Also fill the center of parsed kernel with 1. This shouldn't change the
behavior of the old backends since they will modify the center of the
kernels.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Delay focus updates until critical section. Rational is that focus
events might arrive when the focused window hasn't been managed by
compton, result in that that window not being focused.
This commit makes compton mark focus update events, and only update
focus in critical section, after we managed all the new windows.
Fixes#177
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
glGetIntegerv(GL_NUM_EXTENSIONS) is only available since GL4.
(However glGetStringi(GL_EXTENSIONS) is available since GL3)
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Remove more unused stuff.
Also removed session::xinerama_scrs, since all the information we need
is covered by xinerama_scr_regs. Convert uses of xinerama_scrs to use
xinerama_scr_regs.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
It used to be unmap when we receive its MapNotify, but now since we discard
events received before we grab X server, that event it lost. But it
turns out we can just unmap it when it's first created, no need to wait
for the MapNotify.
Partially fix#160
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
When a window is reparented to the same parent again, we should just
move it to the top of the stack, instead of add it again.
Also a slight refactor restack_win.
Fixes#164
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Window flags can be set whichever backend we use, we will ignore then in
paint_preprocess. free_paint can be called whichever backend we use,
they just do nothing if we use experimental backends.
Others should rather just check if backend_data is NULL.
Making the code slightly more understandable, also making the eventual
removal of these checks easier.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Damage could have already happened at the time we call fill_win on a new
window. It's too difficult, or impossible, for us the find out if that
happened. So just blindly mark window as damaged.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Make sure X grab server is finished before calling x_discard_events in
session_init. Otherwise x_discard_events won't be able to discard all of
the events that were sent before we grab the server.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
A window might be mapped before we had the chance to start managing it.
So check to see if we should map a window after we call fill_win on it.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Create a "critical section" in draw_callback by grabbing the X server.
Delay the managing of newly created windows until we enter the cricial section.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Also, during session_init, discard old events after we grab the server.
If we don't do that, old, out dated CreateNotify will cause us to add
the window with the same id multiple times. Here is an explanation of
how that could happen:
1. compton connects to X
2. a new window created by someone else, X send us a CreateNotify
3. compton is initializing, so it didn't handle the event.
4. compton pulls a list of all windows from X. This will include that
window created in (2)
5. compton starts to handle event. it will handle the CreateNotify sent
in (2), thus adds the same window twice.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Instead of reusing add_win_above with the id of the window on the top
of the stack, we create a new function that takes a list_node.
Fix a problem when the window on the top of the stack is destroyed. Because
add_win_above search for windows from the hash table, it won't find that window
and returns NULL.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Currently compton handles window creation event by immediately query all
the information it needs and create a window struct with that
information. However, this is prone to race conditions.
In the future, we want to react to window creation event by creating a
placeholder in the window stack, and only query window information in a
critical section where the X server is grabbed by us.
This commit split struct win into two struct, one as placeholder, the
other for holding actual window information.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously we do window image update _after_ the to_paint checks.
Therefore the window image might failed to update after we decided to
paint it.
Do window image update before the checks.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Create a shadow image for a mapped window when it's shadow become enabled.
For unmapped window, the shadow image is created during mapping. Also,
if the window has IMAGE_ERROR flag, it won't be render anyway. So shadow
image won't be created in that case either.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If a window is unmapped, then mapped to a different position on screen
before fade out is completed (when fading is not enabled, it has to be
mapped again in the same frame), the old window content is left on screen.
This is because when this happens, we didn't add the unmapping window to
damage.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Using opacity_t for opacity_old resulting in opacity being rounded to 0.
This is fine until the opacity reaches 0, where the opacity != opacity_old
check fails and the damage is not added.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If window is closed when fading is not enabled, the window might be destroyed
before we can add the window extent to damage.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
reg_ignore_valid marks the validity of the reg_ignore of the windows
BENEATH the current window. That is, set `w->reg_ignore_valid = false`
doesn't invalidate `w->reg_ignore`.
When destroying window `w`, we need to invalidate all reg_ignore beneath
`w`. So it is not enough to just set `w->next->reg_ignore_valid =
false`, we also need to invalidate `w->next->reg_ignore`.
Yes, this variable name is confusing. Will be changed in the future.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
When the screen is not redirected, don't update the root image because
the backend is not initialized in this case.
Fixes a crash.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
paint_preprocess calls redir_start if it decides to redirect the screen.
redir_start might change the state of the windows. For example, when the
image of the window fails to bind, redir_start sets the error flag. And
those state changes might exclude the window from being rendered. Thus,
paint_preprocess should be re-run to make sure the new states are
honoured.
Fixes a crash when window images fail to bind after redirecting.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously, when the backend doesn't implement root_change(), we will
deinit, and then init the backend again. However, we didn't free all the
images previously returned by the backend. There is no guarantee that
these images will be valid across deinit/init.
So, make sure they are freed.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Make sure before the mapping operation begins, the window is always in
UNUNMAPPED state. This makes it easier to reason about things, such as
resource availability, etc.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously we try to keep the image data until the rebind succeeds. This
way even if the rebind fails, we still have something to display.
But that triggers a NVIDIA driver bug. Basically you cannot have more
than one names (XIDs) of a window pixmap. NVIDIA driver doesn't like
that, and binding the pixmap with its aliases will fail. This can
sometimes happen if we name the new pixmap before freeing the old one during
rebind, and the pixmap doesn't actually change (i.e. the rebind is
unnecessary, could happen because of X event handling complications).
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Clear the STALE_IMAGE flag set by win_update_bounding_shape to avoid an
unnecessary image data invalidation.
Mainly to workaround a NVIDIA bug. For more detailed explanation, see
commit 8f67c6190c281955846589d8769385f30e7c5d23
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Instead queue up configure notification while the window is unmapped,
just update the window geometry when the window is mapped.
Simplify the logic a little bit.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Use a flag to mark whether the image needs update, and only update once
per frame.
Also refactor some common code into functions, and handle image errors
properly.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Setting glx-swap-method to value other than "undefined" and "buffer-age"
could potentially cause rendering problems. So remove them, the meaning
of the remaining options can be more precisely captured by "use-damage",
so create a new option under that name.
--glx-swap-method is deprecated in favor of the new option --use-damage.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
struct _glx_image_data needs to be freed even when refcount is not 0.
Also fixed a typo in glx_bind_pixmap.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Rectangle textures was used as a fallback when the driver doesn't
support non-power-of-two (NPOT) textures. Since we are using OpenGL 3.0 now,
which includes support for NPOT textures, we don't need this fallback
anymore.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
blur_textures are not initialized when blur is disabled, but we still
try to resize them in gl_resize.
Don't do that.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Choose the best vsync method for the user, instead of asking them to
frustratingly try every one of the options to see what works.
With this commit, the `vsync` option will take only a boolean value.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Refine the API of fill_rectangle and rename it to fill.
Extras:
Keep GL_BLEND enabled; Fixed some texture/fbo leaks
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Implement image processing in gl_compose using a frag shader.
gl_image_op is used to set flags corresponds to what processing is
needed for a texture.
Also implement proper reference counting for textures.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
We only ever invert color of or dim the entire image, so just remove the
ability to do region based inversion and dimming. Making implementing
image_op easier.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
This is basically just adapting old code to the new interface. A lot
of the functions are still stubs, but the basic stuffs is working.
What remains to be done:
* Implement gl_image_op. (It should do the operation lazily, only update
the flags on the texture. Actual processing will be delayed until composition.)
* Implement gl_copy. Now it just return the same image after incrementing the
refcount. It should actually copy the image data structure so it can have a
separate set of flags.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Because when the screen is unredirected, the window resource would have
already been freed. Fix a NULL pointer deref.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If monitor-repaint is not implemented by the backend, compton will
crash. This commit fix the crash by not doing monitor-repaint in that
case.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Rational: current workaround for Composite < 0.2 doesn't work with the
GLX backend, it also doesn't handle window border properly. Plus,
Composite 0.2 came out more than a decade ago, it's safe to assume
everyone has it.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Also remove xrender's prepare since it overwrite part of the screen
outside the damage region, causing rendering problems.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Pass the visible region of the image to the backend, so the backend
could optimize based on that information.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Handle backend initialization failure.
Also actually disable vsync when it's not supported in the new xrender
backend.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If PresentPixmap choose to use the Flip mode, we cannot reuse the same
buffer for rendering for the next frame. Add double buffering for that
case.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Don't use idle fence. There is no reliable & portable way to obtain a
fence Present will like. Intel driver wants a DRI3 fence, NVIDIA driver
doesn't support DRI3, and wants a XSync fence (I am not even sure, since
Present doesn't really work with the NVIDIA driver. And using the fence
on NVIDIA driver causes compton to use 100% CPU).
Just don't bother.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Window size change cause window resource to be released and re-acquired,
so it cannot work on unmapped windows.
Previously we make sure of that by skipping fading and change if the
window is mapped. Now we just make sure that functino is never called
with an unmapped window.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If a window is being faded out because of unmapping, don't handle its
configuration notify. Avoid freeing pixmap of a window that is being
faded out, because that will cause paint to fail.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
win_update_focused transition window into FADING without checking their
current state, resulting into unmapping windows being brought into
FADING state and stops unmapping.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously it is initialized by win_update_bounding_shape, so
configure_win is called when win_data == NULL and state ==
WSTATE_MAPPING, which is an invalid state for a `win` to be in.
Now map_win explicitly calls backend->prepare_win. This is not very
efficient since win_update_bounding_shape will reinit it.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
* Make some functions in win.c pure
* Remove unused code
* Fix misuse of attr_const, const functions shouldn't exam data pointed
to by its pointer arguments.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
When destroying a window, set window->next->reg_ignore_valid to false,
since its reg_ignore is not valid anymore.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Only case where you want to map window in add_win() is during compton
startup, when compton is registering existing windows. Otherwise,
add_win() is always called for newly created windows, so there will
always be a MapNotify coming up for that window. If we map newly created
windows in add_win(), we will try to map it a second time when the MapNotify
arrives.
So, just don't call map_win() from add_win(). For compton startup, we
explicitly call map_win() after calling add_win() in session_init.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If a window is destroyed and starts to fade out when screen is
unredirected, when the screen become redirected again, we will generate
errors when trying to render that window.
Also make sure the assumption that all window are either MAPPED or
UNMAPPED when screen is unredirected holds.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
paint_preprocess takes head of the window list as an argument. The
actual head of the window list might be freed and deleted during processing
of fade, but paint_preprocess will keep using the old head of the list,
thus uses freed memory.
Solution is just don't pass the head as an argument. paint_preprocess
will use session_t::list directly.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Now both the old and the new backends are compiled in, the user can
choose which one to use with a command line switch.
Lower the barrier for testing.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously if we unredirect the screen while a fading is in progress, we
will "resume" the fading when we redirect the screen again.
This is usually fine unless the window being faded out is destroyed. we
still tries to fade it out, but since we don't have the pixmap of the
window anymore (freed by unredirect), we will generate lots of errors
until the window is completely "faded out".
Also this change makes it easy to reason about things. Now we know when
the screen is unredirected, all the windows can either be in MAPPED or
UNMAPPED state, nothing else.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
After converting from struct conv to xorg format, cache the result to
save CPU time.
And remove an unused function.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Instead of storing them as an array of xfixed.
Might cause some performance overhead for the new backend, because
it is allocating a buffer to do the conversion every frame. Will fix
later.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Destruction of Input Only windows is erroneously ignored by commit
cea010a, causing future window with the same window id as the destroyed
Input Only window to be treated by an Input Only window even when they
are not. The end result is that some windows don't show up.
Also improved the comments in win.c.
Fixes#119
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
A side effect of this is that some window properties are allowed to
change during fade-in/out. (e.g. window background can become blurred
during window fade out if window properties changed so blur is requirired).
But it is unclear what is the expected behaviour in this case anyway, so we
choose the one that is simpler to implement.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Window state tracking is basically the back bond of window fading. We
add a new enum field to the win struct to track the state of the window,
instead of using a set of boolean variables. We also remove the fading
callbacks, since it is only used for fading, so the potential of code
reuse is lost. And it makes the code slightly harder to understand.
Also fixed a problem that --no-fading-openclose is not behaving as
advertised (from my observation, enabling this flag disables fading
entirely, instead of just diabling it for open/close).
Also uses double for opacity everywhere internally. Use opacity_t only
when setting X opacity prop.
TODO: Remove win::*_last
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Handle SIGINT using libev's ev_signal handler, instead of using
sigaction().
Fixes#64. But compton might miss signals when it's not in the main loop
(e.g. when compton is reset), and thus not exit "clearly". It should
cause any real problem though, so we don't care.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Put them all into one function, and move the function pointers out of
glx_session_t, making them global variables (because them don't change
after initialized). Remove the function pointer typedefs and replace
them with the ones in glxext.h
We also only lookup the functions once per a lifetime of compton.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
In X, background pixmap is tiled (meaning they are repeated to filled
the window) by default. So, in the glx backend, we mimic this behavior
by binding the background pixmap of the root window (aka, the wallpaper)
to texture that repeats.
Fixes#107
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
If fork is requested, we fork as early as possible, way before anything
is initialized. This way, we don't need to do the gymnastics to make
OpenGL work properly across fork.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Replace most of XFlush with xcb_flush.
Also, in a lot of places, XFlush is used as if it is XSync. Replace
those cases by using xcb's _checked version of functions and
xcb_request_check.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Background: To bind a Xorg window content to a OpenGL FBConfig, which
has to match the color format the Xorg window is using.
Previously, compton just find a FBConfig that has the same depth. This
led to chjj/compton#477, which has been fixed by a ugly hack.
The commit refactor the lookup mechanism to take as much into
consideration as we reasonably can. Hopefully preventing similar
breakages in the future.
Also, some code sharing between the old and new glx backend.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
As a global variable, since they shouldn't change during the period
compton is running. Also limit the scope of the variable to x.c.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously we were using glibc's strtod function to parse floating point
numbers. The problem is, strtod is locale dependent. Meaning 7,7 might
be parsed as two numbers (7 and 7) in one locale, and parsed as one
number (7 point 7) in another locale. This is undesirable.
We need to set the locale to a value we know to make number parsing
consistently. We could use setlocale(), but that is not thread-safe. We
can also use uselocale(), which is thread-safe, but doesn't cover strtod
(Yeah, some of the locale-aware functions only acknowledge the global
locale, not the thread local one).
So in frustration, I just wrote a simple floating point number parser
myself. This parser obviously doesn't cover all cases strtod covers, but
is good enough for our needs.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously the search order is:
~/.config/compton/compton.conf
/etc/xdg/compton/compton.conf
~/.config/compton.conf
/etc/xdg/compton.conf
...
Now the search order is:
~/.config/compton/compton.conf
~/.config/compton.conf
~/.compton.conf
/etc/xdg/compton/compton.conf
...
In other word, compton will now search all possible user config file
path first before searching for a system config file.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
glDrawBuffers doesn't take GL_BACK, which is what we are passing. And we
are using only one buffer argument anyway, no need to use glDrawBuffers.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Xorg uses top left as origin, OpenGL uses lower left. So we need to flip
the y axis, and make sure we are using the right points as origin of
windows.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Buffer age is not a glx specific concept. xrender backend can have
buffer age too, if double buffering is used (required if we want to use
Present). So, make buffer age a generic concept in compton is required
for further backend improvements.
Moved buffer age based damage aggragation out of glx as well.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Previously, compton fails to stop draw_idle in some cases when sw_opti
is enabled.
sw_opti is a feature that limits the draw frequence to vblank frequence.
It adds a delay to drawing when the screen is updated more frequently
than the vblank frequence. However when the delay is not used (i.e. the
screen is updated infrequent enough), compton will start drawing the
frame directly without using the delay. And specically in this case,
compton will fail to stop the draw_idle, causing a callback to be called
once per loop of the mainloop, resulting in high CPU usage.
Fixes#92
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Slightly clean up header inclusion with the help of clang's module system.
It's better for files to include the things you need directly.
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
glStringMarker is usually only available when running under some
kind of GL debugger, and can be used to insert strings into the
GL command string. Writing logs using it can be useful, since it
lets us correspond GL calls with what happens in compton.
More info about the extension can be found here:
https://www.khronos.org/registry/OpenGL/extensions/GREMEDY/GREMEDY_string_marker.txt
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
Yuxuan Shui