diff --git a/x86model/Model/decenc_tests/cat b/x86model/Model/decenc_tests/cat
new file mode 100644
index 0000000..e56f3b4
Binary files /dev/null and b/x86model/Model/decenc_tests/cat differ
diff --git a/x86model/Model/decenc_tests/extract.v b/x86model/Model/decenc_tests/extract.v
new file mode 100644
index 0000000..295a4bb
--- /dev/null
+++ b/x86model/Model/decenc_tests/extract.v
@@ -0,0 +1,58 @@
+(* Extract an OCaml x86 model from the Coq x86 model *)
+
+Require ExtrOcamlString.
+
+(* Map nat to OCaml big_int *)
+Require ExtrOcamlNatBigInt.
+
+(* fix bug in Require ExtrOcamlNatBigInt: missing the outer parentheses *)
+Extract Constant minus => "(fun n m -> Big.max Big.zero (Big.sub n m))".
+
+(* Map Z and pos to OCaml big_int *)
+Require ExtrOcamlZBigInt.
+
+(* bypassing the extraction of opaque constants *)
+Set Extraction AccessOpaque.
+
+Extract Inductive sigT => "(*)"  [ "(,)" ].
+
+Require Import X86Model.X86BG.
+(* Require Import X86Model.Decode. *)
+Require Import X86Model.Parser.
+Require X86Model.X86Semantics.
+Extraction Blacklist String List. 
+Require Encode.
+
+Set Extraction Optimize.
+Set Extraction AutoInline.
+
+(* without this, some files will use an extracted Nat module, which conflicts with
+    Ocaml's default Nat module *)
+Extraction Blacklist Nat.
+
+(* Did not extract X86Semantics.step below since that would call
+   build_dfa on the instruction grammar, which would take
+   approximately 2 minutes.  One way of avoiding this is to change the
+   definition of X86Semantics.step to be parameterized by a dfa
+   builder, similar to how initial_parse_state' does it; that way, the
+   client can supply a dfa builder that just deserializes the dfa from
+   a file, isntead of building it from scratch. *)
+
+Separate Extraction initial_parser_state parse_tokens
+  vinitial_parser_state vparse_tokens
+  initial_naive_parser_state naive_parse_token
+  (* instruction_grammar *)
+  instr_bigrammar BiGrammar.bigrammar_to_grammar
+  X86Semantics.X86_Compile.instr_to_rtl
+  Encode.x86_encode (* encoder from the bigrammar *)
+  Encode.enc_pre_instr_bytes (* a manual encoder *)
+  .
+
+(* Separate Extraction initial_parser_state parse_tokens *)
+(*   vinitial_parser_state vparse_tokens *)
+(*   initial_naive_parser_state naive_parse_token *)
+(*   instruction_grammar X86Semantics.X86_Compile.instr_to_rtl. *)
+
+(* Separate Extraction build_dfa dfa_recognize par2rec X86_PARSER.instr_grammar. *)
+
+(* Extraction "recognizer.ml" build_dfa dfa_recognize par2rec . *)
diff --git a/x86model/Model/decenc_tests/ls b/x86model/Model/decenc_tests/ls
new file mode 100644
index 0000000..d5de5b7
Binary files /dev/null and b/x86model/Model/decenc_tests/ls differ
diff --git a/x86model/Model/decenc_tests/pwd b/x86model/Model/decenc_tests/pwd
new file mode 100644
index 0000000..2ed754a
Binary files /dev/null and b/x86model/Model/decenc_tests/pwd differ
diff --git a/x86model/Model/decenc_tests/tailf b/x86model/Model/decenc_tests/tailf
new file mode 100644
index 0000000..666b284
Binary files /dev/null and b/x86model/Model/decenc_tests/tailf differ