add shift benchmark

[malik672]

benchmark for shift operation to unblock #546
@prestwich

[codspeed-hq]

CodSpeed Performance Report

Merging #547 will degrade performances by 10.18%

_{Comparing malik672:less+than_zero (848bd6a) with main (17c9b3e)}

Summary

⚡ 1 improvement
❌ 2 regressions
✅ 383 untouched
🆕 14 new

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

Benchmarks breakdown

	Benchmark	BASE	HEAD	Change
⚡	most_significant_bits/4096/4096	29 µs	26.1 µs	+11.05%
🆕	shift_left_small/128	N/A	173.7 µs	N/A
🆕	shift_left_small/192	N/A	210.2 µs	N/A
🆕	shift_left_small/256	N/A	246.6 µs	N/A
🆕	shift_left_small/384	N/A	319.5 µs	N/A
🆕	shift_left_small/4096	N/A	4 µs	N/A
🆕	shift_left_small/512	N/A	392.5 µs	N/A
🆕	shift_left_small/64	N/A	134.6 µs	N/A
🆕	shift_right_small/128	N/A	173.7 µs	N/A
🆕	shift_right_small/192	N/A	210.2 µs	N/A
🆕	shift_right_small/256	N/A	246.6 µs	N/A
🆕	shift_right_small/384	N/A	319.6 µs	N/A
🆕	shift_right_small/4096	N/A	38.8 µs	N/A
🆕	shift_right_small/512	N/A	392.5 µs	N/A
🆕	shift_right_small/64	N/A	134.6 µs	N/A
❌	to/f32/4096	26.2 µs	29.1 µs	-10.18%
❌	to/f64/4096	26.3 µs	29.3 µs	-10.14%