kernel: Rewrite pthread emulation (#1440)

* libkernel: Cleanup some function places

* kernel: Refactor thread functions

* kernel: It builds

* kernel: Fix a bunch of bugs, kernel thread heap

* kernel: File cleanup pt1

* File cleanup pt2

* File cleanup pt3

* File cleanup pt4

* kernel: Add missing funcs

* kernel: Add basic exceptions for linux

* gnmdriver: Add workload functions

* kernel: Fix new pthreads code on macOS. (#1441)

* kernel: Downgrade edeadlk to log

* gnmdriver: Add sceGnmSubmitCommandBuffersForWorkload

* exception: Add context register population for macOS. (#1444)

* kernel: Pthread rewrite touchups for Windows

* kernel: Multiplatform thread implementation

* mutex: Remove spamming log

* pthread_spec: Make assert into a log

* pthread_spec: Zero initialize array

* Attempt to fix non-Windows builds

* hotfix: change incorrect NID for scePthreadAttrSetaffinity

* scePthreadAttrSetaffinity implementation

* Attempt to fix Linux

* windows: Address a bunch of address space problems

* address_space: Fix unmap of region surrounded by placeholders

* libs: Reduce logging

* pthread: Implement condvar with waitable atomics and sleepqueue

* sleepq: Separate and make faster

* time: Remove delay execution

* Causes high cpu usage in Tohou Luna Nights

* kernel: Cleanup files again

* pthread: Add missing include

* semaphore: Use binary_semaphore instead of condvar

* Seems more reliable

* libraries/sysmodule: log module on `sceSysmoduleIsLoaded`

* libraries/kernel: implement `scePthreadSetPrio`

---------

Co-authored-by: squidbus <175574877+squidbus@users.noreply.github.com>
Co-authored-by: Daniel R. <47796739+polybiusproxy@users.noreply.github.com>

src/common/debug.h

@@ -41,7 +41,7 @@ enum MarkersPalette : int {
 #define RENDERER_TRACE ZoneScopedC(RendererMarkerColor)
 #define HLE_TRACE ZoneScopedC(HleMarkerColor)
 
-#define TRACE_HINT(str) ZoneText(str.c_str(), str.size())
+#define TRACE_HINT(str) ZoneText(str.data(), str.size())
 
 #define TRACE_WARN(msg)                                                                            \
     [](const auto& msg) { TracyMessageC(msg.c_str(), msg.size(), tracy::Color::DarkOrange); }(msg);

src/common/slab_heap.h

@@ -0,0 +1,163 @@
+// SPDX-FileCopyrightText: Copyright 2024 shadPS4 Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#pragma once
+
+#include <atomic>
+#include "common/assert.h"
+#include "common/spin_lock.h"
+
+namespace Common {
+
+class SlabHeapImpl {
+public:
+    struct Node {
+        Node* next{};
+    };
+
+public:
+    constexpr SlabHeapImpl() = default;
+
+    void Initialize() {
+        ASSERT(m_head == nullptr);
+    }
+
+    Node* GetHead() const {
+        return m_head;
+    }
+
+    void* Allocate() {
+        m_lock.lock();
+
+        Node* ret = m_head;
+        if (ret != nullptr) {
+            m_head = ret->next;
+        }
+
+        m_lock.unlock();
+        return ret;
+    }
+
+    void Free(void* obj) {
+        m_lock.lock();
+
+        Node* node = static_cast<Node*>(obj);
+        node->next = m_head;
+        m_head = node;
+
+        m_lock.unlock();
+    }
+
+private:
+    std::atomic<Node*> m_head{};
+    Common::SpinLock m_lock;
+};
+
+class SlabHeapBase : protected SlabHeapImpl {
+private:
+    size_t m_obj_size{};
+    uintptr_t m_peak{};
+    uintptr_t m_start{};
+    uintptr_t m_end{};
+
+public:
+    constexpr SlabHeapBase() = default;
+
+    bool Contains(uintptr_t address) const {
+        return m_start <= address && address < m_end;
+    }
+
+    void Initialize(size_t obj_size, void* memory, size_t memory_size) {
+        // Ensure we don't initialize a slab using null memory.
+        ASSERT(memory != nullptr);
+
+        // Set our object size.
+        m_obj_size = obj_size;
+
+        // Initialize the base allocator.
+        SlabHeapImpl::Initialize();
+
+        // Set our tracking variables.
+        const size_t num_obj = (memory_size / obj_size);
+        m_start = reinterpret_cast<uintptr_t>(memory);
+        m_end = m_start + num_obj * obj_size;
+        m_peak = m_start;
+
+        // Free the objects.
+        u8* cur = reinterpret_cast<u8*>(m_end);
+
+        for (size_t i = 0; i < num_obj; i++) {
+            cur -= obj_size;
+            SlabHeapImpl::Free(cur);
+        }
+    }
+
+    size_t GetSlabHeapSize() const {
+        return (m_end - m_start) / this->GetObjectSize();
+    }
+
+    size_t GetObjectSize() const {
+        return m_obj_size;
+    }
+
+    void* Allocate() {
+        void* obj = SlabHeapImpl::Allocate();
+        return obj;
+    }
+
+    void Free(void* obj) {
+        // Don't allow freeing an object that wasn't allocated from this heap.
+        const bool contained = this->Contains(reinterpret_cast<uintptr_t>(obj));
+        ASSERT(contained);
+        SlabHeapImpl::Free(obj);
+    }
+
+    size_t GetObjectIndex(const void* obj) const {
+        return (reinterpret_cast<uintptr_t>(obj) - m_start) / this->GetObjectSize();
+    }
+
+    size_t GetPeakIndex() const {
+        return this->GetObjectIndex(reinterpret_cast<const void*>(m_peak));
+    }
+
+    uintptr_t GetSlabHeapAddress() const {
+        return m_start;
+    }
+
+    size_t GetNumRemaining() const {
+        // Only calculate the number of remaining objects under debug configuration.
+        return 0;
+    }
+};
+
+template <typename T>
+class SlabHeap final : public SlabHeapBase {
+private:
+    using BaseHeap = SlabHeapBase;
+
+public:
+    constexpr SlabHeap() = default;
+
+    void Initialize(void* memory, size_t memory_size) {
+        BaseHeap::Initialize(sizeof(T), memory, memory_size);
+    }
+
+    T* Allocate() {
+        T* obj = static_cast<T*>(BaseHeap::Allocate());
+
+        if (obj != nullptr) [[likely]] {
+            std::construct_at(obj);
+        }
+        return obj;
+    }
+
+    void Free(T* obj) {
+        BaseHeap::Free(obj);
+    }
+
+    size_t GetObjectIndex(const T* obj) const {
+        return BaseHeap::GetObjectIndex(obj);
+    }
+};
+
+} // namespace Common

src/common/slot_vector.h

@@ -3,10 +3,7 @@
 
 #pragma once
 
-#include <bit>
-#include <compare>
 #include <numeric>
-#include <type_traits>
 #include <utility>
 #include <vector>
 #include "common/assert.h"

src/common/spin_lock.cpp

@@ -0,0 +1,53 @@
+// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include "common/spin_lock.h"
+
+#if _MSC_VER
+#include <intrin.h>
+#if _M_AMD64
+#define __x86_64__ 1
+#endif
+#if _M_ARM64
+#define __aarch64__ 1
+#endif
+#else
+#if __x86_64__
+#include <xmmintrin.h>
+#endif
+#endif
+
+namespace {
+
+void ThreadPause() {
+#if __x86_64__
+    _mm_pause();
+#elif __aarch64__ && _MSC_VER
+    __yield();
+#elif __aarch64__
+    asm("yield");
+#endif
+}
+
+} // Anonymous namespace
+
+namespace Common {
+
+void SpinLock::lock() {
+    while (lck.test_and_set(std::memory_order_acquire)) {
+        ThreadPause();
+    }
+}
+
+void SpinLock::unlock() {
+    lck.clear(std::memory_order_release);
+}
+
+bool SpinLock::try_lock() {
+    if (lck.test_and_set(std::memory_order_acquire)) {
+        return false;
+    }
+    return true;
+}
+
+} // namespace Common

src/common/spin_lock.h

@@ -0,0 +1,33 @@
+// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#pragma once
+
+#include <atomic>
+
+namespace Common {
+
+/**
+ * SpinLock class
+ * a lock similar to mutex that forces a thread to spin wait instead calling the
+ * supervisor. Should be used on short sequences of code.
+ */
+class SpinLock {
+public:
+    SpinLock() = default;
+
+    SpinLock(const SpinLock&) = delete;
+    SpinLock& operator=(const SpinLock&) = delete;
+
+    SpinLock(SpinLock&&) = delete;
+    SpinLock& operator=(SpinLock&&) = delete;
+
+    void lock();
+    void unlock();
+    [[nodiscard]] bool try_lock();
+
+private:
+    std::atomic_flag lck = ATOMIC_FLAG_INIT;
+};
+
+} // namespace Common