New slatec translations for ecl's string issue

I've regenerated all of the slatec code using the latest version of
f2cl.  This fixes an issue with ecl 11.x where ecl would convert literal
strings to either base-strings or strings depending on the content and
this causes problems with the declarations f2cl had generated.

As a side effect, a couple of the find_root(quad_qags(...)) tests in
rtest14 now pass.  They were failing because quad_qags wanted to print
some warnings, but the type declarations caused an runtime error that
maxima thought was a general failure of quad_qags.

I've tested these changes using the testsuite with cmucl, ccl, ecl, and
sbcl on my Mac, and no unexpected tests fail.

Please test these with your favorite lisp and let me know if there are
any issues.

Ray