Summary: | dev-util/cmake-2.8.6-r4 - stop Xvfb after failed test phase | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Toralf Förster <toralf> |
Component: | [OLD] KDE | Assignee: | Gentoo KDE team <kde> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | DuPol, nikoli, x11 |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | bug-406353.patch |
Description
Toralf Förster
2012-02-29 18:51:37 UTC
Weird it hsould get killed every time einfo "Starting Xvfb on \$DISPLAY=${XDISPLAY} ..." export DISPLAY=:${XDISPLAY} # Do not break on error, but setup $retval, as we need # to kill Xvfb debug-print "${FUNCNAME}: ${VIRTUALX_COMMAND} \"$@\"" ${VIRTUALX_COMMAND} "$@" retval=$? # Now kill Xvfb kill $(cat /tmp/.X${XDISPLAY}-lock) (In reply to comment #1) > Weird it hsould get killed every time > > > einfo "Starting Xvfb on \$DISPLAY=${XDISPLAY} ..." > > export DISPLAY=:${XDISPLAY} > # Do not break on error, but setup $retval, as we need > # to kill Xvfb > debug-print "${FUNCNAME}: ${VIRTUALX_COMMAND} \"$@\"" > ${VIRTUALX_COMMAND} "$@" > retval=$? > > # Now kill Xvfb > kill $(cat /tmp/.X${XDISPLAY}-lock) Yes, but it never reaches this line because ${VIRTUALX_COMMAND} calls die: """ cmake_src_test() { # fix OutDir test # this is altered thanks to our eclass sed -i -e 's:#IGNORE ::g' "${S}"/Tests/OutDir/CMakeLists.txt || die pushd "${CMAKE_BUILD_DIR}" > /dev/null local ctestargs [[ -n ${TEST_VERBOSE} ]] && ctestargs="--extra-verbose --output-on-failure" # Excluded tests: # BootstrapTest: we actualy bootstrap it every time so why test it. # SimpleCOnly_sdcc: sdcc choke on global cflags so just skip the test # as it was never intended to be used this way. "${CMAKE_BUILD_DIR}"/bin/ctest ${ctestargs} \ -E BootstrapTest SimpleCOnly_sdcc \ || die "Tests failed" """ ^^^ Here """ popd > /dev/null """ Created attachment 323008 [details, diff]
bug-406353.patch
I believe the attached patch solves the issue. Thoughts?
> I believe the attached patch solves the issue. Thoughts?
CCing x11, as this patch touches virtualx.eclass.
<johu> chithead: bug 406353 feedback on the patch? <willikins> johu: https://bugs.gentoo.org/406353 "dev-util/cmake-2.8.6-r4 - stop Xvfb after failed test phase"; Gentoo Linux, KDE; UNCO; toralf.foerster:kde <chithead> if you tested it to work, fine with me (In reply to comment #5) > <johu> chithead: bug 406353 feedback on the patch? > <willikins> johu: https://bugs.gentoo.org/406353 "dev-util/cmake-2.8.6-r4 - > stop Xvfb after failed test phase"; Gentoo Linux, KDE; UNCO; > toralf.foerster:kde > <chithead> if you tested it to work, fine with me tested with 2.8.8-r3, patch applied successfully, issue went away Thanks everyone, this patch is now applied in the main tree. This patch to virtualx.eclass broke building firefox for me. The firefox ebuild is EAPI 3 and portage gives an error that nonfatal is unavailable in that EAPI. virtualx.eclass claims to only require EAPI 2 or newer. I updated the eclass to avoid nonfatal when EAPI=2 or 3 is used. |