Remove useless hypotheses from Max_sym and Min_sym.

From Max_is_some and Max_is_ge, one can prove that ge is a total relation.
So the consequents hold even if "ge x y" does not (since "ge y x" then
holds by totality).

lib/coq/int/MinMax.v

@@ -73,16 +73,14 @@ exact Zmin_r.
 Qed.
 
 (* Why3 goal *)
-Lemma Max_sym : forall (x:Z) (y:Z), (y <= x)%Z ->
+Lemma Max_sym : forall (x:Z) (y:Z),
   ((ZArith.BinInt.Z.max x y) = (ZArith.BinInt.Z.max y x)).
-intros x y _.
-apply Zmax_comm.
+exact Zmax_comm.
 Qed.
 
 (* Why3 goal *)
-Lemma Min_sym : forall (x:Z) (y:Z), (y <= x)%Z ->
+Lemma Min_sym : forall (x:Z) (y:Z),
   ((ZArith.BinInt.Z.min x y) = (ZArith.BinInt.Z.min y x)).
-intros x y _.
-apply Zmin_comm.
+exact Zmin_comm.
 Qed.
 

theories/comparison.why

@@ -42,8 +42,8 @@ theory MinMax
   axiom Min_x : forall x y:t. ge y x -> min x y = x
   axiom Min_y : forall x y:t. ge x y -> min x y = y
 
-  lemma Max_sym: forall x y:t. ge x y -> max x y = max y x
-  lemma Min_sym: forall x y:t. ge x y -> min x y = min y x
+  lemma Max_sym: forall x y:t. max x y = max y x
+  lemma Min_sym: forall x y:t. min x y = min y x
 
 (***
   function min (x y : t) : t