Created by Andreas Pokorny and last modified
Get this branch:
bzr branch lp:~andreas-pokorny/mir/treat-pause-like-focus-lost-and-resume-like-focus-gained
Only Andreas Pokorny can upload to this branch. If you are Andreas Pokorny please log in for upload directions.

Branch merges

Related bugs

Related blueprints

Branch information

Andreas Pokorny

Recent revisions

3454. By Andreas Pokorny on 2016-04-15

Make Nested Display handle focus change events as pause/resume indicator

3453. By Andreas Pokorny on 2016-04-14

Pause/Resume of display cause shell to remove focus

Clients should be aware that they will not receive further key input. So any
assumption on the current key state is about to be outdated. The server pauses
on vt switch - so the server removes all input devices and also looses any
knowledge of the key state. With this change the DisplayServer notifies the
shell about this. The AbstractShell reacts to that by removing the focus from
the current client, and on resume the same client receives focus again.

Upcoming changes should also include a key state event alongside the focus

3452. By Kevin DuBois on 2016-04-11

window management: allow the shell more control over input area management. The shell can now allow a surface to have out-of-surface_rect input regions (mostly for the shell's own purposes). Clients are still prevented from setting any input region they want by shell policy.

Approved by Alan Griffiths, Cemil Azizoglu, mir-ci-bot.

3451. By Andreas Pokorny on 2016-04-11

Disable key repeat on arale & mtk-tpd (LP: #1549701)

The arale touchscreen driver simulates a button. When the screen is turned off while the button is held and then released mir will not receive the release event or any notice that the device is off. To avoid that mir keeps on sending repeat events, repeats for the touchscreen are disabled. Fixes: https://bugs.launchpad.net/bugs/1549701.

Approved by mir-ci-bot, Alberto Aguirre, Alan Griffiths, Kevin DuBois, Cemil Azizoglu.

3450. By Daniel van Vugt on 2016-04-11

Improve readability of ClientLatency acceptance tests' source code.

Approved by Daniel van Vugt, mir-ci-bot, Cemil Azizoglu.

3449. By Kevin DuBois on 2016-04-11

android: add support for HWC 1.5 (introduced in android 6.0).

Approved by mir-ci-bot, Cemil Azizoglu.

3448. By Daniel van Vugt on 2016-04-11

acceptance-tests: Record ClientLatency in the right place.

The test increments the virtual vsync counter and records latency at the
same time. However the granularity of our measurements is whole frame
numbers, so recording latency before incrementing the frame number was
causing a latency of zero to be recorded, and N instead of N+1. This
was not correct because we aim for render times closer to zero than
one frame, and even if that's not achieved you should measure the
age of a frame from when it started getting rendered. So this makes
the recorded latency more accurate, shifting all the numbers up by
one. This way any frame rendered in the frame interval immediately
prior has a latency of one and not zero frames.

Approved by mir-ci-bot, Kevin DuBois.

3447. By Daniel van Vugt on 2016-04-11

Add some missing overrides in src/platforms/mesa/server/kms/*.h.

Approved by mir-ci-bot, Cemil Azizoglu.

3446. By Alan Griffiths on 2016-04-08

utils: A helpful launcher (mirrun) for starting mir clients.

Approved by Kevin DuBois, Cemil Azizoglu, Andreas Pokorny, mir-ci-bot.

3445. By Brandon Schaefer on 2016-04-07

Remove the usage of android input channel. Slowly removing all usage of the android input sender/receiver as well as the input channel.

Approved by Kevin DuBois, Cemil Azizoglu, Andreas Pokorny, mir-ci-bot.

Branch metadata

Branch format:
Branch format 7
Repository format:
Bazaar repository format 2a (needs bzr 1.16 or later)
Stacked on:
This branch contains Public information 
Everyone can see this information.