Support macos walltime without profiling in codspeed-rust

[GuillaumeLagrange]

No description provided.

[codspeed-hq]

Merging this PR will degrade performance by 75.45%

⚡ 6 improved benchmarks
❌ 6 regressed benchmarks
✅ 563 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	iter_batched_large_input	44 ns	48 ns	-8.33%
❌	WallTime	large_drop	203.8 µs	829.9 µs	-75.45%
❌	WallTime	iter_with_setup	44 ns	48 ns	-8.33%
❌	WallTime	graph_coloring[5]	1 µs	1.3 µs	-20.39%
⚡	WallTime	system_time	40 ns	38 ns	+5.26%
❌	WallTime	find_highest_set_bit[0]	5 ns	17 ns	-70.59%
⚡	WallTime	hamiltonian_cycle[5]	935 ns	887 ns	+5.41%
⚡	WallTime	slice_into_vec_with_bytes	94 ns	55 ns	+70.91%
⚡	WallTime	init_array[42]	56 ns	47 ns	+19.15%
⚡	WallTime	init_array[1000]	2.2 µs	2.1 µs	+4.52%
❌	Simulation	process_items_with_counter	835.8 ns	865 ns	-3.37%

---

_{Comparing cod-2459-be-able-to-run-all-integrations-on-macos (58b28e6) with main (1339cb3)}