Comment 93 for bug 1480877

Revision history for this message
Tony Espy (awe) wrote :

Here's a new version of the patch with all the access-point related signals removed.

It makes no sense for any of this logic to be present in our system, as none of our code attempts to control the network configuration via Bearer Mgmt. With this patch, the plugin will never receive any NM access-point related signals, and therefore it'll never populate it's access point list.

Also, as pointed out in other comments, the bearer plugin includes blocking DBus calls that are made in response to every single 'PropertiesChanged' signal sent for an access point, which causes a ripple of DBus traffic every time a scan occurs.

@Lorn, if you could look over my patch, I'd appreciate it.

It'd be also nice to get some testing from @faenil and @mterry before we push forward with landing plans. I've confirmed that the CPU spikes in the dbus daemon go away completely with the changes. That said, it'd be good to get some independent results.

Until we have a silo prep'd for Qt, you can just install the libqt5network5_5.4.1+dfsg-2ubuntu10~awe16_armhf.deb from my PPA:

https://launchpad.net/~awe/+archive/ubuntu/ppa/+packages

You also probably want to install dbus-cpp and location-services from:

https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-026

The status of the location-service silo says it needs a rebuild, but the packages in the PPA should be OK for testing. If you have any questions, ask before you install anything.