vmprofiler-cli/src/main.cpp

#include <Windows.h>
#include <filesystem>
#include <fstream>
#include <iostream>
#include <vtil/vtil>

#include <ZydisExportConfig.h>
#include <cli-parser.hpp>
#include <vmp2.hpp>
#include <vmprofiler.hpp>
#include <xtils.hpp>

#define ABS_TO_IMG( addr, mod_base, img_base ) ( addr - mod_base ) + img_base

int __cdecl main( int argc, const char *argv[] )
{
    argparse::argument_parser_t parser( "vmprofiler-cli", "virtual instruction pseudo code generator" );

    parser.add_argument().names( { "--bin", "--vmpbin" } ).description( "unpacked binary protected with VMProtect 2" );
    parser.add_argument().names( { "--vmentry", "--entry" } ).description( "rva to push prior to a vm_entry" );
    parser.add_argument().name( "--showhandlers" ).description( "show all vm handlers..." );
    parser.add_argument().name( "--showhandler" ).description( "show a specific vm handler given its index..." );
    parser.add_argument().name( "--vmp2file" ).description( "path to .vmp2 file..." );
    parser.add_argument()
        .name( "--showblockinstrs" )
        .description( "show the virtual instructions of a specific code block..." );
    parser.add_argument().name( "--showallblocks" ).description( "shows all information for all code blocks..." );
    parser.add_argument()
        .name( "--devirt" )
        .description( "lift to VTIL IR and apply optimizations, then display the output..." );

    parser.enable_help();
    auto err = parser.parse( argc, argv );
    const auto umtils = xtils::um_t::get_instance();

    if ( err )
    {
        std::cout << err << std::endl;
        return -1;
    }

    if ( parser.exists( "help" ) )
    {
        parser.print_help();
        return 0;
    }

    if ( parser.exists( "bin" ) && parser.exists( "vmentry" ) )
    {
        if ( !std::filesystem::exists( parser.get< std::string >( "bin" ) ) )
        {
            std::printf( "> path to protected file is invalid... check your cli args...\n" );
            return -1;
        }

        const auto module_base = reinterpret_cast< std::uintptr_t >(
            LoadLibraryExA( parser.get< std::string >( "bin" ).c_str(), NULL, DONT_RESOLVE_DLL_REFERENCES ) );

        const auto vm_entry_rva = std::strtoull( parser.get< std::string >( "vmentry" ).c_str(), nullptr, 16 );
        const auto image_base = umtils->image_base( parser.get< std::string >( "bin" ).c_str() );
        const auto image_size = NT_HEADER( module_base )->OptionalHeader.SizeOfImage;

        std::printf( "> module base = 0x%p, image base = 0x%p, image size = 0x%p\n", module_base, image_base,
                     image_size );

        vm::ctx_t vmctx( module_base, image_base, image_size, vm_entry_rva );

        if ( !vmctx.init() )
        {
            std::printf( "[!] failed to init vm::ctx_t... make sure all cli arguments are correct!\n" );
            return -1;
        }

        std::puts( "======================== [vm entry] ========================\n" );
        vm::util::print( vmctx.vm_entry );
        std::puts( "======================== [calc jmp] ========================\n" );
        vm::util::print( vmctx.calc_jmp );
        std::puts( "============================================================\n" );
        std::printf( "> vip advancement = %s\n\n",
                     vmctx.exec_type == vmp2::exec_type_t::forward ? "forward" : "backward" );

        if ( parser.exists( "showhandlers" ) )
        {
            for ( auto idx = 0u; idx < vmctx.vm_handlers.size(); ++idx )
            {
                std::printf( "======================== [%s #%d] ========================\n",
                             vmctx.vm_handlers[ idx ].profile ? vmctx.vm_handlers[ idx ].profile->name : "UNK", idx );

                vm::util::print( vmctx.vm_handlers[ idx ].instrs );

                // if there is no imm then there are no transforms...
                if ( !vmctx.vm_handlers[ idx ].imm_size )
                {
                    std::puts( "\n" );
                    continue;
                }

                std::puts( "======================== [transforms] ========================\n" );
                for ( auto &[ mnemonic, instr ] : vmctx.vm_handlers[ idx ].transforms )
                {
                    if ( instr.mnemonic == ZYDIS_MNEMONIC_INVALID )
                        continue;

                    vm::util::print( instr );
                }
                std::puts( "\n" );
            }
        }
        else if ( parser.exists( "showhandler" ) )
        {
            const auto vm_handler_idx = std::strtoul( parser.get< std::string >( "showhandler" ).c_str(), nullptr, 10 );

            if ( vm_handler_idx > 256 )
            {
                std::printf( "> invalid vm handler index... too large...\n" );
                return -1;
            }

            std::printf( "======================== [%s #%d] ========================\n",
                         vmctx.vm_handlers[ vm_handler_idx ].profile ? vmctx.vm_handlers[ vm_handler_idx ].profile->name
                                                                     : "UNK",
                         vm_handler_idx );

            vm::util::print( vmctx.vm_handlers[ vm_handler_idx ].instrs );

            // if there is no imm then there are no transforms...
            if ( !vmctx.vm_handlers[ vm_handler_idx ].imm_size )
            {
                std::puts( "\n" );
                return {};
            }

            std::puts( "======================== [transforms] ========================\n" );
            for ( auto &[ mnemonic, instr ] : vmctx.vm_handlers[ vm_handler_idx ].transforms )
            {
                if ( instr.mnemonic == ZYDIS_MNEMONIC_INVALID )
                    continue;

                vm::util::print( instr );
            }
            std::puts( "\n" );
        }
    }

    if ( !parser.exists( "vmp2file" ) )
        return {};

    std::vector< std::uint8_t > vmp2file;
    if ( !umtils->open_binary_file( parser.get< std::string >( "vmp2file" ), vmp2file ) )
    {
        std::printf( "[!] failed to open vmp2 file...\n" );
        return -1;
    }

    const auto file_header = reinterpret_cast< vmp2::v3::file_header * >( vmp2file.data() );

    if ( file_header->version != vmp2::version_t::v3 )
    {
        std::printf( "[!] invalid vmp2 file version... this build uses v3...\n" );
        return -1;
    }

    auto first_block = reinterpret_cast< vmp2::v3::code_block_t * >( reinterpret_cast< std::uintptr_t >( file_header ) +
                                                                     file_header->code_block_offset );

    if ( parser.exists( "showallblocks" ) )
    {
        for ( auto [ code_block, code_block_num ] = std::tuple{ first_block, 0u };
              code_block_num < file_header->code_block_count;
              code_block = reinterpret_cast< vmp2::v3::code_block_t * >(
                  reinterpret_cast< std::uintptr_t >( code_block ) + code_block->next_block_offset ),
                                                ++code_block_num )
        {

            std::printf( "[code block #%d] begin = 0x%p, virtual instruction count = %d\n", code_block_num,
                         ABS_TO_IMG( code_block->vip_begin, file_header->module_base, file_header->image_base ),
                         code_block->vinstr_count );

            if ( code_block->jcc.has_jcc )
                std::printf(
                    "\tcode block branches to 0x%p and 0x%p\n",
                    ABS_TO_IMG( code_block->jcc.block_addr[ 0 ], file_header->module_base, file_header->image_base ),
                    ABS_TO_IMG( code_block->jcc.block_addr[ 1 ], file_header->module_base, file_header->image_base ) );
        }
    }

    if ( parser.exists( "devirt" ) )
    {
        vtil::basic_block *rtn = nullptr, *first = nullptr;
        for ( auto [ code_block, code_block_num ] = std::tuple{ first_block, 0u };
              code_block_num < file_header->code_block_count;
              code_block = reinterpret_cast< vmp2::v3::code_block_t * >(
                  reinterpret_cast< std::uintptr_t >( code_block ) + code_block->next_block_offset ),
                                                ++code_block_num )
        {
            if ( !rtn )
            {
                rtn = vtil::basic_block::begin(
                    ABS_TO_IMG( code_block->vip_begin, file_header->module_base, file_header->image_base ) );
            }
            else
            {
                first = rtn;

                rtn =
                    rtn->fork( ABS_TO_IMG( code_block->vip_begin, file_header->module_base, file_header->image_base ) );
            }

            for ( auto idx = 0u; idx < code_block->vinstr_count; ++idx )
            {
                auto vinstr = &code_block->vinstr[ idx ];
                if ( vinstr->mnemonic_t == vm::handler::INVALID )
                {
                    std::printf( "[WARNING] vm handler #%d is not implimented, the VTIL output may be incorrect!\n",
                                 vinstr->opcode );
                    rtn->nop();
                    continue;
                }

                const auto result = std::find_if(
                    vm::lifters::all.begin(), vm::lifters::all.end(),
                    [ & ]( vm::lifters::lifter_t *lifter ) -> bool { return lifter->first == vinstr->mnemonic_t; } );

                if ( result == vm::lifters::all.end() )
                {
                    std::printf(
                        "[WARNING] vm handler #%d lifter is not implimented, the VTIL output may be incorrect!\n",
                        vinstr->opcode );

                    rtn->nop();
                    continue;
                }

                // lift the virtual instruction...
                ( *result )->second( rtn, vinstr, code_block );
            }
        }

        vtil::optimizer::apply_all( first );
        vtil::debug::dump( first );
        return {};
    }

    if ( parser.exists( "showblockinstrs" ) )
    {
        const auto block_img_addr = parser.get< std::string >( "showblockinstrs" );

        for ( auto [ code_block, code_block_num ] = std::tuple{ first_block, 0u };
              code_block_num < file_header->code_block_count;
              code_block = reinterpret_cast< vmp2::v3::code_block_t * >(
                  reinterpret_cast< std::uintptr_t >( code_block ) + code_block->next_block_offset ),
                                                ++code_block_num )
        {
            if ( ABS_TO_IMG( code_block->vip_begin, file_header->module_base, file_header->image_base ) ==
                 std::strtoull( block_img_addr.c_str(), nullptr, 16 ) )
            {
                std::printf( "[code block #%d] begin = 0x%p, virtual instruction count = %d\n", code_block_num,
                             ABS_TO_IMG( code_block->vip_begin, file_header->module_base, file_header->image_base ),
                             code_block->vinstr_count );

                std::printf( "> -----------------------------------------------------------------------\n" );
                std::printf( "> opcode | virtual instructions             | virtual instruction pointer\n" );
                std::printf( "> -----------------------------------------------------------------------\n" );
                for ( auto idx = 0u; idx < code_block->vinstr_count; ++idx )
                {
                    const auto vinstr = &code_block->vinstr[ idx ];
                    const auto vm_profile = vm::handler::get_profile( vinstr->mnemonic_t );
                    if ( vinstr->operand.has_imm )
                    {
                        std::printf(
                            "> %-6x | %-15s %-15p | 0x%p\n", vinstr->opcode, vm_profile ? vm_profile->name : "UNK",
                            vinstr->operand.imm.u,
                            ABS_TO_IMG( vinstr->trace_data.vip, file_header->module_base, file_header->image_base ) );
                    }
                    else
                    {
                        std::printf(
                            "> %-6x | %-32s | 0x%p\n", vinstr->opcode, vm_profile ? vm_profile->name : "UNK",
                            ABS_TO_IMG( vinstr->trace_data.vip, file_header->module_base, file_header->image_base ) );
                    }
                }
                break;
            }
        }
    }
}