//! The Vulkan session presenter (punktfunk-planning `linux-client-rearchitecture.md`, //! Phase 1): an SDL3 window + ash swapchain that presents the shared session pump's //! decoded frames, captures input on the `ui_stream` state-machine contract, and reports //! the unified stats window on stdout. No UI toolkit anywhere in the dependency tree. //! //! Three frame paths: software (`CpuFrame` RGBA staging upload), Vulkan Video (the //! decoder's VkImage on THIS device — plane views + the CICP-driven CSC pass), and on //! Linux additionally VAAPI hardware (NV12 dmabuf imported per-plane — `dmabuf.rs`), //! all composited by a letterboxed blit. Devices without the import extensions, and any //! import/present failure streak, demote the decoder to software via the session pump's //! `force_software` contract, same as the GTK presenter. //! //! Builds on Linux AND Windows; `dmabuf` is Linux-only (DRM-PRIME does not exist on //! Windows) and `d3d11` is its Windows counterpart (D3D11VA shared-texture import) — //! the decode chain there is Vulkan → D3D11VA → software. // Unsafe-proof program: every `unsafe {}` in this crate carries a `// SAFETY:` proof. #![deny(clippy::undocumented_unsafe_blocks)] // THE VULKAN CONTRACT, stated once - most `// SAFETY:` proofs in this crate are an instance of it. // // Nearly every `unsafe` here is an `ash` call, which is `unsafe` because Vulkan is a C API, not // because each call carries its own bespoke obligation. Three shapes recur, and only one of them // has a real precondition worth restating per site: // // * CREATE / ALLOCATE - `create_*`, `allocate_*`. The device is live (this type owns it), and the // `vk::*CreateInfo` builders are locals that outlive the synchronous call. The handle returned is // owned by the value being constructed and destroyed in its `Drop`. // * RECORD - `cmd_*`, `begin/end_command_buffer`, `update_descriptor_sets`. Recorded into a command // buffer this code owns and has begun, referencing handles it also owns. Nothing executes until // submit, so a recording error is not yet a memory error. // * DESTROY - `destroy_*`, `free_*`, `unmap_memory`. THIS is the one with a real obligation: the // GPU must not still be using the object. That is established on the path, not by the call - a // fence wait, a `queue_wait_idle`, or the swapchain having been retired - and the per-site proofs // say so, because getting it wrong is a use-after-free the type system cannot catch. // // A block doing something OUTSIDE these three shapes gets a real, specific proof; if you add one and // find yourself writing "as above", it probably belongs in one of them. #[cfg(any(target_os = "linux", windows))] pub mod csc; #[cfg(any(target_os = "linux", windows))] pub mod cursor; #[cfg(windows)] pub mod d3d11; #[cfg(target_os = "linux")] pub mod dmabuf; #[cfg(any(target_os = "linux", windows))] pub mod input; #[cfg(any(target_os = "linux", windows))] pub mod keymap_sdl; #[cfg(any(target_os = "linux", windows))] pub mod overlay; #[cfg(any(target_os = "linux", windows))] mod run; #[cfg(any(target_os = "linux", windows))] pub mod touch; #[cfg(any(target_os = "linux", windows))] pub mod vk; #[cfg(windows)] mod win32; #[cfg(any(target_os = "linux", windows))] pub use run::{run_browse, run_session, ActionOutcome, Outcome, SessionOpts};