Bug#901823: ocaml-atd FTBFS on i386: tests failure

[Stéphane Glondu <glondu@debian.org>]

Le 18/06/2018 à 22:31, Adrian Bunk a écrit :
> Source: ocaml-atd
> Version: 1.12.0-1
> Severity: serious
> 
> https://buildd.debian.org/status/fetch.php?pkg=ocaml-atd&arch=i386&ver=1.12.0-1&stamp=1527061667&raw=0
> 
> ...
>    dh_auto_test -a
> 	make -j4 check
> make[1]: Entering directory '/<<PKGBUILDDIR>>'
> jbuilder runtest
> File "atd/test/jbuild", line 6, characters 11-14:
> Warning: This field is useless without a (public_names ...) field.
> File "atdgen/test/jbuild", line 62, characters 11-14:
> Warning: This field is useless without a (public_names ...) field.
>       ocamlc atdgen/test/.test_atdgen_main.eobjs/test_atdgen_main.{cmi,cmo,cmt}
> File "atdgen/test/test_atdgen_main.ml", line 913, characters 15-22:
> Warning 52: Code should not depend on the actual values of
> this constructor's arguments. They are only for information
> and may change in future versions. (See manual section 8.5)
> File "atdgen/test/test_atdgen_main.ml", line 919, characters 15-23:
> Warning 52: Code should not depend on the actual values of
> this constructor's arguments. They are only for information
> and may change in future versions. (See manual section 8.5)
>     ocamlopt atdgen/test/.test_atdgen_main.eobjs/test_atdgen_main.{cmx,o}
> File "atdgen/test/test_atdgen_main.ml", line 913, characters 15-22:
> Warning 52: Code should not depend on the actual values of
> this constructor's arguments. They are only for information
> and may change in future versions. (See manual section 8.5)
> File "atdgen/test/test_atdgen_main.ml", line 919, characters 15-23:
> Warning 52: Code should not depend on the actual values of
> this constructor's arguments. They are only for information
> and may change in future versions. (See manual section 8.5)
> test_atdgen_main alias atdgen/test/runtest (got signal SEGV)
> (cd _build/default/atdgen/test && ./test_atdgen_main.exe)
> ----- ocaml internals -----
> Makefile:7: recipe for target 'tests' failed
> make[1]: *** [tests] Error 1

I had a look, and the test looks obviously buggy to me, but I can
understand why it triggers only on i386.

I see there is a new upstream release. Johannes, are you planning to
package it?

Cheers,

-- 
Stéphane