Greg Pomerantz
93b27d724b
The 19-26 ms mid-scroll stalls (Mali-only, fixed content offsets) were the Skia renderer's final synchronous GrDirectContext::flush() blocking the UI thread on Mali driver back-pressure — proven by per-stage timestamps in the vendored renderer (stall frames: flush 19-24 ms, walk 1-3 ms, 1 binding re-eval, image imports 0.1 ms each). - vendor/i-slint-renderer-skia (pinned 1.18.1, [patch.crates-io]): remove that one flush. Redundant on the wgpu backend: the surface already calls flush_and_submit before queue.present, same Vulkan queue, in-order. Phone: 0 stalls across 4 runs (max 3.6-7.4 ms vs the 8.3 ms budget), display verified clean. - docs/BUGS/slint-skia-final-flush-jank.md: full evidence, prior art (slint-ui/slint#12030 closest), open question for a future upstream filing (not filed yet). - Profiling also retired the O(n)-walk theory: Slint 1.18 prunes the item walk to ~133-152 items; the windowed Flickable built on that theory was already reverted (ListView is smooth with the fix). - Strip all jank-hunt instrumentation: FR/SCR/TICK/APLYD/STG logging, stg markers, .gallery_* file knobs, jsamp thread sampler, DbgScroll programmatic scroll, nblog perf-file mirroring. Also fixes the 'launches into unstoppable fast scroll' bug: a leftover .gallery_cmd was re-read on every launch (file never deleted) and overrode user swipes — the hook is gone. - Keep: nblog (logd backpressure is real), zero-dimension decode guards, keep-alive (default-off). - HANDOFF: replace stale windowed-Flickable section with the actual root cause/fix; 15 unit tests pass. Verified on phone (aarch64, 8580 photos) and emulator (x86_64).
31 lines
1.3 KiB
Rust
31 lines
1.3 KiB
Rust
// Copyright © SixtyFPS GmbH <info@slint.dev>
|
|
// SPDX-License-Identifier: GPL-3.0-only OR LicenseRef-Slint-Royalty-free-2.0 OR LicenseRef-Slint-Software-3.0
|
|
|
|
// cSpell: ignore dispatchable
|
|
//! Raw values of the Vulkan handles wgpu-hal hands out.
|
|
//!
|
|
//! wgpu-hal exposes its Vulkan objects as `ash::vk` handle types without re-exporting `ash`,
|
|
//! and Skia's `vk` API takes the raw values.
|
|
//! Every `ash::vk` handle is a `#[repr(transparent)]` wrapper mirroring the C type:
|
|
//! a pointer for dispatchable handles (`VkInstance`, `VkQueue`, ...)
|
|
//! and a `u64` for non-dispatchable ones (`VkImage`, ...).
|
|
//! Reading the wrapper is what `ash::vk::Handle::as_raw` does, minus the dependency.
|
|
|
|
/// Returns the raw value of a dispatchable handle.
|
|
///
|
|
/// # Safety
|
|
/// `handle` must be one of `ash::vk`'s dispatchable handle types.
|
|
pub unsafe fn dispatchable<H: Copy>(handle: H) -> u64 {
|
|
const { assert!(size_of::<H>() == size_of::<*mut u8>()) };
|
|
unsafe { core::mem::transmute_copy::<H, *mut u8>(&handle) as u64 }
|
|
}
|
|
|
|
/// Returns the raw value of a non-dispatchable handle.
|
|
///
|
|
/// # Safety
|
|
/// `handle` must be one of `ash::vk`'s non-dispatchable handle types.
|
|
pub unsafe fn non_dispatchable<H: Copy>(handle: H) -> u64 {
|
|
const { assert!(size_of::<H>() == size_of::<u64>()) };
|
|
unsafe { core::mem::transmute_copy::<H, u64>(&handle) }
|
|
}
|