| Summary: | dev-qt/qtbluetooth-5.9.2 does not build bluez support even if its installed due to not detecting qtdbus | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Jan Boysen <trekkie> |
| Component: | Current packages | Assignee: | Qt Bug Alias <qt> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | wiselord1983 |
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | AMD64 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: | emerge --infp dump | ||
We already mangle qconfig.pri in the eclass, I guess we'll need to start manging qmodule.pri too. I can confirm the bug on Qt-5.9.3. The solution with modifying qmodule.pro works. I produced an initial patch set to address this: https://github.com/gentoo/qt/pull/165 The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/proj/qt.git/commit/?id=e553539e7ee050d094da9a4673b2e5ac71aae422 commit e553539e7ee050d094da9a4673b2e5ac71aae422 Author: Michael Palimaka <kensington@gentoo.org> AuthorDate: 2018-01-26 01:29:27 +0000 Commit: Michael Palimaka <kensington@gentoo.org> CommitDate: 2018-02-02 11:53:57 +0000 qt5-build.eclass: introduce for QT5_GENTOO_PRIVATE_CONFIG This behaves similarly to QT5_GENTOO_CONFIG, but for the new qmodule.pri instead. Bug: https://bugs.gentoo.org/634622 eclass/qt5-build.eclass | 116 +++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 106 insertions(+), 10 deletions(-)} The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=74d8dc2351328d2193fe3440600d1642f947db30 commit 74d8dc2351328d2193fe3440600d1642f947db30 Author: Michael Palimaka <kensington@gentoo.org> AuthorDate: 2018-01-26 01:29:27 +0000 Commit: Michael Palimaka <kensington@gentoo.org> CommitDate: 2018-02-03 13:36:57 +0000 qt5-build.eclass: introduce QT5_GENTOO_PRIVATE_CONFIG This behaves similarly to QT5_GENTOO_CONFIG, but for the new qmodule.pri instead. Bug: https://bugs.gentoo.org/634622 eclass/qt5-build.eclass | 116 +++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 106 insertions(+), 10 deletions(-)} This should be fixed in 5.9.4. |
Created attachment 499054 [details] emerge --infp dump It seems like QT5.9 introduced a new way of determining installed/configures QT components. dev-qt/qtcore is configured with -no-dbus as qtdbus is a seperate ebuild i guess. This results in QT.global_private.disabled_features in /usr/lib64/qt5/mkspecs/qmodule.pri contains dbus and dbus-linked even if dev-qt/qtdbus and dev-qt/qdbus are both installed. And in the end qtbluetooth only build dummy interface as it does not detect dbus support in QT. After I altered /usr/lib64/qt5/mkspecs/qmodule.pri and moved dbus from QT.global_private.disabled_features to QT.global_private.enabled_features qtbluetooth was fine and detected BlueZ and build its interface instead of a dummy interface.