// This file is part of river, a dynamic tiling wayland compositor. // // Copyright 2020 The River Developers // // This program is free software: you can redistribute it and/or modify // it under the terms of the GNU General Public License as published by // the Free Software Foundation, version 3. // // This program is distributed in the hope that it will be useful, // but WITHOUT ANY WARRANTY; without even the implied warranty of // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the // GNU General Public License for more details. // // You should have received a copy of the GNU General Public License // along with this program. If not, see . const Self = @This(); const build_options = @import("build_options"); const std = @import("std"); const assert = std.debug.assert; const mem = std.mem; const wlr = @import("wlroots"); const wl = @import("wayland").server.wl; const zwlr = @import("wayland").server.zwlr; const server = &@import("main.zig").server; const util = @import("util.zig"); const DragIcon = @import("DragIcon.zig"); const LayerSurface = @import("LayerSurface.zig"); const LockSurface = @import("LockSurface.zig"); const Output = @import("Output.zig"); const SceneNodeData = @import("SceneNodeData.zig"); const View = @import("View.zig"); const XwaylandOverrideRedirect = @import("XwaylandOverrideRedirect.zig"); scene: *wlr.Scene, /// All windows, status bars, drowdown menus, etc. that can recieve pointer events and similar. interactive_content: *wlr.SceneTree, /// Drag icons, which cannot recieve e.g. pointer events and are therefore kept in a separate tree. drag_icons: *wlr.SceneTree, /// All direct children of the interactive_content scene node layers: struct { /// Parent tree for output trees which have their position updated when /// outputs are moved in the layout. outputs: *wlr.SceneTree, /// Xwayland override redirect windows are a legacy wart that decide where /// to place themselves in layout coordinates. Unfortunately this is how /// X11 decided to make dropdown menus and the like possible. xwayland_override_redirect: if (build_options.xwayland) *wlr.SceneTree else void, }, /// This is kind of like an imaginary output where views start and end their life. /// It is also used to store views and tags when no actual outputs are available. hidden: struct { /// This tree is always disabled. tree: *wlr.SceneTree, tags: u32 = 1 << 0, pending: struct { focus_stack: wl.list.Head(View, .pending_focus_stack_link), wm_stack: wl.list.Head(View, .pending_wm_stack_link), }, inflight: struct { focus_stack: wl.list.Head(View, .inflight_focus_stack_link), wm_stack: wl.list.Head(View, .inflight_wm_stack_link), }, }, new_output: wl.Listener(*wlr.Output) = wl.Listener(*wlr.Output).init(handleNewOutput), output_layout: *wlr.OutputLayout, layout_change: wl.Listener(*wlr.OutputLayout) = wl.Listener(*wlr.OutputLayout).init(handleLayoutChange), output_manager: *wlr.OutputManagerV1, manager_apply: wl.Listener(*wlr.OutputConfigurationV1) = wl.Listener(*wlr.OutputConfigurationV1).init(handleManagerApply), manager_test: wl.Listener(*wlr.OutputConfigurationV1) = wl.Listener(*wlr.OutputConfigurationV1).init(handleManagerTest), power_manager: *wlr.OutputPowerManagerV1, power_manager_set_mode: wl.Listener(*wlr.OutputPowerManagerV1.event.SetMode) = wl.Listener(*wlr.OutputPowerManagerV1.event.SetMode).init(handlePowerManagerSetMode), /// A list of all outputs all_outputs: std.TailQueue(*Output) = .{}, /// A list of all active outputs. See Output.active outputs: std.TailQueue(Output) = .{}, /// Number of layout demands before sending configures to clients. inflight_layout_demands: u32 = 0, /// Number of inflight configures sent in the current transaction. inflight_configures: u32 = 0, transaction_timeout: *wl.EventSource, /// Set to true if applyPending() is called while a transaction is inflight. /// If true when a transaction completes will cause applyPending() to be called again. pending_state_dirty: bool = false, pub fn init(self: *Self) !void { const output_layout = try wlr.OutputLayout.create(); errdefer output_layout.destroy(); const scene = try wlr.Scene.create(); errdefer scene.tree.node.destroy(); const interactive_content = try scene.tree.createSceneTree(); const drag_icons = try scene.tree.createSceneTree(); const hidden_tree = try scene.tree.createSceneTree(); hidden_tree.node.setEnabled(false); const outputs = try interactive_content.createSceneTree(); const xwayland_override_redirect = if (build_options.xwayland) try interactive_content.createSceneTree(); try scene.attachOutputLayout(output_layout); _ = try wlr.XdgOutputManagerV1.create(server.wl_server, output_layout); const event_loop = server.wl_server.getEventLoop(); const transaction_timeout = try event_loop.addTimer(*Self, handleTransactionTimeout, self); errdefer transaction_timeout.remove(); self.* = .{ .scene = scene, .interactive_content = interactive_content, .drag_icons = drag_icons, .layers = .{ .outputs = outputs, .xwayland_override_redirect = xwayland_override_redirect, }, .hidden = .{ .tree = hidden_tree, .pending = .{ .focus_stack = undefined, .wm_stack = undefined, }, .inflight = .{ .focus_stack = undefined, .wm_stack = undefined, }, }, .output_layout = output_layout, .output_manager = try wlr.OutputManagerV1.create(server.wl_server), .power_manager = try wlr.OutputPowerManagerV1.create(server.wl_server), .transaction_timeout = transaction_timeout, }; self.hidden.pending.focus_stack.init(); self.hidden.pending.wm_stack.init(); self.hidden.inflight.focus_stack.init(); self.hidden.inflight.wm_stack.init(); server.backend.events.new_output.add(&self.new_output); self.output_manager.events.apply.add(&self.manager_apply); self.output_manager.events.@"test".add(&self.manager_test); self.output_layout.events.change.add(&self.layout_change); self.power_manager.events.set_mode.add(&self.power_manager_set_mode); } pub fn deinit(self: *Self) void { self.scene.tree.node.destroy(); self.output_layout.destroy(); self.transaction_timeout.remove(); } pub const AtResult = struct { surface: ?*wlr.Surface, sx: f64, sy: f64, node: union(enum) { view: *View, layer_surface: *LayerSurface, lock_surface: *LockSurface, xwayland_override_redirect: if (build_options.xwayland) *XwaylandOverrideRedirect else noreturn, }, }; /// Return information about what is currently rendered in the interactive_content /// tree at the given layout coordinates, taking surface input regions into account. pub fn at(self: Self, lx: f64, ly: f64) ?AtResult { var sx: f64 = undefined; var sy: f64 = undefined; const node_at = self.interactive_content.node.at(lx, ly, &sx, &sy) orelse return null; const surface: ?*wlr.Surface = blk: { if (node_at.type == .buffer) { const scene_buffer = wlr.SceneBuffer.fromNode(node_at); if (wlr.SceneSurface.fromBuffer(scene_buffer)) |scene_surface| { break :blk scene_surface.surface; } } break :blk null; }; if (SceneNodeData.get(node_at)) |scene_node_data| { return .{ .surface = surface, .sx = sx, .sy = sy, .node = switch (scene_node_data.data) { .view => |view| .{ .view = view }, .layer_surface => |layer_surface| .{ .layer_surface = layer_surface }, .lock_surface => |lock_surface| .{ .lock_surface = lock_surface }, .xwayland_override_redirect => |xwayland_override_redirect| .{ .xwayland_override_redirect = xwayland_override_redirect, }, }, }; } else { return null; } } fn handleNewOutput(_: *wl.Listener(*wlr.Output), wlr_output: *wlr.Output) void { const log = std.log.scoped(.output_manager); log.debug("new output {s}", .{wlr_output.name}); Output.create(wlr_output) catch |err| { switch (err) { error.OutOfMemory => log.err("out of memory", .{}), error.InitRenderFailed => log.err("failed to initialize renderer for output {s}", .{wlr_output.name}), } wlr_output.destroy(); }; } /// Remove the output from self.outputs and evacuate views if it is a member of /// the list. The node is not freed pub fn removeOutput(root: *Self, output: *Output) void { { const node = @fieldParentPtr(std.TailQueue(Output).Node, "data", output); // If the node has already been removed, do nothing var output_it = root.outputs.first; while (output_it) |n| : (output_it = n.next) { if (n == node) break; } else return; root.outputs.remove(node); } if (output.inflight.layout_demand) |layout_demand| { layout_demand.deinit(); output.inflight.layout_demand = null; } while (output.layouts.first) |node| node.data.destroy(); { var it = output.inflight.focus_stack.iterator(.forward); while (it.next()) |view| { view.inflight.output = null; view.current.output = null; view.tree.node.reparent(root.hidden.tree); view.popup_tree.node.reparent(root.hidden.tree); } root.hidden.inflight.focus_stack.prependList(&output.inflight.focus_stack); root.hidden.inflight.wm_stack.prependList(&output.inflight.wm_stack); } // Use the first output in the list as fallback. If the last real output // is being removed store the views in Root.hidden. const fallback_output = if (root.outputs.first) |node| &node.data else null; if (fallback_output) |fallback| { var it = output.pending.focus_stack.safeIterator(.reverse); while (it.next()) |view| view.setPendingOutput(fallback); } else { var it = output.pending.focus_stack.iterator(.forward); while (it.next()) |view| view.pending.output = null; root.hidden.pending.focus_stack.prependList(&output.pending.focus_stack); root.hidden.pending.wm_stack.prependList(&output.pending.wm_stack); // Store the focused output tags if we are hotplugged down to // 0 real outputs so they can be restored on gaining a new output. root.hidden.tags = output.pending.tags; } // Close all layer surfaces on the removed output for ([_]zwlr.LayerShellV1.Layer{ .overlay, .top, .bottom, .background }) |layer| { const tree = output.layerSurfaceTree(layer); var it = tree.children.safeIterator(.forward); while (it.next()) |scene_node| { assert(scene_node.type == .tree); if (@intToPtr(?*SceneNodeData, scene_node.data)) |node_data| { node_data.data.layer_surface.wlr_layer_surface.destroy(); } } } // If any seat has the removed output focused, focus the fallback one var seat_it = server.input_manager.seats.first; while (seat_it) |seat_node| : (seat_it = seat_node.next) { const seat = &seat_node.data; if (seat.focused_output == output) { seat.focusOutput(fallback_output); } } output.status.deinit(); root.applyPending(); } /// Add the output to self.outputs and the output layout if it has not /// already been added. pub fn addOutput(root: *Self, output: *Output) void { const node = @fieldParentPtr(std.TailQueue(Output).Node, "data", output); // If we have already added the output, do nothing and return var output_it = root.outputs.first; while (output_it) |n| : (output_it = n.next) if (n == node) return; root.outputs.append(node); // This arranges outputs from left-to-right in the order they appear. The // wlr-output-management protocol may be used to modify this arrangement. // This also creates a wl_output global which is advertised to clients. root.output_layout.addAuto(output.wlr_output); const layout_output = root.output_layout.get(output.wlr_output).?; output.tree.node.setEnabled(true); output.tree.node.setPosition(layout_output.x, layout_output.y); // If we previously had no outputs move all views to the new output and focus it. if (root.outputs.len == 1) { output.pending.tags = root.hidden.tags; { var it = root.hidden.pending.focus_stack.safeIterator(.reverse); while (it.next()) |view| view.setPendingOutput(output); assert(root.hidden.pending.focus_stack.empty()); assert(root.hidden.pending.wm_stack.empty()); assert(root.hidden.inflight.focus_stack.empty()); assert(root.hidden.inflight.wm_stack.empty()); } { // Focus the new output with all seats var it = server.input_manager.seats.first; while (it) |seat_node| : (it = seat_node.next) { const seat = &seat_node.data; seat.focusOutput(output); } } root.applyPending(); } } /// Trigger asynchronous application of pending state for all outputs and views. /// Changes will not be applied to the scene graph until the layout generator /// generates a new layout for all outputs and all affected clients ack a /// configure and commit a new buffer. pub fn applyPending(root: *Self) void { { // Changes to the pending state may require a focus update to keep // state consistent. Instead of having focus(null) calls spread all // around the codebase and risk forgetting one, always ensure focus // state is synchronized here. var it = server.input_manager.seats.first; while (it) |node| : (it = node.next) node.data.focus(null); } // If there is already a transaction inflight, wait until it completes. if (root.inflight_layout_demands > 0 or root.inflight_configures > 0) { root.pending_state_dirty = true; return; } root.pending_state_dirty = false; { var it = root.hidden.pending.focus_stack.iterator(.forward); while (it.next()) |view| { assert(view.pending.output == null); view.inflight.output = null; view.inflight_focus_stack_link.remove(); root.hidden.inflight.focus_stack.append(view); } } { var it = root.hidden.pending.wm_stack.iterator(.forward); while (it.next()) |view| { view.inflight_wm_stack_link.remove(); root.hidden.inflight.wm_stack.append(view); } } { var output_it = root.outputs.first; while (output_it) |node| : (output_it = node.next) { const output = &node.data; // Iterate the focus stack in order to ensure the currently focused/most // recently focused view that requests fullscreen is given fullscreen. var fullscreen_found = false; { var it = output.pending.focus_stack.iterator(.forward); while (it.next()) |view| { assert(view.pending.output == output); if (view.current.float and !view.pending.float) { // If switching from float to non-float, save the dimensions. view.float_box = view.current.box; } else if (!view.current.float and view.pending.float) { // If switching from non-float to float, apply the saved float dimensions. view.pending.box = view.float_box; } if (!fullscreen_found and view.pending.fullscreen and view.pending.tags & output.pending.tags != 0) { fullscreen_found = true; if (output.inflight.fullscreen != view) { if (output.inflight.fullscreen) |old| { old.setFullscreen(false); old.pending.box = old.post_fullscreen_box; old.inflight.box = old.pending.box; } output.inflight.fullscreen = view; view.setFullscreen(true); view.post_fullscreen_box = view.pending.box; view.pending.box = .{ .x = 0, .y = 0, .width = undefined, .height = undefined, }; output.wlr_output.effectiveResolution( &view.pending.box.width, &view.pending.box.height, ); view.inflight.box = view.pending.box; } } view.inflight_focus_stack_link.remove(); output.inflight.focus_stack.append(view); view.inflight = view.pending; } } if (!fullscreen_found) { output.inflight.fullscreen = null; } { var it = output.pending.wm_stack.iterator(.forward); while (it.next()) |view| { view.inflight_wm_stack_link.remove(); output.inflight.wm_stack.append(view); } } output.inflight.tags = output.pending.tags; } } { // Layout demands can't be sent until after the inflight stacks of // all outputs have been updated. var output_it = root.outputs.first; while (output_it) |node| : (output_it = node.next) { const output = &node.data; assert(output.inflight.layout_demand == null); if (output.layout) |layout| { var layout_count: u32 = 0; { var it = output.inflight.wm_stack.iterator(.forward); while (it.next()) |view| { if (!view.inflight.float and !view.inflight.fullscreen and view.inflight.tags & output.inflight.tags != 0) { layout_count += 1; } } } if (layout_count > 0) { // TODO don't do this if the count has not changed layout.startLayoutDemand(layout_count); } } } } if (root.inflight_layout_demands == 0) { root.sendConfigures(); } } /// This function is used to inform the transaction system that a layout demand /// has either been completed or timed out. If it was the last pending layout /// demand in the current sequence, a transaction is started. pub fn notifyLayoutDemandDone(root: *Self) void { root.inflight_layout_demands -= 1; if (root.inflight_layout_demands == 0) { root.sendConfigures(); } } fn sendConfigures(root: *Self) void { assert(root.inflight_layout_demands == 0); assert(root.inflight_configures == 0); // Iterate over all views of all outputs var output_it = root.outputs.first; while (output_it) |output_node| : (output_it = output_node.next) { const output = &output_node.data; var focus_stack_it = output.inflight.focus_stack.iterator(.forward); while (focus_stack_it.next()) |view| { if (view.needsConfigure()) { view.configure(); // We don't give a damn about frame perfection for xwayland views if (!build_options.xwayland or view.impl != .xwayland_view) { root.inflight_configures += 1; view.saveSurfaceTree(); view.sendFrameDone(); } } } } if (root.inflight_configures > 0) { std.log.scoped(.transaction).debug("started transaction with {} pending configure(s)", .{ root.inflight_configures, }); root.transaction_timeout.timerUpdate(200) catch { std.log.scoped(.transaction).err("failed to update timer", .{}); root.commitTransaction(); }; } else { root.commitTransaction(); } } fn handleTransactionTimeout(self: *Self) c_int { assert(self.inflight_layout_demands == 0); std.log.scoped(.transaction).err("timeout occurred, some imperfect frames may be shown", .{}); self.inflight_configures = 0; self.commitTransaction(); return 0; } pub fn notifyConfigured(self: *Self) void { assert(self.inflight_layout_demands == 0); self.inflight_configures -= 1; if (self.inflight_configures == 0) { // Disarm the timer, as we didn't timeout self.transaction_timeout.timerUpdate(0) catch std.log.scoped(.transaction).err("error disarming timer", .{}); self.commitTransaction(); } } /// Apply the inflight state and drop stashed buffers. This means that /// the next frame drawn will be the post-transaction state of the /// layout. Should only be called after all clients have configured for /// the new layout. If called early imperfect frames may be drawn. fn commitTransaction(root: *Self) void { assert(root.inflight_layout_demands == 0); assert(root.inflight_configures == 0); { var it = root.hidden.inflight.focus_stack.safeIterator(.forward); while (it.next()) |view| { assert(view.inflight.output == null); view.current.output = null; view.tree.node.reparent(root.hidden.tree); view.popup_tree.node.reparent(root.hidden.tree); view.updateCurrent(); } } var output_it = root.outputs.first; while (output_it) |output_node| : (output_it = output_node.next) { const output = &output_node.data; if (output.inflight.tags != output.current.tags) { std.log.scoped(.output).debug( "changing current focus: {b:0>10} to {b:0>10}", .{ output.current.tags, output.inflight.tags }, ); } output.current.tags = output.inflight.tags; var focus_stack_it = output.inflight.focus_stack.iterator(.forward); while (focus_stack_it.next()) |view| { assert(view.inflight.output == output); view.inflight_serial = null; if (view.current.output != output or (output.current.fullscreen == view and output.inflight.fullscreen != view)) { view.tree.node.reparent(output.layers.views); view.popup_tree.node.reparent(output.layers.popups); } const enabled = view.current.tags & output.current.tags != 0; view.tree.node.setEnabled(enabled); view.popup_tree.node.setEnabled(enabled); if (output.inflight.fullscreen != view) { // TODO this approach for syncing the order will likely cause over-damaging. view.tree.node.lowerToBottom(); } view.updateCurrent(); } if (output.inflight.fullscreen != output.current.fullscreen) { if (output.inflight.fullscreen) |view| { assert(view.inflight.output == output); assert(view.current.output == output); view.tree.node.reparent(output.layers.fullscreen); } output.current.fullscreen = output.inflight.fullscreen; output.layers.fullscreen.node.setEnabled(output.current.fullscreen != null); } output.status.handleTransactionCommit(output); } { var it = server.input_manager.seats.first; while (it) |node| : (it = node.next) node.data.cursor.updateState(); } { // This must be done after updating cursor state in case the view was the target of move/resize. var it = root.hidden.inflight.focus_stack.safeIterator(.forward); while (it.next()) |view| { if (view.destroying) view.destroy(); } } server.idle_inhibitor_manager.idleInhibitCheckActive(); if (root.pending_state_dirty) { root.applyPending(); } } /// Send the new output configuration to all wlr-output-manager clients fn handleLayoutChange(listener: *wl.Listener(*wlr.OutputLayout), _: *wlr.OutputLayout) void { const self = @fieldParentPtr(Self, "layout_change", listener); const config = self.currentOutputConfig() catch { std.log.scoped(.output_manager).err("out of memory", .{}); return; }; self.output_manager.setConfiguration(config); } fn handleManagerApply( listener: *wl.Listener(*wlr.OutputConfigurationV1), config: *wlr.OutputConfigurationV1, ) void { const self = @fieldParentPtr(Self, "manager_apply", listener); defer config.destroy(); self.processOutputConfig(config, .apply); // Send the config that was actually applied const applied_config = self.currentOutputConfig() catch { std.log.scoped(.output_manager).err("out of memory", .{}); return; }; self.output_manager.setConfiguration(applied_config); } fn handleManagerTest( listener: *wl.Listener(*wlr.OutputConfigurationV1), config: *wlr.OutputConfigurationV1, ) void { const self = @fieldParentPtr(Self, "manager_test", listener); defer config.destroy(); self.processOutputConfig(config, .test_only); } fn processOutputConfig( self: *Self, config: *wlr.OutputConfigurationV1, action: enum { test_only, apply }, ) void { // Ignore layout change events this function generates while applying the config self.layout_change.link.remove(); defer self.output_layout.events.change.add(&self.layout_change); var success = true; var it = config.heads.iterator(.forward); while (it.next()) |head| { const wlr_output = head.state.output; const output = @intToPtr(*Output, wlr_output.data); var proposed_state = wlr.Output.State.init(); head.state.apply(&proposed_state); switch (action) { .test_only => { if (!wlr_output.testState(&proposed_state)) success = false; }, .apply => { if (wlr_output.commitState(&proposed_state)) { if (head.state.enabled) { // Just updates the output's position if it is already in the layout self.output_layout.add(output.wlr_output, head.state.x, head.state.y); output.tree.node.setEnabled(true); output.tree.node.setPosition(head.state.x, head.state.y); output.arrangeLayers(); } else { self.removeOutput(output); self.output_layout.remove(output.wlr_output); output.tree.node.setEnabled(false); } } else { std.log.scoped(.output_manager).err("failed to apply config to output {s}", .{ output.wlr_output.name, }); success = false; } }, } } if (action == .apply) self.applyPending(); if (success) { config.sendSucceeded(); } else { config.sendFailed(); } } fn currentOutputConfig(self: *Self) !*wlr.OutputConfigurationV1 { // TODO there no real reason this needs to allocate memory every time it is called. // consider improving this wlroots api or reimplementing in zig-wlroots/river. const config = try wlr.OutputConfigurationV1.create(); // this destroys all associated config heads as well errdefer config.destroy(); var it = self.all_outputs.first; while (it) |node| : (it = node.next) { const output = node.data; const head = try wlr.OutputConfigurationV1.Head.create(config, output.wlr_output); // If the output is not part of the layout (and thus disabled) // the box will be zeroed out. var box: wlr.Box = undefined; self.output_layout.getBox(output.wlr_output, &box); head.state.x = box.x; head.state.y = box.y; } return config; } fn handlePowerManagerSetMode( _: *wl.Listener(*wlr.OutputPowerManagerV1.event.SetMode), event: *wlr.OutputPowerManagerV1.event.SetMode, ) void { const enable = event.mode == .on; const log_text = if (enable) "Enabling" else "Disabling"; std.log.scoped(.output_manager).debug( "{s} dpms for output {s}", .{ log_text, event.output.name }, ); event.output.enable(enable); event.output.commit() catch { std.log.scoped(.server).err("output commit failed for {s}", .{event.output.name}); }; }