Files
uu-remote-linux/src/privacy_screen/guard.rs
T
Jeff Zhang 8e30371317 feat(host): KDE Wayland privacy screen (RPC 7) (#146)
* feat(host): KDE Wayland 被控端防窥模式(隐私 / 关屏)

viewer RPC 7 打开防窥时,采集和输入切到 KWin 虚拟输出(portal ScreenCast VIRTUAL),
再用 kscreen-doctor 禁用物理屏;隐私模式随后重新启用物理屏并全屏覆盖桌面壁纸。
防窥期间通过 KWin InputDevice 禁用本地键鼠,指针/按键/文字改走 uinput
(KDE 不处理虚拟输出共享上的 portal 输入)。

- privacy_type 实测:1 关闭、2 隐私(壁纸)、3 关屏;SettingMetric 4/7 上报真实状态
- 恢复记录 + privacy-guard 子进程:主机以任何方式结束(含 SIGKILL)都恢复屏幕和输入
- src/privacy_screen/ 分层:控制器/协议/记录/守护/遮挡窗口共享,桌面实现走 Backend trait(kde/)
- portal:Session::set_virtual,闭包恢复按当前源重开;编码器随 portal 尺寸变化重建

Refs #140

* fix(privacy): 按 Mac 客户端协议应答防窥请求,修复授权竞态

- FeatureFlag ff_private_screen=2(ENHANCED_PRIVACY),声明支持关屏
- 先回 RPC 7 应答再切换,避免客户端超时报 9998;切换失败发 ReportError -7
- 应答 RPC 25 update_screen_saver_req(RpcResponse 21),隐私模式不再被中止
- SettingMetric 4/7 上报开关与最近一次模式(host_settings.json privacy_type),状态变化推送 SystemStateChange 5
- portal 会话串行打开,打开失败回填一次性授权:快速重连时防窥切换与闭包恢复并发曾清空保存的授权

字段号与枚举取自官方 Mac 客户端的 SwiftProtobuf name map。
已知限制:Mac 客户端防窥菜单切换后仍显示"不启用",实际模式正确。

Refs #140

* fix(privacy): 处理 #146 评审意见

- 预检 uinput:injector 未打开 uinput(/dev/uinput 不可用或 UU_INPUT_BACKEND=native)时,
  在改动输出和设备之前拒绝防窥(虚拟输出共享上 KDE 忽略 portal 输入)
- leave() 中输入设备恢复失败在恢复输出后向上返回,保留 privacy-power.json 与 guard
- record::ownership() 只缓存成功持锁,guard 轮询和 recover() 可重试
- 遮挡窗口就绪握手:privacy-wallpaper 建好窗口后输出 ready,5 秒内未就绪则报错并回滚
- rebase 到 main(#145):DMA-BUF 打开路径跳过虚拟输出、失败同样回填授权;
  #145 的 uinput 指针路径也按隐私模式的虚拟输出区域换算坐标
- README 表格压缩为一句,细节在 docs/privacy-screen.md

Refs #140

* docs(privacy): 注明 ff_system_metrics=3 对应的枚举

Refs #140

* fix(privacy): resolve lifecycle and restoration review findings
2026-11-02 01:14:10 +08:00

32 lines
1.6 KiB
Rust

//! Desktops may persist what a backend changes (KWin saves input-device flags), so a killed host
//! could leave local input or screens off for good. This child recovers from the record when the
//! host's end of its stdin pipe closes, however the host ends; after a normal restore the record is
//! gone and it just exits.
use std::{process::{Child, Command, Stdio}, sync::Mutex};
static GUARD: Mutex<Option<Child>> = Mutex::new(None);
pub fn start() -> Result<(), String> {
let mut guard = GUARD.lock().map_err(|_| "guard state poisoned")?;
if guard.is_some() { return Ok(()); }
let exe = std::env::current_exe().map_err(|e| e.kind().to_string())?;
let child = Command::new(exe).arg("privacy-guard").stdin(Stdio::piped()).stdout(Stdio::null()).spawn()
.map_err(|e| format!("privacy guard: {}", e.kind()))?;
*guard = Some(child);
Ok(())
}
pub fn stop() {
if let Some(mut child) = GUARD.lock().ok().and_then(|mut g| g.take()) { drop(child.stdin.take()); let _ = child.wait(); }
}
/// `privacy-guard`: waits for the host to go away, then recovers.
pub fn main() -> i32 {
use std::io::Read;
let _ = std::io::stdin().read_to_end(&mut Vec::new());
// A normal restore removes the record before closing the pipe.
if !super::record::pending() { return 0; }
// The host's flock is released with its last descriptor; give a dying host a moment.
super::tool::wait_until(|| super::record::ownership().is_ok());
if super::record::pending() { log::warn!("Privacy guard: host ended without restoring; recovering"); super::recover(); }
0
}