[m-dev.] testing "mmc --make-interface"

Julien Fischer jfischer at opturion.com
Wed Apr 8 14:44:10 AEST 2020

On Wed, 8 Apr 2020, Peter Wang wrote:

> On Wed, 08 Apr 2020 04:38:31 +1000 (AEST), "Zoltan Somogyi" <zoltan.somogyi at runbox.com> wrote:
>> To help that process, I propose that we add two new test
>> directories:
>> - valid_make_int, which checks that running "mmc --make-interface"
>>   on a module succeeds.
>> - invalid_make_int, which checks that running "mmc --make-interface"
>>   on a module fails, and fails with the expected error output.
> That sounds fine.

That's fine by me as well.

>> In both cases, if the test program has submodules, then we would run
>> "mmake test.depend" and "mmake test.int3s" beforehand. (We will
>> need to do something else for the grade  that don't support mmake,
>> but I don't yet know what.)
> As far as I can tell from a quick test, the grade doesn't prevent those
> targets working.

For testing all grades "support" mmake in the sense that tests in grades that must
use mmc --make are run using mmake's --use-mmc-make option.  That option
should handle the above targets properly.


More information about the developers mailing list