Added a new target message to comply with newer versions of GNU Make#9068
Conversation
#fixes 9067 Unit tests were failing on a system with GNU Make v 4.1
|
fixes #9028 |
| 'directory', | ||
| glob.iglob(os.path.join(DATA_PATH, 'make', 'negative', '*')) | ||
| ) | ||
| @pytest.mark.regression('9067') |
There was a problem hiding this comment.
Does this regression actually make sense here? I'm confused why only certain negative tests failed and not all of them for newer versions of Make.
There was a problem hiding this comment.
I marked the regression (even though I didn't add a new test or modified an existing one) to leave information on a related issue.
I'm confused why only certain negative tests failed and not all of them for newer versions of Make.
I am not 😄 There are 3 tests, where one checks that Spack ignores directories without a Makefile. That one passed because _has_make_target returned a result before invoking make.
There was a problem hiding this comment.
To be clearer, the test that succeeds returns a value here:
spack/lib/spack/spack/package.py
Lines 1153 to 1159 in f7068fa
adamjstewart
left a comment
There was a problem hiding this comment.
Thanks for the fix! I never would've tracked this down.
I'm not sold on this 😉 |
…pack#9068) #fixes 9067 Unit tests were failing on a system with GNU Make v 4.1
fixes #9028
fixes #9067
Unit tests were failing on a system with GNU Make v 4.1