| Summary: | sys-fs/encfs-1.9.2: keyword and stabilization request | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Felix Janda <felix.janda> |
| Component: | Stabilization | Assignee: | Anthony Basile <blueness> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | jstein, media-video |
| Priority: | Normal | Keywords: | STABLEREQ |
| Version: | unspecified | Flags: | stable-bot:
sanity-check+
|
| Hardware: | All | ||
| OS: | Linux | ||
| See Also: | https://bugs.gentoo.org/show_bug.cgi?id=630090 | ||
| Whiteboard: | |||
| Package list: |
sys-fs/encfs-1.9.2
dev-libs/tinyxml2-3.0.0
|
Runtime testing required: | --- |
| Bug Depends on: | |||
| Bug Blocks: | 634160, 643478 | ||
|
Description
Felix Janda
2017-09-09 14:42:33 UTC
This requires the stabilization of dev-libs/tinyxml2:0/3. I'm adding the Media Video team because they're maintaining tinyxml2. An automated check of this bug failed - the following atom is unknown: dev-libs/tinyxml2:0/3 Please verify the atom list. Keywording/stabilization on x86 stopped due to bug 630090. An automated check of this bug failed - the following atom is unknown: dev-libs/tinyxml2:0/3 Please verify the atom list. A test failure with an unstable perl version cannot block stabilisation here. (In reply to Andreas Sturmlechner from comment #5) > A test failure with an unstable perl version cannot block stabilisation here. thanks, i'm tired of arguing about stabilization with test failures. i really don't want to add RESTRICT="test" cheaply to a bunch of packages just to get them past stabilization. The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=d925e7b50dc41bb2beeeefcf501a87fed32afe82 commit d925e7b50dc41bb2beeeefcf501a87fed32afe82 Author: Andreas Sturmlechner <asturm@gentoo.org> AuthorDate: 2017-10-09 19:43:38 +0000 Commit: Andreas Sturmlechner <asturm@gentoo.org> CommitDate: 2017-11-29 18:26:23 +0000 profiles: arm: Mask kde-plasma/plasma-meta[crypt] Keywording is stalled and thus arm fails to meet minimum version. Bug: https://bugs.gentoo.org/630486 Bug: https://bugs.gentoo.org/633882 profiles/arch/arm/package.use.mask | 5 +++++ 1 file changed, 5 insertions(+)} ping x86 Apparently it was forgotten to add amd64 to CC list. Still failing tests for x86:
> 2/2 Test #2: scriptedtests ....................***Failed 1.04 sec
>
> 2/2 Testing: scriptedtests
> 2/2 Test: scriptedtests
> Command: "/usr/bin/perl" "-MTest::Harness" "-e" "$$Test::Harness::verbose=1; runtests @ARGV;" "/var/tmp/portage/sys-fs/encfs-1.9.2/work/encfs-1.9.2/tests/normal.t.pl" "/var/tmp/portage/sys-fs/encfs-1.9.2/work/encfs-1.9.2/tests/reverse.t.pl"
> Directory: /var/tmp/portage/sys-fs/encfs-1.9.2/work/encfs-1.9.2
> "scriptedtests" start time: Jan 08 20:10 CET
> Output:
> ----------------------------------------------------------
>
> runTests: mode=standard
> Bailout called. Further testing stopped:
>
> # Failed test 'encfs command returns 0'
> # at /var/tmp/portage/sys-fs/encfs-1.9.2/work/encfs-1.9.2/tests/normal.t.pl line 312.
> FAILED--Further testing stopped.
> <end of output>
> Test time = 1.04 sec
It is a problem with the test suite and our out of source build. If I copy encfs*_build dir back to $S, i.e. tests are passing: > $ ./test.sh > tests/normal.t.pl ... > runTests: mode=standard > tests/normal.t.pl ... 54/116 > runTests: mode=paranoia > tests/normal.t.pl ... ok > tests/reverse.t.pl .. ok > All tests successful. > Files=2, Tests=141, 27 wallclock secs ( 0.05 usr 0.01 sys + 19.87 cusr 2.10 csys = 22.03 CPU) > Result: PASS > portage@vm-gentoo-x86 ~/sys-fs/encfs-1.9.2/work/encfs-1.9.2 $ echo $? > 0 The following change to the ebuild fixes the problem for: --- a/sys-fs/encfs/encfs-1.9.2.ebuild +++ b/sys-fs/encfs/encfs-1.9.2.ebuild @@ -25,6 +25,9 @@ DEPEND=" sys-devel/gettext virtual/pkgconfig" +# Build dir is hardcoded in test suite +BUILD_DIR="${S}/build" + src_configure() { local mycmakeargs=( -DENABLE_NLS="$(usex nls)" > >>> Source compiled. > >>> Test phase: sys-fs/encfs-1.9.2 > >>> Working in BUILD_DIR: "/var/tmp/portage/sys-fs/encfs-1.9.2/work/encfs-1.9.2/build" > ctest -j 5 --test-load 7.95 > Invalid value for 'Test Load' : 7.95 > Test project /var/tmp/portage/sys-fs/encfs-1.9.2/work/encfs-1.9.2/build > Start 1: checkops > Start 2: scriptedtests > 1/2 Test #1: checkops ......................... Passed 0.13 sec > 2/2 Test #2: scriptedtests .................... Passed 27.88 sec > > 100% tests passed, 0 tests failed out of 2 > > Total Test time (real) = 27.89 sec > * Tests succeeded. > >>> Completed testing sys-fs/encfs-1.9.2 > > >>> Install encfs-1.9.2 into /var/tmp/portage/sys-fs/encfs-1.9.2/image/ category sys-fs > (In reply to Thomas Deutschmann from comment #11) > > The following change to the ebuild fixes the problem for: > > --- a/sys-fs/encfs/encfs-1.9.2.ebuild > +++ b/sys-fs/encfs/encfs-1.9.2.ebuild > @@ -25,6 +25,9 @@ DEPEND=" > sys-devel/gettext > virtual/pkgconfig" > > +# Build dir is hardcoded in test suite > +BUILD_DIR="${S}/build" > + > src_configure() { > local mycmakeargs=( > -DENABLE_NLS="$(usex nls)" > > This is a problem but its not the biggest problem with the tests. The problem is with test 2 which tries to do an encfs mount (read the perl). This can fail for spurious reasons: 1) On a hardened kernel this is forbiddent 2) On gentoo-sources you might not have FUSE_FS configured 3) While you could use linux-mod.eclass to check for FUSE_FS, the problem is people often switch kernels, so I'm not a fan of checking for kernel modules from an ebuild. I'll add the above along with RESTRICT="test". I wanted to avoid doing that because the tests do work under some circumstances. amd64 stable x86 stable ~arm restored, maybe sparc team wants it too. dropping ~sparc. The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=5393cdf47136677cbf8455f937b992f2082dc55b commit 5393cdf47136677cbf8455f937b992f2082dc55b Author: Rolf Eike Beer <eike@sf-mail.de> AuthorDate: 2018-04-29 19:39:47 +0000 Commit: Sergei Trofimovich <slyfox@gentoo.org> CommitDate: 2018-04-29 19:41:40 +0000 sys-fs/encfs: keyworded 1.9.2 for sparc Bug: https://bugs.gentoo.org/630486 Package-Manager: Portage-2.3.24, Repoman-2.3.6 RepoMan-Options: --include-arches="sparc" sys-fs/encfs/encfs-1.9.2.ebuild | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)} The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=f825d22a2e3ccb85bc5a1fa0c8a06b2f8d1f0b61 commit f825d22a2e3ccb85bc5a1fa0c8a06b2f8d1f0b61 Author: Sergei Trofimovich <slyfox@gentoo.org> AuthorDate: 2018-04-29 20:25:17 +0000 Commit: Sergei Trofimovich <slyfox@gentoo.org> CommitDate: 2018-04-29 20:25:29 +0000 sys-fs/encfs: keyworded newer versions for sparc Reported-by: [Arfrever] Bug: https://bugs.gentoo.org/630486 Package-Manager: Portage-2.3.31, Repoman-2.3.9 sys-fs/encfs/encfs-1.9.4.ebuild | 2 +- sys-fs/encfs/encfs-1.9.5.ebuild | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)} sparc stable. (In reply to Rolf Eike Beer from comment #19) > sparc stable. meh. at that time, we already had 1.9.4 stable, but now sparc is blocking 1.9.2 cleanup. |
