ECL on Mac OS X Yosemite fails test-program

[Robert P. Goldman]

Now create an program without UIOP
; Registering #<system "hello-no-uiop">
;;;
;;; Compiling /Users/rpg/lisp/asdf/test/file1.lisp.
;;; OPTIMIZE levels: Safety=3, Space=0, Speed=2, Debug=3
;;;
;;; End of Pass 1.
;;; Finished compiling /Users/rpg/lisp/asdf/test/file1.lisp.
;;;
/private/var/folders/xq/ll1d80qs7hnbw3253yc1_x6m0000gn/T/eclinitNnEej4.c:64:1: warning: control reaches end of non-void function [-Wreturn-type]
}
^
1 warning generated.
RUN
  (ASDF/ACTION:OUTPUT-FILE 'ASDF/BUNDLE:PROGRAM-OP 'ASDF-TEST::HELLO-NO-UIOP) => #P"/Users/rpg/lisp/asdf/build/fasls/ecl-16.0.0-unknown-macosx-x64/asdf/test/hello-no-uiop"
TEST ABORTED: Filesystem error with pathname #P"/var/folders/xq/ll1d80qs7hnbw3253yc1_x6m0000gn/T/tmpUQC0VAHN.tmp".
Either
 1) the file does not exist, or
 2) we are not allowed to access the file, or
 3) the pathname points to a broken symbolic link.

By the way, in my build of ECL, I got many warnings of the above form.

[Faré (fahree)]

I believe this is an ECL bug, which I have reported upstream:
https://gitlab.com/embeddable-common-lisp/ecl/issues/149

Indeed, if I add a #+ecl t in the force-shell test of the final run-program function, it all works for me, whereas if I don't, the new (to ecl) use-run-program backend fails erratically with an error that suggests I got NIL as an exit code where 0 was expected.

[pipping (pipping)]

Is this still an issue?

[Robert P. Goldman (rpgoldman)]

This still happens, but not reliably for me. I am starting to wonder if there's something in the filesystem that triggers it. I don't get a NIL exit code, I get a failure in compilation with clang (exit code 1) with the hello string not defined.