perf(pvirtq): combine u15 offset and u1 wrap counter in more places

[mkroening]

This PR makes use of the offset + wrap bitfield for saving one byte per index. The performance improvement might not be measurable, but this also makes the code more readable by combining these related fields.

Depends on #2046.

[github-actions]

Benchmark Results

Benchmark	Current: 57f95d4	Previous: 9d1e4dd	Performance Ratio
startup_benchmark Build Time	114.72 s	111.98 s	1.02
startup_benchmark File Size	0.91 MB	0.91 MB	1.00
Startup Time - 1 core	0.94 s (±0.02 s)	0.94 s (±0.02 s)	1.00
Startup Time - 2 cores	0.94 s (±0.02 s)	0.94 s (±0.03 s)	1.00
Startup Time - 4 cores	0.94 s (±0.03 s)	0.94 s (±0.02 s)	1.00
multithreaded_benchmark Build Time	115.29 s	112.12 s	1.03
multithreaded_benchmark File Size	1.02 MB	1.02 MB	1.00
Multithreaded Pi Efficiency - 2 Threads	89.87 % (±7.80 %)	88.00 % (±7.30 %)	1.02
Multithreaded Pi Efficiency - 4 Threads	45.54 % (±3.47 %)	43.95 % (±3.32 %)	1.04
Multithreaded Pi Efficiency - 8 Threads	25.92 % (±2.62 %)	25.25 % (±2.26 %)	1.03
micro_benchmarks Build Time	294.80 s	315.31 s	0.93
micro_benchmarks File Size	1.02 MB	1.02 MB	1.00
Scheduling time - 1 thread	187.76 ticks (±30.58 ticks)	181.72 ticks (±30.16 ticks)	1.03
Scheduling time - 2 threads	103.66 ticks (±21.44 ticks)	107.77 ticks (±18.94 ticks)	0.96
Micro - Time for syscall (getpid)	10.81 ticks (±6.33 ticks)	13.22 ticks (±5.19 ticks)	0.82
Memcpy speed - (built_in) block size 4096	58059.52 MByte/s (±41478.23 MByte/s)	55204.30 MByte/s (±40404.17 MByte/s)	1.05
Memcpy speed - (built_in) block size 1048576	13798.77 MByte/s (±11407.26 MByte/s)	14269.94 MByte/s (±12083.95 MByte/s)	0.97
Memcpy speed - (built_in) block size 16777216	8440.61 MByte/s (±6978.73 MByte/s)	7485.41 MByte/s (±6068.00 MByte/s)	1.13
Memset speed - (built_in) block size 4096	57061.47 MByte/s (±41206.31 MByte/s)	55494.03 MByte/s (±40595.71 MByte/s)	1.03
Memset speed - (built_in) block size 1048576	14212.20 MByte/s (±11676.23 MByte/s)	14670.46 MByte/s (±12334.51 MByte/s)	0.97
Memset speed - (built_in) block size 16777216	8647.22 MByte/s (±7111.15 MByte/s)	7599.70 MByte/s (±6128.76 MByte/s)	1.14
Memcpy speed - (rust) block size 4096	48221.83 MByte/s (±35963.38 MByte/s)	51056.01 MByte/s (±38411.57 MByte/s)	0.94
Memcpy speed - (rust) block size 1048576	13676.87 MByte/s (±11230.07 MByte/s)	13857.34 MByte/s (±11384.86 MByte/s)	0.99
Memcpy speed - (rust) block size 16777216	8183.51 MByte/s (±6689.26 MByte/s)	7529.44 MByte/s (±6166.85 MByte/s)	1.09
Memset speed - (rust) block size 4096	49285.30 MByte/s (±36845.92 MByte/s)	51991.84 MByte/s (±39176.17 MByte/s)	0.95
Memset speed - (rust) block size 1048576	13969.29 MByte/s (±11407.85 MByte/s)	14087.35 MByte/s (±11505.86 MByte/s)	0.99
Memset speed - (rust) block size 16777216	8404.33 MByte/s (±6846.10 MByte/s)	7599.50 MByte/s (±6197.41 MByte/s)	1.11
alloc_benchmarks Build Time	295.19 s	312.10 s	0.95
alloc_benchmarks File Size	0.98 MB	0.98 MB	1.00
Allocations - Allocation success	100.00 %	100.00 %	1
Allocations - Deallocation success	100.00 %	100.00 %	1
Allocations - Pre-fail Allocations	100.00 %	100.00 %	1
Allocations - Average Allocation time	19837.35 Ticks (±1087.80 Ticks)	20044.74 Ticks (±1076.26 Ticks)	0.99
Allocations - Average Allocation time (no fail)	19837.35 Ticks (±1087.80 Ticks)	20044.74 Ticks (±1076.26 Ticks)	0.99
Allocations - Average Deallocation time	2869.62 Ticks (±1189.04 Ticks)	2980.59 Ticks (±1256.07 Ticks)	0.96
mutex_benchmark Build Time	293.94 s	296.51 s	0.99
mutex_benchmark File Size	1.02 MB	1.02 MB	1.00
Mutex Stress Test Average Time per Iteration - 1 Threads	36.78 ns (±4.34 ns)	36.10 ns (±4.90 ns)	1.02
Mutex Stress Test Average Time per Iteration - 2 Threads	31.70 ns (±3.53 ns)	29.58 ns (±2.65 ns)	1.07

This comment was automatically generated by workflow using github-action-benchmark.

[Gelbpunkt]

Not sure about perf here since bitfields should have generally worse access times due to the necessary bitshifts as opposed to doing them once and using an intermediate struct for holding the values. I don't really mind either way, but this version feels like we're at least not reinventing virtio-spec