veriniseHEADmain

5 files changed, 208 insertions, 2 deletions

diff --git a/src/cs2/hack.cpp b/src/cs2/hack.cpp
index c98337a..a9be030 100644
--- a/src/cs2/hack.cpp
+++ b/src/cs2/hack.cpp
@@ -20,7 +20,7 @@ PROCESS64* hack_init() {
 }
 
 bool hack_run( PROCESS64* p ) {
-  perf_run_metric( perf_loop_start );
+  perf_run_metric( perf_loop_begin );
   CS2* cs = (CS2*)p;
 
   for( I32 i = 0; i < 64; ++i ) {
diff --git a/src/heavens-gate.vcxproj b/src/heavens-gate.vcxproj
index e2e0742..ed749c5 100644
--- a/src/heavens-gate.vcxproj
+++ b/src/heavens-gate.vcxproj
@@ -236,6 +236,7 @@
     <ClInclude Include="vec3.h" />
     <ClInclude Include="winintern.h" />
     <ClInclude Include="x86.h" />
+    <ClInclude Include="xorstr.hpp" />
   </ItemGroup>
   <ItemGroup>
     <Image Include="..\resource\heavens-gate-ico.ico" />
diff --git a/src/heavens-gate.vcxproj.filters b/src/heavens-gate.vcxproj.filters
index 25a909f..ecd163a 100644
--- a/src/heavens-gate.vcxproj.filters
+++ b/src/heavens-gate.vcxproj.filters
@@ -96,6 +96,9 @@
     <ClInclude Include="color.h">
       <Filter>Util</Filter>
     </ClInclude>
+    <ClInclude Include="xorstr.hpp">
+      <Filter>Util</Filter>
+    </ClInclude>
   </ItemGroup>
   <ItemGroup>
     <Filter Include="Console">
diff --git a/src/menu.cpp b/src/menu.cpp
index 27c8bae..c94fde6 100644
--- a/src/menu.cpp
+++ b/src/menu.cpp
@@ -6,6 +6,8 @@
 
 #include "cs2/hack.h"
 
+#include "xorstr.hpp"
+
 PROCESS64* cs2p;
 
 I32 perf_tickrate = 2048;
@@ -351,7 +353,7 @@ void menu_show_ui( PROCESS64 *p ) {
   show_paging( 1 );
 
   con_set_bottomline_text(
-    "LOCALPLAYER: %08X | FLAGS: %08X | menu",
+    xorstr( "LOCALPLAYER: %08X | FLAGS: %08X | menu" ).get(),
     0x0,0x0
   );
 }
\ No newline at end of file
diff --git a/src/xorstr.hpp b/src/xorstr.hpp
new file mode 100644
index 0000000..1bcea88
--- /dev/null
+++ b/src/xorstr.hpp
@@ -0,0 +1,200 @@
+#ifndef JM_XORSTR_HPP
+#define JM_XORSTR_HPP
+
+#if defined(_M_ARM64) || defined(__aarch64__) || defined(_M_ARM) || defined(__arm__)
+#include <arm_neon.h>
+#elif defined(_M_X64) || defined(__amd64__) || defined(_M_IX86) || defined(__i386__)
+#include <immintrin.h>
+#else
+#error Unsupported platform
+#endif
+
+#include <cstdint>
+#include <cstddef>
+#include <utility>
+#include <type_traits>
+
+#define JM_XORSTR_DISABLE_AVX_INTRINSICS
+
+#define xorstr(str) ::jm::xor_string([]() { return str; }, std::integral_constant<std::size_t, sizeof(str) / sizeof(*str)>{}, std::make_index_sequence<::jm::detail::_buffer_size<sizeof(str)>()>{})
+#define xorstr_(str) xorstr(str).crypt_get()
+
+#ifdef _MSC_VER
+#define XORSTR_FORCEINLINE __forceinline
+#else
+#define XORSTR_FORCEINLINE __attribute__((always_inline)) inline
+#endif
+
+namespace jm {
+    namespace detail {
+        template<std::size_t Size>
+        XORSTR_FORCEINLINE constexpr std::size_t _buffer_size() { return ((Size / 16) + (Size % 16 != 0)) * 2; }
+
+        template<std::uint32_t Seed>
+        XORSTR_FORCEINLINE constexpr std::uint32_t key4() noexcept
+        {
+            std::uint32_t value = Seed;
+            for(char c : __TIME__)
+                value = static_cast<std::uint32_t>((value ^ c) * 16777619ull);
+            return value;
+        }
+
+        template<std::size_t S>
+        XORSTR_FORCEINLINE constexpr std::uint64_t key8()
+        {
+            constexpr auto first_part  = key4<2166136261 + S>();
+            constexpr auto second_part = key4<first_part>();
+            return (static_cast<std::uint64_t>(first_part) << 32) | second_part;
+        }
+
+        template<std::size_t N, class CharT>
+        XORSTR_FORCEINLINE constexpr std::uint64_t
+        load_xored_str8(std::uint64_t key, std::size_t idx, const CharT* str) noexcept
+        {
+            using cast_type = typename std::make_unsigned<CharT>::type;
+            constexpr auto value_size = sizeof(CharT);
+            constexpr auto idx_offset = 8 / value_size;
+
+            std::uint64_t value = key;
+            for(std::size_t i = 0; i < idx_offset && i + idx * idx_offset < N; ++i)
+                value ^= (std::uint64_t{ static_cast<cast_type>(str[i + idx * idx_offset]) } << ((i % idx_offset) * 8 * value_size));
+
+            return value;
+        }
+
+        XORSTR_FORCEINLINE std::uint64_t load_from_reg(std::uint64_t value) noexcept
+        {
+#if defined(__clang__) || defined(__GNUC__)
+            asm("" : "=r"(value) : "0"(value) :);
+            return value;
+#else
+            volatile std::uint64_t reg = value;
+            return reg;
+#endif
+        }
+    }
+
+    template<class CharT, std::size_t Size, class Keys, class Indices>
+    class xor_string;
+
+    template<class CharT, std::size_t Size, std::uint64_t... Keys, std::size_t... Indices>
+    class xor_string<CharT, Size, std::integer_sequence<std::uint64_t, Keys...>, std::index_sequence<Indices...>> {
+#ifndef JM_XORSTR_DISABLE_AVX_INTRINSICS
+        constexpr static inline std::uint64_t alignment = ((Size > 16) ? 32 : 16);    
+#else
+        constexpr static inline std::uint64_t alignment = 16;
+#endif
+
+        alignas(alignment) std::uint64_t _storage[sizeof...(Keys)];
+
+    public:
+        using value_type    = CharT;
+        using size_type     = std::size_t;
+        using pointer       = CharT*;
+        using const_pointer = const CharT*;
+
+        template<class L>
+        XORSTR_FORCEINLINE xor_string(L l, std::integral_constant<std::size_t, Size>, std::index_sequence<Indices...>) noexcept
+            : _storage{ ::jm::detail::load_from_reg((std::integral_constant<std::uint64_t, detail::load_xored_str8<Size>(Keys, Indices, l())>::value))... }
+        {}
+
+        XORSTR_FORCEINLINE constexpr size_type size() const noexcept { return Size - 1; }
+
+        XORSTR_FORCEINLINE void crypt() noexcept
+        {
+#if defined(__clang__)
+            alignas(alignment)
+                std::uint64_t arr[]{ ::jm::detail::load_from_reg(Keys)... };
+            std::uint64_t*    keys =
+                (std::uint64_t*)::jm::detail::load_from_reg((std::uint64_t)arr);
+#else
+            alignas(alignment) std::uint64_t keys[]{ ::jm::detail::load_from_reg(Keys)... };
+#endif
+
+#if defined(_M_ARM64) || defined(__aarch64__) || defined(_M_ARM) || defined(__arm__)
+#if defined(__clang__)
+            ((Indices >= sizeof(_storage) / 16 ? static_cast<void>(0) : __builtin_neon_vst1q_v(
+                                    reinterpret_cast<uint64_t*>(_storage) + Indices * 2,
+                                    veorq_u64(__builtin_neon_vld1q_v(reinterpret_cast<const uint64_t*>(_storage) + Indices * 2, 51),
+                                              __builtin_neon_vld1q_v(reinterpret_cast<const uint64_t*>(keys) + Indices * 2, 51)),
+                                    51)), ...);
+#else // GCC, MSVC
+            ((Indices >= sizeof(_storage) / 16 ? static_cast<void>(0) : vst1q_u64(
+                        reinterpret_cast<uint64_t*>(_storage) + Indices * 2,
+                        veorq_u64(vld1q_u64(reinterpret_cast<const uint64_t*>(_storage) + Indices * 2),
+                                  vld1q_u64(reinterpret_cast<const uint64_t*>(keys) + Indices * 2)))), ...);
+#endif
+#elif !defined(JM_XORSTR_DISABLE_AVX_INTRINSICS)
+            ((Indices >= sizeof(_storage) / 32 ? static_cast<void>(0) : _mm256_store_si256(
+                reinterpret_cast<__m256i*>(_storage) + Indices,
+                _mm256_xor_si256(
+                    _mm256_load_si256(reinterpret_cast<const __m256i*>(_storage) + Indices),
+                    _mm256_load_si256(reinterpret_cast<const __m256i*>(keys) + Indices)))), ...);
+
+            if constexpr(sizeof(_storage) % 32 != 0)
+                _mm_store_si128(
+                    reinterpret_cast<__m128i*>(_storage + sizeof...(Keys) - 2),
+                    _mm_xor_si128(_mm_load_si128(reinterpret_cast<const __m128i*>(_storage + sizeof...(Keys) - 2)),
+                                  _mm_load_si128(reinterpret_cast<const __m128i*>(keys + sizeof...(Keys) - 2))));
+#else
+        ((Indices >= sizeof(_storage) / 16 ? static_cast<void>(0) : _mm_store_si128(
+            reinterpret_cast<__m128i*>(_storage) + Indices,
+            _mm_xor_si128(_mm_load_si128(reinterpret_cast<const __m128i*>(_storage) + Indices),
+                          _mm_load_si128(reinterpret_cast<const __m128i*>(keys) + Indices)))), ...);
+#endif
+        }
+
+        XORSTR_FORCEINLINE const_pointer get() const noexcept { return reinterpret_cast<const_pointer>(_storage); }
+
+        XORSTR_FORCEINLINE pointer get() noexcept { return reinterpret_cast<pointer>(_storage); }
+
+        XORSTR_FORCEINLINE pointer crypt_get() noexcept
+        {
+#if defined(__clang__)
+            alignas(alignment)
+                std::uint64_t arr[]{ ::jm::detail::load_from_reg(Keys)... };
+            std::uint64_t*    keys =
+                (std::uint64_t*)::jm::detail::load_from_reg((std::uint64_t)arr);
+#else
+            alignas(alignment) std::uint64_t keys[]{ ::jm::detail::load_from_reg(Keys)... };
+#endif
+
+#if defined(_M_ARM64) || defined(__aarch64__) || defined(_M_ARM) || defined(__arm__)
+#if defined(__clang__)
+            ((Indices >= sizeof(_storage) / 16 ? static_cast<void>(0) : __builtin_neon_vst1q_v(
+                                    reinterpret_cast<uint64_t*>(_storage) + Indices * 2,
+                                    veorq_u64(__builtin_neon_vld1q_v(reinterpret_cast<const uint64_t*>(_storage) + Indices * 2, 51),
+                                              __builtin_neon_vld1q_v(reinterpret_cast<const uint64_t*>(keys) + Indices * 2, 51)),
+                                    51)), ...);
+#else // GCC, MSVC
+            ((Indices >= sizeof(_storage) / 16 ? static_cast<void>(0) : vst1q_u64(
+                        reinterpret_cast<uint64_t*>(_storage) + Indices * 2,
+                        veorq_u64(vld1q_u64(reinterpret_cast<const uint64_t*>(_storage) + Indices * 2),
+                                  vld1q_u64(reinterpret_cast<const uint64_t*>(keys) + Indices * 2)))), ...);
+#endif
+#elif !defined(JM_XORSTR_DISABLE_AVX_INTRINSICS)
+            ((Indices >= sizeof(_storage) / 32 ? static_cast<void>(0) : _mm256_store_si256(
+                reinterpret_cast<__m256i*>(_storage) + Indices,
+                _mm256_xor_si256(
+                    _mm256_load_si256(reinterpret_cast<const __m256i*>(_storage) + Indices),
+                    _mm256_load_si256(reinterpret_cast<const __m256i*>(keys) + Indices)))), ...);
+
+            if constexpr(sizeof(_storage) % 32 != 0)
+                _mm_store_si128(
+                    reinterpret_cast<__m128i*>(_storage + sizeof...(Keys) - 2),
+                    _mm_xor_si128(_mm_load_si128(reinterpret_cast<const __m128i*>(_storage + sizeof...(Keys) - 2)),
+                                  _mm_load_si128(reinterpret_cast<const __m128i*>(keys + sizeof...(Keys) - 2))));
+#else
+        ((Indices >= sizeof(_storage) / 16 ? static_cast<void>(0) : _mm_store_si128(
+            reinterpret_cast<__m128i*>(_storage) + Indices,
+            _mm_xor_si128(_mm_load_si128(reinterpret_cast<const __m128i*>(_storage) + Indices),
+                          _mm_load_si128(reinterpret_cast<const __m128i*>(keys) + Indices)))), ...);
+#endif
+            return (pointer)(_storage);
+        }
+    };
+
+    template<class L, std::size_t Size, std::size_t... Indices>
+    xor_string(L l, std::integral_constant<std::size_t, Size>, std::index_sequence<Indices...>) -> xor_string<std::remove_const_t<std::remove_reference_t<decltype(l()[0])>>, Size, std::integer_sequence<std::uint64_t, detail::key8<Indices>()...>, std::index_sequence<Indices...>>;
+}
+#endif
\ No newline at end of file