bevy_city crash in buffer_vec after GPU data update through shared memory

[mockersf]

Bevy version and features

• main since Allow mesh collection workers to update GPU data directly via shared memory. #22988

What you did

cargo run --release --package bevy_city

What went wrong

thread 'Compute Task Pool (3)' (20073179) panicked at bevy/crates/bevy_render/src/render_resource/buffer_vec.rs:351:41:
index out of bounds: the len is 12232 but the index is 12232
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Encountered a panic in system `bevy_pbr::render::mesh::collect_meshes_for_gpu_building`!
Encountered a panic in system `bevy_render::run_render_schedule`!

[mockersf]

also broke example mixed_lighting with a different panic:

thread 'Compute Task Pool (1)' (20267961) panicked at bevy/crates/bevy_render/src/batching/gpu_preprocessing.rs:383:9:
push exceeded pre-allocated capacity
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Encountered a panic in system `bevy_pbr::render::mesh::collect_meshes_for_gpu_building`!
Encountered a panic in system `bevy_render::run_render_schedule`!

[eswartz]

Also see this in parallax_mapping, though I was just randomly re-running this example on main.

[mockersf]

it's happening in around 25 examples, some randomly, and some seems to be a bit OS dependent, but the panics are the same