Merge branch 'dev' into 'master'

added vmp2 file format version 4 logic...

See merge request vmp2/vmprofiler!16

CMakeLists.txt

@@ -41,6 +41,7 @@ set(vmprofiler_SOURCES "")
 
 list(APPEND vmprofiler_SOURCES
 	"src/calc_jmp.cpp"
+	"src/scn.cpp"
 	"src/vmctx.cpp"
 	"src/vmhandler.cpp"
 	"src/vminstrs.cpp"
@@ -65,6 +66,7 @@ list(APPEND vmprofiler_SOURCES
 	"src/vmprofiles/write.cpp"
 	"src/vmutils.cpp"
 	"include/calc_jmp.hpp"
+	"include/scn.hpp"
 	"include/transform.hpp"
 	"include/vmctx.hpp"
 	"include/vmhandlers.hpp"

include/scn.hpp

@@ -0,0 +1,27 @@
+#include <nt/image.hpp>
+
+/// <summary>
+/// small namespace that contains function wrappers to determine the validity of linear virtual addresses...
+/// </summary>
+namespace scn
+{
+    /// <summary>
+    /// determines if a pointer lands inside of a section that is readonly...
+    ///
+    /// this also checks to make sure the section is not discardable...
+    /// </summary>
+    /// <param name="module_base">linear virtual address of the module....</param>
+    /// <param name="ptr">linear virtual address</param>
+    /// <returns>returns true if ptr lands inside of a readonly section of the module</returns>
+    bool read_only( std::uint64_t module_base, std::uint64_t ptr );
+
+    /// <summary>
+    /// determines if a pointer lands inside of a section that is executable...
+    ///
+    /// this also checks to make sure the section is not discardable...
+    /// </summary>
+    /// <param name="module_base"></param>
+    /// <param name="ptr"></param>
+    /// <returns></returns>
+    bool executable( std::uint64_t module_base, std::uint64_t ptr );
+} // namespace scn

include/vminstrs.hpp

@@ -3,6 +3,7 @@
 #include <vmctx.hpp>
 #include <vmhandlers.hpp>
 #include <vmp2.hpp>
+#include <scn.hpp>
 
 /// <summary>
 /// contains all functions related to virtual instructions...
@@ -72,6 +73,15 @@ namespace vm::instrs
     /// <returns>if last lconstdw is found, return filled in jcc_data structure...</returns>
     std::optional< jcc_data > get_jcc_data( vm::ctx_t &ctx, code_block_t &code_block );
 
+    /// <summary>
+    /// this algo is used to try and find a jmp tables address and all of its entries for a given code block...
+    /// </summary>
+    /// <param name="ctx">vm context</param>
+    /// <param name="code_block">code block that has an absolute jmp... this routine is going to see if this code block
+    /// actually is jmp table stub</param> <returns>if a jmp table is found then we decrypt all entries and return a
+    /// vector of them..</returns>
+    std::optional< std::vector< std::uint64_t > > get_switch_cases( vm::ctx_t &ctx, code_block_t &code_block );
+
     /// <summary>
     /// the top of the stack will contain the lower 32bits of the RVA to the virtual instructions
     /// that will be jumping too... the RVA is image based (not module based, but optional header image

include/vmp2.hpp

@@ -16,7 +16,8 @@ namespace vmp2
         invalid,
         v1 = 0x101,
         v2 = 0x102,
-        v3 = 0x103
+        v3 = 0x103,
+        v4 = 0x104
     };
 
     namespace v1
@@ -173,14 +174,15 @@ namespace vm
         {
             none,
             branching,
-            absolute
+            absolute,
+            switch_case
         };
 
         struct jcc_data
         {
             bool has_jcc;
             jcc_type type;
-            std::uintptr_t block_addr[ 2 ];
+            std::vector< std::uintptr_t > block_addr;
         };
 
         struct code_block_t
@@ -225,3 +227,50 @@ namespace vmp2
         };
     } // namespace v3
 } // namespace vmp2
+
+#pragma pack( push, 1 )
+namespace vmp2
+{
+    namespace v4
+    {
+        struct file_header
+        {
+            u32 magic; // VMP2
+            u64 epoch_time;
+            version_t version;
+
+            u64 module_base;
+            u64 image_base;
+            u64 vm_entry_rva;
+
+            u32 module_offset;
+            u32 module_size;
+
+            u32 rtn_count;
+            u32 rtn_offset;
+        };
+
+        struct code_block_t
+        {
+            std::uintptr_t vip_begin;
+            std::uintptr_t next_block_offset;
+            std::uint32_t vinstr_count;
+
+            bool has_jcc;
+            vm::instrs::jcc_type jcc_type;
+            std::uint32_t num_block_addrs;
+
+            std::uintptr_t branch_addr[ 1 ];
+            vm::instrs::virt_instr_t vinstr[ 1 ];
+        };
+
+        struct rtn_t
+        {
+            u32 size;
+            u32 vm_enter_offset;
+            u32 code_block_count;
+            vmp2::v4::code_block_t code_blocks[ 1 ];
+        };
+    } // namespace v4
+#pragma pack( pop )
+} // namespace vmp2

src/scn.cpp

@@ -0,0 +1,34 @@
+#include <scn.hpp>
+
+namespace scn
+{
+    bool read_only( std::uint64_t module_base, std::uint64_t ptr )
+    {
+        auto win_image = reinterpret_cast< win::image_t<> * >( module_base );
+        auto section_count = win_image->get_file_header()->num_sections;
+        auto sections = win_image->get_nt_headers()->get_sections();
+
+        for ( auto idx = 0u; idx < section_count; ++idx )
+            if ( ptr >= sections[ idx ].virtual_address + module_base &&
+                 ptr < sections[ idx ].virtual_address + sections[ idx ].virtual_size + module_base )
+                return !( sections[ idx ].characteristics.mem_discardable ) &&
+                       !( sections[ idx ].characteristics.mem_write );
+
+        return false;
+    }
+
+    bool executable( std::uint64_t module_base, std::uint64_t ptr )
+    {
+        auto win_image = reinterpret_cast< win::image_t<> * >( module_base );
+        auto section_count = win_image->get_file_header()->num_sections;
+        auto sections = win_image->get_nt_headers()->get_sections();
+
+        for ( auto idx = 0u; idx < section_count; ++idx )
+            if ( ptr >= sections[ idx ].virtual_address + module_base &&
+                 ptr < sections[ idx ].virtual_address + sections[ idx ].virtual_size + module_base )
+                return !( sections[ idx ].characteristics.mem_discardable ) &&
+                       sections[ idx ].characteristics.mem_execute;
+
+        return false;
+    }
+} // namespace scn

src/vminstrs.cpp

@@ -207,6 +207,81 @@ namespace vm::instrs
         return result;
     }
 
+    std::optional< std::vector< std::uint64_t > > get_switch_cases( vm::ctx_t &ctx, code_block_t &code_block )
+    {
+        // find the last LCONSTDW in this code block... it is the XOR decryption key...
+        auto lconstdw_xor_key = std::find_if( code_block.vinstrs.rbegin(), code_block.vinstrs.rend(),
+                                              []( const vm::instrs::virt_instr_t &vinstr ) -> bool {
+                                                  auto profile = vm::handler::get_profile( vinstr.mnemonic_t );
+                                                  return profile && profile->mnemonic == vm::handler::LCONSTDW;
+                                              } );
+
+        if ( lconstdw_xor_key == code_block.vinstrs.rend() )
+            return {};
+
+        // extract the address in which we are jmp'ing too, this gets compared lower in the algo...
+        auto jmp_addr = code_block.vinstrs.back().trace_data.vsp.qword[ 0 ];
+
+        // find the SREGDW that sets a virtual register to the encrypted rva we are jmping too...
+        auto sregdw_jmp_addr = std::find_if( lconstdw_xor_key, code_block.vinstrs.rend(),
+                                             [ & ]( const vm::instrs::virt_instr_t &vinstr ) -> bool {
+                                                 if ( vinstr.mnemonic_t == vm::handler::SREGDW )
+                                                 {
+                                                     if ( ( ( ( std::uint32_t )vinstr.trace_data.vsp.qword[ 0 ] ) ^
+                                                            lconstdw_xor_key->operand.imm.u ) == jmp_addr )
+                                                     {
+                                                         return true;
+                                                     }
+                                                 }
+                                                 return false;
+                                             } );
+
+        if ( sregdw_jmp_addr == code_block.vinstrs.rend() )
+            return {};
+
+        // find the last READDW (the one above SREGDW...)
+        auto readdw_jmp_tbl = std::find_if( sregdw_jmp_addr, code_block.vinstrs.rend(),
+                                            [ & ]( const vm::instrs::virt_instr_t &vinstr ) -> bool {
+                                                return vinstr.mnemonic_t == vm::handler::READDW;
+                                            } );
+
+        if ( readdw_jmp_tbl == code_block.vinstrs.rend() )
+            return {};
+
+        // find the last ADDQ which when computed results in the READDW address found above...
+        auto addq_jmp_tbl_addr = std::find_if(
+            readdw_jmp_tbl, code_block.vinstrs.rend(), [ & ]( const vm::instrs::virt_instr_t &vinstr ) -> bool {
+                return vinstr.mnemonic_t == vm::handler::ADDQ &&
+                       vinstr.trace_data.vsp.qword[ 0 ] + vinstr.trace_data.vsp.qword[ 1 ] ==
+                           readdw_jmp_tbl->trace_data.vsp.qword[ 0 ];
+            } );
+
+        if ( addq_jmp_tbl_addr == code_block.vinstrs.rend() )
+            return {};
+
+        // sanity check...
+        if ( !scn::executable( ctx.module_base, ctx.module_base + addq_jmp_tbl_addr->trace_data.vsp.qword[ 1 ] ) )
+            return {};
+
+        auto jmp_table =
+            reinterpret_cast< std::uint32_t * >( ctx.module_base + addq_jmp_tbl_addr->trace_data.vsp.qword[ 1 ] );
+
+        std::vector< std::uint64_t > result;
+        for ( auto idx = 0u;; ++idx )
+        {
+            auto code_addr = code_block_addr( ctx, jmp_table[ idx ] ^ lconstdw_xor_key->operand.imm.u );
+
+            // keep decrypting entries until we decrypt a value that doesnt land inside of an executable section... if
+            // this allows too many cases we will need to check to see if the first virtual instruction of this block is
+            // an SREGQ...
+            if ( scn::executable( ctx.module_base, code_addr ) )
+                result.push_back( code_addr );
+            else // we finished decrypting the table...
+                break;
+        }
+        return result;
+    }
+
     std::optional< jcc_data > get_jcc_data( vm::ctx_t &vmctx, code_block_t &code_block )
     {
         // there is no branch for this as this is a vmexit...
@@ -224,15 +299,9 @@ namespace vm::instrs
         if ( result == code_block.vinstrs.rend() )
             return jcc_data{ false, jcc_type::none };
 
-        jcc_data jcc;
         const auto xor_key = static_cast< std::uint32_t >( result->operand.imm.u );
         const auto &last_trace = code_block.vinstrs.back().trace_data;
 
-        // since result is already a variable and is a reverse itr
-        // im going to be using rbegin and rend here again...
-        //
-        // look for PUSHVSP virtual instructions with two encrypted virtual
-        // instruction rva's ontop of the virtual stack...
         result = std::find_if(
             code_block.vinstrs.rbegin(), code_block.vinstrs.rend(),
             [ & ]( const vm::instrs::virt_instr_t &vinstr ) -> bool {
@@ -242,13 +311,8 @@ namespace vm::instrs
                     const auto possible_block_1 = code_block_addr( vmctx, vinstr.trace_data.vsp.qword[ 0 ] ^ xor_key ),
                                possible_block_2 = code_block_addr( vmctx, vinstr.trace_data.vsp.qword[ 1 ] ^ xor_key );
 
-                    // if this returns too many false positives we might have to get
-                    // our hands dirty and look into trying to emulate each branch
-                    // to see if the first instruction is an SREGQ...
-                    return possible_block_1 > vmctx.module_base &&
-                           possible_block_1 < vmctx.module_base + vmctx.image_size &&
-                           possible_block_2 > vmctx.module_base &&
-                           possible_block_2 < vmctx.module_base + vmctx.image_size;
+                    return scn::executable( vmctx.module_base, possible_block_1 ) &&
+                           scn::executable( vmctx.module_base, possible_block_2 );
                 }
                 return false;
             } );
@@ -256,20 +320,32 @@ namespace vm::instrs
         // if there is not two branches...
         if ( result == code_block.vinstrs.rend() )
         {
-            jcc.block_addr[ 0 ] = code_block_addr( vmctx, last_trace );
-            jcc.has_jcc = true;
-            jcc.type = jcc_type::absolute;
-        }
-        // else there are two branches...
-        else
-        {
-            jcc.block_addr[ 0 ] = code_block_addr( vmctx, result->trace_data.vsp.qword[ 0 ] ^ xor_key );
-            jcc.block_addr[ 1 ] = code_block_addr( vmctx, result->trace_data.vsp.qword[ 1 ] ^ xor_key );
-
-            jcc.has_jcc = true;
-            jcc.type = jcc_type::branching;
+            // see if this code block is actually a jmp table for a switch case....
+            auto result = get_switch_cases( vmctx, code_block );
+            if ( result.has_value() )
+            {
+                auto vec = result.value();
+                jcc_data jcc;
+                jcc.has_jcc = true;
+                jcc.type = jcc_type::switch_case;
+                jcc.block_addr = vec;
+                return jcc;
+            }
+            else
+            {
+                jcc_data jcc;
+                jcc.block_addr.push_back( code_block_addr( vmctx, last_trace ) );
+                jcc.has_jcc = true;
+                jcc.type = jcc_type::absolute;
+                return jcc;
+            }
         }
 
+        jcc_data jcc;
+        jcc.block_addr.push_back( code_block_addr( vmctx, result->trace_data.vsp.qword[ 0 ] ^ xor_key ) );
+        jcc.block_addr.push_back( code_block_addr( vmctx, result->trace_data.vsp.qword[ 1 ] ^ xor_key ) );
+        jcc.has_jcc = true;
+        jcc.type = jcc_type::branching;
         return jcc;
     }