Bin2json features

.gitignore

@@ -17,8 +17,4 @@ darwin/*
 *.docdir
 
 setup.data
-setup.log
-
-META
-*.mldylib
-*.mllib
+setup.log

README.md

@@ -1,7 +1,12 @@
-# Welcome
+# Rdr 2.0 - Welcome
 
 [![Floobits Status](https://floobits.com/m4b/rdr.svg)](https://floobits.com/m4b/rdr/redirect)
 
+**UPDATE**:
+
+> `rdr` is now version 2.0, which adds a new "byte coverage" algorithm for returning understood sections/segments with respect to a binary.  Check it out with `rdr -c <path to binary>`.
+> Moreover, `rdr` is now a binary _and_ a library, which you all can link against and use in your own projects, if you so desire.  [elf2json](http://github.com/m4b/elf2json) is a new program demonstrating this.
+
 Welcome to the `rdr` project.
 
 `rdr` is an OCaml tool/library for doing cross-platform analysis of binaries.  I typically use it for looking up symbol names, finding the address offset, and then running `gdb` or `lldb` to mess around (you should be using both if you even know what you're doing).

_oasis

@@ -1,16 +1,42 @@
 OASISFormat: 0.4
 Plugins: META (0.4), DevFiles (0.4)
 Name:        rdr
-Version:     1.2
+Version:     2.0.1
 Synopsis:    Lightweight, cross platform binary parsing and analysis library with no dependencies
 Authors:     m4b
+Homepage:    http://github.com/m4b/rdr
+Maintainers: <[email protected]>
 License:     BSD-3-clause
+Description:`rdr` is an OCaml tool/library for doing cross-platform analysis of binaries,
+             by printing headers, locating entry points, showing import and export
+             symbols, their binary offsets and size, etc.
+
+             It also features a symbol map which allows fast lookups for arbitrary
+             symbols, and their associated data, on your system
+             (the default search location are binaries in /usr/lib).
+
+             The latest release also makes `rdr` a package which you can link against
+             and use in your own projects.
+
+             See the README at http://github.com/m4b/rdr for more details.
+
+             Features:
+
+             * 64-bit Linux and Mach-o binary analysis
+             * Searchable symbol-map of all the symbols on your system, including binary
+               offset, size, and exporting library
+             * Print imports and exports of binaries
+             * Make pretty graphs, at the binary or symbol map level
+             * Byte Coverage algorithm which marks byte sequences as understood (or not)
+               and provides other meta-data
+
 
 Library "utils"
   Path: lib/utils
   BuildTools: ocamlbuild
   FindLibParent: rdr  
   FindLibName: utils
+  CompiledObject: best
   Modules:
     RdrUtils,
     Binary,
@@ -19,47 +45,38 @@ Library "utils"
     ByteCoverage,
     Generics
 
-Library "goblin"
-  Path: lib/goblin
-  BuildTools: ocamlbuild
-  FindLibParent: rdr
-  FindLibName: goblin
-  Modules:
-    Goblin,
-    GoblinSymbol,
-    GoblinExport,
-    GoblinImport
-  BuildDepends:
-    rdr.utils
-
 Library "mach"
   Path: lib/mach
   BuildTools: ocamlbuild
   FindLibParent: rdr
   FindLibName: mach
+  CompiledObject: best
   Modules:
     Mach,
     MachBindOpcodes,
     MachCpuTypes,
     MachFat,
     MachLoadCommand,
+    MachLoadCommandTypes,
     MachConstants,
     MachExports,
     MachHeader,
     MachImports,
-    MachSegment64,
-    MachNlist,
+    MachSection,
+    MachSymbolTable,
     MachRebaseOpcodes,
-    MachVersion
+    MachVersion,
+    MachCoverage,
+    MachLoadCommandMacro
   BuildDepends:
-    rdr.utils,
-    rdr.goblin
+    rdr.utils
 
 Library "elf"
   Path: lib/elf
   FindLibParent: rdr  
   FindLibName: elf
   BuildTools: ocamlbuild
+  CompiledObject: best
   Modules:
    Elf,
    ElfHeader,
@@ -73,10 +90,29 @@ Library "elf"
   BuildDepends:
    rdr.utils
 
+Library "goblin"
+  Path: lib/goblin
+  BuildTools: ocamlbuild
+  FindLibParent: rdr
+  FindLibName: goblin
+  CompiledObject: best
+  Modules:
+    Goblin,
+    GoblinSymbol,
+    GoblinExport,
+    GoblinImport,
+    GoblinMach,
+    GoblinElf
+  BuildDepends:
+    rdr.utils,
+    rdr.mach,
+    rdr.elf
+
 Library "rdr"
   Path: lib
   FindLibName: rdr
   BuildTools: ocamlbuild
+  CompiledObject: best
   Modules:
     LibRdr
   BuildDepends:
@@ -89,6 +125,7 @@ Library "rdrutils"
   Path: src
   BuildTools: ocamlbuild
   Install: false
+  CompiledObject: best  
   Modules:
     Config,
     Command,

_tags

@@ -1,5 +1,5 @@
 # OASIS_START
-# DO NOT EDIT (digest: 561c0ff1381a99f143e3c3b7c8381ca7)
+# DO NOT EDIT (digest: 28b9a39932677973ac2e4a14f90407c6)
 # Ignore VCS directories, you can use the same kind of rule outside
 # OASIS_START/STOP if you want to exclude directories that contains
 # useless stuff for the build process
@@ -16,16 +16,17 @@ true: annot, bin_annot
 "_darcs": not_hygienic
 # Library utils
 "lib/utils/utils.cmxs": use_utils
-# Library goblin
-"lib/goblin/goblin.cmxs": use_goblin
-<lib/goblin/*.ml{,i,y}>: use_utils
 # Library mach
 "lib/mach/mach.cmxs": use_mach
-<lib/mach/*.ml{,i,y}>: use_goblin
 <lib/mach/*.ml{,i,y}>: use_utils
 # Library elf
 "lib/elf/elf.cmxs": use_elf
 <lib/elf/*.ml{,i,y}>: use_utils
+# Library goblin
+"lib/goblin/goblin.cmxs": use_goblin
+<lib/goblin/*.ml{,i,y}>: use_elf
+<lib/goblin/*.ml{,i,y}>: use_mach
+<lib/goblin/*.ml{,i,y}>: use_utils
 # Library rdr
 "lib/rdr.cmxs": use_rdr
 <lib/*.ml{,i,y}>: use_elf

lib/LibRdr.ml

@@ -2,3 +2,97 @@ module Elf = Elf
 module Mach = Mach
 module Goblin = Goblin
 module Utils = RdrUtils
+
+module Object = struct
+  type t = | Mach of bytes | Elf of bytes | PE of bytes | Unknown of string * string
+  let get ?verbose:(verbose=false) filename =
+    let ic = open_in_bin filename in
+    if (in_channel_length ic < 4) then
+      (* 4 bytes, less than any magic number we're looking for *)
+      begin
+        close_in ic;
+        Unknown (filename, "less than 4 bytes")
+      end
+    else
+      (* BEGIN Binary cases *)
+      let magic = Input.input_i32be ic in
+      if (verbose) then
+        Printf.printf "opening %s with magic: 0x%x\n" filename magic;
+      (* MACH FAT *)
+      if (magic = Mach.Fat.kFAT_MAGIC) (* cafe babe *) then
+        let nfat_arch = Input.input_i32be ic in
+        if (nfat_arch > 4) then (* hack to avoid java class file errors which have same magic num *)
+          begin
+            close_in ic;
+            Unknown (filename, "mach fat too many archs (probably a java class file)")
+          end
+        else
+          let sizeof_arch_bytes = nfat_arch * Mach.Fat.sizeof_fat_arch in
+          let fat_arch_bytes = Bytes.create sizeof_arch_bytes in
+          really_input ic fat_arch_bytes 0 sizeof_arch_bytes;
+          let offset =
+	    Mach.Fat.get_x86_64_binary_offset fat_arch_bytes nfat_arch in
+          match offset with
+          | Some (offset, size) ->
+	    seek_in ic offset;
+	    let magic = Input.input_i32be ic in
+	    if (magic = Mach.Header.kMH_CIGAM_64) then
+              begin
+                seek_in ic offset;
+                let binary = Bytes.create size in
+                really_input ic binary 0 size;
+                close_in ic;
+                Mach binary
+              end
+	    else
+              begin
+                close_in ic; Unknown (filename, "mach fat has no 64 bit binaries")
+              end
+          | None ->
+	    close_in ic;
+	    Printf.eprintf "<Rdr.Object> ERROR, bad binary: %s\n" filename;
+	    Unknown (filename, "mach fat has no binaries")
+	    (* backwards cause we read the 32bit int big E style *)
+            (* MACH *)
+      else if (magic = Mach.Header.kMH_CIGAM_64) then
+        begin
+	  seek_in ic 0;
+	  let binary = Bytes.create (in_channel_length ic) in
+	  really_input ic binary 0 (in_channel_length ic);
+	  close_in ic;
+	  Mach binary
+        end
+        (* ELF *)
+      else if (magic = Elf.Header.kMAGIC_ELF) then
+        begin
+	  seek_in ic 0;
+	  let binary = Bytes.create (in_channel_length ic) in
+	  really_input ic binary 0 (in_channel_length ic);
+	  close_in ic;
+	  if (Elf.Header.check_64bit binary) then
+	    Elf binary
+	  else
+	    Unknown (filename, "elf binary is not 64-bit")
+        end
+        (* PE *)
+        (*
+      else if (magic = Elf.Header.kMAGIC_ELF) then
+        begin
+seek_in ic 0;
+let binary = Bytes.create (in_channel_length ic) in
+really_input ic binary 0 (in_channel_length ic);
+close_in ic;
+if (Elf.Header.check_64bit binary) then
+Elf binary
+else
+Unknown filename
+        end
+       *)
+      else
+        begin
+	  close_in ic;
+	  if (verbose) then
+	    Printf.printf "ignoring binary: %s\n" filename;
+	  Unknown (filename, "unknown magic number")
+        end
+end

lib/META

@@ -0,0 +1,59 @@
+# OASIS_START
+# DO NOT EDIT (digest: 19245a18a78bafcbdce567c9476d0a71)
+version = "2.0.1"
+description =
+"Lightweight, cross platform binary parsing and analysis library with no dependencies"
+requires = "rdr.goblin rdr.utils rdr.mach rdr.elf"
+archive(byte) = "rdr.cma"
+archive(byte, plugin) = "rdr.cma"
+archive(native) = "rdr.cmxa"
+archive(native, plugin) = "rdr.cmxs"
+exists_if = "rdr.cma"
+package "utils" (
+ version = "2.0.1"
+ description =
+ "Lightweight, cross platform binary parsing and analysis library with no dependencies"
+ archive(byte) = "utils.cma"
+ archive(byte, plugin) = "utils.cma"
+ archive(native) = "utils.cmxa"
+ archive(native, plugin) = "utils.cmxs"
+ exists_if = "utils.cma"
+)
+
+package "mach" (
+ version = "2.0.1"
+ description =
+ "Lightweight, cross platform binary parsing and analysis library with no dependencies"
+ requires = "rdr.utils"
+ archive(byte) = "mach.cma"
+ archive(byte, plugin) = "mach.cma"
+ archive(native) = "mach.cmxa"
+ archive(native, plugin) = "mach.cmxs"
+ exists_if = "mach.cma"
+)
+
+package "goblin" (
+ version = "2.0.1"
+ description =
+ "Lightweight, cross platform binary parsing and analysis library with no dependencies"
+ requires = "rdr.utils rdr.mach rdr.elf"
+ archive(byte) = "goblin.cma"
+ archive(byte, plugin) = "goblin.cma"
+ archive(native) = "goblin.cmxa"
+ archive(native, plugin) = "goblin.cmxs"
+ exists_if = "goblin.cma"
+)
+
+package "elf" (
+ version = "2.0.1"
+ description =
+ "Lightweight, cross platform binary parsing and analysis library with no dependencies"
+ requires = "rdr.utils"
+ archive(byte) = "elf.cma"
+ archive(byte, plugin) = "elf.cma"
+ archive(native) = "elf.cmxa"
+ archive(native, plugin) = "elf.cmxs"
+ exists_if = "elf.cma"
+)
+# OASIS_STOP
+