summaryrefslogtreecommitdiff
path: root/dev-qt/qttools/qttools-6.9999.ebuild
diff options
context:
space:
mode:
authorIonen Wolkens <ionen@gentoo.org>2024-03-11 18:21:38 -0400
committerIonen Wolkens <ionen@gentoo.org>2024-03-12 07:59:29 -0400
commitf8498f5c80f766ba74f6eead3519e7465682ab39 (patch)
tree82ed81b1ecac0935e4728f27b7e838049217e441 /dev-qt/qttools/qttools-6.9999.ebuild
parent06c632041fa1ba8fb7fcd7b74f22491b66044f27 (diff)
downloadgentoo-f8498f5c80f766ba74f6eead3519e7465682ab39.tar.gz
gentoo-f8498f5c80f766ba74f6eead3519e7465682ab39.tar.bz2
gentoo-f8498f5c80f766ba74f6eead3519e7465682ab39.zip
dev-qt/qtbase: tentatively enable unity build in live
Better known as jumbo-build from chromium/qtwebengine. This affects most dev-qt/* revdeps too, similarly to having it set in qt6-build.eclass. Build time reduction seems to be around 30% which can be significant for the bigger packages and slower machines. Not adding IUSE=jumbo-build given there should be little reason for a user to turn this off. It *can* make some sense on chromium due to the increase in memory usage, but that is rather insignificant on other packages and this has reasonable size defaults. Can still MYCMAKEARGS=-DQT_UNITY_BUILD=OFF if really must. Could potentially introduce build failures on bumps or some untested USE combinations depending on how often/well Qt upstream tests this option. Can temporarily disable it per-package when needed (e.g. qttools is known failing with USE=designer for one so far), not really worth worrying about getting these actually fixed asap unless it is qtbase given then we'd lose it for the whole stack unless move the option to qt6-build.eclass. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
Diffstat (limited to 'dev-qt/qttools/qttools-6.9999.ebuild')
0 files changed, 0 insertions, 0 deletions