winit-sonoma-fix/src/platform/linux/mod.rs

486 lines
14 KiB
Rust
Raw Normal View History

#![cfg(any(target_os = "linux", target_os = "dragonfly", target_os = "freebsd", target_os = "openbsd"))]
2015-04-24 17:51:23 +10:00
use std::collections::VecDeque;
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
use std::{env, mem};
use std::ffi::CStr;
use std::os::raw::*;
use std::sync::Arc;
2015-05-04 15:32:02 +10:00
use sctk::reexports::client::ConnectError;
2018-05-08 07:36:21 +10:00
// `std::os::raw::c_void` and `libc::c_void` are NOT interchangeable!
use libc;
2018-05-08 07:36:21 +10:00
use {
CreationError,
CursorState,
EventsLoopClosed,
Icon,
MouseCursor,
ControlFlow,
WindowAttributes,
};
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
use window::MonitorId as RootMonitorId;
2018-05-08 07:36:21 +10:00
use self::x11::{XConnection, XError};
2017-03-04 06:56:27 +11:00
use self::x11::ffi::XVisualInfo;
pub use self::x11::XNotSupported;
2017-03-04 06:56:27 +11:00
mod dlopen;
pub mod wayland;
pub mod x11;
2017-06-22 05:10:23 +10:00
/// Environment variable specifying which backend should be used on unix platform.
///
/// Legal values are x11 and wayland. If this variable is set only the named backend
/// will be tried by winit. If it is not set, winit will try to connect to a wayland connection,
/// and if it fails will fallback on x11.
///
/// If this variable is set with any other value, winit will panic.
const BACKEND_PREFERENCE_ENV_VAR: &str = "WINIT_UNIX_BACKEND";
#[derive(Clone, Default)]
pub struct PlatformSpecificWindowBuilderAttributes {
pub visual_infos: Option<XVisualInfo>,
pub screen_id: Option<i32>,
pub resize_increments: Option<(u32, u32)>,
pub base_size: Option<(u32, u32)>,
}
lazy_static!(
pub static ref X11_BACKEND: Result<Arc<XConnection>, XNotSupported> = {
XConnection::new(Some(x_error_callback)).map(Arc::new)
};
);
pub enum Window {
X(x11::Window),
2017-03-11 09:56:31 +11:00
Wayland(wayland::Window)
}
#[derive(Debug, Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)]
pub enum WindowId {
X(x11::WindowId),
Wayland(wayland::WindowId)
}
#[derive(Debug, Copy, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)]
pub enum DeviceId {
X(x11::DeviceId),
Wayland(wayland::DeviceId)
}
#[derive(Debug, Clone)]
pub enum MonitorId {
X(x11::MonitorId),
Wayland(wayland::MonitorId),
}
impl MonitorId {
#[inline]
pub fn get_name(&self) -> Option<String> {
match self {
&MonitorId::X(ref m) => m.get_name(),
&MonitorId::Wayland(ref m) => m.get_name(),
}
}
#[inline]
pub fn get_native_identifier(&self) -> u32 {
match self {
&MonitorId::X(ref m) => m.get_native_identifier(),
&MonitorId::Wayland(ref m) => m.get_native_identifier(),
}
}
#[inline]
pub fn get_dimensions(&self) -> (u32, u32) {
match self {
&MonitorId::X(ref m) => m.get_dimensions(),
&MonitorId::Wayland(ref m) => m.get_dimensions(),
}
}
Move fullscreen modes to not touch physical resolutions (#270) * Fix X11 screen resolution change using XrandR The previous XF86 resolution switching was broken and everything seems to have moved on to xrandr. Use that instead while cleaning up the code a bit as well. * Use XRandR for actual multiscreen support in X11 * Use actual monitor names in X11 * Get rid of ptr::read usage in X11 * Use a bog standard Vec instead of VecDeque * Get rid of the XRandR mode switching stuff Wayland has made the decision that apps shouldn't change screen resolutions and just take the screens as they've been setup. In the modern world where GPU scaling is cheap and LCD panels are scaling anyway it makes no sense to make "physical" resolution changes when software should be taking care of it. This massively simplifies the code and makes it easier to extend to more niche setups like MST and videowalls. * Rename fullscreen options to match new semantics * Implement XRandR 1.5 support * Get rid of the FullScreen enum Moving to just having two states None and Some(MonitorId) and then being able to set full screen in the current monitor with something like: window.set_fullscreen(Some(window.current_monitor())); * Implement Window::get_current_monitor() Do it by iterating over the available monitors and finding which has the biggest overlap with the window. For this MonitorId needs a new get_position() that needs to be implemented for all platforms. * Add unimplemented get_position() to all MonitorId * Make get_current_monitor() platform specific * Add unimplemented get_current_monitor() to all * Implement proper primary monitor selection in X11 * Shut up some warnings * Remove libxxf86vm package from travis Since we're no longer using XF86 there's no need to keep the package around for CI. * Don't use new struct syntax * Fix indentation * Adjust Android/iOS fullscreen/maximized On Android and iOS we can assume single screen apps that are already fullscreen and maximized so there are a few methods that are implemented by just returning a fixed value or not doing anything. * Mark OSX/Win fullscreen/maximized unimplemented()! These would be safe as no-ops but we should make it explicit so there is more of an incentive to actually implement them.
2017-09-07 18:33:46 +10:00
#[inline]
pub fn get_position(&self) -> (i32, i32) {
Move fullscreen modes to not touch physical resolutions (#270) * Fix X11 screen resolution change using XrandR The previous XF86 resolution switching was broken and everything seems to have moved on to xrandr. Use that instead while cleaning up the code a bit as well. * Use XRandR for actual multiscreen support in X11 * Use actual monitor names in X11 * Get rid of ptr::read usage in X11 * Use a bog standard Vec instead of VecDeque * Get rid of the XRandR mode switching stuff Wayland has made the decision that apps shouldn't change screen resolutions and just take the screens as they've been setup. In the modern world where GPU scaling is cheap and LCD panels are scaling anyway it makes no sense to make "physical" resolution changes when software should be taking care of it. This massively simplifies the code and makes it easier to extend to more niche setups like MST and videowalls. * Rename fullscreen options to match new semantics * Implement XRandR 1.5 support * Get rid of the FullScreen enum Moving to just having two states None and Some(MonitorId) and then being able to set full screen in the current monitor with something like: window.set_fullscreen(Some(window.current_monitor())); * Implement Window::get_current_monitor() Do it by iterating over the available monitors and finding which has the biggest overlap with the window. For this MonitorId needs a new get_position() that needs to be implemented for all platforms. * Add unimplemented get_position() to all MonitorId * Make get_current_monitor() platform specific * Add unimplemented get_current_monitor() to all * Implement proper primary monitor selection in X11 * Shut up some warnings * Remove libxxf86vm package from travis Since we're no longer using XF86 there's no need to keep the package around for CI. * Don't use new struct syntax * Fix indentation * Adjust Android/iOS fullscreen/maximized On Android and iOS we can assume single screen apps that are already fullscreen and maximized so there are a few methods that are implemented by just returning a fixed value or not doing anything. * Mark OSX/Win fullscreen/maximized unimplemented()! These would be safe as no-ops but we should make it explicit so there is more of an incentive to actually implement them.
2017-09-07 18:33:46 +10:00
match self {
&MonitorId::X(ref m) => m.get_position(),
&MonitorId::Wayland(ref m) => m.get_position(),
}
}
#[inline]
pub fn get_hidpi_factor(&self) -> f32 {
match self {
&MonitorId::X(ref m) => m.get_hidpi_factor(),
&MonitorId::Wayland(ref m) => m.get_hidpi_factor(),
}
}
}
impl Window {
#[inline]
2018-05-08 07:36:21 +10:00
pub fn new(
events_loop: &EventsLoop,
attribs: WindowAttributes,
pl_attribs: PlatformSpecificWindowBuilderAttributes,
) -> Result<Self, CreationError> {
match *events_loop {
2018-05-08 07:36:21 +10:00
EventsLoop::Wayland(ref events_loop) => {
wayland::Window::new(events_loop, attribs).map(Window::Wayland)
},
2018-05-08 07:36:21 +10:00
EventsLoop::X(ref events_loop) => {
x11::Window::new(events_loop, attribs, pl_attribs).map(Window::X)
},
}
}
#[inline]
pub fn id(&self) -> WindowId {
2017-03-04 20:48:44 +11:00
match self {
&Window::X(ref w) => WindowId::X(w.id()),
&Window::Wayland(ref w) => WindowId::Wayland(w.id())
2017-03-04 20:48:44 +11:00
}
}
#[inline]
pub fn set_title(&self, title: &str) {
match self {
&Window::X(ref w) => w.set_title(title),
&Window::Wayland(ref w) => w.set_title(title)
}
}
#[inline]
pub fn show(&self) {
match self {
&Window::X(ref w) => w.show(),
&Window::Wayland(ref w) => w.show()
}
}
#[inline]
pub fn hide(&self) {
match self {
&Window::X(ref w) => w.hide(),
&Window::Wayland(ref w) => w.hide()
}
}
#[inline]
pub fn get_position(&self) -> Option<(i32, i32)> {
match self {
&Window::X(ref w) => w.get_position(),
&Window::Wayland(ref w) => w.get_position()
}
}
#[inline]
pub fn get_inner_position(&self) -> Option<(i32, i32)> {
match self {
&Window::X(ref m) => m.get_inner_position(),
&Window::Wayland(ref m) => m.get_inner_position(),
}
}
#[inline]
pub fn set_position(&self, x: i32, y: i32) {
match self {
&Window::X(ref w) => w.set_position(x, y),
&Window::Wayland(ref w) => w.set_position(x, y)
}
}
#[inline]
pub fn get_inner_size(&self) -> Option<(u32, u32)> {
match self {
&Window::X(ref w) => w.get_inner_size(),
&Window::Wayland(ref w) => w.get_inner_size()
}
}
#[inline]
pub fn get_outer_size(&self) -> Option<(u32, u32)> {
match self {
&Window::X(ref w) => w.get_outer_size(),
&Window::Wayland(ref w) => w.get_outer_size()
}
}
#[inline]
pub fn set_inner_size(&self, x: u32, y: u32) {
match self {
&Window::X(ref w) => w.set_inner_size(x, y),
&Window::Wayland(ref w) => w.set_inner_size(x, y)
}
}
#[inline]
pub fn set_min_dimensions(&self, dimensions: Option<(u32, u32)>) {
match self {
&Window::X(ref w) => w.set_min_dimensions(dimensions),
&Window::Wayland(ref w) => w.set_min_dimensions(dimensions)
}
}
#[inline]
pub fn set_max_dimensions(&self, dimensions: Option<(u32, u32)>) {
match self {
&Window::X(ref w) => w.set_max_dimensions(dimensions),
&Window::Wayland(ref w) => w.set_max_dimensions(dimensions)
}
}
#[inline]
pub fn set_cursor(&self, cursor: MouseCursor) {
match self {
&Window::X(ref w) => w.set_cursor(cursor),
&Window::Wayland(ref w) => w.set_cursor(cursor)
}
}
#[inline]
pub fn set_cursor_state(&self, state: CursorState) -> Result<(), String> {
match self {
&Window::X(ref w) => w.set_cursor_state(state),
&Window::Wayland(ref w) => w.set_cursor_state(state)
}
}
#[inline]
pub fn hidpi_factor(&self) -> f32 {
match self {
&Window::X(ref w) => w.hidpi_factor(),
&Window::Wayland(ref w) => w.hidpi_factor()
}
}
#[inline]
pub fn set_cursor_position(&self, x: i32, y: i32) -> Result<(), ()> {
match self {
&Window::X(ref w) => w.set_cursor_position(x, y),
&Window::Wayland(ref w) => w.set_cursor_position(x, y)
}
}
#[inline]
pub fn platform_display(&self) -> *mut libc::c_void {
match self {
&Window::X(ref w) => w.platform_display(),
&Window::Wayland(ref w) => w.get_display().c_ptr() as *mut _
}
}
#[inline]
pub fn platform_window(&self) -> *mut libc::c_void {
match self {
&Window::X(ref w) => w.platform_window(),
&Window::Wayland(ref w) => w.get_surface().c_ptr() as *mut _
}
}
#[inline]
pub fn set_maximized(&self, maximized: bool) {
match self {
&Window::X(ref w) => w.set_maximized(maximized),
&Window::Wayland(ref w) => w.set_maximized(maximized),
}
}
#[inline]
Move fullscreen modes to not touch physical resolutions (#270) * Fix X11 screen resolution change using XrandR The previous XF86 resolution switching was broken and everything seems to have moved on to xrandr. Use that instead while cleaning up the code a bit as well. * Use XRandR for actual multiscreen support in X11 * Use actual monitor names in X11 * Get rid of ptr::read usage in X11 * Use a bog standard Vec instead of VecDeque * Get rid of the XRandR mode switching stuff Wayland has made the decision that apps shouldn't change screen resolutions and just take the screens as they've been setup. In the modern world where GPU scaling is cheap and LCD panels are scaling anyway it makes no sense to make "physical" resolution changes when software should be taking care of it. This massively simplifies the code and makes it easier to extend to more niche setups like MST and videowalls. * Rename fullscreen options to match new semantics * Implement XRandR 1.5 support * Get rid of the FullScreen enum Moving to just having two states None and Some(MonitorId) and then being able to set full screen in the current monitor with something like: window.set_fullscreen(Some(window.current_monitor())); * Implement Window::get_current_monitor() Do it by iterating over the available monitors and finding which has the biggest overlap with the window. For this MonitorId needs a new get_position() that needs to be implemented for all platforms. * Add unimplemented get_position() to all MonitorId * Make get_current_monitor() platform specific * Add unimplemented get_current_monitor() to all * Implement proper primary monitor selection in X11 * Shut up some warnings * Remove libxxf86vm package from travis Since we're no longer using XF86 there's no need to keep the package around for CI. * Don't use new struct syntax * Fix indentation * Adjust Android/iOS fullscreen/maximized On Android and iOS we can assume single screen apps that are already fullscreen and maximized so there are a few methods that are implemented by just returning a fixed value or not doing anything. * Mark OSX/Win fullscreen/maximized unimplemented()! These would be safe as no-ops but we should make it explicit so there is more of an incentive to actually implement them.
2017-09-07 18:33:46 +10:00
pub fn set_fullscreen(&self, monitor: Option<RootMonitorId>) {
match self {
Move fullscreen modes to not touch physical resolutions (#270) * Fix X11 screen resolution change using XrandR The previous XF86 resolution switching was broken and everything seems to have moved on to xrandr. Use that instead while cleaning up the code a bit as well. * Use XRandR for actual multiscreen support in X11 * Use actual monitor names in X11 * Get rid of ptr::read usage in X11 * Use a bog standard Vec instead of VecDeque * Get rid of the XRandR mode switching stuff Wayland has made the decision that apps shouldn't change screen resolutions and just take the screens as they've been setup. In the modern world where GPU scaling is cheap and LCD panels are scaling anyway it makes no sense to make "physical" resolution changes when software should be taking care of it. This massively simplifies the code and makes it easier to extend to more niche setups like MST and videowalls. * Rename fullscreen options to match new semantics * Implement XRandR 1.5 support * Get rid of the FullScreen enum Moving to just having two states None and Some(MonitorId) and then being able to set full screen in the current monitor with something like: window.set_fullscreen(Some(window.current_monitor())); * Implement Window::get_current_monitor() Do it by iterating over the available monitors and finding which has the biggest overlap with the window. For this MonitorId needs a new get_position() that needs to be implemented for all platforms. * Add unimplemented get_position() to all MonitorId * Make get_current_monitor() platform specific * Add unimplemented get_current_monitor() to all * Implement proper primary monitor selection in X11 * Shut up some warnings * Remove libxxf86vm package from travis Since we're no longer using XF86 there's no need to keep the package around for CI. * Don't use new struct syntax * Fix indentation * Adjust Android/iOS fullscreen/maximized On Android and iOS we can assume single screen apps that are already fullscreen and maximized so there are a few methods that are implemented by just returning a fixed value or not doing anything. * Mark OSX/Win fullscreen/maximized unimplemented()! These would be safe as no-ops but we should make it explicit so there is more of an incentive to actually implement them.
2017-09-07 18:33:46 +10:00
&Window::X(ref w) => w.set_fullscreen(monitor),
&Window::Wayland(ref w) => w.set_fullscreen(monitor)
}
}
#[inline]
pub fn set_decorations(&self, decorations: bool) {
match self {
&Window::X(ref w) => w.set_decorations(decorations),
&Window::Wayland(ref w) => w.set_decorations(decorations)
}
}
Move fullscreen modes to not touch physical resolutions (#270) * Fix X11 screen resolution change using XrandR The previous XF86 resolution switching was broken and everything seems to have moved on to xrandr. Use that instead while cleaning up the code a bit as well. * Use XRandR for actual multiscreen support in X11 * Use actual monitor names in X11 * Get rid of ptr::read usage in X11 * Use a bog standard Vec instead of VecDeque * Get rid of the XRandR mode switching stuff Wayland has made the decision that apps shouldn't change screen resolutions and just take the screens as they've been setup. In the modern world where GPU scaling is cheap and LCD panels are scaling anyway it makes no sense to make "physical" resolution changes when software should be taking care of it. This massively simplifies the code and makes it easier to extend to more niche setups like MST and videowalls. * Rename fullscreen options to match new semantics * Implement XRandR 1.5 support * Get rid of the FullScreen enum Moving to just having two states None and Some(MonitorId) and then being able to set full screen in the current monitor with something like: window.set_fullscreen(Some(window.current_monitor())); * Implement Window::get_current_monitor() Do it by iterating over the available monitors and finding which has the biggest overlap with the window. For this MonitorId needs a new get_position() that needs to be implemented for all platforms. * Add unimplemented get_position() to all MonitorId * Make get_current_monitor() platform specific * Add unimplemented get_current_monitor() to all * Implement proper primary monitor selection in X11 * Shut up some warnings * Remove libxxf86vm package from travis Since we're no longer using XF86 there's no need to keep the package around for CI. * Don't use new struct syntax * Fix indentation * Adjust Android/iOS fullscreen/maximized On Android and iOS we can assume single screen apps that are already fullscreen and maximized so there are a few methods that are implemented by just returning a fixed value or not doing anything. * Mark OSX/Win fullscreen/maximized unimplemented()! These would be safe as no-ops but we should make it explicit so there is more of an incentive to actually implement them.
2017-09-07 18:33:46 +10:00
2018-05-08 07:36:21 +10:00
#[inline]
pub fn set_window_icon(&self, window_icon: Option<Icon>) {
match self {
&Window::X(ref w) => w.set_window_icon(window_icon),
&Window::Wayland(_) => (),
}
}
Move fullscreen modes to not touch physical resolutions (#270) * Fix X11 screen resolution change using XrandR The previous XF86 resolution switching was broken and everything seems to have moved on to xrandr. Use that instead while cleaning up the code a bit as well. * Use XRandR for actual multiscreen support in X11 * Use actual monitor names in X11 * Get rid of ptr::read usage in X11 * Use a bog standard Vec instead of VecDeque * Get rid of the XRandR mode switching stuff Wayland has made the decision that apps shouldn't change screen resolutions and just take the screens as they've been setup. In the modern world where GPU scaling is cheap and LCD panels are scaling anyway it makes no sense to make "physical" resolution changes when software should be taking care of it. This massively simplifies the code and makes it easier to extend to more niche setups like MST and videowalls. * Rename fullscreen options to match new semantics * Implement XRandR 1.5 support * Get rid of the FullScreen enum Moving to just having two states None and Some(MonitorId) and then being able to set full screen in the current monitor with something like: window.set_fullscreen(Some(window.current_monitor())); * Implement Window::get_current_monitor() Do it by iterating over the available monitors and finding which has the biggest overlap with the window. For this MonitorId needs a new get_position() that needs to be implemented for all platforms. * Add unimplemented get_position() to all MonitorId * Make get_current_monitor() platform specific * Add unimplemented get_current_monitor() to all * Implement proper primary monitor selection in X11 * Shut up some warnings * Remove libxxf86vm package from travis Since we're no longer using XF86 there's no need to keep the package around for CI. * Don't use new struct syntax * Fix indentation * Adjust Android/iOS fullscreen/maximized On Android and iOS we can assume single screen apps that are already fullscreen and maximized so there are a few methods that are implemented by just returning a fixed value or not doing anything. * Mark OSX/Win fullscreen/maximized unimplemented()! These would be safe as no-ops but we should make it explicit so there is more of an incentive to actually implement them.
2017-09-07 18:33:46 +10:00
#[inline]
pub fn get_current_monitor(&self) -> RootMonitorId {
match self {
&Window::X(ref w) => RootMonitorId{inner: MonitorId::X(w.get_current_monitor())},
&Window::Wayland(ref w) => RootMonitorId{inner: MonitorId::Wayland(w.get_current_monitor())},
}
}
}
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
unsafe extern "C" fn x_error_callback(
display: *mut x11::ffi::Display,
event: *mut x11::ffi::XErrorEvent,
) -> c_int {
if let Ok(ref xconn) = *X11_BACKEND {
let mut buf: [c_char; 1024] = mem::uninitialized();
(xconn.xlib.XGetErrorText)(
display,
(*event).error_code as c_int,
buf.as_mut_ptr(),
buf.len() as c_int,
);
let description = CStr::from_ptr(buf.as_ptr()).to_string_lossy();
let error = XError {
description: description.into_owned(),
error_code: (*event).error_code,
request_code: (*event).request_code,
minor_code: (*event).minor_code,
};
eprintln!("[winit X11 error] {:#?}", error);
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
*xconn.latest_error.lock() = Some(error);
}
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
// Fun fact: this return value is completely ignored.
0
}
pub enum EventsLoop {
Wayland(wayland::EventsLoop),
X(x11::EventsLoop)
}
2017-10-26 05:03:57 +11:00
#[derive(Clone)]
pub enum EventsLoopProxy {
X(x11::EventsLoopProxy),
Wayland(wayland::EventsLoopProxy),
}
impl EventsLoop {
pub fn new() -> EventsLoop {
if let Ok(env_var) = env::var(BACKEND_PREFERENCE_ENV_VAR) {
match env_var.as_str() {
"x11" => {
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
// TODO: propagate
return EventsLoop::new_x11().expect("Failed to initialize X11 backend");
},
"wayland" => {
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
return EventsLoop::new_wayland()
.expect("Failed to initialize Wayland backend");
},
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
_ => panic!(
"Unknown environment variable value for {}, try one of `x11`,`wayland`",
BACKEND_PREFERENCE_ENV_VAR,
),
}
}
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
let wayland_err = match EventsLoop::new_wayland() {
Ok(event_loop) => return event_loop,
Err(err) => err,
};
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
let x11_err = match EventsLoop::new_x11() {
Ok(event_loop) => return event_loop,
Err(err) => err,
};
X11: General cleanup (#491) * X11: General cleanup This is almost entirely internal changes, and as usual, doesn't actually fix any problems people have complained about. - `XSetInputFocus` can't be called before the window is visible. This was previously handled by looping (with a sleep) and querying for the window's state until it was visible. Now we use `XIfEvent`, which blocks until we receive `VisibilityNotify`. Note that this can't be replaced with an `XSync` (I tried). - We now call `XSync` at the end of window creation and check for errors, assuring that broken windows are never returned. When creating invisible windows, this is the only time the output buffer is flushed during the entire window creation process (AFAIK). For visible windows, `XIfEvent` will generally flush, but window creation has overall been reduced to the minimum number of flushes. - `check_errors().expect()` has been a common pattern throughout the backend, but it seems that people (myself included) didn't make a distinction between using it after synchronous requests and asynchronous requests. Now we only use it after async requests if we flush first, though this still isn't correct (since the request likely hasn't been processed yet). The only real solution (besides forcing a sync *every time*) is to handle asynchronous errors *asynchronously*. For future work, I plan on adding logging, though I don't plan on actually *handling* those errors; that's more of something to hope for in the hypothetical async/await XCB paradise. - We now flush whenever it makes sense to. `util::Flusher` was added to force contributors to be aware of the output buffer. - `Window::get_position`, `Window::get_inner_position`, `Window::get_inner_size`, and `Window::get_outer_size` previously all required *several* round-trips. On my machine, it took an average of around 80µs. They've now been reduced to one round-trip each, which reduces my measurement to 16µs. This was accomplished simply by caching the frame extents, which are expensive to calculate (due to various queries and heuristics), but change infrequently and predictably. I still recommend that application developers use these methods sparingly and generally prefer storing the values from `Resized`/`Moved`, as that's zero overhead. - The above change enabled me to change the `Moved` event to supply window positions, rather than client area positions. Additionally, we no longer generate `Moved` for real (as in, not synthetic) `ConfigureNotify` events. Real `ConfigureNotify` events contain positions relative to the parent window, which are typically constant and useless. Since that position would be completely different from the root-relative positions supplied by synthetic `ConfigureNotify` events (which are the vast majority of them), that meant real `ConfigureNotify` events would *always* be detected as the position having changed, so the resultant `Moved` was multiple levels of misleading. In practice, this meant a garbage `Moved` would be sent every time the window was resized; now a resize has to actually change the window's position to be accompanied by `Moved`. - Every time we processed an `XI_Enter` event, we would leak 4 bytes via `util::query_pointer` (`XIQueryPointer`). `XIButtonState` contains a dynamically-allocated mask field which we weren't freeing. As this event occurs with fairly high frequency, long-running applications could easily accumulate substantial leaks. `util::PointerState::drop` now takes care of this. - The `util` module has been split up into several sub-modules, as it was getting rather lengthy. This accounts for a significant part of this diff, unfortunately. - Atoms are now cached. Xlib caches them too, so `XInternAtom` wouldn't typically be a round-trip anyway, but the added complexity is negligible. - Switched from `std::sync::Mutex` to `parking_lot::Mutex` (within this backend). There appears to be no downside to this, but if anyone finds one, this would be easy to revert. - The WM name and supported hints are now global to the application, and are updated upon `ReparentNotify`, which should detect when the WM was replaced (assuming a reparenting WM was involved, that is). Previously, these values were per-window and would never update, meaning replacing the WM could potentially lead to (admittedly very minor) problems. - The result of `Window2::create_empty_cursor` will now only be used if it actually succeeds. - `Window2::load_cursor` no longer re-allocates the cursor name. - `util::lookup_utf8` previously allocated a 16-byte buffer on the heap. Now it allocates a 1024-byte buffer on the stack, and falls back to dynamic allocation if the buffer is too small. This base buffer size is admittedly gratuitous, but less so if you're using IME. - `with_c_str` was finally removed. - Added `util::Format` enum to help prevent goofs when dealing with format arguments. - `util::get_property`, something I added way back in my first winit PR, only calculated offsets correctly for `util::Format::Char`. This was concealed by the accomodating buffer size, as it would be very rare for the offset to be needed; however, testing with a buffer size of 1, `util::Format::Long` would read from the same offset multiple times, and `util::Format::Short` would miss data. This function now works correctly for all formats, relying on the simple fact that the offset increases by the buffer size on each iteration. We also account for the extra byte that `XGetWindowProperty` allocates at the end of the buffer, and copy data from the buffer instead of moving it and taking ownership of the pointer. - Drag and drop now reliably works in release mode. This is presumably related to the `util::get_property` changes. - `util::change_property` now exists, which should make it easier to add features in the future. - The `EventsLoop` device map is no longer in a mutex. - `XConnection` now implements `Debug`. - Valgrind no longer complains about anything related to winit (with either the system allocator or jemalloc, though "not having valgrind complain about jemalloc" isn't something to strive for). * X11: Add better diagnostics when initialization fails * X11: Handle XIQueryDevice failure * X11: Use correct types in error handler
2018-05-03 23:15:49 +10:00
let err_string = format!(
r#"Failed to initialize any backend!
Wayland status: {:#?}
X11 status: {:#?}
"#,
wayland_err,
x11_err,
);
panic!(err_string);
}
pub fn new_wayland() -> Result<EventsLoop, ConnectError> {
wayland::EventsLoop::new()
.map(EventsLoop::Wayland)
}
pub fn new_x11() -> Result<EventsLoop, XNotSupported> {
match *X11_BACKEND {
Ok(ref x) => Ok(EventsLoop::X(x11::EventsLoop::new(x.clone()))),
Err(ref err) => Err(err.clone()),
}
}
#[inline]
pub fn get_available_monitors(&self) -> VecDeque<MonitorId> {
match *self {
EventsLoop::Wayland(ref evlp) => evlp.get_available_monitors()
.into_iter()
.map(MonitorId::Wayland)
.collect(),
EventsLoop::X(ref evlp) => x11::get_available_monitors(evlp.x_connection())
.into_iter()
.map(MonitorId::X)
.collect(),
}
}
#[inline]
pub fn get_primary_monitor(&self) -> MonitorId {
match *self {
EventsLoop::Wayland(ref evlp) => MonitorId::Wayland(evlp.get_primary_monitor()),
EventsLoop::X(ref evlp) => MonitorId::X(x11::get_primary_monitor(evlp.x_connection())),
}
}
pub fn create_proxy(&self) -> EventsLoopProxy {
match *self {
EventsLoop::Wayland(ref evlp) => EventsLoopProxy::Wayland(evlp.create_proxy()),
EventsLoop::X(ref evlp) => EventsLoopProxy::X(evlp.create_proxy()),
}
}
pub fn poll_events<F>(&mut self, callback: F)
where F: FnMut(::Event)
{
match *self {
EventsLoop::Wayland(ref mut evlp) => evlp.poll_events(callback),
EventsLoop::X(ref mut evlp) => evlp.poll_events(callback)
}
}
pub fn run_forever<F>(&mut self, callback: F)
where F: FnMut(::Event) -> ControlFlow
{
match *self {
EventsLoop::Wayland(ref mut evlp) => evlp.run_forever(callback),
EventsLoop::X(ref mut evlp) => evlp.run_forever(callback)
}
}
#[inline]
pub fn is_wayland(&self) -> bool {
match *self {
EventsLoop::Wayland(_) => true,
EventsLoop::X(_) => false,
}
}
#[inline]
pub fn x_connection(&self) -> Option<&Arc<XConnection>> {
match *self {
EventsLoop::Wayland(_) => None,
EventsLoop::X(ref ev) => Some(ev.x_connection()),
}
}
}
impl EventsLoopProxy {
pub fn wakeup(&self) -> Result<(), EventsLoopClosed> {
match *self {
EventsLoopProxy::Wayland(ref proxy) => proxy.wakeup(),
EventsLoopProxy::X(ref proxy) => proxy.wakeup(),
}
}
}