Support macos walltime without profiling

[GuillaumeLagrange]

No description provided.

[codspeed-hq]

Merging this PR will degrade performance by 75.99%

⚡ 4 improved benchmarks
❌ 3 regressed benchmarks
✅ 568 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	WallTime	iter_batched_ref_large_input	48 ns	44 ns	+9.09%
❌	WallTime	large_drop	203.8 µs	848.8 µs	-75.99%
⚡	WallTime	slice_into_vec_with_bytes	94 ns	55 ns	+70.91%
❌	WallTime	graph_coloring[6]	1.1 µs	1.6 µs	-28.86%
⚡	WallTime	init_array[1000]	2.2 µs	2.1 µs	+5.42%
⚡	WallTime	init_array[42]	56 ns	47 ns	+19.15%
❌	Simulation	process_items_with_counter	835.8 ns	865 ns	-3.37%

---

_{Comparing cod-2461-output-compilation-artifacts-for-cargo-codspeed-codspeed (58b28e6) with main (1339cb3)}