On 07/21/2018 02:11 PM, John Ferlan wrote:
On 07/12/2018 03:37 AM, Michal Privoznik wrote:
The most important part is LIBVIRTD_PATH env var fix. It is used in virFileFindResourceFull() from tests. The libvirtd no longer lives under daemon/.
Then, libvirtd-fail test was still failing (as expected) but not because of missing config file but because it was trying to execute (nonexistent) top_builddir/daemon/libvirtd which fulfilled expected outcome and thus test did not fail.
Thirdly, lcov was told to generate coverage for daemon/ dir too.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> --- Makefile.am | 2 +- run.in | 2 +- tests/libvirtd-fail | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/Makefile.am b/Makefile.am index 1926e21b7a..709064c6a6 100644 --- a/Makefile.am +++ b/Makefile.am @@ -80,7 +80,7 @@ check-access: cov: clean-cov $(MKDIR_P) $(top_builddir)/coverage $(LCOV) -c -o $(top_builddir)/coverage/libvirt.info.tmp \ - -d $(top_builddir)/src -d $(top_builddir)/daemon \ + -d $(top_builddir)/src \
Since daemon is the former name and this appears to be a clean label target for coverage, perhaps we should keep daemon just to clean up "old" trees...
No. This is no a clean label. The rule says: in order to make target "cov" you need to make target "clean-cov" first as it is dependency. So I'm changing the create target not the cleanup. And -d $dir to lcov means "include directory $dir to search for .da files" (whatever they are - doesn't matter now). Michal