Dear Maintainer,

The file /usr/include/f2c.h is provided by the "libf2c2-dev:amd64"
package, currently at version 20240504-1+b2.

It includes these lines:

#ifdef INTEGER_STAR_8   /* Adjust for integer*8. */
typedef int64_t longit;
typedef uint64_t ulongint;

The name "longit" is a typo for "longint".  The f2c translator uses
the correct name "longint" for integer*8.  (I have some thoughts
about whether "longint" is a good name, but that's not what this bug
report is about.)  The result is that when f2c is applied to Fortran
code that uses the "integer*8" type, the resulting C code does not
compile.  Adding the missing 'n' to that line should fix the problem.

In the git repo for this package:
    <https://salsa.debian.org/debian/libf2c2.git>
the problem was introduced 2020-07-08 by
    debian/patches/0008-XintXX_t.patch
in commit 279d35f.

(I'm frankly a bit surprised that this hasn't been fixed in nearly
6 years.)

This bug does not affect me.  I don't use f2c.  I played with it
in response to a recent Usenet post and happened to run across this.

Not surprisingly, I see the same behavior on Ubuntu and not on
Fedora.