Generalize list1i etc. to all signed integer types

* src/lisp.h (list1i, list2i, list3i, list4i):
Accept intmax_t instead of EMACS_INT, and use make_int instead
of make_fixnum.  This should help avoid integer-overflow
problems akin to the Time bug (Bug#37795).

diff --git a/src/lisp.h b/src/lisp.h
index fe20add2d70c54983c2d4a8ea9f80638a6aedca9..04fa1d64eab2597af00030292e0125034ceed416 100644 (file)
--- a/src/lisp.h
+++ b/src/lisp.h
@@ -3862,28 +3862,27 @@ extern void visit_static_gc_roots (struct gc_root_visitor visitor);
 /* Build a frequently used 1/2/3/4-integer lists.  */
 
 INLINE Lisp_Object
-list1i (EMACS_INT x)
+list1i (intmax_t a)
 {
-  return list1 (make_fixnum (x));
+  return list1 (make_int (a));
 }
 
 INLINE Lisp_Object
-list2i (EMACS_INT x, EMACS_INT y)
+list2i (intmax_t a, intmax_t b)
 {
-  return list2 (make_fixnum (x), make_fixnum (y));
+  return list2 (make_int (a), make_int (b));
 }
 
 INLINE Lisp_Object
-list3i (EMACS_INT x, EMACS_INT y, EMACS_INT w)
+list3i (intmax_t a, intmax_t b, intmax_t c)
 {
-  return list3 (make_fixnum (x), make_fixnum (y), make_fixnum (w));
+  return list3 (make_int (a), make_int (b), make_int (c));
 }
 
 INLINE Lisp_Object
-list4i (EMACS_INT x, EMACS_INT y, EMACS_INT w, EMACS_INT h)
+list4i (intmax_t a, intmax_t b, intmax_t c, intmax_t d)
 {
-  return list4 (make_fixnum (x), make_fixnum (y),
-               make_fixnum (w), make_fixnum (h));
+  return list4 (make_int (a), make_int (b), make_int (c), make_int (d));
 }
 
 extern Lisp_Object make_uninit_bool_vector (EMACS_INT);