Merge pull request from s3team/existing-symbols

Use existing symbols if binary is not stripped

Author: yellowbyte

README.md

@@ -1,12 +1,12 @@
-# Uroboros: Infrastructure for Reassembleable Disassembling and Transformation (v1.0)
+# Uroboros: Infrastructure for Reassembleable Disassembling and Transformation
 
-# Installation
+## Installation
 
-## Docker
+### Docker
 
 Uroboros is available as a docker image, you can check the details in [docker](docker).
 
-## Build
+### Build
 
 The OCaml compiler and dependent libraries can be obtained through
 [opam](https://opam.ocaml.org/):
@@ -18,7 +18,7 @@ $ opam install -y ppx_deriving.6.0.3 parmap.1.2.5 batteries.3.9.0
 $ eval $(opam env)
 ```
 
-# Usage
+## Usage
 
 Uroboros can take a 32-bit or 64-bit, statically-linked or dynamically-linked ELF executable as the
 input.  To use Uroboros:
@@ -52,7 +52,7 @@ $ python3 uroboros.py bzip -i 500 -k
 A subfolder will be created in `./src` folder, with input binary name and
 timestamp (e.g., `test_fold_bzip_2021-09-19_05:51:00`).
 
-# Instrumentation
+## Instrumentation
 
 Instrumentation tools process the internal data structure of
 Uroboros. Some examples are shown in the [plugins](src/plugins) folder. You
@@ -64,7 +64,7 @@ To register instrumentation code:
 $ cp plugins/mem_write.ml instrumentation_plugin.ml
 ```
 
-# Publication
+## Publication
 ```
 @inproceedings {190920,
 author = {Shuai Wang and Pei Wang and Dinghao Wu},

docker/README.md

@@ -6,17 +6,17 @@ Use `docker build` to build the docker image from Dockerfile.
 
 ```shell
 # build using Ubuntu 20.04 (default)
-docker build -f ./Docker/Dockerfile -t uroboros:1.0 .
+docker build -f ./Docker/Dockerfile -t uroboros .
 # build using Ubuntu 22.04
-docker build --build-arg UBUNTU_VERSION=22.04 -f ./Docker/Dockerfile -t uroboros:1.0 .
+docker build --build-arg UBUNTU_VERSION=22.04 -f ./Docker/Dockerfile -t uroboros .
 # build using Ubuntu 24.04
-docker build --build-arg UBUNTU_VERSION=24.04 -f ./Docker/Dockerfile -t uroboros:1.0 .
+docker build --build-arg UBUNTU_VERSION=24.04 -f ./Docker/Dockerfile -t uroboros .
 ```
 
 ## Docker usage
 
 run the Docker container with `docker run`.
 
 ```shell
-docker run -it -v <path to uroboros parent>/uroboros:/usr/src uroboros:1.0 bash
+docker run -it -v <path to uroboros parent>/uroboros:/usr/src uroboros bash
 ```

src/ail.ml

@@ -94,7 +94,7 @@ object (self)
     Enum.iter help filelines
 
   method ail_dump =
-    (*currently we just dump the extern function info *)
+    (* currently we just dump the extern function info *)
     let check_sym_func f =
       try
         let s = Char.escaped(f.func_name.[0])^Char.escaped(f.func_name.[1]) in
@@ -116,35 +116,33 @@ object (self)
   method post_process f (arch : string) = 
     ignore (Sys.command ("python3 main_discover.py " ^ " " ^ f ^ " " ^ arch));
     ignore(Sys.command("python3 post_process.py "^arch));
-    ignore(Sys.command("python3 post_process_lib.py"));
-    ()
-  (*
-          self#ehframe_dump;
-          self#excpt_tbl_dump;
-   *)
+    ignore(Sys.command("python3 post_process_lib.py"))
+    (*
+    self#ehframe_dump;
+    self#excpt_tbl_dump;
+    *)
 
   method pre_process =
-    let  _ = Sys.command("python3 pre_process.py") in
-    ()
-
-
+    let  _ = Sys.command("python3 pre_process.py") in ()
 
   method instrProcess_2 f (arch: string) =
     let open Disassemble_process in
     let open Analysis_process in
     let module D = Disam in
     let module A = Analysis in
+    let module S = Symbol_get in
     let () = self#pre_process in
     let (il, fl, re) = D.disassemble f funcs secs arch in
 
-	print_endline "3: analysis";
+    print_endline "3: analysis";
 
     let (fbl, bbl, cfg_t, cg, il', re) = A.analyze_one il fl re in
+    let il' = S.apply il' in
     let open Instrumentation_plugin in
     let module IP = Instrumentation_Plugin in
     let instrumented_il = IP.instrument il' fbl bbl in
 
-	print_endline "4: post processing";
+    print_endline "4: post processing";
     A.post_analyze instrumented_il re;
 
     self#post_process f arch

src/disassemble_process.ml

@@ -53,7 +53,7 @@ module Disam = struct
       ail_parser#process_instrs instr_list arch;
 
       let fl = ail_parser#get_funcs in
-      print_endline "2: disassembly validates --> ";
+      print_endline "2: disassembly validates";
 
       (*let _ = List.iter (
         fun f ->
@@ -73,7 +73,6 @@ module Disam = struct
             ()
         ) func2cfg_table in ()
       else ();
-      
 
       il :=
         if EU.elf_32 () && arch <> "arm" then
@@ -101,4 +100,4 @@ module Disam = struct
     print_endline "\tno disassembly error detects";
     let funcs = ail_parser#get_funcs in
     (!il, funcs, re)
-end
+end

src/post_process.py

@@ -92,7 +92,7 @@ def check_32():
     # adjust it into jmp S_0x4006C1
     elif "jmpq " in l and "*" not in l:
         l = l.replace('jmpq ', 'jmp ')
-    elif "__libc_start_main" in l and is_32 == True:
+    elif "call " in l and "__libc_start_main" in l and is_32 == True:
         if arch == "arm":
             """
             In case of ARM, main symbol is not pushed into stack.
@@ -117,7 +117,7 @@ def check_32():
             main_symbol = lines[i-1].split()[1]
             lines[i-1] = lines[i-1].replace(main_symbol, "main")
             main_symbol = main_symbol[1:].strip()
-    elif is_32 == False and "__libc_start_main" in l:
+    elif is_32 == False and "call " in l and "__libc_start_main" in l:
         if arch == "arm":
             pass
         else: