fix(stats): decode stage measures GPU completion; Alt+Enter fullscreen alias

The Vulkan path's receive_frame returns at SUBMISSION (~0.1 ms) — the
hardware decodes asynchronously, so the decode stat was truthful but
measured the wrong boundary. The pump now ships the frame to the
presenter FIRST, then waits the frame's timeline fence (vkWaitSemaphores
resolved through the shared device's proc chain) and stamps
received→decode-COMPLETE — true NVDEC time at zero pipeline cost, since
the presenter's own GPU wait is what actually gates sampling. Software/
VAAPI keep their synchronous stamps.

Also: Alt+Enter joins F11 as the fullscreen toggle (some keyboards' Fn
layer sends a media key for plain F11 — observed on glass as
'F11 only works with shift'); the shell's shortcuts panel lists both.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
2026-07-07 22:11:23 +02:00
parent a5bb5ec4d5
commit 349d16382e
14 changed files with 276 additions and 26 deletions
+4
View File
@@ -57,6 +57,10 @@ fn main() {
.allowlist_type("VkPhysicalDeviceVulkan13Features")
// AVVulkanFramesContext.img_flags values (plane views need MUTABLE_FORMAT).
.allowlist_type("VkImageCreateFlagBits")
// Timeline-semaphore wait — the pump measures true GPU decode completion.
.allowlist_type("VkSemaphoreWaitInfo")
.allowlist_type("PFN_vkWaitSemaphores")
.allowlist_type("PFN_vkGetDeviceProcAddr")
// …plus nothing else of FFmpeg: the core types these structs reference only
// ever appear behind pointers here, so keep them opaque instead of duplicating
// ffmpeg-sys-next's definitions (callers cast pointers between the crates).