First commit ; includes first TP and minijazz compiler

21 files changed, 558 insertions, 0 deletions

diff --git a/tp1/_tags b/tp1/_tags
new file mode 100644
index 0000000..503ce62
--- /dev/null
+++ b/tp1/_tags
@@ -0,0 +1,3 @@
+true: use_menhir
+<*.ml>: debug
+<*.byte>: use_unix, debug
\ No newline at end of file
diff --git a/tp1/graph.ml b/tp1/graph.ml
new file mode 100644
index 0000000..54128ff
--- /dev/null
+++ b/tp1/graph.ml
@@ -0,0 +1,61 @@
+exception Cycle
+type mark = NotVisited | InProgress | Visited
+
+type 'a graph =
+    { mutable g_nodes : 'a node list }
+and 'a node = {
+  n_label : 'a;
+  mutable n_mark : mark;
+  mutable n_link_to : 'a node list;
+  mutable n_linked_by : 'a node list;
+}
+
+let mk_graph () = { g_nodes = [] }
+
+let add_node g x =
+  let n = { n_label = x; n_mark = NotVisited; n_link_to = []; n_linked_by = [] } in
+  g.g_nodes <- n::g.g_nodes
+
+let node_for_label g x =
+  List.find (fun n -> n.n_label = x) g.g_nodes
+
+let add_edge g id1 id2 =
+  let n1 = node_for_label g id1 in
+  let n2 = node_for_label g id2 in
+  n1.n_link_to <- n2::n1.n_link_to;
+  n2.n_linked_by <- n1::n2.n_linked_by
+
+let clear_marks g =
+  List.iter (fun n -> n.n_mark <- NotVisited) g.g_nodes
+
+let find_roots g =
+  List.filter (fun n -> n.n_linked_by = []) g.g_nodes
+
+let has_cycle g =
+	clear_marks g;
+	let rec visit n =
+		match n.n_mark with
+		| InProgress -> true
+		| Visited -> false
+		| NotVisited ->
+			n.n_mark <- InProgress;
+			let ret = List.fold_left (fun x n -> x || (visit n)) false n.n_link_to in
+			n.n_mark <- Visited;
+			ret
+	in
+	let ret = List.fold_left (fun x n -> x || (if n.n_mark = Visited then false else visit n)) false g.g_nodes
+	in clear_marks g; ret;;
+
+let topological g =
+	clear_marks g;
+	let rec aux acc n =
+		if n.n_mark = Visited
+			then acc
+			else begin
+				n.n_mark <- Visited;
+				n.n_label :: (List.fold_left (fun x n -> aux x n) acc n.n_linked_by)
+			end
+	in
+	let ret = List.fold_left (fun x n -> aux x n) [] g.g_nodes
+	in clear_marks g; List.rev ret;;
+
diff --git a/tp1/graph_test.ml b/tp1/graph_test.ml
new file mode 100644
index 0000000..ac31677
--- /dev/null
+++ b/tp1/graph_test.ml
@@ -0,0 +1,28 @@
+open Graph
+
+let rec check l = match l with
+  | [] | [_] -> true
+  | s1::s2::l -> (String.length s1 <= String.length s2) && (check (s2::l))
+
+let test_good () =
+  let g = mk_graph () in
+  add_node g "1"; add_node g "21"; add_node g "22"; add_node g "333";
+  add_edge g "1" "21"; add_edge g "1" "22";
+  add_edge g "21" "333"; add_edge g "22" "333";
+  let l = topological g in
+  print_string "Test: Tri topologique --> ";
+  if check l then print_endline "OK" else print_endline "FAIL";
+  List.iter print_endline l;
+  print_newline ()
+
+let test_cycle () =
+  let g = mk_graph () in
+  add_node g "1"; add_node g "2"; add_node g "3";
+  add_edge g "1" "2"; add_edge g "2" "3"; add_edge g "3" "1";
+  print_string "Test: Detection de cycle --> ";
+  if has_cycle g then print_endline "OK" else print_endline "FAIL"
+;;
+
+test_cycle ();;
+test_good ();;
+
diff --git a/tp1/netlist.ml b/tp1/netlist.ml
new file mode 100644
index 0000000..b1d7932
--- /dev/null
+++ b/tp1/netlist.ml
@@ -0,0 +1,17 @@
+exception Parse_error of string
+
+let find_file filename =
+  try
+    open_in filename
+  with
+    | _ -> raise (Parse_error "No such file '%s'")
+
+let read_file filename =
+  let ic = find_file filename in
+  let lexbuf = Lexing.from_channel ic in
+  lexbuf.Lexing.lex_curr_p <- { lexbuf.Lexing.lex_curr_p with Lexing.pos_fname = filename };
+  try
+    Netlist_parser.program Netlist_lexer.token lexbuf
+  with
+    | e -> raise (Parse_error ("Syntax error (exception: "^(Printexc.to_string e)^")"))
+
diff --git a/tp1/netlist_ast.ml b/tp1/netlist_ast.ml
new file mode 100644
index 0000000..ae16888
--- /dev/null
+++ b/tp1/netlist_ast.ml
@@ -0,0 +1,42 @@
+type ident = string
+
+module Env = struct
+  include Map.Make(struct
+    type t = ident
+    let compare = compare
+  end)
+
+  let of_list l =
+    List.fold_left (fun env (x, ty) -> add x ty env) empty l
+end
+
+type ty = TBit | TBitArray of int
+type value = VBit of bool | VBitArray of bool array
+
+type binop = Or | Xor | And | Nand
+
+type arg =
+    | Avar of ident
+    | Aconst of value
+
+type exp =
+    | Earg of arg
+    | Ereg of ident
+    | Enot of arg
+    | Ebinop of binop * arg * arg
+    | Emux of arg * arg * arg
+    | Erom of int (*addr size*) * int (*word size*) * arg (*read_addr*)
+    | Eram of int (*addr size*) * int (*word size*)
+        * arg (*read_addr*) * arg (*write_enable*)
+        * arg (*write_addr*) * arg (*data*)
+    | Econcat of arg * arg
+    | Eslice of int * int * arg
+    | Eselect of int * arg
+
+type equation = ident * exp
+
+type program =
+    { p_eqs : equation list;
+      p_inputs : ident list;
+      p_outputs : ident list;
+      p_vars : ty Env.t; }
diff --git a/tp1/netlist_lexer.mll b/tp1/netlist_lexer.mll
new file mode 100644
index 0000000..78b0410
--- /dev/null
+++ b/tp1/netlist_lexer.mll
@@ -0,0 +1,37 @@
+{
+open Netlist_parser
+exception Eof
+
+let keyword_list =
+[
+  "AND", AND;
+  "CONCAT", CONCAT;
+  "IN", IN;
+  "INPUT", INPUT;
+  "MUX", MUX;
+  "NAND", NAND;
+  "NOT", NOT;
+  "OR", OR;
+  "OUTPUT", OUTPUT;
+  "RAM", RAM;
+  "REG", REG;
+  "ROM", ROM;
+  "SELECT", SELECT;
+  "SLICE", SLICE;
+  "VAR", VAR;
+  "XOR", XOR;
+]
+
+}
+
+rule token = parse
+    [' ' '\t' '\n']     { token lexbuf }     (* skip blanks *)
+  | "="            { EQUAL }
+  | ":"            { COLON }
+  | ","            { COMMA }
+  | ['0'-'9']+ as lxm { INT(int_of_string lxm) }
+  | ('_' ? ['A'-'Z' 'a'-'z']('_' ? ['A'-'Z' 'a'-'z' ''' '0'-'9']) * as id)
+      { let s = Lexing.lexeme lexbuf in
+        try List.assoc s keyword_list
+        with Not_found -> NAME id }
+  | eof            { EOF }
diff --git a/tp1/netlist_parser.mly b/tp1/netlist_parser.mly
new file mode 100644
index 0000000..0908703
--- /dev/null
+++ b/tp1/netlist_parser.mly
@@ -0,0 +1,70 @@
+%{
+ open Netlist_ast
+
+ let value_of_int n =
+   let rec aux n =
+     let b =
+       match n mod 10 with
+         | 0 -> false
+         | 1 -> true
+         | i -> Format.eprintf "Unexpected: %d@." i; raise Parsing.Parse_error
+     in
+     if n < 10 then
+       [b]
+     else
+       b::(aux (n / 10))
+   in
+   match aux n with
+     | [] -> Format.eprintf "Empty list@."; raise Parsing.Parse_error
+     | [b] -> VBit b
+     | bl -> VBitArray (Array.of_list (List.rev bl))
+%}
+
+%token <int> INT
+%token <string> NAME
+%token AND MUX NAND OR RAM ROM XOR REG NOT
+%token CONCAT SELECT SLICE
+%token COLON EQUAL COMMA VAR IN INPUT OUTPUT
+%token EOF
+
+%start program             /* the entry point */
+%type <Netlist_ast.program> program
+
+%%
+program:
+  INPUT inp=separated_list(COMMA, NAME)
+    OUTPUT out=separated_list(COMMA, NAME)
+    VAR vars=separated_list(COMMA, var) IN eqs=list(equ) EOF
+    { { p_eqs = eqs; p_vars = Env.of_list vars; p_inputs = inp; p_outputs = out; } }
+
+equ:
+  x=NAME EQUAL e=exp { (x, e) }
+
+exp:
+  | a=arg { Earg a }
+  | NOT x=arg { Enot x }
+  | REG x=NAME { Ereg x }
+  | AND x=arg y=arg { Ebinop(And, x, y) }
+  | OR x=arg y=arg { Ebinop(Or, x, y) }
+  | NAND x=arg y=arg { Ebinop(Nand, x, y) }
+  | XOR x=arg y=arg { Ebinop(Xor, x, y) }
+  | MUX x=arg y=arg z=arg { Emux(x, y, z) }
+  | ROM addr=INT word=INT ra=arg
+    { Erom(addr, word, ra) }
+  | RAM addr=INT word=INT ra=arg we=arg wa=arg data=arg
+    { Eram(addr, word, ra, we, wa, data) }
+  | CONCAT x=arg y=arg
+     { Econcat(x, y) }
+  | SELECT idx=INT x=arg
+     { Eselect (idx, x) }
+  | SLICE min=INT max=INT x=arg
+     { Eslice (min, max, x) }
+
+arg:
+  | n=INT { Aconst (value_of_int n) }
+  | id=NAME { Avar id }
+
+var: x=NAME ty=ty_exp { (x, ty) }
+ty_exp:
+  | /*empty*/ { TBit }
+  | COLON n=INT { TBitArray n }
diff --git a/tp1/netlist_printer.ml b/tp1/netlist_printer.ml
new file mode 100644
index 0000000..d513137
--- /dev/null
+++ b/tp1/netlist_printer.ml
@@ -0,0 +1,82 @@
+open Netlist_ast
+open Format
+
+let rec print_env print lp sep rp ff env =
+  let first = ref true in
+  fprintf ff "%s" lp;
+  Env.iter
+    (fun x ty ->
+      if !first then
+        (first := false; fprintf ff "%a" print (x, ty))
+      else
+        fprintf ff "%s%a" sep print (x, ty)) env;
+  fprintf ff "%s" rp
+
+let rec print_list print lp sep rp ff = function
+  | [] -> ()
+  | x :: l ->
+      fprintf ff "%s%a" lp print x;
+      List.iter (fprintf ff "%s %a" sep print) l;
+      fprintf ff "%s" rp
+
+let print_ty ff ty = match ty with
+  | TBit -> ()
+  | TBitArray n -> fprintf ff " : %d" n
+
+let print_bool ff b =
+  if b then
+    fprintf ff "1"
+  else
+    fprintf ff "0"
+
+let print_value ff v = match v with
+  | VBit b -> print_bool ff b
+  | VBitArray a -> Array.iter (print_bool ff) a
+
+let print_arg ff arg = match arg with
+  | Aconst v -> print_value ff v
+  | Avar id -> fprintf ff "%s" id
+
+let print_op ff op = match op with
+  | And -> fprintf ff "AND"
+  | Nand -> fprintf ff "NAND"
+  | Or -> fprintf ff "OR"
+  | Xor -> fprintf ff "XOR"
+
+let print_exp ff e = match e with
+  | Earg a -> print_arg ff a
+  | Ereg x -> fprintf ff "REG %s" x
+  | Enot x -> fprintf ff "NOT %a" print_arg x
+  | Ebinop(op, x, y) -> fprintf ff  "%a %a %a" print_op op  print_arg x  print_arg y
+  | Emux (c, x, y) -> fprintf ff "MUX %a %a %a " print_arg c  print_arg x  print_arg y
+  | Erom (addr, word, ra) -> fprintf ff "ROM %d %d %a" addr word  print_arg ra
+  | Eram (addr, word, ra, we, wa, data) ->
+      fprintf ff "RAM %d %d %a %a %a %a" addr word
+        print_arg ra  print_arg we
+        print_arg wa  print_arg data
+  | Eselect (idx, x) -> fprintf ff "SELECT %d %a" idx print_arg x
+  | Econcat (x, y) ->  fprintf ff  "CONCAT %a %a" print_arg x  print_arg y
+  | Eslice (min, max, x) -> fprintf ff "SLICE %d %d %a" min max print_arg x
+
+let print_eq ff (x, e) =
+  fprintf ff "%s = %a@." x print_exp e
+
+let print_var ff (x, ty) =
+  fprintf ff "@[%s%a@]" x print_ty ty
+
+let print_vars ff env =
+  fprintf ff "@[<v 2>VAR@,%a@]@.IN@,"
+    (print_env print_var "" ", " "") env
+
+let print_idents ff ids =
+  let print_ident ff s = fprintf ff "%s" s in
+  print_list print_ident """,""" ff ids
+
+let print_program oc p =
+  let ff = formatter_of_out_channel oc in
+  fprintf ff "INPUT %a@." print_idents p.p_inputs;
+  fprintf ff "OUTPUT %a@." print_idents p.p_outputs;
+  print_vars ff p.p_vars;
+  List.iter (print_eq ff) p.p_eqs;
+  (* flush *)
+  fprintf ff "@."
diff --git a/tp1/netlist_simulator.byte b/tp1/netlist_simulator.byte
new file mode 100755
index 0000000..155d39a
--- /dev/null
+++ b/tp1/netlist_simulator.byte
diff --git a/tp1/scheduler.ml b/tp1/scheduler.ml
new file mode 100644
index 0000000..130164b
--- /dev/null
+++ b/tp1/scheduler.ml
@@ -0,0 +1,52 @@
+open Netlist_ast
+
+module Smap = Map.Make(String)
+
+exception Combinational_cycle
+
+let read_exp eq =
+	let add_arg x l = match x with
+		| Avar(f) -> f::l
+		| Aconst(_) -> l
+	in
+	let aux = function
+	| Earg(x) -> add_arg x []
+	| Ereg(i) -> []
+	| Enot(x) -> add_arg x []
+	| Ebinop(_, x, y) -> add_arg x (add_arg y [])
+	| Emux(a, b, c) -> add_arg a (add_arg b (add_arg c []))
+	| Erom(_, _, a) -> add_arg a []
+	| Eram(_, _, a, b, c, d) -> []
+	| Econcat(u, v) -> add_arg u (add_arg v [])
+	| Eslice(_, _, a) -> add_arg a []
+	| Eselect(_, a) -> add_arg a []
+	in
+		aux eq;;
+
+let schedule p =
+	let graph = Graph.mk_graph() in
+	let eq_map = List.fold_left (fun x (vn, eqn) -> Smap.add vn eqn x) Smap.empty p.p_eqs in
+	(* Add variables as graph nodes *)
+	List.iter (fun (k, _) -> Graph.add_node graph k) p.p_eqs;
+	(* Add dependencies as graph edges *)
+	List.iter 
+		(fun (n, e) -> List.iter
+			(fun m -> if Smap.mem m eq_map then Graph.add_edge graph m n else ())
+			(read_exp e))
+		p.p_eqs;
+	(* Verify there are no cycles *)
+	if Graph.has_cycle graph then raise Combinational_cycle;
+	(* Topological sort of graph nodes *)
+	let topo_vars = Graph.topological graph in
+	(* Construct new program with sorted expression list *)
+	{
+		p_eqs = List.fold_right
+			(fun n x -> if Smap.mem n eq_map then (n, Smap.find n eq_map)::x else x) topo_vars [];
+		p_inputs = p.p_inputs;
+		p_outputs = p.p_outputs;
+		p_vars = p.p_vars;
+	}
+	;;
+
+
+
diff --git a/tp1/scheduler_test.ml b/tp1/scheduler_test.ml
new file mode 100644
index 0000000..20c20f5
--- /dev/null
+++ b/tp1/scheduler_test.ml
@@ -0,0 +1,41 @@
+let print_only = ref false
+let number_steps = ref (-1)
+
+let compile filename =
+  try
+    let p = Netlist.read_file filename in
+    let out_name = (Filename.chop_suffix filename ".net") ^ "_sch.net" in
+    let out = open_out out_name in
+    let close_all () =
+      close_out out
+    in
+    begin try
+        let p = Scheduler.schedule p in
+        Netlist_printer.print_program out p;
+      with
+        | Scheduler.Combinational_cycle ->
+            Format.eprintf "The netlist has a combinatory cycle.@.";
+            close_all (); exit 2
+    end;
+    close_all ();
+    if not !print_only then (
+      let simulator =
+        if !number_steps = -1 then
+          "./netlist_simulator.byte"
+        else
+          "./netlist_simulator.byte -n "^(string_of_int !number_steps)
+      in
+      ignore (Unix.system (simulator^" "^out_name))
+    )
+  with
+    | Netlist.Parse_error s -> Format.eprintf "An error accurred: %s@." s; exit 2
+
+let main () =
+  Arg.parse
+    ["-print", Arg.Set print_only, "Only print the result of scheduling";
+     "-n", Arg.Set_int number_steps, "Number of steps to simulate"]
+    compile
+    ""
+;;
+
+main ()
diff --git a/tp1/test/clock_div.mj b/tp1/test/clock_div.mj
new file mode 100644
index 0000000..ad1e919
--- /dev/null
+++ b/tp1/test/clock_div.mj
@@ -0,0 +1,4 @@
+main() = (o) where
+  o = reg(c);
+  c = not (reg (o))
+end where
\ No newline at end of file
diff --git a/tp1/test/clock_div.net b/tp1/test/clock_div.net
new file mode 100644
index 0000000..9a17fc7
--- /dev/null
+++ b/tp1/test/clock_div.net
@@ -0,0 +1,9 @@
+INPUT 
+OUTPUT o
+VAR
+  _l_2, c, o
+IN
+c = NOT _l_2
+o = REG c
+_l_2 = REG o
+
diff --git a/tp1/test/cm2.mj b/tp1/test/cm2.mj
new file mode 100644
index 0000000..8863bf1
--- /dev/null
+++ b/tp1/test/cm2.mj
@@ -0,0 +1,4 @@
+main(x) = (s, r) where
+  s = reg (x xor s);
+  r = x and s
+end where
\ No newline at end of file
diff --git a/tp1/test/cm2.net b/tp1/test/cm2.net
new file mode 100644
index 0000000..e296b96
--- /dev/null
+++ b/tp1/test/cm2.net
@@ -0,0 +1,9 @@
+INPUT x
+OUTPUT s, r
+VAR
+  _l_1, r, s, x
+IN
+r = AND x s
+s = REG _l_1
+_l_1 = XOR x s
+
diff --git a/tp1/test/fulladder.mj b/tp1/test/fulladder.mj
new file mode 100644
index 0000000..c4b6b0e
--- /dev/null
+++ b/tp1/test/fulladder.mj
@@ -0,0 +1,4 @@
+main(a,b,c) = (s, r) where
+  s = (a xor b) xor c;
+  r = (a and b) or ((a xor b) and c);
+end where
\ No newline at end of file
diff --git a/tp1/test/fulladder.net b/tp1/test/fulladder.net
new file mode 100644
index 0000000..b2271c2
--- /dev/null
+++ b/tp1/test/fulladder.net
@@ -0,0 +1,12 @@
+INPUT a, b, c
+OUTPUT s, r
+VAR
+  _l_1, _l_3, _l_4, _l_5, a, b, c, r, s
+IN
+r = OR _l_3 _l_5
+s = XOR _l_1 c
+_l_1 = XOR a b
+_l_3 = AND a b
+_l_4 = XOR a b
+_l_5 = AND _l_4 c
+
diff --git a/tp1/test/nadder.mj b/tp1/test/nadder.mj
new file mode 100644
index 0000000..0c95386
--- /dev/null
+++ b/tp1/test/nadder.mj
@@ -0,0 +1,19 @@
+fulladder(a,b,c) = (s, r) where
+  s = (a ^ b) ^ c;
+  r = (a & b) + ((a ^ b) & c);
+end where
+
+adder<n>(a:[n], b:[n], c_in) = (o:[n], c_out) where
+  if n = 0 then
+    o = [];
+    c_out = 0
+  else
+    (s_n1, c_n1) = adder<n-1>(a[1..], b[1..], c_in);
+    (s_n, c_out) = fulladder(a[0], b[0], c_n1);
+    o = s_n . s_n1
+  end if
+end where
+
+main(a, b) = (o, c) where
+  (o, c) = adder<1>(a,b,0)
+end where
\ No newline at end of file
diff --git a/tp1/test/nadder.net b/tp1/test/nadder.net
new file mode 100644
index 0000000..bf87051
--- /dev/null
+++ b/tp1/test/nadder.net
@@ -0,0 +1,17 @@
+INPUT a, b
+OUTPUT o, c
+VAR
+  _l_10_50, _l_11_49, _l_16_22, _l_17_21, _l_7_52, _l_9_51, a, b, c, 
+  c_n1_27, o, s_n_26
+IN
+o = s_n_26
+c = OR _l_9_51 _l_11_49
+s_n_26 = XOR _l_7_52 c_n1_27
+_l_7_52 = XOR _l_16_22 _l_17_21
+_l_9_51 = AND _l_16_22 _l_17_21
+_l_10_50 = XOR _l_16_22 _l_17_21
+_l_11_49 = AND _l_10_50 c_n1_27
+c_n1_27 = 0
+_l_16_22 = SELECT 0 a
+_l_17_21 = SELECT 0 b
+
diff --git a/tp1/test/ram.mj b/tp1/test/ram.mj
new file mode 100644
index 0000000..7c4d1ed
--- /dev/null
+++ b/tp1/test/ram.mj
@@ -0,0 +1,15 @@
+const addr = 2
+const word = 4
+
+or_n<n>(a:[n],b:[n]) = (o:[n]) where
+  if n = 0 then
+    o = []
+  else
+    o = (a[0] and b[0]).(or_n<n-1>(a[1..], b[1..]))
+  end if
+end where
+
+main(ra:[addr], we, wa:[addr], c:[word]) = (o:[word]) where
+  o = ram<addr, word>(ra, we, wa, or_n<word>(o, c))
+end where
+
diff --git a/tp1/test/ram.net b/tp1/test/ram.net
new file mode 100644
index 0000000..ee4570a
--- /dev/null
+++ b/tp1/test/ram.net
@@ -0,0 +1,32 @@
+INPUT ra, we, wa, c
+OUTPUT o
+VAR
+  _l_10_22, _l_10_35, _l_10_48, _l_10_61, _l_11_21, _l_11_34, _l_11_47, 
+  _l_11_60, _l_12_20 : 3, _l_12_33 : 2, _l_12_46 : 1, _l_13_19 : 3, _l_13_32 : 2, 
+  _l_13_45 : 1, _l_14_18 : 3, _l_14_31 : 2, _l_14_44 : 1, _l_16 : 4, 
+  _l_9_23, _l_9_36, _l_9_49, _l_9_62, c : 4, o : 4, ra : 2, wa : 2, we
+IN
+o = RAM 2 4 ra we wa _l_16
+_l_16 = CONCAT _l_11_21 _l_14_18
+_l_9_23 = SELECT 0 o
+_l_10_22 = SELECT 0 c
+_l_11_21 = AND _l_9_23 _l_10_22
+_l_12_20 = SLICE 1 3 o
+_l_13_19 = SLICE 1 3 c
+_l_14_18 = CONCAT _l_11_34 _l_14_31
+_l_9_36 = SELECT 0 _l_12_20
+_l_10_35 = SELECT 0 _l_13_19
+_l_11_34 = AND _l_9_36 _l_10_35
+_l_12_33 = SLICE 1 2 _l_12_20
+_l_13_32 = SLICE 1 2 _l_13_19
+_l_14_31 = CONCAT _l_11_47 _l_14_44
+_l_9_49 = SELECT 0 _l_12_33
+_l_10_48 = SELECT 0 _l_13_32
+_l_11_47 = AND _l_9_49 _l_10_48
+_l_12_46 = SLICE 1 1 _l_12_33
+_l_13_45 = SLICE 1 1 _l_13_32
+_l_14_44 = _l_11_60
+_l_9_62 = SELECT 0 _l_12_46
+_l_10_61 = SELECT 0 _l_13_45
+_l_11_60 = AND _l_9_62 _l_10_61
+