentThread)
info.curPC = currentMIPS->pc;
else
info.curPC = t->context.pc;
info.isCurrent = (*iter == currentThread);
threadList.push_back(info);
}
return threadList;
}
void __KernelChangeThreadState(SceUID threadId, ThreadStatus newStatus)
{
u32 error;
Thread *t = kernelObjects.Get<Thread>(threadId, error);
if (!t)
return;
__KernelChangeThreadState(t, newStatus);
}