if we keep a list of e clients... either make sure any client deleted
is removed from the list on deletion... OR use ref/unref to ensure it
stays alive as long as it's in the list.
This reverts commit 91b3f2e0e1.
revert wars part 4: the blizzard blitz!
the main point of freezing idlers here was not, in fact, to optimize, but to block an infinite loop which pegged the cpu until screensaver ended. this solution should be less issue-prone for the one person who had issues with the previous fix.
This reverts commit 3067f600ee.
revert wars! - i keep hitting problems - the one i still see is that i
come back to a machine that has blanked for a while - i launch some
app (terminology, sylpheed, chromium - doesn't matter) and no window
appears. psorcess is running. no matter how many times i launch it ...
no new process appears. this is a major bug. stopping the idler is an
optimization not a necessity.
so since this e main idler freeze/thaw i've noticed several times, i
come back to my machine after screen-off time period, i wke it up with
a mouse wiggle or keyboard press and try run terminology - no windwo
appears. i can run it all i want - it never shows up but the processes
are there. i've seen it happen to sylpheed where its fetch window
doesn't appear. i've had myserious menu edje objects on the top-left
with only a single item with no bg. i've had e even unable to restart
on ctrl+alt+end.
so i disabled the idler freeze/thaw as i suspected this is what the
root cause is, and sinc ethen the above problems stopped manifesting.
i can only conclude it's a deep and nasty bi-product of stopping the e
main idler, so don't do it. :) better be a bit less efficient than
buggy. either way setting manual rendering and dropping the animator
framerate should do almost all the things needed anyway.
in some cases, it may be desirable to block shape cutting for an amount of time to reduce cpu calculations. this should be set when performing any non-interactive compositor effects
variable takes the name of the output module s/wl_// ; example: E_WL_FORCE=x11
this module allows easy loading of a wayland output module to override any detection which may occur by force-loading the module during startup
in a later commit, an actual output detection mechanism needs to be added to run through all the possible output modes
* try to clear up build system for separating out ecore-x
* add #ifdefs for lots of ecore-x stuff
* break out some internal e wl functions for reuse in api
* store wl surface buffers as an inlist
* add protocol-specific client compositor data
** move lots of X client attributes here
* add pixmap type checks to a number of X-specific things, such as grabinput, to block them for non-X clients
* rearrange startup order to work with wayland
* move X screensaver code to e_comp_x
* flag modules still requiring X with -DNEED_X
given that compositing is always enabled and that it's crucial to be able to change these settings, potentially having them unavailable is not a great way to live
huge fustercluck commit because there wasn't really a way to separate out the changes. better to just rip it all out at once.
* compositor and window management completely rewritten. this was the goal for E19, but it pretty much required everything existing to be scrapped since it wasn't optimized, streamlined, or sensible. now instead of having the compositor strapped to the window manager like an outboard motor, it's housed more like an automobile engine.
** various comp structs have been merged into other places (eg. E_Comp_Zone is now just part of E_Zone where applicable), leading to a large deduplication of attributes
** awful E_Comp_Win is totally dead, having been replaced with e_comp_object smart objects which work just like normal canvas objects
** protocol-specific window management and compositor functionality is now kept exclusively in backend files
** e_pixmap api provides generic client finding and rendering api
** screen/xinerama screens are now provided directly by compositor on startup and re-set on change
** e_comp_render_update finally replaced with eina_tiler
** wayland compositor no longer creates X windows
** compositor e_layout removed entirely
* e_container is gone. this was made unnecessary in E18, but I kept it to avoid having too much code churn in one release. its sole purpose was to catch some events and handle window stacking, both of which are now just done by the compositor infra
* e_manager is just for screensaver and keybind stuff now, possibly remove later?
* e_border is gone along with a lot of its api. e_client has replaced it, and e_client has been rewritten completely; some parts may be similar, but the design now relies upon having a functional compositor
** window configuration/focus functions are all removed. all windows are now managed solely with evas_object_X functions on the "frame" member of a client, just as any other canvas object can be managed.
*** do NOT set interceptors on a client's comp_object. seriously.
* startup order rewritten: compositor now starts much earlier, other things just use attrs and members of the compositor
* ecore_x_pointer_xy_get usage replaced with ecore_evas_pointer_xy_get
* e_popup is totally gone, existing usage replaced by e_comp_object_util_add where applicable, otherwise just placed normally on the canvas
* deskmirror is (more) broken for now
* illume is totally fucked
* Ecore_X_Window replaced with Ecore_Window in most cases
* edge binding XWindows replaced with regular canvas objects
* some E_Win functionality has changed such that delete callbacks are now correctly called in ALL cases. various dialogs have been updated to not crash as a result
comp files and descriptions:
e_comp.c - overall compositor functions, rendering/update loop, shape cutting
e_comp_x.c - X window management and compositor functionality
e_comp_wl.c - Wayland surface management and compositor functionality
e_comp_canvas.c - general compositor canvas functions and utilities
e_comp_object.c - E_Client->frame member for managing clients as Evas_Objects, utility functions for adding objects to the compositor rendering systems
additional authors: ivan.briano@intel.com
feature: new compositor
removal: e_border, e_container, e_popup
so i found something interesting. the grab option causes some erratic
framerate in clients IF thoise clients have to sync to the xserver and
themselves try and render more smoothly. it masically means animators
get delayed due to unusually long render times because of this. this
fixes that. this also adds a comp config version number allowing the
code to upgrade a users config to turn this off and thus allows for
future upgrades too.
without this, it was possible to break the compositor by enabling options which prevent unfullscreening a window on vdesk flip, which causes nocomp mode to remain permanently enabled, even after the window has closed
shelves are almost always around. that means that unless you magically
can do partial rendering, comp updates re-render shelf AND all
contents. for systems this means much mofre gl overhead thatn with
e17. to do a mid-way house, force zoomaps on for shelves, until evas
can auto stuff smart objects into buffers on its own based on behavior.
this is currently GUARANTEED to break with override shaped windows, which means that if such a window is ever opened, it will NEVER receive input unless it is placed over another client window. this is due to a limitation in evas where it is impossible to (de)select input regions which forced me to choose between either having shaped override windows (such as desktop recorders) block all mouse input to the screen OR have normal override windows (gtk menus and popups...) not able to generate their own input shape at all
initial border pixmaps are now available for windows on other desktops immediately after startup, which should:
* improve initial redraw on first desk flip
* make comp mirrors available immediately for non-current desk windows
** fixes ibar menu instances
** fixes deskmirror redraws
* move all shape rectangle stuff into e_container_shape
* use input rects for cutting comp shape when available
* set shape changed and render queue when container shape change callbacks are called
* create fewer comp wins (small memory improvement)
it may happen that a container layer window cannot be found (someone
deleted it other than e - which is possible), and e hasn't expected
this. comp will have removed the comp win entry but containers will
not know. this is either some client being stupid, or malicious, but
protect against it, complain with an ERR and march on so we don't crash.
by making a job to delay population we play a race condition game with
initting of other bits of e. this means the comp may or may not have
been filled by the container layer winds before other internal events
and actions start making comp wins and setting layers. this means that
things like e_comp_win_find() for a con->layers[n].win may fail... as
this has yet to be added... and thus.. we crash in startup. it's a
race condition.
so this removed the delay job and makes population of the comp
explicit. it puts the population into the hnds of the e_main init
after a manager has been created and after a container is created, and
thus laer wins created, so it can walk the x window stack correctly,
but before anything else is done. this means that we will be properly
inited at this point without a race condition.
Mike Blumenkrantz
Carsten Haitzler
Mike Blumenkrantz
discomfitor
Mike Blumenkrantz
Chris Michael
Chris Michael
discomfitor
Jérémy Zurcher
Igor Murzov