Add unit tests for scalars.

Also add a secp256k1_scalar_is_one function.

src/scalar_impl.h

@@ -54,12 +54,17 @@ void static secp256k1_scalar_inverse(secp256k1_scalar_t *r, const secp256k1_scal
 
 void static secp256k1_scalar_negate(secp256k1_scalar_t *r, const secp256k1_scalar_t *a) {
     secp256k1_num_sub(&r->n, &secp256k1_ge_consts->order, &a->n);
+    secp256k1_num_mod(&r->n, &secp256k1_ge_consts->order);
 }
 
 int static secp256k1_scalar_is_zero(const secp256k1_scalar_t *a) {
     return secp256k1_num_is_zero(&a->n);
 }
 
+int static secp256k1_scalar_is_one(const secp256k1_scalar_t *a) {
+    return secp256k1_num_bits(&a->n) == 1;
+}
+
 int static secp256k1_scalar_is_high(const secp256k1_scalar_t *a) {
     return secp256k1_num_cmp(&a->n, &secp256k1_ge_consts->half_order) > 0;
 }