Merge lp://staging/~thomas-voss/dbus-cpp/enable-dual-landings into lp://staging/dbus-cpp

Proposed by Thomas Voß
Status: Merged
Approved by: Thomas Voß
Approved revision: 110
Merged at revision: 102
Proposed branch: lp://staging/~thomas-voss/dbus-cpp/enable-dual-landings
Merge into: lp://staging/dbus-cpp
Diff against target: 1472 lines (+519/-483)
32 files modified
CMakeLists.txt (+29/-4)
debian/VERSION (+1/-0)
debian/VERSION.vivid (+1/-0)
debian/bileto_pre_release_hook (+69/-0)
debian/changelog (+47/-0)
debian/control (+9/-3)
debian/control.in (+73/-0)
debian/get-versions.sh (+67/-0)
debian/libdbus-cpp5.symbols.32bit (+0/-198)
debian/libdbus-cpp5.symbols.64bit (+0/-199)
debian/libdbus-cpp5.symbols.amd64 (+0/-2)
debian/libdbus-cpp5.symbols.arm64 (+0/-2)
debian/libdbus-cpp5.symbols.armhf (+0/-2)
debian/libdbus-cpp5.symbols.coverage (+0/-5)
debian/libdbus-cpp5.symbols.i386 (+0/-2)
debian/libdbus-cpp5.symbols.powerpc (+0/-2)
debian/libdbus-cpp5.symbols.ppc64el (+0/-2)
debian/rules (+17/-1)
include/core/dbus/impl/object.h (+95/-34)
include/core/dbus/impl/property.h (+21/-0)
include/core/dbus/impl/signal.h (+8/-5)
include/core/dbus/message.h (+2/-0)
include/core/dbus/object.h (+8/-7)
include/core/dbus/property.h (+8/-0)
include/core/dbus/signal.h (+5/-2)
src/core/dbus/bus.cpp (+13/-1)
src/core/dbus/match_rule.cpp (+1/-8)
src/core/dbus/message.cpp (+19/-0)
src/core/dbus/service.cpp (+1/-1)
tests/async_execution_load_test.cpp (+1/-1)
tests/executor_test.cpp (+4/-2)
tests/message_test.cpp (+20/-0)
To merge this branch: bzr merge lp://staging/~thomas-voss/dbus-cpp/enable-dual-landings
Reviewer Review Type Date Requested Status
Łukasz Zemczak packaging Approve
PS Jenkins bot continuous-integration Needs Fixing
Review via email: mp+278251@code.staging.launchpad.net

Commit message

Enable dual landings for vivid and all later distributions (xenial at the time of this writing).

Description of the change

Enable dual landings for vivid and all later distributions (xenial at the time of this writing).

To post a comment you must log in.
105. By Thomas Voß

Add missing file.

Revision history for this message
PS Jenkins bot (ps-jenkins) wrote :
review: Needs Fixing (continuous-integration)
Revision history for this message
PS Jenkins bot (ps-jenkins) wrote :
review: Needs Fixing (continuous-integration)
Revision history for this message
Łukasz Zemczak (sil2100) wrote :

13:04 <sil2100> The only problem I see so far is with the versioning
13:05 <sil2100> Currently the train does not support dual-landing with two different
                versioning schemes
13:06 <sil2100> And when building for instance dbus-cpp with your script, the package
                version for both xenial and vivid will be 5.0.0, but the contents inside
                will have 4.3.0 for vivid and 5.0.0 for xenial
13:06 <sil2100> Meaning that the package version is lying and trying to tell
                vivid-overlay users that it's shipping a different version

Sadly, I'm not sure if this could be done from override_dh_auto_clean as the rest. For now the only idea I have is to support such cases in the CI Train, but that's just my first guess.

review: Needs Fixing (packaging)
106. By Thomas Voß

[ CI Train Bot ]
* debian/libdbus-cpp4.symbols.32bit: update to released version.
* debian/libdbus-cpp4.symbols.64bit: update to released version.
[ Thomas Voß ]
* Add a proper operator<< for dbus::Message::Type. (LP: #1524131)
* Only run dbus_shutdown if explicitly requested by env var. (LP:
  #1422304, #1526877)
[ CI Train Bot ]
* New rebuild forced.
[ Thomas Voß ]
* Ensure that Signal with non-void argument types correctly narrow
  their match rules. (LP: #1480877)

107. By Thomas Voß

Fix build on Y, accounting for toolchain bump.

108. By Thomas Voß

Switch to bileto_pre_release_hook

Revision history for this message
Łukasz Zemczak (sil2100) wrote :

One small inline-comment below. I was wondering where libdbus-cpp.vivid-soversion is used, as only libdbus-cpp.soversion seems to be parsed. I guess if we want to be correct, when DBUS_CPP_SOVERSION is defined the right version based on series should be used?

review: Needs Information
109. By Thomas Voß

Address reviewer comments.

110. By Thomas Voß

Make executor test more stable on slow builders.

Revision history for this message
Łukasz Zemczak (sil2100) wrote :

Ok, looks good now, just one really minor inline comment below. Of course the same thing remains: we need some abi compliance checking implemented when symbols files removal. This would need a bug assigned to someone before we can land that. ABI compliance will be even more important soon as we are starting to migrate things from universe to main for ubuntu-touch, and a library without symbol files is unlikely to go through the MIR process smoothly.

That being said, if my comments will be addressed (the inline one and the bug about ABI-checking), this merge is good to go.

review: Approve (packaging)

Preview Diff

[H/L] Next/Prev Comment, [J/K] Next/Prev File, [N/P] Next/Prev Hunk
The diff is not available at this time. You can reload the page or download it.

Subscribers

People subscribed via source and target branches