Created attachment 413696 [details] dev-util:pkgconf-0.9.11:20151004-100515.log Happens on HPPA, PPC64 and AMD64. [ebuild N ~] dev-util/pkgconf-0.9.11::gentoo USE="pkg-config -strict" 0 KiB
Created attachment 414050 [details, diff] pkgconf-0.9.11-tests-multilib.patch Also happens on Lilblue (uclibc on AMD64). The problem is the test runner script is not aware of out of source directory builds. Should mgorny@gentoo.org added to CC? He is the author of the script. Good to know some Gentoo devs have FEATURES="test" in make.conf. It looks it's not always the case.
Created attachment 414052 [details, diff] pkgconf-0.9.11.ebuild.patch
Something's really messed up with the tests, and I don't know the cause. I'm proxy-maintaining the ebuild for now and comments in the ebuild say the tests should be disabled until 0.9.13, so I'm keeping it that way. I'm not sure if I should reassign the bug to myself or leave it to jdhore.
I've opened a pull request bumping pkgconf to a version with this bug fixed: https://github.com/gentoo/gentoo/pull/2410
Affected version all long gone
commit 24f61911d09baf5131b996280731107b64b05d9a Author: Alon Bar-Lev <alonbl@gentoo.org> Date: Sun Nov 1 18:55:18 2015 +0200 dev-util/pkgconf: restrict tests as not working Package-Manager: portage-2.2.20.1 diff --git a/dev-util/pkgconf/pkgconf-0.9.11.ebuild b/dev-util/pkgconf/pkgconf-0.9.11.ebuild index ff55bedf4ef..1294e1be804 100644 --- a/dev-util/pkgconf/pkgconf-0.9.11.ebuild +++ b/dev-util/pkgconf/pkgconf-0.9.11.ebuild @@ -20,6 +20,8 @@ LICENSE="BSD-1" SLOT="0" IUSE="+pkg-config strict" +RESTRICT="test" # at least until 0.9.13 + DEPEND="" RDEPEND="${DEPEND} pkg-config? (
I'm not trying to cause trouble, but I don't understand why this is still open. 0.9.11 was removed here: https://gitweb.gentoo.org/repo/gentoo.git/commit/dev-util/pkgconf/pkgconf-0.9.11.ebuild?id=796c9d664b4958bd2ca9195cfcea439f21c7d415
(In reply to Ben Kohler from comment #7) > 0.9.11 It doesn't matter what version happens to appear in the Summary. The bug wasn't fixed but worked around (and at the time no back-reference to this bug report was made, and the comment promises the test restriction was still under review).
If you're simply not going to fix it, then don't pretend the bug report is obsolete.
(In reply to Jeroen Roovers from comment #9) > If you're simply not going to fix it, then don't pretend the bug report is > obsolete. The bug has been fixed long ago. Check the repo before posting such accusations.
commit 080246df65f36207c60eff54397bf432a860e73d Author: Mike Gilbert <floppym@gentoo.org> Date: Sun May 7 14:26:18 2017 -0400 dev-util/pkgconf: enable testing Package-Manager: Portage-2.3.5_p31, Repoman-2.3.2_p61
(In reply to Mike Gilbert from comment #10) > Check the repo before posting such accusations. 1) I didn't accuse anyone. 2) You couldn't point that out in a comment?
(In reply to Jeroen Roovers from comment #12) You did accuse me of trying to sweep the problem under the rug. You couldn't have done your homework?
(In reply to Mike Gilbert from comment #13) > (In reply to Jeroen Roovers from comment #12) > > You did accuse me of trying to sweep the problem under the rug. No, that's how you read it. I don't know what you read into what I did or said.
(In reply to Jeroen Roovers from comment #9) > If you're simply not going to fix it, then don't pretend the bug report is > obsolete. I infer a couple things from this statement: 1. You assume the problem is still a problem. 2. You assume that I have no intention of fixing it. Neither of these was true, which you would have known if you had taken a minute to check the current state of the package. You were able to dredge up a commit from 2015, so I'm a bit confused on how you missed that the problem had been fixed already. In the best case, you did actually check it and somehow came to the wrong conclusion. In the worst case, you didn't check it, or stopped looking after finding the commit from 2015 and jumped to a conclusion based on incomplete/incorrect information. More generally, my prior experience with you tells me that you have a tendency to jump to incorrect conclusions, and that really makes you difficult to work with. I will acknowledge that my re-closing of the bug without comment was a bit rude, and I really only did it because I saw your name on the comments. I will try not to do that in the future.
