From 2eed4954c762bb8050e40798c3d9f1d3998324d1 Mon Sep 17 00:00:00 2001 From: Amlal El Mahrouss Date: Fri, 23 May 2025 03:48:06 +0200 Subject: feat!(LibCompiler): Codebase and diagram has been improved. Signed-off-by: Amlal El Mahrouss --- dev/LibCompiler/AssemblyInterface.h | 205 --- dev/LibCompiler/Backend/32x0.h | 24 +- dev/LibCompiler/Backend/64x0.h | 42 +- dev/LibCompiler/Backend/Aarch64.h | 2 +- dev/LibCompiler/Backend/Amd64.h | 49 - dev/LibCompiler/Backend/X64.h | 49 + dev/LibCompiler/CodeGen.h | 206 ++++ dev/LibCompiler/CompilerFrontend.h | 151 --- dev/LibCompiler/Defines.h | 15 +- dev/LibCompiler/Detail/AsmUtils.h | 93 -- dev/LibCompiler/Detail/ClUtils.h | 68 - dev/LibCompiler/ErrorID.h | 1 - dev/LibCompiler/ErrorOr.h | 5 + dev/LibCompiler/Frontend.h | 121 ++ dev/LibCompiler/Macros.h | 2 +- dev/LibCompiler/PEF.h | 2 +- dev/LibCompiler/PString.h | 82 ++ dev/LibCompiler/Ref.h | 5 +- dev/LibCompiler/StringView.h | 77 -- dev/LibCompiler/Util/LCAsmUtils.h | 93 ++ dev/LibCompiler/Util/LCClUtils.h | 68 + dev/LibCompiler/Version.h | 4 +- dev/LibCompiler/XCOFF.h | 2 +- dev/LibCompiler/src/Asm/Assembler32x0.cc | 49 + dev/LibCompiler/src/Asm/Assembler64x0.cc | 873 +++++++++++++ dev/LibCompiler/src/Asm/AssemblerAMD64.cc | 1203 ++++++++++++++++++ dev/LibCompiler/src/Asm/AssemblerARM64.cc | 585 +++++++++ dev/LibCompiler/src/Asm/AssemblerPowerPC.cc | 909 ++++++++++++++ dev/LibCompiler/src/Assembler32x0.cc | 49 - dev/LibCompiler/src/Assembler64x0.cc | 873 ------------- dev/LibCompiler/src/AssemblerAMD64.cc | 1203 ------------------ dev/LibCompiler/src/AssemblerARM64.cc | 585 --------- dev/LibCompiler/src/AssemblerPowerPC.cc | 909 -------------- .../src/AssemblyFactory+AssemblyInterface.cc | 55 - dev/LibCompiler/src/CCompiler64x0.cc | 1286 ------------------- dev/LibCompiler/src/CCompilerARM64.cc | 1284 ------------------- dev/LibCompiler/src/CCompilerPower64.cc | 1303 -------------------- dev/LibCompiler/src/CPlusPlusCompilerAMD64.cc | 921 -------------- .../src/CPlusPlusCompilerPreProcessor.cc | 900 -------------- dev/LibCompiler/src/Cl/CCompiler64x0.cc | 1286 +++++++++++++++++++ dev/LibCompiler/src/Cl/CCompilerARM64.cc | 1284 +++++++++++++++++++ dev/LibCompiler/src/Cl/CCompilerPower64.cc | 1303 ++++++++++++++++++++ dev/LibCompiler/src/Cl/CPlusPlusCompilerAMD64.cc | 921 ++++++++++++++ dev/LibCompiler/src/CodeGen.cc | 57 + dev/LibCompiler/src/DynamicLinkerPEF.cc | 692 ----------- dev/LibCompiler/src/Frontend.cc | 51 + dev/LibCompiler/src/Linker/DynamicLinkerPEF.cc | 695 +++++++++++ .../src/PP/CPlusPlusCompilerPreProcessor.cc | 903 ++++++++++++++ dev/LibCompiler/src/PString.cc | 207 ++++ dev/LibCompiler/src/StringView.cc | 207 ---- dev/LibDebugger/Version.h | 15 + docs/drawio/LIBCOMPILER_DESIGN.drawio | 38 +- lc-osx-san.json | 3 +- lc-osx.json | 3 +- lc-posix.json | 3 +- tools/asm.cc | 2 +- tools/cppdrv.cc | 2 +- tools/cxxdrv.cc | 2 +- tools/dbg.cc | 2 +- tools/kdbg.cc | 2 +- tools/ld64.cc | 2 +- 61 files changed, 11042 insertions(+), 10991 deletions(-) delete mode 100644 dev/LibCompiler/AssemblyInterface.h delete mode 100644 dev/LibCompiler/Backend/Amd64.h create mode 100644 dev/LibCompiler/Backend/X64.h create mode 100644 dev/LibCompiler/CodeGen.h delete mode 100644 dev/LibCompiler/CompilerFrontend.h delete mode 100644 dev/LibCompiler/Detail/AsmUtils.h delete mode 100644 dev/LibCompiler/Detail/ClUtils.h create mode 100644 dev/LibCompiler/Frontend.h create mode 100644 dev/LibCompiler/PString.h delete mode 100644 dev/LibCompiler/StringView.h create mode 100644 dev/LibCompiler/Util/LCAsmUtils.h create mode 100644 dev/LibCompiler/Util/LCClUtils.h create mode 100644 dev/LibCompiler/src/Asm/Assembler32x0.cc create mode 100644 dev/LibCompiler/src/Asm/Assembler64x0.cc create mode 100644 dev/LibCompiler/src/Asm/AssemblerAMD64.cc create mode 100644 dev/LibCompiler/src/Asm/AssemblerARM64.cc create mode 100644 dev/LibCompiler/src/Asm/AssemblerPowerPC.cc delete mode 100644 dev/LibCompiler/src/Assembler32x0.cc delete mode 100644 dev/LibCompiler/src/Assembler64x0.cc delete mode 100644 dev/LibCompiler/src/AssemblerAMD64.cc delete mode 100644 dev/LibCompiler/src/AssemblerARM64.cc delete mode 100644 dev/LibCompiler/src/AssemblerPowerPC.cc delete mode 100644 dev/LibCompiler/src/AssemblyFactory+AssemblyInterface.cc delete mode 100644 dev/LibCompiler/src/CCompiler64x0.cc delete mode 100644 dev/LibCompiler/src/CCompilerARM64.cc delete mode 100644 dev/LibCompiler/src/CCompilerPower64.cc delete mode 100644 dev/LibCompiler/src/CPlusPlusCompilerAMD64.cc delete mode 100644 dev/LibCompiler/src/CPlusPlusCompilerPreProcessor.cc create mode 100644 dev/LibCompiler/src/Cl/CCompiler64x0.cc create mode 100644 dev/LibCompiler/src/Cl/CCompilerARM64.cc create mode 100644 dev/LibCompiler/src/Cl/CCompilerPower64.cc create mode 100644 dev/LibCompiler/src/Cl/CPlusPlusCompilerAMD64.cc create mode 100644 dev/LibCompiler/src/CodeGen.cc delete mode 100644 dev/LibCompiler/src/DynamicLinkerPEF.cc create mode 100644 dev/LibCompiler/src/Frontend.cc create mode 100644 dev/LibCompiler/src/Linker/DynamicLinkerPEF.cc create mode 100644 dev/LibCompiler/src/PP/CPlusPlusCompilerPreProcessor.cc create mode 100644 dev/LibCompiler/src/PString.cc delete mode 100644 dev/LibCompiler/src/StringView.cc create mode 100644 dev/LibDebugger/Version.h diff --git a/dev/LibCompiler/AssemblyInterface.h b/dev/LibCompiler/AssemblyInterface.h deleted file mode 100644 index 5206005..0000000 --- a/dev/LibCompiler/AssemblyInterface.h +++ /dev/null @@ -1,205 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -#pragma once - -#include -#include -#include - -#define LC_ASSEMBLY_INTERFACE : public LibCompiler::AssemblyInterface -#define LC_ENCODER : public LibCompiler::EncoderInterface - -namespace LibCompiler { -class AssemblyFactory; -class AssemblyInterface; - -/// @brief Simple assembly factory -class AssemblyFactory final { - public: - explicit AssemblyFactory() = default; - ~AssemblyFactory() = default; - - LIBCOMPILER_COPY_DEFAULT(AssemblyFactory); - - public: - enum { - kArchInvalid = 0, - kArchAMD64 = 100, - kArch32x0, - kArch64x0, - kArchRISCV, - kArchPowerPC, - kArchAARCH64, - kArchUnknown, - }; - - Int32 Compile(std::string sourceFile, const Int32& arch) noexcept; - - void Mount(AssemblyInterface* mountPtr) noexcept; - AssemblyInterface* Unmount() noexcept; - - private: - AssemblyInterface* fMounted{nullptr}; -}; - -/// @brief Assembly to binary generator class. -/// @note This interface creates according to the CPU target of the child class. -class AssemblyInterface { - public: - explicit AssemblyInterface() = default; - virtual ~AssemblyInterface() = default; - - LIBCOMPILER_COPY_DEFAULT(AssemblyInterface); - - virtual UInt32 Arch() noexcept { return AssemblyFactory::kArchAMD64; } - - /// @brief compile to object file. - /// @note Example C++ -> MASM -> AE object. - virtual Int32 CompileToFormat(std::string src, Int32 arch) = 0; -}; - -union NumberCastBase { - NumberCastBase() = default; - ~NumberCastBase() = default; -}; - -union NumberCast64 final { - NumberCast64() = default; - explicit NumberCast64(UInt64 raw) : raw(raw) {} - - ~NumberCast64() { raw = 0; } - - CharType number[8]; - UInt64 raw; -}; - -union NumberCast32 final { - NumberCast32() = default; - explicit NumberCast32(UInt32 raw) : raw(raw) {} - - ~NumberCast32() { raw = 0; } - - CharType number[4]; - UInt32 raw; -}; - -union NumberCast16 final { - NumberCast16() = default; - explicit NumberCast16(UInt16 raw) : raw(raw) {} - - ~NumberCast16() { raw = 0; } - - CharType number[2]; - UInt16 raw; -}; - -union NumberCast8 final { - NumberCast8() = default; - explicit NumberCast8(UInt8 raw) : raw(raw) {} - - ~NumberCast8() { raw = 0; } - - CharType number; - UInt8 raw; -}; - -class EncoderInterface { - public: - explicit EncoderInterface() = default; - virtual ~EncoderInterface() = default; - - LIBCOMPILER_COPY_DEFAULT(EncoderInterface); - - virtual std::string CheckLine(std::string line, std::string file) = 0; - virtual bool WriteLine(std::string line, std::string file) = 0; - virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) = 0; -}; - -#ifdef __ASM_NEED_AMD64__ - -class EncoderAMD64 final : public EncoderInterface { - public: - explicit EncoderAMD64() = default; - ~EncoderAMD64() override = default; - - LIBCOMPILER_COPY_DEFAULT(EncoderAMD64); - - virtual std::string CheckLine(std::string line, std::string file) override; - virtual bool WriteLine(std::string line, std::string file) override; - virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; - - virtual bool WriteNumber16(const std::size_t& pos, std::string& from_what); - virtual bool WriteNumber32(const std::size_t& pos, std::string& from_what); - virtual bool WriteNumber8(const std::size_t& pos, std::string& from_what); -}; - -#endif // __ASM_NEED_AMD64__ - -#ifdef __ASM_NEED_ARM64__ - -class EncoderARM64 final : public EncoderInterface { - public: - explicit EncoderARM64() = default; - ~EncoderARM64() override = default; - - LIBCOMPILER_COPY_DEFAULT(EncoderARM64); - - virtual std::string CheckLine(std::string line, std::string file) override; - virtual bool WriteLine(std::string line, std::string file) override; - virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; -}; - -#endif // __ASM_NEED_ARM64__ - -#ifdef __ASM_NEED_64x0__ - -class Encoder64x0 final : public EncoderInterface { - public: - explicit Encoder64x0() = default; - ~Encoder64x0() override = default; - - LIBCOMPILER_COPY_DEFAULT(Encoder64x0); - - virtual std::string CheckLine(std::string line, std::string file) override; - virtual bool WriteLine(std::string line, std::string file) override; - virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; -}; - -#endif // __ASM_NEED_64x0__ - -#ifdef __ASM_NEED_32x0__ - -class Encoder32x0 final : public EncoderInterface { - public: - explicit Encoder32x0() = default; - ~Encoder32x0() override = default; - - LIBCOMPILER_COPY_DEFAULT(Encoder32x0); - - virtual std::string CheckLine(std::string line, std::string file) override; - virtual bool WriteLine(std::string line, std::string file) override; - virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; -}; - -#endif // __ASM_NEED_32x0__ - -#ifdef __ASM_NEED_PPC__ - -class EncoderPowerPC final : public EncoderInterface { - public: - explicit EncoderPowerPC() = default; - ~EncoderPowerPC() override = default; - - LIBCOMPILER_COPY_DEFAULT(EncoderPowerPC); - - virtual std::string CheckLine(std::string line, std::string file) override; - virtual bool WriteLine(std::string line, std::string file) override; - virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; -}; - -#endif // __ASM_NEED_32x0__ -} // namespace LibCompiler diff --git a/dev/LibCompiler/Backend/32x0.h b/dev/LibCompiler/Backend/32x0.h index 49bb978..240d885 100644 --- a/dev/LibCompiler/Backend/32x0.h +++ b/dev/LibCompiler/Backend/32x0.h @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ @@ -11,7 +11,7 @@ // @brief 32x0 support. // @file Backend/32x0.h -#define kAsmOpcodeDecl(__NAME, __OPCODE, __FUNCT3, __FUNCT7) \ +#define LC_ASM_OPCODE(__NAME, __OPCODE, __FUNCT3, __FUNCT7) \ {.fName = __NAME, .fOpcode = __OPCODE, .fFunct3 = __FUNCT3, .fFunct7 = __FUNCT7}, #define kAsmImmediate 0x01 @@ -37,18 +37,18 @@ struct CpuCode32x0 { #define kAsmByteStr ".byte" /* 8-bit */ inline std::vector kOpcodes32x0 = { - kAsmOpcodeDecl("nop", 0b0100011, 0b000, kAsmNoArgs) // nothing to do. (1C) - kAsmOpcodeDecl("jmp", 0b1110011, 0b001, kAsmJump) // jump to branch (2C) - kAsmOpcodeDecl("mov", 0b0100011, 0b101, kAsmImmediate) // move registers (3C) - kAsmOpcodeDecl("psh", 0b0111011, 0b000, kAsmImmediate) // push to sp (2C) - kAsmOpcodeDecl("pop", 0b0111011, 0b001, kAsmImmediate) // pop from sp. (1C) - kAsmOpcodeDecl("lea", 0b0111011, 0b010, + LC_ASM_OPCODE("nop", 0b0100011, 0b000, kAsmNoArgs) // nothing to do. (1C) + LC_ASM_OPCODE("jmp", 0b1110011, 0b001, kAsmJump) // jump to branch (2C) + LC_ASM_OPCODE("mov", 0b0100011, 0b101, kAsmImmediate) // move registers (3C) + LC_ASM_OPCODE("psh", 0b0111011, 0b000, kAsmImmediate) // push to sp (2C) + LC_ASM_OPCODE("pop", 0b0111011, 0b001, kAsmImmediate) // pop from sp. (1C) + LC_ASM_OPCODE("lea", 0b0111011, 0b010, kAsmImmediate) // setup stack and call, store address to CR (1C). - kAsmOpcodeDecl("ret", 0b0111011, 0b110, + LC_ASM_OPCODE("ret", 0b0111011, 0b110, kAsmImmediate) // return from procedure (2C). - kAsmOpcodeDecl("uc", 0b0111111, 0b000, kAsmSyscall) // user call (1C) - kAsmOpcodeDecl("kc", 0b0111111, 0b001, kAsmSyscall) // kernel call (1C) - kAsmOpcodeDecl("int", 0b0111111, 0b010, kAsmSyscall) // raise interrupt (1C) + LC_ASM_OPCODE("uc", 0b0111111, 0b000, kAsmSyscall) // user call (1C) + LC_ASM_OPCODE("kc", 0b0111111, 0b001, kAsmSyscall) // kernel call (1C) + LC_ASM_OPCODE("int", 0b0111111, 0b010, kAsmSyscall) // raise interrupt (1C) }; // \brief 64x0 register prefix diff --git a/dev/LibCompiler/Backend/64x0.h b/dev/LibCompiler/Backend/64x0.h index f929995..f7f0332 100644 --- a/dev/LibCompiler/Backend/64x0.h +++ b/dev/LibCompiler/Backend/64x0.h @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ @@ -12,7 +12,7 @@ // @brief 64x0 support. // @file Backend/64x0.h -#define kAsmOpcodeDecl(__NAME, __OPCODE, __FUNCT3, __FUNCT7) \ +#define LC_ASM_OPCODE(__NAME, __OPCODE, __FUNCT3, __FUNCT7) \ {.fName = __NAME, .fOpcode = __OPCODE, .fFunct3 = __FUNCT3, .fFunct7 = __FUNCT7}, #define kAsmImmediate 0x01 @@ -32,29 +32,29 @@ struct CpuOpcode64x0 { }; inline std::vector kOpcodes64x0 = { - kAsmOpcodeDecl("nop", 0b0000000, 0b0000000, kAsmNoArgs) // no-operation. - kAsmOpcodeDecl("np", 0b0000000, 0b0000000, kAsmNoArgs) // no-operation. - kAsmOpcodeDecl("jlr", 0b1110011, 0b0000111, + LC_ASM_OPCODE("nop", 0b0000000, 0b0000000, kAsmNoArgs) // no-operation. + LC_ASM_OPCODE("np", 0b0000000, 0b0000000, kAsmNoArgs) // no-operation. + LC_ASM_OPCODE("jlr", 0b1110011, 0b0000111, kAsmJump) // jump to linked return register - kAsmOpcodeDecl("jrl", 0b1110011, 0b0001111, + LC_ASM_OPCODE("jrl", 0b1110011, 0b0001111, kAsmJump) // jump from return register. - kAsmOpcodeDecl("mv", 0b0100011, 0b101, kAsmRegToReg) kAsmOpcodeDecl( - "bg", 0b1100111, 0b111, kAsmRegToReg) kAsmOpcodeDecl("bl", 0b1100111, 0b011, kAsmRegToReg) - kAsmOpcodeDecl("beq", 0b1100111, 0b000, kAsmRegToReg) - kAsmOpcodeDecl("bne", 0b1100111, 0b001, kAsmRegToReg) - kAsmOpcodeDecl("bge", 0b1100111, 0b101, kAsmRegToReg) - kAsmOpcodeDecl("ble", 0b1100111, 0b100, kAsmRegToReg) - kAsmOpcodeDecl("stw", 0b0001111, 0b100, kAsmImmediate) - kAsmOpcodeDecl("ldw", 0b0001111, 0b100, kAsmImmediate) - kAsmOpcodeDecl("lda", 0b0001111, 0b101, kAsmImmediate) - kAsmOpcodeDecl("sta", 0b0001111, 0b001, kAsmImmediate) + LC_ASM_OPCODE("mv", 0b0100011, 0b101, kAsmRegToReg) LC_ASM_OPCODE( + "bg", 0b1100111, 0b111, kAsmRegToReg) LC_ASM_OPCODE("bl", 0b1100111, 0b011, kAsmRegToReg) + LC_ASM_OPCODE("beq", 0b1100111, 0b000, kAsmRegToReg) + LC_ASM_OPCODE("bne", 0b1100111, 0b001, kAsmRegToReg) + LC_ASM_OPCODE("bge", 0b1100111, 0b101, kAsmRegToReg) + LC_ASM_OPCODE("ble", 0b1100111, 0b100, kAsmRegToReg) + LC_ASM_OPCODE("stw", 0b0001111, 0b100, kAsmImmediate) + LC_ASM_OPCODE("ldw", 0b0001111, 0b100, kAsmImmediate) + LC_ASM_OPCODE("lda", 0b0001111, 0b101, kAsmImmediate) + LC_ASM_OPCODE("sta", 0b0001111, 0b001, kAsmImmediate) // add/sub without carry flag - kAsmOpcodeDecl("add", 0b0101011, 0b100, kAsmImmediate) - kAsmOpcodeDecl("sub", 0b0101011, 0b101, kAsmImmediate) + LC_ASM_OPCODE("add", 0b0101011, 0b100, kAsmImmediate) + LC_ASM_OPCODE("sub", 0b0101011, 0b101, kAsmImmediate) // add/sub with carry flag - kAsmOpcodeDecl("addc", 0b0101011, 0b110, kAsmImmediate) - kAsmOpcodeDecl("subc", 0b0101011, 0b111, kAsmImmediate) - kAsmOpcodeDecl("sc", 0b1110011, 0b00, kAsmSyscall)}; + LC_ASM_OPCODE("addc", 0b0101011, 0b110, kAsmImmediate) + LC_ASM_OPCODE("subc", 0b0101011, 0b111, kAsmImmediate) + LC_ASM_OPCODE("sc", 0b1110011, 0b00, kAsmSyscall)}; // \brief 64x0 register prefix // example: r32, r0 diff --git a/dev/LibCompiler/Backend/Aarch64.h b/dev/LibCompiler/Backend/Aarch64.h index dcafa0a..2676306 100644 --- a/dev/LibCompiler/Backend/Aarch64.h +++ b/dev/LibCompiler/Backend/Aarch64.h @@ -1,6 +1,6 @@ /* ------------------------------------------- -Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved +Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/dev/LibCompiler/Backend/Amd64.h b/dev/LibCompiler/Backend/Amd64.h deleted file mode 100644 index ae8458f..0000000 --- a/dev/LibCompiler/Backend/Amd64.h +++ /dev/null @@ -1,49 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -#pragma once - -#include - -// @brief AMD64 support. -// @file Backend/Amd64.h - -#define kAsmOpcodeDecl(__NAME, __OPCODE) {.fName = __NAME, .fOpcode = __OPCODE}, - -typedef char i64_character_t; -typedef uint8_t i64_byte_t; -typedef uint16_t i64_hword_t; -typedef uint32_t i64_word_t; - -#define kAsmRegisterPrefix "r" - -struct CpuOpcodeAMD64 { - std::string fName; - i64_byte_t fPrefixBytes[4]; - i64_hword_t fOpcode; - i64_hword_t fModReg; - i64_word_t fDisplacment; - i64_word_t fImmediate; -}; - -/// these two are edge cases -#define kAsmIntOpcode 0xCC -#define kasmIntOpcodeAlt 0xCD - -#define kAsmJumpOpcode 0x0F80 -#define kJumpLimit 30 -#define kJumpLimitStandard 0xE3 -#define kJumpLimitStandardLimit 0xEB - -inline std::vector kOpcodesAMD64 = { - kAsmOpcodeDecl("int", 0xCD) kAsmOpcodeDecl("into", 0xCE) kAsmOpcodeDecl("intd", 0xF1) - kAsmOpcodeDecl("int3", 0xC3) kAsmOpcodeDecl("iret", 0xCF) kAsmOpcodeDecl("retf", 0xCB) - kAsmOpcodeDecl("retn", 0xC3) kAsmOpcodeDecl("ret", 0xC3) kAsmOpcodeDecl("sti", 0xfb) - kAsmOpcodeDecl("cli", 0xfa) kAsmOpcodeDecl("hlt", 0xf4) kAsmOpcodeDecl("nop", 0x90) - kAsmOpcodeDecl("mov", 0x48) kAsmOpcodeDecl("call", 0xFF) - kAsmOpcodeDecl("syscall", 0x0F) kAsmOpcodeDecl("xor", 0x48)}; - -#define kAsmRegisterLimit 16 diff --git a/dev/LibCompiler/Backend/X64.h b/dev/LibCompiler/Backend/X64.h new file mode 100644 index 0000000..f489515 --- /dev/null +++ b/dev/LibCompiler/Backend/X64.h @@ -0,0 +1,49 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#pragma once + +#include + +// @brief AMD64 support. +// @file Backend/X64.h + +#define LC_ASM_OPCODE(__NAME, __OPCODE) {.fName = __NAME, .fOpcode = __OPCODE}, + +typedef char i64_character_t; +typedef uint8_t i64_byte_t; +typedef uint16_t i64_hword_t; +typedef uint32_t i64_word_t; + +#define kAsmRegisterPrefix "r" + +struct CpuOpcodeAMD64 { + std::string fName; + i64_byte_t fPrefixBytes[4]; + i64_hword_t fOpcode; + i64_hword_t fModReg; + i64_word_t fDisplacment; + i64_word_t fImmediate; +}; + +/// these two are edge cases +#define kAsmIntOpcode 0xCC +#define kasmIntOpcodeAlt 0xCD + +#define kAsmJumpOpcode 0x0F80 +#define kJumpLimit 30 +#define kJumpLimitStandard 0xE3 +#define kJumpLimitStandardLimit 0xEB + +inline std::vector kOpcodesAMD64 = { + LC_ASM_OPCODE("int", 0xCD) LC_ASM_OPCODE("into", 0xCE) LC_ASM_OPCODE("intd", 0xF1) + LC_ASM_OPCODE("int3", 0xC3) LC_ASM_OPCODE("iret", 0xCF) LC_ASM_OPCODE("retf", 0xCB) + LC_ASM_OPCODE("retn", 0xC3) LC_ASM_OPCODE("ret", 0xC3) LC_ASM_OPCODE("sti", 0xfb) + LC_ASM_OPCODE("cli", 0xfa) LC_ASM_OPCODE("hlt", 0xf4) LC_ASM_OPCODE("nop", 0x90) + LC_ASM_OPCODE("mov", 0x48) LC_ASM_OPCODE("call", 0xFF) + LC_ASM_OPCODE("syscall", 0x0F) LC_ASM_OPCODE("xor", 0x48)}; + +#define kAsmRegisterLimit 16 diff --git a/dev/LibCompiler/CodeGen.h b/dev/LibCompiler/CodeGen.h new file mode 100644 index 0000000..1e955d8 --- /dev/null +++ b/dev/LibCompiler/CodeGen.h @@ -0,0 +1,206 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#pragma once + +#include +#include +#include + +#define LC_ASSEMBLY_INTERFACE : public LibCompiler::AssemblyInterface +#define LC_ENCODER : public LibCompiler::EncoderInterface + +namespace LibCompiler { +class AssemblyFactory; +class AssemblyInterface; + +/// @brief Simple assembly factory +class AssemblyFactory final { + public: + explicit AssemblyFactory() = default; + ~AssemblyFactory() = default; + + LIBCOMPILER_COPY_DEFAULT(AssemblyFactory); + + public: + enum { + kArchInvalid = 0, + kArchAMD64 = 100, + kArch32x0, + kArch64x0, + kArchRISCV, + kArchPowerPC, + kArchAARCH64, + kArchUnknown, + kArchCount = kArchUnknown - kArchAMD64, + }; + + Int32 Compile(std::string sourceFile, const Int32& arch) noexcept; + + void Mount(AssemblyInterface* mountPtr) noexcept; + AssemblyInterface* Unmount() noexcept; + + private: + AssemblyInterface* fMounted{nullptr}; +}; + +/// @brief Assembly to binary generator class. +/// @note This interface creates according to the CPU target of the child class. +class AssemblyInterface { + public: + explicit AssemblyInterface() = default; + virtual ~AssemblyInterface() = default; + + LIBCOMPILER_COPY_DEFAULT(AssemblyInterface); + + virtual UInt32 Arch() noexcept { return AssemblyFactory::kArchAMD64; } + + /// @brief compile to object file. + /// @note Example C++ -> MASM -> AE object. + virtual Int32 CompileToFormat(std::string src, Int32 arch) = 0; +}; + +union NumberCastBase { + NumberCastBase() = default; + ~NumberCastBase() = default; +}; + +union NumberCast64 final { + NumberCast64() = default; + explicit NumberCast64(UInt64 raw) : raw(raw) {} + + ~NumberCast64() { raw = 0; } + + CharType number[8]; + UInt64 raw; +}; + +union NumberCast32 final { + NumberCast32() = default; + explicit NumberCast32(UInt32 raw) : raw(raw) {} + + ~NumberCast32() { raw = 0; } + + CharType number[4]; + UInt32 raw; +}; + +union NumberCast16 final { + NumberCast16() = default; + explicit NumberCast16(UInt16 raw) : raw(raw) {} + + ~NumberCast16() { raw = 0; } + + CharType number[2]; + UInt16 raw; +}; + +union NumberCast8 final { + NumberCast8() = default; + explicit NumberCast8(UInt8 raw) : raw(raw) {} + + ~NumberCast8() { raw = 0; } + + CharType number; + UInt8 raw; +}; + +class EncoderInterface { + public: + explicit EncoderInterface() = default; + virtual ~EncoderInterface() = default; + + LIBCOMPILER_COPY_DEFAULT(EncoderInterface); + + virtual std::string CheckLine(std::string line, std::string file) = 0; + virtual bool WriteLine(std::string line, std::string file) = 0; + virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) = 0; +}; + +#ifdef __ASM_NEED_AMD64__ + +class EncoderAMD64 final : public EncoderInterface { + public: + explicit EncoderAMD64() = default; + ~EncoderAMD64() override = default; + + LIBCOMPILER_COPY_DEFAULT(EncoderAMD64); + + virtual std::string CheckLine(std::string line, std::string file) override; + virtual bool WriteLine(std::string line, std::string file) override; + virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; + + virtual bool WriteNumber16(const std::size_t& pos, std::string& from_what); + virtual bool WriteNumber32(const std::size_t& pos, std::string& from_what); + virtual bool WriteNumber8(const std::size_t& pos, std::string& from_what); +}; + +#endif // __ASM_NEED_AMD64__ + +#ifdef __ASM_NEED_ARM64__ + +class EncoderARM64 final : public EncoderInterface { + public: + explicit EncoderARM64() = default; + ~EncoderARM64() override = default; + + LIBCOMPILER_COPY_DEFAULT(EncoderARM64); + + virtual std::string CheckLine(std::string line, std::string file) override; + virtual bool WriteLine(std::string line, std::string file) override; + virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; +}; + +#endif // __ASM_NEED_ARM64__ + +#ifdef __ASM_NEED_64x0__ + +class Encoder64x0 final : public EncoderInterface { + public: + explicit Encoder64x0() = default; + ~Encoder64x0() override = default; + + LIBCOMPILER_COPY_DEFAULT(Encoder64x0); + + virtual std::string CheckLine(std::string line, std::string file) override; + virtual bool WriteLine(std::string line, std::string file) override; + virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; +}; + +#endif // __ASM_NEED_64x0__ + +#ifdef __ASM_NEED_32x0__ + +class Encoder32x0 final : public EncoderInterface { + public: + explicit Encoder32x0() = default; + ~Encoder32x0() override = default; + + LIBCOMPILER_COPY_DEFAULT(Encoder32x0); + + virtual std::string CheckLine(std::string line, std::string file) override; + virtual bool WriteLine(std::string line, std::string file) override; + virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; +}; + +#endif // __ASM_NEED_32x0__ + +#ifdef __ASM_NEED_PPC__ + +class EncoderPowerPC final : public EncoderInterface { + public: + explicit EncoderPowerPC() = default; + ~EncoderPowerPC() override = default; + + LIBCOMPILER_COPY_DEFAULT(EncoderPowerPC); + + virtual std::string CheckLine(std::string line, std::string file) override; + virtual bool WriteLine(std::string line, std::string file) override; + virtual bool WriteNumber(const std::size_t& pos, std::string& from_what) override; +}; + +#endif // __ASM_NEED_32x0__ +} // namespace LibCompiler diff --git a/dev/LibCompiler/CompilerFrontend.h b/dev/LibCompiler/CompilerFrontend.h deleted file mode 100644 index d9dd7a4..0000000 --- a/dev/LibCompiler/CompilerFrontend.h +++ /dev/null @@ -1,151 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -#pragma once - -#include - -#define LC_COMPILER_FRONTEND : public LibCompiler::CompilerFrontendInterface - -namespace LibCompiler { -inline static auto kInvalidFrontend = "?"; - -/// @brief Compiler backend, implements a frontend, such as C, C++... -/// See Toolchain, for some examples. -class CompilerFrontendInterface { - public: - explicit CompilerFrontendInterface() = default; - virtual ~CompilerFrontendInterface() = default; - - LIBCOMPILER_COPY_DEFAULT(CompilerFrontendInterface); - - // NOTE: cast this to your user defined ast. - typedef void* AstType; - - //! @brief Compile a syntax tree ouf of the text. - //! Also takes the source file name for metadata. - - virtual bool Compile(std::string text, std::string file) = 0; - - //! @brief What language are we dealing with? - virtual const char* Language() { return kInvalidFrontend; } - - virtual bool IsValid() { return strcmp(this->Language(), kInvalidFrontend) > 0; } -}; - -struct SyntaxLeafList; -struct SyntaxLeafList; -struct CompilerKeyword; - -/// we want to do that because to separate keywords. -enum KeywordKind { - kKeywordKindNamespace, - kKeywordKindFunctionStart, - kKeywordKindFunctionEnd, - kKeywordKindVariable, - kKeywordKindVariablePtr, - kKeywordKindType, - kKeywordKindTypePtr, - kKeywordKindExpressionBegin, - kKeywordKindExpressionEnd, - kKeywordKindArgSeparator, - kKeywordKindBodyStart, - kKeywordKindBodyEnd, - kKeywordKindClass, - kKeywordKindPtrAccess, - kKeywordKindAccess, - kKeywordKindIf, - kKeywordKindElse, - kKeywordKindElseIf, - kKeywordKindVariableAssign, - kKeywordKindVariableDec, - kKeywordKindVariableInc, - kKeywordKindConstant, - kKeywordKindTypedef, - kKeywordKindEndInstr, - kKeywordKindSpecifier, - kKeywordKindInvalid, - kKeywordKindReturn, - kKeywordKindCommentInline, - kKeywordKindCommentMultiLineStart, - kKeywordKindCommentMultiLineEnd, - kKeywordKindEq, - kKeywordKindNotEq, - kKeywordKindGreaterEq, - kKeywordKindLessEq, - kKeywordKindPtr, -}; - -/// \brief Compiler keyword information struct. -struct CompilerKeyword { - std::string keyword_name; - KeywordKind keyword_kind = kKeywordKindInvalid; -}; -struct SyntaxLeafList final { - struct SyntaxLeaf final { - Int32 fUserType; -#ifdef LC_USE_STRUCTS - CompilerKeyword fUserData; -#else - std::string fUserData; -#endif - - SyntaxLeaf() = default; - - std::string fUserValue; - struct SyntaxLeaf* fNext; - }; - - std::vector fLeafList; - SizeType fNumLeafs; - - size_t SizeOf() { return fNumLeafs; } - std::vector& Get() { return fLeafList; } - SyntaxLeaf& At(size_t index) { return fLeafList[index]; } -}; - -/// find the perfect matching word in a haystack. -/// \param haystack base string -/// \param needle the string we search for. -/// \return if we found it or not. -inline bool find_word(std::string haystack, std::string needle) noexcept { - auto index = haystack.find(needle); - - // check for needle validity. - if (index == std::string::npos) return false; - - // declare lambda - auto not_part_of_word = [&](int index) { - if (std::isspace(haystack[index]) || std::ispunct(haystack[index])) return true; - - if (index <= 0 || index >= haystack.size()) return true; - - return false; - }; - - return not_part_of_word(index - 1) && not_part_of_word(index + needle.size()); -} - -/// find a word within strict conditions and returns a range of it. -/// \param haystack -/// \param needle -/// \return position of needle. -inline std::size_t find_word_range(std::string haystack, std::string needle) noexcept { - auto index = haystack.find(needle); - - // check for needle validity. - if (index == std::string::npos) return false; - - if (!isalnum((haystack[index + needle.size() + 1])) && - !isdigit(haystack[index + needle.size() + 1]) && - !isalnum((haystack[index - needle.size() - 1])) && - !isdigit(haystack[index - needle.size() - 1])) { - return index; - } - - return std::string::npos; -} -} // namespace LibCompiler diff --git a/dev/LibCompiler/Defines.h b/dev/LibCompiler/Defines.h index eacf9c0..1735606 100644 --- a/dev/LibCompiler/Defines.h +++ b/dev/LibCompiler/Defines.h @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ @@ -99,22 +99,16 @@ #define LC_IMPORT_C extern "C" #define LC_IMPORT extern - -#include -#include -#include -#include - namespace LibCompiler { inline constexpr int kBaseYear = 1900; -typedef std::string String; +typedef std::string STLString; -inline String current_date() noexcept { +inline STLString current_date() noexcept { auto time_data = time(nullptr); auto time_struct = gmtime(&time_data); - String fmt = std::to_string(kBaseYear + time_struct->tm_year); + STLString fmt = std::to_string(kBaseYear + time_struct->tm_year); fmt += "-"; fmt += std::to_string(time_struct->tm_mon + 1); @@ -144,7 +138,6 @@ inline bool to_str(CharType* str, Int32 limit, Int32 base) noexcept { return true; } -using String = std::basic_string; } // namespace LibCompiler #define PACKED __attribute__((packed)) diff --git a/dev/LibCompiler/Detail/AsmUtils.h b/dev/LibCompiler/Detail/AsmUtils.h deleted file mode 100644 index 77805c1..0000000 --- a/dev/LibCompiler/Detail/AsmUtils.h +++ /dev/null @@ -1,93 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -#pragma once - -#include -#include - -#include - -using namespace LibCompiler; - -/// @brief Get Number from lineBuffer. -/// @param lineBuffer the lineBuffer to fetch from. -/// @param numberKey where to seek that number. -/// @return -static NumberCast32 GetNumber32(std::string lineBuffer, std::string numberKey) { - auto pos = lineBuffer.find(numberKey) + numberKey.size(); - - while (lineBuffer[pos] == ' ') { - ++pos; - } - - switch (lineBuffer[pos + 1]) { - case 'x': { - if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + lineBuffer, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 16)); - - if (kVerbose) { - kStdOut << "asm: found a base 16 number here: " << lineBuffer.substr(pos) << "\n"; - } - - return numOffset; - } - case 'b': { - if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number:" + lineBuffer, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "asm: found a base 2 number here:" << lineBuffer.substr(pos) << "\n"; - } - - return numOffset; - } - case 'o': { - if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + lineBuffer, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "asm: found a base 8 number here:" << lineBuffer.substr(pos) << "\n"; - } - - return numOffset; - } - default: { - if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + lineBuffer, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 10)); - - if (kVerbose) { - kStdOut << "asm: found a base 10 number here:" << lineBuffer.substr(pos) << "\n"; - } - - return numOffset; - } - } -} diff --git a/dev/LibCompiler/Detail/ClUtils.h b/dev/LibCompiler/Detail/ClUtils.h deleted file mode 100644 index b33f66b..0000000 --- a/dev/LibCompiler/Detail/ClUtils.h +++ /dev/null @@ -1,68 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -#pragma once - -#include -#include -#include -#include - -#define kZero64Section ".zero64" -#define kCode64Section ".code64" -#define kData64Section ".data64" - -#define kZero128Section ".zero128" -#define kCode128Section ".code128" -#define kData128Section ".data128" - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" -#define kYellow "\e[0;33m" - -#define kStdOut (std::cout << kRed << "drv: " << kWhite) -#define kStdErr (std::cout << kYellow << "drv: " << kWhite) - -inline static UInt32 kErrorLimit = 10; -inline static UInt32 kAcceptableErrors = 0; -inline static bool kVerbose = false; -inline static bool kOutputAsBinary = false; - -namespace Detail { -inline void print_error(std::string reason, std::string file) noexcept { - if (reason[0] == '\n') reason.erase(0, 1); - - kStdErr << reason << kBlank << std::endl; - - if (kAcceptableErrors > kErrorLimit) std::exit(LIBCOMPILER_EXEC_ERROR); - - ++kAcceptableErrors; -} - -inline void print_warning(std::string reason, std::string file) noexcept { - if (reason[0] == '\n') reason.erase(0, 1); - - kStdOut << kYellow << reason << kBlank << std::endl; -} - -/// @internal -/// @brief Handler for SIGSEGV signal. -inline void drv_segfault_handler(std::int32_t id) { - switch (id) { - case SIGSEGV: { - kStdErr << "SIGSEGV: Please report this on the GitHub issues page." << kBlank << std::endl; - break; - } - case SIGABRT: { - kStdErr << "SIGABRT: Please report this on the GitHub issues page." << kBlank << std::endl; - break; - } - } - - std::exit(LIBCOMPILER_EXEC_ERROR); -} -} // namespace Detail diff --git a/dev/LibCompiler/ErrorID.h b/dev/LibCompiler/ErrorID.h index 45bb393..c82fb16 100644 --- a/dev/LibCompiler/ErrorID.h +++ b/dev/LibCompiler/ErrorID.h @@ -10,7 +10,6 @@ #pragma once #include -#include #define LIBCOMPILER_SUCCESS 0 #define LIBCOMPILER_EXEC_ERROR -30 diff --git a/dev/LibCompiler/ErrorOr.h b/dev/LibCompiler/ErrorOr.h index ca93dd2..77015b1 100644 --- a/dev/LibCompiler/ErrorOr.h +++ b/dev/LibCompiler/ErrorOr.h @@ -10,6 +10,7 @@ #pragma once #include +#include #include namespace LibCompiler { @@ -33,6 +34,10 @@ class ErrorOr final { Ref Leak() { return mRef; } + Int32 Error() { return mId; } + + BOOL HasError() { return mId != LIBCOMPILER_SUCCESS; } + operator bool() { return mRef; } private: diff --git a/dev/LibCompiler/Frontend.h b/dev/LibCompiler/Frontend.h new file mode 100644 index 0000000..119b9fe --- /dev/null +++ b/dev/LibCompiler/Frontend.h @@ -0,0 +1,121 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#pragma once + +#include + +#define LC_COMPILER_FRONTEND : public LibCompiler::CompilerFrontendInterface + +namespace LibCompiler { +inline static auto kInvalidFrontend = "?"; + +/// @brief Compiler backend, implements a frontend, such as C, C++... +/// See Toolchain, for some examples. +class CompilerFrontendInterface { + public: + explicit CompilerFrontendInterface() = default; + virtual ~CompilerFrontendInterface() = default; + + LIBCOMPILER_COPY_DEFAULT(CompilerFrontendInterface); + + // NOTE: cast this to your user defined ast. + typedef void* AstType; + + //! @brief Compile a syntax tree ouf of the text. + //! Also takes the source file name for metadata. + + virtual bool Compile(std::string text, std::string file) = 0; + + //! @brief What language are we dealing with? + virtual const char* Language() { return kInvalidFrontend; } + + virtual bool IsValid() { return strcmp(this->Language(), kInvalidFrontend) > 0; } +}; + +struct SyntaxLeafList; +struct SyntaxLeafList; +struct CompilerKeyword; + +/// we want to do that because to separate keywords. +enum KeywordKind { + kKeywordKindNamespace, + kKeywordKindFunctionStart, + kKeywordKindFunctionEnd, + kKeywordKindVariable, + kKeywordKindVariablePtr, + kKeywordKindType, + kKeywordKindTypePtr, + kKeywordKindExpressionBegin, + kKeywordKindExpressionEnd, + kKeywordKindArgSeparator, + kKeywordKindBodyStart, + kKeywordKindBodyEnd, + kKeywordKindClass, + kKeywordKindPtrAccess, + kKeywordKindAccess, + kKeywordKindIf, + kKeywordKindElse, + kKeywordKindElseIf, + kKeywordKindVariableAssign, + kKeywordKindVariableDec, + kKeywordKindVariableInc, + kKeywordKindConstant, + kKeywordKindTypedef, + kKeywordKindEndInstr, + kKeywordKindSpecifier, + kKeywordKindInvalid, + kKeywordKindReturn, + kKeywordKindCommentInline, + kKeywordKindCommentMultiLineStart, + kKeywordKindCommentMultiLineEnd, + kKeywordKindEq, + kKeywordKindNotEq, + kKeywordKindGreaterEq, + kKeywordKindLessEq, + kKeywordKindPtr, +}; + +/// \brief Compiler keyword information struct. +struct CompilerKeyword { + std::string keyword_name; + KeywordKind keyword_kind = kKeywordKindInvalid; +}; +struct SyntaxLeafList final { + struct SyntaxLeaf final { + Int32 fUserType; +#ifdef LC_USE_STRUCTS + CompilerKeyword fUserData; +#else + std::string fUserData; +#endif + + SyntaxLeaf() = default; + + std::string fUserValue; + struct SyntaxLeaf* fNext; + }; + + std::vector fLeafList; + SizeType fNumLeafs; + + size_t SizeOf() { return fNumLeafs; } + std::vector& Get() { return fLeafList; } + SyntaxLeaf& At(size_t index) { return fLeafList[index]; } +}; + +/// find the perfect matching word in a haystack. +/// \param haystack base string +/// \param needle the string we search for. +/// \return if we found it or not. +bool find_word(std::string haystack, std::string needle) noexcept; + +/// find a word within strict conditions and returns a range of it. +/// \param haystack +/// \param needle +/// \return position of needle. +std::size_t find_word_range(std::string haystack, std::string needle) noexcept; +} // namespace LibCompiler diff --git a/dev/LibCompiler/Macros.h b/dev/LibCompiler/Macros.h index cce89d6..5027ffa 100644 --- a/dev/LibCompiler/Macros.h +++ b/dev/LibCompiler/Macros.h @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/dev/LibCompiler/PEF.h b/dev/LibCompiler/PEF.h index fe6cfe1..f2662c8 100644 --- a/dev/LibCompiler/PEF.h +++ b/dev/LibCompiler/PEF.h @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/dev/LibCompiler/PString.h b/dev/LibCompiler/PString.h new file mode 100644 index 0000000..a54be06 --- /dev/null +++ b/dev/LibCompiler/PString.h @@ -0,0 +1,82 @@ +/* + * ======================================================== + * + * LibCompiler + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +#pragma once + +#include +#include + +namespace LibCompiler { +class StringBuilder; +class PString; + +/** + * @brief PString class, contains a C string and manages it. + * @note No need to manage it it's getting deleted by default. + */ + +class PString final { + public: + explicit PString() = delete; + + explicit PString(SizeType Sz) noexcept : m_Sz(Sz) { + m_Data = new CharType[Sz]; + assert(m_Data); + } + + ~PString() noexcept { + if (m_Data) { + memset(m_Data, 0, m_Sz); + delete[] m_Data; + + m_Data = nullptr; + } + } + + LIBCOMPILER_COPY_DEFAULT(PString); + + CharType* Data(); + const CharType* CData() const; + SizeType Length() const; + + bool operator==(const CharType* rhs) const; + bool operator!=(const CharType* rhs) const; + + bool operator==(const PString& rhs) const; + bool operator!=(const PString& rhs) const; + + PString& operator+=(const CharType* rhs); + PString& operator+=(const PString& rhs); + + operator bool() { return m_Data && m_Data[0] != 0; } + + bool operator!() { return !m_Data || m_Data[0] == 0; } + + private: + CharType* m_Data{nullptr}; + SizeType m_Sz{0}; + SizeType m_Cur{0}; + + friend class StringBuilder; +}; + +/** + * @brief StringBuilder class + * @note These results shall call be delete[] after they're used. + */ +struct StringBuilder final { + static PString Construct(const CharType* data); + static const char* FromInt(const char* fmt, int n); + static const char* FromBool(const char* fmt, bool n); + static const char* Format(const char* fmt, const char* from); + static bool Equals(const char* lhs, const char* rhs); +}; + +using PStringOr = ErrorOr; +} // namespace LibCompiler diff --git a/dev/LibCompiler/Ref.h b/dev/LibCompiler/Ref.h index c5188b5..ffc11c0 100644 --- a/dev/LibCompiler/Ref.h +++ b/dev/LibCompiler/Ref.h @@ -22,14 +22,15 @@ class Ref final { ~Ref() { if (m_Strong) { - (*m_Class).~T(); + delete m_Class; + m_Class = nullptr; } } LIBCOMPILER_COPY_DEFAULT(Ref); public: - explicit Ref(T cls, const Bool& strong = false) : m_Class(&cls), m_Strong(strong) {} + explicit Ref(T* cls, const Bool& strong = false) : m_Class(cls), m_Strong(strong) {} Ref& operator=(T ref) { *m_Class = ref; diff --git a/dev/LibCompiler/StringView.h b/dev/LibCompiler/StringView.h deleted file mode 100644 index 18003a3..0000000 --- a/dev/LibCompiler/StringView.h +++ /dev/null @@ -1,77 +0,0 @@ -/* - * ======================================================== - * - * LibCompiler - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -#pragma once - -#include -#include - -namespace LibCompiler { -/** - * @brief StringView class, contains a C string and manages it. - * @note No need to manage it it's getting deleted by default. - */ - -class StringView final { - public: - explicit StringView() = delete; - - explicit StringView(SizeType Sz) noexcept : m_Sz(Sz) { - m_Data = new CharType[Sz]; - assert(m_Data); - } - - ~StringView() noexcept { - if (m_Data) { - memset(m_Data, 0, m_Sz); - delete[] m_Data; - - m_Data = nullptr; - } - } - - LIBCOMPILER_COPY_DEFAULT(StringView); - - CharType* Data(); - const CharType* CData() const; - SizeType Length() const; - - bool operator==(const CharType* rhs) const; - bool operator!=(const CharType* rhs) const; - - bool operator==(const StringView& rhs) const; - bool operator!=(const StringView& rhs) const; - - StringView& operator+=(const CharType* rhs); - StringView& operator+=(const StringView& rhs); - - operator bool() { return m_Data && m_Data[0] != 0; } - - bool operator!() { return !m_Data || m_Data[0] == 0; } - - private: - CharType* m_Data{nullptr}; - SizeType m_Sz{0}; - SizeType m_Cur{0}; - - friend class StringBuilder; -}; - -/** - * @brief StringBuilder class - * @note These results shall call delete[] after they're used. - */ -struct StringBuilder final { - static StringView Construct(const CharType* data); - static const char* FromInt(const char* fmt, int n); - static const char* FromBool(const char* fmt, bool n); - static const char* Format(const char* fmt, const char* from); - static bool Equals(const char* lhs, const char* rhs); -}; -} // namespace LibCompiler diff --git a/dev/LibCompiler/Util/LCAsmUtils.h b/dev/LibCompiler/Util/LCAsmUtils.h new file mode 100644 index 0000000..7728444 --- /dev/null +++ b/dev/LibCompiler/Util/LCAsmUtils.h @@ -0,0 +1,93 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#pragma once + +#include +#include + +#include + +using namespace LibCompiler; + +/// @brief Get Number from lineBuffer. +/// @param lineBuffer the lineBuffer to fetch from. +/// @param numberKey where to seek that number. +/// @return +static NumberCast32 GetNumber32(std::string lineBuffer, std::string numberKey) { + auto pos = lineBuffer.find(numberKey) + numberKey.size(); + + while (lineBuffer[pos] == ' ') { + ++pos; + } + + switch (lineBuffer[pos + 1]) { + case 'x': { + if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + lineBuffer, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 16)); + + if (kVerbose) { + kStdOut << "asm: found a base 16 number here: " << lineBuffer.substr(pos) << "\n"; + } + + return numOffset; + } + case 'b': { + if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number:" + lineBuffer, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "asm: found a base 2 number here:" << lineBuffer.substr(pos) << "\n"; + } + + return numOffset; + } + case 'o': { + if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + lineBuffer, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "asm: found a base 8 number here:" << lineBuffer.substr(pos) << "\n"; + } + + return numOffset; + } + default: { + if (auto res = strtol(lineBuffer.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + lineBuffer, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + NumberCast32 numOffset(strtol(lineBuffer.substr(pos).c_str(), nullptr, 10)); + + if (kVerbose) { + kStdOut << "asm: found a base 10 number here:" << lineBuffer.substr(pos) << "\n"; + } + + return numOffset; + } + } +} diff --git a/dev/LibCompiler/Util/LCClUtils.h b/dev/LibCompiler/Util/LCClUtils.h new file mode 100644 index 0000000..c88339d --- /dev/null +++ b/dev/LibCompiler/Util/LCClUtils.h @@ -0,0 +1,68 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#pragma once + +#include +#include +#include +#include + +#define kZero64Section ".zero64" +#define kCode64Section ".code64" +#define kData64Section ".data64" + +#define kZero128Section ".zero128" +#define kCode128Section ".code128" +#define kData128Section ".data128" + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" +#define kYellow "\e[0;33m" + +#define kStdOut (std::cout << kRed << "drv: " << kWhite) +#define kStdErr (std::cout << kYellow << "drv: " << kWhite) + +inline static UInt32 kErrorLimit = 10; +inline static UInt32 kAcceptableErrors = 0; +inline static bool kVerbose = false; +inline static bool kOutputAsBinary = false; + +namespace Detail { +inline void print_error(std::string reason, std::string file) noexcept { + if (reason[0] == '\n') reason.erase(0, 1); + + kStdErr << reason << kBlank << std::endl; + + if (kAcceptableErrors > kErrorLimit) std::exit(LIBCOMPILER_EXEC_ERROR); + + ++kAcceptableErrors; +} + +inline void print_warning(std::string reason, std::string file) noexcept { + if (reason[0] == '\n') reason.erase(0, 1); + + kStdOut << kYellow << reason << kBlank << std::endl; +} + +/// @internal +/// @brief Handler for SIGSEGV signal. +inline void drv_segfault_handler(std::int32_t id) { + switch (id) { + case SIGSEGV: { + kStdErr << "SIGSEGV: Please report this on the GitHub issues page." << kBlank << std::endl; + break; + } + case SIGABRT: { + kStdErr << "SIGABRT: Please report this on the GitHub issues page." << kBlank << std::endl; + break; + } + } + + std::exit(LIBCOMPILER_EXEC_ERROR); +} +} // namespace Detail diff --git a/dev/LibCompiler/Version.h b/dev/LibCompiler/Version.h index 601bb28..b94be42 100644 --- a/dev/LibCompiler/Version.h +++ b/dev/LibCompiler/Version.h @@ -1,12 +1,12 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ #pragma once -#define kDistVersion "v0.0.1" +#define kDistVersion "v0.0.1-libcompiler" #define kDistVersionBCD 0x0001 #define ToString(X) Stringify(X) diff --git a/dev/LibCompiler/XCOFF.h b/dev/LibCompiler/XCOFF.h index 047a20f..7aa13d6 100644 --- a/dev/LibCompiler/XCOFF.h +++ b/dev/LibCompiler/XCOFF.h @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved File: XCOFF.hpp Purpose: XCOFF for NeKernel. diff --git a/dev/LibCompiler/src/Asm/Assembler32x0.cc b/dev/LibCompiler/src/Asm/Assembler32x0.cc new file mode 100644 index 0000000..683b866 --- /dev/null +++ b/dev/LibCompiler/src/Asm/Assembler32x0.cc @@ -0,0 +1,49 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +/// bugs: 0 + +///////////////////////////////////////////////////////////////////////////////////////// + +// @file 32asm.cxx +// @author EL Mahrouss Amlal +// @brief 32x0 Assembler. + +// REMINDER: when dealing with an undefined symbol use (string +// size):LinkerFindSymbol:(string) so that ld will look for it. + +///////////////////////////////////////////////////////////////////////////////////////// + +#define __ASM_NEED_32x0__ 1 + +#include +#include +#include +#include + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" +#define kYellow "\e[0;33m" + +#define kStdOut (std::cout << kWhite) +#define kStdErr (std::cout << kRed) + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief 32x0 Assembler entrypoint, the program/module starts here. + +///////////////////////////////////////////////////////////////////////////////////////// + +LIBCOMPILER_MODULE(NEAssemblerMain32000) { + return 0; +} diff --git a/dev/LibCompiler/src/Asm/Assembler64x0.cc b/dev/LibCompiler/src/Asm/Assembler64x0.cc new file mode 100644 index 0000000..9d07d99 --- /dev/null +++ b/dev/LibCompiler/src/Asm/Assembler64x0.cc @@ -0,0 +1,873 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +/// bugs: 0 + +///////////////////////////////////////////////////////////////////////////////////////// + +// @file Assembler64x0.cxx +// @author EL Mahrouss Amlal +// @brief 64x000 Assembler. + +// REMINDER: when dealing with an undefined symbol use (string +// size):LinkerFindSymbol:(string) so that ld will look for it. + +///////////////////////////////////////////////////////////////////////////////////////// + +#define __ASM_NEED_64x0__ 1 + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +static char kOutputArch = LibCompiler::kPefArch64000; + +/// @note The 64x0 is VLSIW, so we need to jump to 4 bytes. +constexpr auto k64x0IPAlignment = 0x4U; + +static std::size_t kCounter = 1UL; + +static std::uintptr_t kOrigin = kPefBaseOrigin; +static std::vector> kOriginLabel; + +static std::vector kBytes; + +static LibCompiler::AERecordHeader kCurrentRecord{ + .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; + +static std::vector kRecords; +static std::vector kUndefinedSymbols; + +static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; +static const std::string kRelocSymbol = ":RuntimeSymbol:"; + +// \brief forward decl. +static bool asm_read_attributes(std::string line); + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief 64x0 assembler entrypoint, the program/module starts here. + +///////////////////////////////////////////////////////////////////////////////////////// + +LIBCOMPILER_MODULE(AssemblerMain64x0) { + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + for (size_t i = 1; i < argc; ++i) { + if (argv[i][0] == '-') { + if (strcmp(argv[i], "-64x0-ver") == 0 || strcmp(argv[i], "-64x0-v") == 0) { + kStdOut + << "Assembler64x0: 64x0 Assembler.\nAssembler64x0: v1.10\nAssembler64x0: Copyright (c) " + "Amlal El Mahrouss\n"; + return 0; + } else if (strcmp(argv[i], "-64x0-h") == 0) { + kStdOut << "Assembler64x0: 64x0 Assembler.\nAssembler64x0: Copyright (c) 2024 Mahrouss " + "Logic.\n"; + kStdOut << "--version: Print program version.\n"; + kStdOut << "--verbose: Print verbose output.\n"; + kStdOut << "--binary: Output as flat binary.\n"; + kStdOut << "--64xxx: Compile for a subset of the X64000.\n"; + + return 0; + } else if (strcmp(argv[i], "-64x0-binary") == 0) { + kOutputAsBinary = true; + continue; + } else if (strcmp(argv[i], "-64x0-verbose") == 0) { + kVerbose = true; + continue; + } + + kStdOut << "Assembler64x0: ignore " << argv[i] << "\n"; + continue; + } + + if (!std::filesystem::exists(argv[i])) { + kStdOut << "Assembler64x0: can't open: " << argv[i] << std::endl; + goto asm_fail_exit; + } + + std::string object_output(argv[i]); + + for (auto& ext : kAsmFileExts) { + if (object_output.find(ext) != std::string::npos) { + object_output.erase(object_output.find(ext), std::strlen(ext)); + } + } + + object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; + + std::ifstream file_ptr(argv[i]); + std::ofstream file_ptr_out(object_output, std::ofstream::binary); + + if (file_ptr_out.bad()) { + if (kVerbose) { + kStdOut << "Assembler64x0: error: " << strerror(errno) << "\n"; + } + } + + std::string line; + + LibCompiler::AEHeader hdr{0}; + + memset(hdr.fPad, kAENullType, kAEPad); + + hdr.fMagic[0] = kAEMag0; + hdr.fMagic[1] = kAEMag1; + hdr.fSize = sizeof(LibCompiler::AEHeader); + hdr.fArch = kOutputArch; + + ///////////////////////////////////////////////////////////////////////////////////////// + + // COMPILATION LOOP + + ///////////////////////////////////////////////////////////////////////////////////////// + + LibCompiler::Encoder64x0 asm64; + + while (std::getline(file_ptr, line)) { + if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { + Detail::print_error(ln, argv[i]); + continue; + } + + try { + asm_read_attributes(line); + asm64.WriteLine(line, argv[i]); + } catch (const std::exception& e) { + if (kVerbose) { + std::string what = e.what(); + Detail::print_warning("exit because of: " + what, "LibCompiler"); + } + + std::filesystem::remove(object_output); + goto asm_fail_exit; + } + } + + if (!kOutputAsBinary) { + if (kVerbose) { + kStdOut << "Assembler64x0: Writing object file...\n"; + } + + // this is the final step, write everything to the file. + + auto pos = file_ptr_out.tellp(); + + hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); + + file_ptr_out << hdr; + + if (kRecords.empty()) { + kStdErr << "Assembler64x0: At least one record is needed to write an object " + "file.\nAssembler64x0: Make one using `public_segment .code64 foo_bar`.\n"; + + std::filesystem::remove(object_output); + return 1; + } + + kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + std::size_t record_count = 0UL; + + for (auto& rec : kRecords) { + if (kVerbose) kStdOut << "Assembler64x0: Wrote record " << rec.fName << " to file...\n"; + + rec.fFlags |= LibCompiler::kKindRelocationAtRuntime; + rec.fOffset = record_count; + ++record_count; + + file_ptr_out << rec; + } + + // increment once again, so that we won't lie about the kUndefinedSymbols. + ++record_count; + + for (auto& sym : kUndefinedSymbols) { + LibCompiler::AERecordHeader _record_hdr{0}; + + if (kVerbose) kStdOut << "Assembler64x0: Wrote symbol " << sym << " to file...\n"; + + _record_hdr.fKind = kAENullType; + _record_hdr.fSize = sym.size(); + _record_hdr.fOffset = record_count; + + ++record_count; + + memset(_record_hdr.fPad, kAENullType, kAEPad); + memcpy(_record_hdr.fName, sym.c_str(), sym.size()); + + file_ptr_out << _record_hdr; + + ++kCounter; + } + + auto pos_end = file_ptr_out.tellp(); + + file_ptr_out.seekp(pos); + + hdr.fStartCode = pos_end; + hdr.fCodeSize = kBytes.size(); + + file_ptr_out << hdr; + + file_ptr_out.seekp(pos_end); + } else { + if (kVerbose) { + kStdOut << "Assembler64x0: Write raw binary...\n"; + } + } + + // byte from byte, we write this. + for (auto& byte : kBytes) { + file_ptr_out.write(reinterpret_cast(&byte), sizeof(byte)); + } + + if (kVerbose) kStdOut << "Assembler64x0: Wrote file with program in it.\n"; + + file_ptr_out.flush(); + file_ptr_out.close(); + + if (kVerbose) kStdOut << "Assembler64x0: Exit succeeded.\n"; + + return 0; + } + +asm_fail_exit: + + if (kVerbose) kStdOut << "Assembler64x0: Exit failed.\n"; + + return 1; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for attributes +// returns true if any was found. + +///////////////////////////////////////////////////////////////////////////////////////// + +static bool asm_read_attributes(std::string line) { + // extern_segment is the opposite of public_segment, it signals to the ld + // that we need this symbol. + if (LibCompiler::find_word(line, "extern_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid extern_segment directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_extern_segment_bin"); + } + + auto name = line.substr(line.find("extern_segment") + strlen("extern_segment")); + + /// sanity check to avoid stupid linker errors. + if (name.size() == 0) { + Detail::print_error("Invalid extern_segment", "power-as"); + throw std::runtime_error("invalid_extern_segment"); + } + + std::string result = std::to_string(name.size()); + result += kUndefinedSymbol; + + // mangle this + for (char& j : name) { + if (j == ' ' || j == ',') j = '$'; + } + + result += name; + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that ld can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, result.c_str(), result.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + // public_segment is a special keyword used by Assembler64x0 to tell the AE output stage to + // mark this section as a header. it currently supports .code64, .data64., + // .zero64 + else if (LibCompiler::find_word(line, "public_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid public_segment directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_public_segment_bin"); + } + + auto name = line.substr(line.find("public_segment") + strlen("public_segment")); + + std::string name_copy = name; + + for (char& j : name) { + if (j == ' ') j = '$'; + } + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + + name_copy.erase(name_copy.find(".code64"), strlen(".code64")); + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + + name_copy.erase(name_copy.find(".data64"), strlen(".data64")); + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + + name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that ld can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); + + kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); + ++kOrigin; + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, name.c_str(), name.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + + return false; +} + +// \brief algorithms and helpers. + +namespace Detail::algorithm { +// \brief authorize a brief set of characters. +static inline bool is_not_alnum_space(char c) { + return !(isalpha(c) || isdigit(c) || (c == ' ') || (c == '\t') || (c == ',') || (c == '(') || + (c == ')') || (c == '"') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || + (c == '_') || (c == ':') || (c == '@') || (c == '.')); +} + +bool is_valid_64x0(std::string str) { + return std::find_if(str.begin(), str.end(), is_not_alnum_space) == str.end(); +} +} // namespace Detail::algorithm + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for line (syntax check) + +///////////////////////////////////////////////////////////////////////////////////////// + +std::string LibCompiler::Encoder64x0::CheckLine(std::string line, std::string file) { + std::string err_str; + + if (line.empty() || LibCompiler::find_word(line, "extern_segment") || + LibCompiler::find_word(line, "public_segment") || line.find('#') != std::string::npos || + LibCompiler::find_word(line, ";")) { + if (line.find('#') != std::string::npos) { + line.erase(line.find('#')); + } else if (line.find(';') != std::string::npos) { + line.erase(line.find(';')); + } else { + // now check the line for validity + if (!Detail::algorithm::is_valid_64x0(line)) { + err_str = "Line contains non alphanumeric characters.\nhere -> "; + err_str += line; + } + } + + return err_str; + } + + if (!Detail::algorithm::is_valid_64x0(line)) { + err_str = "Line contains non alphanumeric characters.\nhere -> "; + err_str += line; + + return err_str; + } + + // check for a valid instruction format. + + if (line.find(',') != std::string::npos) { + if (line.find(',') + 1 == line.size()) { + err_str += "\nInstruction lacks right register, here -> "; + err_str += line.substr(line.find(',')); + + return err_str; + } else { + bool nothing_on_right = true; + + if (line.find(',') + 1 > line.size()) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + + auto substr = line.substr(line.find(',') + 1); + + for (auto& ch : substr) { + if (ch != ' ' && ch != '\t') { + nothing_on_right = false; + } + } + + // this means we found nothing after that ',' . + if (nothing_on_right) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + } + } + + // these do take an argument. + std::vector operands_inst = {"stw", "ldw", "lda", "sta"}; + + // these don't. + std::vector filter_inst = {"jlr", "jrl", "int"}; + + for (auto& opcode64x0 : kOpcodes64x0) { + if (line.find(opcode64x0.fName) != std::string::npos) { + if (opcode64x0.fFunct7 == kAsmNoArgs) return err_str; + + for (auto& op : operands_inst) { + // if only the instruction was found. + if (line == op) { + err_str += "\nMalformed "; + err_str += op; + err_str += " instruction, here -> "; + err_str += line; + } + } + + // if it is like that -> addr1, 0x0 + if (auto it = std::find(filter_inst.begin(), filter_inst.end(), opcode64x0.fName); + it == filter_inst.cend()) { + if (LibCompiler::find_word(line, opcode64x0.fName)) { + if (!isspace(line[line.find(opcode64x0.fName) + strlen(opcode64x0.fName)])) { + err_str += "\nMissing space between "; + err_str += opcode64x0.fName; + err_str += " and operands.\nhere -> "; + err_str += line; + } + } + } + + return err_str; + } + } + + err_str += "Unrecognized instruction: " + line; + + return err_str; +} + +bool LibCompiler::Encoder64x0::WriteNumber(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_hex_number"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); + + for (char& i : num.number) { + kBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "Assembler64x0: found a base 16 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; + } + case 'b': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "Assembler64x0: found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + kBytes.push_back(i); + } + + return true; + } + case 'o': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "Assembler64x0: found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + kBytes.push_back(i); + } + + return true; + } + default: { + break; + } + } + + /* check for errno and stuff like that */ + if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + return false; + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); + + for (char& i : num.number) { + kBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "Assembler64x0: found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Read and write an instruction to the output array. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool LibCompiler::Encoder64x0::WriteLine(std::string line, std::string file) { + if (LibCompiler::find_word(line, "public_segment ")) return true; + + for (auto& opcode64x0 : kOpcodes64x0) { + // strict check here + if (LibCompiler::find_word(line, opcode64x0.fName) && Detail::algorithm::is_valid_64x0(line)) { + std::string name(opcode64x0.fName); + std::string jump_label, cpy_jump_label; + + kBytes.emplace_back(opcode64x0.fOpcode); + kBytes.emplace_back(opcode64x0.fFunct3); + kBytes.emplace_back(opcode64x0.fFunct7); + + // check funct7 type. + switch (opcode64x0.fFunct7) { + // reg to reg means register to register transfer operation. + case kAsmRegToReg: + case kAsmImmediate: { + // \brief how many registers we found. + std::size_t found_some = 0UL; + + for (size_t line_index = 0UL; line_index < line.size(); line_index++) { + if (line[line_index] == kAsmRegisterPrefix[0] && isdigit(line[line_index + 1])) { + std::string register_syntax = kAsmRegisterPrefix; + register_syntax += line[line_index + 1]; + + if (isdigit(line[line_index + 2])) register_syntax += line[line_index + 2]; + + std::string reg_str; + reg_str += line[line_index + 1]; + + if (isdigit(line[line_index + 2])) reg_str += line[line_index + 2]; + + // it ranges from r0 to r19 + // something like r190 doesn't exist in the instruction set. + if (kOutputArch == LibCompiler::kPefArch64000) { + if (isdigit(line[line_index + 3]) && isdigit(line[line_index + 2])) { + reg_str += line[line_index + 3]; + Detail::print_error("invalid register index, r" + reg_str + + "\nnote: The 64x0 accepts registers from r0 to r20.", + file); + throw std::runtime_error("invalid_register_index"); + } + } + + // finally cast to a size_t + std::size_t reg_index = strtol(reg_str.c_str(), nullptr, 10); + + if (reg_index > kAsmRegisterLimit) { + Detail::print_error("invalid register index, r" + reg_str, file); + throw std::runtime_error("invalid_register_index"); + } + + kBytes.emplace_back(reg_index); + ++found_some; + + if (kVerbose) { + kStdOut << "Assembler64x0: Register found: " << register_syntax << "\n"; + kStdOut << "Assembler64x0: Register amount in instruction: " << found_some << "\n"; + } + } + } + + // we're not in immediate addressing, reg to reg. + if (opcode64x0.fFunct7 != kAsmImmediate) { + // remember! register to register! + if (found_some == 1) { + Detail::print_error( + "Too few registers.\ntip: each Assembler64x0 register " + "starts with 'r'.\nline: " + + line, + file); + throw std::runtime_error("not_a_register"); + } + } + + if (found_some < 1 && name != "ldw" && name != "lda" && name != "stw") { + Detail::print_error("invalid combination of opcode and registers.\nline: " + line, + file); + throw std::runtime_error("invalid_comb_op_reg"); + } else if (found_some == 1 && name == "add") { + Detail::print_error("invalid combination of opcode and registers.\nline: " + line, + file); + throw std::runtime_error("invalid_comb_op_reg"); + } else if (found_some == 1 && name == "sub") { + Detail::print_error("invalid combination of opcode and registers.\nline: " + line, + file); + throw std::runtime_error("invalid_comb_op_reg"); + } + + if (found_some > 0 && name == "pop") { + Detail::print_error( + "invalid combination for opcode 'pop'.\ntip: it expects " + "nothing.\nline: " + + line, + file); + throw std::runtime_error("invalid_comb_op_pop"); + } + } + default: + break; + } + + // try to fetch a number from the name + if (name == "stw" || name == "ldw" || name == "lda" || name == "sta") { + auto where_string = name; + + // if we load something, we'd need it's symbol/literal + if (name == "stw" || name == "sta" || name == "ldw" || name == "lda" || name == "sta") + where_string = ","; + + jump_label = line; + + auto found_sym = false; + + while (jump_label.find(where_string) != std::string::npos) { + jump_label = jump_label.substr(jump_label.find(where_string) + where_string.size()); + + while (jump_label.find(" ") != std::string::npos) { + jump_label.erase(jump_label.find(" "), 1); + } + + if (jump_label[0] != kAsmRegisterPrefix[0] && !isdigit(jump_label[1])) { + if (found_sym) { + Detail::print_error( + "invalid combination of opcode and operands.\nhere -> " + jump_label, file); + throw std::runtime_error("invalid_comb_op_ops"); + } else { + // death trap installed. + found_sym = true; + } + } + } + + cpy_jump_label = jump_label; + + // replace any spaces with $ + if (jump_label[0] == ' ') { + while (jump_label.find(' ') != std::string::npos) { + if (isalnum(jump_label[0]) || isdigit(jump_label[0])) break; + + jump_label.erase(jump_label.find(' '), 1); + } + } + + if (!this->WriteNumber(0, jump_label)) { + // sta expects this: sta 0x000000, r0 + if (name == "sta") { + Detail::print_error("invalid combination of opcode and operands.\nHere ->" + line, + file); + throw std::runtime_error("invalid_comb_op_ops"); + } + } else { + if (name == "sta" && cpy_jump_label.find("extern_segment ") != std::string::npos) { + Detail::print_error("invalid usage extern_segment on 'sta', here: " + line, file); + throw std::runtime_error("invalid_sta_usage"); + } + } + + goto asm_write_label; + } + + // This is the case where we jump to a label, it is also used as a goto. + if (name == "lda" || name == "sta") { + asm_write_label: + if (cpy_jump_label.find('\n') != std::string::npos) + cpy_jump_label.erase(cpy_jump_label.find('\n'), 1); + + if (cpy_jump_label.find("extern_segment") != std::string::npos) { + cpy_jump_label.erase(cpy_jump_label.find("extern_segment"), strlen("extern_segment")); + + if (name == "sta") { + Detail::print_error("extern_segment is not allowed on a sta operation.", file); + throw std::runtime_error("extern_segment_sta_op"); + } else { + goto asm_end_label_cpy; + } + } + + if (name == "lda" || name == "sta") { + for (auto& label : kOriginLabel) { + if (cpy_jump_label == label.first) { + if (kVerbose) { + kStdOut << "Assembler64x0: Replace label " << cpy_jump_label + << " to address: " << label.second << std::endl; + } + + LibCompiler::NumberCast64 num(label.second); + + for (auto& num : num.number) { + kBytes.push_back(num); + } + + goto asm_end_label_cpy; + } + } + + if (cpy_jump_label[0] == '0') { + switch (cpy_jump_label[1]) { + case 'x': + case 'o': + case 'b': + if (this->WriteNumber(0, cpy_jump_label)) goto asm_end_label_cpy; + + break; + default: + break; + } + + if (isdigit(cpy_jump_label[0])) { + if (this->WriteNumber(0, cpy_jump_label)) goto asm_end_label_cpy; + + break; + } + } + } + + if (cpy_jump_label.size() < 1) { + Detail::print_error("label is empty, can't jump on it.", file); + throw std::runtime_error("label_empty"); + } + + /// don't go any further if: + /// load word (ldw) or store word. (stw) + + if (name == "ldw" || name == "stw") break; + + auto mld_reloc_str = std::to_string(cpy_jump_label.size()); + mld_reloc_str += kUndefinedSymbol; + mld_reloc_str += cpy_jump_label; + + bool ignore_back_slash = false; + + for (auto& reloc_chr : mld_reloc_str) { + if (reloc_chr == '\\') { + ignore_back_slash = true; + continue; + } + + if (ignore_back_slash) { + ignore_back_slash = false; + continue; + } + + kBytes.push_back(reloc_chr); + } + + kBytes.push_back('\0'); + goto asm_end_label_cpy; + } + + asm_end_label_cpy: + kOrigin += k64x0IPAlignment; + + break; + } + } + + return true; +} + +// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/Asm/AssemblerAMD64.cc b/dev/LibCompiler/src/Asm/AssemblerAMD64.cc new file mode 100644 index 0000000..4a4ca41 --- /dev/null +++ b/dev/LibCompiler/src/Asm/AssemblerAMD64.cc @@ -0,0 +1,1203 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @file AssemblerAMD64.cc +/// @author EL Mahrouss Amlal +/// @brief AMD64 Assembler. +/// REMINDER: when dealing with an undefined symbol use (string +/// size):LinkerFindSymbol:(string) so that ld will look for it. + +///////////////////////////////////////////////////////////////////////////////////////// + +/// bugs: 0 + +/// feature request: 1 +/// Encode registers in mov, add, xor... + +///////////////////////////////////////////////////////////////////////////////////////// + +#define __ASM_NEED_AMD64__ 1 + +#define kAssemblerPragmaSymStr "#" +#define kAssemblerPragmaSym '#' + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" +#define kYellow "\e[0;33m" + +static char kOutputArch = LibCompiler::kPefArchAMD64; + +constexpr auto kIPAlignement = 0x1U; + +static std::size_t kCounter = 1UL; + +static std::uintptr_t kOrigin = kPefBaseOrigin; +static std::vector> kOriginLabel; + +/// @brief keep it simple by default. +static std::int32_t kRegisterBitWidth = 16U; + +static std::vector kAppBytes; + +static LibCompiler::AERecordHeader kCurrentRecord{ + .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; + +static std::vector kRecords; +static std::vector kDefinedSymbols; +static std::vector kUndefinedSymbols; + +static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; + +// \brief forward decl. +static bool asm_read_attributes(std::string line); + +#include + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief AMD64 assembler entrypoint, the program/module starts here. + +///////////////////////////////////////////////////////////////////////////////////////// + +LIBCOMPILER_MODULE(AssemblerMainAMD64) { + //////////////// CPU OPCODES BEGIN //////////////// + + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + std::string opcodes_jump[kJumpLimit] = {"ja", "jae", "jb", "jbe", "jc", "je", "jg", "jge", + "jl", "jle", "jna", "jnae", "jnb", "jnbe", "jnc", "jne", + "jng", "jnge", "jnl", "jnle", "jno", "jnp", "jns", "jnz", + "jo", "jp", "jpe", "jpo", "js", "jz"}; + + for (i64_hword_t i = 0; i < kJumpLimit; i++) { + CpuOpcodeAMD64 code{.fName = opcodes_jump[i], + .fOpcode = static_cast(kAsmJumpOpcode + i)}; + kOpcodesAMD64.push_back(code); + } + + CpuOpcodeAMD64 code{.fName = "jcxz", .fOpcode = 0xE3}; + kOpcodesAMD64.push_back(code); + + for (i64_hword_t i = kJumpLimitStandard; i < kJumpLimitStandardLimit; i++) { + CpuOpcodeAMD64 code{.fName = "jmp", .fOpcode = i}; + kOpcodesAMD64.push_back(code); + } + + CpuOpcodeAMD64 lahf{.fName = "lahf", .fOpcode = 0x9F}; + kOpcodesAMD64.push_back(lahf); + + CpuOpcodeAMD64 lds{.fName = "lds", .fOpcode = 0xC5}; + kOpcodesAMD64.push_back(lds); + + CpuOpcodeAMD64 lea{.fName = "lea", .fOpcode = 0x8D}; + kOpcodesAMD64.push_back(lea); + + CpuOpcodeAMD64 nop{.fName = "nop", .fOpcode = 0x90}; + kOpcodesAMD64.push_back(nop); + + //////////////// CPU OPCODES END //////////////// + + for (size_t i = 1; i < argc; ++i) { + if (argv[i][0] == '-') { + if (strcmp(argv[i], "--amd64:ver") == 0 || strcmp(argv[i], "--amd64:v") == 0) { + kStdOut << "AssemblerAMD64: AMD64 Assembler Driver.\nAssemblerAMD64: " + "v1.10\nAssemblerAMD64: Copyright " + "(c) Amlal El Mahrouss\n"; + return 0; + } else if (strcmp(argv[i], "--amd64:h") == 0) { + kStdOut << "AssemblerAMD64: AMD64 Assembler Driver.\nAssemblerAMD64: Copyright (c) 2024 " + "Amlal El Mahrouss\n"; + kStdOut << "--version: Print program version.\n"; + kStdOut << "--verbose: Print verbose output.\n"; + kStdOut << "--binary: Output as flat binary.\n"; + + return 0; + } else if (strcmp(argv[i], "--amd64:binary") == 0) { + kOutputAsBinary = true; + continue; + } else if (strcmp(argv[i], "--amd64:verbose") == 0) { + kVerbose = true; + continue; + } + + kStdOut << "AssemblerAMD64: ignore " << argv[i] << "\n"; + continue; + } + + if (!std::filesystem::exists(argv[i])) { + kStdOut << "AssemblerAMD64: can't open: " << argv[i] << std::endl; + goto asm_fail_exit; + } + + std::string object_output(argv[i]); + std::string asm_input(argv[i]); + + for (auto& ext : kAsmFileExts) { + if (object_output.ends_with(ext)) { + object_output.erase(object_output.find(ext), std::strlen(ext)); + break; + } + } + + object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; + + std::ifstream file_ptr(argv[i]); + std::ofstream file_ptr_out(object_output, std::ofstream::binary); + + kStdOut << "AssemblerAMD64: Assembling: " << argv[i] << "\n"; + + if (file_ptr_out.bad()) { + if (kVerbose) { + kStdOut << "AssemblerAMD64: error: " << strerror(errno) << "\n"; + } + + return 1; + } + + std::string line; + + LibCompiler::AEHeader hdr{0}; + + memset(hdr.fPad, kAENullType, kAEPad); + + hdr.fMagic[0] = kAEMag0; + hdr.fMagic[1] = kAEMag1; + hdr.fSize = sizeof(LibCompiler::AEHeader); + hdr.fArch = kOutputArch; + + ///////////////////////////////////////////////////////////////////////////////////////// + + // COMPILATION LOOP + + ///////////////////////////////////////////////////////////////////////////////////////// + + LibCompiler::EncoderAMD64 asm64; + + if (kVerbose) { + kStdOut << "Compiling: " + asm_input << "\n"; + kStdOut << "From: " + line << "\n"; + } + + while (std::getline(file_ptr, line)) { + if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { + Detail::print_error(ln, argv[i]); + continue; + } + + try { + asm_read_attributes(line); + asm64.WriteLine(line, argv[i]); + } catch (const std::exception& e) { + if (kVerbose) { + std::string what = e.what(); + Detail::print_warning("exit because of: " + what, "LibCompiler"); + } + + try { + std::filesystem::remove(object_output); + } catch (...) { + } + + goto asm_fail_exit; + } + } + + if (!kOutputAsBinary) { + if (kVerbose) { + kStdOut << "AssemblerAMD64: Writing object file...\n"; + } + + // this is the final step, write everything to the file. + + auto pos = file_ptr_out.tellp(); + + hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); + + file_ptr_out << hdr; + + if (kRecords.empty()) { + kStdErr << "AssemblerAMD64: At least one record is needed to write an object " + "file.\nAssemblerAMD64: Make one using `public_segment .code64 foo_bar`.\n"; + + std::filesystem::remove(object_output); + return 1; + } + + kRecords[kRecords.size() - 1].fSize = kAppBytes.size(); + + std::size_t record_count = 0UL; + + for (auto& rec : kRecords) { + if (kVerbose) kStdOut << "AssemblerAMD64: Wrote record " << rec.fName << " to file...\n"; + + rec.fFlags |= LibCompiler::kKindRelocationAtRuntime; + rec.fOffset = record_count; + ++record_count; + + file_ptr_out << rec; + } + + // increment once again, so that we won't lie about the kUndefinedSymbols. + ++record_count; + + for (auto& sym : kUndefinedSymbols) { + LibCompiler::AERecordHeader _record_hdr{0}; + + if (kVerbose) kStdOut << "AssemblerAMD64: Wrote symbol " << sym << " to file...\n"; + + _record_hdr.fKind = kAENullType; + _record_hdr.fSize = sym.size(); + _record_hdr.fOffset = record_count; + + ++record_count; + + memset(_record_hdr.fPad, kAENullType, kAEPad); + memcpy(_record_hdr.fName, sym.c_str(), sym.size()); + + file_ptr_out << _record_hdr; + + ++kCounter; + } + + auto pos_end = file_ptr_out.tellp(); + + file_ptr_out.seekp(pos); + + hdr.fStartCode = pos_end; + hdr.fCodeSize = kAppBytes.size(); + + file_ptr_out << hdr; + + file_ptr_out.seekp(pos_end); + } else { + if (kVerbose) { + kStdOut << "AssemblerAMD64: Write raw binary...\n"; + } + } + + // byte from byte, we write this. + for (auto& byte : kAppBytes) { + if (byte == 0) continue; + + if (byte == 0xFF) { + byte = 0; + } + + file_ptr_out << reinterpret_cast(&byte)[0]; + } + + if (kVerbose) kStdOut << "AssemblerAMD64: Wrote file with program in it.\n"; + + file_ptr_out.flush(); + file_ptr_out.close(); + + if (kVerbose) kStdOut << "AssemblerAMD64: Exit succeeded.\n"; + + return 0; + } + +asm_fail_exit: + + if (kVerbose) kStdOut << "AssemblerAMD64: Exit failed.\n"; + + return 1; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for attributes +// returns true if any was found. + +///////////////////////////////////////////////////////////////////////////////////////// + +static bool asm_read_attributes(std::string line) { + // extern_segment is the opposite of public_segment, it signals to the ld + // that we need this symbol. + if (LibCompiler::find_word(line, "extern_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_extern_segment_bin"); + } + + auto name = line.substr(line.find("extern_segment") + strlen("extern_segment") + 1); + + if (name.size() == 0) { + Detail::print_error("Invalid extern_segment", "power-as"); + throw std::runtime_error("invalid_extern_segment"); + } + + std::string result = std::to_string(name.size()); + result += kUndefinedSymbol; + + // mangle this + for (char& j : name) { + if (j == ' ' || j == ',') j = '$'; + } + + result += name; + + if (name.find(kPefCode64) != std::string::npos) { + // data is treated as code. + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(kPefData64) != std::string::npos) { + // no code will be executed from here. + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(kPefZero64) != std::string::npos) { + // this is a bss section. + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that ld can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kAppBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, result.c_str(), result.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + // public_segment is a special keyword used by AssemblerAMD64 to tell the AE output stage to + // mark this section as a header. it currently supports .code64, .data64 and + // .zero64. + else if (LibCompiler::find_word(line, "public_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_public_segment_bin"); + } + + auto name = line.substr(line.find("public_segment") + strlen("public_segment") + 1); + + std::string name_copy = name; + + for (char& j : name) { + if (j == ' ') j = '$'; + } + + if (std::find(kDefinedSymbols.begin(), kDefinedSymbols.end(), name) != kDefinedSymbols.end()) { + Detail::print_error("Symbol already defined.", "LibCompiler"); + throw std::runtime_error("invalid_public_segment_bin"); + } + + kDefinedSymbols.push_back(name); + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + + name_copy.erase(name_copy.find(".code64"), strlen(".code64")); + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + + name_copy.erase(name_copy.find(".data64"), strlen(".data64")); + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + + name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that ld can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); + + kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); + ++kOrigin; + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kAppBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, name.c_str(), name.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + + return false; +} + +// \brief algorithms and helpers. + +namespace Detail::algorithm { +// \brief authorize a brief set of characters. +static inline bool is_not_valid(char c) { + if ((isalpha(c) || isdigit(c)) || + ((c == ' ') || (c == '\t') || (c == ',') || (c == '(') || (c == ')') || (c == '"') || + (c == '*') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || (c == '_') || + (c == ':') || (c == '@') || (c == '.') || (c == '#') || (c == ';'))) + return false; + + return true; +} + +bool is_valid_amd64(std::string str) { + return std::find_if(str.begin(), str.end(), is_not_valid) == str.end(); +} +} // namespace Detail::algorithm + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for line (syntax check) + +///////////////////////////////////////////////////////////////////////////////////////// + +std::string LibCompiler::EncoderAMD64::CheckLine(std::string line, std::string file) { + std::string err_str; + + if (line.empty() || LibCompiler::find_word(line, "extern_segment") || + LibCompiler::find_word(line, "public_segment") || + LibCompiler::find_word(line, kAssemblerPragmaSymStr) || LibCompiler::find_word(line, ";") || + line[0] == kAssemblerPragmaSym) { + if (line.find(';') != std::string::npos) { + line.erase(line.find(';')); + } else { + // now check the line for validity + if (!Detail::algorithm::is_valid_amd64(line)) { + err_str = "Line contains non valid characters.\nhere -> "; + err_str += line; + } + } + + return err_str; + } + + // check for a valid instruction format. + + if (line.find(',') != std::string::npos) { + if (line.find(',') + 1 == line.size()) { + err_str += "\nInstruction lacks right register, here -> "; + err_str += line.substr(line.find(',')); + + return err_str; + } else { + bool nothing_on_right = true; + + if (line.find(',') + 1 > line.size()) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + + auto substr = line.substr(line.find(',') + 1); + + for (auto& ch : substr) { + if (ch != ' ' && ch != '\t') { + nothing_on_right = false; + } + } + + // this means we found nothing after that ',' . + if (nothing_on_right) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + } + } + for (auto& opcodeAMD64 : kOpcodesAMD64) { + if (LibCompiler::find_word(line, opcodeAMD64.fName)) { + return err_str; + } + } + + err_str += "\nUnrecognized instruction -> " + line; + + return err_str; +} + +bool LibCompiler::EncoderAMD64::WriteNumber(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + LibCompiler::NumberCast64 num = + LibCompiler::NumberCast64(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) + << "\n"; + } + + return true; + } + case 'b': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + LibCompiler::NumberCast64 num = + LibCompiler::NumberCast64(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + return true; + } + case 'o': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + LibCompiler::NumberCast64 num = + LibCompiler::NumberCast64(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + return true; + } + default: { + break; + } + } + + /* check for errno and stuff like that */ + if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + return false; + } + } + + LibCompiler::NumberCast64 num = + LibCompiler::NumberCast64(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +bool LibCompiler::EncoderAMD64::WriteNumber32(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); + res += kOrigin; + + if (errno != 0) { + return false; + } + + LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) + << "\n"; + } + + return true; + } + case 'b': { + auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); + res += kOrigin; + + if (errno != 0) { + return false; + } + + LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + return true; + } + case 'o': { + auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); + res += kOrigin; + + if (errno != 0) { + return false; + } + + LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + return true; + } + default: { + break; + } + } + + auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 10); + res += kOrigin; + + if (errno != 0) { + return false; + } + + LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +bool LibCompiler::EncoderAMD64::WriteNumber16(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + LibCompiler::NumberCast16 num = + LibCompiler::NumberCast16(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) + << "\n"; + } + + return true; + } + case 'b': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + LibCompiler::NumberCast16 num = + LibCompiler::NumberCast16(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + return true; + } + case 'o': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + LibCompiler::NumberCast16 num = + LibCompiler::NumberCast16(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + return true; + } + default: { + break; + } + } + + /* check for errno and stuff like that */ + if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + return false; + } + } + + LibCompiler::NumberCast16 num = + LibCompiler::NumberCast16(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); + + for (char& i : num.number) { + if (i == 0) i = 0xFF; + + kAppBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +bool LibCompiler::EncoderAMD64::WriteNumber8(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + LibCompiler::NumberCast8 num = + LibCompiler::NumberCast8(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); + + kAppBytes.push_back(num.number); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) + << "\n"; + } + + return true; + } + case 'b': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + LibCompiler::NumberCast8 num = + LibCompiler::NumberCast8(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + kAppBytes.push_back(num.number); + + return true; + } + case 'o': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + LibCompiler::NumberCast8 num = + LibCompiler::NumberCast8(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + kAppBytes.push_back(num.number); + + return true; + } + default: { + break; + } + } + + /* check for errno and stuff like that */ + if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + return false; + } + } + + LibCompiler::NumberCast8 num = + LibCompiler::NumberCast8(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); + + kAppBytes.push_back(num.number); + + if (kVerbose) { + kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Read and write an instruction to the output array. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool LibCompiler::EncoderAMD64::WriteLine(std::string line, std::string file) { + if (LibCompiler::find_word(line, "public_segment ")) return true; + + struct RegMapAMD64 { + std::string fName; + i64_byte_t fModRM; + }; + + std::vector kRegisterList{ + {.fName = "ax", .fModRM = 0x0}, {.fName = "cx", .fModRM = 1}, + {.fName = "dx", .fModRM = 0x2}, {.fName = "bx", .fModRM = 3}, + {.fName = "sp", .fModRM = 0x4}, {.fName = "bp", .fModRM = 5}, + {.fName = "si", .fModRM = 0x6}, {.fName = "di", .fModRM = 7}, + }; + + BOOL foundInstruction = false; + + for (auto& opcodeAMD64 : kOpcodesAMD64) { + // strict check here + if (LibCompiler::find_word(line, opcodeAMD64.fName) && + Detail::algorithm::is_valid_amd64(line)) { + foundInstruction = true; + std::string name(opcodeAMD64.fName); + + /// Move instruction handler. + if (line.find(name) != std::string::npos) { + if (name == "mov" || name == "xor") { + std::string substr = line.substr(line.find(name) + name.size()); + + uint64_t bits = kRegisterBitWidth; + + if (substr.find(",") == std::string::npos) { + Detail::print_error("Syntax error: missing right operand.", "LibCompiler"); + throw std::runtime_error("syntax_err"); + } + + bool onlyOneReg = true; + + std::vector currentRegList; + + for (auto& reg : kRegisterList) { + std::vector regExt = {'e', 'r'}; + + for (auto& ext : regExt) { + std::string registerName; + + if (bits > 16) registerName.push_back(ext); + + registerName += reg.fName; + + while (line.find(registerName) != std::string::npos) { + line.erase(line.find(registerName), registerName.size()); + + if (bits == 16) { + if (registerName[0] == 'r') { + Detail::print_error("invalid size for register, current bit width is: " + + std::to_string(kRegisterBitWidth), + file); + throw std::runtime_error("invalid_reg_size"); + } + } + + currentRegList.push_back({.fName = registerName, .fModRM = reg.fModRM}); + } + } + } + + if (currentRegList.size() > 1) onlyOneReg = false; + + bool hasRBasedRegs = false; + + if (!onlyOneReg) { + /// very tricky to understand. + /// but this checks for a r8 through r15 register. + if (currentRegList[0].fName[0] == 'r' || currentRegList[1].fName[0] == 'r') { + if (isdigit(currentRegList[0].fName[1]) && isdigit(currentRegList[1].fName[1])) { + kAppBytes.emplace_back(0x4d); + hasRBasedRegs = true; + } else if (isdigit(currentRegList[0].fName[1]) || + isdigit(currentRegList[1].fName[1])) { + kAppBytes.emplace_back(0x4c); + hasRBasedRegs = true; + } + } + } + + if (name == "mov") { + if (bits == 64 || bits == 32) { + if (!hasRBasedRegs && bits >= 32) { + kAppBytes.emplace_back(opcodeAMD64.fOpcode); + } + + if (!onlyOneReg) kAppBytes.emplace_back(0x89); + } else if (bits == 16) { + if (hasRBasedRegs) { + Detail::print_error("Invalid combination of operands and registers.", + "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } else { + kAppBytes.emplace_back(0x66); + kAppBytes.emplace_back(0x89); + } + } + } else { + if (!hasRBasedRegs && bits >= 32) { + kAppBytes.emplace_back(opcodeAMD64.fOpcode); + } + + kAppBytes.emplace_back(0x31); + } + + if (onlyOneReg) { + auto num = GetNumber32(line, ","); + + for (auto& num_idx : num.number) { + if (num_idx == 0) num_idx = 0xFF; + } + + auto modrm = (0x3 << 6 | currentRegList[0].fModRM); + + kAppBytes.emplace_back(0xC7); // prefixed before placing the modrm and then the number. + kAppBytes.emplace_back(modrm); + + if (name != "xor") { + kAppBytes.emplace_back(num.number[0]); + kAppBytes.emplace_back(num.number[1]); + kAppBytes.emplace_back(num.number[2]); + kAppBytes.emplace_back(num.number[3]); + } + + break; + } + + if (currentRegList[1].fName[0] == 'r' && currentRegList[0].fName[0] == 'e') { + Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } + + if (currentRegList[0].fName[0] == 'r' && currentRegList[1].fName[0] == 'e') { + Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } + + if (bits == 16) { + if (currentRegList[0].fName[0] == 'r' || currentRegList[0].fName[0] == 'e') { + Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } + + if (currentRegList[1].fName[0] == 'r' || currentRegList[1].fName[0] == 'e') { + Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } + } else { + if (currentRegList[0].fName[0] != 'r' || currentRegList[0].fName[0] == 'e') { + Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } + + if (currentRegList[1].fName[0] != 'r' || currentRegList[1].fName[0] == 'e') { + Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); + throw std::runtime_error("comb_op_reg"); + } + } + + /// encode register using the modrm encoding. + + auto modrm = (0x3 << 6 | currentRegList[1].fModRM << 3 | currentRegList[0].fModRM); + + kAppBytes.emplace_back(modrm); + + break; + } + } + + if (name == "int" || name == "into" || name == "intd") { + kAppBytes.emplace_back(opcodeAMD64.fOpcode); + this->WriteNumber8(line.find(name) + name.size() + 1, line); + + break; + } else if (name == "jmp" || name == "call") { + kAppBytes.emplace_back(opcodeAMD64.fOpcode); + + if (!this->WriteNumber32(line.find(name) + name.size() + 1, line)) { + throw std::runtime_error("BUG: WriteNumber32"); + } + + break; + } else if (name == "syscall") { + kAppBytes.emplace_back(opcodeAMD64.fOpcode); + kAppBytes.emplace_back(0x05); + + break; + } else { + kAppBytes.emplace_back(opcodeAMD64.fOpcode); + + break; + } + } + } + + if (line[0] == kAssemblerPragmaSym) { + if (foundInstruction) { + Detail::print_error("Syntax error: " + line, file); + throw std::runtime_error("syntax_err"); + } + + if (line.find("bits 64") != std::string::npos) { + kRegisterBitWidth = 64U; + } else if (line.find("bits 32") != std::string::npos) { + kRegisterBitWidth = 32U; + } else if (line.find("bits 16") != std::string::npos) { + kRegisterBitWidth = 16U; + } else if (line.find("org") != std::string::npos) { + size_t base[] = {10, 16, 2, 7}; + + for (size_t i = 0; i < 4; i++) { + if (kOrigin = strtol((line.substr(line.find("org") + strlen("org") + 1)).c_str(), nullptr, + base[i]); + kOrigin) { + if (errno != 0) { + continue; + } else { + if (kVerbose) { + kStdOut << "AssemblerAMD64: Origin Set: " << kOrigin << std::endl; + } + + break; + } + } + } + } + } + /// write a dword + else if (line.find(".dword") != std::string::npos) { + this->WriteNumber32(line.find(".dword") + strlen(".dword") + 1, line); + } + /// write a long + else if (line.find(".long") != std::string::npos) { + this->WriteNumber(line.find(".long") + strlen(".long") + 1, line); + } + /// write a 16-bit number + else if (line.find(".word") != std::string::npos) { + this->WriteNumber16(line.find(".word") + strlen(".word") + 1, line); + } + + kOrigin += kIPAlignement; + + return true; +} + +// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/Asm/AssemblerARM64.cc b/dev/LibCompiler/src/Asm/AssemblerARM64.cc new file mode 100644 index 0000000..5014a73 --- /dev/null +++ b/dev/LibCompiler/src/Asm/AssemblerARM64.cc @@ -0,0 +1,585 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @file AssemblerARM64.cxx +/// @author EL Mahrouss Amlal +/// @brief 'ACORN' Assembler. + +/// REMINDER: when dealing with an undefined symbol use (string +/// size):LinkerFindSymbol:(string) so that li will look for it. + +///////////////////////////////////////////////////////////////////////////////////////// + +#define __ASM_NEED_ARM64__ 1 + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" +#define kYellow "\e[0;33m" + +constexpr auto cPowerIPAlignment = 0x1U; + +static CharType kOutputArch = LibCompiler::kPefArchARM64; + +static std::size_t kCounter = 1UL; + +static std::uintptr_t kOrigin = kPefBaseOrigin; +static std::vector> kOriginLabel; + +static std::vector kBytes; + +static LibCompiler::AERecordHeader kCurrentRecord{ + .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; + +static std::vector kRecords; +static std::vector kUndefinedSymbols; + +static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; +static const std::string kRelocSymbol = ":RuntimeSymbol:"; + +// \brief forward decl. +static bool asm_read_attributes(std::string line); + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @brief POWER assembler entrypoint, the program/module starts here. + +///////////////////////////////////////////////////////////////////////////////////////// + +LIBCOMPILER_MODULE(AssemblerMainARM64) { + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + for (size_t i = 1; i < argc; ++i) { + if (argv[i][0] == '-') { + if (strcmp(argv[i], "--ver") == 0 || strcmp(argv[i], "--v") == 0) { + kStdOut << "AssemblerPower: AARCH64 Assembler Driver.\nAssemblerPower: " << kDistVersion + << "\nAssemblerPower: " + "Copyright (c) " + "Amlal El Mahrouss\n"; + return 0; + } else if (strcmp(argv[i], "--h") == 0) { + kStdOut << "AssemblerPower: AARCH64 Assembler Driver.\nAssemblerPower: Copyright (c) 2024 " + "Amlal El Mahrouss\n"; + kStdOut << "--version,/v: print program version.\n"; + kStdOut << "--verbose: print verbose output.\n"; + kStdOut << "--binary: output as flat binary.\n"; + + return 0; + } else if (strcmp(argv[i], "--binary") == 0) { + kOutputAsBinary = true; + continue; + } else if (strcmp(argv[i], "--verbose") == 0) { + kVerbose = true; + continue; + } + + kStdOut << "AssemblerPower: ignore " << argv[i] << "\n"; + continue; + } + + if (!std::filesystem::exists(argv[i])) { + kStdOut << "AssemblerPower: can't open: " << argv[i] << std::endl; + goto asm_fail_exit; + } + + std::string object_output(argv[i]); + + for (auto& ext : kAsmFileExts) { + if (object_output.find(ext) != std::string::npos) { + object_output.erase(object_output.find(ext), std::strlen(ext)); + } + } + + object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; + + std::ifstream file_ptr(argv[i]); + std::ofstream file_ptr_out(object_output, std::ofstream::binary); + + if (file_ptr_out.bad()) { + if (kVerbose) { + kStdOut << "AssemblerPower: error: " << strerror(errno) << "\n"; + } + } + + std::string line; + + LibCompiler::AEHeader hdr{0}; + + memset(hdr.fPad, kAENullType, kAEPad); + + hdr.fMagic[0] = kAEMag0; + hdr.fMagic[1] = kAEMag1; + hdr.fSize = sizeof(LibCompiler::AEHeader); + hdr.fArch = kOutputArch; + + ///////////////////////////////////////////////////////////////////////////////////////// + + // COMPILATION LOOP + + ///////////////////////////////////////////////////////////////////////////////////////// + + LibCompiler::EncoderARM64 asm64; + + while (std::getline(file_ptr, line)) { + if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { + Detail::print_error(ln, argv[i]); + continue; + } + + try { + asm_read_attributes(line); + asm64.WriteLine(line, argv[i]); + } catch (const std::exception& e) { + if (kVerbose) { + std::string what = e.what(); + Detail::print_warning("exit because of: " + what, "LibCompiler"); + } + + std::filesystem::remove(object_output); + goto asm_fail_exit; + } + } + + if (!kOutputAsBinary) { + if (kVerbose) { + kStdOut << "AssemblerARM64: Writing object file...\n"; + } + + // this is the final step, write everything to the file. + + auto pos = file_ptr_out.tellp(); + + hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); + + file_ptr_out << hdr; + + if (kRecords.empty()) { + kStdErr << "AssemblerARM64: At least one record is needed to write an object " + "file.\nAssemblerARM64: Make one using `public_segment .code64 foo_bar`.\n"; + + std::filesystem::remove(object_output); + return 1; + } + + kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + std::size_t record_count = 0UL; + + for (auto& record_hdr : kRecords) { + record_hdr.fFlags |= LibCompiler::kKindRelocationAtRuntime; + record_hdr.fOffset = record_count; + ++record_count; + + file_ptr_out << record_hdr; + + if (kVerbose) kStdOut << "AssemblerARM64: Wrote record " << record_hdr.fName << "...\n"; + } + + // increment once again, so that we won't lie about the kUndefinedSymbols. + ++record_count; + + for (auto& sym : kUndefinedSymbols) { + LibCompiler::AERecordHeader undefined_sym{0}; + + if (kVerbose) kStdOut << "AssemblerARM64: Wrote symbol " << sym << " to file...\n"; + + undefined_sym.fKind = kAENullType; + undefined_sym.fSize = sym.size(); + undefined_sym.fOffset = record_count; + + ++record_count; + + memset(undefined_sym.fPad, kAENullType, kAEPad); + memcpy(undefined_sym.fName, sym.c_str(), sym.size()); + + file_ptr_out << undefined_sym; + + ++kCounter; + } + + auto pos_end = file_ptr_out.tellp(); + + file_ptr_out.seekp(pos); + + hdr.fStartCode = pos_end; + hdr.fCodeSize = kBytes.size(); + + file_ptr_out << hdr; + + file_ptr_out.seekp(pos_end); + } else { + if (kVerbose) { + kStdOut << "AssemblerARM64: Write raw binary...\n"; + } + } + + // byte from byte, we write this. + for (auto& byte : kBytes) { + file_ptr_out.write(reinterpret_cast(&byte), sizeof(byte)); + } + + if (kVerbose) kStdOut << "AssemblerARM64: Wrote file with program in it.\n"; + + file_ptr_out.flush(); + file_ptr_out.close(); + + if (kVerbose) kStdOut << "AssemblerARM64: Exit succeeded.\n"; + + return 0; + } + +asm_fail_exit: + + if (kVerbose) kStdOut << "AssemblerARM64: Exit failed.\n"; + + return LIBCOMPILER_EXEC_ERROR; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for attributes +// returns true if any was found. + +///////////////////////////////////////////////////////////////////////////////////////// + +static bool asm_read_attributes(std::string line) { + // extern_segment is the opposite of public_segment, it signals to the li + // that we need this symbol. + if (LibCompiler::find_word(line, "extern_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid extern_segment directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_extern_segment_bin"); + } + + auto name = line.substr(line.find("extern_segment") + strlen("extern_segment") + 1); + + if (name.size() == 0) { + Detail::print_error("Invalid extern_segment", "LibCompiler"); + throw std::runtime_error("invalid_extern_segment"); + } + + std::string result = std::to_string(name.size()); + result += kUndefinedSymbol; + + // mangle this + for (char& j : name) { + if (j == ' ' || j == ',') j = '$'; + } + + result += name; + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that li can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, result.c_str(), result.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + // public_segment is a special keyword used by Assembler to tell the AE output stage to + // mark this section as a header. it currently supports .code64, .data64., + // .zero64 + else if (LibCompiler::find_word(line, "public_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid public_segment directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_public_segment_bin"); + } + + auto name = line.substr(line.find("public_segment") + strlen("public_segment")); + + std::string name_copy = name; + + for (char& j : name) { + if (j == ' ') j = '$'; + } + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + + name_copy.erase(name_copy.find(".code64"), strlen(".code64")); + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + + name_copy.erase(name_copy.find(".data64"), strlen(".data64")); + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + + name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that li can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); + + kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); + ++kOrigin; + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, name.c_str(), name.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + + return false; +} + +// \brief algorithms and helpers. + +namespace Detail::algorithm { +// \brief authorize a brief set of characters. +static inline bool is_not_alnum_space(char c) { + return !(isalpha(c) || isdigit(c) || (c == ' ') || (c == '\t') || (c == ',') || (c == '(') || + (c == ')') || (c == '"') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || + (c == '_') || (c == ':') || (c == '@') || (c == '.')); +} + +bool is_valid_arm64(std::string str) { + return std::find_if(str.begin(), str.end(), is_not_alnum_space) == str.end(); +} +} // namespace Detail::algorithm + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for line (syntax check) + +///////////////////////////////////////////////////////////////////////////////////////// + +std::string LibCompiler::EncoderARM64::CheckLine(std::string line, std::string file) { + std::string err_str; + + if (line.empty() || LibCompiler::find_word(line, "extern_segment") || + LibCompiler::find_word(line, "public_segment") || line.find('#') != std::string::npos || + LibCompiler::find_word(line, ";")) { + if (line.find('#') != std::string::npos) { + line.erase(line.find('#')); + } else if (line.find(';') != std::string::npos) { + line.erase(line.find(';')); + } else { + /// does the line contains valid input? + if (!Detail::algorithm::is_valid_arm64(line)) { + err_str = "Line contains non alphanumeric characters.\nhere -> "; + err_str += line; + } + } + + return err_str; + } + + if (!Detail::algorithm::is_valid_arm64(line)) { + err_str = "Line contains non alphanumeric characters.\nhere -> "; + err_str += line; + + return err_str; + } + + // check for a valid instruction format. + + if (line.find(',') != std::string::npos) { + if (line.find(',') + 1 == line.size()) { + err_str += "\nInstruction lacks right register, here -> "; + err_str += line.substr(line.find(',')); + + return err_str; + } else { + bool nothing_on_right = true; + + if (line.find(',') + 1 > line.size()) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + + auto substr = line.substr(line.find(',') + 1); + + for (auto& ch : substr) { + if (ch != ' ' && ch != '\t') { + nothing_on_right = false; + } + } + + // this means we found nothing after that ',' . + if (nothing_on_right) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + } + } + + return err_str; +} + +bool LibCompiler::EncoderARM64::WriteNumber(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); + + for (char& i : num.number) { + kBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerARM64: found a base 16 number here: " << jump_label.substr(pos) + << "\n"; + } + + return true; + } + case 'b': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "AssemblerARM64: found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + kBytes.push_back(i); + } + + return true; + } + case 'o': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "AssemblerARM64: found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + kBytes.push_back(i); + } + + return true; + } + default: { + break; + } + } + + /* check for errno and stuff like that */ + if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + return false; + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); + + for (char& i : num.number) { + kBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerARM64: found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @brief Read and write an instruction to the output array. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool LibCompiler::EncoderARM64::WriteLine(std::string line, std::string file) { + if (LibCompiler::find_word(line, "public_segment")) return false; + + if (!Detail::algorithm::is_valid_arm64(line)) return false; + + return true; +} + +// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/Asm/AssemblerPowerPC.cc b/dev/LibCompiler/src/Asm/AssemblerPowerPC.cc new file mode 100644 index 0000000..1502205 --- /dev/null +++ b/dev/LibCompiler/src/Asm/AssemblerPowerPC.cc @@ -0,0 +1,909 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @file AssemblerPower.cxx +/// @author EL Mahrouss Amlal +/// @brief POWER Assembler. + +/// REMINDER: when dealing with an undefined symbol use (string +/// size):LinkerFindSymbol:(string) so that li will look for it. + +///////////////////////////////////////////////////////////////////////////////////////// + +#define __ASM_NEED_PPC__ 1 + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" +#define kYellow "\e[0;33m" + +constexpr auto cPowerIPAlignment = 0x4U; + +static CharType kOutputArch = LibCompiler::kPefArchPowerPC; + +static std::size_t kCounter = 1UL; + +static std::uintptr_t kOrigin = kPefBaseOrigin; +static std::vector> kOriginLabel; + +static std::vector kBytes; + +static LibCompiler::AERecordHeader kCurrentRecord{ + .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; + +static std::vector kRecords; +static std::vector kUndefinedSymbols; + +static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; +static const std::string kRelocSymbol = ":RuntimeSymbol:"; + +// \brief forward decl. +static bool asm_read_attributes(std::string line); + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @brief POWER assembler entrypoint, the program/module starts here. + +///////////////////////////////////////////////////////////////////////////////////////// + +LIBCOMPILER_MODULE(AssemblerMainPower64) { + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + for (size_t i = 1; i < argc; ++i) { + if (argv[i][0] == '-') { + if (strcmp(argv[i], "--ver") == 0 || strcmp(argv[i], "--v") == 0) { + kStdOut << "AssemblerPower: POWER64 Assembler Driver.\nAssemblerPower: " << kDistVersion + << "\nAssemblerPower: " + "Copyright (c) " + "Amlal El Mahrouss\n"; + return 0; + } else if (strcmp(argv[i], "--h") == 0) { + kStdOut << "AssemblerPower: POWER64 Assembler Driver.\nAssemblerPower: Copyright (c) 2024 " + "Amlal El Mahrouss\n"; + kStdOut << "--version,/v: print program version.\n"; + kStdOut << "--verbose: print verbose output.\n"; + kStdOut << "--binary: output as flat binary.\n"; + + return 0; + } else if (strcmp(argv[i], "--binary") == 0) { + kOutputAsBinary = true; + continue; + } else if (strcmp(argv[i], "--verbose") == 0) { + kVerbose = true; + continue; + } + + kStdOut << "AssemblerPower: ignore " << argv[i] << "\n"; + continue; + } + + if (!std::filesystem::exists(argv[i])) { + kStdOut << "AssemblerPower: can't open: " << argv[i] << std::endl; + goto asm_fail_exit; + } + + std::string object_output(argv[i]); + + for (auto& ext : kAsmFileExts) { + if (object_output.find(ext) != std::string::npos) { + object_output.erase(object_output.find(ext), std::strlen(ext)); + } + } + + object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; + + std::ifstream file_ptr(argv[i]); + std::ofstream file_ptr_out(object_output, std::ofstream::binary); + + if (file_ptr_out.bad()) { + if (kVerbose) { + kStdOut << "AssemblerPower: error: " << strerror(errno) << "\n"; + } + } + + std::string line; + + LibCompiler::AEHeader hdr{0}; + + memset(hdr.fPad, kAENullType, kAEPad); + + hdr.fMagic[0] = kAEMag0; + hdr.fMagic[1] = kAEMag1; + hdr.fSize = sizeof(LibCompiler::AEHeader); + hdr.fArch = kOutputArch; + + ///////////////////////////////////////////////////////////////////////////////////////// + + // COMPILATION LOOP + + ///////////////////////////////////////////////////////////////////////////////////////// + + LibCompiler::EncoderPowerPC asm64; + + while (std::getline(file_ptr, line)) { + if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { + Detail::print_error(ln, argv[i]); + continue; + } + + try { + asm_read_attributes(line); + asm64.WriteLine(line, argv[i]); + } catch (const std::exception& e) { + if (kVerbose) { + std::string what = e.what(); + Detail::print_warning("exit because of: " + what, "LibCompiler"); + } + + std::filesystem::remove(object_output); + goto asm_fail_exit; + } + } + + if (!kOutputAsBinary) { + if (kVerbose) { + kStdOut << "AssemblerPower: Writing object file...\n"; + } + + // this is the final step, write everything to the file. + + auto pos = file_ptr_out.tellp(); + + hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); + + file_ptr_out << hdr; + + if (kRecords.empty()) { + kStdErr << "AssemblerPower: At least one record is needed to write an object " + "file.\nAssemblerPower: Make one using `public_segment .code64 foo_bar`.\n"; + + std::filesystem::remove(object_output); + return 1; + } + + kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + std::size_t record_count = 0UL; + + for (auto& record_hdr : kRecords) { + record_hdr.fFlags |= LibCompiler::kKindRelocationAtRuntime; + record_hdr.fOffset = record_count; + ++record_count; + + file_ptr_out << record_hdr; + + if (kVerbose) kStdOut << "AssemblerPower: Wrote record " << record_hdr.fName << "...\n"; + } + + // increment once again, so that we won't lie about the kUndefinedSymbols. + ++record_count; + + for (auto& sym : kUndefinedSymbols) { + LibCompiler::AERecordHeader undefined_sym{0}; + + if (kVerbose) kStdOut << "AssemblerPower: Wrote symbol " << sym << " to file...\n"; + + undefined_sym.fKind = kAENullType; + undefined_sym.fSize = sym.size(); + undefined_sym.fOffset = record_count; + + ++record_count; + + memset(undefined_sym.fPad, kAENullType, kAEPad); + memcpy(undefined_sym.fName, sym.c_str(), sym.size()); + + file_ptr_out << undefined_sym; + + ++kCounter; + } + + auto pos_end = file_ptr_out.tellp(); + + file_ptr_out.seekp(pos); + + hdr.fStartCode = pos_end; + hdr.fCodeSize = kBytes.size(); + + file_ptr_out << hdr; + + file_ptr_out.seekp(pos_end); + } else { + if (kVerbose) { + kStdOut << "AssemblerPower: Write raw binary...\n"; + } + } + + // byte from byte, we write this. + for (auto& byte : kBytes) { + file_ptr_out.write(reinterpret_cast(&byte), sizeof(byte)); + } + + if (kVerbose) kStdOut << "AssemblerPower: Wrote file with program in it.\n"; + + file_ptr_out.flush(); + file_ptr_out.close(); + + if (kVerbose) kStdOut << "AssemblerPower: Exit succeeded.\n"; + + return 0; + } + +asm_fail_exit: + + if (kVerbose) kStdOut << "AssemblerPower: Exit failed.\n"; + + return LIBCOMPILER_EXEC_ERROR; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for attributes +// returns true if any was found. + +///////////////////////////////////////////////////////////////////////////////////////// + +static bool asm_read_attributes(std::string line) { + // extern_segment is the opposite of public_segment, it signals to the li + // that we need this symbol. + if (LibCompiler::find_word(line, "extern_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid extern_segment directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_extern_segment_bin"); + } + + auto name = line.substr(line.find("extern_segment") + strlen("extern_segment") + 1); + + if (name.size() == 0) { + Detail::print_error("Invalid extern_segment", "LibCompiler"); + throw std::runtime_error("invalid_extern_segment"); + } + + std::string result = std::to_string(name.size()); + result += kUndefinedSymbol; + + // mangle this + for (char& j : name) { + if (j == ' ' || j == ',') j = '$'; + } + + result += name; + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that li can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, result.c_str(), result.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + // public_segment is a special keyword used by AssemblerPower to tell the AE output stage to + // mark this section as a header. it currently supports .code64, .data64., + // .zero64 + else if (LibCompiler::find_word(line, "public_segment")) { + if (kOutputAsBinary) { + Detail::print_error("Invalid public_segment directive in flat binary mode.", "LibCompiler"); + throw std::runtime_error("invalid_public_segment_bin"); + } + + auto name = line.substr(line.find("public_segment") + strlen("public_segment")); + + std::string name_copy = name; + + for (char& j : name) { + if (j == ' ') j = '$'; + } + + if (name.find(".code64") != std::string::npos) { + // data is treated as code. + + name_copy.erase(name_copy.find(".code64"), strlen(".code64")); + kCurrentRecord.fKind = LibCompiler::kPefCode; + } else if (name.find(".data64") != std::string::npos) { + // no code will be executed from here. + + name_copy.erase(name_copy.find(".data64"), strlen(".data64")); + kCurrentRecord.fKind = LibCompiler::kPefData; + } else if (name.find(".zero64") != std::string::npos) { + // this is a bss section. + + name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); + kCurrentRecord.fKind = LibCompiler::kPefZero; + } + + // this is a special case for the start stub. + // we want this so that li can find it. + + if (name == kPefStart) { + kCurrentRecord.fKind = LibCompiler::kPefCode; + } + + while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); + + kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); + ++kOrigin; + + // now we can tell the code size of the previous kCurrentRecord. + + if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); + + memset(kCurrentRecord.fName, 0, kAESymbolLen); + memcpy(kCurrentRecord.fName, name.c_str(), name.size()); + + ++kCounter; + + memset(kCurrentRecord.fPad, kAENullType, kAEPad); + + kRecords.emplace_back(kCurrentRecord); + + return true; + } + + return false; +} + +// \brief algorithms and helpers. + +namespace Detail::algorithm { +// \brief authorize a brief set of characters. +static inline bool is_not_alnum_space(char c) { + return !(isalpha(c) || isdigit(c) || (c == ' ') || (c == '\t') || (c == ',') || (c == '(') || + (c == ')') || (c == '"') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || + (c == '_') || (c == ':') || (c == '@') || (c == '.')); +} + +bool is_valid_power64(std::string str) { + return std::find_if(str.begin(), str.end(), is_not_alnum_space) == str.end(); +} +} // namespace Detail::algorithm + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Check for line (syntax check) + +///////////////////////////////////////////////////////////////////////////////////////// + +std::string LibCompiler::EncoderPowerPC::CheckLine(std::string line, std::string file) { + std::string err_str; + + if (line.empty() || LibCompiler::find_word(line, "extern_segment") || + LibCompiler::find_word(line, "public_segment") || line.find('#') != std::string::npos || + LibCompiler::find_word(line, ";")) { + if (line.find('#') != std::string::npos) { + line.erase(line.find('#')); + } else if (line.find(';') != std::string::npos) { + line.erase(line.find(';')); + } else { + /// does the line contains valid input? + if (!Detail::algorithm::is_valid_power64(line)) { + err_str = "Line contains non alphanumeric characters.\nhere -> "; + err_str += line; + } + } + + return err_str; + } + + if (!Detail::algorithm::is_valid_power64(line)) { + err_str = "Line contains non alphanumeric characters.\nhere -> "; + err_str += line; + + return err_str; + } + + // check for a valid instruction format. + + if (line.find(',') != std::string::npos) { + if (line.find(',') + 1 == line.size()) { + err_str += "\nInstruction lacks right register, here -> "; + err_str += line.substr(line.find(',')); + + return err_str; + } else { + bool nothing_on_right = true; + + if (line.find(',') + 1 > line.size()) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + + auto substr = line.substr(line.find(',') + 1); + + for (auto& ch : substr) { + if (ch != ' ' && ch != '\t') { + nothing_on_right = false; + } + } + + // this means we found nothing after that ',' . + if (nothing_on_right) { + err_str += "\nInstruction not complete, here -> "; + err_str += line; + + return err_str; + } + } + } + + // these do take an argument. + std::vector operands_inst = {"stw", "li"}; + + // these don't. + std::vector filter_inst = {"blr", "bl", "sc"}; + + for (auto& opcode_risc : kOpcodesPowerPC) { + if (LibCompiler::find_word(line, opcode_risc.name)) { + for (auto& op : operands_inst) { + // if only the instruction was found. + if (line == op) { + err_str += "\nMalformed "; + err_str += op; + err_str += " instruction, here -> "; + err_str += line; + } + } + + // if it is like that -> addr1, 0x0 + if (auto it = std::find(filter_inst.begin(), filter_inst.end(), opcode_risc.name); + it == filter_inst.cend()) { + if (LibCompiler::find_word(line, opcode_risc.name)) { + if (!isspace(line[line.find(opcode_risc.name) + strlen(opcode_risc.name)])) { + err_str += "\nMissing space between "; + err_str += opcode_risc.name; + err_str += " and operands.\nhere -> "; + err_str += line; + } + } + } + + return err_str; + } + } + + err_str += "Unrecognized instruction: " + line; + + return err_str; +} + +bool LibCompiler::EncoderPowerPC::WriteNumber(const std::size_t& pos, std::string& jump_label) { + if (!isdigit(jump_label[pos])) return false; + + switch (jump_label[pos + 1]) { + case 'x': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { + if (errno != 0) { + Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_hex"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); + + for (char& i : num.number) { + kBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerPower: found a base 16 number here: " << jump_label.substr(pos) + << "\n"; + } + + return true; + } + case 'b': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { + if (errno != 0) { + Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_bin"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); + + if (kVerbose) { + kStdOut << "AssemblerPower: found a base 2 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + kBytes.push_back(i); + } + + return true; + } + case 'o': { + if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { + if (errno != 0) { + Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); + throw std::runtime_error("invalid_octal"); + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); + + if (kVerbose) { + kStdOut << "AssemblerPower: found a base 8 number here: " << jump_label.substr(pos) << "\n"; + } + + for (char& i : num.number) { + kBytes.push_back(i); + } + + return true; + } + default: { + break; + } + } + + /* check for errno and stuff like that */ + if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { + if (errno != 0) { + return false; + } + } + + LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); + + for (char& i : num.number) { + kBytes.push_back(i); + } + + if (kVerbose) { + kStdOut << "AssemblerPower: found a base 10 number here: " << jump_label.substr(pos) << "\n"; + } + + return true; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @brief Read and write an instruction to the output array. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool LibCompiler::EncoderPowerPC::WriteLine(std::string line, std::string file) { + if (LibCompiler::find_word(line, "public_segment")) return false; + if (!Detail::algorithm::is_valid_power64(line)) return false; + + for (auto& opcode_risc : kOpcodesPowerPC) { + // strict check here + if (LibCompiler::find_word(line, opcode_risc.name)) { + std::string name(opcode_risc.name); + std::string jump_label, cpy_jump_label; + std::vector found_registers_index; + + // check funct7 type. + switch (opcode_risc.ops->type) { + default: { + NumberCast32 num(opcode_risc.opcode); + + for (auto ch : num.number) { + kBytes.emplace_back(ch); + } + break; + } + case BADDR: + case PCREL: { + auto num = GetNumber32(line, name); + + kBytes.emplace_back(num.number[0]); + kBytes.emplace_back(num.number[1]); + kBytes.emplace_back(num.number[2]); + kBytes.emplace_back(0x48); + + break; + } + /// General purpose, float, vector operations. Everything that involve + /// registers. + case G0REG: + case FREG: + case VREG: + case GREG: { + // \brief how many registers we found. + std::size_t found_some_count = 0UL; + std::size_t register_count = 0UL; + std::string opcodeName = opcode_risc.name; + std::size_t register_sum = 0; + + NumberCast64 num(opcode_risc.opcode); + + for (size_t line_index = 0UL; line_index < line.size(); line_index++) { + if (line[line_index] == kAsmRegisterPrefix[0] && isdigit(line[line_index + 1])) { + std::string register_syntax = kAsmRegisterPrefix; + register_syntax += line[line_index + 1]; + + if (isdigit(line[line_index + 2])) register_syntax += line[line_index + 2]; + + std::string reg_str; + reg_str += line[line_index + 1]; + + if (isdigit(line[line_index + 2])) reg_str += line[line_index + 2]; + + // it ranges from r0 to r19 + // something like r190 doesn't exist in the instruction set. + if (isdigit(line[line_index + 3]) && isdigit(line[line_index + 2])) { + reg_str += line[line_index + 3]; + Detail::print_error("invalid register index, r" + reg_str + + "\nnote: The POWER accepts registers from r0 to r32.", + file); + throw std::runtime_error("invalid_register_index"); + } + + // finally cast to a size_t + std::size_t reg_index = strtol(reg_str.c_str(), nullptr, 10); + + if (reg_index > kAsmRegisterLimit) { + Detail::print_error("invalid register index, r" + reg_str, file); + throw std::runtime_error("invalid_register_index"); + } + + if (opcodeName == "li") { + char numIndex = 0; + + for (size_t i = 0; i != reg_index; i++) { + numIndex += 0x20; + } + + auto num = GetNumber32(line, reg_str); + + kBytes.push_back(num.number[0]); + kBytes.push_back(num.number[1]); + kBytes.push_back(numIndex); + kBytes.push_back(0x38); + + // check if bigger than two. + for (size_t i = 2; i < 4; i++) { + if (num.number[i] > 0) { + Detail::print_warning("number overflow on li operation.", file); + break; + } + } + + break; + } + + if ((opcodeName[0] == 's' && opcodeName[1] == 't')) { + if (register_sum == 0) { + for (size_t indexReg = 0UL; indexReg < reg_index; ++indexReg) { + register_sum += 0x20; + } + } else { + register_sum += reg_index; + } + } + + if (opcodeName == "mr") { + switch (register_count) { + case 0: { + kBytes.push_back(0x78); + + char numIndex = 0x3; + + for (size_t i = 0; i != reg_index; i++) { + numIndex += 0x8; + } + + kBytes.push_back(numIndex); + + break; + } + case 1: { + char numIndex = 0x1; + + for (size_t i = 0; i != reg_index; i++) { + numIndex += 0x20; + } + + for (size_t i = 0; i != reg_index; i++) { + kBytes[kBytes.size() - 1] += 0x8; + } + + kBytes[kBytes.size() - 1] -= 0x8; + + kBytes.push_back(numIndex); + + if (reg_index >= 10 && reg_index < 20) + kBytes.push_back(0x7d); + else if (reg_index >= 20 && reg_index < 30) + kBytes.push_back(0x7e); + else if (reg_index >= 30) + kBytes.push_back(0x7f); + else + kBytes.push_back(0x7c); + + break; + } + default: + break; + } + + ++register_count; + ++found_some_count; + } + + if (opcodeName == "addi") { + if (found_some_count == 2 || found_some_count == 0) + kBytes.emplace_back(reg_index); + else if (found_some_count == 1) + kBytes.emplace_back(0x00); + + ++found_some_count; + + if (found_some_count > 3) { + Detail::print_error("Too much registers. -> " + line, file); + throw std::runtime_error("too_much_regs"); + } + } + + if (opcodeName.find("cmp") != std::string::npos) { + ++found_some_count; + + if (found_some_count > 3) { + Detail::print_error("Too much registers. -> " + line, file); + throw std::runtime_error("too_much_regs"); + } + } + + if (opcodeName.find("mf") != std::string::npos || + opcodeName.find("mt") != std::string::npos) { + char numIndex = 0; + + for (size_t i = 0; i != reg_index; i++) { + numIndex += 0x20; + } + + num.number[2] += numIndex; + + ++found_some_count; + + if (found_some_count > 1) { + Detail::print_error("Too much registers. -> " + line, file); + throw std::runtime_error("too_much_regs"); + } + + if (kVerbose) { + kStdOut << "AssemblerPower: Found register: " << register_syntax << "\n"; + kStdOut << "AssemblerPower: Amount of registers in instruction: " + << found_some_count << "\n"; + } + + if (reg_index >= 10 && reg_index < 20) + num.number[3] = 0x7d; + else if (reg_index >= 20 && reg_index < 30) + num.number[3] = 0x7e; + else if (reg_index >= 30) + num.number[3] = 0x7f; + else + num.number[3] = 0x7c; + + for (auto ch : num.number) { + kBytes.emplace_back(ch); + } + } + + found_registers_index.push_back(reg_index); + } + } + + if (opcodeName == "addi") { + kBytes.emplace_back(0x38); + } + + if (opcodeName.find("cmp") != std::string::npos) { + char rightReg = 0x0; + + for (size_t i = 0; i != found_registers_index[1]; i++) { + rightReg += 0x08; + } + + kBytes.emplace_back(0x00); + kBytes.emplace_back(rightReg); + kBytes.emplace_back(found_registers_index[0]); + kBytes.emplace_back(0x7c); + } + + if ((opcodeName[0] == 's' && opcodeName[1] == 't')) { + size_t offset = 0UL; + + if (line.find('+') != std::string::npos) { + auto number = GetNumber32(line.substr(line.find("+")), "+"); + offset = number.raw; + } + + kBytes.push_back(offset); + kBytes.push_back(0x00); + kBytes.push_back(register_sum); + + kBytes.emplace_back(0x90); + } + + if (opcodeName == "mr") { + if (register_count == 1) { + Detail::print_error("Too few registers. -> " + line, file); + throw std::runtime_error("too_few_registers"); + } + } + + // we're not in immediate addressing, reg to reg. + if (opcode_risc.ops->type != GREG) { + // remember! register to register! + if (found_some_count == 1) { + Detail::print_error( + "Unrecognized register found.\ntip: each AssemblerPower register " + "starts with 'r'.\nline: " + + line, + file); + + throw std::runtime_error("not_a_register"); + } + } + + if (found_some_count < 1 && name[0] != 'l' && name[0] != 's') { + Detail::print_error("invalid combination of opcode and registers.\nline: " + line, + file); + throw std::runtime_error("invalid_comb_op_reg"); + } + + break; + } + } + + kOrigin += cPowerIPAlignment; + break; + } + } + + return true; +} + +// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/Assembler32x0.cc b/dev/LibCompiler/src/Assembler32x0.cc deleted file mode 100644 index 44850f6..0000000 --- a/dev/LibCompiler/src/Assembler32x0.cc +++ /dev/null @@ -1,49 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -/// bugs: 0 - -///////////////////////////////////////////////////////////////////////////////////////// - -// @file 32asm.cxx -// @author EL Mahrouss Amlal -// @brief 32x0 Assembler. - -// REMINDER: when dealing with an undefined symbol use (string -// size):LinkerFindSymbol:(string) so that ld will look for it. - -///////////////////////////////////////////////////////////////////////////////////////// - -#define __ASM_NEED_32x0__ 1 - -#include -#include -#include -#include - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" -#define kYellow "\e[0;33m" - -#define kStdOut (std::cout << kWhite) -#define kStdErr (std::cout << kRed) - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief 32x0 Assembler entrypoint, the program/module starts here. - -///////////////////////////////////////////////////////////////////////////////////////// - -LIBCOMPILER_MODULE(NEAssemblerMain32000) { - return 0; -} diff --git a/dev/LibCompiler/src/Assembler64x0.cc b/dev/LibCompiler/src/Assembler64x0.cc deleted file mode 100644 index eb5b4dd..0000000 --- a/dev/LibCompiler/src/Assembler64x0.cc +++ /dev/null @@ -1,873 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -/// bugs: 0 - -///////////////////////////////////////////////////////////////////////////////////////// - -// @file Assembler64x0.cxx -// @author EL Mahrouss Amlal -// @brief 64x000 Assembler. - -// REMINDER: when dealing with an undefined symbol use (string -// size):LinkerFindSymbol:(string) so that ld will look for it. - -///////////////////////////////////////////////////////////////////////////////////////// - -#define __ASM_NEED_64x0__ 1 - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -static char kOutputArch = LibCompiler::kPefArch64000; - -/// @note The 64x0 is VLSIW, so we need to jump to 4 bytes. -constexpr auto k64x0IPAlignment = 0x4U; - -static std::size_t kCounter = 1UL; - -static std::uintptr_t kOrigin = kPefBaseOrigin; -static std::vector> kOriginLabel; - -static std::vector kBytes; - -static LibCompiler::AERecordHeader kCurrentRecord{ - .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; - -static std::vector kRecords; -static std::vector kUndefinedSymbols; - -static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; -static const std::string kRelocSymbol = ":RuntimeSymbol:"; - -// \brief forward decl. -static bool asm_read_attributes(std::string line); - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief 64x0 assembler entrypoint, the program/module starts here. - -///////////////////////////////////////////////////////////////////////////////////////// - -LIBCOMPILER_MODULE(AssemblerMain64x0) { - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - for (size_t i = 1; i < argc; ++i) { - if (argv[i][0] == '-') { - if (strcmp(argv[i], "-64x0-ver") == 0 || strcmp(argv[i], "-64x0-v") == 0) { - kStdOut - << "Assembler64x0: 64x0 Assembler.\nAssembler64x0: v1.10\nAssembler64x0: Copyright (c) " - "Amlal El Mahrouss\n"; - return 0; - } else if (strcmp(argv[i], "-64x0-h") == 0) { - kStdOut << "Assembler64x0: 64x0 Assembler.\nAssembler64x0: Copyright (c) 2024 Mahrouss " - "Logic.\n"; - kStdOut << "--version: Print program version.\n"; - kStdOut << "--verbose: Print verbose output.\n"; - kStdOut << "--binary: Output as flat binary.\n"; - kStdOut << "--64xxx: Compile for a subset of the X64000.\n"; - - return 0; - } else if (strcmp(argv[i], "-64x0-binary") == 0) { - kOutputAsBinary = true; - continue; - } else if (strcmp(argv[i], "-64x0-verbose") == 0) { - kVerbose = true; - continue; - } - - kStdOut << "Assembler64x0: ignore " << argv[i] << "\n"; - continue; - } - - if (!std::filesystem::exists(argv[i])) { - kStdOut << "Assembler64x0: can't open: " << argv[i] << std::endl; - goto asm_fail_exit; - } - - std::string object_output(argv[i]); - - for (auto& ext : kAsmFileExts) { - if (object_output.find(ext) != std::string::npos) { - object_output.erase(object_output.find(ext), std::strlen(ext)); - } - } - - object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; - - std::ifstream file_ptr(argv[i]); - std::ofstream file_ptr_out(object_output, std::ofstream::binary); - - if (file_ptr_out.bad()) { - if (kVerbose) { - kStdOut << "Assembler64x0: error: " << strerror(errno) << "\n"; - } - } - - std::string line; - - LibCompiler::AEHeader hdr{0}; - - memset(hdr.fPad, kAENullType, kAEPad); - - hdr.fMagic[0] = kAEMag0; - hdr.fMagic[1] = kAEMag1; - hdr.fSize = sizeof(LibCompiler::AEHeader); - hdr.fArch = kOutputArch; - - ///////////////////////////////////////////////////////////////////////////////////////// - - // COMPILATION LOOP - - ///////////////////////////////////////////////////////////////////////////////////////// - - LibCompiler::Encoder64x0 asm64; - - while (std::getline(file_ptr, line)) { - if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { - Detail::print_error(ln, argv[i]); - continue; - } - - try { - asm_read_attributes(line); - asm64.WriteLine(line, argv[i]); - } catch (const std::exception& e) { - if (kVerbose) { - std::string what = e.what(); - Detail::print_warning("exit because of: " + what, "LibCompiler"); - } - - std::filesystem::remove(object_output); - goto asm_fail_exit; - } - } - - if (!kOutputAsBinary) { - if (kVerbose) { - kStdOut << "Assembler64x0: Writing object file...\n"; - } - - // this is the final step, write everything to the file. - - auto pos = file_ptr_out.tellp(); - - hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); - - file_ptr_out << hdr; - - if (kRecords.empty()) { - kStdErr << "Assembler64x0: At least one record is needed to write an object " - "file.\nAssembler64x0: Make one using `public_segment .code64 foo_bar`.\n"; - - std::filesystem::remove(object_output); - return 1; - } - - kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - std::size_t record_count = 0UL; - - for (auto& rec : kRecords) { - if (kVerbose) kStdOut << "Assembler64x0: Wrote record " << rec.fName << " to file...\n"; - - rec.fFlags |= LibCompiler::kKindRelocationAtRuntime; - rec.fOffset = record_count; - ++record_count; - - file_ptr_out << rec; - } - - // increment once again, so that we won't lie about the kUndefinedSymbols. - ++record_count; - - for (auto& sym : kUndefinedSymbols) { - LibCompiler::AERecordHeader _record_hdr{0}; - - if (kVerbose) kStdOut << "Assembler64x0: Wrote symbol " << sym << " to file...\n"; - - _record_hdr.fKind = kAENullType; - _record_hdr.fSize = sym.size(); - _record_hdr.fOffset = record_count; - - ++record_count; - - memset(_record_hdr.fPad, kAENullType, kAEPad); - memcpy(_record_hdr.fName, sym.c_str(), sym.size()); - - file_ptr_out << _record_hdr; - - ++kCounter; - } - - auto pos_end = file_ptr_out.tellp(); - - file_ptr_out.seekp(pos); - - hdr.fStartCode = pos_end; - hdr.fCodeSize = kBytes.size(); - - file_ptr_out << hdr; - - file_ptr_out.seekp(pos_end); - } else { - if (kVerbose) { - kStdOut << "Assembler64x0: Write raw binary...\n"; - } - } - - // byte from byte, we write this. - for (auto& byte : kBytes) { - file_ptr_out.write(reinterpret_cast(&byte), sizeof(byte)); - } - - if (kVerbose) kStdOut << "Assembler64x0: Wrote file with program in it.\n"; - - file_ptr_out.flush(); - file_ptr_out.close(); - - if (kVerbose) kStdOut << "Assembler64x0: Exit succeeded.\n"; - - return 0; - } - -asm_fail_exit: - - if (kVerbose) kStdOut << "Assembler64x0: Exit failed.\n"; - - return 1; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for attributes -// returns true if any was found. - -///////////////////////////////////////////////////////////////////////////////////////// - -static bool asm_read_attributes(std::string line) { - // extern_segment is the opposite of public_segment, it signals to the ld - // that we need this symbol. - if (LibCompiler::find_word(line, "extern_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid extern_segment directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_extern_segment_bin"); - } - - auto name = line.substr(line.find("extern_segment") + strlen("extern_segment")); - - /// sanity check to avoid stupid linker errors. - if (name.size() == 0) { - Detail::print_error("Invalid extern_segment", "power-as"); - throw std::runtime_error("invalid_extern_segment"); - } - - std::string result = std::to_string(name.size()); - result += kUndefinedSymbol; - - // mangle this - for (char& j : name) { - if (j == ' ' || j == ',') j = '$'; - } - - result += name; - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that ld can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, result.c_str(), result.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - // public_segment is a special keyword used by Assembler64x0 to tell the AE output stage to - // mark this section as a header. it currently supports .code64, .data64., - // .zero64 - else if (LibCompiler::find_word(line, "public_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid public_segment directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_public_segment_bin"); - } - - auto name = line.substr(line.find("public_segment") + strlen("public_segment")); - - std::string name_copy = name; - - for (char& j : name) { - if (j == ' ') j = '$'; - } - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - - name_copy.erase(name_copy.find(".code64"), strlen(".code64")); - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - - name_copy.erase(name_copy.find(".data64"), strlen(".data64")); - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - - name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that ld can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); - - kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); - ++kOrigin; - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, name.c_str(), name.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - - return false; -} - -// \brief algorithms and helpers. - -namespace Detail::algorithm { -// \brief authorize a brief set of characters. -static inline bool is_not_alnum_space(char c) { - return !(isalpha(c) || isdigit(c) || (c == ' ') || (c == '\t') || (c == ',') || (c == '(') || - (c == ')') || (c == '"') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || - (c == '_') || (c == ':') || (c == '@') || (c == '.')); -} - -bool is_valid_64x0(std::string str) { - return std::find_if(str.begin(), str.end(), is_not_alnum_space) == str.end(); -} -} // namespace Detail::algorithm - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for line (syntax check) - -///////////////////////////////////////////////////////////////////////////////////////// - -std::string LibCompiler::Encoder64x0::CheckLine(std::string line, std::string file) { - std::string err_str; - - if (line.empty() || LibCompiler::find_word(line, "extern_segment") || - LibCompiler::find_word(line, "public_segment") || line.find('#') != std::string::npos || - LibCompiler::find_word(line, ";")) { - if (line.find('#') != std::string::npos) { - line.erase(line.find('#')); - } else if (line.find(';') != std::string::npos) { - line.erase(line.find(';')); - } else { - // now check the line for validity - if (!Detail::algorithm::is_valid_64x0(line)) { - err_str = "Line contains non alphanumeric characters.\nhere -> "; - err_str += line; - } - } - - return err_str; - } - - if (!Detail::algorithm::is_valid_64x0(line)) { - err_str = "Line contains non alphanumeric characters.\nhere -> "; - err_str += line; - - return err_str; - } - - // check for a valid instruction format. - - if (line.find(',') != std::string::npos) { - if (line.find(',') + 1 == line.size()) { - err_str += "\nInstruction lacks right register, here -> "; - err_str += line.substr(line.find(',')); - - return err_str; - } else { - bool nothing_on_right = true; - - if (line.find(',') + 1 > line.size()) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - - auto substr = line.substr(line.find(',') + 1); - - for (auto& ch : substr) { - if (ch != ' ' && ch != '\t') { - nothing_on_right = false; - } - } - - // this means we found nothing after that ',' . - if (nothing_on_right) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - } - } - - // these do take an argument. - std::vector operands_inst = {"stw", "ldw", "lda", "sta"}; - - // these don't. - std::vector filter_inst = {"jlr", "jrl", "int"}; - - for (auto& opcode64x0 : kOpcodes64x0) { - if (line.find(opcode64x0.fName) != std::string::npos) { - if (opcode64x0.fFunct7 == kAsmNoArgs) return err_str; - - for (auto& op : operands_inst) { - // if only the instruction was found. - if (line == op) { - err_str += "\nMalformed "; - err_str += op; - err_str += " instruction, here -> "; - err_str += line; - } - } - - // if it is like that -> addr1, 0x0 - if (auto it = std::find(filter_inst.begin(), filter_inst.end(), opcode64x0.fName); - it == filter_inst.cend()) { - if (LibCompiler::find_word(line, opcode64x0.fName)) { - if (!isspace(line[line.find(opcode64x0.fName) + strlen(opcode64x0.fName)])) { - err_str += "\nMissing space between "; - err_str += opcode64x0.fName; - err_str += " and operands.\nhere -> "; - err_str += line; - } - } - } - - return err_str; - } - } - - err_str += "Unrecognized instruction: " + line; - - return err_str; -} - -bool LibCompiler::Encoder64x0::WriteNumber(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_hex_number"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); - - for (char& i : num.number) { - kBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "Assembler64x0: found a base 16 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; - } - case 'b': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "Assembler64x0: found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - kBytes.push_back(i); - } - - return true; - } - case 'o': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "Assembler64x0: found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - kBytes.push_back(i); - } - - return true; - } - default: { - break; - } - } - - /* check for errno and stuff like that */ - if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - return false; - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); - - for (char& i : num.number) { - kBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "Assembler64x0: found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Read and write an instruction to the output array. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool LibCompiler::Encoder64x0::WriteLine(std::string line, std::string file) { - if (LibCompiler::find_word(line, "public_segment ")) return true; - - for (auto& opcode64x0 : kOpcodes64x0) { - // strict check here - if (LibCompiler::find_word(line, opcode64x0.fName) && Detail::algorithm::is_valid_64x0(line)) { - std::string name(opcode64x0.fName); - std::string jump_label, cpy_jump_label; - - kBytes.emplace_back(opcode64x0.fOpcode); - kBytes.emplace_back(opcode64x0.fFunct3); - kBytes.emplace_back(opcode64x0.fFunct7); - - // check funct7 type. - switch (opcode64x0.fFunct7) { - // reg to reg means register to register transfer operation. - case kAsmRegToReg: - case kAsmImmediate: { - // \brief how many registers we found. - std::size_t found_some = 0UL; - - for (size_t line_index = 0UL; line_index < line.size(); line_index++) { - if (line[line_index] == kAsmRegisterPrefix[0] && isdigit(line[line_index + 1])) { - std::string register_syntax = kAsmRegisterPrefix; - register_syntax += line[line_index + 1]; - - if (isdigit(line[line_index + 2])) register_syntax += line[line_index + 2]; - - std::string reg_str; - reg_str += line[line_index + 1]; - - if (isdigit(line[line_index + 2])) reg_str += line[line_index + 2]; - - // it ranges from r0 to r19 - // something like r190 doesn't exist in the instruction set. - if (kOutputArch == LibCompiler::kPefArch64000) { - if (isdigit(line[line_index + 3]) && isdigit(line[line_index + 2])) { - reg_str += line[line_index + 3]; - Detail::print_error("invalid register index, r" + reg_str + - "\nnote: The 64x0 accepts registers from r0 to r20.", - file); - throw std::runtime_error("invalid_register_index"); - } - } - - // finally cast to a size_t - std::size_t reg_index = strtol(reg_str.c_str(), nullptr, 10); - - if (reg_index > kAsmRegisterLimit) { - Detail::print_error("invalid register index, r" + reg_str, file); - throw std::runtime_error("invalid_register_index"); - } - - kBytes.emplace_back(reg_index); - ++found_some; - - if (kVerbose) { - kStdOut << "Assembler64x0: Register found: " << register_syntax << "\n"; - kStdOut << "Assembler64x0: Register amount in instruction: " << found_some << "\n"; - } - } - } - - // we're not in immediate addressing, reg to reg. - if (opcode64x0.fFunct7 != kAsmImmediate) { - // remember! register to register! - if (found_some == 1) { - Detail::print_error( - "Too few registers.\ntip: each Assembler64x0 register " - "starts with 'r'.\nline: " + - line, - file); - throw std::runtime_error("not_a_register"); - } - } - - if (found_some < 1 && name != "ldw" && name != "lda" && name != "stw") { - Detail::print_error("invalid combination of opcode and registers.\nline: " + line, - file); - throw std::runtime_error("invalid_comb_op_reg"); - } else if (found_some == 1 && name == "add") { - Detail::print_error("invalid combination of opcode and registers.\nline: " + line, - file); - throw std::runtime_error("invalid_comb_op_reg"); - } else if (found_some == 1 && name == "sub") { - Detail::print_error("invalid combination of opcode and registers.\nline: " + line, - file); - throw std::runtime_error("invalid_comb_op_reg"); - } - - if (found_some > 0 && name == "pop") { - Detail::print_error( - "invalid combination for opcode 'pop'.\ntip: it expects " - "nothing.\nline: " + - line, - file); - throw std::runtime_error("invalid_comb_op_pop"); - } - } - default: - break; - } - - // try to fetch a number from the name - if (name == "stw" || name == "ldw" || name == "lda" || name == "sta") { - auto where_string = name; - - // if we load something, we'd need it's symbol/literal - if (name == "stw" || name == "sta" || name == "ldw" || name == "lda" || name == "sta") - where_string = ","; - - jump_label = line; - - auto found_sym = false; - - while (jump_label.find(where_string) != std::string::npos) { - jump_label = jump_label.substr(jump_label.find(where_string) + where_string.size()); - - while (jump_label.find(" ") != std::string::npos) { - jump_label.erase(jump_label.find(" "), 1); - } - - if (jump_label[0] != kAsmRegisterPrefix[0] && !isdigit(jump_label[1])) { - if (found_sym) { - Detail::print_error( - "invalid combination of opcode and operands.\nhere -> " + jump_label, file); - throw std::runtime_error("invalid_comb_op_ops"); - } else { - // death trap installed. - found_sym = true; - } - } - } - - cpy_jump_label = jump_label; - - // replace any spaces with $ - if (jump_label[0] == ' ') { - while (jump_label.find(' ') != std::string::npos) { - if (isalnum(jump_label[0]) || isdigit(jump_label[0])) break; - - jump_label.erase(jump_label.find(' '), 1); - } - } - - if (!this->WriteNumber(0, jump_label)) { - // sta expects this: sta 0x000000, r0 - if (name == "sta") { - Detail::print_error("invalid combination of opcode and operands.\nHere ->" + line, - file); - throw std::runtime_error("invalid_comb_op_ops"); - } - } else { - if (name == "sta" && cpy_jump_label.find("extern_segment ") != std::string::npos) { - Detail::print_error("invalid usage extern_segment on 'sta', here: " + line, file); - throw std::runtime_error("invalid_sta_usage"); - } - } - - goto asm_write_label; - } - - // This is the case where we jump to a label, it is also used as a goto. - if (name == "lda" || name == "sta") { - asm_write_label: - if (cpy_jump_label.find('\n') != std::string::npos) - cpy_jump_label.erase(cpy_jump_label.find('\n'), 1); - - if (cpy_jump_label.find("extern_segment") != std::string::npos) { - cpy_jump_label.erase(cpy_jump_label.find("extern_segment"), strlen("extern_segment")); - - if (name == "sta") { - Detail::print_error("extern_segment is not allowed on a sta operation.", file); - throw std::runtime_error("extern_segment_sta_op"); - } else { - goto asm_end_label_cpy; - } - } - - if (name == "lda" || name == "sta") { - for (auto& label : kOriginLabel) { - if (cpy_jump_label == label.first) { - if (kVerbose) { - kStdOut << "Assembler64x0: Replace label " << cpy_jump_label - << " to address: " << label.second << std::endl; - } - - LibCompiler::NumberCast64 num(label.second); - - for (auto& num : num.number) { - kBytes.push_back(num); - } - - goto asm_end_label_cpy; - } - } - - if (cpy_jump_label[0] == '0') { - switch (cpy_jump_label[1]) { - case 'x': - case 'o': - case 'b': - if (this->WriteNumber(0, cpy_jump_label)) goto asm_end_label_cpy; - - break; - default: - break; - } - - if (isdigit(cpy_jump_label[0])) { - if (this->WriteNumber(0, cpy_jump_label)) goto asm_end_label_cpy; - - break; - } - } - } - - if (cpy_jump_label.size() < 1) { - Detail::print_error("label is empty, can't jump on it.", file); - throw std::runtime_error("label_empty"); - } - - /// don't go any further if: - /// load word (ldw) or store word. (stw) - - if (name == "ldw" || name == "stw") break; - - auto mld_reloc_str = std::to_string(cpy_jump_label.size()); - mld_reloc_str += kUndefinedSymbol; - mld_reloc_str += cpy_jump_label; - - bool ignore_back_slash = false; - - for (auto& reloc_chr : mld_reloc_str) { - if (reloc_chr == '\\') { - ignore_back_slash = true; - continue; - } - - if (ignore_back_slash) { - ignore_back_slash = false; - continue; - } - - kBytes.push_back(reloc_chr); - } - - kBytes.push_back('\0'); - goto asm_end_label_cpy; - } - - asm_end_label_cpy: - kOrigin += k64x0IPAlignment; - - break; - } - } - - return true; -} - -// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/AssemblerAMD64.cc b/dev/LibCompiler/src/AssemblerAMD64.cc deleted file mode 100644 index f017f0d..0000000 --- a/dev/LibCompiler/src/AssemblerAMD64.cc +++ /dev/null @@ -1,1203 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @file AssemblerAMD64.cc -/// @author EL Mahrouss Amlal -/// @brief AMD64 Assembler. -/// REMINDER: when dealing with an undefined symbol use (string -/// size):LinkerFindSymbol:(string) so that ld will look for it. - -///////////////////////////////////////////////////////////////////////////////////////// - -/// bugs: 0 - -/// feature request: 1 -/// Encode registers in mov, add, xor... - -///////////////////////////////////////////////////////////////////////////////////////// - -#define __ASM_NEED_AMD64__ 1 - -#define kAssemblerPragmaSymStr "#" -#define kAssemblerPragmaSym '#' - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" -#define kYellow "\e[0;33m" - -static char kOutputArch = LibCompiler::kPefArchAMD64; - -constexpr auto kIPAlignement = 0x1U; - -static std::size_t kCounter = 1UL; - -static std::uintptr_t kOrigin = kPefBaseOrigin; -static std::vector> kOriginLabel; - -/// @brief keep it simple by default. -static std::int32_t kRegisterBitWidth = 16U; - -static std::vector kAppBytes; - -static LibCompiler::AERecordHeader kCurrentRecord{ - .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; - -static std::vector kRecords; -static std::vector kDefinedSymbols; -static std::vector kUndefinedSymbols; - -static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; - -// \brief forward decl. -static bool asm_read_attributes(std::string line); - -#include - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief AMD64 assembler entrypoint, the program/module starts here. - -///////////////////////////////////////////////////////////////////////////////////////// - -LIBCOMPILER_MODULE(AssemblerMainAMD64) { - //////////////// CPU OPCODES BEGIN //////////////// - - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - std::string opcodes_jump[kJumpLimit] = {"ja", "jae", "jb", "jbe", "jc", "je", "jg", "jge", - "jl", "jle", "jna", "jnae", "jnb", "jnbe", "jnc", "jne", - "jng", "jnge", "jnl", "jnle", "jno", "jnp", "jns", "jnz", - "jo", "jp", "jpe", "jpo", "js", "jz"}; - - for (i64_hword_t i = 0; i < kJumpLimit; i++) { - CpuOpcodeAMD64 code{.fName = opcodes_jump[i], - .fOpcode = static_cast(kAsmJumpOpcode + i)}; - kOpcodesAMD64.push_back(code); - } - - CpuOpcodeAMD64 code{.fName = "jcxz", .fOpcode = 0xE3}; - kOpcodesAMD64.push_back(code); - - for (i64_hword_t i = kJumpLimitStandard; i < kJumpLimitStandardLimit; i++) { - CpuOpcodeAMD64 code{.fName = "jmp", .fOpcode = i}; - kOpcodesAMD64.push_back(code); - } - - CpuOpcodeAMD64 lahf{.fName = "lahf", .fOpcode = 0x9F}; - kOpcodesAMD64.push_back(lahf); - - CpuOpcodeAMD64 lds{.fName = "lds", .fOpcode = 0xC5}; - kOpcodesAMD64.push_back(lds); - - CpuOpcodeAMD64 lea{.fName = "lea", .fOpcode = 0x8D}; - kOpcodesAMD64.push_back(lea); - - CpuOpcodeAMD64 nop{.fName = "nop", .fOpcode = 0x90}; - kOpcodesAMD64.push_back(nop); - - //////////////// CPU OPCODES END //////////////// - - for (size_t i = 1; i < argc; ++i) { - if (argv[i][0] == '-') { - if (strcmp(argv[i], "--amd64:ver") == 0 || strcmp(argv[i], "--amd64:v") == 0) { - kStdOut << "AssemblerAMD64: AMD64 Assembler Driver.\nAssemblerAMD64: " - "v1.10\nAssemblerAMD64: Copyright " - "(c) Amlal El Mahrouss\n"; - return 0; - } else if (strcmp(argv[i], "--amd64:h") == 0) { - kStdOut << "AssemblerAMD64: AMD64 Assembler Driver.\nAssemblerAMD64: Copyright (c) 2024 " - "Amlal El Mahrouss\n"; - kStdOut << "--version: Print program version.\n"; - kStdOut << "--verbose: Print verbose output.\n"; - kStdOut << "--binary: Output as flat binary.\n"; - - return 0; - } else if (strcmp(argv[i], "--amd64:binary") == 0) { - kOutputAsBinary = true; - continue; - } else if (strcmp(argv[i], "--amd64:verbose") == 0) { - kVerbose = true; - continue; - } - - kStdOut << "AssemblerAMD64: ignore " << argv[i] << "\n"; - continue; - } - - if (!std::filesystem::exists(argv[i])) { - kStdOut << "AssemblerAMD64: can't open: " << argv[i] << std::endl; - goto asm_fail_exit; - } - - std::string object_output(argv[i]); - std::string asm_input(argv[i]); - - for (auto& ext : kAsmFileExts) { - if (object_output.ends_with(ext)) { - object_output.erase(object_output.find(ext), std::strlen(ext)); - break; - } - } - - object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; - - std::ifstream file_ptr(argv[i]); - std::ofstream file_ptr_out(object_output, std::ofstream::binary); - - kStdOut << "AssemblerAMD64: Assembling: " << argv[i] << "\n"; - - if (file_ptr_out.bad()) { - if (kVerbose) { - kStdOut << "AssemblerAMD64: error: " << strerror(errno) << "\n"; - } - - return 1; - } - - std::string line; - - LibCompiler::AEHeader hdr{0}; - - memset(hdr.fPad, kAENullType, kAEPad); - - hdr.fMagic[0] = kAEMag0; - hdr.fMagic[1] = kAEMag1; - hdr.fSize = sizeof(LibCompiler::AEHeader); - hdr.fArch = kOutputArch; - - ///////////////////////////////////////////////////////////////////////////////////////// - - // COMPILATION LOOP - - ///////////////////////////////////////////////////////////////////////////////////////// - - LibCompiler::EncoderAMD64 asm64; - - if (kVerbose) { - kStdOut << "Compiling: " + asm_input << "\n"; - kStdOut << "From: " + line << "\n"; - } - - while (std::getline(file_ptr, line)) { - if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { - Detail::print_error(ln, argv[i]); - continue; - } - - try { - asm_read_attributes(line); - asm64.WriteLine(line, argv[i]); - } catch (const std::exception& e) { - if (kVerbose) { - std::string what = e.what(); - Detail::print_warning("exit because of: " + what, "LibCompiler"); - } - - try { - std::filesystem::remove(object_output); - } catch (...) { - } - - goto asm_fail_exit; - } - } - - if (!kOutputAsBinary) { - if (kVerbose) { - kStdOut << "AssemblerAMD64: Writing object file...\n"; - } - - // this is the final step, write everything to the file. - - auto pos = file_ptr_out.tellp(); - - hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); - - file_ptr_out << hdr; - - if (kRecords.empty()) { - kStdErr << "AssemblerAMD64: At least one record is needed to write an object " - "file.\nAssemblerAMD64: Make one using `public_segment .code64 foo_bar`.\n"; - - std::filesystem::remove(object_output); - return 1; - } - - kRecords[kRecords.size() - 1].fSize = kAppBytes.size(); - - std::size_t record_count = 0UL; - - for (auto& rec : kRecords) { - if (kVerbose) kStdOut << "AssemblerAMD64: Wrote record " << rec.fName << " to file...\n"; - - rec.fFlags |= LibCompiler::kKindRelocationAtRuntime; - rec.fOffset = record_count; - ++record_count; - - file_ptr_out << rec; - } - - // increment once again, so that we won't lie about the kUndefinedSymbols. - ++record_count; - - for (auto& sym : kUndefinedSymbols) { - LibCompiler::AERecordHeader _record_hdr{0}; - - if (kVerbose) kStdOut << "AssemblerAMD64: Wrote symbol " << sym << " to file...\n"; - - _record_hdr.fKind = kAENullType; - _record_hdr.fSize = sym.size(); - _record_hdr.fOffset = record_count; - - ++record_count; - - memset(_record_hdr.fPad, kAENullType, kAEPad); - memcpy(_record_hdr.fName, sym.c_str(), sym.size()); - - file_ptr_out << _record_hdr; - - ++kCounter; - } - - auto pos_end = file_ptr_out.tellp(); - - file_ptr_out.seekp(pos); - - hdr.fStartCode = pos_end; - hdr.fCodeSize = kAppBytes.size(); - - file_ptr_out << hdr; - - file_ptr_out.seekp(pos_end); - } else { - if (kVerbose) { - kStdOut << "AssemblerAMD64: Write raw binary...\n"; - } - } - - // byte from byte, we write this. - for (auto& byte : kAppBytes) { - if (byte == 0) continue; - - if (byte == 0xFF) { - byte = 0; - } - - file_ptr_out << reinterpret_cast(&byte)[0]; - } - - if (kVerbose) kStdOut << "AssemblerAMD64: Wrote file with program in it.\n"; - - file_ptr_out.flush(); - file_ptr_out.close(); - - if (kVerbose) kStdOut << "AssemblerAMD64: Exit succeeded.\n"; - - return 0; - } - -asm_fail_exit: - - if (kVerbose) kStdOut << "AssemblerAMD64: Exit failed.\n"; - - return 1; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for attributes -// returns true if any was found. - -///////////////////////////////////////////////////////////////////////////////////////// - -static bool asm_read_attributes(std::string line) { - // extern_segment is the opposite of public_segment, it signals to the ld - // that we need this symbol. - if (LibCompiler::find_word(line, "extern_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_extern_segment_bin"); - } - - auto name = line.substr(line.find("extern_segment") + strlen("extern_segment") + 1); - - if (name.size() == 0) { - Detail::print_error("Invalid extern_segment", "power-as"); - throw std::runtime_error("invalid_extern_segment"); - } - - std::string result = std::to_string(name.size()); - result += kUndefinedSymbol; - - // mangle this - for (char& j : name) { - if (j == ' ' || j == ',') j = '$'; - } - - result += name; - - if (name.find(kPefCode64) != std::string::npos) { - // data is treated as code. - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(kPefData64) != std::string::npos) { - // no code will be executed from here. - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(kPefZero64) != std::string::npos) { - // this is a bss section. - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that ld can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kAppBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, result.c_str(), result.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - // public_segment is a special keyword used by AssemblerAMD64 to tell the AE output stage to - // mark this section as a header. it currently supports .code64, .data64 and - // .zero64. - else if (LibCompiler::find_word(line, "public_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_public_segment_bin"); - } - - auto name = line.substr(line.find("public_segment") + strlen("public_segment") + 1); - - std::string name_copy = name; - - for (char& j : name) { - if (j == ' ') j = '$'; - } - - if (std::find(kDefinedSymbols.begin(), kDefinedSymbols.end(), name) != kDefinedSymbols.end()) { - Detail::print_error("Symbol already defined.", "LibCompiler"); - throw std::runtime_error("invalid_public_segment_bin"); - } - - kDefinedSymbols.push_back(name); - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - - name_copy.erase(name_copy.find(".code64"), strlen(".code64")); - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - - name_copy.erase(name_copy.find(".data64"), strlen(".data64")); - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - - name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that ld can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); - - kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); - ++kOrigin; - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kAppBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, name.c_str(), name.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - - return false; -} - -// \brief algorithms and helpers. - -namespace Detail::algorithm { -// \brief authorize a brief set of characters. -static inline bool is_not_valid(char c) { - if ((isalpha(c) || isdigit(c)) || - ((c == ' ') || (c == '\t') || (c == ',') || (c == '(') || (c == ')') || (c == '"') || - (c == '*') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || (c == '_') || - (c == ':') || (c == '@') || (c == '.') || (c == '#') || (c == ';'))) - return false; - - return true; -} - -bool is_valid_amd64(std::string str) { - return std::find_if(str.begin(), str.end(), is_not_valid) == str.end(); -} -} // namespace Detail::algorithm - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for line (syntax check) - -///////////////////////////////////////////////////////////////////////////////////////// - -std::string LibCompiler::EncoderAMD64::CheckLine(std::string line, std::string file) { - std::string err_str; - - if (line.empty() || LibCompiler::find_word(line, "extern_segment") || - LibCompiler::find_word(line, "public_segment") || - LibCompiler::find_word(line, kAssemblerPragmaSymStr) || LibCompiler::find_word(line, ";") || - line[0] == kAssemblerPragmaSym) { - if (line.find(';') != std::string::npos) { - line.erase(line.find(';')); - } else { - // now check the line for validity - if (!Detail::algorithm::is_valid_amd64(line)) { - err_str = "Line contains non valid characters.\nhere -> "; - err_str += line; - } - } - - return err_str; - } - - // check for a valid instruction format. - - if (line.find(',') != std::string::npos) { - if (line.find(',') + 1 == line.size()) { - err_str += "\nInstruction lacks right register, here -> "; - err_str += line.substr(line.find(',')); - - return err_str; - } else { - bool nothing_on_right = true; - - if (line.find(',') + 1 > line.size()) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - - auto substr = line.substr(line.find(',') + 1); - - for (auto& ch : substr) { - if (ch != ' ' && ch != '\t') { - nothing_on_right = false; - } - } - - // this means we found nothing after that ',' . - if (nothing_on_right) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - } - } - for (auto& opcodeAMD64 : kOpcodesAMD64) { - if (LibCompiler::find_word(line, opcodeAMD64.fName)) { - return err_str; - } - } - - err_str += "\nUnrecognized instruction -> " + line; - - return err_str; -} - -bool LibCompiler::EncoderAMD64::WriteNumber(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - LibCompiler::NumberCast64 num = - LibCompiler::NumberCast64(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) - << "\n"; - } - - return true; - } - case 'b': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - LibCompiler::NumberCast64 num = - LibCompiler::NumberCast64(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - return true; - } - case 'o': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - LibCompiler::NumberCast64 num = - LibCompiler::NumberCast64(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - return true; - } - default: { - break; - } - } - - /* check for errno and stuff like that */ - if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - return false; - } - } - - LibCompiler::NumberCast64 num = - LibCompiler::NumberCast64(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -bool LibCompiler::EncoderAMD64::WriteNumber32(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); - res += kOrigin; - - if (errno != 0) { - return false; - } - - LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) - << "\n"; - } - - return true; - } - case 'b': { - auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); - res += kOrigin; - - if (errno != 0) { - return false; - } - - LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - return true; - } - case 'o': { - auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); - res += kOrigin; - - if (errno != 0) { - return false; - } - - LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - return true; - } - default: { - break; - } - } - - auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 10); - res += kOrigin; - - if (errno != 0) { - return false; - } - - LibCompiler::NumberCast32 num = LibCompiler::NumberCast32(res); - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -bool LibCompiler::EncoderAMD64::WriteNumber16(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - LibCompiler::NumberCast16 num = - LibCompiler::NumberCast16(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) - << "\n"; - } - - return true; - } - case 'b': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - LibCompiler::NumberCast16 num = - LibCompiler::NumberCast16(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - return true; - } - case 'o': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - LibCompiler::NumberCast16 num = - LibCompiler::NumberCast16(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - return true; - } - default: { - break; - } - } - - /* check for errno and stuff like that */ - if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - return false; - } - } - - LibCompiler::NumberCast16 num = - LibCompiler::NumberCast16(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); - - for (char& i : num.number) { - if (i == 0) i = 0xFF; - - kAppBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -bool LibCompiler::EncoderAMD64::WriteNumber8(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - LibCompiler::NumberCast8 num = - LibCompiler::NumberCast8(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); - - kAppBytes.push_back(num.number); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 16 number here: " << jump_label.substr(pos) - << "\n"; - } - - return true; - } - case 'b': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - LibCompiler::NumberCast8 num = - LibCompiler::NumberCast8(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - kAppBytes.push_back(num.number); - - return true; - } - case 'o': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - LibCompiler::NumberCast8 num = - LibCompiler::NumberCast8(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - kAppBytes.push_back(num.number); - - return true; - } - default: { - break; - } - } - - /* check for errno and stuff like that */ - if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - return false; - } - } - - LibCompiler::NumberCast8 num = - LibCompiler::NumberCast8(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); - - kAppBytes.push_back(num.number); - - if (kVerbose) { - kStdOut << "AssemblerAMD64: Found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Read and write an instruction to the output array. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool LibCompiler::EncoderAMD64::WriteLine(std::string line, std::string file) { - if (LibCompiler::find_word(line, "public_segment ")) return true; - - struct RegMapAMD64 { - std::string fName; - i64_byte_t fModRM; - }; - - std::vector kRegisterList{ - {.fName = "ax", .fModRM = 0x0}, {.fName = "cx", .fModRM = 1}, - {.fName = "dx", .fModRM = 0x2}, {.fName = "bx", .fModRM = 3}, - {.fName = "sp", .fModRM = 0x4}, {.fName = "bp", .fModRM = 5}, - {.fName = "si", .fModRM = 0x6}, {.fName = "di", .fModRM = 7}, - }; - - BOOL foundInstruction = false; - - for (auto& opcodeAMD64 : kOpcodesAMD64) { - // strict check here - if (LibCompiler::find_word(line, opcodeAMD64.fName) && - Detail::algorithm::is_valid_amd64(line)) { - foundInstruction = true; - std::string name(opcodeAMD64.fName); - - /// Move instruction handler. - if (line.find(name) != std::string::npos) { - if (name == "mov" || name == "xor") { - std::string substr = line.substr(line.find(name) + name.size()); - - uint64_t bits = kRegisterBitWidth; - - if (substr.find(",") == std::string::npos) { - Detail::print_error("Syntax error: missing right operand.", "LibCompiler"); - throw std::runtime_error("syntax_err"); - } - - bool onlyOneReg = true; - - std::vector currentRegList; - - for (auto& reg : kRegisterList) { - std::vector regExt = {'e', 'r'}; - - for (auto& ext : regExt) { - std::string registerName; - - if (bits > 16) registerName.push_back(ext); - - registerName += reg.fName; - - while (line.find(registerName) != std::string::npos) { - line.erase(line.find(registerName), registerName.size()); - - if (bits == 16) { - if (registerName[0] == 'r') { - Detail::print_error("invalid size for register, current bit width is: " + - std::to_string(kRegisterBitWidth), - file); - throw std::runtime_error("invalid_reg_size"); - } - } - - currentRegList.push_back({.fName = registerName, .fModRM = reg.fModRM}); - } - } - } - - if (currentRegList.size() > 1) onlyOneReg = false; - - bool hasRBasedRegs = false; - - if (!onlyOneReg) { - /// very tricky to understand. - /// but this checks for a r8 through r15 register. - if (currentRegList[0].fName[0] == 'r' || currentRegList[1].fName[0] == 'r') { - if (isdigit(currentRegList[0].fName[1]) && isdigit(currentRegList[1].fName[1])) { - kAppBytes.emplace_back(0x4d); - hasRBasedRegs = true; - } else if (isdigit(currentRegList[0].fName[1]) || - isdigit(currentRegList[1].fName[1])) { - kAppBytes.emplace_back(0x4c); - hasRBasedRegs = true; - } - } - } - - if (name == "mov") { - if (bits == 64 || bits == 32) { - if (!hasRBasedRegs && bits >= 32) { - kAppBytes.emplace_back(opcodeAMD64.fOpcode); - } - - if (!onlyOneReg) kAppBytes.emplace_back(0x89); - } else if (bits == 16) { - if (hasRBasedRegs) { - Detail::print_error("Invalid combination of operands and registers.", - "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } else { - kAppBytes.emplace_back(0x66); - kAppBytes.emplace_back(0x89); - } - } - } else { - if (!hasRBasedRegs && bits >= 32) { - kAppBytes.emplace_back(opcodeAMD64.fOpcode); - } - - kAppBytes.emplace_back(0x31); - } - - if (onlyOneReg) { - auto num = GetNumber32(line, ","); - - for (auto& num_idx : num.number) { - if (num_idx == 0) num_idx = 0xFF; - } - - auto modrm = (0x3 << 6 | currentRegList[0].fModRM); - - kAppBytes.emplace_back(0xC7); // prefixed before placing the modrm and then the number. - kAppBytes.emplace_back(modrm); - - if (name != "xor") { - kAppBytes.emplace_back(num.number[0]); - kAppBytes.emplace_back(num.number[1]); - kAppBytes.emplace_back(num.number[2]); - kAppBytes.emplace_back(num.number[3]); - } - - break; - } - - if (currentRegList[1].fName[0] == 'r' && currentRegList[0].fName[0] == 'e') { - Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } - - if (currentRegList[0].fName[0] == 'r' && currentRegList[1].fName[0] == 'e') { - Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } - - if (bits == 16) { - if (currentRegList[0].fName[0] == 'r' || currentRegList[0].fName[0] == 'e') { - Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } - - if (currentRegList[1].fName[0] == 'r' || currentRegList[1].fName[0] == 'e') { - Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } - } else { - if (currentRegList[0].fName[0] != 'r' || currentRegList[0].fName[0] == 'e') { - Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } - - if (currentRegList[1].fName[0] != 'r' || currentRegList[1].fName[0] == 'e') { - Detail::print_error("Invalid combination of operands and registers.", "LibCompiler"); - throw std::runtime_error("comb_op_reg"); - } - } - - /// encode register using the modrm encoding. - - auto modrm = (0x3 << 6 | currentRegList[1].fModRM << 3 | currentRegList[0].fModRM); - - kAppBytes.emplace_back(modrm); - - break; - } - } - - if (name == "int" || name == "into" || name == "intd") { - kAppBytes.emplace_back(opcodeAMD64.fOpcode); - this->WriteNumber8(line.find(name) + name.size() + 1, line); - - break; - } else if (name == "jmp" || name == "call") { - kAppBytes.emplace_back(opcodeAMD64.fOpcode); - - if (!this->WriteNumber32(line.find(name) + name.size() + 1, line)) { - throw std::runtime_error("BUG: WriteNumber32"); - } - - break; - } else if (name == "syscall") { - kAppBytes.emplace_back(opcodeAMD64.fOpcode); - kAppBytes.emplace_back(0x05); - - break; - } else { - kAppBytes.emplace_back(opcodeAMD64.fOpcode); - - break; - } - } - } - - if (line[0] == kAssemblerPragmaSym) { - if (foundInstruction) { - Detail::print_error("Syntax error: " + line, file); - throw std::runtime_error("syntax_err"); - } - - if (line.find("bits 64") != std::string::npos) { - kRegisterBitWidth = 64U; - } else if (line.find("bits 32") != std::string::npos) { - kRegisterBitWidth = 32U; - } else if (line.find("bits 16") != std::string::npos) { - kRegisterBitWidth = 16U; - } else if (line.find("org") != std::string::npos) { - size_t base[] = {10, 16, 2, 7}; - - for (size_t i = 0; i < 4; i++) { - if (kOrigin = strtol((line.substr(line.find("org") + strlen("org") + 1)).c_str(), nullptr, - base[i]); - kOrigin) { - if (errno != 0) { - continue; - } else { - if (kVerbose) { - kStdOut << "AssemblerAMD64: Origin Set: " << kOrigin << std::endl; - } - - break; - } - } - } - } - } - /// write a dword - else if (line.find(".dword") != std::string::npos) { - this->WriteNumber32(line.find(".dword") + strlen(".dword") + 1, line); - } - /// write a long - else if (line.find(".long") != std::string::npos) { - this->WriteNumber(line.find(".long") + strlen(".long") + 1, line); - } - /// write a 16-bit number - else if (line.find(".word") != std::string::npos) { - this->WriteNumber16(line.find(".word") + strlen(".word") + 1, line); - } - - kOrigin += kIPAlignement; - - return true; -} - -// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/AssemblerARM64.cc b/dev/LibCompiler/src/AssemblerARM64.cc deleted file mode 100644 index bad974e..0000000 --- a/dev/LibCompiler/src/AssemblerARM64.cc +++ /dev/null @@ -1,585 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @file AssemblerARM64.cxx -/// @author EL Mahrouss Amlal -/// @brief 'ACORN' Assembler. - -/// REMINDER: when dealing with an undefined symbol use (string -/// size):LinkerFindSymbol:(string) so that li will look for it. - -///////////////////////////////////////////////////////////////////////////////////////// - -#define __ASM_NEED_ARM64__ 1 - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" -#define kYellow "\e[0;33m" - -constexpr auto cPowerIPAlignment = 0x1U; - -static CharType kOutputArch = LibCompiler::kPefArchARM64; - -static std::size_t kCounter = 1UL; - -static std::uintptr_t kOrigin = kPefBaseOrigin; -static std::vector> kOriginLabel; - -static std::vector kBytes; - -static LibCompiler::AERecordHeader kCurrentRecord{ - .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; - -static std::vector kRecords; -static std::vector kUndefinedSymbols; - -static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; -static const std::string kRelocSymbol = ":RuntimeSymbol:"; - -// \brief forward decl. -static bool asm_read_attributes(std::string line); - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @brief POWER assembler entrypoint, the program/module starts here. - -///////////////////////////////////////////////////////////////////////////////////////// - -LIBCOMPILER_MODULE(AssemblerMainARM64) { - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - for (size_t i = 1; i < argc; ++i) { - if (argv[i][0] == '-') { - if (strcmp(argv[i], "--ver") == 0 || strcmp(argv[i], "--v") == 0) { - kStdOut << "AssemblerPower: AARCH64 Assembler Driver.\nAssemblerPower: " << kDistVersion - << "\nAssemblerPower: " - "Copyright (c) " - "Amlal El Mahrouss\n"; - return 0; - } else if (strcmp(argv[i], "--h") == 0) { - kStdOut << "AssemblerPower: AARCH64 Assembler Driver.\nAssemblerPower: Copyright (c) 2024 " - "Amlal El Mahrouss\n"; - kStdOut << "--version,/v: print program version.\n"; - kStdOut << "--verbose: print verbose output.\n"; - kStdOut << "--binary: output as flat binary.\n"; - - return 0; - } else if (strcmp(argv[i], "--binary") == 0) { - kOutputAsBinary = true; - continue; - } else if (strcmp(argv[i], "--verbose") == 0) { - kVerbose = true; - continue; - } - - kStdOut << "AssemblerPower: ignore " << argv[i] << "\n"; - continue; - } - - if (!std::filesystem::exists(argv[i])) { - kStdOut << "AssemblerPower: can't open: " << argv[i] << std::endl; - goto asm_fail_exit; - } - - std::string object_output(argv[i]); - - for (auto& ext : kAsmFileExts) { - if (object_output.find(ext) != std::string::npos) { - object_output.erase(object_output.find(ext), std::strlen(ext)); - } - } - - object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; - - std::ifstream file_ptr(argv[i]); - std::ofstream file_ptr_out(object_output, std::ofstream::binary); - - if (file_ptr_out.bad()) { - if (kVerbose) { - kStdOut << "AssemblerPower: error: " << strerror(errno) << "\n"; - } - } - - std::string line; - - LibCompiler::AEHeader hdr{0}; - - memset(hdr.fPad, kAENullType, kAEPad); - - hdr.fMagic[0] = kAEMag0; - hdr.fMagic[1] = kAEMag1; - hdr.fSize = sizeof(LibCompiler::AEHeader); - hdr.fArch = kOutputArch; - - ///////////////////////////////////////////////////////////////////////////////////////// - - // COMPILATION LOOP - - ///////////////////////////////////////////////////////////////////////////////////////// - - LibCompiler::EncoderARM64 asm64; - - while (std::getline(file_ptr, line)) { - if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { - Detail::print_error(ln, argv[i]); - continue; - } - - try { - asm_read_attributes(line); - asm64.WriteLine(line, argv[i]); - } catch (const std::exception& e) { - if (kVerbose) { - std::string what = e.what(); - Detail::print_warning("exit because of: " + what, "LibCompiler"); - } - - std::filesystem::remove(object_output); - goto asm_fail_exit; - } - } - - if (!kOutputAsBinary) { - if (kVerbose) { - kStdOut << "AssemblerARM64: Writing object file...\n"; - } - - // this is the final step, write everything to the file. - - auto pos = file_ptr_out.tellp(); - - hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); - - file_ptr_out << hdr; - - if (kRecords.empty()) { - kStdErr << "AssemblerARM64: At least one record is needed to write an object " - "file.\nAssemblerARM64: Make one using `public_segment .code64 foo_bar`.\n"; - - std::filesystem::remove(object_output); - return 1; - } - - kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - std::size_t record_count = 0UL; - - for (auto& record_hdr : kRecords) { - record_hdr.fFlags |= LibCompiler::kKindRelocationAtRuntime; - record_hdr.fOffset = record_count; - ++record_count; - - file_ptr_out << record_hdr; - - if (kVerbose) kStdOut << "AssemblerARM64: Wrote record " << record_hdr.fName << "...\n"; - } - - // increment once again, so that we won't lie about the kUndefinedSymbols. - ++record_count; - - for (auto& sym : kUndefinedSymbols) { - LibCompiler::AERecordHeader undefined_sym{0}; - - if (kVerbose) kStdOut << "AssemblerARM64: Wrote symbol " << sym << " to file...\n"; - - undefined_sym.fKind = kAENullType; - undefined_sym.fSize = sym.size(); - undefined_sym.fOffset = record_count; - - ++record_count; - - memset(undefined_sym.fPad, kAENullType, kAEPad); - memcpy(undefined_sym.fName, sym.c_str(), sym.size()); - - file_ptr_out << undefined_sym; - - ++kCounter; - } - - auto pos_end = file_ptr_out.tellp(); - - file_ptr_out.seekp(pos); - - hdr.fStartCode = pos_end; - hdr.fCodeSize = kBytes.size(); - - file_ptr_out << hdr; - - file_ptr_out.seekp(pos_end); - } else { - if (kVerbose) { - kStdOut << "AssemblerARM64: Write raw binary...\n"; - } - } - - // byte from byte, we write this. - for (auto& byte : kBytes) { - file_ptr_out.write(reinterpret_cast(&byte), sizeof(byte)); - } - - if (kVerbose) kStdOut << "AssemblerARM64: Wrote file with program in it.\n"; - - file_ptr_out.flush(); - file_ptr_out.close(); - - if (kVerbose) kStdOut << "AssemblerARM64: Exit succeeded.\n"; - - return 0; - } - -asm_fail_exit: - - if (kVerbose) kStdOut << "AssemblerARM64: Exit failed.\n"; - - return LIBCOMPILER_EXEC_ERROR; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for attributes -// returns true if any was found. - -///////////////////////////////////////////////////////////////////////////////////////// - -static bool asm_read_attributes(std::string line) { - // extern_segment is the opposite of public_segment, it signals to the li - // that we need this symbol. - if (LibCompiler::find_word(line, "extern_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid extern_segment directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_extern_segment_bin"); - } - - auto name = line.substr(line.find("extern_segment") + strlen("extern_segment") + 1); - - if (name.size() == 0) { - Detail::print_error("Invalid extern_segment", "LibCompiler"); - throw std::runtime_error("invalid_extern_segment"); - } - - std::string result = std::to_string(name.size()); - result += kUndefinedSymbol; - - // mangle this - for (char& j : name) { - if (j == ' ' || j == ',') j = '$'; - } - - result += name; - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that li can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, result.c_str(), result.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - // public_segment is a special keyword used by Assembler to tell the AE output stage to - // mark this section as a header. it currently supports .code64, .data64., - // .zero64 - else if (LibCompiler::find_word(line, "public_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid public_segment directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_public_segment_bin"); - } - - auto name = line.substr(line.find("public_segment") + strlen("public_segment")); - - std::string name_copy = name; - - for (char& j : name) { - if (j == ' ') j = '$'; - } - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - - name_copy.erase(name_copy.find(".code64"), strlen(".code64")); - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - - name_copy.erase(name_copy.find(".data64"), strlen(".data64")); - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - - name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that li can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); - - kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); - ++kOrigin; - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, name.c_str(), name.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - - return false; -} - -// \brief algorithms and helpers. - -namespace Detail::algorithm { -// \brief authorize a brief set of characters. -static inline bool is_not_alnum_space(char c) { - return !(isalpha(c) || isdigit(c) || (c == ' ') || (c == '\t') || (c == ',') || (c == '(') || - (c == ')') || (c == '"') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || - (c == '_') || (c == ':') || (c == '@') || (c == '.')); -} - -bool is_valid_arm64(std::string str) { - return std::find_if(str.begin(), str.end(), is_not_alnum_space) == str.end(); -} -} // namespace Detail::algorithm - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for line (syntax check) - -///////////////////////////////////////////////////////////////////////////////////////// - -std::string LibCompiler::EncoderARM64::CheckLine(std::string line, std::string file) { - std::string err_str; - - if (line.empty() || LibCompiler::find_word(line, "extern_segment") || - LibCompiler::find_word(line, "public_segment") || line.find('#') != std::string::npos || - LibCompiler::find_word(line, ";")) { - if (line.find('#') != std::string::npos) { - line.erase(line.find('#')); - } else if (line.find(';') != std::string::npos) { - line.erase(line.find(';')); - } else { - /// does the line contains valid input? - if (!Detail::algorithm::is_valid_arm64(line)) { - err_str = "Line contains non alphanumeric characters.\nhere -> "; - err_str += line; - } - } - - return err_str; - } - - if (!Detail::algorithm::is_valid_arm64(line)) { - err_str = "Line contains non alphanumeric characters.\nhere -> "; - err_str += line; - - return err_str; - } - - // check for a valid instruction format. - - if (line.find(',') != std::string::npos) { - if (line.find(',') + 1 == line.size()) { - err_str += "\nInstruction lacks right register, here -> "; - err_str += line.substr(line.find(',')); - - return err_str; - } else { - bool nothing_on_right = true; - - if (line.find(',') + 1 > line.size()) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - - auto substr = line.substr(line.find(',') + 1); - - for (auto& ch : substr) { - if (ch != ' ' && ch != '\t') { - nothing_on_right = false; - } - } - - // this means we found nothing after that ',' . - if (nothing_on_right) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - } - } - - return err_str; -} - -bool LibCompiler::EncoderARM64::WriteNumber(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); - - for (char& i : num.number) { - kBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerARM64: found a base 16 number here: " << jump_label.substr(pos) - << "\n"; - } - - return true; - } - case 'b': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "AssemblerARM64: found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - kBytes.push_back(i); - } - - return true; - } - case 'o': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "AssemblerARM64: found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - kBytes.push_back(i); - } - - return true; - } - default: { - break; - } - } - - /* check for errno and stuff like that */ - if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - return false; - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); - - for (char& i : num.number) { - kBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerARM64: found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @brief Read and write an instruction to the output array. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool LibCompiler::EncoderARM64::WriteLine(std::string line, std::string file) { - if (LibCompiler::find_word(line, "public_segment")) return false; - - if (!Detail::algorithm::is_valid_arm64(line)) return false; - - return true; -} - -// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/AssemblerPowerPC.cc b/dev/LibCompiler/src/AssemblerPowerPC.cc deleted file mode 100644 index c0a7172..0000000 --- a/dev/LibCompiler/src/AssemblerPowerPC.cc +++ /dev/null @@ -1,909 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @file AssemblerPower.cxx -/// @author EL Mahrouss Amlal -/// @brief POWER Assembler. - -/// REMINDER: when dealing with an undefined symbol use (string -/// size):LinkerFindSymbol:(string) so that li will look for it. - -///////////////////////////////////////////////////////////////////////////////////////// - -#define __ASM_NEED_PPC__ 1 - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" -#define kYellow "\e[0;33m" - -constexpr auto cPowerIPAlignment = 0x4U; - -static CharType kOutputArch = LibCompiler::kPefArchPowerPC; - -static std::size_t kCounter = 1UL; - -static std::uintptr_t kOrigin = kPefBaseOrigin; -static std::vector> kOriginLabel; - -static std::vector kBytes; - -static LibCompiler::AERecordHeader kCurrentRecord{ - .fName = "", .fKind = LibCompiler::kPefCode, .fSize = 0, .fOffset = 0}; - -static std::vector kRecords; -static std::vector kUndefinedSymbols; - -static const std::string kUndefinedSymbol = ":UndefinedSymbol:"; -static const std::string kRelocSymbol = ":RuntimeSymbol:"; - -// \brief forward decl. -static bool asm_read_attributes(std::string line); - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @brief POWER assembler entrypoint, the program/module starts here. - -///////////////////////////////////////////////////////////////////////////////////////// - -LIBCOMPILER_MODULE(AssemblerMainPower64) { - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - for (size_t i = 1; i < argc; ++i) { - if (argv[i][0] == '-') { - if (strcmp(argv[i], "--ver") == 0 || strcmp(argv[i], "--v") == 0) { - kStdOut << "AssemblerPower: POWER64 Assembler Driver.\nAssemblerPower: " << kDistVersion - << "\nAssemblerPower: " - "Copyright (c) " - "Amlal El Mahrouss\n"; - return 0; - } else if (strcmp(argv[i], "--h") == 0) { - kStdOut << "AssemblerPower: POWER64 Assembler Driver.\nAssemblerPower: Copyright (c) 2024 " - "Amlal El Mahrouss\n"; - kStdOut << "--version,/v: print program version.\n"; - kStdOut << "--verbose: print verbose output.\n"; - kStdOut << "--binary: output as flat binary.\n"; - - return 0; - } else if (strcmp(argv[i], "--binary") == 0) { - kOutputAsBinary = true; - continue; - } else if (strcmp(argv[i], "--verbose") == 0) { - kVerbose = true; - continue; - } - - kStdOut << "AssemblerPower: ignore " << argv[i] << "\n"; - continue; - } - - if (!std::filesystem::exists(argv[i])) { - kStdOut << "AssemblerPower: can't open: " << argv[i] << std::endl; - goto asm_fail_exit; - } - - std::string object_output(argv[i]); - - for (auto& ext : kAsmFileExts) { - if (object_output.find(ext) != std::string::npos) { - object_output.erase(object_output.find(ext), std::strlen(ext)); - } - } - - object_output += kOutputAsBinary ? kBinaryFileExt : kObjectFileExt; - - std::ifstream file_ptr(argv[i]); - std::ofstream file_ptr_out(object_output, std::ofstream::binary); - - if (file_ptr_out.bad()) { - if (kVerbose) { - kStdOut << "AssemblerPower: error: " << strerror(errno) << "\n"; - } - } - - std::string line; - - LibCompiler::AEHeader hdr{0}; - - memset(hdr.fPad, kAENullType, kAEPad); - - hdr.fMagic[0] = kAEMag0; - hdr.fMagic[1] = kAEMag1; - hdr.fSize = sizeof(LibCompiler::AEHeader); - hdr.fArch = kOutputArch; - - ///////////////////////////////////////////////////////////////////////////////////////// - - // COMPILATION LOOP - - ///////////////////////////////////////////////////////////////////////////////////////// - - LibCompiler::EncoderPowerPC asm64; - - while (std::getline(file_ptr, line)) { - if (auto ln = asm64.CheckLine(line, argv[i]); !ln.empty()) { - Detail::print_error(ln, argv[i]); - continue; - } - - try { - asm_read_attributes(line); - asm64.WriteLine(line, argv[i]); - } catch (const std::exception& e) { - if (kVerbose) { - std::string what = e.what(); - Detail::print_warning("exit because of: " + what, "LibCompiler"); - } - - std::filesystem::remove(object_output); - goto asm_fail_exit; - } - } - - if (!kOutputAsBinary) { - if (kVerbose) { - kStdOut << "AssemblerPower: Writing object file...\n"; - } - - // this is the final step, write everything to the file. - - auto pos = file_ptr_out.tellp(); - - hdr.fCount = kRecords.size() + kUndefinedSymbols.size(); - - file_ptr_out << hdr; - - if (kRecords.empty()) { - kStdErr << "AssemblerPower: At least one record is needed to write an object " - "file.\nAssemblerPower: Make one using `public_segment .code64 foo_bar`.\n"; - - std::filesystem::remove(object_output); - return 1; - } - - kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - std::size_t record_count = 0UL; - - for (auto& record_hdr : kRecords) { - record_hdr.fFlags |= LibCompiler::kKindRelocationAtRuntime; - record_hdr.fOffset = record_count; - ++record_count; - - file_ptr_out << record_hdr; - - if (kVerbose) kStdOut << "AssemblerPower: Wrote record " << record_hdr.fName << "...\n"; - } - - // increment once again, so that we won't lie about the kUndefinedSymbols. - ++record_count; - - for (auto& sym : kUndefinedSymbols) { - LibCompiler::AERecordHeader undefined_sym{0}; - - if (kVerbose) kStdOut << "AssemblerPower: Wrote symbol " << sym << " to file...\n"; - - undefined_sym.fKind = kAENullType; - undefined_sym.fSize = sym.size(); - undefined_sym.fOffset = record_count; - - ++record_count; - - memset(undefined_sym.fPad, kAENullType, kAEPad); - memcpy(undefined_sym.fName, sym.c_str(), sym.size()); - - file_ptr_out << undefined_sym; - - ++kCounter; - } - - auto pos_end = file_ptr_out.tellp(); - - file_ptr_out.seekp(pos); - - hdr.fStartCode = pos_end; - hdr.fCodeSize = kBytes.size(); - - file_ptr_out << hdr; - - file_ptr_out.seekp(pos_end); - } else { - if (kVerbose) { - kStdOut << "AssemblerPower: Write raw binary...\n"; - } - } - - // byte from byte, we write this. - for (auto& byte : kBytes) { - file_ptr_out.write(reinterpret_cast(&byte), sizeof(byte)); - } - - if (kVerbose) kStdOut << "AssemblerPower: Wrote file with program in it.\n"; - - file_ptr_out.flush(); - file_ptr_out.close(); - - if (kVerbose) kStdOut << "AssemblerPower: Exit succeeded.\n"; - - return 0; - } - -asm_fail_exit: - - if (kVerbose) kStdOut << "AssemblerPower: Exit failed.\n"; - - return LIBCOMPILER_EXEC_ERROR; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for attributes -// returns true if any was found. - -///////////////////////////////////////////////////////////////////////////////////////// - -static bool asm_read_attributes(std::string line) { - // extern_segment is the opposite of public_segment, it signals to the li - // that we need this symbol. - if (LibCompiler::find_word(line, "extern_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid extern_segment directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_extern_segment_bin"); - } - - auto name = line.substr(line.find("extern_segment") + strlen("extern_segment") + 1); - - if (name.size() == 0) { - Detail::print_error("Invalid extern_segment", "LibCompiler"); - throw std::runtime_error("invalid_extern_segment"); - } - - std::string result = std::to_string(name.size()); - result += kUndefinedSymbol; - - // mangle this - for (char& j : name) { - if (j == ' ' || j == ',') j = '$'; - } - - result += name; - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that li can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, result.c_str(), result.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - // public_segment is a special keyword used by AssemblerPower to tell the AE output stage to - // mark this section as a header. it currently supports .code64, .data64., - // .zero64 - else if (LibCompiler::find_word(line, "public_segment")) { - if (kOutputAsBinary) { - Detail::print_error("Invalid public_segment directive in flat binary mode.", "LibCompiler"); - throw std::runtime_error("invalid_public_segment_bin"); - } - - auto name = line.substr(line.find("public_segment") + strlen("public_segment")); - - std::string name_copy = name; - - for (char& j : name) { - if (j == ' ') j = '$'; - } - - if (name.find(".code64") != std::string::npos) { - // data is treated as code. - - name_copy.erase(name_copy.find(".code64"), strlen(".code64")); - kCurrentRecord.fKind = LibCompiler::kPefCode; - } else if (name.find(".data64") != std::string::npos) { - // no code will be executed from here. - - name_copy.erase(name_copy.find(".data64"), strlen(".data64")); - kCurrentRecord.fKind = LibCompiler::kPefData; - } else if (name.find(".zero64") != std::string::npos) { - // this is a bss section. - - name_copy.erase(name_copy.find(".zero64"), strlen(".zero64")); - kCurrentRecord.fKind = LibCompiler::kPefZero; - } - - // this is a special case for the start stub. - // we want this so that li can find it. - - if (name == kPefStart) { - kCurrentRecord.fKind = LibCompiler::kPefCode; - } - - while (name_copy.find(" ") != std::string::npos) name_copy.erase(name_copy.find(" "), 1); - - kOriginLabel.push_back(std::make_pair(name_copy, kOrigin)); - ++kOrigin; - - // now we can tell the code size of the previous kCurrentRecord. - - if (!kRecords.empty()) kRecords[kRecords.size() - 1].fSize = kBytes.size(); - - memset(kCurrentRecord.fName, 0, kAESymbolLen); - memcpy(kCurrentRecord.fName, name.c_str(), name.size()); - - ++kCounter; - - memset(kCurrentRecord.fPad, kAENullType, kAEPad); - - kRecords.emplace_back(kCurrentRecord); - - return true; - } - - return false; -} - -// \brief algorithms and helpers. - -namespace Detail::algorithm { -// \brief authorize a brief set of characters. -static inline bool is_not_alnum_space(char c) { - return !(isalpha(c) || isdigit(c) || (c == ' ') || (c == '\t') || (c == ',') || (c == '(') || - (c == ')') || (c == '"') || (c == '\'') || (c == '[') || (c == ']') || (c == '+') || - (c == '_') || (c == ':') || (c == '@') || (c == '.')); -} - -bool is_valid_power64(std::string str) { - return std::find_if(str.begin(), str.end(), is_not_alnum_space) == str.end(); -} -} // namespace Detail::algorithm - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Check for line (syntax check) - -///////////////////////////////////////////////////////////////////////////////////////// - -std::string LibCompiler::EncoderPowerPC::CheckLine(std::string line, std::string file) { - std::string err_str; - - if (line.empty() || LibCompiler::find_word(line, "extern_segment") || - LibCompiler::find_word(line, "public_segment") || line.find('#') != std::string::npos || - LibCompiler::find_word(line, ";")) { - if (line.find('#') != std::string::npos) { - line.erase(line.find('#')); - } else if (line.find(';') != std::string::npos) { - line.erase(line.find(';')); - } else { - /// does the line contains valid input? - if (!Detail::algorithm::is_valid_power64(line)) { - err_str = "Line contains non alphanumeric characters.\nhere -> "; - err_str += line; - } - } - - return err_str; - } - - if (!Detail::algorithm::is_valid_power64(line)) { - err_str = "Line contains non alphanumeric characters.\nhere -> "; - err_str += line; - - return err_str; - } - - // check for a valid instruction format. - - if (line.find(',') != std::string::npos) { - if (line.find(',') + 1 == line.size()) { - err_str += "\nInstruction lacks right register, here -> "; - err_str += line.substr(line.find(',')); - - return err_str; - } else { - bool nothing_on_right = true; - - if (line.find(',') + 1 > line.size()) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - - auto substr = line.substr(line.find(',') + 1); - - for (auto& ch : substr) { - if (ch != ' ' && ch != '\t') { - nothing_on_right = false; - } - } - - // this means we found nothing after that ',' . - if (nothing_on_right) { - err_str += "\nInstruction not complete, here -> "; - err_str += line; - - return err_str; - } - } - } - - // these do take an argument. - std::vector operands_inst = {"stw", "li"}; - - // these don't. - std::vector filter_inst = {"blr", "bl", "sc"}; - - for (auto& opcode_risc : kOpcodesPowerPC) { - if (LibCompiler::find_word(line, opcode_risc.name)) { - for (auto& op : operands_inst) { - // if only the instruction was found. - if (line == op) { - err_str += "\nMalformed "; - err_str += op; - err_str += " instruction, here -> "; - err_str += line; - } - } - - // if it is like that -> addr1, 0x0 - if (auto it = std::find(filter_inst.begin(), filter_inst.end(), opcode_risc.name); - it == filter_inst.cend()) { - if (LibCompiler::find_word(line, opcode_risc.name)) { - if (!isspace(line[line.find(opcode_risc.name) + strlen(opcode_risc.name)])) { - err_str += "\nMissing space between "; - err_str += opcode_risc.name; - err_str += " and operands.\nhere -> "; - err_str += line; - } - } - } - - return err_str; - } - } - - err_str += "Unrecognized instruction: " + line; - - return err_str; -} - -bool LibCompiler::EncoderPowerPC::WriteNumber(const std::size_t& pos, std::string& jump_label) { - if (!isdigit(jump_label[pos])) return false; - - switch (jump_label[pos + 1]) { - case 'x': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16); !res) { - if (errno != 0) { - Detail::print_error("invalid hex number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_hex"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 16)); - - for (char& i : num.number) { - kBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerPower: found a base 16 number here: " << jump_label.substr(pos) - << "\n"; - } - - return true; - } - case 'b': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2); !res) { - if (errno != 0) { - Detail::print_error("invalid binary number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_bin"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 2)); - - if (kVerbose) { - kStdOut << "AssemblerPower: found a base 2 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - kBytes.push_back(i); - } - - return true; - } - case 'o': { - if (auto res = strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7); !res) { - if (errno != 0) { - Detail::print_error("invalid octal number: " + jump_label, "LibCompiler"); - throw std::runtime_error("invalid_octal"); - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos + 2).c_str(), nullptr, 7)); - - if (kVerbose) { - kStdOut << "AssemblerPower: found a base 8 number here: " << jump_label.substr(pos) << "\n"; - } - - for (char& i : num.number) { - kBytes.push_back(i); - } - - return true; - } - default: { - break; - } - } - - /* check for errno and stuff like that */ - if (auto res = strtol(jump_label.substr(pos).c_str(), nullptr, 10); !res) { - if (errno != 0) { - return false; - } - } - - LibCompiler::NumberCast64 num(strtol(jump_label.substr(pos).c_str(), nullptr, 10)); - - for (char& i : num.number) { - kBytes.push_back(i); - } - - if (kVerbose) { - kStdOut << "AssemblerPower: found a base 10 number here: " << jump_label.substr(pos) << "\n"; - } - - return true; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @brief Read and write an instruction to the output array. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool LibCompiler::EncoderPowerPC::WriteLine(std::string line, std::string file) { - if (LibCompiler::find_word(line, "public_segment")) return false; - if (!Detail::algorithm::is_valid_power64(line)) return false; - - for (auto& opcode_risc : kOpcodesPowerPC) { - // strict check here - if (LibCompiler::find_word(line, opcode_risc.name)) { - std::string name(opcode_risc.name); - std::string jump_label, cpy_jump_label; - std::vector found_registers_index; - - // check funct7 type. - switch (opcode_risc.ops->type) { - default: { - NumberCast32 num(opcode_risc.opcode); - - for (auto ch : num.number) { - kBytes.emplace_back(ch); - } - break; - } - case BADDR: - case PCREL: { - auto num = GetNumber32(line, name); - - kBytes.emplace_back(num.number[0]); - kBytes.emplace_back(num.number[1]); - kBytes.emplace_back(num.number[2]); - kBytes.emplace_back(0x48); - - break; - } - /// General purpose, float, vector operations. Everything that involve - /// registers. - case G0REG: - case FREG: - case VREG: - case GREG: { - // \brief how many registers we found. - std::size_t found_some_count = 0UL; - std::size_t register_count = 0UL; - std::string opcodeName = opcode_risc.name; - std::size_t register_sum = 0; - - NumberCast64 num(opcode_risc.opcode); - - for (size_t line_index = 0UL; line_index < line.size(); line_index++) { - if (line[line_index] == kAsmRegisterPrefix[0] && isdigit(line[line_index + 1])) { - std::string register_syntax = kAsmRegisterPrefix; - register_syntax += line[line_index + 1]; - - if (isdigit(line[line_index + 2])) register_syntax += line[line_index + 2]; - - std::string reg_str; - reg_str += line[line_index + 1]; - - if (isdigit(line[line_index + 2])) reg_str += line[line_index + 2]; - - // it ranges from r0 to r19 - // something like r190 doesn't exist in the instruction set. - if (isdigit(line[line_index + 3]) && isdigit(line[line_index + 2])) { - reg_str += line[line_index + 3]; - Detail::print_error("invalid register index, r" + reg_str + - "\nnote: The POWER accepts registers from r0 to r32.", - file); - throw std::runtime_error("invalid_register_index"); - } - - // finally cast to a size_t - std::size_t reg_index = strtol(reg_str.c_str(), nullptr, 10); - - if (reg_index > kAsmRegisterLimit) { - Detail::print_error("invalid register index, r" + reg_str, file); - throw std::runtime_error("invalid_register_index"); - } - - if (opcodeName == "li") { - char numIndex = 0; - - for (size_t i = 0; i != reg_index; i++) { - numIndex += 0x20; - } - - auto num = GetNumber32(line, reg_str); - - kBytes.push_back(num.number[0]); - kBytes.push_back(num.number[1]); - kBytes.push_back(numIndex); - kBytes.push_back(0x38); - - // check if bigger than two. - for (size_t i = 2; i < 4; i++) { - if (num.number[i] > 0) { - Detail::print_warning("number overflow on li operation.", file); - break; - } - } - - break; - } - - if ((opcodeName[0] == 's' && opcodeName[1] == 't')) { - if (register_sum == 0) { - for (size_t indexReg = 0UL; indexReg < reg_index; ++indexReg) { - register_sum += 0x20; - } - } else { - register_sum += reg_index; - } - } - - if (opcodeName == "mr") { - switch (register_count) { - case 0: { - kBytes.push_back(0x78); - - char numIndex = 0x3; - - for (size_t i = 0; i != reg_index; i++) { - numIndex += 0x8; - } - - kBytes.push_back(numIndex); - - break; - } - case 1: { - char numIndex = 0x1; - - for (size_t i = 0; i != reg_index; i++) { - numIndex += 0x20; - } - - for (size_t i = 0; i != reg_index; i++) { - kBytes[kBytes.size() - 1] += 0x8; - } - - kBytes[kBytes.size() - 1] -= 0x8; - - kBytes.push_back(numIndex); - - if (reg_index >= 10 && reg_index < 20) - kBytes.push_back(0x7d); - else if (reg_index >= 20 && reg_index < 30) - kBytes.push_back(0x7e); - else if (reg_index >= 30) - kBytes.push_back(0x7f); - else - kBytes.push_back(0x7c); - - break; - } - default: - break; - } - - ++register_count; - ++found_some_count; - } - - if (opcodeName == "addi") { - if (found_some_count == 2 || found_some_count == 0) - kBytes.emplace_back(reg_index); - else if (found_some_count == 1) - kBytes.emplace_back(0x00); - - ++found_some_count; - - if (found_some_count > 3) { - Detail::print_error("Too much registers. -> " + line, file); - throw std::runtime_error("too_much_regs"); - } - } - - if (opcodeName.find("cmp") != std::string::npos) { - ++found_some_count; - - if (found_some_count > 3) { - Detail::print_error("Too much registers. -> " + line, file); - throw std::runtime_error("too_much_regs"); - } - } - - if (opcodeName.find("mf") != std::string::npos || - opcodeName.find("mt") != std::string::npos) { - char numIndex = 0; - - for (size_t i = 0; i != reg_index; i++) { - numIndex += 0x20; - } - - num.number[2] += numIndex; - - ++found_some_count; - - if (found_some_count > 1) { - Detail::print_error("Too much registers. -> " + line, file); - throw std::runtime_error("too_much_regs"); - } - - if (kVerbose) { - kStdOut << "AssemblerPower: Found register: " << register_syntax << "\n"; - kStdOut << "AssemblerPower: Amount of registers in instruction: " - << found_some_count << "\n"; - } - - if (reg_index >= 10 && reg_index < 20) - num.number[3] = 0x7d; - else if (reg_index >= 20 && reg_index < 30) - num.number[3] = 0x7e; - else if (reg_index >= 30) - num.number[3] = 0x7f; - else - num.number[3] = 0x7c; - - for (auto ch : num.number) { - kBytes.emplace_back(ch); - } - } - - found_registers_index.push_back(reg_index); - } - } - - if (opcodeName == "addi") { - kBytes.emplace_back(0x38); - } - - if (opcodeName.find("cmp") != std::string::npos) { - char rightReg = 0x0; - - for (size_t i = 0; i != found_registers_index[1]; i++) { - rightReg += 0x08; - } - - kBytes.emplace_back(0x00); - kBytes.emplace_back(rightReg); - kBytes.emplace_back(found_registers_index[0]); - kBytes.emplace_back(0x7c); - } - - if ((opcodeName[0] == 's' && opcodeName[1] == 't')) { - size_t offset = 0UL; - - if (line.find('+') != std::string::npos) { - auto number = GetNumber32(line.substr(line.find("+")), "+"); - offset = number.raw; - } - - kBytes.push_back(offset); - kBytes.push_back(0x00); - kBytes.push_back(register_sum); - - kBytes.emplace_back(0x90); - } - - if (opcodeName == "mr") { - if (register_count == 1) { - Detail::print_error("Too few registers. -> " + line, file); - throw std::runtime_error("too_few_registers"); - } - } - - // we're not in immediate addressing, reg to reg. - if (opcode_risc.ops->type != GREG) { - // remember! register to register! - if (found_some_count == 1) { - Detail::print_error( - "Unrecognized register found.\ntip: each AssemblerPower register " - "starts with 'r'.\nline: " + - line, - file); - - throw std::runtime_error("not_a_register"); - } - } - - if (found_some_count < 1 && name[0] != 'l' && name[0] != 's') { - Detail::print_error("invalid combination of opcode and registers.\nline: " + line, - file); - throw std::runtime_error("invalid_comb_op_reg"); - } - - break; - } - } - - kOrigin += cPowerIPAlignment; - break; - } - } - - return true; -} - -// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/AssemblyFactory+AssemblyInterface.cc b/dev/LibCompiler/src/AssemblyFactory+AssemblyInterface.cc deleted file mode 100644 index e6072d9..0000000 --- a/dev/LibCompiler/src/AssemblyFactory+AssemblyInterface.cc +++ /dev/null @@ -1,55 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved - -------------------------------------------- */ - -#include -#include - -/** - * @file AssemblyFactory.cxx - * @author amlal (amlal@nekernel.org) - * @brief Assembler Kit - * @version 0.1 - * @date 2024-01-27 - * - * @copyright Copyright (c) 2024-2025 Amlal El Mahrouss - * - */ - -//! @file Asm.cpp -//! @brief AssemblyKit source implementation. - -namespace LibCompiler { -///! @brief Compile for specific format (ELF, PEF, ZBIN) -Int32 AssemblyFactory::Compile(std::string sourceFile, const Int32& arch) noexcept { - if (sourceFile.length() < 1 || !fMounted) return LIBCOMPILER_UNIMPLEMENTED; - - if (arch != fMounted->Arch()) return LIBCOMPILER_INVALID_ARCH; - - try { - return fMounted->CompileToFormat(sourceFile, arch); - } catch (std::exception& e) { - return LIBCOMPILER_EXEC_ERROR; - } -} - -///! @brief mount assembly backend. -void AssemblyFactory::Mount(AssemblyInterface* mountPtr) noexcept { - if (mountPtr) { - fMounted = mountPtr; - } -} - -///! @brief Unmount assembler. -AssemblyInterface* AssemblyFactory::Unmount() noexcept { - auto mount_prev = fMounted; - - if (mount_prev) { - fMounted = nullptr; - } - - return mount_prev; -} -} // namespace LibCompiler diff --git a/dev/LibCompiler/src/CCompiler64x0.cc b/dev/LibCompiler/src/CCompiler64x0.cc deleted file mode 100644 index 55adc75..0000000 --- a/dev/LibCompiler/src/CCompiler64x0.cc +++ /dev/null @@ -1,1286 +0,0 @@ -/* - * ======================================================== - * - * cc - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -/// BUGS: 0 -/// TODO: none - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -/* C driver */ -/* This is part of the LibCompiler. */ -/* (c) Amlal El Mahrouss */ - -/// @author EL Mahrouss Amlal (amlel) -/// @file 64x0-cc.cxx -/// @brief 64x0 C Compiler. - -/// TODO: support structures, else if, else, . and -> - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -#define kExitOK (0) - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" - -///////////////////////////////////// - -// INTERNAL STUFF OF THE C COMPILER - -///////////////////////////////////// - -namespace Detail { -// \brief Register map structure, used to keep track of each variable's registers. -struct CompilerRegisterMap final { - std::string fName; - std::string fReg; -}; - -// \brief Map for C structs -// \author amlel -struct CompilerStructMap final { - // 'my_foo' - std::string fName; - - // if instance: stores a valid register. - std::string fReg; - - // offset count - std::size_t fOffsetsCnt; - - // offset array. - std::vector> fOffsets; -}; - -struct CompilerState final { - std::vector fSyntaxTreeList; - std::vector kStackFrame; - std::vector kStructMap; - LibCompiler::SyntaxLeafList* fSyntaxTree{nullptr}; - std::unique_ptr fOutputAssembly; - std::string fLastFile; - std::string fLastError; - bool fVerbose; -}; -} // namespace Detail - -static Detail::CompilerState kState; -static std::string kIfFunction = ""; - -namespace Detail { -/// @brief prints an error into stdout. -/// @param reason the reason of the error. -/// @param file where does it originate from? -void print_error(std::string reason, std::string file) noexcept; - -struct CompilerType final { - std::string fName; - std::string fValue; -}; -} // namespace Detail - -///////////////////////////////////////////////////////////////////////////////////////// - -// Target architecture. -static int kMachine = 0; - -///////////////////////////////////////// - -// REGISTERS ACCORDING TO USED ASSEMBLER - -///////////////////////////////////////// - -static size_t kRegisterCnt = kAsmRegisterLimit; -static size_t kStartUsable = 2; -static size_t kUsableLimit = 15; -static size_t kRegisterCounter = kStartUsable; -static std::string kRegisterPrefix = kAsmRegisterPrefix; - -///////////////////////////////////////// - -// COMPILER PARSING UTILITIES/STATES. - -///////////////////////////////////////// - -static std::vector kFileList; -static LibCompiler::AssemblyFactory kFactory; -static bool kInStruct = false; -static bool kOnWhileLoop = false; -static bool kOnForLoop = false; -static bool kInBraces = false; -static bool kIfFound = false; -static size_t kBracesCount = 0UL; - -/* @brief C compiler backend for C */ -class CompilerFrontend64x0 final : public LibCompiler::CompilerFrontendInterface { - public: - explicit CompilerFrontend64x0() = default; - ~CompilerFrontend64x0() override = default; - - LIBCOMPILER_COPY_DEFAULT(CompilerFrontend64x0); - - std::string Check(const char* text, const char* file); - bool Compile(std::string text, std::string file) override; - - const char* Language() override { return "64k C"; } -}; - -static CompilerFrontend64x0* kCompilerFrontend = nullptr; -static std::vector kCompilerVariables; -static std::vector kCompilerFunctions; -static std::vector kCompilerTypes; - -namespace Detail { -union number_cast final { - public: - number_cast(UInt64 _Raw) : _Raw(_Raw) {} - - public: - char _Num[8]; - UInt64 _Raw; -}; - -union double_cast final { - public: - double_cast(float _Raw) : _Raw(_Raw) {} - - public: - char _Sign; - char _Lh[8]; - char _Rh[23]; - - float _Raw; -}; -} // namespace Detail - -///////////////////////////////////////////////////////////////////////////////////////// - -// @name Compile -// @brief Generate MASM from a C assignement. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool CompilerFrontend64x0::Compile(std::string text_, std::string file) { - std::string text = text_; - - bool typeFound = false; - bool fnFound = false; - - // setup generator. - std::random_device rd; - - auto seed_data = std::array{}; - std::generate(std::begin(seed_data), std::end(seed_data), std::ref(rd)); - std::seed_seq seq(std::begin(seed_data), std::end(seed_data)); - std::mt19937 generator(seq); - - // start parsing - for (size_t text_index = 0; text_index < text.size(); ++text_index) { - auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); - - auto gen = uuids::uuid_random_generator{generator}; - uuids::uuid out = gen(); - - Detail::number_cast time_off = (UInt64) out.as_bytes().data(); - - if (!typeFound) { - auto substr = text.substr(text_index); - std::string match_type; - - for (size_t y = 0; y < substr.size(); ++y) { - if (substr[y] == ' ') { - while (match_type.find(' ') != std::string::npos) { - match_type.erase(match_type.find(' ')); - } - - for (auto& clType : kCompilerTypes) { - if (clType.fName == match_type) { - match_type.clear(); - - std::string buf; - - buf += clType.fValue; - buf += ' '; - - if (substr.find('=') != std::string::npos) { - break; - } - - if (text.find('(') != std::string::npos) { - syntaxLeaf.fUserValue = buf; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - } - - typeFound = true; - break; - } - } - - break; - } - - match_type += substr[y]; - } - } - - if (text[text_index] == '{') { - if (kInStruct) { - continue; - } - - kInBraces = true; - ++kBracesCount; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - } - - // return keyword handler - if (text[text_index] == 'r') { - std::string return_keyword; - return_keyword += "return"; - - std::size_t index = 0UL; - - std::string value; - - for (size_t return_index = text_index; return_index < text.size(); ++return_index) { - if (text[return_index] != return_keyword[index]) { - for (size_t value_index = return_index; value_index < text.size(); ++value_index) { - if (text[value_index] == ';') break; - - value += text[value_index]; - } - - break; - } - - ++index; - } - - if (index == return_keyword.size()) { - if (!value.empty()) { - if (value.find('(') != std::string::npos) { - value.erase(value.find('(')); - } - - if (!isdigit(value[value.find('(') + 2])) { - std::string tmp = value; - bool reg_to_reg = false; - - value.clear(); - - value += " extern_segment"; - value += tmp; - } - - syntaxLeaf.fUserValue = "\tldw r19, "; - - // make it pretty. - if (value.find('\t') != std::string::npos) value.erase(value.find('\t'), 1); - - syntaxLeaf.fUserValue += value + "\n"; - } - - syntaxLeaf.fUserValue += "\tjlr"; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - break; - } - } - - if (text[text_index] == 'i' && text[text_index + 1] == 'f') { - auto expr = text.substr(text_index + 2); - text.erase(text_index, 2); - - if (expr.find("{") != std::string::npos) { - expr.erase(expr.find("{")); - } - - if (expr.find("(") != std::string::npos) expr.erase(expr.find("(")); - - if (expr.find(")") != std::string::npos) expr.erase(expr.find(")")); - - kIfFunction = "__LIBCOMPILER_IF_PROC_"; - kIfFunction += std::to_string(time_off._Raw); - - syntaxLeaf.fUserValue = "\tlda r12, extern_segment "; - syntaxLeaf.fUserValue += kIfFunction + - "\n\t#r12 = Code to jump on, r11 right cond, r10 left cond.\n\tbeq " - "r10, r11, r12\ndword public_segment .code64 " + - kIfFunction + "\n"; - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - kIfFound = true; - } - - // Parse expressions and instructions here. - // what does this mean? - // we encounter an assignment, or we reached the end of an expression. - if (text[text_index] == '=' || text[text_index] == ';') { - if (fnFound) continue; - if (kIfFound) continue; - - if (text[text_index] == ';' && kInStruct) continue; - - if (text.find("typedef ") != std::string::npos) continue; - - if (text[text_index] == '=' && kInStruct) { - Detail::print_error("assignement of value in struct " + text, file); - continue; - } - - if (text[text_index] == ';' && kInStruct) { - bool space_found_ = false; - std::string sym; - - for (auto& ch : text) { - if (ch == ' ') { - space_found_ = true; - } - - if (ch == ';') break; - - if (space_found_) sym.push_back(ch); - } - - kState.kStructMap[kState.kStructMap.size() - 1].fOffsets.push_back( - std::make_pair(kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4, sym)); - - kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt = - kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4; - - continue; - } - - if (text[text_index] == '=' && kInStruct) { - continue; - } - - if (text[text_index + 1] == '=' || text[text_index - 1] == '!' || - text[text_index - 1] == '<' || text[text_index - 1] == '>') { - continue; - } - - std::string substr; - - if (text.find('=') != std::string::npos && kInBraces && !kIfFound) { - if (text.find("*") != std::string::npos) { - if (text.find("=") > text.find("*")) - substr += "\tlda "; - else - substr += "\tldw "; - } else { - substr += "\tldw "; - } - } else if (text.find('=') != std::string::npos && !kInBraces) { - substr += "stw public_segment .data64 "; - } - - int first_encountered = 0; - - std::string str_name; - - for (size_t text_index_2 = 0; text_index_2 < text.size(); ++text_index_2) { - if (text[text_index_2] == '\"') { - ++text_index_2; - - // want to add this, so that the parser recognizes that this is a - // string. - substr += '"'; - - for (; text_index_2 < text.size(); ++text_index_2) { - if (text[text_index_2] == '\"') break; - - substr += text[text_index_2]; - } - } - - if (text[text_index_2] == '{' || text[text_index_2] == '}') continue; - - if (text[text_index_2] == ';') { - break; - } - - if (text[text_index_2] == ' ' || text[text_index_2] == '\t') { - if (first_encountered != 2) { - if (text[text_index] != '=' && - substr.find("public_segment .data64") == std::string::npos && !kInStruct) - substr += "public_segment .data64 "; - } - - ++first_encountered; - - continue; - } - - if (text[text_index_2] == '=') { - if (!kInBraces) { - substr.replace(substr.find("public_segment .data64"), strlen("public_segment .data64"), - "public_segment .zero64 "); - } - - substr += ","; - continue; - } - - substr += text[text_index_2]; - } - - for (auto& clType : kCompilerTypes) { - if (substr.find(clType.fName) != std::string::npos) { - if (substr.find(clType.fName) > substr.find('"')) continue; - - substr.erase(substr.find(clType.fName), clType.fName.size()); - } else if (substr.find(clType.fValue) != std::string::npos) { - if (substr.find(clType.fValue) > substr.find('"')) continue; - - if (clType.fName == "const") continue; - - substr.erase(substr.find(clType.fValue), clType.fValue.size()); - } - } - - if (substr.find("extern") != std::string::npos) { - substr.replace(substr.find("extern"), strlen("extern"), "extern_segment "); - - if (substr.find("public_segment .data64") != std::string::npos) - substr.erase(substr.find("public_segment .data64"), strlen("public_segment .data64")); - } - - auto var_to_find = std::find_if( - kCompilerVariables.cbegin(), kCompilerVariables.cend(), - [&](Detail::CompilerType type) { return type.fName.find(substr) != std::string::npos; }); - - if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; - - std::string reg = kAsmRegisterPrefix; - reg += std::to_string(kRegisterCounter); - - if (var_to_find == kCompilerVariables.cend()) { - ++kRegisterCounter; - - kState.kStackFrame.push_back({.fName = substr, .fReg = reg}); - kCompilerVariables.push_back({.fName = substr}); - } - - syntaxLeaf.fUserValue += substr; - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - if (text[text_index] == '=') break; - } - - // function handler. - - if (text[text_index] == '(' && !fnFound && !kIfFound) { - std::string substr; - std::string args_buffer; - std::string args; - - bool type_crossed = false; - - for (size_t idx = text.find('(') + 1; idx < text.size(); ++idx) { - if (text[idx] == ',') continue; - - if (text[idx] == ' ') continue; - - if (text[idx] == ')') break; - } - - for (char substr_first_index : text) { - if (substr_first_index != ',') - args_buffer += substr_first_index; - else - args_buffer += '$'; - - if (substr_first_index == ';') { - args_buffer = args_buffer.erase(0, args_buffer.find('(')); - args_buffer = args_buffer.erase(args_buffer.find(';'), 1); - args_buffer = args_buffer.erase(args_buffer.find(')'), 1); - args_buffer = args_buffer.erase(args_buffer.find('('), 1); - - if (!args_buffer.empty()) args += "\tldw r6, "; - - std::string register_type; - std::size_t index = 7UL; - - while (args_buffer.find("$") != std::string::npos) { - register_type = kRegisterPrefix; - register_type += std::to_string(index); - - ++index; - - args_buffer.replace(args_buffer.find('$'), 1, "\n\tldw " + register_type + ","); - } - - args += args_buffer; - args += "\n\tlda r19, "; - } - } - - for (char _text_i : text) { - if (_text_i == '\t' || _text_i == ' ') { - if (!type_crossed) { - substr.clear(); - type_crossed = true; - } - - continue; - } - - if (_text_i == '(') break; - - substr += _text_i; - } - - if (kInBraces) { - syntaxLeaf.fUserValue = args; - syntaxLeaf.fUserValue += substr; - syntaxLeaf.fUserValue += "\n\tjrl\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - fnFound = true; - } else { - syntaxLeaf.fUserValue.clear(); - - syntaxLeaf.fUserValue += "public_segment .code64 "; - - syntaxLeaf.fUserValue += substr; - syntaxLeaf.fUserValue += "\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - fnFound = true; - } - - kCompilerFunctions.push_back(text); - } - - if (text[text_index] == '-' && text[text_index + 1] == '-') { - text = text.replace(text.find("--"), strlen("--"), ""); - - for (int _text_i = 0; _text_i < text.size(); ++_text_i) { - if (text[_text_i] == '\t' || text[_text_i] == ' ') text.erase(_text_i, 1); - } - - syntaxLeaf.fUserValue += "sub "; - syntaxLeaf.fUserValue += text; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - break; - } - - if (text[text_index] == '}') { - kRegisterCounter = kStartUsable; - - --kBracesCount; - - if (kBracesCount < 1) { - kInBraces = false; - kBracesCount = 0; - } - - if (kIfFound) kIfFound = false; - - if (kInStruct) kInStruct = false; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - } - - syntaxLeaf.fUserValue.clear(); - } - - auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); - syntaxLeaf.fUserValue = "\n"; - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - return true; -} - -static bool kShouldHaveBraces = false; -static std::string kFnName; - -std::string CompilerFrontend64x0::Check(const char* text, const char* file) { - std::string err_str; - std::string ln = text; - - if (ln.empty()) { - return err_str; - } - - bool non_ascii_found = false; - - for (int i = 0; i < ln.size(); ++i) { - if (isalnum(ln[i])) { - non_ascii_found = true; - break; - } - } - - if (kShouldHaveBraces && ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - } - - if (!non_ascii_found) return err_str; - - size_t string_index = 1UL; - - if (ln.find('\'') != std::string::npos) { - string_index = ln.find('\'') + 1; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == '\'') { - if (ln[string_index + 1] != ';') { - ln.erase(string_index, 1); - } - - return err_str; - } - } - } else if (ln.find('"') != std::string::npos) { - string_index = ln.find('"') + 1; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == '"') { - if (ln[string_index + 1] != ';') { - ln.erase(string_index, 1); - } else { - break; - } - } - } - } else if (ln.find('"') == std::string::npos && ln.find('\'') == std::string::npos) { - std::vector forbidden_words; - - forbidden_words.push_back("\\"); - forbidden_words.push_back("?"); - forbidden_words.push_back("@"); - forbidden_words.push_back("~"); - forbidden_words.push_back("::"); - forbidden_words.push_back("--*"); - forbidden_words.push_back("*/"); - - // add them to avoid stupid mistakes. - forbidden_words.push_back("namespace"); - forbidden_words.push_back("class"); - forbidden_words.push_back("extern \"C\""); - - for (auto& forbidden : forbidden_words) { - if (ln.find(forbidden) != std::string::npos) { - err_str += "\nForbidden character detected: "; - err_str += forbidden; - - return err_str; - } - } - } - - struct CompilerVariableRange final { - std::string fBegin; - std::string fEnd; - }; - - const std::vector variables_list = { - {.fBegin = "static ", .fEnd = "="}, {.fBegin = "=", .fEnd = ";"}, - {.fBegin = "if(", .fEnd = "="}, {.fBegin = "if (", .fEnd = "="}, - {.fBegin = "if(", .fEnd = "<"}, {.fBegin = "if (", .fEnd = "<"}, - {.fBegin = "if(", .fEnd = ">"}, {.fBegin = "if (", .fEnd = ">"}, - {.fBegin = "if(", .fEnd = ")"}, {.fBegin = "if (", .fEnd = ")"}, - - {.fBegin = "else(", .fEnd = "="}, {.fBegin = "else (", .fEnd = "="}, - {.fBegin = "else(", .fEnd = "<"}, {.fBegin = "else (", .fEnd = "<"}, - {.fBegin = "else(", .fEnd = ">"}, {.fBegin = "else (", .fEnd = ">"}, - {.fBegin = "else(", .fEnd = ")"}, {.fBegin = "else (", .fEnd = ")"}, - }; - - for (auto& variable : variables_list) { - if (ln.find(variable.fBegin) != std::string::npos) { - string_index = ln.find(variable.fBegin) + variable.fBegin.size(); - - while (ln[string_index] == ' ') ++string_index; - - std::string keyword; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == variable.fEnd[0]) { - std::string varname = ""; - - for (size_t index_keyword = ln.find(' '); ln[index_keyword] != variable.fBegin[0]; - ++index_keyword) { - if (ln[index_keyword] == ' ') { - continue; - } - - if (isdigit(ln[index_keyword])) { - goto cc_next_loop; - } - - varname += ln[index_keyword]; - } - - if (varname.find(' ') != std::string::npos) { - varname.erase(0, varname.find(' ')); - - if (variable.fBegin == "extern") { - varname.erase(0, varname.find(' ')); - } - } - - if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; - - std::string reg = kAsmRegisterPrefix; - reg += std::to_string(kRegisterCounter); - - kCompilerVariables.push_back({.fValue = varname}); - goto cc_check_done; - } - - keyword.push_back(ln[string_index]); - } - - goto cc_next_loop; - - cc_check_done: - - // skip digit value. - if (isdigit(keyword[0]) || keyword[0] == '"') { - goto cc_next_loop; - } - - while (keyword.find(' ') != std::string::npos) keyword.erase(keyword.find(' '), 1); - - for (auto& var : kCompilerVariables) { - if (var.fValue.find(keyword) != std::string::npos) { - err_str.clear(); - goto cc_next; - } - } - - for (auto& fn : kCompilerFunctions) { - if (fn.find(keyword[0]) != std::string::npos) { - auto where_begin = fn.find(keyword[0]); - auto keyword_begin = 0UL; - auto failed = false; - - for (; where_begin < keyword.size(); ++where_begin) { - if (fn[where_begin] == '(' && keyword[keyword_begin] == '(') break; - - if (fn[where_begin] != keyword[keyword_begin]) { - failed = true; - break; - } - - ++keyword_begin; - } - - if (!failed) { - err_str.clear(); - goto cc_next; - } else { - continue; - } - } - } - - cc_error_value: - if (keyword.find("->") != std::string::npos) return err_str; - - if (keyword.find(".") != std::string::npos) return err_str; - - if (isalnum(keyword[0])) err_str += "\nUndefined value: " + keyword; - - return err_str; - } - - cc_next_loop: - continue; - } - -cc_next: - - // extern does not declare anything, it extern_segments a variable. - // so that's why it's not declare upper. - if (LibCompiler::find_word(ln, "extern")) { - auto substr = ln.substr(ln.find("extern") + strlen("extern")); - kCompilerVariables.push_back({.fValue = substr}); - } - - if (kShouldHaveBraces && ln.find('{') == std::string::npos) { - err_str += "Missing '{' for function "; - err_str += kFnName; - err_str += "\n"; - - kShouldHaveBraces = false; - kFnName.clear(); - } else if (kShouldHaveBraces && ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - kFnName.clear(); - } - - bool type_not_found = true; - - if (ln.find('\'') != std::string::npos) { - ln.replace(ln.find('\''), 3, "0"); - } - - auto first = ln.find('"'); - if (first != std::string::npos) { - auto second = 0UL; - bool found_second_quote = false; - - for (size_t i = first + 1; i < ln.size(); ++i) { - if (ln[i] == '\"') { - found_second_quote = true; - second = i; - - break; - } - } - - if (!found_second_quote) { - err_str += "Missing terminating \"."; - err_str += " here -> " + ln.substr(ln.find('"'), second); - } - } - - if (ln.find(')') != std::string::npos && ln.find(';') == std::string::npos) { - if (ln.find('{') == std::string::npos) { - kFnName = ln; - kShouldHaveBraces = true; - - goto skip_braces_check; - } else if (ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - } - } - -skip_braces_check: - - for (auto& key : kCompilerTypes) { - if (LibCompiler::find_word(ln, key.fName)) { - if (isdigit(ln[ln.find(key.fName) + key.fName.size() + 1])) { - err_str += "\nNumber cannot be set for "; - err_str += key.fName; - err_str += "'s name. here -> "; - err_str += ln; - } - - if (ln.find(key.fName) == 0 || ln[ln.find(key.fName) - 1] == ' ' || - ln[ln.find(key.fName) - 1] == '\t') { - type_not_found = false; - - if (ln[ln.find(key.fName) + key.fName.size()] != ' ') { - type_not_found = true; - - if (ln[ln.find(key.fName) + key.fName.size()] == '\t') type_not_found = false; - - goto next; - } else if (ln[ln.find(key.fName) + key.fName.size()] != '\t') { - type_not_found = true; - - if (ln[ln.find(key.fName) + key.fName.size()] == ' ') type_not_found = false; - } - } - - next: - - if (ln.find(';') == std::string::npos) { - if (ln.find('(') != std::string::npos) { - if (ln.find('=') == std::string::npos) continue; - } - - err_str += "\nMissing ';', here -> "; - err_str += ln; - } else { - continue; - } - - if (ln.find('=') != std::string::npos) { - if (ln.find('(') != std::string::npos) { - if (ln.find(')') == std::string::npos) { - err_str += "\nMissing ')', after '(' here -> "; - err_str += ln.substr(ln.find('(')); - } - } - } - } - } - - if (kInBraces && ln.find("struct") != std::string::npos && - ln.find("union") != std::string::npos && ln.find("enum") != std::string::npos && - ln.find('=') != std::string::npos) { - if (ln.find(';') == std::string::npos) { - err_str += "\nMissing ';' after struct/union/enum declaration, here -> "; - err_str += ln; - } - } - - if (ln.find(';') != std::string::npos && ln.find("for") == std::string::npos) { - if (ln.find(';') + 1 != ln.size()) { - for (int i = 0; i < ln.substr(ln.find(';') + 1).size(); ++i) { - if ((ln.substr(ln.find(';') + 1)[i] != ' ') || (ln.substr(ln.find(';') + 1)[i] != '\t')) { - if (auto err = this->Check(ln.substr(ln.find(';') + 1).c_str(), file); !err.empty()) { - err_str += "\nUnexpected text after ';' -> "; - err_str += ln.substr(ln.find(';')); - err_str += err; - } - } - } - } - } - - if (ln.find('(') != std::string::npos) { - if (ln.find(';') == std::string::npos && !LibCompiler::find_word(ln, "|") && - !LibCompiler::find_word(ln, "||") && !LibCompiler::find_word(ln, "&") && - !LibCompiler::find_word(ln, "&&") && !LibCompiler::find_word(ln, "~")) { - bool found_func = false; - size_t i = ln.find('('); - std::vector opens; - std::vector closes; - - for (; i < ln.size(); ++i) { - if (ln[i] == ')') { - closes.push_back(1); - } - - if (ln[i] == '(') { - opens.push_back(1); - } - } - - if (closes.size() != opens.size()) err_str += "Unterminated (), here -> " + ln; - - bool space_found = false; - - for (int i = 0; i < ln.size(); ++i) { - if (ln[i] == ')' && !space_found) { - space_found = true; - continue; - } - - if (space_found) { - if (ln[i] == ' ' && isalnum(ln[i + 1])) { - err_str += "\nBad function format here -> "; - err_str += ln; - } - } - } - } - - if (ln.find('(') < 1) { - err_str += "\nMissing identifier before '(' here -> "; - err_str += ln; - } else { - if (type_not_found && ln.find(';') == std::string::npos && - ln.find("if") == std::string::npos && ln.find("|") == std::string::npos && - ln.find("&") == std::string::npos && ln.find("(") == std::string::npos && - ln.find(")") == std::string::npos) { - err_str += "\n Missing ';' or type, here -> "; - err_str += ln; - } - } - - if (ln.find(')') == std::string::npos) { - err_str += "\nMissing ')', after '(' here -> "; - err_str += ln.substr(ln.find('(')); - } - } else { - if (ln.find("for") != std::string::npos || ln.find("while") != std::string::npos) { - err_str += "\nMissing '(', after \"for\", here -> "; - err_str += ln; - } - } - - if (ln.find('}') != std::string::npos && !kInBraces) { - if (!kInStruct && ln.find(';') == std::string::npos) { - err_str += "\nMismatched '}', here -> "; - err_str += ln; - } - } - - if (!ln.empty()) { - if (ln.find(';') == std::string::npos && ln.find('{') == std::string::npos && - ln.find('}') == std::string::npos && ln.find(')') == std::string::npos && - ln.find('(') == std::string::npos && ln.find(',') == std::string::npos) { - if (ln.size() <= 2) return err_str; - - err_str += "\nMissing ';', here -> "; - err_str += ln; - } - } - - return err_str; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -/** - * @brief C To Assembly mount-point. - */ - -///////////////////////////////////////////////////////////////////////////////////////// - -class AssemblyCCInterface final LC_ASSEMBLY_INTERFACE { - public: - explicit AssemblyCCInterface() = default; - ~AssemblyCCInterface() override = default; - - LIBCOMPILER_COPY_DEFAULT(AssemblyCCInterface); - - UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArch64x0; } - - Int32 CompileToFormat(std::string src, Int32 arch) override { - if (kCompilerFrontend == nullptr) return 1; - - /* @brief copy contents wihtout extension */ - std::string src_file = src.data(); - std::ifstream src_fp = std::ifstream(src_file, std::ios::in); - std::string dest; - - for (auto& ch : src_file) { - if (ch == '.') { - break; - } - - dest += ch; - } - - /* According to PEF ABI. */ - std::vector exts = kAsmFileExts; - dest += exts[4]; - - kState.fOutputAssembly = std::make_unique(dest); - - auto fmt = LibCompiler::current_date(); - - (*kState.fOutputAssembly) << "# Path: " << src_file << "\n"; - (*kState.fOutputAssembly) << "# Language: 64x0 Assembly (Generated from ANSI C)\n"; - (*kState.fOutputAssembly) << "# Date: " << fmt << "\n\n"; - - LibCompiler::SyntaxLeafList syntax; - - kState.fSyntaxTreeList.push_back(syntax); - kState.fSyntaxTree = &kState.fSyntaxTreeList[kState.fSyntaxTreeList.size() - 1]; - - std::string line_src; - - while (std::getline(src_fp, line_src)) { - if (auto err = kCompilerFrontend->Check(line_src.c_str(), src.data()); err.empty()) { - kCompilerFrontend->Compile(line_src, src.data()); - } else { - Detail::print_error(err, src.data()); - } - } - - if (kAcceptableErrors > 0) return 1; - - std::vector keywords = {"ldw", "stw", "lda", "sta", "add", "sub", "mv"}; - - /// - /// Replace, optimize, fix assembly output. - /// - - for (auto& leaf : kState.fSyntaxTree->fLeafList) { - std::vector access_keywords = {"->", "."}; - - for (auto& access_ident : access_keywords) { - if (LibCompiler::find_word(leaf.fUserValue, access_ident)) { - for (auto& struc : kState.kStructMap) { - /// TODO: - } - } - } - - for (auto& keyword : keywords) { - if (LibCompiler::find_word(leaf.fUserValue, keyword)) { - std::size_t cnt = 0UL; - - for (auto& reg : kState.kStackFrame) { - std::string needle; - - for (size_t i = 0; i < reg.fName.size(); i++) { - if (reg.fName[i] == ' ') { - ++i; - - for (; i < reg.fName.size(); i++) { - if (reg.fName[i] == ',') { - break; - } - - if (reg.fName[i] == ' ') continue; - - needle += reg.fName[i]; - } - - break; - } - } - - if (LibCompiler::find_word(leaf.fUserValue, needle)) { - if (leaf.fUserValue.find("extern_segment " + needle) != std::string::npos) { - std::string range = "extern_segment " + needle; - leaf.fUserValue.replace(leaf.fUserValue.find("extern_segment " + needle), - range.size(), needle); - } - - if (leaf.fUserValue.find("ldw r6") != std::string::npos) { - std::string::difference_type countComma = - std::count(leaf.fUserValue.begin(), leaf.fUserValue.end(), ','); - - if (countComma == 1) { - leaf.fUserValue.replace(leaf.fUserValue.find("ldw"), strlen("ldw"), "mv"); - } - } - - leaf.fUserValue.replace(leaf.fUserValue.find(needle), needle.size(), reg.fReg); - - ++cnt; - } - } - - if (cnt > 1 && keyword != "mv" && keyword != "add" && keyword != "sub") { - leaf.fUserValue.replace(leaf.fUserValue.find(keyword), keyword.size(), "mv"); - } - } - } - } - - for (auto& leaf : kState.fSyntaxTree->fLeafList) { - (*kState.fOutputAssembly) << leaf.fUserValue; - } - - kState.fSyntaxTree = nullptr; - - kState.fOutputAssembly->flush(); - kState.fOutputAssembly.reset(); - - return kExitOK; - } -}; - -///////////////////////////////////////////////////////////////////////////////////////// - -#include - -#define kPrintF printf -#define kSplashCxx() kPrintF(kWhite "NE C Driver, %s, (c) Amlal El Mahrouss\n", kDistVersion) - -static void cc_print_help() { - kSplashCxx(); -} - -///////////////////////////////////////////////////////////////////////////////////////// - -#define kExt ".c" - -LIBCOMPILER_MODULE(NeOSCompilerCLang64x0) { - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - kCompilerTypes.push_back({.fName = "void", .fValue = "void"}); - kCompilerTypes.push_back({.fName = "char", .fValue = "byte"}); - kCompilerTypes.push_back({.fName = "short", .fValue = "hword"}); - kCompilerTypes.push_back({.fName = "int", .fValue = "dword"}); - kCompilerTypes.push_back({.fName = "long", .fValue = "qword"}); - kCompilerTypes.push_back({.fName = "*", .fValue = "offset"}); - - bool skip = false; - - kFactory.Mount(new AssemblyCCInterface()); - kMachine = LibCompiler::AssemblyFactory::kArch64x0; - kCompilerFrontend = new CompilerFrontend64x0(); - - for (auto index = 1UL; index < argc; ++index) { - if (skip) { - skip = false; - continue; - } - - if (argv[index][0] == '-') { - if (strcmp(argv[index], "--v") == 0 || strcmp(argv[index], "--version") == 0) { - kSplashCxx(); - return kExitOK; - } - - if (strcmp(argv[index], "--verbose") == 0) { - kState.fVerbose = true; - - continue; - } - - if (strcmp(argv[index], "--h") == 0 || strcmp(argv[index], "--help") == 0) { - cc_print_help(); - - return kExitOK; - } - - if (strcmp(argv[index], "--dialect") == 0) { - if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; - - return kExitOK; - } - - if (strcmp(argv[index], "--fmax-exceptions") == 0) { - try { - kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); - } - // catch anything here - catch (...) { - kErrorLimit = 0; - } - - skip = true; - - continue; - } - - std::string err = "Unknown command: "; - err += argv[index]; - - Detail::print_error(err, "cc"); - - continue; - } - - kFileList.emplace_back(argv[index]); - - std::string srcFile = argv[index]; - - if (strstr(argv[index], kExt) == nullptr) { - if (kState.fVerbose) { - Detail::print_error(srcFile + " is not a valid C source.\n", "cc"); - } - - return 1; - } - - if (kFactory.Compile(srcFile, kMachine) != kExitOK) return 1; - } - - return kExitOK; -} - -// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/CCompilerARM64.cc b/dev/LibCompiler/src/CCompilerARM64.cc deleted file mode 100644 index 7a342ae..0000000 --- a/dev/LibCompiler/src/CCompilerARM64.cc +++ /dev/null @@ -1,1284 +0,0 @@ -/* - * ======================================================== - * - * cc - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -/// BUGS: 0 -/// TODO: none - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -/* C driver */ -/* This is part of the LibCompiler. */ -/* (c) Amlal El Mahrouss */ - -/// @author EL Mahrouss Amlal (amlel) -/// @file ARM64-cc.cxx -/// @brief ARM64 C Compiler. - -/// TODO: support structures, else if, else, . and -> - -///////////////////// - -// ANSI ESCAPE CODES - -///////////////////// - -#define kExitOK (0) - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" - -///////////////////////////////////// - -// INTERNAL STUFF OF THE C COMPILER - -///////////////////////////////////// - -namespace Detail { -// \brief Register map structure, used to keep track of each variable's registers. -struct CompilerRegisterMap final { - std::string fName; - std::string fReg; -}; - -// \brief Map for C structs -// \author amlel -struct CompilerStructMap final { - // 'my_foo' - std::string fName; - - // if instance: stores a valid register. - std::string fReg; - - // offset count - std::size_t fOffsetsCnt; - - // offset array. - std::vector> fOffsets; -}; - -struct CompilerState final { - std::vector fSyntaxTreeList; - std::vector kStackFrame; - std::vector kStructMap; - LibCompiler::SyntaxLeafList* fSyntaxTree{nullptr}; - std::unique_ptr fOutputAssembly; - std::string fLastFile; - std::string fLastError; - bool fVerbose; -}; -} // namespace Detail - -static Detail::CompilerState kState; -static std::string kIfFunction = ""; - -namespace Detail { -/// @brief prints an error into stdout. -/// @param reason the reason of the error. -/// @param file where does it originate from? -void print_error(std::string reason, std::string file) noexcept; - -struct CompilerType final { - std::string fName; - std::string fValue; -}; -} // namespace Detail - -///////////////////////////////////////////////////////////////////////////////////////// - -// Target architecture. -static int kMachine = 0; - -///////////////////////////////////////// - -// REGISTERS ACCORDING TO USED ASSEMBLER - -///////////////////////////////////////// - -static size_t kRegisterCnt = kAsmRegisterLimit; -static size_t kStartUsable = 8; -static size_t kUsableLimit = 15; -static size_t kRegisterCounter = kStartUsable; -static std::string kRegisterPrefix = kAsmRegisterPrefix; - -///////////////////////////////////////// - -// COMPILER PARSING UTILITIES/STATES. - -///////////////////////////////////////// - -static std::vector kFileList; -static LibCompiler::AssemblyFactory kFactory; -static bool kInStruct = false; -static bool kOnWhileLoop = false; -static bool kOnForLoop = false; -static bool kInBraces = false; -static bool kIfFound = false; -static size_t kBracesCount = 0UL; - -/* @brief C compiler backend for C */ -class CompilerFrontendARM64 final : public LibCompiler::CompilerFrontendInterface { - public: - explicit CompilerFrontendARM64() = default; - ~CompilerFrontendARM64() override = default; - - LIBCOMPILER_COPY_DEFAULT(CompilerFrontendARM64); - - std::string Check(const char* text, const char* file); - bool Compile(std::string text, std::string file) override; - - const char* Language() override { return "64k C"; } -}; - -static CompilerFrontendARM64* kCompilerFrontend = nullptr; -static std::vector kCompilerVariables; -static std::vector kCompilerFunctions; -static std::vector kCompilerTypes; - -namespace Detail { -union number_cast final { - public: - number_cast(UInt64 _Raw) : _Raw(_Raw) {} - - public: - char _Num[8]; - UInt64 _Raw; -}; - -union double_cast final { - public: - double_cast(float _Raw) : _Raw(_Raw) {} - - public: - char _Sign; - char _Lh[8]; - char _Rh[23]; - - float _Raw; -}; -} // namespace Detail - -///////////////////////////////////////////////////////////////////////////////////////// - -// @name Compile -// @brief Generate MASM from a C assignement. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool CompilerFrontendARM64::Compile(std::string text, std::string file) { - bool typeFound = false; - bool fnFound = false; - - // setup generator. - std::random_device rd; - - auto seed_data = std::array{}; - std::generate(std::begin(seed_data), std::end(seed_data), std::ref(rd)); - std::seed_seq seq(std::begin(seed_data), std::end(seed_data)); - std::mt19937 generator(seq); - - // start parsing - for (size_t text_index = 0; text_index < text.size(); ++text_index) { - auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); - - auto gen = uuids::uuid_random_generator{generator}; - uuids::uuid out = gen(); - - Detail::number_cast time_off = (UInt64) out.as_bytes().data(); - - if (!typeFound) { - auto substr = text.substr(text_index); - std::string match_type; - - for (size_t y = 0; y < substr.size(); ++y) { - if (substr[y] == ' ') { - while (match_type.find(' ') != std::string::npos) { - match_type.erase(match_type.find(' ')); - } - - for (auto& clType : kCompilerTypes) { - if (clType.fName == match_type) { - match_type.clear(); - - std::string buf; - - buf += clType.fValue; - buf += ' '; - - if (substr.find('=') != std::string::npos) { - break; - } - - if (text.find('(') != std::string::npos) { - syntaxLeaf.fUserValue = buf; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - } - - typeFound = true; - break; - } - } - - break; - } - - match_type += substr[y]; - } - } - - if (text[text_index] == '{') { - if (kInStruct) { - continue; - } - - kInBraces = true; - ++kBracesCount; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - } - - // return keyword handler - if (text[text_index] == 'r') { - std::string return_keyword; - return_keyword += "return"; - - std::size_t index = 0UL; - - std::string value; - - for (size_t return_index = text_index; return_index < text.size(); ++return_index) { - if (text[return_index] != return_keyword[index]) { - for (size_t value_index = return_index; value_index < text.size(); ++value_index) { - if (text[value_index] == ';') break; - - value += text[value_index]; - } - - break; - } - - ++index; - } - - if (index == return_keyword.size()) { - if (!value.empty()) { - if (value.find('(') != std::string::npos) { - value.erase(value.find('(')); - } - - if (!isdigit(value[value.find('(') + 2])) { - std::string tmp = value; - bool reg_to_reg = false; - - value.clear(); - - value += " extern_segment"; - value += tmp; - } - - syntaxLeaf.fUserValue = "\tldw r19, "; - - // make it pretty. - if (value.find('\t') != std::string::npos) value.erase(value.find('\t'), 1); - - syntaxLeaf.fUserValue += value + "\n"; - } - - syntaxLeaf.fUserValue += "\tjlr"; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - break; - } - } - - if (text[text_index] == 'i' && text[text_index + 1] == 'f') { - auto expr = text.substr(text_index + 2); - text.erase(text_index, 2); - - if (expr.find("{") != std::string::npos) { - expr.erase(expr.find("{")); - } - - if (expr.find("(") != std::string::npos) expr.erase(expr.find("(")); - - if (expr.find(")") != std::string::npos) expr.erase(expr.find(")")); - - kIfFunction = "__LIBCOMPILER_IF_PROC_"; - kIfFunction += std::to_string(time_off._Raw); - - syntaxLeaf.fUserValue = "\tlda r12, extern_segment "; - syntaxLeaf.fUserValue += kIfFunction + - "\n\t#r12 = Code to jump on, r11 right cond, r10 left cond.\n\tbeq " - "r10, r11, r12\ndword public_segment .code64 " + - kIfFunction + "\n"; - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - kIfFound = true; - } - - // Parse expressions and instructions here. - // what does this mean? - // we encounter an assignment, or we reached the end of an expression. - if (text[text_index] == '=' || text[text_index] == ';') { - if (fnFound) continue; - if (kIfFound) continue; - - if (text[text_index] == ';' && kInStruct) continue; - - if (text.find("typedef ") != std::string::npos) continue; - - if (text[text_index] == '=' && kInStruct) { - Detail::print_error("assignement of value in struct " + text, file); - continue; - } - - if (text[text_index] == ';' && kInStruct) { - bool space_found_ = false; - std::string sym; - - for (auto& ch : text) { - if (ch == ' ') { - space_found_ = true; - } - - if (ch == ';') break; - - if (space_found_) sym.push_back(ch); - } - - kState.kStructMap[kState.kStructMap.size() - 1].fOffsets.push_back( - std::make_pair(kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4, sym)); - - kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt = - kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4; - - continue; - } - - if (text[text_index] == '=' && kInStruct) { - continue; - } - - if (text[text_index + 1] == '=' || text[text_index - 1] == '!' || - text[text_index - 1] == '<' || text[text_index - 1] == '>') { - continue; - } - - std::string substr; - - if (text.find('=') != std::string::npos && kInBraces && !kIfFound) { - if (text.find("*") != std::string::npos) { - if (text.find("=") > text.find("*")) - substr += "\tlda "; - else - substr += "\tldw "; - } else { - substr += "\tldw "; - } - } else if (text.find('=') != std::string::npos && !kInBraces) { - substr += "stw public_segment .data64 "; - } - - int first_encountered = 0; - - std::string str_name; - - for (size_t text_index_2 = 0; text_index_2 < text.size(); ++text_index_2) { - if (text[text_index_2] == '\"') { - ++text_index_2; - - // want to add this, so that the parser recognizes that this is a - // string. - substr += '"'; - - for (; text_index_2 < text.size(); ++text_index_2) { - if (text[text_index_2] == '\"') break; - - substr += text[text_index_2]; - } - } - - if (text[text_index_2] == '{' || text[text_index_2] == '}') continue; - - if (text[text_index_2] == ';') { - break; - } - - if (text[text_index_2] == ' ' || text[text_index_2] == '\t') { - if (first_encountered != 2) { - if (text[text_index] != '=' && - substr.find("public_segment .data64") == std::string::npos && !kInStruct) - substr += "public_segment .data64 "; - } - - ++first_encountered; - - continue; - } - - if (text[text_index_2] == '=') { - if (!kInBraces) { - substr.replace(substr.find("public_segment .data64"), strlen("public_segment .data64"), - "public_segment .zero64 "); - } - - substr += ","; - continue; - } - - substr += text[text_index_2]; - } - - for (auto& clType : kCompilerTypes) { - if (substr.find(clType.fName) != std::string::npos) { - if (substr.find(clType.fName) > substr.find('"')) continue; - - substr.erase(substr.find(clType.fName), clType.fName.size()); - } else if (substr.find(clType.fValue) != std::string::npos) { - if (substr.find(clType.fValue) > substr.find('"')) continue; - - if (clType.fName == "const") continue; - - substr.erase(substr.find(clType.fValue), clType.fValue.size()); - } - } - - if (substr.find("extern") != std::string::npos) { - substr.replace(substr.find("extern"), strlen("extern"), "extern_segment "); - - if (substr.find("public_segment .data64") != std::string::npos) - substr.erase(substr.find("public_segment .data64"), strlen("public_segment .data64")); - } - - auto var_to_find = std::find_if( - kCompilerVariables.cbegin(), kCompilerVariables.cend(), - [&](Detail::CompilerType type) { return type.fName.find(substr) != std::string::npos; }); - - if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; - - std::string reg = kAsmRegisterPrefix; - reg += std::to_string(kRegisterCounter); - - if (var_to_find == kCompilerVariables.cend()) { - ++kRegisterCounter; - - kState.kStackFrame.push_back({.fName = substr, .fReg = reg}); - kCompilerVariables.push_back({.fName = substr}); - } - - syntaxLeaf.fUserValue += substr; - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - if (text[text_index] == '=') break; - } - - // function handler. - - if (text[text_index] == '(' && !fnFound && !kIfFound) { - std::string substr; - std::string args_buffer; - std::string args; - - bool type_crossed = false; - - for (size_t idx = text.find('(') + 1; idx < text.size(); ++idx) { - if (text[idx] == ',') continue; - - if (text[idx] == ' ') continue; - - if (text[idx] == ')') break; - } - - for (char substr_first_index : text) { - if (substr_first_index != ',') - args_buffer += substr_first_index; - else - args_buffer += '$'; - - if (substr_first_index == ';') { - args_buffer = args_buffer.erase(0, args_buffer.find('(')); - args_buffer = args_buffer.erase(args_buffer.find(';'), 1); - args_buffer = args_buffer.erase(args_buffer.find(')'), 1); - args_buffer = args_buffer.erase(args_buffer.find('('), 1); - - if (!args_buffer.empty()) args += "\tldw r6, "; - - std::string register_type; - std::size_t index = 7UL; - - while (args_buffer.find("$") != std::string::npos) { - register_type = kRegisterPrefix; - register_type += std::to_string(index); - - ++index; - - args_buffer.replace(args_buffer.find('$'), 1, "\n\tldw " + register_type + ","); - } - - args += args_buffer; - args += "\n\tlda r19, "; - } - } - - for (char _text_i : text) { - if (_text_i == '\t' || _text_i == ' ') { - if (!type_crossed) { - substr.clear(); - type_crossed = true; - } - - continue; - } - - if (_text_i == '(') break; - - substr += _text_i; - } - - if (kInBraces) { - syntaxLeaf.fUserValue = args; - syntaxLeaf.fUserValue += substr; - syntaxLeaf.fUserValue += "\n\tjrl\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - fnFound = true; - } else { - syntaxLeaf.fUserValue.clear(); - - syntaxLeaf.fUserValue += "public_segment .code64 "; - - syntaxLeaf.fUserValue += substr; - syntaxLeaf.fUserValue += "\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - fnFound = true; - } - - kCompilerFunctions.push_back(text); - } - - if (text[text_index] == '-' && text[text_index + 1] == '-') { - text = text.replace(text.find("--"), strlen("--"), ""); - - for (int _text_i = 0; _text_i < text.size(); ++_text_i) { - if (text[_text_i] == '\t' || text[_text_i] == ' ') text.erase(_text_i, 1); - } - - syntaxLeaf.fUserValue += "sub "; - syntaxLeaf.fUserValue += text; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - break; - } - - if (text[text_index] == '}') { - kRegisterCounter = kStartUsable; - - --kBracesCount; - - if (kBracesCount < 1) { - kInBraces = false; - kBracesCount = 0; - } - - if (kIfFound) kIfFound = false; - - if (kInStruct) kInStruct = false; - - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - } - - syntaxLeaf.fUserValue.clear(); - } - - auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); - syntaxLeaf.fUserValue = "\n"; - kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); - - return true; -} - -static bool kShouldHaveBraces = false; -static std::string kFnName; - -std::string CompilerFrontendARM64::Check(const char* text, const char* file) { - std::string err_str; - std::string ln = text; - - if (ln.empty()) { - return err_str; - } - - bool non_ascii_found = false; - - for (int i = 0; i < ln.size(); ++i) { - if (isalnum(ln[i])) { - non_ascii_found = true; - break; - } - } - - if (kShouldHaveBraces && ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - } - - if (!non_ascii_found) return err_str; - - size_t string_index = 1UL; - - if (ln.find('\'') != std::string::npos) { - string_index = ln.find('\'') + 1; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == '\'') { - if (ln[string_index + 1] != ';') { - ln.erase(string_index, 1); - } - - return err_str; - } - } - } else if (ln.find('"') != std::string::npos) { - string_index = ln.find('"') + 1; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == '"') { - if (ln[string_index + 1] != ';') { - ln.erase(string_index, 1); - } else { - break; - } - } - } - } else if (ln.find('"') == std::string::npos && ln.find('\'') == std::string::npos) { - std::vector forbidden_words; - - forbidden_words.push_back("\\"); - forbidden_words.push_back("?"); - forbidden_words.push_back("@"); - forbidden_words.push_back("~"); - forbidden_words.push_back("::"); - forbidden_words.push_back("--*"); - forbidden_words.push_back("*/"); - - // add them to avoid stupid mistakes. - forbidden_words.push_back("namespace"); - forbidden_words.push_back("class"); - forbidden_words.push_back("extern \"C\""); - - for (auto& forbidden : forbidden_words) { - if (ln.find(forbidden) != std::string::npos) { - err_str += "\nForbidden character detected: "; - err_str += forbidden; - - return err_str; - } - } - } - - struct CompilerVariableRange final { - std::string fBegin; - std::string fEnd; - }; - - const std::vector variables_list = { - {.fBegin = "static ", .fEnd = "="}, {.fBegin = "=", .fEnd = ";"}, - {.fBegin = "if(", .fEnd = "="}, {.fBegin = "if (", .fEnd = "="}, - {.fBegin = "if(", .fEnd = "<"}, {.fBegin = "if (", .fEnd = "<"}, - {.fBegin = "if(", .fEnd = ">"}, {.fBegin = "if (", .fEnd = ">"}, - {.fBegin = "if(", .fEnd = ")"}, {.fBegin = "if (", .fEnd = ")"}, - - {.fBegin = "else(", .fEnd = "="}, {.fBegin = "else (", .fEnd = "="}, - {.fBegin = "else(", .fEnd = "<"}, {.fBegin = "else (", .fEnd = "<"}, - {.fBegin = "else(", .fEnd = ">"}, {.fBegin = "else (", .fEnd = ">"}, - {.fBegin = "else(", .fEnd = ")"}, {.fBegin = "else (", .fEnd = ")"}, - }; - - for (auto& variable : variables_list) { - if (ln.find(variable.fBegin) != std::string::npos) { - string_index = ln.find(variable.fBegin) + variable.fBegin.size(); - - while (ln[string_index] == ' ') ++string_index; - - std::string keyword; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == variable.fEnd[0]) { - std::string varname = ""; - - for (size_t index_keyword = ln.find(' '); ln[index_keyword] != variable.fBegin[0]; - ++index_keyword) { - if (ln[index_keyword] == ' ') { - continue; - } - - if (isdigit(ln[index_keyword])) { - goto cc_next_loop; - } - - varname += ln[index_keyword]; - } - - if (varname.find(' ') != std::string::npos) { - varname.erase(0, varname.find(' ')); - - if (variable.fBegin == "extern") { - varname.erase(0, varname.find(' ')); - } - } - - if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; - - std::string reg = kAsmRegisterPrefix; - reg += std::to_string(kRegisterCounter); - - kCompilerVariables.push_back({.fValue = varname}); - goto cc_check_done; - } - - keyword.push_back(ln[string_index]); - } - - goto cc_next_loop; - - cc_check_done: - - // skip digit value. - if (isdigit(keyword[0]) || keyword[0] == '"') { - goto cc_next_loop; - } - - while (keyword.find(' ') != std::string::npos) keyword.erase(keyword.find(' '), 1); - - for (auto& var : kCompilerVariables) { - if (var.fValue.find(keyword) != std::string::npos) { - err_str.clear(); - goto cc_next; - } - } - - for (auto& fn : kCompilerFunctions) { - if (fn.find(keyword[0]) != std::string::npos) { - auto where_begin = fn.find(keyword[0]); - auto keyword_begin = 0UL; - auto failed = false; - - for (; where_begin < keyword.size(); ++where_begin) { - if (fn[where_begin] == '(' && keyword[keyword_begin] == '(') break; - - if (fn[where_begin] != keyword[keyword_begin]) { - failed = true; - break; - } - - ++keyword_begin; - } - - if (!failed) { - err_str.clear(); - goto cc_next; - } else { - continue; - } - } - } - - cc_error_value: - if (keyword.find("->") != std::string::npos) return err_str; - - if (keyword.find(".") != std::string::npos) return err_str; - - if (isalnum(keyword[0])) err_str += "\nUndefined value: " + keyword; - - return err_str; - } - - cc_next_loop: - continue; - } - -cc_next: - - // extern does not declare anything, it extern_segments a variable. - // so that's why it's not declare upper. - if (LibCompiler::find_word(ln, "extern")) { - auto substr = ln.substr(ln.find("extern") + strlen("extern")); - kCompilerVariables.push_back({.fValue = substr}); - } - - if (kShouldHaveBraces && ln.find('{') == std::string::npos) { - err_str += "Missing '{' for function "; - err_str += kFnName; - err_str += "\n"; - - kShouldHaveBraces = false; - kFnName.clear(); - } else if (kShouldHaveBraces && ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - kFnName.clear(); - } - - bool type_not_found = true; - - if (ln.find('\'') != std::string::npos) { - ln.replace(ln.find('\''), 3, "0"); - } - - auto first = ln.find('"'); - if (first != std::string::npos) { - auto second = 0UL; - bool found_second_quote = false; - - for (size_t i = first + 1; i < ln.size(); ++i) { - if (ln[i] == '\"') { - found_second_quote = true; - second = i; - - break; - } - } - - if (!found_second_quote) { - err_str += "Missing terminating \"."; - err_str += " here -> " + ln.substr(ln.find('"'), second); - } - } - - if (ln.find(')') != std::string::npos && ln.find(';') == std::string::npos) { - if (ln.find('{') == std::string::npos) { - kFnName = ln; - kShouldHaveBraces = true; - - goto skip_braces_check; - } else if (ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - } - } - -skip_braces_check: - - for (auto& key : kCompilerTypes) { - if (LibCompiler::find_word(ln, key.fName)) { - if (isdigit(ln[ln.find(key.fName) + key.fName.size() + 1])) { - err_str += "\nNumber cannot be set for "; - err_str += key.fName; - err_str += "'s name. here -> "; - err_str += ln; - } - - if (ln.find(key.fName) == 0 || ln[ln.find(key.fName) - 1] == ' ' || - ln[ln.find(key.fName) - 1] == '\t') { - type_not_found = false; - - if (ln[ln.find(key.fName) + key.fName.size()] != ' ') { - type_not_found = true; - - if (ln[ln.find(key.fName) + key.fName.size()] == '\t') type_not_found = false; - - goto next; - } else if (ln[ln.find(key.fName) + key.fName.size()] != '\t') { - type_not_found = true; - - if (ln[ln.find(key.fName) + key.fName.size()] == ' ') type_not_found = false; - } - } - - next: - - if (ln.find(';') == std::string::npos) { - if (ln.find('(') != std::string::npos) { - if (ln.find('=') == std::string::npos) continue; - } - - err_str += "\nMissing ';', here -> "; - err_str += ln; - } else { - continue; - } - - if (ln.find('=') != std::string::npos) { - if (ln.find('(') != std::string::npos) { - if (ln.find(')') == std::string::npos) { - err_str += "\nMissing ')', after '(' here -> "; - err_str += ln.substr(ln.find('(')); - } - } - } - } - } - - if (kInBraces && ln.find("struct") != std::string::npos && - ln.find("union") != std::string::npos && ln.find("enum") != std::string::npos && - ln.find('=') != std::string::npos) { - if (ln.find(';') == std::string::npos) { - err_str += "\nMissing ';' after struct/union/enum declaration, here -> "; - err_str += ln; - } - } - - if (ln.find(';') != std::string::npos && ln.find("for") == std::string::npos) { - if (ln.find(';') + 1 != ln.size()) { - for (int i = 0; i < ln.substr(ln.find(';') + 1).size(); ++i) { - if ((ln.substr(ln.find(';') + 1)[i] != ' ') || (ln.substr(ln.find(';') + 1)[i] != '\t')) { - if (auto err = this->Check(ln.substr(ln.find(';') + 1).c_str(), file); !err.empty()) { - err_str += "\nUnexpected text after ';' -> "; - err_str += ln.substr(ln.find(';')); - err_str += err; - } - } - } - } - } - - if (ln.find('(') != std::string::npos) { - if (ln.find(';') == std::string::npos && !LibCompiler::find_word(ln, "|") && - !LibCompiler::find_word(ln, "||") && !LibCompiler::find_word(ln, "&") && - !LibCompiler::find_word(ln, "&&") && !LibCompiler::find_word(ln, "~")) { - bool found_func = false; - size_t i = ln.find('('); - std::vector opens; - std::vector closes; - - for (; i < ln.size(); ++i) { - if (ln[i] == ')') { - closes.push_back(1); - } - - if (ln[i] == '(') { - opens.push_back(1); - } - } - - if (closes.size() != opens.size()) err_str += "Unterminated (), here -> " + ln; - - bool space_found = false; - - for (int i = 0; i < ln.size(); ++i) { - if (ln[i] == ')' && !space_found) { - space_found = true; - continue; - } - - if (space_found) { - if (ln[i] == ' ' && isalnum(ln[i + 1])) { - err_str += "\nBad function format here -> "; - err_str += ln; - } - } - } - } - - if (ln.find('(') < 1) { - err_str += "\nMissing identifier before '(' here -> "; - err_str += ln; - } else { - if (type_not_found && ln.find(';') == std::string::npos && - ln.find("if") == std::string::npos && ln.find("|") == std::string::npos && - ln.find("&") == std::string::npos && ln.find("(") == std::string::npos && - ln.find(")") == std::string::npos) { - err_str += "\n Missing ';' or type, here -> "; - err_str += ln; - } - } - - if (ln.find(')') == std::string::npos) { - err_str += "\nMissing ')', after '(' here -> "; - err_str += ln.substr(ln.find('(')); - } - } else { - if (ln.find("for") != std::string::npos || ln.find("while") != std::string::npos) { - err_str += "\nMissing '(', after \"for\", here -> "; - err_str += ln; - } - } - - if (ln.find('}') != std::string::npos && !kInBraces) { - if (!kInStruct && ln.find(';') == std::string::npos) { - err_str += "\nMismatched '}', here -> "; - err_str += ln; - } - } - - if (!ln.empty()) { - if (ln.find(';') == std::string::npos && ln.find('{') == std::string::npos && - ln.find('}') == std::string::npos && ln.find(')') == std::string::npos && - ln.find('(') == std::string::npos && ln.find(',') == std::string::npos) { - if (ln.size() <= 2) return err_str; - - err_str += "\nMissing ';', here -> "; - err_str += ln; - } - } - - return err_str; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -/** - * @brief C To Assembly mount-point. - */ - -///////////////////////////////////////////////////////////////////////////////////////// - -class AssemblyCCInterface final LC_ASSEMBLY_INTERFACE { - public: - explicit AssemblyCCInterface() = default; - ~AssemblyCCInterface() override = default; - - LIBCOMPILER_COPY_DEFAULT(AssemblyCCInterface); - - UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArchAARCH64; } - - Int32 CompileToFormat(std::string src, Int32 arch) override { - if (kCompilerFrontend == nullptr) return 1; - - /* @brief copy contents wihtout extension */ - std::string src_file = src.data(); - std::ifstream src_fp = std::ifstream(src_file, std::ios::in); - std::string dest; - - for (auto& ch : src_file) { - if (ch == '.') { - break; - } - - dest += ch; - } - - /* According to PEF ABI. */ - std::vector exts = kAsmFileExts; - dest += exts[4]; - - kState.fOutputAssembly = std::make_unique(dest); - - auto fmt = LibCompiler::current_date(); - - (*kState.fOutputAssembly) << "# Path: " << src_file << "\n"; - (*kState.fOutputAssembly) << "# Language: ARM64 Assembly (Generated from ANSI C)\n"; - (*kState.fOutputAssembly) << "# Date: " << fmt << "\n\n"; - - LibCompiler::SyntaxLeafList syntax; - - kState.fSyntaxTreeList.push_back(syntax); - kState.fSyntaxTree = &kState.fSyntaxTreeList[kState.fSyntaxTreeList.size() - 1]; - - std::string line_src; - - while (std::getline(src_fp, line_src)) { - if (auto err = kCompilerFrontend->Check(line_src.c_str(), src.data()); err.empty()) { - kCompilerFrontend->Compile(line_src, src.data()); - } else { - Detail::print_error(err, src.data()); - } - } - - if (kAcceptableErrors > 0) return 1; - - std::vector keywords = {"ldw", "stw", "lda", "sta", "add", "sub", "mv"}; - - /// - /// Replace, optimize, fix assembly output. - /// - - for (auto& leaf : kState.fSyntaxTree->fLeafList) { - std::vector access_keywords = {"->", "."}; - - for (auto& access_ident : access_keywords) { - if (LibCompiler::find_word(leaf.fUserValue, access_ident)) { - for (auto& struc : kState.kStructMap) { - /// TODO: - } - } - } - - for (auto& keyword : keywords) { - if (LibCompiler::find_word(leaf.fUserValue, keyword)) { - std::size_t cnt = 0UL; - - for (auto& reg : kState.kStackFrame) { - std::string needle; - - for (size_t i = 0; i < reg.fName.size(); i++) { - if (reg.fName[i] == ' ') { - ++i; - - for (; i < reg.fName.size(); i++) { - if (reg.fName[i] == ',') { - break; - } - - if (reg.fName[i] == ' ') continue; - - needle += reg.fName[i]; - } - - break; - } - } - - if (LibCompiler::find_word(leaf.fUserValue, needle)) { - if (leaf.fUserValue.find("extern_segment " + needle) != std::string::npos) { - std::string range = "extern_segment " + needle; - leaf.fUserValue.replace(leaf.fUserValue.find("extern_segment " + needle), - range.size(), needle); - } - - if (leaf.fUserValue.find("ldw r6") != std::string::npos) { - std::string::difference_type countComma = - std::count(leaf.fUserValue.begin(), leaf.fUserValue.end(), ','); - - if (countComma == 1) { - leaf.fUserValue.replace(leaf.fUserValue.find("ldw"), strlen("ldw"), "mv"); - } - } - - leaf.fUserValue.replace(leaf.fUserValue.find(needle), needle.size(), reg.fReg); - - ++cnt; - } - } - - if (cnt > 1 && keyword != "mv" && keyword != "add" && keyword != "sub") { - leaf.fUserValue.replace(leaf.fUserValue.find(keyword), keyword.size(), "mv"); - } - } - } - } - - for (auto& leaf : kState.fSyntaxTree->fLeafList) { - (*kState.fOutputAssembly) << leaf.fUserValue; - } - - kState.fSyntaxTree = nullptr; - - kState.fOutputAssembly->flush(); - kState.fOutputAssembly.reset(); - - return kExitOK; - } -}; - -///////////////////////////////////////////////////////////////////////////////////////// - -#include - -#define kPrintF printf -#define kSplashCxx() kPrintF(kWhite "NE C Driver, %s, (c) Amlal El Mahrouss\n", kDistVersion) - -static void cc_print_help() { - kSplashCxx(); -} - -///////////////////////////////////////////////////////////////////////////////////////// - -#define kCExtension ".c" - -LIBCOMPILER_MODULE(NeOSCompilerCLangARM64) { - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - kCompilerTypes.push_back({.fName = "void", .fValue = "void"}); - kCompilerTypes.push_back({.fName = "char", .fValue = "byte"}); - kCompilerTypes.push_back({.fName = "short", .fValue = "hword"}); - kCompilerTypes.push_back({.fName = "int", .fValue = "dword"}); - kCompilerTypes.push_back({.fName = "long", .fValue = "qword"}); - kCompilerTypes.push_back({.fName = "*", .fValue = "offset"}); - - bool skip = false; - - kFactory.Mount(new AssemblyCCInterface()); - kMachine = LibCompiler::AssemblyFactory::kArchAARCH64; - kCompilerFrontend = new CompilerFrontendARM64(); - - for (auto index = 1UL; index < argc; ++index) { - if (skip) { - skip = false; - continue; - } - - if (argv[index][0] == '-') { - if (strcmp(argv[index], "--v") == 0 || strcmp(argv[index], "--version") == 0) { - kSplashCxx(); - return kExitOK; - } - - if (strcmp(argv[index], "--verbose") == 0) { - kState.fVerbose = true; - - continue; - } - - if (strcmp(argv[index], "--h") == 0 || strcmp(argv[index], "--help") == 0) { - cc_print_help(); - - return kExitOK; - } - - if (strcmp(argv[index], "--dialect") == 0) { - if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; - - return kExitOK; - } - - if (strcmp(argv[index], "--fmax-exceptions") == 0) { - try { - kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); - } - // catch anything here - catch (...) { - kErrorLimit = 0; - } - - skip = true; - - continue; - } - - std::string err = "Unknown command: "; - err += argv[index]; - - Detail::print_error(err, "cc"); - - continue; - } - - kFileList.emplace_back(argv[index]); - - std::string srcFile = argv[index]; - - if (strstr(argv[index], kCExtension) == nullptr) { - if (kState.fVerbose) { - Detail::print_error(srcFile + " is not a valid C source.\n", "cc"); - } - - return 1; - } - - if (kFactory.Compile(srcFile, kMachine) != kExitOK) return 1; - } - - return kExitOK; -} - -// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/CCompilerPower64.cc b/dev/LibCompiler/src/CCompilerPower64.cc deleted file mode 100644 index 91cd32f..0000000 --- a/dev/LibCompiler/src/CCompilerPower64.cc +++ /dev/null @@ -1,1303 +0,0 @@ -/* - * ======================================================== - * - * CompilerPower64 - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#define kExitOK 0 - -/// @author EL Mahrouss Amlal (amlal@nekernel.org) -/// @file cc.cxx -/// @brief POWER64 C Compiler. - -///////////////////// - -/// ANSI ESCAPE CODES - -///////////////////// - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" - -///////////////////////////////////// - -/// INTERNAL STRUCTURES OF THE C COMPILER - -///////////////////////////////////// - -namespace Detail { -// \brief name to register struct. -struct CompilerRegisterMap final { - std::string fName; - std::string fReg; -}; - -// \brief Map for C structs -// \author amlel -struct CompilerStructMap final { - /// 'struct::my_foo' - std::string fName; - - /// if instance: stores a valid register. - std::string fReg; - - /// offset count - std::size_t fOffsetsCnt; - - /// offset array. - std::vector> fOffsets; -}; - -struct CompilerState final { - std::vector fSyntaxTreeList; - std::vector kStackFrame; - std::vector kStructMap; - LibCompiler::SyntaxLeafList* fSyntaxTree{nullptr}; - std::unique_ptr fOutputAssembly; - std::string fLastFile; - std::string fLastError; - bool fVerbose; -}; -} // namespace Detail - -static Detail::CompilerState kState; -static std::string kIfFunction = ""; - -namespace Detail { -/// @brief prints an error into stdout. -/// @param reason the reason of the error. -/// @param file where does it originate from? -void print_error(std::string reason, std::string file) noexcept; - -struct CompilerType final { - std::string fName; - std::string fValue; -}; -} // namespace Detail - -///////////////////////////////////////////////////////////////////////////////////////// - -// Target architecture. -static int kMachine = 0; - -///////////////////////////////////////// - -// REGISTERS ACCORDING TO USED ASSEMBLER - -///////////////////////////////////////// - -static size_t kRegisterCnt = kAsmRegisterLimit; -static size_t kStartUsable = 2; -static size_t kUsableLimit = 15; -static size_t kRegisterCounter = kStartUsable; -static std::string kRegisterPrefix = kAsmRegisterPrefix; - -///////////////////////////////////////// - -// COMPILER PARSING UTILITIES/STATES. - -///////////////////////////////////////// - -static std::vector kFileList; -static LibCompiler::AssemblyFactory kFactory; -static bool kInStruct = false; -static bool kOnWhileLoop = false; -static bool kOnForLoop = false; -static bool kInBraces = false; -static bool kIfFound = false; -static size_t kBracesCount = 0UL; - -/* @brief C compiler backend for C */ -class CompilerFrontendPower64 final : public LibCompiler::CompilerFrontendInterface { - public: - explicit CompilerFrontendPower64() = default; - ~CompilerFrontendPower64() override = default; - - LIBCOMPILER_COPY_DEFAULT(CompilerFrontendPower64); - - std::string Check(const char* text, const char* file); - bool Compile(std::string text, std::string file) override; - - const char* Language() override { return "POWER C"; } -}; - -static CompilerFrontendPower64* kCompilerFrontend = nullptr; -static std::vector kCompilerVariables; -static std::vector kCompilerFunctions; -static std::vector kCompilerTypes; - -namespace Detail { -union number_cast final { - public: - number_cast(UInt64 _Raw) : _Raw(_Raw) {} - - public: - char _Num[8]; - UInt64 _Raw; -}; - -union double_cast final { - public: - double_cast(float _Raw) : _Raw(_Raw) {} - - public: - char _Sign; - char _Lh[8]; - char _Rh[23]; - - float _Raw; -}; -} // namespace Detail - -///////////////////////////////////////////////////////////////////////////////////////// - -// @name Compile -// @brief Generate MASM from a C assignement. - -///////////////////////////////////////////////////////////////////////////////////////// - -bool CompilerFrontendPower64::Compile(std::string text_, std::string file) { - std::string text = text_; - - bool typeFound = false; - bool fnFound = false; - - // setup generator. - std::random_device rd; - - auto seed_data = std::array{}; - std::generate(std::begin(seed_data), std::end(seed_data), std::ref(rd)); - std::seed_seq seq(std::begin(seed_data), std::end(seed_data)); - std::mt19937 generator(seq); - - // start parsing - for (size_t text_index = 0; text_index < text.size(); ++text_index) { - auto syntax_leaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); - - auto gen = uuids::uuid_random_generator{generator}; - uuids::uuid out = gen(); - - Detail::number_cast time_off = (UInt64) out.as_bytes().data(); - - if (!typeFound) { - auto substr = text.substr(text_index); - std::string match_type; - - for (size_t y = 0; y < substr.size(); ++y) { - if (substr[y] == ' ') { - while (match_type.find(' ') != std::string::npos) { - match_type.erase(match_type.find(' ')); - } - - for (auto& clType : kCompilerTypes) { - if (clType.fName == match_type) { - match_type.clear(); - - std::string buf; - - buf += clType.fValue; - buf += ' '; - - if (substr.find('=') != std::string::npos) { - break; - } - - if (text.find('(') != std::string::npos) { - syntax_leaf.fUserValue = buf; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - } - - typeFound = true; - break; - } - } - - break; - } - - match_type += substr[y]; - } - } - - if (text[text_index] == '{') { - if (kInStruct) { - continue; - } - - kInBraces = true; - ++kBracesCount; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - } - - // return keyword handler - if (text[text_index] == 'r') { - std::string return_keyword; - return_keyword += "return"; - - std::size_t index = 0UL; - - std::string value; - - for (size_t return_index = text_index; return_index < text.size(); ++return_index) { - if (text[return_index] != return_keyword[index]) { - for (size_t value_index = return_index; value_index < text.size(); ++value_index) { - if (text[value_index] == ';') break; - - value += text[value_index]; - } - - break; - } - - ++index; - } - - if (index == return_keyword.size()) { - if (!value.empty()) { - if (value.find('(') != std::string::npos) { - value.erase(value.find('(')); - } - - if (!isdigit(value[value.find('(') + 2])) { - std::string tmp = value; - bool reg_to_reg = false; - - value.clear(); - - value += " extern_segment"; - value += tmp; - } - - syntax_leaf.fUserValue = "\tmr r31, "; - - // make it pretty. - while (value.find('\t') != std::string::npos) value.erase(value.find('\t'), 1); - - while (value.find(' ') != std::string::npos) value.erase(value.find(' '), 1); - - while (value.find("extern_segment") != std::string::npos) - value.erase(value.find("extern_segment"), strlen("extern_segment")); - - bool found = false; - - for (auto& reg : kState.kStackFrame) { - if (value.find(reg.fName) != std::string::npos) { - found = true; - syntax_leaf.fUserValue += reg.fReg; - break; - } - } - - if (!found) syntax_leaf.fUserValue += "r0"; - } - - syntax_leaf.fUserValue += "\n\tblr"; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - - break; - } - } - - if (text[text_index] == 'i' && text[text_index + 1] == 'f') { - auto expr = text.substr(text_index + 2); - text.erase(text_index, 2); - - if (expr.find("{") != std::string::npos) { - expr.erase(expr.find("{")); - } - - if (expr.find("(") != std::string::npos) expr.erase(expr.find("(")); - - if (expr.find(")") != std::string::npos) expr.erase(expr.find(")")); - - kIfFunction = "__LIBCOMPILER_IF_PROC_"; - kIfFunction += std::to_string(time_off._Raw); - - syntax_leaf.fUserValue = - "\tcmpw " - "r10, r11"; - - syntax_leaf.fUserValue += "\n\tbeq extern_segment " + kIfFunction + - " \ndword public_segment .code64 " + kIfFunction + "\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - - kIfFound = true; - } - - // Parse expressions and instructions here. - // what does this mean? - // we encounter an assignment, or we reached the end of an expression. - if (text[text_index] == '=' || text[text_index] == ';') { - if (fnFound) continue; - if (kIfFound) continue; - - if (text[text_index] == ';' && kInStruct) continue; - - if (text.find("typedef ") != std::string::npos) continue; - - if (text[text_index] == '=' && kInStruct) { - Detail::print_error("assignement of value inside a struct " + text, file); - continue; - } - - if (text[text_index] == ';' && kInStruct) { - bool space_found_ = false; - std::string sym; - - for (auto& ch : text) { - if (ch == ' ') { - space_found_ = true; - } - - if (ch == ';') break; - - if (space_found_) sym.push_back(ch); - } - - kState.kStructMap[kState.kStructMap.size() - 1].fOffsets.push_back( - std::make_pair(kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4, sym)); - - kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt = - kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4; - - continue; - } - - if (text[text_index] == '=' && kInStruct) { - continue; - } - - if (text[text_index + 1] == '=' || text[text_index - 1] == '!' || - text[text_index - 1] == '<' || text[text_index - 1] == '>') { - continue; - } - - std::string substr; - - if (text.find('=') != std::string::npos && kInBraces && !kIfFound) { - if (text.find("*") != std::string::npos) { - if (text.find("=") > text.find("*")) - substr += "\tli "; - else - substr += "\tli "; - } else { - substr += "\tli "; - } - } else if (text.find('=') != std::string::npos && !kInBraces) { - substr += "stw public_segment .data64 "; - } - - int first_encountered = 0; - - std::string str_name; - - for (size_t text_index_2 = 0; text_index_2 < text.size(); ++text_index_2) { - if (text[text_index_2] == '\"') { - ++text_index_2; - - // want to add this, so that the parser recognizes that this is a - // string. - substr += '"'; - - for (; text_index_2 < text.size(); ++text_index_2) { - if (text[text_index_2] == '\"') break; - - substr += text[text_index_2]; - } - } - - if (text[text_index_2] == '{' || text[text_index_2] == '}') continue; - - if (text[text_index_2] == ';') { - break; - } - - if (text[text_index_2] == ' ' || text[text_index_2] == '\t') { - if (first_encountered != 2) { - if (text[text_index] != '=' && - substr.find("public_segment .data64") == std::string::npos && !kInStruct) - substr += "public_segment .data64 "; - } - - ++first_encountered; - - continue; - } - - if (text[text_index_2] == '=') { - if (!kInBraces) { - substr.replace(substr.find("public_segment .data64"), strlen("public_segment .data64"), - "public_segment .zero64 "); - } - - substr += ","; - continue; - } - - substr += text[text_index_2]; - } - - for (auto& clType : kCompilerTypes) { - if (substr.find(clType.fName) != std::string::npos) { - if (substr.find(clType.fName) > substr.find('"')) continue; - - substr.erase(substr.find(clType.fName), clType.fName.size()); - } else if (substr.find(clType.fValue) != std::string::npos) { - if (substr.find(clType.fValue) > substr.find('"')) continue; - - if (clType.fName == "const") continue; - - substr.erase(substr.find(clType.fValue), clType.fValue.size()); - } - } - - if (substr.find("extern") != std::string::npos) { - substr.replace(substr.find("extern"), strlen("extern"), "extern_segment "); - - if (substr.find("public_segment .data64") != std::string::npos) - substr.erase(substr.find("public_segment .data64"), strlen("public_segment .data64")); - } - - auto var_to_find = std::find_if( - kCompilerVariables.cbegin(), kCompilerVariables.cend(), - [&](Detail::CompilerType type) { return type.fName.find(substr) != std::string::npos; }); - - kCompilerVariables.push_back({.fName = substr}); - - if (text[text_index] == ';') break; - - std::string reg = kAsmRegisterPrefix; - - ++kRegisterCounter; - reg += std::to_string(kRegisterCounter); - - auto newSubstr = substr.substr(substr.find(" ")); - - std::string symbol; - - for (size_t start = 0; start < newSubstr.size(); ++start) { - if (newSubstr[start] == ',') break; - - if (newSubstr[start] == ' ') continue; - - symbol += (newSubstr[start]); - } - - kState.kStackFrame.push_back({.fName = symbol, .fReg = reg}); - - syntax_leaf.fUserValue += "\n\tli " + reg + substr.substr(substr.find(',')); - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - } - - // function handler. - - if (text[text_index] == '(' && !fnFound && !kIfFound) { - std::string substr; - std::string args_buffer; - std::string args; - - bool type_crossed = false; - - for (size_t idx = text.find('(') + 1; idx < text.size(); ++idx) { - if (text[idx] == ',') continue; - - if (text[idx] == ' ') continue; - - if (text[idx] == ')') break; - } - - for (char substr_first_index : text) { - if (substr_first_index != ',') - args_buffer += substr_first_index; - else - args_buffer += '$'; - - if (substr_first_index == ';') { - args_buffer = args_buffer.erase(0, args_buffer.find('(')); - args_buffer = args_buffer.erase(args_buffer.find(';'), 1); - args_buffer = args_buffer.erase(args_buffer.find(')'), 1); - args_buffer = args_buffer.erase(args_buffer.find('('), 1); - - if (!args_buffer.empty()) args += "\tldw r6, "; - - std::string register_type; - std::size_t index = 7UL; - - while (args_buffer.find("$") != std::string::npos) { - register_type = kRegisterPrefix; - register_type += std::to_string(index); - - ++index; - - args_buffer.replace(args_buffer.find('$'), 1, "\n\tldw " + register_type + ","); - } - - args += args_buffer; - args += "\n\tli r31, "; - } - } - - for (char _text_i : text) { - if (_text_i == '\t' || _text_i == ' ') { - if (!type_crossed) { - substr.clear(); - type_crossed = true; - } - - continue; - } - - if (_text_i == '(') break; - - substr += _text_i; - } - - if (kInBraces) { - syntax_leaf.fUserValue = args; - syntax_leaf.fUserValue += substr; - - syntax_leaf.fUserValue += "\n\tblr\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - - fnFound = true; - } else { - syntax_leaf.fUserValue.clear(); - - syntax_leaf.fUserValue += "public_segment .code64 "; - - syntax_leaf.fUserValue += substr; - syntax_leaf.fUserValue += "\n"; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - - fnFound = true; - } - - kCompilerFunctions.push_back(text); - } - - if (text[text_index] == '-' && text[text_index + 1] == '-') { - text = text.replace(text.find("--"), strlen("--"), ""); - - for (int _text_i = 0; _text_i < text.size(); ++_text_i) { - if (text[_text_i] == '\t' || text[_text_i] == ' ') text.erase(_text_i, 1); - } - - syntax_leaf.fUserValue += "dec "; - syntax_leaf.fUserValue += text; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - break; - } - - if (text[text_index] == '}') { - kRegisterCounter = kStartUsable; - - --kBracesCount; - - if (kBracesCount < 1) { - kInBraces = false; - kBracesCount = 0; - } - - if (kIfFound) kIfFound = false; - - if (kInStruct) kInStruct = false; - - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - } - - syntax_leaf.fUserValue.clear(); - } - - auto syntax_leaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); - syntax_leaf.fUserValue = "\n"; - kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); - - return true; -} - -static bool kShouldHaveBraces = false; -static std::string kFnName; - -std::string CompilerFrontendPower64::Check(const char* text, const char* file) { - std::string err_str; - std::string ln = text; - - if (ln.empty()) { - return err_str; - } - - bool non_ascii_found = false; - - for (int i = 0; i < ln.size(); ++i) { - if (isalnum(ln[i])) { - non_ascii_found = true; - break; - } - } - - if (kShouldHaveBraces && ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - } - - if (!non_ascii_found) return err_str; - - size_t string_index = 1UL; - - if (ln.find('\'') != std::string::npos) { - string_index = ln.find('\'') + 1; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == '\'') { - if (ln[string_index + 1] != ';') { - ln.erase(string_index, 1); - } - - return err_str; - } - } - } else if (ln.find('"') != std::string::npos) { - string_index = ln.find('"') + 1; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == '"') { - if (ln[string_index + 1] != ';') { - ln.erase(string_index, 1); - } else { - break; - } - } - } - } else if (ln.find('"') == std::string::npos && ln.find('\'') == std::string::npos) { - std::vector forbidden_words; - - forbidden_words.push_back("\\"); - forbidden_words.push_back("?"); - forbidden_words.push_back("@"); - forbidden_words.push_back("~"); - forbidden_words.push_back("::"); - forbidden_words.push_back("--*"); - forbidden_words.push_back("*/"); - - // add them to avoid stupid mistakes. - forbidden_words.push_back("namespace"); - forbidden_words.push_back("class"); - forbidden_words.push_back("extern \"C\""); - - for (auto& forbidden : forbidden_words) { - if (ln.find(forbidden) != std::string::npos) { - err_str += "\nForbidden character detected: "; - err_str += forbidden; - - return err_str; - } - } - } - - struct CompilerVariableRange final { - std::string fBegin; - std::string fEnd; - }; - - const std::vector variables_list = { - {.fBegin = "static ", .fEnd = "="}, {.fBegin = "=", .fEnd = ";"}, - {.fBegin = "if(", .fEnd = "="}, {.fBegin = "if (", .fEnd = "="}, - {.fBegin = "if(", .fEnd = "<"}, {.fBegin = "if (", .fEnd = "<"}, - {.fBegin = "if(", .fEnd = ">"}, {.fBegin = "if (", .fEnd = ">"}, - {.fBegin = "if(", .fEnd = ")"}, {.fBegin = "if (", .fEnd = ")"}, - - {.fBegin = "else(", .fEnd = "="}, {.fBegin = "else (", .fEnd = "="}, - {.fBegin = "else(", .fEnd = "<"}, {.fBegin = "else (", .fEnd = "<"}, - {.fBegin = "else(", .fEnd = ">"}, {.fBegin = "else (", .fEnd = ">"}, - {.fBegin = "else(", .fEnd = ")"}, {.fBegin = "else (", .fEnd = ")"}, - }; - - for (auto& variable : variables_list) { - if (ln.find(variable.fBegin) != std::string::npos) { - string_index = ln.find(variable.fBegin) + variable.fBegin.size(); - - while (ln[string_index] == ' ') ++string_index; - - std::string keyword; - - for (; string_index < ln.size(); ++string_index) { - if (ln[string_index] == variable.fEnd[0]) { - std::string varname = ""; - - for (size_t index_keyword = ln.find(' '); ln[index_keyword] != variable.fBegin[0]; - ++index_keyword) { - if (ln[index_keyword] == ' ') { - continue; - } - - if (isdigit(ln[index_keyword])) { - goto cc_next_loop; - } - - varname += ln[index_keyword]; - } - - if (varname.find(' ') != std::string::npos) { - varname.erase(0, varname.find(' ')); - - if (variable.fBegin == "extern") { - varname.erase(0, varname.find(' ')); - } - } - - if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; - - std::string reg = kAsmRegisterPrefix; - reg += std::to_string(kRegisterCounter); - - kCompilerVariables.push_back({.fValue = varname}); - goto cc_check_done; - } - - keyword.push_back(ln[string_index]); - } - - goto cc_next_loop; - - cc_check_done: - - // skip digit value. - if (isdigit(keyword[0]) || keyword[0] == '"') { - goto cc_next_loop; - } - - while (keyword.find(' ') != std::string::npos) keyword.erase(keyword.find(' '), 1); - - for (auto& var : kCompilerVariables) { - if (var.fValue.find(keyword) != std::string::npos) { - err_str.clear(); - goto cc_next; - } - } - - for (auto& fn : kCompilerFunctions) { - if (fn.find(keyword[0]) != std::string::npos) { - auto where_begin = fn.find(keyword[0]); - auto keyword_begin = 0UL; - auto failed = false; - - for (; where_begin < keyword.size(); ++where_begin) { - if (fn[where_begin] == '(' && keyword[keyword_begin] == '(') break; - - if (fn[where_begin] != keyword[keyword_begin]) { - failed = true; - break; - } - - ++keyword_begin; - } - - if (!failed) { - err_str.clear(); - goto cc_next; - } else { - continue; - } - } - } - - cc_error_value: - if (keyword.find("->") != std::string::npos) return err_str; - - if (keyword.find(".") != std::string::npos) return err_str; - - if (isalnum(keyword[0])) err_str += "\nUndefined value: " + keyword; - - return err_str; - } - - cc_next_loop: - continue; - } - -cc_next: - - // extern does not declare anything, it extern_segments a variable. - // so that's why it's not declare upper. - if (LibCompiler::find_word(ln, "extern")) { - auto substr = ln.substr(ln.find("extern") + strlen("extern")); - kCompilerVariables.push_back({.fValue = substr}); - } - - if (kShouldHaveBraces && ln.find('{') == std::string::npos) { - err_str += "Missing '{' for function "; - err_str += kFnName; - err_str += "\n"; - - kShouldHaveBraces = false; - kFnName.clear(); - } else if (kShouldHaveBraces && ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - kFnName.clear(); - } - - bool type_not_found = true; - - if (ln.find('\'') != std::string::npos) { - ln.replace(ln.find('\''), 3, "0"); - } - - auto first = ln.find('"'); - if (first != std::string::npos) { - auto second = 0UL; - bool found_second_quote = false; - - for (size_t i = first + 1; i < ln.size(); ++i) { - if (ln[i] == '\"') { - found_second_quote = true; - second = i; - - break; - } - } - - if (!found_second_quote) { - err_str += "Missing terminating \"."; - err_str += " here -> " + ln.substr(ln.find('"'), second); - } - } - - if (ln.find(')') != std::string::npos && ln.find(';') == std::string::npos) { - if (ln.find('{') == std::string::npos) { - kFnName = ln; - kShouldHaveBraces = true; - - goto skip_braces_check; - } else if (ln.find('{') != std::string::npos) { - kShouldHaveBraces = false; - } - } - -skip_braces_check: - - for (auto& key : kCompilerTypes) { - if (LibCompiler::find_word(ln, key.fName)) { - if (isdigit(ln[ln.find(key.fName) + key.fName.size() + 1])) { - err_str += "\nNumber cannot be set for "; - err_str += key.fName; - err_str += "'s name. here -> "; - err_str += ln; - } - - if (ln.find(key.fName) == 0 || ln[ln.find(key.fName) - 1] == ' ' || - ln[ln.find(key.fName) - 1] == '\t') { - type_not_found = false; - - if (ln[ln.find(key.fName) + key.fName.size()] != ' ') { - type_not_found = true; - - if (ln[ln.find(key.fName) + key.fName.size()] == '\t') type_not_found = false; - - goto next; - } else if (ln[ln.find(key.fName) + key.fName.size()] != '\t') { - type_not_found = true; - - if (ln[ln.find(key.fName) + key.fName.size()] == ' ') type_not_found = false; - } - } - - next: - - if (ln.find(';') == std::string::npos) { - if (ln.find('(') != std::string::npos) { - if (ln.find('=') == std::string::npos) continue; - } - - err_str += "\nMissing ';', here -> "; - err_str += ln; - } else { - continue; - } - - if (ln.find('=') != std::string::npos) { - if (ln.find('(') != std::string::npos) { - if (ln.find(')') == std::string::npos) { - err_str += "\nMissing ')', after '(' here -> "; - err_str += ln.substr(ln.find('(')); - } - } - } - } - } - - if (kInBraces && ln.find("struct") != std::string::npos && - ln.find("union") != std::string::npos && ln.find("enum") != std::string::npos && - ln.find('=') != std::string::npos) { - if (ln.find(';') == std::string::npos) { - err_str += "\nMissing ';' after struct/union/enum declaration, here -> "; - err_str += ln; - } - } - - if (ln.find(';') != std::string::npos && ln.find("for") == std::string::npos) { - if (ln.find(';') + 1 != ln.size()) { - for (int i = 0; i < ln.substr(ln.find(';') + 1).size(); ++i) { - if ((ln.substr(ln.find(';') + 1)[i] != ' ') || (ln.substr(ln.find(';') + 1)[i] != '\t')) { - if (auto err = this->Check(ln.substr(ln.find(';') + 1).c_str(), file); !err.empty()) { - err_str += "\nUnexpected text after ';' -> "; - err_str += ln.substr(ln.find(';')); - err_str += err; - } - } - } - } - } - - if (ln.find('(') != std::string::npos) { - if (ln.find(';') == std::string::npos && !LibCompiler::find_word(ln, "|") && - !LibCompiler::find_word(ln, "||") && !LibCompiler::find_word(ln, "&") && - !LibCompiler::find_word(ln, "&&") && !LibCompiler::find_word(ln, "~")) { - bool found_func = false; - size_t i = ln.find('('); - std::vector opens; - std::vector closes; - - for (; i < ln.size(); ++i) { - if (ln[i] == ')') { - closes.push_back(1); - } - - if (ln[i] == '(') { - opens.push_back(1); - } - } - - if (closes.size() != opens.size()) err_str += "Unterminated (), here -> " + ln; - - bool space_found = false; - - for (int i = 0; i < ln.size(); ++i) { - if (ln[i] == ')' && !space_found) { - space_found = true; - continue; - } - - if (space_found) { - if (ln[i] == ' ' && isalnum(ln[i + 1])) { - err_str += "\nBad function format here -> "; - err_str += ln; - } - } - } - } - - if (ln.find('(') < 1) { - err_str += "\nMissing identifier before '(' here -> "; - err_str += ln; - } else { - if (type_not_found && ln.find(';') == std::string::npos && - ln.find("if") == std::string::npos && ln.find("|") == std::string::npos && - ln.find("&") == std::string::npos && ln.find("(") == std::string::npos && - ln.find(")") == std::string::npos) { - err_str += "\n Missing ';' or type, here -> "; - err_str += ln; - } - } - - if (ln.find(')') == std::string::npos) { - err_str += "\nMissing ')', after '(' here -> "; - err_str += ln.substr(ln.find('(')); - } - } else { - if (ln.find("for") != std::string::npos || ln.find("while") != std::string::npos) { - err_str += "\nMissing '(', after \"for\", here -> "; - err_str += ln; - } - } - - if (ln.find('}') != std::string::npos && !kInBraces) { - if (!kInStruct && ln.find(';') == std::string::npos) { - err_str += "\nMismatched '}', here -> "; - err_str += ln; - } - } - - if (!ln.empty()) { - if (ln.find(';') == std::string::npos && ln.find('{') == std::string::npos && - ln.find('}') == std::string::npos && ln.find(')') == std::string::npos && - ln.find('(') == std::string::npos && ln.find(',') == std::string::npos) { - if (ln.size() <= 2) return err_str; - - err_str += "\nMissing ';', here -> "; - err_str += ln; - } - } - - return err_str; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -/** - * @brief C To Assembly mount-point. - */ - -///////////////////////////////////////////////////////////////////////////////////////// - -class AssemblyMountpointCLang final LC_ASSEMBLY_INTERFACE { - public: - explicit AssemblyMountpointCLang() = default; - ~AssemblyMountpointCLang() override = default; - - LIBCOMPILER_COPY_DEFAULT(AssemblyMountpointCLang); - - UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArchPowerPC; } - - Int32 CompileToFormat(std::string src, Int32 arch) override { - if (kCompilerFrontend == nullptr) return 1; - - /* @brief copy contents wihtout extension */ - std::string src_file = src.data(); - std::ifstream src_fp = std::ifstream(src_file, std::ios::in); - std::string dest; - - for (auto& ch : src_file) { - if (ch == '.') { - break; - } - - dest += ch; - } - - /* According to PEF ABI. */ - std::vector exts = kAsmFileExts; - dest += exts[4]; - - kState.fOutputAssembly = std::make_unique(dest); - - auto fmt = LibCompiler::current_date(); - - (*kState.fOutputAssembly) << "# Path: " << src_file << "\n"; - (*kState.fOutputAssembly) << "# Language: POWER Assembly (Generated from C)\n"; - (*kState.fOutputAssembly) << "# Date: " << fmt << "\n\n"; - - LibCompiler::SyntaxLeafList syntax; - - kState.fSyntaxTreeList.push_back(syntax); - kState.fSyntaxTree = &kState.fSyntaxTreeList[kState.fSyntaxTreeList.size() - 1]; - - std::string line_src; - - while (std::getline(src_fp, line_src)) { - if (auto err = kCompilerFrontend->Check(line_src.c_str(), src.data()); err.empty()) { - kCompilerFrontend->Compile(line_src, src.data()); - } else { - Detail::print_error(err, src.data()); - } - } - - if (kAcceptableErrors > 0) return 1; - - std::vector keywords = {"ld", "stw", "add", "sub", "or"}; - - /// - /// Replace, optimize, fix assembly output. - /// - - for (auto& leaf : kState.fSyntaxTree->fLeafList) { - std::vector access_keywords = {"->", "."}; - - for (auto& access_ident : access_keywords) { - if (LibCompiler::find_word(leaf.fUserValue, access_ident)) { - for (auto& struc : kState.kStructMap) { - /// TODO: - } - } - } - - for (auto& keyword : keywords) { - if (LibCompiler::find_word(leaf.fUserValue, keyword)) { - std::size_t cnt = 0UL; - - for (auto& reg : kState.kStackFrame) { - std::string needle; - - for (size_t i = 0; i < reg.fName.size(); i++) { - if (reg.fName[i] == ' ') { - ++i; - - for (; i < reg.fName.size(); i++) { - if (reg.fName[i] == ',') { - break; - } - - if (reg.fName[i] == ' ') continue; - - needle += reg.fName[i]; - } - - break; - } - } - - if (LibCompiler::find_word(leaf.fUserValue, needle)) { - if (leaf.fUserValue.find("extern_segment ") != std::string::npos) { - std::string range = "extern_segment "; - leaf.fUserValue.replace(leaf.fUserValue.find(range), range.size(), ""); - } - - if (leaf.fUserValue.find("ldw r6") != std::string::npos) { - std::string::difference_type countComma = - std::count(leaf.fUserValue.begin(), leaf.fUserValue.end(), ','); - - if (countComma == 1) { - leaf.fUserValue.replace(leaf.fUserValue.find("ldw"), strlen("ldw"), "mr"); - } - } - - leaf.fUserValue.replace(leaf.fUserValue.find(needle), needle.size(), reg.fReg); - - ++cnt; - } - } - - if (cnt > 1 && keyword != "mr" && keyword != "add" && keyword != "dec") { - leaf.fUserValue.replace(leaf.fUserValue.find(keyword), keyword.size(), "mr"); - } - } - } - } - - for (auto& leaf : kState.fSyntaxTree->fLeafList) { - (*kState.fOutputAssembly) << leaf.fUserValue; - } - - kState.fSyntaxTree = nullptr; - - kState.fOutputAssembly->flush(); - kState.fOutputAssembly.reset(); - - return kExitOK; - } -}; - -///////////////////////////////////////////////////////////////////////////////////////// - -#include - -#define kPrintF printf -#define kSplashCxx() kPrintF(kWhite "cc, %s, (c) Amlal El Mahrouss\n", kDistVersion) - -static void cc_print_help() { - kSplashCxx(); -} - -///////////////////////////////////////////////////////////////////////////////////////// - -#define kExt ".c" - -LIBCOMPILER_MODULE(NeOSCompilerCLangPowerPC) { - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - kCompilerTypes.push_back({.fName = "void", .fValue = "void"}); - kCompilerTypes.push_back({.fName = "char", .fValue = "byte"}); - kCompilerTypes.push_back({.fName = "short", .fValue = "hword"}); - kCompilerTypes.push_back({.fName = "int", .fValue = "dword"}); - kCompilerTypes.push_back({.fName = "long", .fValue = "qword"}); - kCompilerTypes.push_back({.fName = "*", .fValue = "offset"}); - - bool skip = false; - - kFactory.Mount(new AssemblyMountpointCLang()); - kMachine = LibCompiler::AssemblyFactory::kArchPowerPC; - kCompilerFrontend = new CompilerFrontendPower64(); - - for (auto index = 1UL; index < argc; ++index) { - if (skip) { - skip = false; - continue; - } - - if (argv[index][0] == '-') { - if (strcmp(argv[index], "-v") == 0 || strcmp(argv[index], "-version") == 0) { - kSplashCxx(); - return kExitOK; - } - - if (strcmp(argv[index], "-verbose") == 0) { - kState.fVerbose = true; - - continue; - } - - if (strcmp(argv[index], "-h") == 0 || strcmp(argv[index], "-help") == 0) { - cc_print_help(); - - return kExitOK; - } - - if (strcmp(argv[index], "-dialect") == 0) { - if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; - - return kExitOK; - } - - if (strcmp(argv[index], "-fmax-exceptions") == 0) { - try { - kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); - } - // catch anything here - catch (...) { - kErrorLimit = 0; - } - - skip = true; - - continue; - } - - std::string err = "Unknown command: "; - err += argv[index]; - - Detail::print_error(err, "cc"); - - continue; - } - - kFileList.emplace_back(argv[index]); - - std::string srcFile = argv[index]; - - if (strstr(argv[index], kExt) == nullptr) { - if (kState.fVerbose) { - Detail::print_error(srcFile + " is not a valid C source.\n", "cc"); - } - - return 1; - } - - if (kFactory.Compile(srcFile, kMachine) != kExitOK) return 1; - } - - return kExitOK; -} - -// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/CPlusPlusCompilerAMD64.cc b/dev/LibCompiler/src/CPlusPlusCompilerAMD64.cc deleted file mode 100644 index cc6fa67..0000000 --- a/dev/LibCompiler/src/CPlusPlusCompilerAMD64.cc +++ /dev/null @@ -1,921 +0,0 @@ -/* - * ======================================================== - * - * cxxdrv - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -/// BUGS: 1 - -#define kPrintF printf - -#define kExitOK (EXIT_SUCCESS) -#define kExitNO (EXIT_FAILURE) - -// extern_segment, @autodelete { ... }, fn foo() -> auto { ... } - -#include -#include -#include -#include - -/* NeKernel C++ Compiler Driver */ -/* This is part of the LibCompiler. */ -/* (c) Amlal El Mahrouss */ - -/// @author EL Mahrouss Amlal (amlal@nekernel.org) -/// @file CPlusPlusCompilerAMD64.cxx -/// @brief Optimized C++ Compiler Driver. -/// @todo Throw error for scoped inside scoped variables when they get referenced outside. -/// @todo Add class/struct/enum support. - -/////////////////////// - -// ANSI ESCAPE CODES // - -/////////////////////// - -#define kBlank "\e[0;30m" -#define kRed "\e[0;31m" -#define kWhite "\e[0;97m" - -///////////////////////////////////// - -// INTERNALS OF THE C++ COMPILER - -///////////////////////////////////// - -/// @internal -namespace Detail { -std::filesystem::path expand_home(const std::filesystem::path& p) { - if (!p.empty() && p.string()[0] == '~') { - const char* home = std::getenv("HOME"); // For Unix-like systems - - if (!home) { - home = std::getenv("USERPROFILE"); // For Windows - } - - if (home) { - return std::filesystem::path(home) / p.relative_path().string().substr(1); - } else { - throw std::runtime_error("Home directory not found in environment variables"); - } - } - return p; -} - -struct CompilerRegisterMap final { - std::string fName; - std::string fReg; -}; - -// \brief Offset based struct/class -struct CompilerStructMap final { - std::string fName; - std::string fReg; - - // offset counter - std::size_t fOffsetsCnt; - - // offset array - std::vector> fOffsets; -}; - -struct CompilerState final { - std::vector fStackMapVector; - std::vector fStructMapVector; - std::string fOutputValue; - std::string fLastFile; - std::string fLastError; -}; -} // namespace Detail - -static Detail::CompilerState kState; - -static Int32 kOnClassScope = 0; - -///////////////////////////////////////////////////////////////////////////////////////// - -// Target architecture. -static int kMachine = LibCompiler::AssemblyFactory::kArchAMD64; - -///////////////////////////////////////// - -// ARGUMENTS REGISTERS (R8, R15) - -///////////////////////////////////////// - -static std::vector kKeywords; - -///////////////////////////////////////// - -// COMPILER PARSING UTILITIES/STATES. - -///////////////////////////////////////// - -static LibCompiler::AssemblyFactory kFactory; -static Boolean kInStruct = false; -static Boolean kOnWhileLoop = false; -static Boolean kOnForLoop = false; -static Boolean kInBraces = false; -static size_t kBracesCount = 0UL; - -/* @brief C++ compiler backend for the NeKernel C++ driver */ -class CompilerFrontendCPlusPlus final : public LibCompiler::CompilerFrontendInterface { - public: - explicit CompilerFrontendCPlusPlus() = default; - ~CompilerFrontendCPlusPlus() override = default; - - LIBCOMPILER_COPY_DEFAULT(CompilerFrontendCPlusPlus); - - Boolean Compile(const std::string text, std::string file) override; - - const char* Language() override; -}; - -/// @internal compiler variables - -static CompilerFrontendCPlusPlus* kCompilerFrontend = nullptr; - -static std::vector kRegisterMap; - -static std::vector kRegisterList = { - "rbx", "rsi", "r10", "r11", "r12", "r13", "r14", "r15", "xmm12", "xmm13", "xmm14", "xmm15", -}; - -/// @brief The PEF calling convention (caller must save rax, rbp) -/// @note callee must return via **rax**. -static std::vector kRegisterConventionCallList = { - "r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15", -}; - -static std::size_t kFunctionEmbedLevel = 0UL; - -/// detail namespaces - -const char* CompilerFrontendCPlusPlus::Language() { - return "AMD64 C++"; -} - -static std::uintptr_t kOrigin = 0x1000000; -static std::vector> kOriginMap; - -///////////////////////////////////////////////////////////////////////////////////////// - -/// @name Compile -/// @brief Generate assembly from a C++ source. - -///////////////////////////////////////////////////////////////////////////////////////// - -Boolean CompilerFrontendCPlusPlus::Compile(std::string text, std::string file) { - if (text.length() < 1) return false; - - std::size_t index = 0UL; - std::vector> keywords_list; - - for (auto& keyword : kKeywords) { - if (text.find(keyword.keyword_name) != std::string::npos) { - switch (keyword.keyword_kind) { - case LibCompiler::kKeywordKindCommentInline: { - break; - } - default: - break; - } - - std::size_t pos = text.find(keyword.keyword_name); - if (pos == std::string::npos) continue; - - // Safe guard: can't go before start of string - if (pos > 0 && text[pos - 1] == '+' && - keyword.keyword_kind == LibCompiler::kKeywordKindVariableAssign) - continue; - - if (pos > 0 && text[pos - 1] == '-' && - keyword.keyword_kind == LibCompiler::kKeywordKindVariableAssign) - continue; - - // Safe guard: don't go out of range - if ((pos + keyword.keyword_name.size()) < text.size() && - text[pos + keyword.keyword_name.size()] == '=' && - keyword.keyword_kind == LibCompiler::kKeywordKindVariableAssign) - continue; - - keywords_list.emplace_back(std::make_pair(keyword, index)); - ++index; - } - } - - LibCompiler::SyntaxLeafList::SyntaxLeaf syntax_tree; - - for (auto& keyword : keywords_list) { - if (text.find(keyword.first.keyword_name) == std::string::npos) continue; - - switch (keyword.first.keyword_kind) { - case LibCompiler::KeywordKind::kKeywordKindClass: { - ++kOnClassScope; - break; - } - case LibCompiler::KeywordKind::kKeywordKindIf: { - std::size_t keywordPos = text.find(keyword.first.keyword_name); - std::size_t openParen = text.find("(", keywordPos); - std::size_t closeParen = text.find(")", openParen); - - if (keywordPos == std::string::npos || openParen == std::string::npos || - closeParen == std::string::npos || closeParen <= openParen) { - Detail::print_error("Malformed if expression: " + text, file); - return false; - } - - auto expr = text.substr(openParen + 1, closeParen - openParen - 1); - - if (expr.find(">=") != std::string::npos) { - auto left = text.substr( - text.find(keyword.first.keyword_name) + keyword.first.keyword_name.size() + 2, - expr.find("<=") + strlen("<=")); - auto right = text.substr(expr.find(">=") + strlen(">="), text.find(")") - 1); - - size_t i = right.size() - 1; - - if (i < 1) break; - - try { - while (!std::isalnum(right[i])) { - right.erase(i, 1); - --i; - } - - right.erase(0, i); - } catch (...) { - right.erase(0, i); - } - - i = left.size() - 1; - try { - while (!std::isalnum(left[i])) { - left.erase(i, 1); - --i; - } - - left.erase(0, i); - } catch (...) { - left.erase(0, i); - } - - if (!isdigit(left[0]) || !isdigit(right[0])) { - auto indexRight = 0UL; - - auto& valueOfVar = !isdigit(left[0]) ? left : right; - - if (!valueOfVar.empty()) { - for (auto pairRight : kRegisterMap) { - ++indexRight; - - if (pairRight != valueOfVar) { - auto& valueOfVarOpposite = isdigit(left[0]) ? left : right; - - syntax_tree.fUserValue += - "mov " + kRegisterList[indexRight + 1] + ", " + valueOfVarOpposite + "\n"; - syntax_tree.fUserValue += "cmp " + kRegisterList[kRegisterMap.size() - 1] + "," + - kRegisterList[indexRight + 1] + "\n"; - - goto done_iterarting_on_if; - } - - auto& valueOfVarOpposite = isdigit(left[0]) ? left : right; - - syntax_tree.fUserValue += - "mov " + kRegisterList[indexRight + 1] + ", " + valueOfVarOpposite + "\n"; - syntax_tree.fUserValue += "cmp " + kRegisterList[kRegisterMap.size() - 1] + ", " + - kRegisterList[indexRight + 1] + "\n"; - - break; - } - } - } - - done_iterarting_on_if: - - std::string fnName = text; - fnName.erase(fnName.find(keyword.first.keyword_name)); - - for (auto& ch : fnName) { - if (ch == ' ') ch = '_'; - } - - syntax_tree.fUserValue += - "jge __OFFSET_ON_TRUE_LC\nsegment .code64 __OFFSET_ON_TRUE_LC:\n"; - } - - break; - } - case LibCompiler::KeywordKind::kKeywordKindFunctionStart: { - for (auto& ch : text) { - if (isdigit(ch)) { - goto dont_accept; - } - } - - goto accept; - - dont_accept: - return false; - - accept: - std::string fnName = text; - size_t indexFnName = 0; - - // this one is for the type. - for (auto& ch : text) { - ++indexFnName; - - if (ch == '\t') break; - if (ch == ' ') break; - } - - fnName = text.substr(indexFnName); - - if (text.find("return ") != std::string::npos) { - text.erase(0, text.find("return ")); - break; - } - - if (text.ends_with(";") && text.find("return") == std::string::npos) - goto lc_write_assembly; - else if (text.size() <= indexFnName) - Detail::print_error("Invalid function name: " + fnName, file); - - indexFnName = 0; - - for (auto& ch : fnName) { - if (ch == ' ' || ch == '\t') { - if (fnName[indexFnName - 1] != ')') - Detail::print_error("Invalid function name: " + fnName, file); - } - - ++indexFnName; - } - - if (fnName.find("(") != LibCompiler::String::npos) { - fnName.erase(fnName.find("(")); - } - - syntax_tree.fUserValue = "public_segment .code64 __LIBCOMPILER_" + fnName + "\n"; - ++kFunctionEmbedLevel; - - kOriginMap.push_back({"__LIBCOMPILER_" + fnName, kOrigin}); - - break; - - lc_write_assembly: - auto it = std::find_if(kOriginMap.begin(), kOriginMap.end(), - [&fnName](std::pair pair) -> bool { - return fnName == pair.first; - }); - - if (it != kOriginMap.end()) { - std::stringstream ss; - ss << std::hex << it->second; - - syntax_tree.fUserValue = "jmp " + ss.str() + "\n"; - kOrigin += 1UL; - } - } - case LibCompiler::KeywordKind::kKeywordKindFunctionEnd: { - if (kOnClassScope) --kOnClassScope; - - if (text.ends_with(";")) break; - - --kFunctionEmbedLevel; - - if (kRegisterMap.size() > kRegisterList.size()) { - --kFunctionEmbedLevel; - } - - if (kFunctionEmbedLevel < 1) kRegisterMap.clear(); - - break; - } - case LibCompiler::KeywordKind::kKeywordKindEndInstr: - case LibCompiler::KeywordKind::kKeywordKindVariableInc: - case LibCompiler::KeywordKind::kKeywordKindVariableDec: - case LibCompiler::KeywordKind::kKeywordKindVariableAssign: { - std::string valueOfVar = ""; - - if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindVariableInc) { - valueOfVar = text.substr(text.find("+=") + 2); - } else if (keyword.first.keyword_kind == - LibCompiler::KeywordKind::kKeywordKindVariableDec) { - valueOfVar = text.substr(text.find("-=") + 2); - } else if (keyword.first.keyword_kind == - LibCompiler::KeywordKind::kKeywordKindVariableAssign) { - valueOfVar = text.substr(text.find("=") + 1); - } else if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindEndInstr) { - break; - } - - while (valueOfVar.find(";") != std::string::npos && - keyword.first.keyword_kind != LibCompiler::KeywordKind::kKeywordKindEndInstr) { - valueOfVar.erase(valueOfVar.find(";")); - } - - std::string varName = text; - - if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindVariableInc) { - varName.erase(varName.find("+=")); - } else if (keyword.first.keyword_kind == - LibCompiler::KeywordKind::kKeywordKindVariableDec) { - varName.erase(varName.find("-=")); - } else if (keyword.first.keyword_kind == - LibCompiler::KeywordKind::kKeywordKindVariableAssign) { - varName.erase(varName.find("=")); - } else if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindEndInstr) { - varName.erase(varName.find(";")); - } - - static Boolean typeFound = false; - - for (auto& keyword : kKeywords) { - if (keyword.keyword_kind == LibCompiler::kKeywordKindType) { - if (text.find(keyword.keyword_name) != std::string::npos) { - if (text[text.find(keyword.keyword_name)] == ' ') { - typeFound = false; - continue; - } - - typeFound = true; - } - } - } - - std::string instr = "mov "; - - if (typeFound && - keyword.first.keyword_kind != LibCompiler::KeywordKind::kKeywordKindVariableInc && - keyword.first.keyword_kind != LibCompiler::KeywordKind::kKeywordKindVariableDec) { - if (kRegisterMap.size() > kRegisterList.size()) { - ++kFunctionEmbedLevel; - } - - while (varName.find(" ") != std::string::npos) { - varName.erase(varName.find(" "), 1); - } - - while (varName.find("\t") != std::string::npos) { - varName.erase(varName.find("\t"), 1); - } - - for (size_t i = 0; !isalnum(valueOfVar[i]); i++) { - if (i > valueOfVar.size()) break; - - valueOfVar.erase(i, 1); - } - - constexpr auto kTrueVal = "true"; - constexpr auto kFalseVal = "false"; - - if (valueOfVar == kTrueVal) { - valueOfVar = "1"; - } else if (valueOfVar == kFalseVal) { - valueOfVar = "0"; - } - - std::size_t indexRight = 0UL; - - for (auto pairRight : kRegisterMap) { - ++indexRight; - - if (pairRight != valueOfVar) { - if (valueOfVar[0] == '\"') { - syntax_tree.fUserValue = "segment .data64 __LIBCOMPILER_LOCAL_VAR_" + varName + - ": db " + valueOfVar + ", 0\n\n"; - syntax_tree.fUserValue += instr + kRegisterList[kRegisterMap.size() - 1] + ", " + - "__LIBCOMPILER_LOCAL_VAR_" + varName + "\n"; - kOrigin += 1UL; - } else { - syntax_tree.fUserValue = - instr + kRegisterList[kRegisterMap.size() - 1] + ", " + valueOfVar + "\n"; - kOrigin += 1UL; - } - - goto done; - } - } - - if (((int) indexRight - 1) < 0) { - if (valueOfVar[0] == '\"') { - syntax_tree.fUserValue = "segment .data64 __LIBCOMPILER_LOCAL_VAR_" + varName + - ": db " + valueOfVar + ", 0\n"; - syntax_tree.fUserValue += instr + kRegisterList[kRegisterMap.size()] + ", " + - "__LIBCOMPILER_LOCAL_VAR_" + varName + "\n"; - kOrigin += 1UL; - } else { - syntax_tree.fUserValue = - instr + kRegisterList[kRegisterMap.size()] + ", " + valueOfVar + "\n"; - kOrigin += 1UL; - } - - goto done; - } - - if (valueOfVar[0] != '\"' && valueOfVar[0] != '\'' && !isdigit(valueOfVar[0])) { - for (auto pair : kRegisterMap) { - if (pair == valueOfVar) goto done; - } - - Detail::print_error("Variable not declared: " + varName, file); - return false; - } - - done: - for (auto& keyword : kKeywords) { - if (keyword.keyword_kind == LibCompiler::kKeywordKindType && - varName.find(keyword.keyword_name) != std::string::npos) { - varName.erase(varName.find(keyword.keyword_name), keyword.keyword_name.size()); - break; - } - } - - kRegisterMap.push_back(varName); - - break; - } - - if (kKeywords[keyword.second - 1].keyword_kind == LibCompiler::kKeywordKindType || - kKeywords[keyword.second - 1].keyword_kind == LibCompiler::kKeywordKindTypePtr) { - syntax_tree.fUserValue = "\n"; - continue; - } - - if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindEndInstr) { - syntax_tree.fUserValue = "\n"; - continue; - } - - if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindVariableInc) { - instr = "add "; - } else if (keyword.first.keyword_kind == - LibCompiler::KeywordKind::kKeywordKindVariableDec) { - instr = "sub "; - } - - std::string varErrCpy = varName; - - while (varName.find(" ") != std::string::npos) { - varName.erase(varName.find(" "), 1); - } - - while (varName.find("\t") != std::string::npos) { - varName.erase(varName.find("\t"), 1); - } - - std::size_t indxReg = 0UL; - - for (size_t i = 0; !isalnum(valueOfVar[i]); i++) { - if (i > valueOfVar.size()) break; - - valueOfVar.erase(i, 1); - } - - while (valueOfVar.find(" ") != std::string::npos) { - valueOfVar.erase(valueOfVar.find(" "), 1); - } - - while (valueOfVar.find("\t") != std::string::npos) { - valueOfVar.erase(valueOfVar.find("\t"), 1); - } - - constexpr auto kTrueVal = "true"; - constexpr auto kFalseVal = "false"; - - /// interpet boolean values, since we're on C++ - - if (valueOfVar == kTrueVal) { - valueOfVar = "1"; - } else if (valueOfVar == kFalseVal) { - valueOfVar = "0"; - } - - for (auto pair : kRegisterMap) { - ++indxReg; - - if (pair != varName) continue; - - std::size_t indexRight = 0ul; - - for (auto pairRight : kRegisterMap) { - ++indexRight; - - if (pairRight != varName) { - syntax_tree.fUserValue = - instr + kRegisterList[kRegisterMap.size()] + ", " + valueOfVar + "\n"; - kOrigin += 1UL; - continue; - } - - syntax_tree.fUserValue = - instr + kRegisterList[indexRight - 1] + ", " + valueOfVar + "\n"; - kOrigin += 1UL; - break; - } - - break; - } - - if (syntax_tree.fUserValue.empty()) { - Detail::print_error("Variable not declared: " + varName, file); - } - - break; - } - case LibCompiler::KeywordKind::kKeywordKindReturn: { - try { - auto pos = text.find("return") + strlen("return") + 1; - std::string subText = text.substr(pos); - subText = subText.erase(subText.find(";")); - size_t indxReg = 0UL; - - if (subText[0] != '\"' && subText[0] != '\'') { - if (!isdigit(subText[0])) { - for (auto pair : kRegisterMap) { - ++indxReg; - - if (pair != subText) continue; - - syntax_tree.fUserValue = "mov rax, " + kRegisterList[indxReg - 1] + "\nret\n"; - kOrigin += 1UL; - - break; - } - } else { - syntax_tree.fUserValue = "mov rax, " + subText + "\nret\n"; - kOrigin += 1UL; - - break; - } - } else { - syntax_tree.fUserValue = "__LIBCOMPILER_LOCAL_RETURN_STRING: db " + subText + - ", 0\nmov rcx, __LIBCOMPILER_LOCAL_RETURN_STRING\n"; - syntax_tree.fUserValue += "mov rax, rcx\nret\n"; - kOrigin += 1UL; - - break; - } - - if (syntax_tree.fUserValue.empty()) { - if (subText.find("(") != std::string::npos) { - subText.erase(subText.find("(")); - - auto it = - std::find_if(kOriginMap.begin(), kOriginMap.end(), - [&subText](std::pair pair) -> bool { - return pair.first.find(subText) != std::string::npos; - }); - - if (it == kOriginMap.end()) - Detail::print_error("Invalid return value: " + subText, file); - - std::stringstream ss; - ss << it->second; - - syntax_tree.fUserValue = "jmp " + ss.str() + "\nret\n"; - kOrigin += 1UL; - break; - } - } - - break; - } catch (...) { - syntax_tree.fUserValue = "ret\n"; - kOrigin += 1UL; - } - } - default: { - continue; - } - } - } - - kState.fOutputValue = syntax_tree.fUserValue; - return true; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -/** - * @brief C++ assembler class. - */ - -///////////////////////////////////////////////////////////////////////////////////////// - -class AssemblyCPlusPlusInterface final LC_ASSEMBLY_INTERFACE { - public: - explicit AssemblyCPlusPlusInterface() = default; - ~AssemblyCPlusPlusInterface() override = default; - - LIBCOMPILER_COPY_DEFAULT(AssemblyCPlusPlusInterface); - - UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArchAMD64; } - - Int32 CompileToFormat(std::string src, Int32 arch) override { - if (kCompilerFrontend == nullptr) return kExitNO; - - std::string dest = src; - dest += ".pp.masm"; - - std::ofstream out_fp(dest); - - std::ifstream src_fp = std::ifstream(src); - - std::string line_source; - - while (std::getline(src_fp, line_source)) { - if (kVerbose) { - kStdOut << line_source << std::endl; - kStdOut << line_source.length() << " bytes\n"; - } - - kCompilerFrontend->Compile(line_source, src); - - out_fp << kState.fOutputValue; - kState.fOutputValue.clear(); - } - - if (kVerbose) { - kStdOut << "Done compiling " << src << " to " << dest << "\n"; - } - - return kExitOK; - } -}; - -///////////////////////////////////////////////////////////////////////////////////////// - -///////////////////////////////////////////////////////////////////////////////////////// - -#define kExtListCxx \ - { ".cpp", ".cxx", ".cc", ".c++", ".cp" } - -LIBCOMPILER_MODULE(CompilerCPlusPlusAMD64) { - Boolean skip = false; - - kKeywords.push_back({.keyword_name = "if", .keyword_kind = LibCompiler::kKeywordKindIf}); - kKeywords.push_back({.keyword_name = "else", .keyword_kind = LibCompiler::kKeywordKindElse}); - kKeywords.push_back({.keyword_name = "else if", .keyword_kind = LibCompiler::kKeywordKindElseIf}); - - kKeywords.push_back({.keyword_name = "class", .keyword_kind = LibCompiler::kKeywordKindClass}); - kKeywords.push_back({.keyword_name = "struct", .keyword_kind = LibCompiler::kKeywordKindClass}); - kKeywords.push_back( - {.keyword_name = "namespace", .keyword_kind = LibCompiler::kKeywordKindNamespace}); - kKeywords.push_back( - {.keyword_name = "typedef", .keyword_kind = LibCompiler::kKeywordKindTypedef}); - kKeywords.push_back({.keyword_name = "using", .keyword_kind = LibCompiler::kKeywordKindTypedef}); - kKeywords.push_back({.keyword_name = "{", .keyword_kind = LibCompiler::kKeywordKindBodyStart}); - kKeywords.push_back({.keyword_name = "}", .keyword_kind = LibCompiler::kKeywordKindBodyEnd}); - kKeywords.push_back({.keyword_name = "auto", .keyword_kind = LibCompiler::kKeywordKindVariable}); - kKeywords.push_back({.keyword_name = "int", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "bool", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "unsigned", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "short", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "char", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "long", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "float", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "double", .keyword_kind = LibCompiler::kKeywordKindType}); - kKeywords.push_back({.keyword_name = "void", .keyword_kind = LibCompiler::kKeywordKindType}); - - kKeywords.push_back( - {.keyword_name = "auto*", .keyword_kind = LibCompiler::kKeywordKindVariablePtr}); - kKeywords.push_back({.keyword_name = "int*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back({.keyword_name = "bool*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back( - {.keyword_name = "unsigned*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back({.keyword_name = "short*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back({.keyword_name = "char*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back({.keyword_name = "long*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back({.keyword_name = "float*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back( - {.keyword_name = "double*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - kKeywords.push_back({.keyword_name = "void*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); - - kKeywords.push_back( - {.keyword_name = "(", .keyword_kind = LibCompiler::kKeywordKindFunctionStart}); - kKeywords.push_back({.keyword_name = ")", .keyword_kind = LibCompiler::kKeywordKindFunctionEnd}); - kKeywords.push_back( - {.keyword_name = "=", .keyword_kind = LibCompiler::kKeywordKindVariableAssign}); - kKeywords.push_back({.keyword_name = "+=", .keyword_kind = LibCompiler::kKeywordKindVariableInc}); - kKeywords.push_back({.keyword_name = "-=", .keyword_kind = LibCompiler::kKeywordKindVariableDec}); - kKeywords.push_back({.keyword_name = "const", .keyword_kind = LibCompiler::kKeywordKindConstant}); - kKeywords.push_back({.keyword_name = "*", .keyword_kind = LibCompiler::kKeywordKindPtr}); - kKeywords.push_back({.keyword_name = "->", .keyword_kind = LibCompiler::kKeywordKindPtrAccess}); - kKeywords.push_back({.keyword_name = ".", .keyword_kind = LibCompiler::kKeywordKindAccess}); - kKeywords.push_back({.keyword_name = ",", .keyword_kind = LibCompiler::kKeywordKindArgSeparator}); - kKeywords.push_back({.keyword_name = ";", .keyword_kind = LibCompiler::kKeywordKindEndInstr}); - kKeywords.push_back({.keyword_name = ":", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); - kKeywords.push_back( - {.keyword_name = "public:", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); - kKeywords.push_back( - {.keyword_name = "private:", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); - kKeywords.push_back( - {.keyword_name = "protected:", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); - kKeywords.push_back( - {.keyword_name = "final", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); - kKeywords.push_back({.keyword_name = "return", .keyword_kind = LibCompiler::kKeywordKindReturn}); - kKeywords.push_back( - {.keyword_name = "/*", .keyword_kind = LibCompiler::kKeywordKindCommentMultiLineStart}); - kKeywords.push_back( - {.keyword_name = "*/", .keyword_kind = LibCompiler::kKeywordKindCommentMultiLineEnd}); - kKeywords.push_back( - {.keyword_name = "//", .keyword_kind = LibCompiler::kKeywordKindCommentInline}); - kKeywords.push_back({.keyword_name = "==", .keyword_kind = LibCompiler::kKeywordKindEq}); - kKeywords.push_back({.keyword_name = "!=", .keyword_kind = LibCompiler::kKeywordKindNotEq}); - kKeywords.push_back({.keyword_name = ">=", .keyword_kind = LibCompiler::kKeywordKindGreaterEq}); - kKeywords.push_back({.keyword_name = "<=", .keyword_kind = LibCompiler::kKeywordKindLessEq}); - - kErrorLimit = 100; - - kCompilerFrontend = new CompilerFrontendCPlusPlus(); - kFactory.Mount(new AssemblyCPlusPlusInterface()); - - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - for (auto index = 1UL; index < argc; ++index) { - if (argv[index][0] == '-') { - if (skip) { - skip = false; - continue; - } - - if (strcmp(argv[index], "-cxx-verbose") == 0) { - kVerbose = true; - - continue; - } - - if (strcmp(argv[index], "-cxx-dialect") == 0) { - if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; - - return LIBCOMPILER_SUCCESS; - } - - if (strcmp(argv[index], "-cxx-max-err") == 0) { - try { - kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); - } - // catch anything here - catch (...) { - kErrorLimit = 0; - } - - skip = true; - - continue; - } - - std::string err = "Unknown option: "; - err += argv[index]; - - Detail::print_error(err, "cxxdrv"); - - continue; - } - - std::string argv_i = argv[index]; - - std::vector exts = kExtListCxx; - BOOL found = false; - - for (std::string ext : exts) { - if (argv_i.ends_with(ext)) { - found = true; - - if (kFactory.Compile(argv_i, kMachine) != kExitOK) { - return LIBCOMPILER_INVALID_DATA; - } - } - } - - if (!found) { - if (kVerbose) { - Detail::print_error(argv_i + " is not a valid C++ source.", "cxxdrv"); - } - - return LIBCOMPILER_INVALID_DATA; - } - } - - kFactory.Unmount(); - - delete kCompilerFrontend; - kCompilerFrontend = nullptr; - - kRegisterMap.clear(); - kOriginMap.clear(); - - return LIBCOMPILER_SUCCESS; -} - -// -// Last rev 8-1-24 -// diff --git a/dev/LibCompiler/src/CPlusPlusCompilerPreProcessor.cc b/dev/LibCompiler/src/CPlusPlusCompilerPreProcessor.cc deleted file mode 100644 index 9981f37..0000000 --- a/dev/LibCompiler/src/CPlusPlusCompilerPreProcessor.cc +++ /dev/null @@ -1,900 +0,0 @@ -/* - * ======================================================== - * - * bpp - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -/// BUGS: 0 - -#include -#include -#include -#include -#include -#include -#include -#include - -#define kMacroPrefix '#' - -/// @author EL Mahrouss Amlal (amlel) -/// @file bpp.cxx -/// @brief Preprocessor. - -typedef Int32 (*bpp_parser_fn_t)(LibCompiler::String& line, std::ifstream& hdr_file, - std::ofstream& pp_out); - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief Preprocessor internal types. - -///////////////////////////////////////////////////////////////////////////////////////// - -namespace Detail { -enum { - kInvalid = 0, - kEqual = 100, - kGreaterEqThan, - kLesserEqThan, - kGreaterThan, - kLesserThan, - kNotEqual, - kCount = 6, -}; - -struct bpp_macro_condition final { - int32_t fType; - LibCompiler::String fTypeName; - - void Print() { - std::cout << "type: " << fType << "\n"; - std::cout << "type_name: " << fTypeName << "\n"; - } -}; - -struct bpp_macro final { - std::vector fArgs; - LibCompiler::String fName; - LibCompiler::String fValue; - - void Print() { - std::cout << "name: " << fName << "\n"; - std::cout << "value: " << fValue << "\n"; - - for (auto& arg : fArgs) { - std::cout << "arg: " << arg << "\n"; - } - } -}; -} // namespace Detail - -static std::vector kFiles; -static std::vector kMacros; -static std::vector kIncludes; - -static LibCompiler::String kWorkingDir = ""; - -///////////////////////////////////////////////////////////////////////////////////////// - -// @name bpp_parse_if_condition -// @brief parse #if condition - -///////////////////////////////////////////////////////////////////////////////////////// - -int32_t bpp_parse_if_condition(Detail::bpp_macro_condition& cond, Detail::bpp_macro& macro, - bool& inactive_code, bool& defined, LibCompiler::String& macro_str) { - if (cond.fType == Detail::kEqual) { - auto substr_macro = macro_str.substr(macro_str.find(macro.fName) + macro.fName.size()); - - if (substr_macro.find(macro.fValue) != LibCompiler::String::npos) { - if (macro.fValue == "0") { - defined = false; - inactive_code = true; - - return 1; - } - - defined = true; - inactive_code = false; - - return 1; - } - } else if (cond.fType == Detail::kNotEqual) { - auto substr_macro = macro_str.substr(macro_str.find(macro.fName) + macro.fName.size()); - - if (substr_macro.find(macro.fName) != LibCompiler::String::npos) { - if (substr_macro.find(macro.fValue) != LibCompiler::String::npos) { - defined = false; - inactive_code = true; - - return 1; - } - - defined = true; - inactive_code = false; - - return 1; - } - - return 0; - } - - auto substr_macro = macro_str.substr(macro_str.find(macro.fName) + macro.fName.size()); - - LibCompiler::String number; - - for (auto& macro_num : kMacros) { - if (substr_macro.find(macro_num.fName) != LibCompiler::String::npos) { - for (size_t i = 0; i < macro_num.fName.size(); ++i) { - if (isdigit(macro_num.fValue[i])) { - number += macro_num.fValue[i]; - } else { - number.clear(); - break; - } - } - - break; - } - } - - size_t y = 2; - - /* last try */ - for (; y < macro_str.size(); y++) { - if (isdigit(macro_str[y])) { - for (size_t x = y; x < macro_str.size(); x++) { - if (macro_str[x] == ' ') break; - - number += macro_str[x]; - } - - break; - } - } - - size_t rhs = atol(macro.fValue.c_str()); - size_t lhs = atol(number.c_str()); - - if (lhs == 0) { - number.clear(); - ++y; - - for (; y < macro_str.size(); y++) { - if (isdigit(macro_str[y])) { - for (size_t x = y; x < macro_str.size(); x++) { - if (macro_str[x] == ' ') break; - - number += macro_str[x]; - } - - break; - } - } - - lhs = atol(number.c_str()); - } - - if (cond.fType == Detail::kGreaterThan) { - if (lhs < rhs) { - defined = true; - inactive_code = false; - - return 1; - } - - return 0; - } - - if (cond.fType == Detail::kGreaterEqThan) { - if (lhs <= rhs) { - defined = true; - inactive_code = false; - - return 1; - } - - return 0; - } - - if (cond.fType == Detail::kLesserEqThan) { - if (lhs >= rhs) { - defined = true; - inactive_code = false; - - return 1; - } - - return 0; - } - - if (cond.fType == Detail::kLesserThan) { - if (lhs > rhs) { - defined = true; - inactive_code = false; - - return 1; - } - - return 0; - } - - return 0; -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief stores every included file here. - -///////////////////////////////////////////////////////////////////////////////////////// - -std::vector kAllIncludes; - -///////////////////////////////////////////////////////////////////////////////////////// - -// @name bpp_parse_file -// @brief parse file to preprocess it. - -///////////////////////////////////////////////////////////////////////////////////////// - -void bpp_parse_file(std::ifstream& hdr_file, std::ofstream& pp_out) { - LibCompiler::String hdr_line; - LibCompiler::String line_after_include; - - bool inactive_code = false; - bool defined = false; - - try { - while (std::getline(hdr_file, hdr_line)) { - if (inactive_code) { - if (hdr_line.find("#endif") == LibCompiler::String::npos) { - continue; - } else if (hdr_line[0] == kMacroPrefix && - hdr_line.find("#endif") != LibCompiler::String::npos) { - inactive_code = false; - } - - if (hdr_line.find("*/") != LibCompiler::String::npos) { - continue; - } - } - - if (hdr_line.find("--/") != LibCompiler::String::npos) { - hdr_line.erase(hdr_line.find("--/")); - } - - if (hdr_line.find("--*") != LibCompiler::String::npos) { - inactive_code = true; - // get rid of comment. - hdr_line.erase(hdr_line.find("--*")); - } - - /// BPP 'brief' documentation. - if (hdr_line.find("@brief") != LibCompiler::String::npos) { - hdr_line.erase(hdr_line.find("@brief")); - - // TODO: Write an .html or append to it. - } - - if (hdr_line[0] == kMacroPrefix && hdr_line.find("endif") != LibCompiler::String::npos) { - if (!defined && inactive_code) { - inactive_code = false; - defined = false; - - continue; - } - - continue; - } - - if (!defined && inactive_code) { - continue; - } - - if (defined && inactive_code) { - continue; - } - - for (auto macro : kMacros) { - if (LibCompiler::find_word(hdr_line, macro.fName)) { - if (hdr_line.substr(hdr_line.find(macro.fName)).find(macro.fName + '(') != - LibCompiler::String::npos) { - if (!macro.fArgs.empty()) { - LibCompiler::String symbol_val = macro.fValue; - std::vector args; - - size_t x_arg_indx = 0; - - LibCompiler::String line_after_define = hdr_line; - LibCompiler::String str_arg; - - if (line_after_define.find("(") != LibCompiler::String::npos) { - line_after_define.erase(0, line_after_define.find("(") + 1); - - for (auto& subc : line_after_define) { - if (subc == ' ' || subc == '\t') continue; - - if (subc == ',' || subc == ')') { - if (str_arg.empty()) continue; - - args.push_back(str_arg); - - str_arg.clear(); - - continue; - } - - str_arg.push_back(subc); - } - } - - for (auto arg : macro.fArgs) { - if (symbol_val.find(macro.fArgs[x_arg_indx]) != LibCompiler::String::npos) { - symbol_val.replace(symbol_val.find(macro.fArgs[x_arg_indx]), - macro.fArgs[x_arg_indx].size(), args[x_arg_indx]); - ++x_arg_indx; - } else { - throw std::runtime_error("cppdrv: Internal error."); - } - } - - auto len = macro.fName.size(); - len += symbol_val.size(); - len += 2; // ( and ) - - hdr_line.erase(hdr_line.find(")"), 1); - - hdr_line.replace(hdr_line.find(hdr_line.substr(hdr_line.find(macro.fName + '('))), - len, symbol_val); - } else { - auto value = macro.fValue; - - hdr_line.replace(hdr_line.find(macro.fName), macro.fName.size(), value); - } - } - } - } - - if (hdr_line[0] == kMacroPrefix && hdr_line.find("define ") != LibCompiler::String::npos) { - auto line_after_define = hdr_line.substr(hdr_line.find("define ") + strlen("define ")); - - LibCompiler::String macro_value; - LibCompiler::String macro_key; - - std::size_t pos = 0UL; - - std::vector args; - bool on_args = false; - - for (auto& ch : line_after_define) { - ++pos; - - if (ch == '(') { - on_args = true; - continue; - } - - if (ch == ')') { - on_args = false; - continue; - } - - if (ch == '\\') continue; - - if (on_args) continue; - - if (ch == ' ') { - for (size_t i = pos; i < line_after_define.size(); i++) { - macro_value += line_after_define[i]; - } - - break; - } - - macro_key += ch; - } - - LibCompiler::String str; - - if (line_after_define.find("(") != LibCompiler::String::npos) { - line_after_define.erase(0, line_after_define.find("(") + 1); - - for (auto& subc : line_after_define) { - if (subc == ',' || subc == ')') { - if (str.empty()) continue; - - args.push_back(str); - - str.clear(); - - continue; - } - - str.push_back(subc); - } - } - - Detail::bpp_macro macro; - - macro.fArgs = args; - macro.fName = macro_key; - macro.fValue = macro_value; - - kMacros.emplace_back(macro); - - continue; - } - - if (hdr_line[0] != kMacroPrefix) { - if (inactive_code) { - continue; - } - - pp_out << hdr_line << std::endl; - - continue; - } - - if (hdr_line[0] == kMacroPrefix && hdr_line.find("ifndef") != LibCompiler::String::npos) { - auto line_after_ifndef = hdr_line.substr(hdr_line.find("ifndef") + strlen("ifndef") + 1); - LibCompiler::String macro; - - for (auto& ch : line_after_ifndef) { - if (ch == ' ') { - break; - } - - macro += ch; - } - - if (macro == "0") { - defined = true; - inactive_code = false; - continue; - } - - if (macro == "1") { - defined = false; - inactive_code = true; - - continue; - } - - bool found = false; - - defined = true; - inactive_code = false; - - for (auto& macro_ref : kMacros) { - if (hdr_line.find(macro_ref.fName) != LibCompiler::String::npos) { - found = true; - break; - } - } - - if (found) { - defined = false; - inactive_code = true; - - continue; - } - } else if (hdr_line[0] == kMacroPrefix && - hdr_line.find("else") != LibCompiler::String::npos) { - if (!defined && inactive_code) { - inactive_code = false; - defined = true; - - continue; - } else { - defined = false; - inactive_code = true; - - continue; - } - } else if (hdr_line[0] == kMacroPrefix && - hdr_line.find("ifdef") != LibCompiler::String::npos) { - auto line_after_ifdef = hdr_line.substr(hdr_line.find("ifdef") + strlen("ifdef") + 1); - LibCompiler::String macro; - - for (auto& ch : line_after_ifdef) { - if (ch == ' ') { - break; - } - - macro += ch; - } - - if (macro == "0") { - defined = false; - inactive_code = true; - - continue; - } - - if (macro == "1") { - defined = true; - inactive_code = false; - - continue; - } - - defined = false; - inactive_code = true; - - for (auto& macro_ref : kMacros) { - if (hdr_line.find(macro_ref.fName) != LibCompiler::String::npos) { - defined = true; - inactive_code = false; - - break; - } - } - } else if (hdr_line[0] == kMacroPrefix && hdr_line.find("if") != LibCompiler::String::npos) { - inactive_code = true; - - std::vector bpp_macro_condition_list = { - { - .fType = Detail::kEqual, - .fTypeName = "==", - }, - { - .fType = Detail::kNotEqual, - .fTypeName = "!=", - }, - { - .fType = Detail::kLesserThan, - .fTypeName = "<", - }, - { - .fType = Detail::kGreaterThan, - .fTypeName = ">", - }, - { - .fType = Detail::kLesserEqThan, - .fTypeName = "<=", - }, - { - .fType = Detail::kGreaterEqThan, - .fTypeName = ">=", - }, - }; - - int32_t good_to_go = 0; - - for (auto& macro_condition : bpp_macro_condition_list) { - if (hdr_line.find(macro_condition.fTypeName) != LibCompiler::String::npos) { - for (auto& found_macro : kMacros) { - if (hdr_line.find(found_macro.fName) != LibCompiler::String::npos) { - good_to_go = bpp_parse_if_condition(macro_condition, found_macro, inactive_code, - defined, hdr_line); - - break; - } - } - } - } - - if (good_to_go) continue; - - auto line_after_if = hdr_line.substr(hdr_line.find("if") + strlen("if") + 1); - LibCompiler::String macro; - - for (auto& ch : line_after_if) { - if (ch == ' ') { - break; - } - - macro += ch; - } - - if (macro == "0") { - defined = false; - inactive_code = true; - continue; - } - - if (macro == "1") { - defined = true; - inactive_code = false; - - continue; - } - - // last try, is it defined to be one? - for (auto& macro_ref : kMacros) { - if (macro_ref.fName.find(macro) != LibCompiler::String::npos && macro_ref.fValue == "1") { - inactive_code = false; - defined = true; - - break; - } - } - } else if (hdr_line[0] == kMacroPrefix && - hdr_line.find("warning") != LibCompiler::String::npos) { - auto line_after_warning = hdr_line.substr(hdr_line.find("warning") + strlen("warning") + 1); - LibCompiler::String message; - - for (auto& ch : line_after_warning) { - if (ch == '\r' || ch == '\n') { - break; - } - - message += ch; - } - - std::cout << "warn: " << message << std::endl; - } else if (hdr_line[0] == kMacroPrefix && - hdr_line.find("error") != LibCompiler::String::npos) { - auto line_after_warning = hdr_line.substr(hdr_line.find("error") + strlen("error") + 1); - LibCompiler::String message; - - for (auto& ch : line_after_warning) { - if (ch == '\r' || ch == '\n') { - break; - } - - message += ch; - } - - throw std::runtime_error("error: " + message); - } else if (hdr_line[0] == kMacroPrefix && - hdr_line.find("include ") != LibCompiler::String::npos) { - line_after_include = hdr_line.substr(hdr_line.find("include ") + strlen("include ")); - - kIncludeFile: - auto it = std::find(kAllIncludes.cbegin(), kAllIncludes.cend(), line_after_include); - - if (it != kAllIncludes.cend()) { - continue; - } - - LibCompiler::String path; - - kAllIncludes.push_back(line_after_include); - - bool enable = false; - bool not_local = false; - - for (auto& ch : line_after_include) { - if (ch == ' ') continue; - - if (ch == '<') { - not_local = true; - enable = true; - - continue; - } - - if (ch == '\"') { - not_local = false; - enable = true; - continue; - } - - if (enable) { - path += ch; - } - } - - if (not_local) { - bool open = false; - - if (path.ends_with('>')) { - path.erase(path.find('>')); - } - - if (path.ends_with('"')) { - path.erase(path.find('"')); - } - - for (auto& include : kIncludes) { - LibCompiler::String header_path = include; - header_path.push_back('/'); - header_path += path; - - std::ifstream header(header_path); - - if (!header.is_open()) continue; - - open = true; - - bpp_parse_file(header, pp_out); - - break; - } - - if (!open) { - throw std::runtime_error("cppdrv: no such include file: " + path); - } - } else { - std::ifstream header(path); - - if (!header.is_open()) throw std::runtime_error("cppdrv: no such include file: " + path); - - bpp_parse_file(header, pp_out); - } - } else { - std::cerr << ("cppdrv: unknown pre-processor directive, " + hdr_line) << "\n"; - continue; - } - } - } catch (std::out_of_range& oor) { - return; - } -} - -///////////////////////////////////////////////////////////////////////////////////////// - -// @brief main entrypoint of app. - -///////////////////////////////////////////////////////////////////////////////////////// - -LIBCOMPILER_MODULE(CPlusPlusPreprocessorMain) { - try { - bool skip = false; - bool double_skip = false; - - Detail::bpp_macro macro_1; - - macro_1.fName = "__true"; - macro_1.fValue = "1"; - - kMacros.push_back(macro_1); - - Detail::bpp_macro macro_unreachable; - - macro_unreachable.fName = "__unreachable"; - macro_unreachable.fValue = "__libcompiler_unreachable"; - - kMacros.push_back(macro_unreachable); - - Detail::bpp_macro macro_unused; - - macro_unreachable.fName = "__unused"; - macro_unreachable.fValue = "__libcompiler_unused"; - - kMacros.push_back(macro_unused); - - Detail::bpp_macro macro_0; - - macro_0.fName = "__false"; - macro_0.fValue = "0"; - - kMacros.push_back(macro_0); - - Detail::bpp_macro macro_zka; - - macro_zka.fName = "__LIBCOMPILER__"; - macro_zka.fValue = "1"; - - kMacros.push_back(macro_zka); - - Detail::bpp_macro macro_cxx; - - macro_cxx.fName = "__cplusplus"; - macro_cxx.fValue = "202302L"; - - kMacros.push_back(macro_cxx); - - Detail::bpp_macro macro_size_t; - macro_size_t.fName = "__SIZE_TYPE__"; - macro_size_t.fValue = "unsigned long long int"; - - kMacros.push_back(macro_size_t); - - macro_size_t.fName = "__UINT32_TYPE__"; - macro_size_t.fValue = "unsigned int"; - - kMacros.push_back(macro_size_t); - - macro_size_t.fName = "__UINTPTR_TYPE__"; - macro_size_t.fValue = "unsigned long long int"; - - kMacros.push_back(macro_size_t); - - for (auto index = 1UL; index < argc; ++index) { - if (skip) { - skip = false; - continue; - } - - if (double_skip) { - ++index; - double_skip = false; - continue; - } - - if (argv[index][0] == '-') { - if (strcmp(argv[index], "--cppdrv:ver") == 0) { - printf("%s\n", - "NeKernel Preprocessor Driver v1.11, (c) Amlal El Mahrouss 2024-2025 all rights " - "reserved."); - - return LIBCOMPILER_SUCCESS; - } - - if (strcmp(argv[index], "--cppdrv:?") == 0) { - printf("%s\n", - "NeKernel Preprocessor Driver v1.11, (c) Amlal El Mahrouss 2024-2025 all rights " - "reserved."); - printf("%s\n", "--cppdrv:working-dir : set directory to working path."); - printf("%s\n", "--cppdrv:include-dir : add directory to include path."); - printf("%s\n", "--cppdrv:def : define a macro."); - printf("%s\n", "--cppdrv:ver: print the version."); - printf("%s\n", "--cppdrv:?: show help (this current command)."); - - return LIBCOMPILER_SUCCESS; - } - - if (strcmp(argv[index], "--cppdrv:include-dir") == 0) { - LibCompiler::String inc = argv[index + 1]; - - skip = true; - - kIncludes.push_back(inc); - } - - if (strcmp(argv[index], "--cppdrv:working-dir") == 0) { - LibCompiler::String inc = argv[index + 1]; - skip = true; - kWorkingDir = inc; - } - - if (strcmp(argv[index], "--cppdrv:def") == 0 && argv[index + 1] != nullptr && - argv[index + 2] != nullptr) { - LibCompiler::String macro_key = argv[index + 1]; - - LibCompiler::String macro_value; - bool is_string = false; - - for (int argv_find_len = 0; argv_find_len < strlen(argv[index]); ++argv_find_len) { - if (!isdigit(argv[index][argv_find_len])) { - is_string = true; - macro_value += "\""; - - break; - } - } - - macro_value += argv[index + 2]; - - if (is_string) macro_value += "\""; - - Detail::bpp_macro macro; - macro.fName = macro_key; - macro.fValue = macro_value; - - kMacros.push_back(macro); - - double_skip = true; - } - - continue; - } - - kFiles.emplace_back(argv[index]); - } - - if (kFiles.empty()) return LIBCOMPILER_EXEC_ERROR; - - for (auto& file : kFiles) { - if (!std::filesystem::exists(file)) continue; - - std::ifstream file_descriptor(file); - std::ofstream file_descriptor_pp(file + ".pp"); - - bpp_parse_file(file_descriptor, file_descriptor_pp); - } - - return LIBCOMPILER_SUCCESS; - } catch (const std::runtime_error& e) { - std::cout << e.what() << '\n'; - } - - return LIBCOMPILER_EXEC_ERROR; -} - -// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/Cl/CCompiler64x0.cc b/dev/LibCompiler/src/Cl/CCompiler64x0.cc new file mode 100644 index 0000000..6ca680b --- /dev/null +++ b/dev/LibCompiler/src/Cl/CCompiler64x0.cc @@ -0,0 +1,1286 @@ +/* + * ======================================================== + * + * cc + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +/// BUGS: 0 +/// TODO: none + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* C driver */ +/* This is part of the LibCompiler. */ +/* (c) Amlal El Mahrouss */ + +/// @author EL Mahrouss Amlal (amlel) +/// @file 64x0-cc.cxx +/// @brief 64x0 C Compiler. + +/// TODO: support structures, else if, else, . and -> + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +#define kExitOK (0) + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" + +///////////////////////////////////// + +// INTERNAL STUFF OF THE C COMPILER + +///////////////////////////////////// + +namespace Detail { +// \brief Register map structure, used to keep track of each variable's registers. +struct CompilerRegisterMap final { + std::string fName; + std::string fReg; +}; + +// \brief Map for C structs +// \author amlel +struct CompilerStructMap final { + // 'my_foo' + std::string fName; + + // if instance: stores a valid register. + std::string fReg; + + // offset count + std::size_t fOffsetsCnt; + + // offset array. + std::vector> fOffsets; +}; + +struct CompilerState final { + std::vector fSyntaxTreeList; + std::vector kStackFrame; + std::vector kStructMap; + LibCompiler::SyntaxLeafList* fSyntaxTree{nullptr}; + std::unique_ptr fOutputAssembly; + std::string fLastFile; + std::string fLastError; + bool fVerbose; +}; +} // namespace Detail + +static Detail::CompilerState kState; +static std::string kIfFunction = ""; + +namespace Detail { +/// @brief prints an error into stdout. +/// @param reason the reason of the error. +/// @param file where does it originate from? +void print_error(std::string reason, std::string file) noexcept; + +struct CompilerType final { + std::string fName; + std::string fValue; +}; +} // namespace Detail + +///////////////////////////////////////////////////////////////////////////////////////// + +// Target architecture. +static int kMachine = 0; + +///////////////////////////////////////// + +// REGISTERS ACCORDING TO USED ASSEMBLER + +///////////////////////////////////////// + +static size_t kRegisterCnt = kAsmRegisterLimit; +static size_t kStartUsable = 2; +static size_t kUsableLimit = 15; +static size_t kRegisterCounter = kStartUsable; +static std::string kRegisterPrefix = kAsmRegisterPrefix; + +///////////////////////////////////////// + +// COMPILER PARSING UTILITIES/STATES. + +///////////////////////////////////////// + +static std::vector kFileList; +static LibCompiler::AssemblyFactory kFactory; +static bool kInStruct = false; +static bool kOnWhileLoop = false; +static bool kOnForLoop = false; +static bool kInBraces = false; +static bool kIfFound = false; +static size_t kBracesCount = 0UL; + +/* @brief C compiler backend for C */ +class CompilerFrontend64x0 final : public LibCompiler::CompilerFrontendInterface { + public: + explicit CompilerFrontend64x0() = default; + ~CompilerFrontend64x0() override = default; + + LIBCOMPILER_COPY_DEFAULT(CompilerFrontend64x0); + + std::string Check(const char* text, const char* file); + bool Compile(std::string text, std::string file) override; + + const char* Language() override { return "64k C"; } +}; + +static CompilerFrontend64x0* kCompilerFrontend = nullptr; +static std::vector kCompilerVariables; +static std::vector kCompilerFunctions; +static std::vector kCompilerTypes; + +namespace Detail { +union number_cast final { + public: + number_cast(UInt64 _Raw) : _Raw(_Raw) {} + + public: + char _Num[8]; + UInt64 _Raw; +}; + +union double_cast final { + public: + double_cast(float _Raw) : _Raw(_Raw) {} + + public: + char _Sign; + char _Lh[8]; + char _Rh[23]; + + float _Raw; +}; +} // namespace Detail + +///////////////////////////////////////////////////////////////////////////////////////// + +// @name Compile +// @brief Generate MASM from a C assignement. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool CompilerFrontend64x0::Compile(std::string text_, std::string file) { + std::string text = text_; + + bool typeFound = false; + bool fnFound = false; + + // setup generator. + std::random_device rd; + + auto seed_data = std::array{}; + std::generate(std::begin(seed_data), std::end(seed_data), std::ref(rd)); + std::seed_seq seq(std::begin(seed_data), std::end(seed_data)); + std::mt19937 generator(seq); + + // start parsing + for (size_t text_index = 0; text_index < text.size(); ++text_index) { + auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); + + auto gen = uuids::uuid_random_generator{generator}; + uuids::uuid out = gen(); + + Detail::number_cast time_off = (UInt64) out.as_bytes().data(); + + if (!typeFound) { + auto substr = text.substr(text_index); + std::string match_type; + + for (size_t y = 0; y < substr.size(); ++y) { + if (substr[y] == ' ') { + while (match_type.find(' ') != std::string::npos) { + match_type.erase(match_type.find(' ')); + } + + for (auto& clType : kCompilerTypes) { + if (clType.fName == match_type) { + match_type.clear(); + + std::string buf; + + buf += clType.fValue; + buf += ' '; + + if (substr.find('=') != std::string::npos) { + break; + } + + if (text.find('(') != std::string::npos) { + syntaxLeaf.fUserValue = buf; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + } + + typeFound = true; + break; + } + } + + break; + } + + match_type += substr[y]; + } + } + + if (text[text_index] == '{') { + if (kInStruct) { + continue; + } + + kInBraces = true; + ++kBracesCount; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + } + + // return keyword handler + if (text[text_index] == 'r') { + std::string return_keyword; + return_keyword += "return"; + + std::size_t index = 0UL; + + std::string value; + + for (size_t return_index = text_index; return_index < text.size(); ++return_index) { + if (text[return_index] != return_keyword[index]) { + for (size_t value_index = return_index; value_index < text.size(); ++value_index) { + if (text[value_index] == ';') break; + + value += text[value_index]; + } + + break; + } + + ++index; + } + + if (index == return_keyword.size()) { + if (!value.empty()) { + if (value.find('(') != std::string::npos) { + value.erase(value.find('(')); + } + + if (!isdigit(value[value.find('(') + 2])) { + std::string tmp = value; + bool reg_to_reg = false; + + value.clear(); + + value += " extern_segment"; + value += tmp; + } + + syntaxLeaf.fUserValue = "\tldw r19, "; + + // make it pretty. + if (value.find('\t') != std::string::npos) value.erase(value.find('\t'), 1); + + syntaxLeaf.fUserValue += value + "\n"; + } + + syntaxLeaf.fUserValue += "\tjlr"; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + break; + } + } + + if (text[text_index] == 'i' && text[text_index + 1] == 'f') { + auto expr = text.substr(text_index + 2); + text.erase(text_index, 2); + + if (expr.find("{") != std::string::npos) { + expr.erase(expr.find("{")); + } + + if (expr.find("(") != std::string::npos) expr.erase(expr.find("(")); + + if (expr.find(")") != std::string::npos) expr.erase(expr.find(")")); + + kIfFunction = "__LIBCOMPILER_IF_PROC_"; + kIfFunction += std::to_string(time_off._Raw); + + syntaxLeaf.fUserValue = "\tlda r12, extern_segment "; + syntaxLeaf.fUserValue += kIfFunction + + "\n\t#r12 = Code to jump on, r11 right cond, r10 left cond.\n\tbeq " + "r10, r11, r12\ndword public_segment .code64 " + + kIfFunction + "\n"; + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + kIfFound = true; + } + + // Parse expressions and instructions here. + // what does this mean? + // we encounter an assignment, or we reached the end of an expression. + if (text[text_index] == '=' || text[text_index] == ';') { + if (fnFound) continue; + if (kIfFound) continue; + + if (text[text_index] == ';' && kInStruct) continue; + + if (text.find("typedef ") != std::string::npos) continue; + + if (text[text_index] == '=' && kInStruct) { + Detail::print_error("assignement of value in struct " + text, file); + continue; + } + + if (text[text_index] == ';' && kInStruct) { + bool space_found_ = false; + std::string sym; + + for (auto& ch : text) { + if (ch == ' ') { + space_found_ = true; + } + + if (ch == ';') break; + + if (space_found_) sym.push_back(ch); + } + + kState.kStructMap[kState.kStructMap.size() - 1].fOffsets.push_back( + std::make_pair(kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4, sym)); + + kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt = + kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4; + + continue; + } + + if (text[text_index] == '=' && kInStruct) { + continue; + } + + if (text[text_index + 1] == '=' || text[text_index - 1] == '!' || + text[text_index - 1] == '<' || text[text_index - 1] == '>') { + continue; + } + + std::string substr; + + if (text.find('=') != std::string::npos && kInBraces && !kIfFound) { + if (text.find("*") != std::string::npos) { + if (text.find("=") > text.find("*")) + substr += "\tlda "; + else + substr += "\tldw "; + } else { + substr += "\tldw "; + } + } else if (text.find('=') != std::string::npos && !kInBraces) { + substr += "stw public_segment .data64 "; + } + + int first_encountered = 0; + + std::string str_name; + + for (size_t text_index_2 = 0; text_index_2 < text.size(); ++text_index_2) { + if (text[text_index_2] == '\"') { + ++text_index_2; + + // want to add this, so that the parser recognizes that this is a + // string. + substr += '"'; + + for (; text_index_2 < text.size(); ++text_index_2) { + if (text[text_index_2] == '\"') break; + + substr += text[text_index_2]; + } + } + + if (text[text_index_2] == '{' || text[text_index_2] == '}') continue; + + if (text[text_index_2] == ';') { + break; + } + + if (text[text_index_2] == ' ' || text[text_index_2] == '\t') { + if (first_encountered != 2) { + if (text[text_index] != '=' && + substr.find("public_segment .data64") == std::string::npos && !kInStruct) + substr += "public_segment .data64 "; + } + + ++first_encountered; + + continue; + } + + if (text[text_index_2] == '=') { + if (!kInBraces) { + substr.replace(substr.find("public_segment .data64"), strlen("public_segment .data64"), + "public_segment .zero64 "); + } + + substr += ","; + continue; + } + + substr += text[text_index_2]; + } + + for (auto& clType : kCompilerTypes) { + if (substr.find(clType.fName) != std::string::npos) { + if (substr.find(clType.fName) > substr.find('"')) continue; + + substr.erase(substr.find(clType.fName), clType.fName.size()); + } else if (substr.find(clType.fValue) != std::string::npos) { + if (substr.find(clType.fValue) > substr.find('"')) continue; + + if (clType.fName == "const") continue; + + substr.erase(substr.find(clType.fValue), clType.fValue.size()); + } + } + + if (substr.find("extern") != std::string::npos) { + substr.replace(substr.find("extern"), strlen("extern"), "extern_segment "); + + if (substr.find("public_segment .data64") != std::string::npos) + substr.erase(substr.find("public_segment .data64"), strlen("public_segment .data64")); + } + + auto var_to_find = std::find_if( + kCompilerVariables.cbegin(), kCompilerVariables.cend(), + [&](Detail::CompilerType type) { return type.fName.find(substr) != std::string::npos; }); + + if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; + + std::string reg = kAsmRegisterPrefix; + reg += std::to_string(kRegisterCounter); + + if (var_to_find == kCompilerVariables.cend()) { + ++kRegisterCounter; + + kState.kStackFrame.push_back({.fName = substr, .fReg = reg}); + kCompilerVariables.push_back({.fName = substr}); + } + + syntaxLeaf.fUserValue += substr; + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + if (text[text_index] == '=') break; + } + + // function handler. + + if (text[text_index] == '(' && !fnFound && !kIfFound) { + std::string substr; + std::string args_buffer; + std::string args; + + bool type_crossed = false; + + for (size_t idx = text.find('(') + 1; idx < text.size(); ++idx) { + if (text[idx] == ',') continue; + + if (text[idx] == ' ') continue; + + if (text[idx] == ')') break; + } + + for (char substr_first_index : text) { + if (substr_first_index != ',') + args_buffer += substr_first_index; + else + args_buffer += '$'; + + if (substr_first_index == ';') { + args_buffer = args_buffer.erase(0, args_buffer.find('(')); + args_buffer = args_buffer.erase(args_buffer.find(';'), 1); + args_buffer = args_buffer.erase(args_buffer.find(')'), 1); + args_buffer = args_buffer.erase(args_buffer.find('('), 1); + + if (!args_buffer.empty()) args += "\tldw r6, "; + + std::string register_type; + std::size_t index = 7UL; + + while (args_buffer.find("$") != std::string::npos) { + register_type = kRegisterPrefix; + register_type += std::to_string(index); + + ++index; + + args_buffer.replace(args_buffer.find('$'), 1, "\n\tldw " + register_type + ","); + } + + args += args_buffer; + args += "\n\tlda r19, "; + } + } + + for (char _text_i : text) { + if (_text_i == '\t' || _text_i == ' ') { + if (!type_crossed) { + substr.clear(); + type_crossed = true; + } + + continue; + } + + if (_text_i == '(') break; + + substr += _text_i; + } + + if (kInBraces) { + syntaxLeaf.fUserValue = args; + syntaxLeaf.fUserValue += substr; + syntaxLeaf.fUserValue += "\n\tjrl\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + fnFound = true; + } else { + syntaxLeaf.fUserValue.clear(); + + syntaxLeaf.fUserValue += "public_segment .code64 "; + + syntaxLeaf.fUserValue += substr; + syntaxLeaf.fUserValue += "\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + fnFound = true; + } + + kCompilerFunctions.push_back(text); + } + + if (text[text_index] == '-' && text[text_index + 1] == '-') { + text = text.replace(text.find("--"), strlen("--"), ""); + + for (int _text_i = 0; _text_i < text.size(); ++_text_i) { + if (text[_text_i] == '\t' || text[_text_i] == ' ') text.erase(_text_i, 1); + } + + syntaxLeaf.fUserValue += "sub "; + syntaxLeaf.fUserValue += text; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + break; + } + + if (text[text_index] == '}') { + kRegisterCounter = kStartUsable; + + --kBracesCount; + + if (kBracesCount < 1) { + kInBraces = false; + kBracesCount = 0; + } + + if (kIfFound) kIfFound = false; + + if (kInStruct) kInStruct = false; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + } + + syntaxLeaf.fUserValue.clear(); + } + + auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); + syntaxLeaf.fUserValue = "\n"; + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + return true; +} + +static bool kShouldHaveBraces = false; +static std::string kFnName; + +std::string CompilerFrontend64x0::Check(const char* text, const char* file) { + std::string err_str; + std::string ln = text; + + if (ln.empty()) { + return err_str; + } + + bool non_ascii_found = false; + + for (int i = 0; i < ln.size(); ++i) { + if (isalnum(ln[i])) { + non_ascii_found = true; + break; + } + } + + if (kShouldHaveBraces && ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + } + + if (!non_ascii_found) return err_str; + + size_t string_index = 1UL; + + if (ln.find('\'') != std::string::npos) { + string_index = ln.find('\'') + 1; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == '\'') { + if (ln[string_index + 1] != ';') { + ln.erase(string_index, 1); + } + + return err_str; + } + } + } else if (ln.find('"') != std::string::npos) { + string_index = ln.find('"') + 1; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == '"') { + if (ln[string_index + 1] != ';') { + ln.erase(string_index, 1); + } else { + break; + } + } + } + } else if (ln.find('"') == std::string::npos && ln.find('\'') == std::string::npos) { + std::vector forbidden_words; + + forbidden_words.push_back("\\"); + forbidden_words.push_back("?"); + forbidden_words.push_back("@"); + forbidden_words.push_back("~"); + forbidden_words.push_back("::"); + forbidden_words.push_back("--*"); + forbidden_words.push_back("*/"); + + // add them to avoid stupid mistakes. + forbidden_words.push_back("namespace"); + forbidden_words.push_back("class"); + forbidden_words.push_back("extern \"C\""); + + for (auto& forbidden : forbidden_words) { + if (ln.find(forbidden) != std::string::npos) { + err_str += "\nForbidden character detected: "; + err_str += forbidden; + + return err_str; + } + } + } + + struct CompilerVariableRange final { + std::string fBegin; + std::string fEnd; + }; + + const std::vector variables_list = { + {.fBegin = "static ", .fEnd = "="}, {.fBegin = "=", .fEnd = ";"}, + {.fBegin = "if(", .fEnd = "="}, {.fBegin = "if (", .fEnd = "="}, + {.fBegin = "if(", .fEnd = "<"}, {.fBegin = "if (", .fEnd = "<"}, + {.fBegin = "if(", .fEnd = ">"}, {.fBegin = "if (", .fEnd = ">"}, + {.fBegin = "if(", .fEnd = ")"}, {.fBegin = "if (", .fEnd = ")"}, + + {.fBegin = "else(", .fEnd = "="}, {.fBegin = "else (", .fEnd = "="}, + {.fBegin = "else(", .fEnd = "<"}, {.fBegin = "else (", .fEnd = "<"}, + {.fBegin = "else(", .fEnd = ">"}, {.fBegin = "else (", .fEnd = ">"}, + {.fBegin = "else(", .fEnd = ")"}, {.fBegin = "else (", .fEnd = ")"}, + }; + + for (auto& variable : variables_list) { + if (ln.find(variable.fBegin) != std::string::npos) { + string_index = ln.find(variable.fBegin) + variable.fBegin.size(); + + while (ln[string_index] == ' ') ++string_index; + + std::string keyword; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == variable.fEnd[0]) { + std::string varname = ""; + + for (size_t index_keyword = ln.find(' '); ln[index_keyword] != variable.fBegin[0]; + ++index_keyword) { + if (ln[index_keyword] == ' ') { + continue; + } + + if (isdigit(ln[index_keyword])) { + goto cc_next_loop; + } + + varname += ln[index_keyword]; + } + + if (varname.find(' ') != std::string::npos) { + varname.erase(0, varname.find(' ')); + + if (variable.fBegin == "extern") { + varname.erase(0, varname.find(' ')); + } + } + + if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; + + std::string reg = kAsmRegisterPrefix; + reg += std::to_string(kRegisterCounter); + + kCompilerVariables.push_back({.fValue = varname}); + goto cc_check_done; + } + + keyword.push_back(ln[string_index]); + } + + goto cc_next_loop; + + cc_check_done: + + // skip digit value. + if (isdigit(keyword[0]) || keyword[0] == '"') { + goto cc_next_loop; + } + + while (keyword.find(' ') != std::string::npos) keyword.erase(keyword.find(' '), 1); + + for (auto& var : kCompilerVariables) { + if (var.fValue.find(keyword) != std::string::npos) { + err_str.clear(); + goto cc_next; + } + } + + for (auto& fn : kCompilerFunctions) { + if (fn.find(keyword[0]) != std::string::npos) { + auto where_begin = fn.find(keyword[0]); + auto keyword_begin = 0UL; + auto failed = false; + + for (; where_begin < keyword.size(); ++where_begin) { + if (fn[where_begin] == '(' && keyword[keyword_begin] == '(') break; + + if (fn[where_begin] != keyword[keyword_begin]) { + failed = true; + break; + } + + ++keyword_begin; + } + + if (!failed) { + err_str.clear(); + goto cc_next; + } else { + continue; + } + } + } + + cc_error_value: + if (keyword.find("->") != std::string::npos) return err_str; + + if (keyword.find(".") != std::string::npos) return err_str; + + if (isalnum(keyword[0])) err_str += "\nUndefined value: " + keyword; + + return err_str; + } + + cc_next_loop: + continue; + } + +cc_next: + + // extern does not declare anything, it extern_segments a variable. + // so that's why it's not declare upper. + if (LibCompiler::find_word(ln, "extern")) { + auto substr = ln.substr(ln.find("extern") + strlen("extern")); + kCompilerVariables.push_back({.fValue = substr}); + } + + if (kShouldHaveBraces && ln.find('{') == std::string::npos) { + err_str += "Missing '{' for function "; + err_str += kFnName; + err_str += "\n"; + + kShouldHaveBraces = false; + kFnName.clear(); + } else if (kShouldHaveBraces && ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + kFnName.clear(); + } + + bool type_not_found = true; + + if (ln.find('\'') != std::string::npos) { + ln.replace(ln.find('\''), 3, "0"); + } + + auto first = ln.find('"'); + if (first != std::string::npos) { + auto second = 0UL; + bool found_second_quote = false; + + for (size_t i = first + 1; i < ln.size(); ++i) { + if (ln[i] == '\"') { + found_second_quote = true; + second = i; + + break; + } + } + + if (!found_second_quote) { + err_str += "Missing terminating \"."; + err_str += " here -> " + ln.substr(ln.find('"'), second); + } + } + + if (ln.find(')') != std::string::npos && ln.find(';') == std::string::npos) { + if (ln.find('{') == std::string::npos) { + kFnName = ln; + kShouldHaveBraces = true; + + goto skip_braces_check; + } else if (ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + } + } + +skip_braces_check: + + for (auto& key : kCompilerTypes) { + if (LibCompiler::find_word(ln, key.fName)) { + if (isdigit(ln[ln.find(key.fName) + key.fName.size() + 1])) { + err_str += "\nNumber cannot be set for "; + err_str += key.fName; + err_str += "'s name. here -> "; + err_str += ln; + } + + if (ln.find(key.fName) == 0 || ln[ln.find(key.fName) - 1] == ' ' || + ln[ln.find(key.fName) - 1] == '\t') { + type_not_found = false; + + if (ln[ln.find(key.fName) + key.fName.size()] != ' ') { + type_not_found = true; + + if (ln[ln.find(key.fName) + key.fName.size()] == '\t') type_not_found = false; + + goto next; + } else if (ln[ln.find(key.fName) + key.fName.size()] != '\t') { + type_not_found = true; + + if (ln[ln.find(key.fName) + key.fName.size()] == ' ') type_not_found = false; + } + } + + next: + + if (ln.find(';') == std::string::npos) { + if (ln.find('(') != std::string::npos) { + if (ln.find('=') == std::string::npos) continue; + } + + err_str += "\nMissing ';', here -> "; + err_str += ln; + } else { + continue; + } + + if (ln.find('=') != std::string::npos) { + if (ln.find('(') != std::string::npos) { + if (ln.find(')') == std::string::npos) { + err_str += "\nMissing ')', after '(' here -> "; + err_str += ln.substr(ln.find('(')); + } + } + } + } + } + + if (kInBraces && ln.find("struct") != std::string::npos && + ln.find("union") != std::string::npos && ln.find("enum") != std::string::npos && + ln.find('=') != std::string::npos) { + if (ln.find(';') == std::string::npos) { + err_str += "\nMissing ';' after struct/union/enum declaration, here -> "; + err_str += ln; + } + } + + if (ln.find(';') != std::string::npos && ln.find("for") == std::string::npos) { + if (ln.find(';') + 1 != ln.size()) { + for (int i = 0; i < ln.substr(ln.find(';') + 1).size(); ++i) { + if ((ln.substr(ln.find(';') + 1)[i] != ' ') || (ln.substr(ln.find(';') + 1)[i] != '\t')) { + if (auto err = this->Check(ln.substr(ln.find(';') + 1).c_str(), file); !err.empty()) { + err_str += "\nUnexpected text after ';' -> "; + err_str += ln.substr(ln.find(';')); + err_str += err; + } + } + } + } + } + + if (ln.find('(') != std::string::npos) { + if (ln.find(';') == std::string::npos && !LibCompiler::find_word(ln, "|") && + !LibCompiler::find_word(ln, "||") && !LibCompiler::find_word(ln, "&") && + !LibCompiler::find_word(ln, "&&") && !LibCompiler::find_word(ln, "~")) { + bool found_func = false; + size_t i = ln.find('('); + std::vector opens; + std::vector closes; + + for (; i < ln.size(); ++i) { + if (ln[i] == ')') { + closes.push_back(1); + } + + if (ln[i] == '(') { + opens.push_back(1); + } + } + + if (closes.size() != opens.size()) err_str += "Unterminated (), here -> " + ln; + + bool space_found = false; + + for (int i = 0; i < ln.size(); ++i) { + if (ln[i] == ')' && !space_found) { + space_found = true; + continue; + } + + if (space_found) { + if (ln[i] == ' ' && isalnum(ln[i + 1])) { + err_str += "\nBad function format here -> "; + err_str += ln; + } + } + } + } + + if (ln.find('(') < 1) { + err_str += "\nMissing identifier before '(' here -> "; + err_str += ln; + } else { + if (type_not_found && ln.find(';') == std::string::npos && + ln.find("if") == std::string::npos && ln.find("|") == std::string::npos && + ln.find("&") == std::string::npos && ln.find("(") == std::string::npos && + ln.find(")") == std::string::npos) { + err_str += "\n Missing ';' or type, here -> "; + err_str += ln; + } + } + + if (ln.find(')') == std::string::npos) { + err_str += "\nMissing ')', after '(' here -> "; + err_str += ln.substr(ln.find('(')); + } + } else { + if (ln.find("for") != std::string::npos || ln.find("while") != std::string::npos) { + err_str += "\nMissing '(', after \"for\", here -> "; + err_str += ln; + } + } + + if (ln.find('}') != std::string::npos && !kInBraces) { + if (!kInStruct && ln.find(';') == std::string::npos) { + err_str += "\nMismatched '}', here -> "; + err_str += ln; + } + } + + if (!ln.empty()) { + if (ln.find(';') == std::string::npos && ln.find('{') == std::string::npos && + ln.find('}') == std::string::npos && ln.find(')') == std::string::npos && + ln.find('(') == std::string::npos && ln.find(',') == std::string::npos) { + if (ln.size() <= 2) return err_str; + + err_str += "\nMissing ';', here -> "; + err_str += ln; + } + } + + return err_str; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +/** + * @brief C To Assembly mount-point. + */ + +///////////////////////////////////////////////////////////////////////////////////////// + +class AssemblyCCInterface final LC_ASSEMBLY_INTERFACE { + public: + explicit AssemblyCCInterface() = default; + ~AssemblyCCInterface() override = default; + + LIBCOMPILER_COPY_DEFAULT(AssemblyCCInterface); + + UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArch64x0; } + + Int32 CompileToFormat(std::string src, Int32 arch) override { + if (kCompilerFrontend == nullptr) return 1; + + /* @brief copy contents wihtout extension */ + std::string src_file = src.data(); + std::ifstream src_fp = std::ifstream(src_file, std::ios::in); + std::string dest; + + for (auto& ch : src_file) { + if (ch == '.') { + break; + } + + dest += ch; + } + + /* According to PEF ABI. */ + std::vector exts = kAsmFileExts; + dest += exts[4]; + + kState.fOutputAssembly = std::make_unique(dest); + + auto fmt = LibCompiler::current_date(); + + (*kState.fOutputAssembly) << "# Path: " << src_file << "\n"; + (*kState.fOutputAssembly) << "# Language: 64x0 Assembly (Generated from ANSI C)\n"; + (*kState.fOutputAssembly) << "# Date: " << fmt << "\n\n"; + + LibCompiler::SyntaxLeafList syntax; + + kState.fSyntaxTreeList.push_back(syntax); + kState.fSyntaxTree = &kState.fSyntaxTreeList[kState.fSyntaxTreeList.size() - 1]; + + std::string line_src; + + while (std::getline(src_fp, line_src)) { + if (auto err = kCompilerFrontend->Check(line_src.c_str(), src.data()); err.empty()) { + kCompilerFrontend->Compile(line_src, src.data()); + } else { + Detail::print_error(err, src.data()); + } + } + + if (kAcceptableErrors > 0) return 1; + + std::vector keywords = {"ldw", "stw", "lda", "sta", "add", "sub", "mv"}; + + /// + /// Replace, optimize, fix assembly output. + /// + + for (auto& leaf : kState.fSyntaxTree->fLeafList) { + std::vector access_keywords = {"->", "."}; + + for (auto& access_ident : access_keywords) { + if (LibCompiler::find_word(leaf.fUserValue, access_ident)) { + for (auto& struc : kState.kStructMap) { + /// TODO: + } + } + } + + for (auto& keyword : keywords) { + if (LibCompiler::find_word(leaf.fUserValue, keyword)) { + std::size_t cnt = 0UL; + + for (auto& reg : kState.kStackFrame) { + std::string needle; + + for (size_t i = 0; i < reg.fName.size(); i++) { + if (reg.fName[i] == ' ') { + ++i; + + for (; i < reg.fName.size(); i++) { + if (reg.fName[i] == ',') { + break; + } + + if (reg.fName[i] == ' ') continue; + + needle += reg.fName[i]; + } + + break; + } + } + + if (LibCompiler::find_word(leaf.fUserValue, needle)) { + if (leaf.fUserValue.find("extern_segment " + needle) != std::string::npos) { + std::string range = "extern_segment " + needle; + leaf.fUserValue.replace(leaf.fUserValue.find("extern_segment " + needle), + range.size(), needle); + } + + if (leaf.fUserValue.find("ldw r6") != std::string::npos) { + std::string::difference_type countComma = + std::count(leaf.fUserValue.begin(), leaf.fUserValue.end(), ','); + + if (countComma == 1) { + leaf.fUserValue.replace(leaf.fUserValue.find("ldw"), strlen("ldw"), "mv"); + } + } + + leaf.fUserValue.replace(leaf.fUserValue.find(needle), needle.size(), reg.fReg); + + ++cnt; + } + } + + if (cnt > 1 && keyword != "mv" && keyword != "add" && keyword != "sub") { + leaf.fUserValue.replace(leaf.fUserValue.find(keyword), keyword.size(), "mv"); + } + } + } + } + + for (auto& leaf : kState.fSyntaxTree->fLeafList) { + (*kState.fOutputAssembly) << leaf.fUserValue; + } + + kState.fSyntaxTree = nullptr; + + kState.fOutputAssembly->flush(); + kState.fOutputAssembly.reset(); + + return kExitOK; + } +}; + +///////////////////////////////////////////////////////////////////////////////////////// + +#include + +#define kPrintF printf +#define kSplashCxx() kPrintF(kWhite "NE C Driver, %s, (c) Amlal El Mahrouss\n", kDistVersion) + +static void cc_print_help() { + kSplashCxx(); +} + +///////////////////////////////////////////////////////////////////////////////////////// + +#define kExt ".c" + +LIBCOMPILER_MODULE(NeOSCompilerCLang64x0) { + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + kCompilerTypes.push_back({.fName = "void", .fValue = "void"}); + kCompilerTypes.push_back({.fName = "char", .fValue = "byte"}); + kCompilerTypes.push_back({.fName = "short", .fValue = "hword"}); + kCompilerTypes.push_back({.fName = "int", .fValue = "dword"}); + kCompilerTypes.push_back({.fName = "long", .fValue = "qword"}); + kCompilerTypes.push_back({.fName = "*", .fValue = "offset"}); + + bool skip = false; + + kFactory.Mount(new AssemblyCCInterface()); + kMachine = LibCompiler::AssemblyFactory::kArch64x0; + kCompilerFrontend = new CompilerFrontend64x0(); + + for (auto index = 1UL; index < argc; ++index) { + if (skip) { + skip = false; + continue; + } + + if (argv[index][0] == '-') { + if (strcmp(argv[index], "--v") == 0 || strcmp(argv[index], "--version") == 0) { + kSplashCxx(); + return kExitOK; + } + + if (strcmp(argv[index], "--verbose") == 0) { + kState.fVerbose = true; + + continue; + } + + if (strcmp(argv[index], "--h") == 0 || strcmp(argv[index], "--help") == 0) { + cc_print_help(); + + return kExitOK; + } + + if (strcmp(argv[index], "--dialect") == 0) { + if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; + + return kExitOK; + } + + if (strcmp(argv[index], "--fmax-exceptions") == 0) { + try { + kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); + } + // catch anything here + catch (...) { + kErrorLimit = 0; + } + + skip = true; + + continue; + } + + std::string err = "Unknown command: "; + err += argv[index]; + + Detail::print_error(err, "cc"); + + continue; + } + + kFileList.emplace_back(argv[index]); + + std::string srcFile = argv[index]; + + if (strstr(argv[index], kExt) == nullptr) { + if (kState.fVerbose) { + Detail::print_error(srcFile + " is not a valid C source.\n", "cc"); + } + + return 1; + } + + if (kFactory.Compile(srcFile, kMachine) != kExitOK) return 1; + } + + return kExitOK; +} + +// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/Cl/CCompilerARM64.cc b/dev/LibCompiler/src/Cl/CCompilerARM64.cc new file mode 100644 index 0000000..9abdc8c --- /dev/null +++ b/dev/LibCompiler/src/Cl/CCompilerARM64.cc @@ -0,0 +1,1284 @@ +/* + * ======================================================== + * + * cc + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +/// BUGS: 0 +/// TODO: none + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* C driver */ +/* This is part of the LibCompiler. */ +/* (c) Amlal El Mahrouss */ + +/// @author EL Mahrouss Amlal (amlel) +/// @file ARM64-cc.cxx +/// @brief ARM64 C Compiler. + +/// TODO: support structures, else if, else, . and -> + +///////////////////// + +// ANSI ESCAPE CODES + +///////////////////// + +#define kExitOK (0) + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" + +///////////////////////////////////// + +// INTERNAL STUFF OF THE C COMPILER + +///////////////////////////////////// + +namespace Detail { +// \brief Register map structure, used to keep track of each variable's registers. +struct CompilerRegisterMap final { + std::string fName; + std::string fReg; +}; + +// \brief Map for C structs +// \author amlel +struct CompilerStructMap final { + // 'my_foo' + std::string fName; + + // if instance: stores a valid register. + std::string fReg; + + // offset count + std::size_t fOffsetsCnt; + + // offset array. + std::vector> fOffsets; +}; + +struct CompilerState final { + std::vector fSyntaxTreeList; + std::vector kStackFrame; + std::vector kStructMap; + LibCompiler::SyntaxLeafList* fSyntaxTree{nullptr}; + std::unique_ptr fOutputAssembly; + std::string fLastFile; + std::string fLastError; + bool fVerbose; +}; +} // namespace Detail + +static Detail::CompilerState kState; +static std::string kIfFunction = ""; + +namespace Detail { +/// @brief prints an error into stdout. +/// @param reason the reason of the error. +/// @param file where does it originate from? +void print_error(std::string reason, std::string file) noexcept; + +struct CompilerType final { + std::string fName; + std::string fValue; +}; +} // namespace Detail + +///////////////////////////////////////////////////////////////////////////////////////// + +// Target architecture. +static int kMachine = 0; + +///////////////////////////////////////// + +// REGISTERS ACCORDING TO USED ASSEMBLER + +///////////////////////////////////////// + +static size_t kRegisterCnt = kAsmRegisterLimit; +static size_t kStartUsable = 8; +static size_t kUsableLimit = 15; +static size_t kRegisterCounter = kStartUsable; +static std::string kRegisterPrefix = kAsmRegisterPrefix; + +///////////////////////////////////////// + +// COMPILER PARSING UTILITIES/STATES. + +///////////////////////////////////////// + +static std::vector kFileList; +static LibCompiler::AssemblyFactory kFactory; +static bool kInStruct = false; +static bool kOnWhileLoop = false; +static bool kOnForLoop = false; +static bool kInBraces = false; +static bool kIfFound = false; +static size_t kBracesCount = 0UL; + +/* @brief C compiler backend for C */ +class CompilerFrontendARM64 final : public LibCompiler::CompilerFrontendInterface { + public: + explicit CompilerFrontendARM64() = default; + ~CompilerFrontendARM64() override = default; + + LIBCOMPILER_COPY_DEFAULT(CompilerFrontendARM64); + + std::string Check(const char* text, const char* file); + bool Compile(std::string text, std::string file) override; + + const char* Language() override { return "64k C"; } +}; + +static CompilerFrontendARM64* kCompilerFrontend = nullptr; +static std::vector kCompilerVariables; +static std::vector kCompilerFunctions; +static std::vector kCompilerTypes; + +namespace Detail { +union number_cast final { + public: + number_cast(UInt64 _Raw) : _Raw(_Raw) {} + + public: + char _Num[8]; + UInt64 _Raw; +}; + +union double_cast final { + public: + double_cast(float _Raw) : _Raw(_Raw) {} + + public: + char _Sign; + char _Lh[8]; + char _Rh[23]; + + float _Raw; +}; +} // namespace Detail + +///////////////////////////////////////////////////////////////////////////////////////// + +// @name Compile +// @brief Generate MASM from a C assignement. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool CompilerFrontendARM64::Compile(std::string text, std::string file) { + bool typeFound = false; + bool fnFound = false; + + // setup generator. + std::random_device rd; + + auto seed_data = std::array{}; + std::generate(std::begin(seed_data), std::end(seed_data), std::ref(rd)); + std::seed_seq seq(std::begin(seed_data), std::end(seed_data)); + std::mt19937 generator(seq); + + // start parsing + for (size_t text_index = 0; text_index < text.size(); ++text_index) { + auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); + + auto gen = uuids::uuid_random_generator{generator}; + uuids::uuid out = gen(); + + Detail::number_cast time_off = (UInt64) out.as_bytes().data(); + + if (!typeFound) { + auto substr = text.substr(text_index); + std::string match_type; + + for (size_t y = 0; y < substr.size(); ++y) { + if (substr[y] == ' ') { + while (match_type.find(' ') != std::string::npos) { + match_type.erase(match_type.find(' ')); + } + + for (auto& clType : kCompilerTypes) { + if (clType.fName == match_type) { + match_type.clear(); + + std::string buf; + + buf += clType.fValue; + buf += ' '; + + if (substr.find('=') != std::string::npos) { + break; + } + + if (text.find('(') != std::string::npos) { + syntaxLeaf.fUserValue = buf; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + } + + typeFound = true; + break; + } + } + + break; + } + + match_type += substr[y]; + } + } + + if (text[text_index] == '{') { + if (kInStruct) { + continue; + } + + kInBraces = true; + ++kBracesCount; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + } + + // return keyword handler + if (text[text_index] == 'r') { + std::string return_keyword; + return_keyword += "return"; + + std::size_t index = 0UL; + + std::string value; + + for (size_t return_index = text_index; return_index < text.size(); ++return_index) { + if (text[return_index] != return_keyword[index]) { + for (size_t value_index = return_index; value_index < text.size(); ++value_index) { + if (text[value_index] == ';') break; + + value += text[value_index]; + } + + break; + } + + ++index; + } + + if (index == return_keyword.size()) { + if (!value.empty()) { + if (value.find('(') != std::string::npos) { + value.erase(value.find('(')); + } + + if (!isdigit(value[value.find('(') + 2])) { + std::string tmp = value; + bool reg_to_reg = false; + + value.clear(); + + value += " extern_segment"; + value += tmp; + } + + syntaxLeaf.fUserValue = "\tldw r19, "; + + // make it pretty. + if (value.find('\t') != std::string::npos) value.erase(value.find('\t'), 1); + + syntaxLeaf.fUserValue += value + "\n"; + } + + syntaxLeaf.fUserValue += "\tjlr"; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + break; + } + } + + if (text[text_index] == 'i' && text[text_index + 1] == 'f') { + auto expr = text.substr(text_index + 2); + text.erase(text_index, 2); + + if (expr.find("{") != std::string::npos) { + expr.erase(expr.find("{")); + } + + if (expr.find("(") != std::string::npos) expr.erase(expr.find("(")); + + if (expr.find(")") != std::string::npos) expr.erase(expr.find(")")); + + kIfFunction = "__LIBCOMPILER_IF_PROC_"; + kIfFunction += std::to_string(time_off._Raw); + + syntaxLeaf.fUserValue = "\tlda r12, extern_segment "; + syntaxLeaf.fUserValue += kIfFunction + + "\n\t#r12 = Code to jump on, r11 right cond, r10 left cond.\n\tbeq " + "r10, r11, r12\ndword public_segment .code64 " + + kIfFunction + "\n"; + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + kIfFound = true; + } + + // Parse expressions and instructions here. + // what does this mean? + // we encounter an assignment, or we reached the end of an expression. + if (text[text_index] == '=' || text[text_index] == ';') { + if (fnFound) continue; + if (kIfFound) continue; + + if (text[text_index] == ';' && kInStruct) continue; + + if (text.find("typedef ") != std::string::npos) continue; + + if (text[text_index] == '=' && kInStruct) { + Detail::print_error("assignement of value in struct " + text, file); + continue; + } + + if (text[text_index] == ';' && kInStruct) { + bool space_found_ = false; + std::string sym; + + for (auto& ch : text) { + if (ch == ' ') { + space_found_ = true; + } + + if (ch == ';') break; + + if (space_found_) sym.push_back(ch); + } + + kState.kStructMap[kState.kStructMap.size() - 1].fOffsets.push_back( + std::make_pair(kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4, sym)); + + kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt = + kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4; + + continue; + } + + if (text[text_index] == '=' && kInStruct) { + continue; + } + + if (text[text_index + 1] == '=' || text[text_index - 1] == '!' || + text[text_index - 1] == '<' || text[text_index - 1] == '>') { + continue; + } + + std::string substr; + + if (text.find('=') != std::string::npos && kInBraces && !kIfFound) { + if (text.find("*") != std::string::npos) { + if (text.find("=") > text.find("*")) + substr += "\tlda "; + else + substr += "\tldw "; + } else { + substr += "\tldw "; + } + } else if (text.find('=') != std::string::npos && !kInBraces) { + substr += "stw public_segment .data64 "; + } + + int first_encountered = 0; + + std::string str_name; + + for (size_t text_index_2 = 0; text_index_2 < text.size(); ++text_index_2) { + if (text[text_index_2] == '\"') { + ++text_index_2; + + // want to add this, so that the parser recognizes that this is a + // string. + substr += '"'; + + for (; text_index_2 < text.size(); ++text_index_2) { + if (text[text_index_2] == '\"') break; + + substr += text[text_index_2]; + } + } + + if (text[text_index_2] == '{' || text[text_index_2] == '}') continue; + + if (text[text_index_2] == ';') { + break; + } + + if (text[text_index_2] == ' ' || text[text_index_2] == '\t') { + if (first_encountered != 2) { + if (text[text_index] != '=' && + substr.find("public_segment .data64") == std::string::npos && !kInStruct) + substr += "public_segment .data64 "; + } + + ++first_encountered; + + continue; + } + + if (text[text_index_2] == '=') { + if (!kInBraces) { + substr.replace(substr.find("public_segment .data64"), strlen("public_segment .data64"), + "public_segment .zero64 "); + } + + substr += ","; + continue; + } + + substr += text[text_index_2]; + } + + for (auto& clType : kCompilerTypes) { + if (substr.find(clType.fName) != std::string::npos) { + if (substr.find(clType.fName) > substr.find('"')) continue; + + substr.erase(substr.find(clType.fName), clType.fName.size()); + } else if (substr.find(clType.fValue) != std::string::npos) { + if (substr.find(clType.fValue) > substr.find('"')) continue; + + if (clType.fName == "const") continue; + + substr.erase(substr.find(clType.fValue), clType.fValue.size()); + } + } + + if (substr.find("extern") != std::string::npos) { + substr.replace(substr.find("extern"), strlen("extern"), "extern_segment "); + + if (substr.find("public_segment .data64") != std::string::npos) + substr.erase(substr.find("public_segment .data64"), strlen("public_segment .data64")); + } + + auto var_to_find = std::find_if( + kCompilerVariables.cbegin(), kCompilerVariables.cend(), + [&](Detail::CompilerType type) { return type.fName.find(substr) != std::string::npos; }); + + if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; + + std::string reg = kAsmRegisterPrefix; + reg += std::to_string(kRegisterCounter); + + if (var_to_find == kCompilerVariables.cend()) { + ++kRegisterCounter; + + kState.kStackFrame.push_back({.fName = substr, .fReg = reg}); + kCompilerVariables.push_back({.fName = substr}); + } + + syntaxLeaf.fUserValue += substr; + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + if (text[text_index] == '=') break; + } + + // function handler. + + if (text[text_index] == '(' && !fnFound && !kIfFound) { + std::string substr; + std::string args_buffer; + std::string args; + + bool type_crossed = false; + + for (size_t idx = text.find('(') + 1; idx < text.size(); ++idx) { + if (text[idx] == ',') continue; + + if (text[idx] == ' ') continue; + + if (text[idx] == ')') break; + } + + for (char substr_first_index : text) { + if (substr_first_index != ',') + args_buffer += substr_first_index; + else + args_buffer += '$'; + + if (substr_first_index == ';') { + args_buffer = args_buffer.erase(0, args_buffer.find('(')); + args_buffer = args_buffer.erase(args_buffer.find(';'), 1); + args_buffer = args_buffer.erase(args_buffer.find(')'), 1); + args_buffer = args_buffer.erase(args_buffer.find('('), 1); + + if (!args_buffer.empty()) args += "\tldw r6, "; + + std::string register_type; + std::size_t index = 7UL; + + while (args_buffer.find("$") != std::string::npos) { + register_type = kRegisterPrefix; + register_type += std::to_string(index); + + ++index; + + args_buffer.replace(args_buffer.find('$'), 1, "\n\tldw " + register_type + ","); + } + + args += args_buffer; + args += "\n\tlda r19, "; + } + } + + for (char _text_i : text) { + if (_text_i == '\t' || _text_i == ' ') { + if (!type_crossed) { + substr.clear(); + type_crossed = true; + } + + continue; + } + + if (_text_i == '(') break; + + substr += _text_i; + } + + if (kInBraces) { + syntaxLeaf.fUserValue = args; + syntaxLeaf.fUserValue += substr; + syntaxLeaf.fUserValue += "\n\tjrl\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + fnFound = true; + } else { + syntaxLeaf.fUserValue.clear(); + + syntaxLeaf.fUserValue += "public_segment .code64 "; + + syntaxLeaf.fUserValue += substr; + syntaxLeaf.fUserValue += "\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + fnFound = true; + } + + kCompilerFunctions.push_back(text); + } + + if (text[text_index] == '-' && text[text_index + 1] == '-') { + text = text.replace(text.find("--"), strlen("--"), ""); + + for (int _text_i = 0; _text_i < text.size(); ++_text_i) { + if (text[_text_i] == '\t' || text[_text_i] == ' ') text.erase(_text_i, 1); + } + + syntaxLeaf.fUserValue += "sub "; + syntaxLeaf.fUserValue += text; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + break; + } + + if (text[text_index] == '}') { + kRegisterCounter = kStartUsable; + + --kBracesCount; + + if (kBracesCount < 1) { + kInBraces = false; + kBracesCount = 0; + } + + if (kIfFound) kIfFound = false; + + if (kInStruct) kInStruct = false; + + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + } + + syntaxLeaf.fUserValue.clear(); + } + + auto syntaxLeaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); + syntaxLeaf.fUserValue = "\n"; + kState.fSyntaxTree->fLeafList.push_back(syntaxLeaf); + + return true; +} + +static bool kShouldHaveBraces = false; +static std::string kFnName; + +std::string CompilerFrontendARM64::Check(const char* text, const char* file) { + std::string err_str; + std::string ln = text; + + if (ln.empty()) { + return err_str; + } + + bool non_ascii_found = false; + + for (int i = 0; i < ln.size(); ++i) { + if (isalnum(ln[i])) { + non_ascii_found = true; + break; + } + } + + if (kShouldHaveBraces && ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + } + + if (!non_ascii_found) return err_str; + + size_t string_index = 1UL; + + if (ln.find('\'') != std::string::npos) { + string_index = ln.find('\'') + 1; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == '\'') { + if (ln[string_index + 1] != ';') { + ln.erase(string_index, 1); + } + + return err_str; + } + } + } else if (ln.find('"') != std::string::npos) { + string_index = ln.find('"') + 1; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == '"') { + if (ln[string_index + 1] != ';') { + ln.erase(string_index, 1); + } else { + break; + } + } + } + } else if (ln.find('"') == std::string::npos && ln.find('\'') == std::string::npos) { + std::vector forbidden_words; + + forbidden_words.push_back("\\"); + forbidden_words.push_back("?"); + forbidden_words.push_back("@"); + forbidden_words.push_back("~"); + forbidden_words.push_back("::"); + forbidden_words.push_back("--*"); + forbidden_words.push_back("*/"); + + // add them to avoid stupid mistakes. + forbidden_words.push_back("namespace"); + forbidden_words.push_back("class"); + forbidden_words.push_back("extern \"C\""); + + for (auto& forbidden : forbidden_words) { + if (ln.find(forbidden) != std::string::npos) { + err_str += "\nForbidden character detected: "; + err_str += forbidden; + + return err_str; + } + } + } + + struct CompilerVariableRange final { + std::string fBegin; + std::string fEnd; + }; + + const std::vector variables_list = { + {.fBegin = "static ", .fEnd = "="}, {.fBegin = "=", .fEnd = ";"}, + {.fBegin = "if(", .fEnd = "="}, {.fBegin = "if (", .fEnd = "="}, + {.fBegin = "if(", .fEnd = "<"}, {.fBegin = "if (", .fEnd = "<"}, + {.fBegin = "if(", .fEnd = ">"}, {.fBegin = "if (", .fEnd = ">"}, + {.fBegin = "if(", .fEnd = ")"}, {.fBegin = "if (", .fEnd = ")"}, + + {.fBegin = "else(", .fEnd = "="}, {.fBegin = "else (", .fEnd = "="}, + {.fBegin = "else(", .fEnd = "<"}, {.fBegin = "else (", .fEnd = "<"}, + {.fBegin = "else(", .fEnd = ">"}, {.fBegin = "else (", .fEnd = ">"}, + {.fBegin = "else(", .fEnd = ")"}, {.fBegin = "else (", .fEnd = ")"}, + }; + + for (auto& variable : variables_list) { + if (ln.find(variable.fBegin) != std::string::npos) { + string_index = ln.find(variable.fBegin) + variable.fBegin.size(); + + while (ln[string_index] == ' ') ++string_index; + + std::string keyword; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == variable.fEnd[0]) { + std::string varname = ""; + + for (size_t index_keyword = ln.find(' '); ln[index_keyword] != variable.fBegin[0]; + ++index_keyword) { + if (ln[index_keyword] == ' ') { + continue; + } + + if (isdigit(ln[index_keyword])) { + goto cc_next_loop; + } + + varname += ln[index_keyword]; + } + + if (varname.find(' ') != std::string::npos) { + varname.erase(0, varname.find(' ')); + + if (variable.fBegin == "extern") { + varname.erase(0, varname.find(' ')); + } + } + + if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; + + std::string reg = kAsmRegisterPrefix; + reg += std::to_string(kRegisterCounter); + + kCompilerVariables.push_back({.fValue = varname}); + goto cc_check_done; + } + + keyword.push_back(ln[string_index]); + } + + goto cc_next_loop; + + cc_check_done: + + // skip digit value. + if (isdigit(keyword[0]) || keyword[0] == '"') { + goto cc_next_loop; + } + + while (keyword.find(' ') != std::string::npos) keyword.erase(keyword.find(' '), 1); + + for (auto& var : kCompilerVariables) { + if (var.fValue.find(keyword) != std::string::npos) { + err_str.clear(); + goto cc_next; + } + } + + for (auto& fn : kCompilerFunctions) { + if (fn.find(keyword[0]) != std::string::npos) { + auto where_begin = fn.find(keyword[0]); + auto keyword_begin = 0UL; + auto failed = false; + + for (; where_begin < keyword.size(); ++where_begin) { + if (fn[where_begin] == '(' && keyword[keyword_begin] == '(') break; + + if (fn[where_begin] != keyword[keyword_begin]) { + failed = true; + break; + } + + ++keyword_begin; + } + + if (!failed) { + err_str.clear(); + goto cc_next; + } else { + continue; + } + } + } + + cc_error_value: + if (keyword.find("->") != std::string::npos) return err_str; + + if (keyword.find(".") != std::string::npos) return err_str; + + if (isalnum(keyword[0])) err_str += "\nUndefined value: " + keyword; + + return err_str; + } + + cc_next_loop: + continue; + } + +cc_next: + + // extern does not declare anything, it extern_segments a variable. + // so that's why it's not declare upper. + if (LibCompiler::find_word(ln, "extern")) { + auto substr = ln.substr(ln.find("extern") + strlen("extern")); + kCompilerVariables.push_back({.fValue = substr}); + } + + if (kShouldHaveBraces && ln.find('{') == std::string::npos) { + err_str += "Missing '{' for function "; + err_str += kFnName; + err_str += "\n"; + + kShouldHaveBraces = false; + kFnName.clear(); + } else if (kShouldHaveBraces && ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + kFnName.clear(); + } + + bool type_not_found = true; + + if (ln.find('\'') != std::string::npos) { + ln.replace(ln.find('\''), 3, "0"); + } + + auto first = ln.find('"'); + if (first != std::string::npos) { + auto second = 0UL; + bool found_second_quote = false; + + for (size_t i = first + 1; i < ln.size(); ++i) { + if (ln[i] == '\"') { + found_second_quote = true; + second = i; + + break; + } + } + + if (!found_second_quote) { + err_str += "Missing terminating \"."; + err_str += " here -> " + ln.substr(ln.find('"'), second); + } + } + + if (ln.find(')') != std::string::npos && ln.find(';') == std::string::npos) { + if (ln.find('{') == std::string::npos) { + kFnName = ln; + kShouldHaveBraces = true; + + goto skip_braces_check; + } else if (ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + } + } + +skip_braces_check: + + for (auto& key : kCompilerTypes) { + if (LibCompiler::find_word(ln, key.fName)) { + if (isdigit(ln[ln.find(key.fName) + key.fName.size() + 1])) { + err_str += "\nNumber cannot be set for "; + err_str += key.fName; + err_str += "'s name. here -> "; + err_str += ln; + } + + if (ln.find(key.fName) == 0 || ln[ln.find(key.fName) - 1] == ' ' || + ln[ln.find(key.fName) - 1] == '\t') { + type_not_found = false; + + if (ln[ln.find(key.fName) + key.fName.size()] != ' ') { + type_not_found = true; + + if (ln[ln.find(key.fName) + key.fName.size()] == '\t') type_not_found = false; + + goto next; + } else if (ln[ln.find(key.fName) + key.fName.size()] != '\t') { + type_not_found = true; + + if (ln[ln.find(key.fName) + key.fName.size()] == ' ') type_not_found = false; + } + } + + next: + + if (ln.find(';') == std::string::npos) { + if (ln.find('(') != std::string::npos) { + if (ln.find('=') == std::string::npos) continue; + } + + err_str += "\nMissing ';', here -> "; + err_str += ln; + } else { + continue; + } + + if (ln.find('=') != std::string::npos) { + if (ln.find('(') != std::string::npos) { + if (ln.find(')') == std::string::npos) { + err_str += "\nMissing ')', after '(' here -> "; + err_str += ln.substr(ln.find('(')); + } + } + } + } + } + + if (kInBraces && ln.find("struct") != std::string::npos && + ln.find("union") != std::string::npos && ln.find("enum") != std::string::npos && + ln.find('=') != std::string::npos) { + if (ln.find(';') == std::string::npos) { + err_str += "\nMissing ';' after struct/union/enum declaration, here -> "; + err_str += ln; + } + } + + if (ln.find(';') != std::string::npos && ln.find("for") == std::string::npos) { + if (ln.find(';') + 1 != ln.size()) { + for (int i = 0; i < ln.substr(ln.find(';') + 1).size(); ++i) { + if ((ln.substr(ln.find(';') + 1)[i] != ' ') || (ln.substr(ln.find(';') + 1)[i] != '\t')) { + if (auto err = this->Check(ln.substr(ln.find(';') + 1).c_str(), file); !err.empty()) { + err_str += "\nUnexpected text after ';' -> "; + err_str += ln.substr(ln.find(';')); + err_str += err; + } + } + } + } + } + + if (ln.find('(') != std::string::npos) { + if (ln.find(';') == std::string::npos && !LibCompiler::find_word(ln, "|") && + !LibCompiler::find_word(ln, "||") && !LibCompiler::find_word(ln, "&") && + !LibCompiler::find_word(ln, "&&") && !LibCompiler::find_word(ln, "~")) { + bool found_func = false; + size_t i = ln.find('('); + std::vector opens; + std::vector closes; + + for (; i < ln.size(); ++i) { + if (ln[i] == ')') { + closes.push_back(1); + } + + if (ln[i] == '(') { + opens.push_back(1); + } + } + + if (closes.size() != opens.size()) err_str += "Unterminated (), here -> " + ln; + + bool space_found = false; + + for (int i = 0; i < ln.size(); ++i) { + if (ln[i] == ')' && !space_found) { + space_found = true; + continue; + } + + if (space_found) { + if (ln[i] == ' ' && isalnum(ln[i + 1])) { + err_str += "\nBad function format here -> "; + err_str += ln; + } + } + } + } + + if (ln.find('(') < 1) { + err_str += "\nMissing identifier before '(' here -> "; + err_str += ln; + } else { + if (type_not_found && ln.find(';') == std::string::npos && + ln.find("if") == std::string::npos && ln.find("|") == std::string::npos && + ln.find("&") == std::string::npos && ln.find("(") == std::string::npos && + ln.find(")") == std::string::npos) { + err_str += "\n Missing ';' or type, here -> "; + err_str += ln; + } + } + + if (ln.find(')') == std::string::npos) { + err_str += "\nMissing ')', after '(' here -> "; + err_str += ln.substr(ln.find('(')); + } + } else { + if (ln.find("for") != std::string::npos || ln.find("while") != std::string::npos) { + err_str += "\nMissing '(', after \"for\", here -> "; + err_str += ln; + } + } + + if (ln.find('}') != std::string::npos && !kInBraces) { + if (!kInStruct && ln.find(';') == std::string::npos) { + err_str += "\nMismatched '}', here -> "; + err_str += ln; + } + } + + if (!ln.empty()) { + if (ln.find(';') == std::string::npos && ln.find('{') == std::string::npos && + ln.find('}') == std::string::npos && ln.find(')') == std::string::npos && + ln.find('(') == std::string::npos && ln.find(',') == std::string::npos) { + if (ln.size() <= 2) return err_str; + + err_str += "\nMissing ';', here -> "; + err_str += ln; + } + } + + return err_str; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +/** + * @brief C To Assembly mount-point. + */ + +///////////////////////////////////////////////////////////////////////////////////////// + +class AssemblyCCInterface final LC_ASSEMBLY_INTERFACE { + public: + explicit AssemblyCCInterface() = default; + ~AssemblyCCInterface() override = default; + + LIBCOMPILER_COPY_DEFAULT(AssemblyCCInterface); + + UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArchAARCH64; } + + Int32 CompileToFormat(std::string src, Int32 arch) override { + if (kCompilerFrontend == nullptr) return 1; + + /* @brief copy contents wihtout extension */ + std::string src_file = src.data(); + std::ifstream src_fp = std::ifstream(src_file, std::ios::in); + std::string dest; + + for (auto& ch : src_file) { + if (ch == '.') { + break; + } + + dest += ch; + } + + /* According to PEF ABI. */ + std::vector exts = kAsmFileExts; + dest += exts[4]; + + kState.fOutputAssembly = std::make_unique(dest); + + auto fmt = LibCompiler::current_date(); + + (*kState.fOutputAssembly) << "# Path: " << src_file << "\n"; + (*kState.fOutputAssembly) << "# Language: ARM64 Assembly (Generated from ANSI C)\n"; + (*kState.fOutputAssembly) << "# Date: " << fmt << "\n\n"; + + LibCompiler::SyntaxLeafList syntax; + + kState.fSyntaxTreeList.push_back(syntax); + kState.fSyntaxTree = &kState.fSyntaxTreeList[kState.fSyntaxTreeList.size() - 1]; + + std::string line_src; + + while (std::getline(src_fp, line_src)) { + if (auto err = kCompilerFrontend->Check(line_src.c_str(), src.data()); err.empty()) { + kCompilerFrontend->Compile(line_src, src.data()); + } else { + Detail::print_error(err, src.data()); + } + } + + if (kAcceptableErrors > 0) return 1; + + std::vector keywords = {"ldw", "stw", "lda", "sta", "add", "sub", "mv"}; + + /// + /// Replace, optimize, fix assembly output. + /// + + for (auto& leaf : kState.fSyntaxTree->fLeafList) { + std::vector access_keywords = {"->", "."}; + + for (auto& access_ident : access_keywords) { + if (LibCompiler::find_word(leaf.fUserValue, access_ident)) { + for (auto& struc : kState.kStructMap) { + /// TODO: + } + } + } + + for (auto& keyword : keywords) { + if (LibCompiler::find_word(leaf.fUserValue, keyword)) { + std::size_t cnt = 0UL; + + for (auto& reg : kState.kStackFrame) { + std::string needle; + + for (size_t i = 0; i < reg.fName.size(); i++) { + if (reg.fName[i] == ' ') { + ++i; + + for (; i < reg.fName.size(); i++) { + if (reg.fName[i] == ',') { + break; + } + + if (reg.fName[i] == ' ') continue; + + needle += reg.fName[i]; + } + + break; + } + } + + if (LibCompiler::find_word(leaf.fUserValue, needle)) { + if (leaf.fUserValue.find("extern_segment " + needle) != std::string::npos) { + std::string range = "extern_segment " + needle; + leaf.fUserValue.replace(leaf.fUserValue.find("extern_segment " + needle), + range.size(), needle); + } + + if (leaf.fUserValue.find("ldw r6") != std::string::npos) { + std::string::difference_type countComma = + std::count(leaf.fUserValue.begin(), leaf.fUserValue.end(), ','); + + if (countComma == 1) { + leaf.fUserValue.replace(leaf.fUserValue.find("ldw"), strlen("ldw"), "mv"); + } + } + + leaf.fUserValue.replace(leaf.fUserValue.find(needle), needle.size(), reg.fReg); + + ++cnt; + } + } + + if (cnt > 1 && keyword != "mv" && keyword != "add" && keyword != "sub") { + leaf.fUserValue.replace(leaf.fUserValue.find(keyword), keyword.size(), "mv"); + } + } + } + } + + for (auto& leaf : kState.fSyntaxTree->fLeafList) { + (*kState.fOutputAssembly) << leaf.fUserValue; + } + + kState.fSyntaxTree = nullptr; + + kState.fOutputAssembly->flush(); + kState.fOutputAssembly.reset(); + + return kExitOK; + } +}; + +///////////////////////////////////////////////////////////////////////////////////////// + +#include + +#define kPrintF printf +#define kSplashCxx() kPrintF(kWhite "NE C Driver, %s, (c) Amlal El Mahrouss\n", kDistVersion) + +static void cc_print_help() { + kSplashCxx(); +} + +///////////////////////////////////////////////////////////////////////////////////////// + +#define kCExtension ".c" + +LIBCOMPILER_MODULE(NeOSCompilerCLangARM64) { + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + kCompilerTypes.push_back({.fName = "void", .fValue = "void"}); + kCompilerTypes.push_back({.fName = "char", .fValue = "byte"}); + kCompilerTypes.push_back({.fName = "short", .fValue = "hword"}); + kCompilerTypes.push_back({.fName = "int", .fValue = "dword"}); + kCompilerTypes.push_back({.fName = "long", .fValue = "qword"}); + kCompilerTypes.push_back({.fName = "*", .fValue = "offset"}); + + bool skip = false; + + kFactory.Mount(new AssemblyCCInterface()); + kMachine = LibCompiler::AssemblyFactory::kArchAARCH64; + kCompilerFrontend = new CompilerFrontendARM64(); + + for (auto index = 1UL; index < argc; ++index) { + if (skip) { + skip = false; + continue; + } + + if (argv[index][0] == '-') { + if (strcmp(argv[index], "--v") == 0 || strcmp(argv[index], "--version") == 0) { + kSplashCxx(); + return kExitOK; + } + + if (strcmp(argv[index], "--verbose") == 0) { + kState.fVerbose = true; + + continue; + } + + if (strcmp(argv[index], "--h") == 0 || strcmp(argv[index], "--help") == 0) { + cc_print_help(); + + return kExitOK; + } + + if (strcmp(argv[index], "--dialect") == 0) { + if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; + + return kExitOK; + } + + if (strcmp(argv[index], "--fmax-exceptions") == 0) { + try { + kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); + } + // catch anything here + catch (...) { + kErrorLimit = 0; + } + + skip = true; + + continue; + } + + std::string err = "Unknown command: "; + err += argv[index]; + + Detail::print_error(err, "cc"); + + continue; + } + + kFileList.emplace_back(argv[index]); + + std::string srcFile = argv[index]; + + if (strstr(argv[index], kCExtension) == nullptr) { + if (kState.fVerbose) { + Detail::print_error(srcFile + " is not a valid C source.\n", "cc"); + } + + return 1; + } + + if (kFactory.Compile(srcFile, kMachine) != kExitOK) return 1; + } + + return kExitOK; +} + +// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/Cl/CCompilerPower64.cc b/dev/LibCompiler/src/Cl/CCompilerPower64.cc new file mode 100644 index 0000000..d3dd81e --- /dev/null +++ b/dev/LibCompiler/src/Cl/CCompilerPower64.cc @@ -0,0 +1,1303 @@ +/* + * ======================================================== + * + * CompilerPower64 + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define kExitOK 0 + +/// @author EL Mahrouss Amlal (amlal@nekernel.org) +/// @file cc.cxx +/// @brief POWER64 C Compiler. + +///////////////////// + +/// ANSI ESCAPE CODES + +///////////////////// + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" + +///////////////////////////////////// + +/// INTERNAL STRUCTURES OF THE C COMPILER + +///////////////////////////////////// + +namespace Detail { +// \brief name to register struct. +struct CompilerRegisterMap final { + std::string fName; + std::string fReg; +}; + +// \brief Map for C structs +// \author amlel +struct CompilerStructMap final { + /// 'struct::my_foo' + std::string fName; + + /// if instance: stores a valid register. + std::string fReg; + + /// offset count + std::size_t fOffsetsCnt; + + /// offset array. + std::vector> fOffsets; +}; + +struct CompilerState final { + std::vector fSyntaxTreeList; + std::vector kStackFrame; + std::vector kStructMap; + LibCompiler::SyntaxLeafList* fSyntaxTree{nullptr}; + std::unique_ptr fOutputAssembly; + std::string fLastFile; + std::string fLastError; + bool fVerbose; +}; +} // namespace Detail + +static Detail::CompilerState kState; +static std::string kIfFunction = ""; + +namespace Detail { +/// @brief prints an error into stdout. +/// @param reason the reason of the error. +/// @param file where does it originate from? +void print_error(std::string reason, std::string file) noexcept; + +struct CompilerType final { + std::string fName; + std::string fValue; +}; +} // namespace Detail + +///////////////////////////////////////////////////////////////////////////////////////// + +// Target architecture. +static int kMachine = 0; + +///////////////////////////////////////// + +// REGISTERS ACCORDING TO USED ASSEMBLER + +///////////////////////////////////////// + +static size_t kRegisterCnt = kAsmRegisterLimit; +static size_t kStartUsable = 2; +static size_t kUsableLimit = 15; +static size_t kRegisterCounter = kStartUsable; +static std::string kRegisterPrefix = kAsmRegisterPrefix; + +///////////////////////////////////////// + +// COMPILER PARSING UTILITIES/STATES. + +///////////////////////////////////////// + +static std::vector kFileList; +static LibCompiler::AssemblyFactory kFactory; +static bool kInStruct = false; +static bool kOnWhileLoop = false; +static bool kOnForLoop = false; +static bool kInBraces = false; +static bool kIfFound = false; +static size_t kBracesCount = 0UL; + +/* @brief C compiler backend for C */ +class CompilerFrontendPower64 final : public LibCompiler::CompilerFrontendInterface { + public: + explicit CompilerFrontendPower64() = default; + ~CompilerFrontendPower64() override = default; + + LIBCOMPILER_COPY_DEFAULT(CompilerFrontendPower64); + + std::string Check(const char* text, const char* file); + bool Compile(std::string text, std::string file) override; + + const char* Language() override { return "POWER C"; } +}; + +static CompilerFrontendPower64* kCompilerFrontend = nullptr; +static std::vector kCompilerVariables; +static std::vector kCompilerFunctions; +static std::vector kCompilerTypes; + +namespace Detail { +union number_cast final { + public: + number_cast(UInt64 _Raw) : _Raw(_Raw) {} + + public: + char _Num[8]; + UInt64 _Raw; +}; + +union double_cast final { + public: + double_cast(float _Raw) : _Raw(_Raw) {} + + public: + char _Sign; + char _Lh[8]; + char _Rh[23]; + + float _Raw; +}; +} // namespace Detail + +///////////////////////////////////////////////////////////////////////////////////////// + +// @name Compile +// @brief Generate MASM from a C assignement. + +///////////////////////////////////////////////////////////////////////////////////////// + +bool CompilerFrontendPower64::Compile(std::string text_, std::string file) { + std::string text = text_; + + bool typeFound = false; + bool fnFound = false; + + // setup generator. + std::random_device rd; + + auto seed_data = std::array{}; + std::generate(std::begin(seed_data), std::end(seed_data), std::ref(rd)); + std::seed_seq seq(std::begin(seed_data), std::end(seed_data)); + std::mt19937 generator(seq); + + // start parsing + for (size_t text_index = 0; text_index < text.size(); ++text_index) { + auto syntax_leaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); + + auto gen = uuids::uuid_random_generator{generator}; + uuids::uuid out = gen(); + + Detail::number_cast time_off = (UInt64) out.as_bytes().data(); + + if (!typeFound) { + auto substr = text.substr(text_index); + std::string match_type; + + for (size_t y = 0; y < substr.size(); ++y) { + if (substr[y] == ' ') { + while (match_type.find(' ') != std::string::npos) { + match_type.erase(match_type.find(' ')); + } + + for (auto& clType : kCompilerTypes) { + if (clType.fName == match_type) { + match_type.clear(); + + std::string buf; + + buf += clType.fValue; + buf += ' '; + + if (substr.find('=') != std::string::npos) { + break; + } + + if (text.find('(') != std::string::npos) { + syntax_leaf.fUserValue = buf; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + } + + typeFound = true; + break; + } + } + + break; + } + + match_type += substr[y]; + } + } + + if (text[text_index] == '{') { + if (kInStruct) { + continue; + } + + kInBraces = true; + ++kBracesCount; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + } + + // return keyword handler + if (text[text_index] == 'r') { + std::string return_keyword; + return_keyword += "return"; + + std::size_t index = 0UL; + + std::string value; + + for (size_t return_index = text_index; return_index < text.size(); ++return_index) { + if (text[return_index] != return_keyword[index]) { + for (size_t value_index = return_index; value_index < text.size(); ++value_index) { + if (text[value_index] == ';') break; + + value += text[value_index]; + } + + break; + } + + ++index; + } + + if (index == return_keyword.size()) { + if (!value.empty()) { + if (value.find('(') != std::string::npos) { + value.erase(value.find('(')); + } + + if (!isdigit(value[value.find('(') + 2])) { + std::string tmp = value; + bool reg_to_reg = false; + + value.clear(); + + value += " extern_segment"; + value += tmp; + } + + syntax_leaf.fUserValue = "\tmr r31, "; + + // make it pretty. + while (value.find('\t') != std::string::npos) value.erase(value.find('\t'), 1); + + while (value.find(' ') != std::string::npos) value.erase(value.find(' '), 1); + + while (value.find("extern_segment") != std::string::npos) + value.erase(value.find("extern_segment"), strlen("extern_segment")); + + bool found = false; + + for (auto& reg : kState.kStackFrame) { + if (value.find(reg.fName) != std::string::npos) { + found = true; + syntax_leaf.fUserValue += reg.fReg; + break; + } + } + + if (!found) syntax_leaf.fUserValue += "r0"; + } + + syntax_leaf.fUserValue += "\n\tblr"; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + + break; + } + } + + if (text[text_index] == 'i' && text[text_index + 1] == 'f') { + auto expr = text.substr(text_index + 2); + text.erase(text_index, 2); + + if (expr.find("{") != std::string::npos) { + expr.erase(expr.find("{")); + } + + if (expr.find("(") != std::string::npos) expr.erase(expr.find("(")); + + if (expr.find(")") != std::string::npos) expr.erase(expr.find(")")); + + kIfFunction = "__LIBCOMPILER_IF_PROC_"; + kIfFunction += std::to_string(time_off._Raw); + + syntax_leaf.fUserValue = + "\tcmpw " + "r10, r11"; + + syntax_leaf.fUserValue += "\n\tbeq extern_segment " + kIfFunction + + " \ndword public_segment .code64 " + kIfFunction + "\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + + kIfFound = true; + } + + // Parse expressions and instructions here. + // what does this mean? + // we encounter an assignment, or we reached the end of an expression. + if (text[text_index] == '=' || text[text_index] == ';') { + if (fnFound) continue; + if (kIfFound) continue; + + if (text[text_index] == ';' && kInStruct) continue; + + if (text.find("typedef ") != std::string::npos) continue; + + if (text[text_index] == '=' && kInStruct) { + Detail::print_error("assignement of value inside a struct " + text, file); + continue; + } + + if (text[text_index] == ';' && kInStruct) { + bool space_found_ = false; + std::string sym; + + for (auto& ch : text) { + if (ch == ' ') { + space_found_ = true; + } + + if (ch == ';') break; + + if (space_found_) sym.push_back(ch); + } + + kState.kStructMap[kState.kStructMap.size() - 1].fOffsets.push_back( + std::make_pair(kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4, sym)); + + kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt = + kState.kStructMap[kState.kStructMap.size() - 1].fOffsetsCnt + 4; + + continue; + } + + if (text[text_index] == '=' && kInStruct) { + continue; + } + + if (text[text_index + 1] == '=' || text[text_index - 1] == '!' || + text[text_index - 1] == '<' || text[text_index - 1] == '>') { + continue; + } + + std::string substr; + + if (text.find('=') != std::string::npos && kInBraces && !kIfFound) { + if (text.find("*") != std::string::npos) { + if (text.find("=") > text.find("*")) + substr += "\tli "; + else + substr += "\tli "; + } else { + substr += "\tli "; + } + } else if (text.find('=') != std::string::npos && !kInBraces) { + substr += "stw public_segment .data64 "; + } + + int first_encountered = 0; + + std::string str_name; + + for (size_t text_index_2 = 0; text_index_2 < text.size(); ++text_index_2) { + if (text[text_index_2] == '\"') { + ++text_index_2; + + // want to add this, so that the parser recognizes that this is a + // string. + substr += '"'; + + for (; text_index_2 < text.size(); ++text_index_2) { + if (text[text_index_2] == '\"') break; + + substr += text[text_index_2]; + } + } + + if (text[text_index_2] == '{' || text[text_index_2] == '}') continue; + + if (text[text_index_2] == ';') { + break; + } + + if (text[text_index_2] == ' ' || text[text_index_2] == '\t') { + if (first_encountered != 2) { + if (text[text_index] != '=' && + substr.find("public_segment .data64") == std::string::npos && !kInStruct) + substr += "public_segment .data64 "; + } + + ++first_encountered; + + continue; + } + + if (text[text_index_2] == '=') { + if (!kInBraces) { + substr.replace(substr.find("public_segment .data64"), strlen("public_segment .data64"), + "public_segment .zero64 "); + } + + substr += ","; + continue; + } + + substr += text[text_index_2]; + } + + for (auto& clType : kCompilerTypes) { + if (substr.find(clType.fName) != std::string::npos) { + if (substr.find(clType.fName) > substr.find('"')) continue; + + substr.erase(substr.find(clType.fName), clType.fName.size()); + } else if (substr.find(clType.fValue) != std::string::npos) { + if (substr.find(clType.fValue) > substr.find('"')) continue; + + if (clType.fName == "const") continue; + + substr.erase(substr.find(clType.fValue), clType.fValue.size()); + } + } + + if (substr.find("extern") != std::string::npos) { + substr.replace(substr.find("extern"), strlen("extern"), "extern_segment "); + + if (substr.find("public_segment .data64") != std::string::npos) + substr.erase(substr.find("public_segment .data64"), strlen("public_segment .data64")); + } + + auto var_to_find = std::find_if( + kCompilerVariables.cbegin(), kCompilerVariables.cend(), + [&](Detail::CompilerType type) { return type.fName.find(substr) != std::string::npos; }); + + kCompilerVariables.push_back({.fName = substr}); + + if (text[text_index] == ';') break; + + std::string reg = kAsmRegisterPrefix; + + ++kRegisterCounter; + reg += std::to_string(kRegisterCounter); + + auto newSubstr = substr.substr(substr.find(" ")); + + std::string symbol; + + for (size_t start = 0; start < newSubstr.size(); ++start) { + if (newSubstr[start] == ',') break; + + if (newSubstr[start] == ' ') continue; + + symbol += (newSubstr[start]); + } + + kState.kStackFrame.push_back({.fName = symbol, .fReg = reg}); + + syntax_leaf.fUserValue += "\n\tli " + reg + substr.substr(substr.find(',')); + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + } + + // function handler. + + if (text[text_index] == '(' && !fnFound && !kIfFound) { + std::string substr; + std::string args_buffer; + std::string args; + + bool type_crossed = false; + + for (size_t idx = text.find('(') + 1; idx < text.size(); ++idx) { + if (text[idx] == ',') continue; + + if (text[idx] == ' ') continue; + + if (text[idx] == ')') break; + } + + for (char substr_first_index : text) { + if (substr_first_index != ',') + args_buffer += substr_first_index; + else + args_buffer += '$'; + + if (substr_first_index == ';') { + args_buffer = args_buffer.erase(0, args_buffer.find('(')); + args_buffer = args_buffer.erase(args_buffer.find(';'), 1); + args_buffer = args_buffer.erase(args_buffer.find(')'), 1); + args_buffer = args_buffer.erase(args_buffer.find('('), 1); + + if (!args_buffer.empty()) args += "\tldw r6, "; + + std::string register_type; + std::size_t index = 7UL; + + while (args_buffer.find("$") != std::string::npos) { + register_type = kRegisterPrefix; + register_type += std::to_string(index); + + ++index; + + args_buffer.replace(args_buffer.find('$'), 1, "\n\tldw " + register_type + ","); + } + + args += args_buffer; + args += "\n\tli r31, "; + } + } + + for (char _text_i : text) { + if (_text_i == '\t' || _text_i == ' ') { + if (!type_crossed) { + substr.clear(); + type_crossed = true; + } + + continue; + } + + if (_text_i == '(') break; + + substr += _text_i; + } + + if (kInBraces) { + syntax_leaf.fUserValue = args; + syntax_leaf.fUserValue += substr; + + syntax_leaf.fUserValue += "\n\tblr\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + + fnFound = true; + } else { + syntax_leaf.fUserValue.clear(); + + syntax_leaf.fUserValue += "public_segment .code64 "; + + syntax_leaf.fUserValue += substr; + syntax_leaf.fUserValue += "\n"; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + + fnFound = true; + } + + kCompilerFunctions.push_back(text); + } + + if (text[text_index] == '-' && text[text_index + 1] == '-') { + text = text.replace(text.find("--"), strlen("--"), ""); + + for (int _text_i = 0; _text_i < text.size(); ++_text_i) { + if (text[_text_i] == '\t' || text[_text_i] == ' ') text.erase(_text_i, 1); + } + + syntax_leaf.fUserValue += "dec "; + syntax_leaf.fUserValue += text; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + break; + } + + if (text[text_index] == '}') { + kRegisterCounter = kStartUsable; + + --kBracesCount; + + if (kBracesCount < 1) { + kInBraces = false; + kBracesCount = 0; + } + + if (kIfFound) kIfFound = false; + + if (kInStruct) kInStruct = false; + + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + } + + syntax_leaf.fUserValue.clear(); + } + + auto syntax_leaf = LibCompiler::SyntaxLeafList::SyntaxLeaf(); + syntax_leaf.fUserValue = "\n"; + kState.fSyntaxTree->fLeafList.push_back(syntax_leaf); + + return true; +} + +static bool kShouldHaveBraces = false; +static std::string kFnName; + +std::string CompilerFrontendPower64::Check(const char* text, const char* file) { + std::string err_str; + std::string ln = text; + + if (ln.empty()) { + return err_str; + } + + bool non_ascii_found = false; + + for (int i = 0; i < ln.size(); ++i) { + if (isalnum(ln[i])) { + non_ascii_found = true; + break; + } + } + + if (kShouldHaveBraces && ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + } + + if (!non_ascii_found) return err_str; + + size_t string_index = 1UL; + + if (ln.find('\'') != std::string::npos) { + string_index = ln.find('\'') + 1; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == '\'') { + if (ln[string_index + 1] != ';') { + ln.erase(string_index, 1); + } + + return err_str; + } + } + } else if (ln.find('"') != std::string::npos) { + string_index = ln.find('"') + 1; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == '"') { + if (ln[string_index + 1] != ';') { + ln.erase(string_index, 1); + } else { + break; + } + } + } + } else if (ln.find('"') == std::string::npos && ln.find('\'') == std::string::npos) { + std::vector forbidden_words; + + forbidden_words.push_back("\\"); + forbidden_words.push_back("?"); + forbidden_words.push_back("@"); + forbidden_words.push_back("~"); + forbidden_words.push_back("::"); + forbidden_words.push_back("--*"); + forbidden_words.push_back("*/"); + + // add them to avoid stupid mistakes. + forbidden_words.push_back("namespace"); + forbidden_words.push_back("class"); + forbidden_words.push_back("extern \"C\""); + + for (auto& forbidden : forbidden_words) { + if (ln.find(forbidden) != std::string::npos) { + err_str += "\nForbidden character detected: "; + err_str += forbidden; + + return err_str; + } + } + } + + struct CompilerVariableRange final { + std::string fBegin; + std::string fEnd; + }; + + const std::vector variables_list = { + {.fBegin = "static ", .fEnd = "="}, {.fBegin = "=", .fEnd = ";"}, + {.fBegin = "if(", .fEnd = "="}, {.fBegin = "if (", .fEnd = "="}, + {.fBegin = "if(", .fEnd = "<"}, {.fBegin = "if (", .fEnd = "<"}, + {.fBegin = "if(", .fEnd = ">"}, {.fBegin = "if (", .fEnd = ">"}, + {.fBegin = "if(", .fEnd = ")"}, {.fBegin = "if (", .fEnd = ")"}, + + {.fBegin = "else(", .fEnd = "="}, {.fBegin = "else (", .fEnd = "="}, + {.fBegin = "else(", .fEnd = "<"}, {.fBegin = "else (", .fEnd = "<"}, + {.fBegin = "else(", .fEnd = ">"}, {.fBegin = "else (", .fEnd = ">"}, + {.fBegin = "else(", .fEnd = ")"}, {.fBegin = "else (", .fEnd = ")"}, + }; + + for (auto& variable : variables_list) { + if (ln.find(variable.fBegin) != std::string::npos) { + string_index = ln.find(variable.fBegin) + variable.fBegin.size(); + + while (ln[string_index] == ' ') ++string_index; + + std::string keyword; + + for (; string_index < ln.size(); ++string_index) { + if (ln[string_index] == variable.fEnd[0]) { + std::string varname = ""; + + for (size_t index_keyword = ln.find(' '); ln[index_keyword] != variable.fBegin[0]; + ++index_keyword) { + if (ln[index_keyword] == ' ') { + continue; + } + + if (isdigit(ln[index_keyword])) { + goto cc_next_loop; + } + + varname += ln[index_keyword]; + } + + if (varname.find(' ') != std::string::npos) { + varname.erase(0, varname.find(' ')); + + if (variable.fBegin == "extern") { + varname.erase(0, varname.find(' ')); + } + } + + if (kRegisterCounter == 5 || kRegisterCounter == 6) ++kRegisterCounter; + + std::string reg = kAsmRegisterPrefix; + reg += std::to_string(kRegisterCounter); + + kCompilerVariables.push_back({.fValue = varname}); + goto cc_check_done; + } + + keyword.push_back(ln[string_index]); + } + + goto cc_next_loop; + + cc_check_done: + + // skip digit value. + if (isdigit(keyword[0]) || keyword[0] == '"') { + goto cc_next_loop; + } + + while (keyword.find(' ') != std::string::npos) keyword.erase(keyword.find(' '), 1); + + for (auto& var : kCompilerVariables) { + if (var.fValue.find(keyword) != std::string::npos) { + err_str.clear(); + goto cc_next; + } + } + + for (auto& fn : kCompilerFunctions) { + if (fn.find(keyword[0]) != std::string::npos) { + auto where_begin = fn.find(keyword[0]); + auto keyword_begin = 0UL; + auto failed = false; + + for (; where_begin < keyword.size(); ++where_begin) { + if (fn[where_begin] == '(' && keyword[keyword_begin] == '(') break; + + if (fn[where_begin] != keyword[keyword_begin]) { + failed = true; + break; + } + + ++keyword_begin; + } + + if (!failed) { + err_str.clear(); + goto cc_next; + } else { + continue; + } + } + } + + cc_error_value: + if (keyword.find("->") != std::string::npos) return err_str; + + if (keyword.find(".") != std::string::npos) return err_str; + + if (isalnum(keyword[0])) err_str += "\nUndefined value: " + keyword; + + return err_str; + } + + cc_next_loop: + continue; + } + +cc_next: + + // extern does not declare anything, it extern_segments a variable. + // so that's why it's not declare upper. + if (LibCompiler::find_word(ln, "extern")) { + auto substr = ln.substr(ln.find("extern") + strlen("extern")); + kCompilerVariables.push_back({.fValue = substr}); + } + + if (kShouldHaveBraces && ln.find('{') == std::string::npos) { + err_str += "Missing '{' for function "; + err_str += kFnName; + err_str += "\n"; + + kShouldHaveBraces = false; + kFnName.clear(); + } else if (kShouldHaveBraces && ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + kFnName.clear(); + } + + bool type_not_found = true; + + if (ln.find('\'') != std::string::npos) { + ln.replace(ln.find('\''), 3, "0"); + } + + auto first = ln.find('"'); + if (first != std::string::npos) { + auto second = 0UL; + bool found_second_quote = false; + + for (size_t i = first + 1; i < ln.size(); ++i) { + if (ln[i] == '\"') { + found_second_quote = true; + second = i; + + break; + } + } + + if (!found_second_quote) { + err_str += "Missing terminating \"."; + err_str += " here -> " + ln.substr(ln.find('"'), second); + } + } + + if (ln.find(')') != std::string::npos && ln.find(';') == std::string::npos) { + if (ln.find('{') == std::string::npos) { + kFnName = ln; + kShouldHaveBraces = true; + + goto skip_braces_check; + } else if (ln.find('{') != std::string::npos) { + kShouldHaveBraces = false; + } + } + +skip_braces_check: + + for (auto& key : kCompilerTypes) { + if (LibCompiler::find_word(ln, key.fName)) { + if (isdigit(ln[ln.find(key.fName) + key.fName.size() + 1])) { + err_str += "\nNumber cannot be set for "; + err_str += key.fName; + err_str += "'s name. here -> "; + err_str += ln; + } + + if (ln.find(key.fName) == 0 || ln[ln.find(key.fName) - 1] == ' ' || + ln[ln.find(key.fName) - 1] == '\t') { + type_not_found = false; + + if (ln[ln.find(key.fName) + key.fName.size()] != ' ') { + type_not_found = true; + + if (ln[ln.find(key.fName) + key.fName.size()] == '\t') type_not_found = false; + + goto next; + } else if (ln[ln.find(key.fName) + key.fName.size()] != '\t') { + type_not_found = true; + + if (ln[ln.find(key.fName) + key.fName.size()] == ' ') type_not_found = false; + } + } + + next: + + if (ln.find(';') == std::string::npos) { + if (ln.find('(') != std::string::npos) { + if (ln.find('=') == std::string::npos) continue; + } + + err_str += "\nMissing ';', here -> "; + err_str += ln; + } else { + continue; + } + + if (ln.find('=') != std::string::npos) { + if (ln.find('(') != std::string::npos) { + if (ln.find(')') == std::string::npos) { + err_str += "\nMissing ')', after '(' here -> "; + err_str += ln.substr(ln.find('(')); + } + } + } + } + } + + if (kInBraces && ln.find("struct") != std::string::npos && + ln.find("union") != std::string::npos && ln.find("enum") != std::string::npos && + ln.find('=') != std::string::npos) { + if (ln.find(';') == std::string::npos) { + err_str += "\nMissing ';' after struct/union/enum declaration, here -> "; + err_str += ln; + } + } + + if (ln.find(';') != std::string::npos && ln.find("for") == std::string::npos) { + if (ln.find(';') + 1 != ln.size()) { + for (int i = 0; i < ln.substr(ln.find(';') + 1).size(); ++i) { + if ((ln.substr(ln.find(';') + 1)[i] != ' ') || (ln.substr(ln.find(';') + 1)[i] != '\t')) { + if (auto err = this->Check(ln.substr(ln.find(';') + 1).c_str(), file); !err.empty()) { + err_str += "\nUnexpected text after ';' -> "; + err_str += ln.substr(ln.find(';')); + err_str += err; + } + } + } + } + } + + if (ln.find('(') != std::string::npos) { + if (ln.find(';') == std::string::npos && !LibCompiler::find_word(ln, "|") && + !LibCompiler::find_word(ln, "||") && !LibCompiler::find_word(ln, "&") && + !LibCompiler::find_word(ln, "&&") && !LibCompiler::find_word(ln, "~")) { + bool found_func = false; + size_t i = ln.find('('); + std::vector opens; + std::vector closes; + + for (; i < ln.size(); ++i) { + if (ln[i] == ')') { + closes.push_back(1); + } + + if (ln[i] == '(') { + opens.push_back(1); + } + } + + if (closes.size() != opens.size()) err_str += "Unterminated (), here -> " + ln; + + bool space_found = false; + + for (int i = 0; i < ln.size(); ++i) { + if (ln[i] == ')' && !space_found) { + space_found = true; + continue; + } + + if (space_found) { + if (ln[i] == ' ' && isalnum(ln[i + 1])) { + err_str += "\nBad function format here -> "; + err_str += ln; + } + } + } + } + + if (ln.find('(') < 1) { + err_str += "\nMissing identifier before '(' here -> "; + err_str += ln; + } else { + if (type_not_found && ln.find(';') == std::string::npos && + ln.find("if") == std::string::npos && ln.find("|") == std::string::npos && + ln.find("&") == std::string::npos && ln.find("(") == std::string::npos && + ln.find(")") == std::string::npos) { + err_str += "\n Missing ';' or type, here -> "; + err_str += ln; + } + } + + if (ln.find(')') == std::string::npos) { + err_str += "\nMissing ')', after '(' here -> "; + err_str += ln.substr(ln.find('(')); + } + } else { + if (ln.find("for") != std::string::npos || ln.find("while") != std::string::npos) { + err_str += "\nMissing '(', after \"for\", here -> "; + err_str += ln; + } + } + + if (ln.find('}') != std::string::npos && !kInBraces) { + if (!kInStruct && ln.find(';') == std::string::npos) { + err_str += "\nMismatched '}', here -> "; + err_str += ln; + } + } + + if (!ln.empty()) { + if (ln.find(';') == std::string::npos && ln.find('{') == std::string::npos && + ln.find('}') == std::string::npos && ln.find(')') == std::string::npos && + ln.find('(') == std::string::npos && ln.find(',') == std::string::npos) { + if (ln.size() <= 2) return err_str; + + err_str += "\nMissing ';', here -> "; + err_str += ln; + } + } + + return err_str; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +/** + * @brief C To Assembly mount-point. + */ + +///////////////////////////////////////////////////////////////////////////////////////// + +class AssemblyMountpointCLang final LC_ASSEMBLY_INTERFACE { + public: + explicit AssemblyMountpointCLang() = default; + ~AssemblyMountpointCLang() override = default; + + LIBCOMPILER_COPY_DEFAULT(AssemblyMountpointCLang); + + UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArchPowerPC; } + + Int32 CompileToFormat(std::string src, Int32 arch) override { + if (kCompilerFrontend == nullptr) return 1; + + /* @brief copy contents wihtout extension */ + std::string src_file = src.data(); + std::ifstream src_fp = std::ifstream(src_file, std::ios::in); + std::string dest; + + for (auto& ch : src_file) { + if (ch == '.') { + break; + } + + dest += ch; + } + + /* According to PEF ABI. */ + std::vector exts = kAsmFileExts; + dest += exts[4]; + + kState.fOutputAssembly = std::make_unique(dest); + + auto fmt = LibCompiler::current_date(); + + (*kState.fOutputAssembly) << "# Path: " << src_file << "\n"; + (*kState.fOutputAssembly) << "# Language: POWER Assembly (Generated from C)\n"; + (*kState.fOutputAssembly) << "# Date: " << fmt << "\n\n"; + + LibCompiler::SyntaxLeafList syntax; + + kState.fSyntaxTreeList.push_back(syntax); + kState.fSyntaxTree = &kState.fSyntaxTreeList[kState.fSyntaxTreeList.size() - 1]; + + std::string line_src; + + while (std::getline(src_fp, line_src)) { + if (auto err = kCompilerFrontend->Check(line_src.c_str(), src.data()); err.empty()) { + kCompilerFrontend->Compile(line_src, src.data()); + } else { + Detail::print_error(err, src.data()); + } + } + + if (kAcceptableErrors > 0) return 1; + + std::vector keywords = {"ld", "stw", "add", "sub", "or"}; + + /// + /// Replace, optimize, fix assembly output. + /// + + for (auto& leaf : kState.fSyntaxTree->fLeafList) { + std::vector access_keywords = {"->", "."}; + + for (auto& access_ident : access_keywords) { + if (LibCompiler::find_word(leaf.fUserValue, access_ident)) { + for (auto& struc : kState.kStructMap) { + /// TODO: + } + } + } + + for (auto& keyword : keywords) { + if (LibCompiler::find_word(leaf.fUserValue, keyword)) { + std::size_t cnt = 0UL; + + for (auto& reg : kState.kStackFrame) { + std::string needle; + + for (size_t i = 0; i < reg.fName.size(); i++) { + if (reg.fName[i] == ' ') { + ++i; + + for (; i < reg.fName.size(); i++) { + if (reg.fName[i] == ',') { + break; + } + + if (reg.fName[i] == ' ') continue; + + needle += reg.fName[i]; + } + + break; + } + } + + if (LibCompiler::find_word(leaf.fUserValue, needle)) { + if (leaf.fUserValue.find("extern_segment ") != std::string::npos) { + std::string range = "extern_segment "; + leaf.fUserValue.replace(leaf.fUserValue.find(range), range.size(), ""); + } + + if (leaf.fUserValue.find("ldw r6") != std::string::npos) { + std::string::difference_type countComma = + std::count(leaf.fUserValue.begin(), leaf.fUserValue.end(), ','); + + if (countComma == 1) { + leaf.fUserValue.replace(leaf.fUserValue.find("ldw"), strlen("ldw"), "mr"); + } + } + + leaf.fUserValue.replace(leaf.fUserValue.find(needle), needle.size(), reg.fReg); + + ++cnt; + } + } + + if (cnt > 1 && keyword != "mr" && keyword != "add" && keyword != "dec") { + leaf.fUserValue.replace(leaf.fUserValue.find(keyword), keyword.size(), "mr"); + } + } + } + } + + for (auto& leaf : kState.fSyntaxTree->fLeafList) { + (*kState.fOutputAssembly) << leaf.fUserValue; + } + + kState.fSyntaxTree = nullptr; + + kState.fOutputAssembly->flush(); + kState.fOutputAssembly.reset(); + + return kExitOK; + } +}; + +///////////////////////////////////////////////////////////////////////////////////////// + +#include + +#define kPrintF printf +#define kSplashCxx() kPrintF(kWhite "cc, %s, (c) Amlal El Mahrouss\n", kDistVersion) + +static void cc_print_help() { + kSplashCxx(); +} + +///////////////////////////////////////////////////////////////////////////////////////// + +#define kExt ".c" + +LIBCOMPILER_MODULE(NeOSCompilerCLangPowerPC) { + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + kCompilerTypes.push_back({.fName = "void", .fValue = "void"}); + kCompilerTypes.push_back({.fName = "char", .fValue = "byte"}); + kCompilerTypes.push_back({.fName = "short", .fValue = "hword"}); + kCompilerTypes.push_back({.fName = "int", .fValue = "dword"}); + kCompilerTypes.push_back({.fName = "long", .fValue = "qword"}); + kCompilerTypes.push_back({.fName = "*", .fValue = "offset"}); + + bool skip = false; + + kFactory.Mount(new AssemblyMountpointCLang()); + kMachine = LibCompiler::AssemblyFactory::kArchPowerPC; + kCompilerFrontend = new CompilerFrontendPower64(); + + for (auto index = 1UL; index < argc; ++index) { + if (skip) { + skip = false; + continue; + } + + if (argv[index][0] == '-') { + if (strcmp(argv[index], "-v") == 0 || strcmp(argv[index], "-version") == 0) { + kSplashCxx(); + return kExitOK; + } + + if (strcmp(argv[index], "-verbose") == 0) { + kState.fVerbose = true; + + continue; + } + + if (strcmp(argv[index], "-h") == 0 || strcmp(argv[index], "-help") == 0) { + cc_print_help(); + + return kExitOK; + } + + if (strcmp(argv[index], "-dialect") == 0) { + if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; + + return kExitOK; + } + + if (strcmp(argv[index], "-fmax-exceptions") == 0) { + try { + kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); + } + // catch anything here + catch (...) { + kErrorLimit = 0; + } + + skip = true; + + continue; + } + + std::string err = "Unknown command: "; + err += argv[index]; + + Detail::print_error(err, "cc"); + + continue; + } + + kFileList.emplace_back(argv[index]); + + std::string srcFile = argv[index]; + + if (strstr(argv[index], kExt) == nullptr) { + if (kState.fVerbose) { + Detail::print_error(srcFile + " is not a valid C source.\n", "cc"); + } + + return 1; + } + + if (kFactory.Compile(srcFile, kMachine) != kExitOK) return 1; + } + + return kExitOK; +} + +// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/Cl/CPlusPlusCompilerAMD64.cc b/dev/LibCompiler/src/Cl/CPlusPlusCompilerAMD64.cc new file mode 100644 index 0000000..c99508c --- /dev/null +++ b/dev/LibCompiler/src/Cl/CPlusPlusCompilerAMD64.cc @@ -0,0 +1,921 @@ +/* + * ======================================================== + * + * cxxdrv + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +/// BUGS: 1 + +#define kPrintF printf + +#define kExitOK (EXIT_SUCCESS) +#define kExitNO (EXIT_FAILURE) + +// extern_segment, @autodelete { ... }, fn foo() -> auto { ... } + +#include +#include +#include +#include + +/* NeKernel C++ Compiler Driver */ +/* This is part of the LibCompiler. */ +/* (c) Amlal El Mahrouss */ + +/// @author EL Mahrouss Amlal (amlal@nekernel.org) +/// @file CPlusPlusCompilerAMD64.cxx +/// @brief Optimized C++ Compiler Driver. +/// @todo Throw error for scoped inside scoped variables when they get referenced outside. +/// @todo Add class/struct/enum support. + +/////////////////////// + +// ANSI ESCAPE CODES // + +/////////////////////// + +#define kBlank "\e[0;30m" +#define kRed "\e[0;31m" +#define kWhite "\e[0;97m" + +///////////////////////////////////// + +// INTERNALS OF THE C++ COMPILER + +///////////////////////////////////// + +/// @internal +namespace Detail { +std::filesystem::path expand_home(const std::filesystem::path& p) { + if (!p.empty() && p.string()[0] == '~') { + const char* home = std::getenv("HOME"); // For Unix-like systems + + if (!home) { + home = std::getenv("USERPROFILE"); // For Windows + } + + if (home) { + return std::filesystem::path(home) / p.relative_path().string().substr(1); + } else { + throw std::runtime_error("Home directory not found in environment variables"); + } + } + return p; +} + +struct CompilerRegisterMap final { + std::string fName; + std::string fReg; +}; + +// \brief Offset based struct/class +struct CompilerStructMap final { + std::string fName; + std::string fReg; + + // offset counter + std::size_t fOffsetsCnt; + + // offset array + std::vector> fOffsets; +}; + +struct CompilerState final { + std::vector fStackMapVector; + std::vector fStructMapVector; + std::string fOutputValue; + std::string fLastFile; + std::string fLastError; +}; +} // namespace Detail + +static Detail::CompilerState kState; + +static Int32 kOnClassScope = 0; + +///////////////////////////////////////////////////////////////////////////////////////// + +// Target architecture. +static int kMachine = LibCompiler::AssemblyFactory::kArchAMD64; + +///////////////////////////////////////// + +// ARGUMENTS REGISTERS (R8, R15) + +///////////////////////////////////////// + +static std::vector kKeywords; + +///////////////////////////////////////// + +// COMPILER PARSING UTILITIES/STATES. + +///////////////////////////////////////// + +static LibCompiler::AssemblyFactory kFactory; +static Boolean kInStruct = false; +static Boolean kOnWhileLoop = false; +static Boolean kOnForLoop = false; +static Boolean kInBraces = false; +static size_t kBracesCount = 0UL; + +/* @brief C++ compiler backend for the NeKernel C++ driver */ +class CompilerFrontendCPlusPlus final : public LibCompiler::CompilerFrontendInterface { + public: + explicit CompilerFrontendCPlusPlus() = default; + ~CompilerFrontendCPlusPlus() override = default; + + LIBCOMPILER_COPY_DEFAULT(CompilerFrontendCPlusPlus); + + Boolean Compile(const std::string text, std::string file) override; + + const char* Language() override; +}; + +/// @internal compiler variables + +static CompilerFrontendCPlusPlus* kCompilerFrontend = nullptr; + +static std::vector kRegisterMap; + +static std::vector kRegisterList = { + "rbx", "rsi", "r10", "r11", "r12", "r13", "r14", "r15", "xmm12", "xmm13", "xmm14", "xmm15", +}; + +/// @brief The PEF calling convention (caller must save rax, rbp) +/// @note callee must return via **rax**. +static std::vector kRegisterConventionCallList = { + "r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15", +}; + +static std::size_t kFunctionEmbedLevel = 0UL; + +/// detail namespaces + +const char* CompilerFrontendCPlusPlus::Language() { + return "AMD64 C++"; +} + +static std::uintptr_t kOrigin = 0x1000000; +static std::vector> kOriginMap; + +///////////////////////////////////////////////////////////////////////////////////////// + +/// @name Compile +/// @brief Generate assembly from a C++ source. + +///////////////////////////////////////////////////////////////////////////////////////// + +Boolean CompilerFrontendCPlusPlus::Compile(std::string text, std::string file) { + if (text.length() < 1) return false; + + std::size_t index = 0UL; + std::vector> keywords_list; + + for (auto& keyword : kKeywords) { + if (text.find(keyword.keyword_name) != std::string::npos) { + switch (keyword.keyword_kind) { + case LibCompiler::kKeywordKindCommentInline: { + break; + } + default: + break; + } + + std::size_t pos = text.find(keyword.keyword_name); + if (pos == std::string::npos) continue; + + // Safe guard: can't go before start of string + if (pos > 0 && text[pos - 1] == '+' && + keyword.keyword_kind == LibCompiler::kKeywordKindVariableAssign) + continue; + + if (pos > 0 && text[pos - 1] == '-' && + keyword.keyword_kind == LibCompiler::kKeywordKindVariableAssign) + continue; + + // Safe guard: don't go out of range + if ((pos + keyword.keyword_name.size()) < text.size() && + text[pos + keyword.keyword_name.size()] == '=' && + keyword.keyword_kind == LibCompiler::kKeywordKindVariableAssign) + continue; + + keywords_list.emplace_back(std::make_pair(keyword, index)); + ++index; + } + } + + LibCompiler::SyntaxLeafList::SyntaxLeaf syntax_tree; + + for (auto& keyword : keywords_list) { + if (text.find(keyword.first.keyword_name) == std::string::npos) continue; + + switch (keyword.first.keyword_kind) { + case LibCompiler::KeywordKind::kKeywordKindClass: { + ++kOnClassScope; + break; + } + case LibCompiler::KeywordKind::kKeywordKindIf: { + std::size_t keywordPos = text.find(keyword.first.keyword_name); + std::size_t openParen = text.find("(", keywordPos); + std::size_t closeParen = text.find(")", openParen); + + if (keywordPos == std::string::npos || openParen == std::string::npos || + closeParen == std::string::npos || closeParen <= openParen) { + Detail::print_error("Malformed if expression: " + text, file); + return false; + } + + auto expr = text.substr(openParen + 1, closeParen - openParen - 1); + + if (expr.find(">=") != std::string::npos) { + auto left = text.substr( + text.find(keyword.first.keyword_name) + keyword.first.keyword_name.size() + 2, + expr.find("<=") + strlen("<=")); + auto right = text.substr(expr.find(">=") + strlen(">="), text.find(")") - 1); + + size_t i = right.size() - 1; + + if (i < 1) break; + + try { + while (!std::isalnum(right[i])) { + right.erase(i, 1); + --i; + } + + right.erase(0, i); + } catch (...) { + right.erase(0, i); + } + + i = left.size() - 1; + try { + while (!std::isalnum(left[i])) { + left.erase(i, 1); + --i; + } + + left.erase(0, i); + } catch (...) { + left.erase(0, i); + } + + if (!isdigit(left[0]) || !isdigit(right[0])) { + auto indexRight = 0UL; + + auto& valueOfVar = !isdigit(left[0]) ? left : right; + + if (!valueOfVar.empty()) { + for (auto pairRight : kRegisterMap) { + ++indexRight; + + if (pairRight != valueOfVar) { + auto& valueOfVarOpposite = isdigit(left[0]) ? left : right; + + syntax_tree.fUserValue += + "mov " + kRegisterList[indexRight + 1] + ", " + valueOfVarOpposite + "\n"; + syntax_tree.fUserValue += "cmp " + kRegisterList[kRegisterMap.size() - 1] + "," + + kRegisterList[indexRight + 1] + "\n"; + + goto done_iterarting_on_if; + } + + auto& valueOfVarOpposite = isdigit(left[0]) ? left : right; + + syntax_tree.fUserValue += + "mov " + kRegisterList[indexRight + 1] + ", " + valueOfVarOpposite + "\n"; + syntax_tree.fUserValue += "cmp " + kRegisterList[kRegisterMap.size() - 1] + ", " + + kRegisterList[indexRight + 1] + "\n"; + + break; + } + } + } + + done_iterarting_on_if: + + std::string fnName = text; + fnName.erase(fnName.find(keyword.first.keyword_name)); + + for (auto& ch : fnName) { + if (ch == ' ') ch = '_'; + } + + syntax_tree.fUserValue += + "jge __OFFSET_ON_TRUE_LC\nsegment .code64 __OFFSET_ON_TRUE_LC:\n"; + } + + break; + } + case LibCompiler::KeywordKind::kKeywordKindFunctionStart: { + for (auto& ch : text) { + if (isdigit(ch)) { + goto dont_accept; + } + } + + goto accept; + + dont_accept: + return false; + + accept: + std::string fnName = text; + size_t indexFnName = 0; + + // this one is for the type. + for (auto& ch : text) { + ++indexFnName; + + if (ch == '\t') break; + if (ch == ' ') break; + } + + fnName = text.substr(indexFnName); + + if (text.find("return ") != std::string::npos) { + text.erase(0, text.find("return ")); + break; + } + + if (text.ends_with(";") && text.find("return") == std::string::npos) + goto lc_write_assembly; + else if (text.size() <= indexFnName) + Detail::print_error("Invalid function name: " + fnName, file); + + indexFnName = 0; + + for (auto& ch : fnName) { + if (ch == ' ' || ch == '\t') { + if (fnName[indexFnName - 1] != ')') + Detail::print_error("Invalid function name: " + fnName, file); + } + + ++indexFnName; + } + + if (fnName.find("(") != LibCompiler::STLString::npos) { + fnName.erase(fnName.find("(")); + } + + syntax_tree.fUserValue = "public_segment .code64 __LIBCOMPILER_" + fnName + "\n"; + ++kFunctionEmbedLevel; + + kOriginMap.push_back({"__LIBCOMPILER_" + fnName, kOrigin}); + + break; + + lc_write_assembly: + auto it = std::find_if(kOriginMap.begin(), kOriginMap.end(), + [&fnName](std::pair pair) -> bool { + return fnName == pair.first; + }); + + if (it != kOriginMap.end()) { + std::stringstream ss; + ss << std::hex << it->second; + + syntax_tree.fUserValue = "jmp " + ss.str() + "\n"; + kOrigin += 1UL; + } + } + case LibCompiler::KeywordKind::kKeywordKindFunctionEnd: { + if (kOnClassScope) --kOnClassScope; + + if (text.ends_with(";")) break; + + --kFunctionEmbedLevel; + + if (kRegisterMap.size() > kRegisterList.size()) { + --kFunctionEmbedLevel; + } + + if (kFunctionEmbedLevel < 1) kRegisterMap.clear(); + + break; + } + case LibCompiler::KeywordKind::kKeywordKindEndInstr: + case LibCompiler::KeywordKind::kKeywordKindVariableInc: + case LibCompiler::KeywordKind::kKeywordKindVariableDec: + case LibCompiler::KeywordKind::kKeywordKindVariableAssign: { + std::string valueOfVar = ""; + + if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindVariableInc) { + valueOfVar = text.substr(text.find("+=") + 2); + } else if (keyword.first.keyword_kind == + LibCompiler::KeywordKind::kKeywordKindVariableDec) { + valueOfVar = text.substr(text.find("-=") + 2); + } else if (keyword.first.keyword_kind == + LibCompiler::KeywordKind::kKeywordKindVariableAssign) { + valueOfVar = text.substr(text.find("=") + 1); + } else if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindEndInstr) { + break; + } + + while (valueOfVar.find(";") != std::string::npos && + keyword.first.keyword_kind != LibCompiler::KeywordKind::kKeywordKindEndInstr) { + valueOfVar.erase(valueOfVar.find(";")); + } + + std::string varName = text; + + if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindVariableInc) { + varName.erase(varName.find("+=")); + } else if (keyword.first.keyword_kind == + LibCompiler::KeywordKind::kKeywordKindVariableDec) { + varName.erase(varName.find("-=")); + } else if (keyword.first.keyword_kind == + LibCompiler::KeywordKind::kKeywordKindVariableAssign) { + varName.erase(varName.find("=")); + } else if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindEndInstr) { + varName.erase(varName.find(";")); + } + + static Boolean typeFound = false; + + for (auto& keyword : kKeywords) { + if (keyword.keyword_kind == LibCompiler::kKeywordKindType) { + if (text.find(keyword.keyword_name) != std::string::npos) { + if (text[text.find(keyword.keyword_name)] == ' ') { + typeFound = false; + continue; + } + + typeFound = true; + } + } + } + + std::string instr = "mov "; + + if (typeFound && + keyword.first.keyword_kind != LibCompiler::KeywordKind::kKeywordKindVariableInc && + keyword.first.keyword_kind != LibCompiler::KeywordKind::kKeywordKindVariableDec) { + if (kRegisterMap.size() > kRegisterList.size()) { + ++kFunctionEmbedLevel; + } + + while (varName.find(" ") != std::string::npos) { + varName.erase(varName.find(" "), 1); + } + + while (varName.find("\t") != std::string::npos) { + varName.erase(varName.find("\t"), 1); + } + + for (size_t i = 0; !isalnum(valueOfVar[i]); i++) { + if (i > valueOfVar.size()) break; + + valueOfVar.erase(i, 1); + } + + constexpr auto kTrueVal = "true"; + constexpr auto kFalseVal = "false"; + + if (valueOfVar == kTrueVal) { + valueOfVar = "1"; + } else if (valueOfVar == kFalseVal) { + valueOfVar = "0"; + } + + std::size_t indexRight = 0UL; + + for (auto pairRight : kRegisterMap) { + ++indexRight; + + if (pairRight != valueOfVar) { + if (valueOfVar[0] == '\"') { + syntax_tree.fUserValue = "segment .data64 __LIBCOMPILER_LOCAL_VAR_" + varName + + ": db " + valueOfVar + ", 0\n\n"; + syntax_tree.fUserValue += instr + kRegisterList[kRegisterMap.size() - 1] + ", " + + "__LIBCOMPILER_LOCAL_VAR_" + varName + "\n"; + kOrigin += 1UL; + } else { + syntax_tree.fUserValue = + instr + kRegisterList[kRegisterMap.size() - 1] + ", " + valueOfVar + "\n"; + kOrigin += 1UL; + } + + goto done; + } + } + + if (((int) indexRight - 1) < 0) { + if (valueOfVar[0] == '\"') { + syntax_tree.fUserValue = "segment .data64 __LIBCOMPILER_LOCAL_VAR_" + varName + + ": db " + valueOfVar + ", 0\n"; + syntax_tree.fUserValue += instr + kRegisterList[kRegisterMap.size()] + ", " + + "__LIBCOMPILER_LOCAL_VAR_" + varName + "\n"; + kOrigin += 1UL; + } else { + syntax_tree.fUserValue = + instr + kRegisterList[kRegisterMap.size()] + ", " + valueOfVar + "\n"; + kOrigin += 1UL; + } + + goto done; + } + + if (valueOfVar[0] != '\"' && valueOfVar[0] != '\'' && !isdigit(valueOfVar[0])) { + for (auto pair : kRegisterMap) { + if (pair == valueOfVar) goto done; + } + + Detail::print_error("Variable not declared: " + varName, file); + return false; + } + + done: + for (auto& keyword : kKeywords) { + if (keyword.keyword_kind == LibCompiler::kKeywordKindType && + varName.find(keyword.keyword_name) != std::string::npos) { + varName.erase(varName.find(keyword.keyword_name), keyword.keyword_name.size()); + break; + } + } + + kRegisterMap.push_back(varName); + + break; + } + + if (kKeywords[keyword.second - 1].keyword_kind == LibCompiler::kKeywordKindType || + kKeywords[keyword.second - 1].keyword_kind == LibCompiler::kKeywordKindTypePtr) { + syntax_tree.fUserValue = "\n"; + continue; + } + + if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindEndInstr) { + syntax_tree.fUserValue = "\n"; + continue; + } + + if (keyword.first.keyword_kind == LibCompiler::KeywordKind::kKeywordKindVariableInc) { + instr = "add "; + } else if (keyword.first.keyword_kind == + LibCompiler::KeywordKind::kKeywordKindVariableDec) { + instr = "sub "; + } + + std::string varErrCpy = varName; + + while (varName.find(" ") != std::string::npos) { + varName.erase(varName.find(" "), 1); + } + + while (varName.find("\t") != std::string::npos) { + varName.erase(varName.find("\t"), 1); + } + + std::size_t indxReg = 0UL; + + for (size_t i = 0; !isalnum(valueOfVar[i]); i++) { + if (i > valueOfVar.size()) break; + + valueOfVar.erase(i, 1); + } + + while (valueOfVar.find(" ") != std::string::npos) { + valueOfVar.erase(valueOfVar.find(" "), 1); + } + + while (valueOfVar.find("\t") != std::string::npos) { + valueOfVar.erase(valueOfVar.find("\t"), 1); + } + + constexpr auto kTrueVal = "true"; + constexpr auto kFalseVal = "false"; + + /// interpet boolean values, since we're on C++ + + if (valueOfVar == kTrueVal) { + valueOfVar = "1"; + } else if (valueOfVar == kFalseVal) { + valueOfVar = "0"; + } + + for (auto pair : kRegisterMap) { + ++indxReg; + + if (pair != varName) continue; + + std::size_t indexRight = 0ul; + + for (auto pairRight : kRegisterMap) { + ++indexRight; + + if (pairRight != varName) { + syntax_tree.fUserValue = + instr + kRegisterList[kRegisterMap.size()] + ", " + valueOfVar + "\n"; + kOrigin += 1UL; + continue; + } + + syntax_tree.fUserValue = + instr + kRegisterList[indexRight - 1] + ", " + valueOfVar + "\n"; + kOrigin += 1UL; + break; + } + + break; + } + + if (syntax_tree.fUserValue.empty()) { + Detail::print_error("Variable not declared: " + varName, file); + } + + break; + } + case LibCompiler::KeywordKind::kKeywordKindReturn: { + try { + auto pos = text.find("return") + strlen("return") + 1; + std::string subText = text.substr(pos); + subText = subText.erase(subText.find(";")); + size_t indxReg = 0UL; + + if (subText[0] != '\"' && subText[0] != '\'') { + if (!isdigit(subText[0])) { + for (auto pair : kRegisterMap) { + ++indxReg; + + if (pair != subText) continue; + + syntax_tree.fUserValue = "mov rax, " + kRegisterList[indxReg - 1] + "\nret\n"; + kOrigin += 1UL; + + break; + } + } else { + syntax_tree.fUserValue = "mov rax, " + subText + "\nret\n"; + kOrigin += 1UL; + + break; + } + } else { + syntax_tree.fUserValue = "__LIBCOMPILER_LOCAL_RETURN_STRING: db " + subText + + ", 0\nmov rcx, __LIBCOMPILER_LOCAL_RETURN_STRING\n"; + syntax_tree.fUserValue += "mov rax, rcx\nret\n"; + kOrigin += 1UL; + + break; + } + + if (syntax_tree.fUserValue.empty()) { + if (subText.find("(") != std::string::npos) { + subText.erase(subText.find("(")); + + auto it = + std::find_if(kOriginMap.begin(), kOriginMap.end(), + [&subText](std::pair pair) -> bool { + return pair.first.find(subText) != std::string::npos; + }); + + if (it == kOriginMap.end()) + Detail::print_error("Invalid return value: " + subText, file); + + std::stringstream ss; + ss << it->second; + + syntax_tree.fUserValue = "jmp " + ss.str() + "\nret\n"; + kOrigin += 1UL; + break; + } + } + + break; + } catch (...) { + syntax_tree.fUserValue = "ret\n"; + kOrigin += 1UL; + } + } + default: { + continue; + } + } + } + + kState.fOutputValue = syntax_tree.fUserValue; + return true; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +/** + * @brief C++ assembler class. + */ + +///////////////////////////////////////////////////////////////////////////////////////// + +class AssemblyCPlusPlusInterface final LC_ASSEMBLY_INTERFACE { + public: + explicit AssemblyCPlusPlusInterface() = default; + ~AssemblyCPlusPlusInterface() override = default; + + LIBCOMPILER_COPY_DEFAULT(AssemblyCPlusPlusInterface); + + UInt32 Arch() noexcept override { return LibCompiler::AssemblyFactory::kArchAMD64; } + + Int32 CompileToFormat(std::string src, Int32 arch) override { + if (kCompilerFrontend == nullptr) return kExitNO; + + std::string dest = src; + dest += ".pp.masm"; + + std::ofstream out_fp(dest); + + std::ifstream src_fp = std::ifstream(src); + + std::string line_source; + + while (std::getline(src_fp, line_source)) { + if (kVerbose) { + kStdOut << line_source << std::endl; + kStdOut << line_source.length() << " bytes\n"; + } + + kCompilerFrontend->Compile(line_source, src); + + out_fp << kState.fOutputValue; + kState.fOutputValue.clear(); + } + + if (kVerbose) { + kStdOut << "Done compiling " << src << " to " << dest << "\n"; + } + + return kExitOK; + } +}; + +///////////////////////////////////////////////////////////////////////////////////////// + +///////////////////////////////////////////////////////////////////////////////////////// + +#define kExtListCxx \ + { ".cpp", ".cxx", ".cc", ".c++", ".cp" } + +LIBCOMPILER_MODULE(CompilerCPlusPlusAMD64) { + Boolean skip = false; + + kKeywords.push_back({.keyword_name = "if", .keyword_kind = LibCompiler::kKeywordKindIf}); + kKeywords.push_back({.keyword_name = "else", .keyword_kind = LibCompiler::kKeywordKindElse}); + kKeywords.push_back({.keyword_name = "else if", .keyword_kind = LibCompiler::kKeywordKindElseIf}); + + kKeywords.push_back({.keyword_name = "class", .keyword_kind = LibCompiler::kKeywordKindClass}); + kKeywords.push_back({.keyword_name = "struct", .keyword_kind = LibCompiler::kKeywordKindClass}); + kKeywords.push_back( + {.keyword_name = "namespace", .keyword_kind = LibCompiler::kKeywordKindNamespace}); + kKeywords.push_back( + {.keyword_name = "typedef", .keyword_kind = LibCompiler::kKeywordKindTypedef}); + kKeywords.push_back({.keyword_name = "using", .keyword_kind = LibCompiler::kKeywordKindTypedef}); + kKeywords.push_back({.keyword_name = "{", .keyword_kind = LibCompiler::kKeywordKindBodyStart}); + kKeywords.push_back({.keyword_name = "}", .keyword_kind = LibCompiler::kKeywordKindBodyEnd}); + kKeywords.push_back({.keyword_name = "auto", .keyword_kind = LibCompiler::kKeywordKindVariable}); + kKeywords.push_back({.keyword_name = "int", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "bool", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "unsigned", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "short", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "char", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "long", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "float", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "double", .keyword_kind = LibCompiler::kKeywordKindType}); + kKeywords.push_back({.keyword_name = "void", .keyword_kind = LibCompiler::kKeywordKindType}); + + kKeywords.push_back( + {.keyword_name = "auto*", .keyword_kind = LibCompiler::kKeywordKindVariablePtr}); + kKeywords.push_back({.keyword_name = "int*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back({.keyword_name = "bool*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back( + {.keyword_name = "unsigned*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back({.keyword_name = "short*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back({.keyword_name = "char*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back({.keyword_name = "long*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back({.keyword_name = "float*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back( + {.keyword_name = "double*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + kKeywords.push_back({.keyword_name = "void*", .keyword_kind = LibCompiler::kKeywordKindTypePtr}); + + kKeywords.push_back( + {.keyword_name = "(", .keyword_kind = LibCompiler::kKeywordKindFunctionStart}); + kKeywords.push_back({.keyword_name = ")", .keyword_kind = LibCompiler::kKeywordKindFunctionEnd}); + kKeywords.push_back( + {.keyword_name = "=", .keyword_kind = LibCompiler::kKeywordKindVariableAssign}); + kKeywords.push_back({.keyword_name = "+=", .keyword_kind = LibCompiler::kKeywordKindVariableInc}); + kKeywords.push_back({.keyword_name = "-=", .keyword_kind = LibCompiler::kKeywordKindVariableDec}); + kKeywords.push_back({.keyword_name = "const", .keyword_kind = LibCompiler::kKeywordKindConstant}); + kKeywords.push_back({.keyword_name = "*", .keyword_kind = LibCompiler::kKeywordKindPtr}); + kKeywords.push_back({.keyword_name = "->", .keyword_kind = LibCompiler::kKeywordKindPtrAccess}); + kKeywords.push_back({.keyword_name = ".", .keyword_kind = LibCompiler::kKeywordKindAccess}); + kKeywords.push_back({.keyword_name = ",", .keyword_kind = LibCompiler::kKeywordKindArgSeparator}); + kKeywords.push_back({.keyword_name = ";", .keyword_kind = LibCompiler::kKeywordKindEndInstr}); + kKeywords.push_back({.keyword_name = ":", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); + kKeywords.push_back( + {.keyword_name = "public:", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); + kKeywords.push_back( + {.keyword_name = "private:", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); + kKeywords.push_back( + {.keyword_name = "protected:", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); + kKeywords.push_back( + {.keyword_name = "final", .keyword_kind = LibCompiler::kKeywordKindSpecifier}); + kKeywords.push_back({.keyword_name = "return", .keyword_kind = LibCompiler::kKeywordKindReturn}); + kKeywords.push_back( + {.keyword_name = "/*", .keyword_kind = LibCompiler::kKeywordKindCommentMultiLineStart}); + kKeywords.push_back( + {.keyword_name = "*/", .keyword_kind = LibCompiler::kKeywordKindCommentMultiLineEnd}); + kKeywords.push_back( + {.keyword_name = "//", .keyword_kind = LibCompiler::kKeywordKindCommentInline}); + kKeywords.push_back({.keyword_name = "==", .keyword_kind = LibCompiler::kKeywordKindEq}); + kKeywords.push_back({.keyword_name = "!=", .keyword_kind = LibCompiler::kKeywordKindNotEq}); + kKeywords.push_back({.keyword_name = ">=", .keyword_kind = LibCompiler::kKeywordKindGreaterEq}); + kKeywords.push_back({.keyword_name = "<=", .keyword_kind = LibCompiler::kKeywordKindLessEq}); + + kErrorLimit = 100; + + kCompilerFrontend = new CompilerFrontendCPlusPlus(); + kFactory.Mount(new AssemblyCPlusPlusInterface()); + + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + for (auto index = 1UL; index < argc; ++index) { + if (argv[index][0] == '-') { + if (skip) { + skip = false; + continue; + } + + if (strcmp(argv[index], "-cxx-verbose") == 0) { + kVerbose = true; + + continue; + } + + if (strcmp(argv[index], "-cxx-dialect") == 0) { + if (kCompilerFrontend) std::cout << kCompilerFrontend->Language() << "\n"; + + return LIBCOMPILER_SUCCESS; + } + + if (strcmp(argv[index], "-cxx-max-err") == 0) { + try { + kErrorLimit = std::strtol(argv[index + 1], nullptr, 10); + } + // catch anything here + catch (...) { + kErrorLimit = 0; + } + + skip = true; + + continue; + } + + std::string err = "Unknown option: "; + err += argv[index]; + + Detail::print_error(err, "cxxdrv"); + + continue; + } + + std::string argv_i = argv[index]; + + std::vector exts = kExtListCxx; + BOOL found = false; + + for (std::string ext : exts) { + if (argv_i.ends_with(ext)) { + found = true; + + if (kFactory.Compile(argv_i, kMachine) != kExitOK) { + return LIBCOMPILER_INVALID_DATA; + } + } + } + + if (!found) { + if (kVerbose) { + Detail::print_error(argv_i + " is not a valid C++ source.", "cxxdrv"); + } + + return LIBCOMPILER_INVALID_DATA; + } + } + + kFactory.Unmount(); + + delete kCompilerFrontend; + kCompilerFrontend = nullptr; + + kRegisterMap.clear(); + kOriginMap.clear(); + + return LIBCOMPILER_SUCCESS; +} + +// +// Last rev 8-1-24 +// diff --git a/dev/LibCompiler/src/CodeGen.cc b/dev/LibCompiler/src/CodeGen.cc new file mode 100644 index 0000000..31a809b --- /dev/null +++ b/dev/LibCompiler/src/CodeGen.cc @@ -0,0 +1,57 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#include +#include + +/** + * @file AssemblyFactory.cxx + * @author amlal (amlal@nekernel.org) + * @brief Assembler Kit + * @version 0.1 + * @date 2024-01-27 + * + * @copyright Copyright (c) 2024-2025 Amlal El Mahrouss + * + */ + +//! @file Asm.cpp +//! @brief AssemblyKit source implementation. + +namespace LibCompiler { +///! @brief Compile for specific format (ELF, PEF, ZBIN) +Int32 AssemblyFactory::Compile(STLString sourceFile, const Int32& arch) noexcept { + if (sourceFile.length() < 1 || !fMounted) return LIBCOMPILER_UNIMPLEMENTED; + + if (!fMounted) return LIBCOMPILER_UNIMPLEMENTED; + + if (arch != fMounted->Arch()) return LIBCOMPILER_INVALID_ARCH; + + try { + return this->fMounted->CompileToFormat(sourceFile, arch); + } catch (std::exception& e) { + return LIBCOMPILER_EXEC_ERROR; + } +} + +///! @brief mount assembly backend. +void AssemblyFactory::Mount(AssemblyInterface* mountPtr) noexcept { + if (mountPtr) { + fMounted = mountPtr; + } +} + +///! @brief Unmount assembler. +AssemblyInterface* AssemblyFactory::Unmount() noexcept { + auto mount_prev = fMounted; + + if (mount_prev) { + fMounted = nullptr; + } + + return mount_prev; +} +} // namespace LibCompiler diff --git a/dev/LibCompiler/src/DynamicLinkerPEF.cc b/dev/LibCompiler/src/DynamicLinkerPEF.cc deleted file mode 100644 index 6421171..0000000 --- a/dev/LibCompiler/src/DynamicLinkerPEF.cc +++ /dev/null @@ -1,692 +0,0 @@ -/* ------------------------------------------- - - Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved - - @file DynamicLinker64PEF.cc - @brief: C++ 64-Bit PEF Linker. - -------------------------------------------- */ - -/// @author EL Mahrouss Amlal (amlal@nekernel.org) -/// @brief NeKernel 64-bit PEF Linker. -/// Last Rev: Sat Apr 19 CET 2025 -/// @note Do not look up for anything with .code64/.data64/.zero64! -/// It will be loaded when the program loader will start the image. - -//! Toolchain Kit. -#include - -//! Error ID -#include - -//! Assembler Kit -#include - -//! Preferred Executable Format -#include - -//! UUID header. -#include - -//! Release macros. -#include - -//! Advanced Executable Object Format. -#include - -//! LibCompiler utils. -#include - -//! I/O stream from std c++ -#include - -#define kLinkerVersionStr \ - "NeKernel 64-Bit Linker (Preferred Executable Format) %s, (c) Amlal El Mahrouss " \ - "2024-2025 " \ - "all rights reserved.\n" - -#define MemoryCopy(DST, SRC, SZ) memcpy(DST, SRC, SZ) -#define StringCompare(DST, SRC) strcmp(DST, SRC) - -#define kPefNoCpu (0U) -#define kPefNoSubCpu (0U) - -#define kLinkerDefaultOrigin kPefBaseOrigin -#define kLinkerId (0x5046FF) -#define kLinkerAbiContainer "__PEFContainer:ABI:" - -#define kPrintF printf -#define kLinkerSplash() kConsoleOut << std::printf(kLinkerVersionStr, kDistVersion) - -/// @brief PEF stack size symbol. -#define kLinkerStackSizeSymbol "__PEFSizeOfReserveStack" - -#define kConsoleOut \ - (std::cout << "\e[0;31m" \ - << "ld64: " \ - << "\e[0;97m") - -namespace Detail { -struct DynamicLinkerBlob final { - std::vector mBlob{}; // PEF code/bss/data blob. - UIntPtr mOffset{0UL}; // the offset of the PEF container header... -}; -} // namespace Detail - -enum { - kABITypeNull = 0, - kABITypeStart = 0x1010, /* The start of ABI list. */ - kABITypeNE = 0x5046, /* PF (NeKernel's PEF ABI) */ - kABITypeInvalid = 0xFFFF, -}; - -static LibCompiler::String kOutput = "a" kPefExt; -static Int32 kAbi = kABITypeNE; -static Int32 kSubArch = kPefNoSubCpu; -static Int32 kArch = LibCompiler::kPefArchInvalid; -static Bool kFatBinaryEnable = false; -static Bool kStartFound = false; -static Bool kDuplicateSymbols = false; - -/* ld64 is to be found, mld is to be found at runtime. */ -static const CharType* kLdDefineSymbol = ":UndefinedSymbol:"; -static const CharType* kLdDynamicSym = ":RuntimeSymbol:"; - -/* object code and list. */ -static std::vector kObjectList; -static std::vector kObjectBytes; - -/// @brief NE 64-bit Linker. -/// @note This linker is made for PEF executable, thus NE based OSes. -LIBCOMPILER_MODULE(DynamicLinker64PEF) { - bool is_executable = true; - - ::signal(SIGSEGV, Detail::drv_segfault_handler); - - /** - * @brief parse flags and trigger options. - */ - for (size_t linker_arg = 1; linker_arg < argc; ++linker_arg) { - if (StringCompare(argv[linker_arg], "-help") == 0) { - kLinkerSplash(); - - kConsoleOut << "-version: Show linker version.\n"; - kConsoleOut << "-help: Show linker help.\n"; - kConsoleOut << "-ld-verbose: Enable linker trace.\n"; - kConsoleOut << "-dylib: Output as a Dynamic PEF.\n"; - kConsoleOut << "-fat: Output as a FAT PEF.\n"; - kConsoleOut << "-32k: Output as a 32x0 PEF.\n"; - kConsoleOut << "-64k: Output as a 64x0 PEF.\n"; - kConsoleOut << "-amd64: Output as a AMD64 PEF.\n"; - kConsoleOut << "-rv64: Output as a RISC-V PEF.\n"; - kConsoleOut << "-power64: Output as a POWER PEF.\n"; - kConsoleOut << "-arm64: Output as a ARM64 PEF.\n"; - kConsoleOut << "-output: Select the output file name.\n"; - - return LIBCOMPILER_SUCCESS; - } else if (StringCompare(argv[linker_arg], "-version") == 0) { - kLinkerSplash(); - - return LIBCOMPILER_SUCCESS; - } else if (StringCompare(argv[linker_arg], "-fat-binary") == 0) { - kFatBinaryEnable = true; - - continue; - } else if (StringCompare(argv[linker_arg], "-64k") == 0) { - kArch = LibCompiler::kPefArch64000; - - continue; - } else if (StringCompare(argv[linker_arg], "-amd64") == 0) { - kArch = LibCompiler::kPefArchAMD64; - - continue; - } else if (StringCompare(argv[linker_arg], "-32k") == 0) { - kArch = LibCompiler::kPefArch32000; - - continue; - } else if (StringCompare(argv[linker_arg], "-power64") == 0) { - kArch = LibCompiler::kPefArchPowerPC; - - continue; - } else if (StringCompare(argv[linker_arg], "-riscv64") == 0) { - kArch = LibCompiler::kPefArchRISCV; - - continue; - } else if (StringCompare(argv[linker_arg], "-arm64") == 0) { - kArch = LibCompiler::kPefArchARM64; - - continue; - } else if (StringCompare(argv[linker_arg], "-ld-verbose") == 0) { - kVerbose = true; - - continue; - } else if (StringCompare(argv[linker_arg], "-dylib") == 0) { - if (kOutput.empty()) { - continue; - } - - if (kOutput.find(kPefExt) != LibCompiler::String::npos) - kOutput.erase(kOutput.find(kPefExt), strlen(kPefExt)); - - kOutput += kPefDylibExt; - - is_executable = false; - - continue; - } else if (StringCompare(argv[linker_arg], "-output") == 0) { - if ((linker_arg + 1) > argc) continue; - - kOutput = argv[linker_arg + 1]; - ++linker_arg; - - continue; - } else { - if (argv[linker_arg][0] == '-') { - kConsoleOut << "unknown flag: " << argv[linker_arg] << "\n"; - return EXIT_FAILURE; - } - - kObjectList.emplace_back(argv[linker_arg]); - - continue; - } - } - - if (kOutput.empty()) { - kConsoleOut << "no output filename set." << std::endl; - return LIBCOMPILER_EXEC_ERROR; - } else if (kObjectList.empty()) { - kConsoleOut << "no input files." << std::endl; - return LIBCOMPILER_EXEC_ERROR; - } else { - namespace FS = std::filesystem; - - // check for existing files, if they don't throw an error. - for (auto& obj : kObjectList) { - if (!FS::exists(obj)) { - // if filesystem doesn't find file - // -> throw error. - kConsoleOut << "no such file: " << obj << std::endl; - return LIBCOMPILER_EXEC_ERROR; - } - } - } - - // PEF expects a valid target architecture when outputing a binary. - if (kArch == LibCompiler::kPefArchInvalid) { - kConsoleOut << "no target architecture set, can't continue." << std::endl; - return LIBCOMPILER_EXEC_ERROR; - } - - LibCompiler::PEFContainer pef_container{}; - - int32_t archs = kArch; - - pef_container.Count = 0UL; - pef_container.Kind = is_executable ? LibCompiler::kPefKindExec : LibCompiler::kPefKindDylib; - pef_container.SubCpu = kSubArch; - pef_container.Linker = kLinkerId; // Amlal El Mahrouss Linker - pef_container.Abi = kAbi; // Multi-Processor UX ABI - pef_container.Magic[0] = kPefMagic[kFatBinaryEnable ? 2 : 0]; - pef_container.Magic[1] = kPefMagic[1]; - pef_container.Magic[2] = kPefMagic[kFatBinaryEnable ? 0 : 2]; - pef_container.Magic[3] = kPefMagic[3]; - pef_container.Version = kPefVersion; - - // specify the start address, can be 0x10000 - pef_container.Start = kLinkerDefaultOrigin; - pef_container.HdrSz = sizeof(LibCompiler::PEFContainer); - pef_container.Checksum = 0UL; - - std::ofstream output_fc(kOutput, std::ofstream::binary); - - if (output_fc.bad()) { - if (kVerbose) { - kConsoleOut << "error: " << strerror(errno) << "\n"; - } - - return LIBCOMPILER_FILE_NOT_FOUND; - } - - //! Read AE to convert as PEF. - - std::vector command_headers; - LibCompiler::Utils::AEReadableProtocol reader_protocol{}; - - for (const auto& objectFile : kObjectList) { - if (!std::filesystem::exists(objectFile)) continue; - - LibCompiler::AEHeader hdr{}; - - reader_protocol.FP = std::ifstream(objectFile, std::ifstream::binary); - reader_protocol.FP >> hdr; - - if (hdr.fMagic[0] == kAEMag0 && hdr.fMagic[1] == kAEMag1 && - hdr.fSize == sizeof(LibCompiler::AEHeader)) { - if (hdr.fArch != kArch) { - if (kVerbose) kConsoleOut << "is this a FAT binary? : "; - - if (!kFatBinaryEnable) { - if (kVerbose) kConsoleOut << "not a FAT binary.\n"; - - kConsoleOut << "object " << objectFile - << " is a different kind of architecture and output isn't " - "treated as a FAT binary." - << std::endl; - - return LIBCOMPILER_FAT_ERROR; - } else { - if (kVerbose) { - kConsoleOut << "Architecture matches what we expect.\n"; - } - } - } - - // append arch type to archs varaible. - archs |= hdr.fArch; - std::size_t cnt = hdr.fCount; - - if (kVerbose) kConsoleOut << "Object header found, record count: " << cnt << "\n"; - - pef_container.Count = cnt; - - char_type* raw_ae_records = new char_type[cnt * sizeof(LibCompiler::AERecordHeader)]; - - if (!raw_ae_records) { - if (kVerbose) kConsoleOut << "Allocation failure for records of count: " << cnt << "\n"; - } - - memset(raw_ae_records, 0, cnt * sizeof(LibCompiler::AERecordHeader)); - - auto* ae_records = reader_protocol.Read(raw_ae_records, cnt); - - auto org = kLinkerDefaultOrigin; - - for (size_t ae_record_index = 0; ae_record_index < cnt; ++ae_record_index) { - LibCompiler::PEFCommandHeader command_header{0}; - std::size_t offset_of_obj = ae_records[ae_record_index].fOffset; - - MemoryCopy(command_header.Name, ae_records[ae_record_index].fName, kPefNameLen); - - LibCompiler::String cmd_hdr_name(command_header.Name); - - // check this header if it's any valid. - if (cmd_hdr_name.find(kPefCode64) == LibCompiler::String::npos && - cmd_hdr_name.find(kPefData64) == LibCompiler::String::npos && - cmd_hdr_name.find(kPefZero64) == LibCompiler::String::npos) { - if (cmd_hdr_name.find(kPefStart) == LibCompiler::String::npos && - *command_header.Name == 0) { - if (cmd_hdr_name.find(kLdDefineSymbol) != LibCompiler::String::npos) { - goto ld_mark_header; - } else { - continue; - } - } - } - - if (cmd_hdr_name.find(kPefStart) != LibCompiler::String::npos && - cmd_hdr_name.find(kPefCode64) != LibCompiler::String::npos) { - kStartFound = true; - } - - ld_mark_header: - command_header.Offset = offset_of_obj; - command_header.Kind = ae_records[ae_record_index].fKind; - command_header.Size = ae_records[ae_record_index].fSize; - command_header.Cpu = hdr.fArch; - command_header.VMAddress = org; /// TODO: - command_header.SubCpu = hdr.fSubArch; - - org += command_header.Size; - - if (kVerbose) { - kConsoleOut << "Record: " << ae_records[ae_record_index].fName << " is marked.\n"; - - kConsoleOut << "Offset: " << command_header.Offset << "\n"; - } - - command_headers.emplace_back(command_header); - } - - delete[] raw_ae_records; - raw_ae_records = nullptr; - - std::vector bytes; - bytes.resize(hdr.fCodeSize); - - reader_protocol.FP.seekg(std::streamsize(hdr.fStartCode)); - reader_protocol.FP.read(bytes.data(), std::streamsize(hdr.fCodeSize)); - - kObjectBytes.push_back({.mBlob = bytes, .mOffset = hdr.fStartCode}); - - // Blob was written, close fp. - - reader_protocol.FP.close(); - - continue; - } - - kConsoleOut << "Not an AE container: " << objectFile << std::endl; - // don't continue, it is a fatal error. - return LIBCOMPILER_EXEC_ERROR; - } - - pef_container.Cpu = archs; - - output_fc << pef_container; - - if (kVerbose) { - kConsoleOut << "Wrote container to: " << output_fc.tellp() << ".\n"; - } - - output_fc.seekp(std::streamsize(pef_container.HdrSz)); - - std::vector not_found; - std::vector symbols; - - // step 2: check for errors (multiple symbols, undefined ones) - - for (auto& command_hdr : command_headers) { - // check if this symbol needs to be resolved. - if (LibCompiler::String(command_hdr.Name).find(kLdDefineSymbol) != LibCompiler::String::npos && - LibCompiler::String(command_hdr.Name).find(kLdDynamicSym) == LibCompiler::String::npos) { - if (kVerbose) kConsoleOut << "Found undefined symbol: " << command_hdr.Name << "\n"; - - if (auto it = - std::find(not_found.begin(), not_found.end(), LibCompiler::String(command_hdr.Name)); - it == not_found.end()) { - not_found.emplace_back(command_hdr.Name); - } - } - - symbols.emplace_back(command_hdr.Name); - } - - // Now try to solve these symbols. - - for (size_t not_found_idx = 0; not_found_idx < command_headers.size(); ++not_found_idx) { - if (const auto it = std::find(not_found.begin(), not_found.end(), - LibCompiler::String(command_headers[not_found_idx].Name)); - it != not_found.end()) { - LibCompiler::String symbol_imp = *it; - - if (symbol_imp.find(kLdDefineSymbol) == LibCompiler::String::npos) continue; - - // erase the lookup prefix. - symbol_imp.erase(0, symbol_imp.find(kLdDefineSymbol) + strlen(kLdDefineSymbol)); - - // demangle everything. - while (symbol_imp.find('$') != LibCompiler::String::npos) - symbol_imp.erase(symbol_imp.find('$'), 1); - - // the reason we do is because, this may not match the symbol, and we need - // to look for other matching symbols. - for (auto& command_hdr : command_headers) { - if (LibCompiler::String(command_hdr.Name).find(symbol_imp) != LibCompiler::String::npos && - LibCompiler::String(command_hdr.Name).find(kLdDefineSymbol) == - LibCompiler::String::npos) { - LibCompiler::String undefined_symbol = command_hdr.Name; - auto result_of_sym = undefined_symbol.substr(undefined_symbol.find(symbol_imp)); - - for (int i = 0; result_of_sym[i] != 0; ++i) { - if (result_of_sym[i] != symbol_imp[i]) goto ld_continue_search; - } - - not_found.erase(it); - - if (kVerbose) kConsoleOut << "Found symbol: " << command_hdr.Name << "\n"; - - break; - } - } - - ld_continue_search: - continue; - } - } - - // step 3: check for errors (recheck if we have those symbols.) - - if (!kStartFound && is_executable) { - if (kVerbose) - kConsoleOut << "Undefined entrypoint: " << kPefStart - << ", you may have forget to link " - "against the C++ runtime library.\n"; - - kConsoleOut << "Undefined entrypoint " << kPefStart << " for executable: " << kOutput << "\n"; - } - - // step 4: write all PEF commands. - - LibCompiler::PEFCommandHeader date_cmd_hdr{}; - - time_t timestamp = time(nullptr); - - LibCompiler::String timeStampStr = "Container:BuildEpoch:"; - timeStampStr += std::to_string(timestamp); - - strncpy(date_cmd_hdr.Name, timeStampStr.c_str(), timeStampStr.size()); - - date_cmd_hdr.Flags = 0; - date_cmd_hdr.Kind = LibCompiler::kPefZero; - date_cmd_hdr.Offset = output_fc.tellp(); - date_cmd_hdr.Size = timeStampStr.size(); - - command_headers.push_back(date_cmd_hdr); - - LibCompiler::PEFCommandHeader abi_cmd_hdr{}; - - LibCompiler::String abi = kLinkerAbiContainer; - - switch (kArch) { - case LibCompiler::kPefArchAMD64: { - abi += "MSFT"; - break; - } - case LibCompiler::kPefArchPowerPC: { - abi += "SYSV"; - break; - } - case LibCompiler::kPefArch32000: - case LibCompiler::kPefArch64000: { - abi += "_NEP"; - break; - } - default: { - abi += "_IDK"; - break; - } - } - - MemoryCopy(abi_cmd_hdr.Name, abi.c_str(), abi.size()); - - abi_cmd_hdr.Size = abi.size(); - abi_cmd_hdr.Offset = output_fc.tellp(); - abi_cmd_hdr.Flags = 0; - abi_cmd_hdr.Kind = LibCompiler::kPefLinkerID; - - command_headers.push_back(abi_cmd_hdr); - - LibCompiler::PEFCommandHeader stack_cmd_hdr{0}; - - stack_cmd_hdr.Cpu = kArch; - stack_cmd_hdr.Flags = 0; - stack_cmd_hdr.Size = sizeof(uintptr_t); - stack_cmd_hdr.Offset = 0; - - MemoryCopy(stack_cmd_hdr.Name, kLinkerStackSizeSymbol, strlen(kLinkerStackSizeSymbol)); - - command_headers.push_back(stack_cmd_hdr); - - LibCompiler::PEFCommandHeader uuid_cmd_hdr{}; - - std::random_device rd; - - auto seedData = std::array{}; - std::generate(std::begin(seedData), std::end(seedData), std::ref(rd)); - std::seed_seq seq(std::begin(seedData), std::end(seedData)); - std::mt19937 generator(seq); - - auto gen = uuids::uuid_random_generator{generator}; - uuids::uuid id = gen(); - auto uuidStr = uuids::to_string(id); - - MemoryCopy(uuid_cmd_hdr.Name, "Container:GUID:4:", strlen("Container:GUID:4:")); - MemoryCopy(uuid_cmd_hdr.Name + strlen("Container:GUID:4:"), uuidStr.c_str(), uuidStr.size()); - - uuid_cmd_hdr.Size = strlen(uuid_cmd_hdr.Name); - uuid_cmd_hdr.Offset = output_fc.tellp(); - uuid_cmd_hdr.Flags = LibCompiler::kPefLinkerID; - uuid_cmd_hdr.Kind = LibCompiler::kPefZero; - - command_headers.push_back(uuid_cmd_hdr); - - // prepare a symbol vector. - std::vector undef_symbols; - std::vector dupl_symbols; - std::vector resolve_symbols; - - constexpr Int32 kPaddingOffset = 16; - - size_t previous_offset = - (command_headers.size() * sizeof(LibCompiler::PEFCommandHeader)) + kPaddingOffset; - - LibCompiler::PEFCommandHeader end_exec_hdr; - - end_exec_hdr.Offset = output_fc.tellp(); - end_exec_hdr.Flags = LibCompiler::kPefLinkerID; - end_exec_hdr.Kind = LibCompiler::kPefZero; - - MemoryCopy(end_exec_hdr.Name, "Container:Exec:END", strlen("Container:Exec:END")); - - end_exec_hdr.Size = strlen(end_exec_hdr.Name); - - command_headers.push_back(end_exec_hdr); - - // Finally write down the command headers. - // And check for any duplications - for (size_t commandHeaderIndex = 0UL; commandHeaderIndex < command_headers.size(); - ++commandHeaderIndex) { - if (LibCompiler::String(command_headers[commandHeaderIndex].Name).find(kLdDefineSymbol) != - LibCompiler::String::npos && - LibCompiler::String(command_headers[commandHeaderIndex].Name).find(kLdDynamicSym) == - LibCompiler::String::npos) { - // ignore :UndefinedSymbol: headers, they do not contain code. - continue; - } - - LibCompiler::String symbol_name = command_headers[commandHeaderIndex].Name; - - if (!symbol_name.empty()) { - undef_symbols.emplace_back(symbol_name); - } - - command_headers[commandHeaderIndex].Offset += previous_offset; - previous_offset += command_headers[commandHeaderIndex].Size; - - LibCompiler::String name = command_headers[commandHeaderIndex].Name; - - /// so this is valid when we get to the entrypoint. - /// it is always a code64 container. And should equal to kPefStart as well. - /// this chunk of code updates the pef_container.Start with the updated offset. - if (name.find(kPefStart) != LibCompiler::String::npos && - name.find(kPefCode64) != LibCompiler::String::npos) { - pef_container.Start = command_headers[commandHeaderIndex].Offset; - auto tellCurPos = output_fc.tellp(); - - output_fc.seekp(0); - output_fc << pef_container; - - output_fc.seekp(tellCurPos); - } - - if (kVerbose) { - kConsoleOut << "Command name: " << name << "\n"; - kConsoleOut << "VMAddress of command content: " << command_headers[commandHeaderIndex].Offset - << "\n"; - } - - output_fc << command_headers[commandHeaderIndex]; - - for (size_t sub_command_header_index = 0UL; sub_command_header_index < command_headers.size(); - ++sub_command_header_index) { - if (sub_command_header_index == commandHeaderIndex) continue; - - if (LibCompiler::String(command_headers[sub_command_header_index].Name) - .find(kLdDefineSymbol) != LibCompiler::String::npos && - LibCompiler::String(command_headers[sub_command_header_index].Name).find(kLdDynamicSym) == - LibCompiler::String::npos) { - if (kVerbose) { - kConsoleOut << "Ignoring :UndefinedSymbol: headers...\n"; - } - - // ignore :UndefinedSymbol: headers, they do not contain code. - continue; - } - - auto& command_hdr = command_headers[sub_command_header_index]; - - if (command_hdr.Name == LibCompiler::String(command_headers[commandHeaderIndex].Name)) { - if (std::find(dupl_symbols.cbegin(), dupl_symbols.cend(), command_hdr.Name) == - dupl_symbols.cend()) { - dupl_symbols.emplace_back(command_hdr.Name); - } - - if (kVerbose) kConsoleOut << "Found duplicate symbols of: " << command_hdr.Name << "\n"; - - kDuplicateSymbols = true; - } - } - } - - if (!dupl_symbols.empty()) { - for (auto& symbol : dupl_symbols) { - kConsoleOut << "Multiple symbols of: " << symbol << " detected, cannot continue.\n"; - } - - return LIBCOMPILER_EXEC_ERROR; - } - - // step 2.5: write program bytes. - - for (auto& struct_of_blob : kObjectBytes) { - output_fc.write(struct_of_blob.mBlob.data(), struct_of_blob.mBlob.size()); - } - - if (kVerbose) { - kConsoleOut << "Wrote contents of: " << kOutput << "\n"; - } - - // step 3: check if we have those symbols - - std::vector unreferenced_symbols; - - for (auto& command_hdr : command_headers) { - if (auto it = - std::find(not_found.begin(), not_found.end(), LibCompiler::String(command_hdr.Name)); - it != not_found.end()) { - unreferenced_symbols.emplace_back(command_hdr.Name); - } - } - - if (!unreferenced_symbols.empty()) { - for (auto& unreferenced_symbol : unreferenced_symbols) { - kConsoleOut << "Undefined symbol " << unreferenced_symbol << "\n"; - } - - return LIBCOMPILER_EXEC_ERROR; - } - - if ((!kStartFound || kDuplicateSymbols) && - (std::filesystem::exists(kOutput) || !unreferenced_symbols.empty())) { - if (kVerbose) { - kConsoleOut << "File: " << kOutput << ", is corrupt, removing file...\n"; - } - - return LIBCOMPILER_EXEC_ERROR; - } - - return LIBCOMPILER_SUCCESS; -} - -// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/Frontend.cc b/dev/LibCompiler/src/Frontend.cc new file mode 100644 index 0000000..69ff6fd --- /dev/null +++ b/dev/LibCompiler/src/Frontend.cc @@ -0,0 +1,51 @@ +/* ------------------------------------------- + + Copyright (C) 2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#include + +namespace LibCompiler { +/// find the perfect matching word in a haystack. +/// \param haystack base string +/// \param needle the string we search for. +/// \return if we found it or not. +bool find_word(std::string haystack, std::string needle) noexcept { + auto index = haystack.find(needle); + + // check for needle validity. + if (index == std::string::npos) return false; + + // declare lambda + auto not_part_of_word = [&](int index) { + if (std::isspace(haystack[index]) || std::ispunct(haystack[index])) return true; + + if (index <= 0 || index >= haystack.size()) return true; + + return false; + }; + + return not_part_of_word(index - 1) && not_part_of_word(index + needle.size()); +} + +/// find a word within strict conditions and returns a range of it. +/// \param haystack +/// \param needle +/// \return position of needle. +std::size_t find_word_range(std::string haystack, std::string needle) noexcept { + auto index = haystack.find(needle); + + // check for needle validity. + if (index == std::string::npos) return false; + + if (!isalnum((haystack[index + needle.size() + 1])) && + !isdigit(haystack[index + needle.size() + 1]) && + !isalnum((haystack[index - needle.size() - 1])) && + !isdigit(haystack[index - needle.size() - 1])) { + return index; + } + + return std::string::npos; +} +} // namespace LibCompiler \ No newline at end of file diff --git a/dev/LibCompiler/src/Linker/DynamicLinkerPEF.cc b/dev/LibCompiler/src/Linker/DynamicLinkerPEF.cc new file mode 100644 index 0000000..0c6f3bf --- /dev/null +++ b/dev/LibCompiler/src/Linker/DynamicLinkerPEF.cc @@ -0,0 +1,695 @@ +/* ------------------------------------------- + + Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved + + @file DynamicLinker64PEF.cc + @brief: C++ 64-Bit PEF Linker. + +------------------------------------------- */ + +/// @author EL Mahrouss Amlal (amlal@nekernel.org) +/// @brief NeKernel 64-bit PEF Linker. +/// Last Rev: Sat Apr 19 CET 2025 +/// @note Do not look up for anything with .code64/.data64/.zero64! +/// It will be loaded when the program loader will start the image. + +//! Toolchain Kit. +#include + +//! Error ID +#include + +//! Assembler Kit +#include + +//! Preferred Executable Format +#include + +//! UUID header. +#include + +//! Release macros. +#include + +//! Advanced Executable Object Format. +#include + +//! LibCompiler utils. +#include + +//! I/O stream from std c++ +#include + +#define kLinkerVersionStr \ + "NeKernel 64-Bit Linker (Preferred Executable Format) %s, (c) Amlal El Mahrouss " \ + "2024-2025 " \ + "all rights reserved.\n" + +#define MemoryCopy(DST, SRC, SZ) memcpy(DST, SRC, SZ) +#define StringCompare(DST, SRC) strcmp(DST, SRC) + +#define kPefNoCpu (0U) +#define kPefNoSubCpu (0U) + +#define kLinkerDefaultOrigin kPefBaseOrigin +#define kLinkerId (0x5046FF) +#define kLinkerAbiContainer "__PEFContainer:ABI:" + +#define kPrintF printf +#define kLinkerSplash() kConsoleOut << std::printf(kLinkerVersionStr, kDistVersion) + +/// @brief PEF stack size symbol. +#define kLinkerStackSizeSymbol "__PEFSizeOfReserveStack" + +#define kConsoleOut \ + (std::cout << "\e[0;31m" \ + << "ld64: " \ + << "\e[0;97m") + +namespace Detail { +struct DynamicLinkerBlob final { + std::vector mBlob{}; // PEF code/bss/data blob. + UIntPtr mOffset{0UL}; // the offset of the PEF container header... +}; +} // namespace Detail + +enum { + kABITypeNull = 0, + kABITypeStart = 0x1010, /* The start of ABI list. */ + kABITypeNE = 0x5046, /* PF (NeKernel's PEF ABI) */ + kABITypeInvalid = 0xFFFF, +}; + +static LibCompiler::STLString kOutput = "a" kPefExt; +static Int32 kAbi = kABITypeNE; +static Int32 kSubArch = kPefNoSubCpu; +static Int32 kArch = LibCompiler::kPefArchInvalid; +static Bool kFatBinaryEnable = false; +static Bool kStartFound = false; +static Bool kDuplicateSymbols = false; + +/* ld64 is to be found, mld is to be found at runtime. */ +static const CharType* kLdDefineSymbol = ":UndefinedSymbol:"; +static const CharType* kLdDynamicSym = ":RuntimeSymbol:"; + +/* object code and list. */ +static std::vector kObjectList; +static std::vector kObjectBytes; + +/// @brief NE 64-bit Linker. +/// @note This linker is made for PEF executable, thus NE based OSes. +LIBCOMPILER_MODULE(DynamicLinker64PEF) { + bool is_executable = true; + + ::signal(SIGSEGV, Detail::drv_segfault_handler); + + /** + * @brief parse flags and trigger options. + */ + for (size_t linker_arg = 1; linker_arg < argc; ++linker_arg) { + if (StringCompare(argv[linker_arg], "-help") == 0) { + kLinkerSplash(); + + kConsoleOut << "-version: Show linker version.\n"; + kConsoleOut << "-help: Show linker help.\n"; + kConsoleOut << "-ld-verbose: Enable linker trace.\n"; + kConsoleOut << "-dylib: Output as a Dynamic PEF.\n"; + kConsoleOut << "-fat: Output as a FAT PEF.\n"; + kConsoleOut << "-32k: Output as a 32x0 PEF.\n"; + kConsoleOut << "-64k: Output as a 64x0 PEF.\n"; + kConsoleOut << "-amd64: Output as a AMD64 PEF.\n"; + kConsoleOut << "-rv64: Output as a RISC-V PEF.\n"; + kConsoleOut << "-power64: Output as a POWER PEF.\n"; + kConsoleOut << "-arm64: Output as a ARM64 PEF.\n"; + kConsoleOut << "-output: Select the output file name.\n"; + + return LIBCOMPILER_SUCCESS; + } else if (StringCompare(argv[linker_arg], "-version") == 0) { + kLinkerSplash(); + + return LIBCOMPILER_SUCCESS; + } else if (StringCompare(argv[linker_arg], "-fat-binary") == 0) { + kFatBinaryEnable = true; + + continue; + } else if (StringCompare(argv[linker_arg], "-64k") == 0) { + kArch = LibCompiler::kPefArch64000; + + continue; + } else if (StringCompare(argv[linker_arg], "-amd64") == 0) { + kArch = LibCompiler::kPefArchAMD64; + + continue; + } else if (StringCompare(argv[linker_arg], "-32k") == 0) { + kArch = LibCompiler::kPefArch32000; + + continue; + } else if (StringCompare(argv[linker_arg], "-power64") == 0) { + kArch = LibCompiler::kPefArchPowerPC; + + continue; + } else if (StringCompare(argv[linker_arg], "-riscv64") == 0) { + kArch = LibCompiler::kPefArchRISCV; + + continue; + } else if (StringCompare(argv[linker_arg], "-arm64") == 0) { + kArch = LibCompiler::kPefArchARM64; + + continue; + } else if (StringCompare(argv[linker_arg], "-ld-verbose") == 0) { + kVerbose = true; + + continue; + } else if (StringCompare(argv[linker_arg], "-dylib") == 0) { + if (kOutput.empty()) { + continue; + } + + if (kOutput.find(kPefExt) != LibCompiler::STLString::npos) + kOutput.erase(kOutput.find(kPefExt), strlen(kPefExt)); + + kOutput += kPefDylibExt; + + is_executable = false; + + continue; + } else if (StringCompare(argv[linker_arg], "-output") == 0) { + if ((linker_arg + 1) > argc) continue; + + kOutput = argv[linker_arg + 1]; + ++linker_arg; + + continue; + } else { + if (argv[linker_arg][0] == '-') { + kConsoleOut << "unknown flag: " << argv[linker_arg] << "\n"; + return EXIT_FAILURE; + } + + kObjectList.emplace_back(argv[linker_arg]); + + continue; + } + } + + if (kOutput.empty()) { + kConsoleOut << "no output filename set." << std::endl; + return LIBCOMPILER_EXEC_ERROR; + } else if (kObjectList.empty()) { + kConsoleOut << "no input files." << std::endl; + return LIBCOMPILER_EXEC_ERROR; + } else { + namespace FS = std::filesystem; + + // check for existing files, if they don't throw an error. + for (auto& obj : kObjectList) { + if (!FS::exists(obj)) { + // if filesystem doesn't find file + // -> throw error. + kConsoleOut << "no such file: " << obj << std::endl; + return LIBCOMPILER_EXEC_ERROR; + } + } + } + + // PEF expects a valid target architecture when outputing a binary. + if (kArch == LibCompiler::kPefArchInvalid) { + kConsoleOut << "no target architecture set, can't continue." << std::endl; + return LIBCOMPILER_EXEC_ERROR; + } + + LibCompiler::PEFContainer pef_container{}; + + int32_t archs = kArch; + + pef_container.Count = 0UL; + pef_container.Kind = is_executable ? LibCompiler::kPefKindExec : LibCompiler::kPefKindDylib; + pef_container.SubCpu = kSubArch; + pef_container.Linker = kLinkerId; // Amlal El Mahrouss Linker + pef_container.Abi = kAbi; // Multi-Processor UX ABI + pef_container.Magic[0] = kPefMagic[kFatBinaryEnable ? 2 : 0]; + pef_container.Magic[1] = kPefMagic[1]; + pef_container.Magic[2] = kPefMagic[kFatBinaryEnable ? 0 : 2]; + pef_container.Magic[3] = kPefMagic[3]; + pef_container.Version = kPefVersion; + + // specify the start address, can be 0x10000 + pef_container.Start = kLinkerDefaultOrigin; + pef_container.HdrSz = sizeof(LibCompiler::PEFContainer); + pef_container.Checksum = 0UL; + + std::ofstream output_fc(kOutput, std::ofstream::binary); + + if (output_fc.bad()) { + if (kVerbose) { + kConsoleOut << "error: " << strerror(errno) << "\n"; + } + + return LIBCOMPILER_FILE_NOT_FOUND; + } + + //! Read AE to convert as PEF. + + std::vector command_headers; + LibCompiler::Utils::AEReadableProtocol reader_protocol{}; + + for (const auto& objectFile : kObjectList) { + if (!std::filesystem::exists(objectFile)) continue; + + LibCompiler::AEHeader hdr{}; + + reader_protocol.FP = std::ifstream(objectFile, std::ifstream::binary); + reader_protocol.FP >> hdr; + + if (hdr.fMagic[0] == kAEMag0 && hdr.fMagic[1] == kAEMag1 && + hdr.fSize == sizeof(LibCompiler::AEHeader)) { + if (hdr.fArch != kArch) { + if (kVerbose) kConsoleOut << "is this a FAT binary? : "; + + if (!kFatBinaryEnable) { + if (kVerbose) kConsoleOut << "not a FAT binary.\n"; + + kConsoleOut << "object " << objectFile + << " is a different kind of architecture and output isn't " + "treated as a FAT binary." + << std::endl; + + return LIBCOMPILER_FAT_ERROR; + } else { + if (kVerbose) { + kConsoleOut << "Architecture matches what we expect.\n"; + } + } + } + + // append arch type to archs varaible. + archs |= hdr.fArch; + std::size_t cnt = hdr.fCount; + + if (kVerbose) kConsoleOut << "Object header found, record count: " << cnt << "\n"; + + pef_container.Count = cnt; + + char_type* raw_ae_records = new char_type[cnt * sizeof(LibCompiler::AERecordHeader)]; + + if (!raw_ae_records) { + if (kVerbose) kConsoleOut << "Allocation failure for records of count: " << cnt << "\n"; + } + + memset(raw_ae_records, 0, cnt * sizeof(LibCompiler::AERecordHeader)); + + auto* ae_records = reader_protocol.Read(raw_ae_records, cnt); + + auto org = kLinkerDefaultOrigin; + + for (size_t ae_record_index = 0; ae_record_index < cnt; ++ae_record_index) { + LibCompiler::PEFCommandHeader command_header{0}; + std::size_t offset_of_obj = ae_records[ae_record_index].fOffset; + + MemoryCopy(command_header.Name, ae_records[ae_record_index].fName, kPefNameLen); + + LibCompiler::STLString cmd_hdr_name(command_header.Name); + + // check this header if it's any valid. + if (cmd_hdr_name.find(kPefCode64) == LibCompiler::STLString::npos && + cmd_hdr_name.find(kPefData64) == LibCompiler::STLString::npos && + cmd_hdr_name.find(kPefZero64) == LibCompiler::STLString::npos) { + if (cmd_hdr_name.find(kPefStart) == LibCompiler::STLString::npos && + *command_header.Name == 0) { + if (cmd_hdr_name.find(kLdDefineSymbol) != LibCompiler::STLString::npos) { + goto ld_mark_header; + } else { + continue; + } + } + } + + if (cmd_hdr_name.find(kPefStart) != LibCompiler::STLString::npos && + cmd_hdr_name.find(kPefCode64) != LibCompiler::STLString::npos) { + kStartFound = true; + } + + ld_mark_header: + command_header.Offset = offset_of_obj; + command_header.Kind = ae_records[ae_record_index].fKind; + command_header.Size = ae_records[ae_record_index].fSize; + command_header.Cpu = hdr.fArch; + command_header.VMAddress = org; /// TODO: + command_header.SubCpu = hdr.fSubArch; + + org += command_header.Size; + + if (kVerbose) { + kConsoleOut << "Record: " << ae_records[ae_record_index].fName << " is marked.\n"; + + kConsoleOut << "Offset: " << command_header.Offset << "\n"; + } + + command_headers.emplace_back(command_header); + } + + delete[] raw_ae_records; + raw_ae_records = nullptr; + + std::vector bytes; + bytes.resize(hdr.fCodeSize); + + reader_protocol.FP.seekg(std::streamsize(hdr.fStartCode)); + reader_protocol.FP.read(bytes.data(), std::streamsize(hdr.fCodeSize)); + + kObjectBytes.push_back({.mBlob = bytes, .mOffset = hdr.fStartCode}); + + // Blob was written, close fp. + + reader_protocol.FP.close(); + + continue; + } + + kConsoleOut << "Not an AE container: " << objectFile << std::endl; + // don't continue, it is a fatal error. + return LIBCOMPILER_EXEC_ERROR; + } + + pef_container.Cpu = archs; + + output_fc << pef_container; + + if (kVerbose) { + kConsoleOut << "Wrote container to: " << output_fc.tellp() << ".\n"; + } + + output_fc.seekp(std::streamsize(pef_container.HdrSz)); + + std::vector not_found; + std::vector symbols; + + // step 2: check for errors (multiple symbols, undefined ones) + + for (auto& command_hdr : command_headers) { + // check if this symbol needs to be resolved. + if (LibCompiler::STLString(command_hdr.Name).find(kLdDefineSymbol) != + LibCompiler::STLString::npos && + LibCompiler::STLString(command_hdr.Name).find(kLdDynamicSym) == + LibCompiler::STLString::npos) { + if (kVerbose) kConsoleOut << "Found undefined symbol: " << command_hdr.Name << "\n"; + + if (auto it = std::find(not_found.begin(), not_found.end(), + LibCompiler::STLString(command_hdr.Name)); + it == not_found.end()) { + not_found.emplace_back(command_hdr.Name); + } + } + + symbols.emplace_back(command_hdr.Name); + } + + // Now try to solve these symbols. + + for (size_t not_found_idx = 0; not_found_idx < command_headers.size(); ++not_found_idx) { + if (const auto it = std::find(not_found.begin(), not_found.end(), + LibCompiler::STLString(command_headers[not_found_idx].Name)); + it != not_found.end()) { + LibCompiler::STLString symbol_imp = *it; + + if (symbol_imp.find(kLdDefineSymbol) == LibCompiler::STLString::npos) continue; + + // erase the lookup prefix. + symbol_imp.erase(0, symbol_imp.find(kLdDefineSymbol) + strlen(kLdDefineSymbol)); + + // demangle everything. + while (symbol_imp.find('$') != LibCompiler::STLString::npos) + symbol_imp.erase(symbol_imp.find('$'), 1); + + // the reason we do is because, this may not match the symbol, and we need + // to look for other matching symbols. + for (auto& command_hdr : command_headers) { + if (LibCompiler::STLString(command_hdr.Name).find(symbol_imp) != + LibCompiler::STLString::npos && + LibCompiler::STLString(command_hdr.Name).find(kLdDefineSymbol) == + LibCompiler::STLString::npos) { + LibCompiler::STLString undefined_symbol = command_hdr.Name; + auto result_of_sym = undefined_symbol.substr(undefined_symbol.find(symbol_imp)); + + for (int i = 0; result_of_sym[i] != 0; ++i) { + if (result_of_sym[i] != symbol_imp[i]) goto ld_continue_search; + } + + not_found.erase(it); + + if (kVerbose) kConsoleOut << "Found symbol: " << command_hdr.Name << "\n"; + + break; + } + } + + ld_continue_search: + continue; + } + } + + // step 3: check for errors (recheck if we have those symbols.) + + if (!kStartFound && is_executable) { + if (kVerbose) + kConsoleOut << "Undefined entrypoint: " << kPefStart + << ", you may have forget to link " + "against the C++ runtime library.\n"; + + kConsoleOut << "Undefined entrypoint " << kPefStart << " for executable: " << kOutput << "\n"; + } + + // step 4: write all PEF commands. + + LibCompiler::PEFCommandHeader date_cmd_hdr{}; + + time_t timestamp = time(nullptr); + + LibCompiler::STLString timeStampStr = "Container:BuildEpoch:"; + timeStampStr += std::to_string(timestamp); + + strncpy(date_cmd_hdr.Name, timeStampStr.c_str(), timeStampStr.size()); + + date_cmd_hdr.Flags = 0; + date_cmd_hdr.Kind = LibCompiler::kPefZero; + date_cmd_hdr.Offset = output_fc.tellp(); + date_cmd_hdr.Size = timeStampStr.size(); + + command_headers.push_back(date_cmd_hdr); + + LibCompiler::PEFCommandHeader abi_cmd_hdr{}; + + LibCompiler::STLString abi = kLinkerAbiContainer; + + switch (kArch) { + case LibCompiler::kPefArchAMD64: { + abi += "MSFT"; + break; + } + case LibCompiler::kPefArchPowerPC: { + abi += "SYSV"; + break; + } + case LibCompiler::kPefArch32000: + case LibCompiler::kPefArch64000: { + abi += "_NEP"; + break; + } + default: { + abi += "_IDK"; + break; + } + } + + MemoryCopy(abi_cmd_hdr.Name, abi.c_str(), abi.size()); + + abi_cmd_hdr.Size = abi.size(); + abi_cmd_hdr.Offset = output_fc.tellp(); + abi_cmd_hdr.Flags = 0; + abi_cmd_hdr.Kind = LibCompiler::kPefLinkerID; + + command_headers.push_back(abi_cmd_hdr); + + LibCompiler::PEFCommandHeader stack_cmd_hdr{0}; + + stack_cmd_hdr.Cpu = kArch; + stack_cmd_hdr.Flags = 0; + stack_cmd_hdr.Size = sizeof(uintptr_t); + stack_cmd_hdr.Offset = 0; + + MemoryCopy(stack_cmd_hdr.Name, kLinkerStackSizeSymbol, strlen(kLinkerStackSizeSymbol)); + + command_headers.push_back(stack_cmd_hdr); + + LibCompiler::PEFCommandHeader uuid_cmd_hdr{}; + + std::random_device rd; + + auto seedData = std::array{}; + std::generate(std::begin(seedData), std::end(seedData), std::ref(rd)); + std::seed_seq seq(std::begin(seedData), std::end(seedData)); + std::mt19937 generator(seq); + + auto gen = uuids::uuid_random_generator{generator}; + uuids::uuid id = gen(); + auto uuidStr = uuids::to_string(id); + + MemoryCopy(uuid_cmd_hdr.Name, "Container:GUID:4:", strlen("Container:GUID:4:")); + MemoryCopy(uuid_cmd_hdr.Name + strlen("Container:GUID:4:"), uuidStr.c_str(), uuidStr.size()); + + uuid_cmd_hdr.Size = strlen(uuid_cmd_hdr.Name); + uuid_cmd_hdr.Offset = output_fc.tellp(); + uuid_cmd_hdr.Flags = LibCompiler::kPefLinkerID; + uuid_cmd_hdr.Kind = LibCompiler::kPefZero; + + command_headers.push_back(uuid_cmd_hdr); + + // prepare a symbol vector. + std::vector undef_symbols; + std::vector dupl_symbols; + std::vector resolve_symbols; + + constexpr Int32 kPaddingOffset = 16; + + size_t previous_offset = + (command_headers.size() * sizeof(LibCompiler::PEFCommandHeader)) + kPaddingOffset; + + LibCompiler::PEFCommandHeader end_exec_hdr; + + end_exec_hdr.Offset = output_fc.tellp(); + end_exec_hdr.Flags = LibCompiler::kPefLinkerID; + end_exec_hdr.Kind = LibCompiler::kPefZero; + + MemoryCopy(end_exec_hdr.Name, "Container:Exec:END", strlen("Container:Exec:END")); + + end_exec_hdr.Size = strlen(end_exec_hdr.Name); + + command_headers.push_back(end_exec_hdr); + + // Finally write down the command headers. + // And check for any duplications + for (size_t commandHeaderIndex = 0UL; commandHeaderIndex < command_headers.size(); + ++commandHeaderIndex) { + if (LibCompiler::STLString(command_headers[commandHeaderIndex].Name).find(kLdDefineSymbol) != + LibCompiler::STLString::npos && + LibCompiler::STLString(command_headers[commandHeaderIndex].Name).find(kLdDynamicSym) == + LibCompiler::STLString::npos) { + // ignore :UndefinedSymbol: headers, they do not contain code. + continue; + } + + LibCompiler::STLString symbol_name = command_headers[commandHeaderIndex].Name; + + if (!symbol_name.empty()) { + undef_symbols.emplace_back(symbol_name); + } + + command_headers[commandHeaderIndex].Offset += previous_offset; + previous_offset += command_headers[commandHeaderIndex].Size; + + LibCompiler::STLString name = command_headers[commandHeaderIndex].Name; + + /// so this is valid when we get to the entrypoint. + /// it is always a code64 container. And should equal to kPefStart as well. + /// this chunk of code updates the pef_container.Start with the updated offset. + if (name.find(kPefStart) != LibCompiler::STLString::npos && + name.find(kPefCode64) != LibCompiler::STLString::npos) { + pef_container.Start = command_headers[commandHeaderIndex].Offset; + auto tellCurPos = output_fc.tellp(); + + output_fc.seekp(0); + output_fc << pef_container; + + output_fc.seekp(tellCurPos); + } + + if (kVerbose) { + kConsoleOut << "Command name: " << name << "\n"; + kConsoleOut << "VMAddress of command content: " << command_headers[commandHeaderIndex].Offset + << "\n"; + } + + output_fc << command_headers[commandHeaderIndex]; + + for (size_t sub_command_header_index = 0UL; sub_command_header_index < command_headers.size(); + ++sub_command_header_index) { + if (sub_command_header_index == commandHeaderIndex) continue; + + if (LibCompiler::STLString(command_headers[sub_command_header_index].Name) + .find(kLdDefineSymbol) != LibCompiler::STLString::npos && + LibCompiler::STLString(command_headers[sub_command_header_index].Name) + .find(kLdDynamicSym) == LibCompiler::STLString::npos) { + if (kVerbose) { + kConsoleOut << "Ignoring :UndefinedSymbol: headers...\n"; + } + + // ignore :UndefinedSymbol: headers, they do not contain code. + continue; + } + + auto& command_hdr = command_headers[sub_command_header_index]; + + if (command_hdr.Name == LibCompiler::STLString(command_headers[commandHeaderIndex].Name)) { + if (std::find(dupl_symbols.cbegin(), dupl_symbols.cend(), command_hdr.Name) == + dupl_symbols.cend()) { + dupl_symbols.emplace_back(command_hdr.Name); + } + + if (kVerbose) kConsoleOut << "Found duplicate symbols of: " << command_hdr.Name << "\n"; + + kDuplicateSymbols = true; + } + } + } + + if (!dupl_symbols.empty()) { + for (auto& symbol : dupl_symbols) { + kConsoleOut << "Multiple symbols of: " << symbol << " detected, cannot continue.\n"; + } + + return LIBCOMPILER_EXEC_ERROR; + } + + // step 2.5: write program bytes. + + for (auto& struct_of_blob : kObjectBytes) { + output_fc.write(struct_of_blob.mBlob.data(), struct_of_blob.mBlob.size()); + } + + if (kVerbose) { + kConsoleOut << "Wrote contents of: " << kOutput << "\n"; + } + + // step 3: check if we have those symbols + + std::vector unreferenced_symbols; + + for (auto& command_hdr : command_headers) { + if (auto it = + std::find(not_found.begin(), not_found.end(), LibCompiler::STLString(command_hdr.Name)); + it != not_found.end()) { + unreferenced_symbols.emplace_back(command_hdr.Name); + } + } + + if (!unreferenced_symbols.empty()) { + for (auto& unreferenced_symbol : unreferenced_symbols) { + kConsoleOut << "Undefined symbol " << unreferenced_symbol << "\n"; + } + + return LIBCOMPILER_EXEC_ERROR; + } + + if ((!kStartFound || kDuplicateSymbols) && + (std::filesystem::exists(kOutput) || !unreferenced_symbols.empty())) { + if (kVerbose) { + kConsoleOut << "File: " << kOutput << ", is corrupt, removing file...\n"; + } + + return LIBCOMPILER_EXEC_ERROR; + } + + return LIBCOMPILER_SUCCESS; +} + +// Last rev 13-1-24 diff --git a/dev/LibCompiler/src/PP/CPlusPlusCompilerPreProcessor.cc b/dev/LibCompiler/src/PP/CPlusPlusCompilerPreProcessor.cc new file mode 100644 index 0000000..2c7c551 --- /dev/null +++ b/dev/LibCompiler/src/PP/CPlusPlusCompilerPreProcessor.cc @@ -0,0 +1,903 @@ +/* + * ======================================================== + * + * bpp + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +/// BUGS: 0 + +#include +#include +#include +#include +#include +#include +#include +#include + +#define kMacroPrefix '#' + +/// @author EL Mahrouss Amlal (amlel) +/// @file bpp.cxx +/// @brief Preprocessor. + +typedef Int32 (*bpp_parser_fn_t)(LibCompiler::STLString& line, std::ifstream& hdr_file, + std::ofstream& pp_out); + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief Preprocessor internal types. + +///////////////////////////////////////////////////////////////////////////////////////// + +namespace Detail { +enum { + kInvalid = 0, + kEqual = 100, + kGreaterEqThan, + kLesserEqThan, + kGreaterThan, + kLesserThan, + kNotEqual, + kCount = 6, +}; + +struct bpp_macro_condition final { + int32_t fType; + LibCompiler::STLString fTypeName; + + void Print() { + std::cout << "type: " << fType << "\n"; + std::cout << "type_name: " << fTypeName << "\n"; + } +}; + +struct bpp_macro final { + std::vector fArgs; + LibCompiler::STLString fName; + LibCompiler::STLString fValue; + + void Print() { + std::cout << "name: " << fName << "\n"; + std::cout << "value: " << fValue << "\n"; + + for (auto& arg : fArgs) { + std::cout << "arg: " << arg << "\n"; + } + } +}; +} // namespace Detail + +static std::vector kFiles; +static std::vector kMacros; +static std::vector kIncludes; + +static LibCompiler::STLString kWorkingDir = ""; + +///////////////////////////////////////////////////////////////////////////////////////// + +// @name bpp_parse_if_condition +// @brief parse #if condition + +///////////////////////////////////////////////////////////////////////////////////////// + +int32_t bpp_parse_if_condition(Detail::bpp_macro_condition& cond, Detail::bpp_macro& macro, + bool& inactive_code, bool& defined, + LibCompiler::STLString& macro_str) { + if (cond.fType == Detail::kEqual) { + auto substr_macro = macro_str.substr(macro_str.find(macro.fName) + macro.fName.size()); + + if (substr_macro.find(macro.fValue) != LibCompiler::STLString::npos) { + if (macro.fValue == "0") { + defined = false; + inactive_code = true; + + return 1; + } + + defined = true; + inactive_code = false; + + return 1; + } + } else if (cond.fType == Detail::kNotEqual) { + auto substr_macro = macro_str.substr(macro_str.find(macro.fName) + macro.fName.size()); + + if (substr_macro.find(macro.fName) != LibCompiler::STLString::npos) { + if (substr_macro.find(macro.fValue) != LibCompiler::STLString::npos) { + defined = false; + inactive_code = true; + + return 1; + } + + defined = true; + inactive_code = false; + + return 1; + } + + return 0; + } + + auto substr_macro = macro_str.substr(macro_str.find(macro.fName) + macro.fName.size()); + + LibCompiler::STLString number; + + for (auto& macro_num : kMacros) { + if (substr_macro.find(macro_num.fName) != LibCompiler::STLString::npos) { + for (size_t i = 0; i < macro_num.fName.size(); ++i) { + if (isdigit(macro_num.fValue[i])) { + number += macro_num.fValue[i]; + } else { + number.clear(); + break; + } + } + + break; + } + } + + size_t y = 2; + + /* last try */ + for (; y < macro_str.size(); y++) { + if (isdigit(macro_str[y])) { + for (size_t x = y; x < macro_str.size(); x++) { + if (macro_str[x] == ' ') break; + + number += macro_str[x]; + } + + break; + } + } + + size_t rhs = atol(macro.fValue.c_str()); + size_t lhs = atol(number.c_str()); + + if (lhs == 0) { + number.clear(); + ++y; + + for (; y < macro_str.size(); y++) { + if (isdigit(macro_str[y])) { + for (size_t x = y; x < macro_str.size(); x++) { + if (macro_str[x] == ' ') break; + + number += macro_str[x]; + } + + break; + } + } + + lhs = atol(number.c_str()); + } + + if (cond.fType == Detail::kGreaterThan) { + if (lhs < rhs) { + defined = true; + inactive_code = false; + + return 1; + } + + return 0; + } + + if (cond.fType == Detail::kGreaterEqThan) { + if (lhs <= rhs) { + defined = true; + inactive_code = false; + + return 1; + } + + return 0; + } + + if (cond.fType == Detail::kLesserEqThan) { + if (lhs >= rhs) { + defined = true; + inactive_code = false; + + return 1; + } + + return 0; + } + + if (cond.fType == Detail::kLesserThan) { + if (lhs > rhs) { + defined = true; + inactive_code = false; + + return 1; + } + + return 0; + } + + return 0; +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief stores every included file here. + +///////////////////////////////////////////////////////////////////////////////////////// + +std::vector kAllIncludes; + +///////////////////////////////////////////////////////////////////////////////////////// + +// @name bpp_parse_file +// @brief parse file to preprocess it. + +///////////////////////////////////////////////////////////////////////////////////////// + +void bpp_parse_file(std::ifstream& hdr_file, std::ofstream& pp_out) { + LibCompiler::STLString hdr_line; + LibCompiler::STLString line_after_include; + + bool inactive_code = false; + bool defined = false; + + try { + while (std::getline(hdr_file, hdr_line)) { + if (inactive_code) { + if (hdr_line.find("#endif") == LibCompiler::STLString::npos) { + continue; + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("#endif") != LibCompiler::STLString::npos) { + inactive_code = false; + } + + if (hdr_line.find("*/") != LibCompiler::STLString::npos) { + continue; + } + } + + if (hdr_line.find("--/") != LibCompiler::STLString::npos) { + hdr_line.erase(hdr_line.find("--/")); + } + + if (hdr_line.find("--*") != LibCompiler::STLString::npos) { + inactive_code = true; + // get rid of comment. + hdr_line.erase(hdr_line.find("--*")); + } + + /// BPP 'brief' documentation. + if (hdr_line.find("@brief") != LibCompiler::STLString::npos) { + hdr_line.erase(hdr_line.find("@brief")); + + // TODO: Write an .html or append to it. + } + + if (hdr_line[0] == kMacroPrefix && hdr_line.find("endif") != LibCompiler::STLString::npos) { + if (!defined && inactive_code) { + inactive_code = false; + defined = false; + + continue; + } + + continue; + } + + if (!defined && inactive_code) { + continue; + } + + if (defined && inactive_code) { + continue; + } + + for (auto macro : kMacros) { + if (LibCompiler::find_word(hdr_line, macro.fName)) { + if (hdr_line.substr(hdr_line.find(macro.fName)).find(macro.fName + '(') != + LibCompiler::STLString::npos) { + if (!macro.fArgs.empty()) { + LibCompiler::STLString symbol_val = macro.fValue; + std::vector args; + + size_t x_arg_indx = 0; + + LibCompiler::STLString line_after_define = hdr_line; + LibCompiler::STLString str_arg; + + if (line_after_define.find("(") != LibCompiler::STLString::npos) { + line_after_define.erase(0, line_after_define.find("(") + 1); + + for (auto& subc : line_after_define) { + if (subc == ' ' || subc == '\t') continue; + + if (subc == ',' || subc == ')') { + if (str_arg.empty()) continue; + + args.push_back(str_arg); + + str_arg.clear(); + + continue; + } + + str_arg.push_back(subc); + } + } + + for (auto arg : macro.fArgs) { + if (symbol_val.find(macro.fArgs[x_arg_indx]) != LibCompiler::STLString::npos) { + symbol_val.replace(symbol_val.find(macro.fArgs[x_arg_indx]), + macro.fArgs[x_arg_indx].size(), args[x_arg_indx]); + ++x_arg_indx; + } else { + throw std::runtime_error("cppdrv: Internal error."); + } + } + + auto len = macro.fName.size(); + len += symbol_val.size(); + len += 2; // ( and ) + + hdr_line.erase(hdr_line.find(")"), 1); + + hdr_line.replace(hdr_line.find(hdr_line.substr(hdr_line.find(macro.fName + '('))), + len, symbol_val); + } else { + auto value = macro.fValue; + + hdr_line.replace(hdr_line.find(macro.fName), macro.fName.size(), value); + } + } + } + } + + if (hdr_line[0] == kMacroPrefix && hdr_line.find("define ") != LibCompiler::STLString::npos) { + auto line_after_define = hdr_line.substr(hdr_line.find("define ") + strlen("define ")); + + LibCompiler::STLString macro_value; + LibCompiler::STLString macro_key; + + std::size_t pos = 0UL; + + std::vector args; + bool on_args = false; + + for (auto& ch : line_after_define) { + ++pos; + + if (ch == '(') { + on_args = true; + continue; + } + + if (ch == ')') { + on_args = false; + continue; + } + + if (ch == '\\') continue; + + if (on_args) continue; + + if (ch == ' ') { + for (size_t i = pos; i < line_after_define.size(); i++) { + macro_value += line_after_define[i]; + } + + break; + } + + macro_key += ch; + } + + LibCompiler::STLString str; + + if (line_after_define.find("(") != LibCompiler::STLString::npos) { + line_after_define.erase(0, line_after_define.find("(") + 1); + + for (auto& subc : line_after_define) { + if (subc == ',' || subc == ')') { + if (str.empty()) continue; + + args.push_back(str); + + str.clear(); + + continue; + } + + str.push_back(subc); + } + } + + Detail::bpp_macro macro; + + macro.fArgs = args; + macro.fName = macro_key; + macro.fValue = macro_value; + + kMacros.emplace_back(macro); + + continue; + } + + if (hdr_line[0] != kMacroPrefix) { + if (inactive_code) { + continue; + } + + pp_out << hdr_line << std::endl; + + continue; + } + + if (hdr_line[0] == kMacroPrefix && hdr_line.find("ifndef") != LibCompiler::STLString::npos) { + auto line_after_ifndef = hdr_line.substr(hdr_line.find("ifndef") + strlen("ifndef") + 1); + LibCompiler::STLString macro; + + for (auto& ch : line_after_ifndef) { + if (ch == ' ') { + break; + } + + macro += ch; + } + + if (macro == "0") { + defined = true; + inactive_code = false; + continue; + } + + if (macro == "1") { + defined = false; + inactive_code = true; + + continue; + } + + bool found = false; + + defined = true; + inactive_code = false; + + for (auto& macro_ref : kMacros) { + if (hdr_line.find(macro_ref.fName) != LibCompiler::STLString::npos) { + found = true; + break; + } + } + + if (found) { + defined = false; + inactive_code = true; + + continue; + } + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("else") != LibCompiler::STLString::npos) { + if (!defined && inactive_code) { + inactive_code = false; + defined = true; + + continue; + } else { + defined = false; + inactive_code = true; + + continue; + } + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("ifdef") != LibCompiler::STLString::npos) { + auto line_after_ifdef = hdr_line.substr(hdr_line.find("ifdef") + strlen("ifdef") + 1); + LibCompiler::STLString macro; + + for (auto& ch : line_after_ifdef) { + if (ch == ' ') { + break; + } + + macro += ch; + } + + if (macro == "0") { + defined = false; + inactive_code = true; + + continue; + } + + if (macro == "1") { + defined = true; + inactive_code = false; + + continue; + } + + defined = false; + inactive_code = true; + + for (auto& macro_ref : kMacros) { + if (hdr_line.find(macro_ref.fName) != LibCompiler::STLString::npos) { + defined = true; + inactive_code = false; + + break; + } + } + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("if") != LibCompiler::STLString::npos) { + inactive_code = true; + + std::vector bpp_macro_condition_list = { + { + .fType = Detail::kEqual, + .fTypeName = "==", + }, + { + .fType = Detail::kNotEqual, + .fTypeName = "!=", + }, + { + .fType = Detail::kLesserThan, + .fTypeName = "<", + }, + { + .fType = Detail::kGreaterThan, + .fTypeName = ">", + }, + { + .fType = Detail::kLesserEqThan, + .fTypeName = "<=", + }, + { + .fType = Detail::kGreaterEqThan, + .fTypeName = ">=", + }, + }; + + int32_t good_to_go = 0; + + for (auto& macro_condition : bpp_macro_condition_list) { + if (hdr_line.find(macro_condition.fTypeName) != LibCompiler::STLString::npos) { + for (auto& found_macro : kMacros) { + if (hdr_line.find(found_macro.fName) != LibCompiler::STLString::npos) { + good_to_go = bpp_parse_if_condition(macro_condition, found_macro, inactive_code, + defined, hdr_line); + + break; + } + } + } + } + + if (good_to_go) continue; + + auto line_after_if = hdr_line.substr(hdr_line.find("if") + strlen("if") + 1); + LibCompiler::STLString macro; + + for (auto& ch : line_after_if) { + if (ch == ' ') { + break; + } + + macro += ch; + } + + if (macro == "0") { + defined = false; + inactive_code = true; + continue; + } + + if (macro == "1") { + defined = true; + inactive_code = false; + + continue; + } + + // last try, is it defined to be one? + for (auto& macro_ref : kMacros) { + if (macro_ref.fName.find(macro) != LibCompiler::STLString::npos && + macro_ref.fValue == "1") { + inactive_code = false; + defined = true; + + break; + } + } + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("warning") != LibCompiler::STLString::npos) { + auto line_after_warning = hdr_line.substr(hdr_line.find("warning") + strlen("warning") + 1); + LibCompiler::STLString message; + + for (auto& ch : line_after_warning) { + if (ch == '\r' || ch == '\n') { + break; + } + + message += ch; + } + + std::cout << "warn: " << message << std::endl; + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("error") != LibCompiler::STLString::npos) { + auto line_after_warning = hdr_line.substr(hdr_line.find("error") + strlen("error") + 1); + LibCompiler::STLString message; + + for (auto& ch : line_after_warning) { + if (ch == '\r' || ch == '\n') { + break; + } + + message += ch; + } + + throw std::runtime_error("error: " + message); + } else if (hdr_line[0] == kMacroPrefix && + hdr_line.find("include ") != LibCompiler::STLString::npos) { + line_after_include = hdr_line.substr(hdr_line.find("include ") + strlen("include ")); + + kIncludeFile: + auto it = std::find(kAllIncludes.cbegin(), kAllIncludes.cend(), line_after_include); + + if (it != kAllIncludes.cend()) { + continue; + } + + LibCompiler::STLString path; + + kAllIncludes.push_back(line_after_include); + + bool enable = false; + bool not_local = false; + + for (auto& ch : line_after_include) { + if (ch == ' ') continue; + + if (ch == '<') { + not_local = true; + enable = true; + + continue; + } + + if (ch == '\"') { + not_local = false; + enable = true; + continue; + } + + if (enable) { + path += ch; + } + } + + if (not_local) { + bool open = false; + + if (path.ends_with('>')) { + path.erase(path.find('>')); + } + + if (path.ends_with('"')) { + path.erase(path.find('"')); + } + + for (auto& include : kIncludes) { + LibCompiler::STLString header_path = include; + header_path.push_back('/'); + header_path += path; + + std::ifstream header(header_path); + + if (!header.is_open()) continue; + + open = true; + + bpp_parse_file(header, pp_out); + + break; + } + + if (!open) { + throw std::runtime_error("cppdrv: no such include file: " + path); + } + } else { + std::ifstream header(path); + + if (!header.is_open()) throw std::runtime_error("cppdrv: no such include file: " + path); + + bpp_parse_file(header, pp_out); + } + } else { + std::cerr << ("cppdrv: unknown pre-processor directive, " + hdr_line) << "\n"; + continue; + } + } + } catch (std::out_of_range& oor) { + return; + } +} + +///////////////////////////////////////////////////////////////////////////////////////// + +// @brief main entrypoint of app. + +///////////////////////////////////////////////////////////////////////////////////////// + +LIBCOMPILER_MODULE(CPlusPlusPreprocessorMain) { + try { + bool skip = false; + bool double_skip = false; + + Detail::bpp_macro macro_1; + + macro_1.fName = "__true"; + macro_1.fValue = "1"; + + kMacros.push_back(macro_1); + + Detail::bpp_macro macro_unreachable; + + macro_unreachable.fName = "__unreachable"; + macro_unreachable.fValue = "__libcompiler_unreachable"; + + kMacros.push_back(macro_unreachable); + + Detail::bpp_macro macro_unused; + + macro_unreachable.fName = "__unused"; + macro_unreachable.fValue = "__libcompiler_unused"; + + kMacros.push_back(macro_unused); + + Detail::bpp_macro macro_0; + + macro_0.fName = "__false"; + macro_0.fValue = "0"; + + kMacros.push_back(macro_0); + + Detail::bpp_macro macro_zka; + + macro_zka.fName = "__LIBCOMPILER__"; + macro_zka.fValue = "1"; + + kMacros.push_back(macro_zka); + + Detail::bpp_macro macro_cxx; + + macro_cxx.fName = "__cplusplus"; + macro_cxx.fValue = "202302L"; + + kMacros.push_back(macro_cxx); + + Detail::bpp_macro macro_size_t; + macro_size_t.fName = "__SIZE_TYPE__"; + macro_size_t.fValue = "unsigned long long int"; + + kMacros.push_back(macro_size_t); + + macro_size_t.fName = "__UINT32_TYPE__"; + macro_size_t.fValue = "unsigned int"; + + kMacros.push_back(macro_size_t); + + macro_size_t.fName = "__UINTPTR_TYPE__"; + macro_size_t.fValue = "unsigned long long int"; + + kMacros.push_back(macro_size_t); + + for (auto index = 1UL; index < argc; ++index) { + if (skip) { + skip = false; + continue; + } + + if (double_skip) { + ++index; + double_skip = false; + continue; + } + + if (argv[index][0] == '-') { + if (strcmp(argv[index], "--cppdrv:ver") == 0) { + printf("%s\n", + "NeKernel Preprocessor Driver v1.11, (c) Amlal El Mahrouss 2024-2025 all rights " + "reserved."); + + return LIBCOMPILER_SUCCESS; + } + + if (strcmp(argv[index], "--cppdrv:?") == 0) { + printf("%s\n", + "NeKernel Preprocessor Driver v1.11, (c) Amlal El Mahrouss 2024-2025 all rights " + "reserved."); + printf("%s\n", "--cppdrv:working-dir : set directory to working path."); + printf("%s\n", "--cppdrv:include-dir : add directory to include path."); + printf("%s\n", "--cppdrv:def : define a macro."); + printf("%s\n", "--cppdrv:ver: print the version."); + printf("%s\n", "--cppdrv:?: show help (this current command)."); + + return LIBCOMPILER_SUCCESS; + } + + if (strcmp(argv[index], "--cppdrv:include-dir") == 0) { + LibCompiler::STLString inc = argv[index + 1]; + + skip = true; + + kIncludes.push_back(inc); + } + + if (strcmp(argv[index], "--cppdrv:working-dir") == 0) { + LibCompiler::STLString inc = argv[index + 1]; + skip = true; + kWorkingDir = inc; + } + + if (strcmp(argv[index], "--cppdrv:def") == 0 && argv[index + 1] != nullptr && + argv[index + 2] != nullptr) { + LibCompiler::STLString macro_key = argv[index + 1]; + + LibCompiler::STLString macro_value; + bool is_string = false; + + for (int argv_find_len = 0; argv_find_len < strlen(argv[index]); ++argv_find_len) { + if (!isdigit(argv[index][argv_find_len])) { + is_string = true; + macro_value += "\""; + + break; + } + } + + macro_value += argv[index + 2]; + + if (is_string) macro_value += "\""; + + Detail::bpp_macro macro; + macro.fName = macro_key; + macro.fValue = macro_value; + + kMacros.push_back(macro); + + double_skip = true; + } + + continue; + } + + kFiles.emplace_back(argv[index]); + } + + if (kFiles.empty()) return LIBCOMPILER_EXEC_ERROR; + + for (auto& file : kFiles) { + if (!std::filesystem::exists(file)) continue; + + std::ifstream file_descriptor(file); + std::ofstream file_descriptor_pp(file + ".pp"); + + bpp_parse_file(file_descriptor, file_descriptor_pp); + } + + return LIBCOMPILER_SUCCESS; + } catch (const std::runtime_error& e) { + std::cout << e.what() << '\n'; + } + + return LIBCOMPILER_EXEC_ERROR; +} + +// Last rev 8-1-24 diff --git a/dev/LibCompiler/src/PString.cc b/dev/LibCompiler/src/PString.cc new file mode 100644 index 0000000..c8c060b --- /dev/null +++ b/dev/LibCompiler/src/PString.cc @@ -0,0 +1,207 @@ +/* + * ======================================================== + * + * LibCompiler + * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. + * + * ======================================================== + */ + +/** + * @file PString.cxx + * @author Amlal (amlal@el-mahrouss-logic.com) + * @brief C++ string manipulation API. + * @version 0.2 + * @date 2024-01-23 + * + * @copyright Copyright (c) Amlal El Mahrouss + * + */ + +#include + +namespace LibCompiler { +CharType* PString::Data() { + return m_Data; +} + +const CharType* PString::CData() const { + return m_Data; +} + +SizeType PString::Length() const { + return strlen(m_Data); +} + +bool PString::operator==(const PString& rhs) const { + if (rhs.Length() != Length()) return false; + + for (SizeType index = 0; index < Length(); ++index) { + if (rhs.m_Data[index] != m_Data[index]) return false; + } + + return true; +} + +bool PString::operator==(const CharType* rhs) const { + if (string_length(rhs) != Length()) return false; + + for (SizeType index = 0; index < string_length(rhs); ++index) { + if (rhs[index] != m_Data[index]) return false; + } + + return true; +} + +bool PString::operator!=(const PString& rhs) const { + if (rhs.Length() != Length()) return false; + + for (SizeType index = 0; index < rhs.Length(); ++index) { + if (rhs.m_Data[index] == m_Data[index]) return false; + } + + return true; +} + +bool PString::operator!=(const CharType* rhs) const { + if (string_length(rhs) != Length()) return false; + + for (SizeType index = 0; index < string_length(rhs); ++index) { + if (rhs[index] == m_Data[index]) return false; + } + + return true; +} + +PString StringBuilder::Construct(const CharType* data) { + if (!data || *data == 0) return PString(0); + + PString view(strlen(data)); + view += data; + + return view; +} + +const char* StringBuilder::FromInt(const char* fmt, int i) { + if (!fmt) return ("-1"); + + auto ret_len = 8 + string_length(fmt); + char* ret = new char[ret_len]; + + if (!ret) return ("-1"); + + memset(ret, 0, ret_len); + + CharType result[sizeof(int64_t)]; + + if (!to_str(result, sizeof(int64_t), i)) { + delete[] ret; + return ("-1"); + } + + const auto fmt_len = string_length(fmt); + const auto res_len = string_length(result); + + for (SizeType idx = 0; idx < fmt_len; ++idx) { + if (fmt[idx] == '%') { + SizeType result_cnt = idx; + + for (auto y_idx = 0; y_idx < res_len; ++y_idx) { + ret[y_idx] = result[result_cnt]; + ++result_cnt; + } + + break; + } + + ret[idx] = fmt[idx]; + } + + return ret; /* Copy that ret into a buffer, Alloca allocates to the stack */ +} + +const char* StringBuilder::FromBool(const char* fmt, bool i) { + if (!fmt) return ("?"); + + const char* boolean_expr = i ? "true" : "false"; + char* ret = new char[i ? 4 : 5 + string_length(fmt)]; + + if (!ret) return ("?"); + + const auto fmt_len = string_length(fmt); + const auto res_len = string_length(boolean_expr); + + for (SizeType idx = 0; idx < fmt_len; ++idx) { + if (fmt[idx] == '%') { + SizeType result_cnt = idx; + + for (auto y_idx = idx; y_idx < res_len; ++y_idx) { + ret[result_cnt] = boolean_expr[y_idx]; + ++result_cnt; + } + + break; + } + + ret[idx] = fmt[idx]; + } + + return ret; +} + +bool StringBuilder::Equals(const char* lhs, const char* rhs) { + if (string_length(rhs) != string_length(lhs)) return false; + + for (SizeType index = 0; index < string_length(rhs); ++index) { + if (rhs[index] != lhs[index]) return false; + } + + return true; +} + +const char* StringBuilder::Format(const char* fmt, const char* fmtRight) { + if (!fmt || !fmtRight) return ("?"); + + char* ret = new char[string_length(fmtRight) + string_length(fmtRight)]; + if (!ret) return ("?"); + + for (SizeType idx = 0; idx < string_length(fmt); ++idx) { + if (fmt[idx] == '%') { + SizeType result_cnt = idx; + + for (SizeType y_idx = 0; y_idx < string_length(fmtRight); ++y_idx) { + ret[result_cnt] = fmtRight[y_idx]; + ++result_cnt; + } + + break; + } + + ret[idx] = fmt[idx]; + } + + return ret; +} + +PString& PString::operator+=(const CharType* rhs) { + if (strlen(rhs) > this->m_Sz) { + throw std::runtime_error("out_of_bounds: PString"); + } + + memcpy(this->m_Data + this->m_Cur, rhs, strlen(rhs)); + this->m_Cur += strlen(rhs); + + return *this; +} + +PString& PString::operator+=(const PString& rhs) { + if (rhs.m_Cur > this->m_Sz) { + throw std::runtime_error("out_of_bounds: PString"); + } + + memcpy(this->m_Data + this->m_Cur, rhs.CData(), strlen(rhs.CData())); + this->m_Cur += strlen(rhs.CData()); + + return *this; +} +} // namespace LibCompiler diff --git a/dev/LibCompiler/src/StringView.cc b/dev/LibCompiler/src/StringView.cc deleted file mode 100644 index 41ee636..0000000 --- a/dev/LibCompiler/src/StringView.cc +++ /dev/null @@ -1,207 +0,0 @@ -/* - * ======================================================== - * - * LibCompiler - * Copyright (C) 2024-2025 Amlal El Mahrouss, all rights reserved. - * - * ======================================================== - */ - -/** - * @file String.cxx - * @author Amlal (amlal@el-mahrouss-logic.com) - * @brief C++ string manipulation API. - * @version 0.2 - * @date 2024-01-23 - * - * @copyright Copyright (c) Amlal El Mahrouss - * - */ - -#include - -namespace LibCompiler { -CharType* StringView::Data() { - return m_Data; -} - -const CharType* StringView::CData() const { - return m_Data; -} - -SizeType StringView::Length() const { - return strlen(m_Data); -} - -bool StringView::operator==(const StringView& rhs) const { - if (rhs.Length() != Length()) return false; - - for (SizeType index = 0; index < Length(); ++index) { - if (rhs.m_Data[index] != m_Data[index]) return false; - } - - return true; -} - -bool StringView::operator==(const CharType* rhs) const { - if (string_length(rhs) != Length()) return false; - - for (SizeType index = 0; index < string_length(rhs); ++index) { - if (rhs[index] != m_Data[index]) return false; - } - - return true; -} - -bool StringView::operator!=(const StringView& rhs) const { - if (rhs.Length() != Length()) return false; - - for (SizeType index = 0; index < rhs.Length(); ++index) { - if (rhs.m_Data[index] == m_Data[index]) return false; - } - - return true; -} - -bool StringView::operator!=(const CharType* rhs) const { - if (string_length(rhs) != Length()) return false; - - for (SizeType index = 0; index < string_length(rhs); ++index) { - if (rhs[index] == m_Data[index]) return false; - } - - return true; -} - -StringView StringBuilder::Construct(const CharType* data) { - if (!data || *data == 0) return StringView(0); - - StringView view(strlen(data)); - view += data; - - return view; -} - -const char* StringBuilder::FromInt(const char* fmt, int i) { - if (!fmt) return ("-1"); - - auto ret_len = 8 + string_length(fmt); - char* ret = new char[ret_len]; - - if (!ret) return ("-1"); - - memset(ret, 0, ret_len); - - CharType result[sizeof(int64_t)]; - - if (!to_str(result, sizeof(int64_t), i)) { - delete[] ret; - return ("-1"); - } - - const auto fmt_len = string_length(fmt); - const auto res_len = string_length(result); - - for (SizeType idx = 0; idx < fmt_len; ++idx) { - if (fmt[idx] == '%') { - SizeType result_cnt = idx; - - for (auto y_idx = 0; y_idx < res_len; ++y_idx) { - ret[y_idx] = result[result_cnt]; - ++result_cnt; - } - - break; - } - - ret[idx] = fmt[idx]; - } - - return ret; /* Copy that ret into a buffer, Alloca allocates to the stack */ -} - -const char* StringBuilder::FromBool(const char* fmt, bool i) { - if (!fmt) return ("?"); - - const char* boolean_expr = i ? "true" : "false"; - char* ret = new char[i ? 4 : 5 + string_length(fmt)]; - - if (!ret) return ("?"); - - const auto fmt_len = string_length(fmt); - const auto res_len = string_length(boolean_expr); - - for (SizeType idx = 0; idx < fmt_len; ++idx) { - if (fmt[idx] == '%') { - SizeType result_cnt = idx; - - for (auto y_idx = idx; y_idx < res_len; ++y_idx) { - ret[result_cnt] = boolean_expr[y_idx]; - ++result_cnt; - } - - break; - } - - ret[idx] = fmt[idx]; - } - - return ret; -} - -bool StringBuilder::Equals(const char* lhs, const char* rhs) { - if (string_length(rhs) != string_length(lhs)) return false; - - for (SizeType index = 0; index < string_length(rhs); ++index) { - if (rhs[index] != lhs[index]) return false; - } - - return true; -} - -const char* StringBuilder::Format(const char* fmt, const char* fmtRight) { - if (!fmt || !fmtRight) return ("?"); - - char* ret = new char[string_length(fmtRight) + string_length(fmtRight)]; - if (!ret) return ("?"); - - for (SizeType idx = 0; idx < string_length(fmt); ++idx) { - if (fmt[idx] == '%') { - SizeType result_cnt = idx; - - for (SizeType y_idx = 0; y_idx < string_length(fmtRight); ++y_idx) { - ret[result_cnt] = fmtRight[y_idx]; - ++result_cnt; - } - - break; - } - - ret[idx] = fmt[idx]; - } - - return ret; -} - -StringView& StringView::operator+=(const CharType* rhs) { - if (strlen(rhs) > this->m_Sz) { - throw std::runtime_error("out_of_bounds: StringView"); - } - - memcpy(this->m_Data + this->m_Cur, rhs, strlen(rhs)); - this->m_Cur += strlen(rhs); - - return *this; -} - -StringView& StringView::operator+=(const StringView& rhs) { - if (rhs.m_Cur > this->m_Sz) { - throw std::runtime_error("out_of_bounds: StringView"); - } - - memcpy(this->m_Data + this->m_Cur, rhs.CData(), strlen(rhs.CData())); - this->m_Cur += strlen(rhs.CData()); - - return *this; -} -} // namespace LibCompiler diff --git a/dev/LibDebugger/Version.h b/dev/LibDebugger/Version.h new file mode 100644 index 0000000..4159191 --- /dev/null +++ b/dev/LibDebugger/Version.h @@ -0,0 +1,15 @@ +/* ------------------------------------------- + + Copyright (C) 2025 Amlal EL Mahrouss, all rights reserved + +------------------------------------------- */ + +#pragma once + +#define kDistVersion "v0.0.1-libdebugger" +#define kDistVersionBCD 0x0001 + +#define ToString(X) Stringify(X) +#define Stringify(X) #X + +#define kDistRelease ToString(kDistReleaseBranch) diff --git a/docs/drawio/LIBCOMPILER_DESIGN.drawio b/docs/drawio/LIBCOMPILER_DESIGN.drawio index 3c81ecc..89cb7d1 100644 --- a/docs/drawio/LIBCOMPILER_DESIGN.drawio +++ b/docs/drawio/LIBCOMPILER_DESIGN.drawio @@ -1,61 +1,61 @@ - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - + diff --git a/lc-osx-san.json b/lc-osx-san.json index 9b2d1c6..4e8f5c8 100644 --- a/lc-osx-san.json +++ b/lc-osx-san.json @@ -8,7 +8,8 @@ "./dev/LibCompiler/src/Detail" ], "sources_path": [ - "dev/LibCompiler/src/*.cc" + "dev/LibCompiler/src/*.cc", + "dev/LibCompiler/src/*/*.cc" ], "output_name": "/usr/local/lib/libCompiler.dylib", "compiler_flags": [ diff --git a/lc-osx.json b/lc-osx.json index 38a4c10..b331cbc 100644 --- a/lc-osx.json +++ b/lc-osx.json @@ -8,7 +8,8 @@ "./dev/LibCompiler/src/Detail" ], "sources_path": [ - "dev/LibCompiler/src/*.cc" + "dev/LibCompiler/src/*.cc", + "dev/LibCompiler/src/*/*.cc" ], "output_name": "/usr/local/lib/libCompiler.dylib", "compiler_flags": [ diff --git a/lc-posix.json b/lc-posix.json index 297d35b..6717707 100644 --- a/lc-posix.json +++ b/lc-posix.json @@ -8,7 +8,8 @@ "./dev/LibCompiler/src/Detail" ], "sources_path": [ - "dev/LibCompiler/src/*.cc" + "dev/LibCompiler/src/*.cc", + "dev/LibCompiler/src/*/*.cc" ], "output_name": "/usr/lib/libCompiler.so", "compiler_flags": [ diff --git a/tools/asm.cc b/tools/asm.cc index 8f98543..6f83c81 100644 --- a/tools/asm.cc +++ b/tools/asm.cc @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/tools/cppdrv.cc b/tools/cppdrv.cc index 7adfd0b..e479222 100644 --- a/tools/cppdrv.cc +++ b/tools/cppdrv.cc @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/tools/cxxdrv.cc b/tools/cxxdrv.cc index d500101..1f58692 100644 --- a/tools/cxxdrv.cc +++ b/tools/cxxdrv.cc @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/tools/dbg.cc b/tools/dbg.cc index b9be862..1d89c21 100644 --- a/tools/dbg.cc +++ b/tools/dbg.cc @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/tools/kdbg.cc b/tools/kdbg.cc index 11901e3..08a0465 100644 --- a/tools/kdbg.cc +++ b/tools/kdbg.cc @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ diff --git a/tools/ld64.cc b/tools/ld64.cc index ebb7376..ebdcb47 100644 --- a/tools/ld64.cc +++ b/tools/ld64.cc @@ -1,6 +1,6 @@ /* ------------------------------------------- - Copyright (C) 2024-2025 Amlal EL Mahrous, all rights reserved + Copyright (C) 2024-2025 Amlal EL Mahrouss, all rights reserved ------------------------------------------- */ -- cgit v1.2.3