~andreas-pokorny/mir/key-repeater-on-timer-fd-utility : changes

~andreas-pokorny/mir/key-repeater-on-timer-fd-utility » Changes from revision 3758

From Revision 3758 to 3739

Rev	Summary	Authors	Date
3758	relative! instead of TFD_TIMER_ABSTIME relative! instead of TFD_TIMER_ABSTIME	Andreas Pokorny	7 years ago
3757	Add a simpler key repeater utility class for stub and evdev ... Add a simpler key repeater utility class for stub and evdev input platform The key repeater is meant to be integrated with the Dispatchable classes and will replace the key repeater that currently abuses the main loop in inject repeated keys inside the mirserver. It uses a new mircommon steady_timer_fd utility that wrapps CLOCK_MONOTONIC timer fds. This is a preparation step to move the repeat handling inside the input platforms, which will evade another mirserver event type fumbling location.	Andreas Pokorny	7 years ago
3756	Deprecate mir_input_device_state_event_device_pressed_keys f... Deprecate mir_input_device_state_event_device_pressed_keys for mir_input_device_state_event_device_pressed_keys_for_index. Approved by mir-ci-bot, Cemil Azizoglu, Andreas Pokorny, Alexandros Frantzis, Alan Griffiths.	Brandon Schaefer	7 years ago
3755	The deprecated function mir_connection_create_spec_for_toolt... The deprecated function mir_connection_create_spec_for_tooltip() failed to set the aux_rect. Reimplemented as a wrapper for mir_connection_create_spec_for_tip(). Fixes: https://bugs.launchpad.net/bugs/1632335. Approved by Brandon Schaefer, mir-ci-bot, Alexandros Frantzis.	Alan Griffiths	7 years ago
3754	Make typo::Renderer::Image safer. Make typo::Renderer::Image safer. Approved by mir-ci-bot, Alan Griffiths.	Daniel van Vugt	7 years ago
3753	Make touch event creation a single step Make touch event creation a single step This avoids another case in which we directly accessed the event type to update a cookie. This is also a preparation step to make event validator not depend on the internal structure of the input event types. Additionally this removes touch contact size from the construction of touch events as it was always pulled out of the major and minor axis sizes. Instead this changes adds pieces to revive the contact orientation support in those places that were touched by the change. Approved by mir-ci-bot, Alan Griffiths, Cemil Azizoglu, Alexandros Frantzis.	Andreas Pokorny	7 years ago
3752	client: add mir_buffer_stream_get_swap_interval and mir_buff... client: add mir_buffer_stream_get_swap_interval and mir_buffer_stream_set_swap_interval. Previously there was no way to set swapinterval on non-default streams. Approved by mir-ci-bot, Daniel van Vugt, Alan Griffiths.	Kevin DuBois	7 years ago
3751	Currently, in order to create an egl native window, we need ... Currently, in order to create an egl native window, we need an EGLNativeSurface. I.e. virtual std::shared_ptr<void> create_egl_native_window(EGLNativeSurface surface) = 0; Sometimes, as in the case of 'render surfaces', we do not have a native surface at the time when we want to create the native window. In that case, we should still be able to create it without one (i.e. with a nullptr), and "install" the native surface later, before it's needed. This is what this MP does. Approved by mir-ci-bot, Chris Halse Rogers.	Cemil Azizoglu	7 years ago
3750	mirvanity: Reset stats using the R key mirvanity: Reset stats using the R key Because sometimes you change displays or compositing method at runtime, and don't want to wait for the old measurements to expire from the ring buffer. Just clear them instantly with a key press. Approved by mir-ci-bot, Alexandros Frantzis.	Daniel van Vugt	7 years ago
3749	Rename namespace mir::examples::typo -> mir::typo Rename namespace mir::examples::typo -> mir::typo Because it's not an example. In fact, libtypo is designed to not depend on Mir-anything at all. It could even become an external project later (with a different name, because the current one is not unique enough). . Approved by mir-ci-bot, Cemil Azizoglu.	Daniel van Vugt	7 years ago
3748	Introducing backend plumbing for getting high precision fram... Introducing backend plumbing for getting high precision frame timing. Presently this is only complete for the android, kms and x11 graphics platforms. Nested and eglstream will come later. . Approved by Cemil Azizoglu, mir-ci-bot.	Daniel van Vugt	7 years ago
3747	Delete a header we don't use anymore. Fixes: https://bugs.la... Delete a header we don't use anymore. Fixes: https://bugs.launchpad.net/bugs/1556205. Approved by Daniel van Vugt, mir-ci-bot.	Alan Griffiths	7 years ago
3746	nested: optimize fullscreen client surfaces by passing their... nested: optimize fullscreen client surfaces by passing their buffers through the nested server to the host server without composing. (LP: #1262116). Fixes: https://bugs.launchpad.net/bugs/1262116. Approved by Daniel van Vugt, Cemil Azizoglu, mir-ci-bot.	Kevin DuBois	7 years ago
3745	mirclient-detail: Add a function to transform the locations ... mirclient-detail: Add a function to transform the locations stored in events This removes another case in which mirserver directly accessed the input event structure. Approved by Kevin DuBois, mir-ci-bot, Brandon Schaefer.	Andreas Pokorny	7 years ago
3744	Make BufferVault swap interval logic more clear, so it's now... Make BufferVault swap interval logic more clear, so it's now easier to fix/enhance/maintain. Previously the caller was required to blindly increase or decrease some invisible value which implicitly toggled between swap interval 0 or 1 behaviour. So that's fragile to begin with. Secondly, we have known issues with overallocation (LP: #1565596) and underallocation (https://code.launchpad.net/~kdub/mir/nested-passthrough/+merge/307315) in this area, so to resolve those the vault will need to know the intended usage pattern rather than being told an exact number of buffers, which is causing the over/under allocation problems. Approved by Kevin DuBois, mir-ci-bot.	Daniel van Vugt	7 years ago
3743	mesa client: suggest using DMA buf to map a MirBuffer to a G... mesa client: suggest using DMA buf to map a MirBuffer to a GL texture. GL mapping of MirBuffer's are needed in the case where a passthrough buffer needs to be mapped (eg, if a small window pops up over a fullscreen window, the fullscreen window has to then be GL-drawn, when it was previously passed-through). Approved by mir-ci-bot, Daniel van Vugt, Cemil Azizoglu, Alan Griffiths, Chris Halse Rogers.	Kevin DuBois	7 years ago
3742	Reduce exposure of MirEvent structures in two platforms and ... Reduce exposure of MirEvent structures in two platforms and mirserver. Approved by Cemil Azizoglu, mir-ci-bot, Alan Griffiths.	Andreas Pokorny	7 years ago
3741	Allow eglInitialize to indicate version 1.4 or above. Allow eglInitialize to indicate version 1.4 or above. Approved by mir-ci-bot, Daniel van Vugt, Chris Halse Rogers.	Andreas Pokorny	7 years ago
3740	compositor: remove mc::BufferStreamSurfaces and mc::BufferBu... compositor: remove mc::BufferStreamSurfaces and mc::BufferBundle. The first was the implementation of mc::BufferStream for server side queueing, which mostly wrapped the logic in the removed mc::BufferQueue. the latter was the interface for the BufferQueue helper class, that was used in a few integration-level tests. These tests have been ported to used mc::Stream. . Approved by mir-ci-bot, Chris Halse Rogers, Cemil Azizoglu.	Kevin DuBois	7 years ago
3739	nested: Set the MirClientHostConnection callbacks (thread-)s... nested: Set the MirClientHostConnection callbacks (thread-)safely (LP: #1628828). Fixes: https://bugs.launchpad.net/bugs/1628828. Approved by Daniel van Vugt, Chris Halse Rogers, mir-ci-bot, Kevin DuBois.	Alexandros Frantzis	7 years ago

Older »