Re: ASDF test failure

Daniel KochmaƄski <[email protected]> Mon, 06 Apr 2026 07:48:49 +0000
Newsgroups gmane.lisp.ecl.general
Message-ID <ykmqZJpOVOBf_cb0wqYAfCnped2F6T-L-ulEFWrLG5z_lObnUYoQetKhT6aL8rhY6qjGQH75Ts5Vk_VxN4FXjEQqu-AVfkmI_m3VD8i1Ahg=@turtleware.eu>
Hey Robert,

> I'm getting a test failure on ASDF with ECL in building a DLL on Linux. H=
oping that someone can suggest an appropriate fix. You can find the full tr=
anscript here: https://gitlab.common-lisp.net/asdf/asdf/-/jobs/114480
>=20
> Here's the snippet with the error compiling through ECL to gcc:
>=20
>     ;;; Compiling /builds/asdf/asdf/test/monodll-user.lisp.
>     ;;; OPTIMIZE levels: Safety=3D3, Space=3D0, Speed=3D2, Debug=3D3
>     ;;;
>     ;;; Internal error:
>     ;;;   ** Error code 1 when executing
>     ;;; (EXT:RUN-PROGRAM "gcc" ("-I." "-I/usr/local/include/" "-D_GNU_SOU=
RCE" "-D_FILE_OFFSET_BITS=3D64" "-g" "-O2" "-fPIC" "-D_THREAD_SAFE" "-Dlinu=
x" "-O2" "-c" "/builds/asdf/asdf/build/fasls/ecl-24.5.10-linux-x64/asdf/tes=
t/monodll-user.c" "-o" "/builds/asdf/asdf/build/fasls/ecl-24.5.10-linux-x64=
/asdf/test/monodll-user.o")):
>     ;;; In file included from /usr/local/include/ecl/ecl.h:81,
>     ;;;                  from /usr/local/include/ecl/ecl-cmp.h:31,
>     ;;;                  from /builds/asdf/asdf/build/fasls/ecl-24.5.10-l=
inux-x64/asdf/test/monodll-user.c:5:
>     ;;; /builds/asdf/asdf/build/fasls/ecl-24.5.10-linux-x64/asdf/test/mon=
odll-user.c: In function 'L2always_42':
>     ;;; /builds/asdf/asdf/build/fasls/ecl-24.5.10-linux-x64/asdf/test/mon=
odll-user.c:21:26: error: implicit declaration of function 'always_42'; did=
 you mean 'L2always_42'? [-Wimplicit-function-declaration]
>     ;;;    21 |    value0 =3D ecl_make_int(always_42());
>     ;;;       |                          ^~~~~~~~~
>     TEST ABORTED: COMPILE-FILE-ERROR while
>                   compiling #<cl-source-file "test-asdf/monodll-user" "fi=
le-type">
>

it seems that you are running the old release of ECL. Last month we've made=
 a new release
ECL 26.3.27 -- perhaps the problem is already fixed there?

If not, please provide details how to run this test; preferably on the issu=
e tracker, but
if you can't do that, then the mailing list will do.

Best regards,
Daniel