feat(uhyve): use v2 interface

[n0toose]

On x86_64, the data address will be stored in the rcx register to support 64-bit addresses.

This also removes the usage of ExitArgs, as it is a single integer.

[n0toose]

Succeeds #2013

[github-actions]

Benchmark Results

Details

Benchmark	Current: 12fe5b1	Previous: e0ce538	Performance Ratio
startup_benchmark Build Time	89.15 s	88.64 s	1.01 ❗
startup_benchmark File Size	0.77 MB	0.78 MB	0.98 ❗
Startup Time - 1 core	0.97 s (±0.02 s)	0.96 s (±0.03 s)	1.01
Startup Time - 2 cores	0.95 s (±0.03 s)	0.96 s (±0.04 s)	0.99
Startup Time - 4 cores	0.96 s (±0.04 s)	0.98 s (±0.03 s)	0.99
multithreaded_benchmark Build Time	91.19 s	90.67 s	1.01 ❗
multithreaded_benchmark File Size	0.89 MB	0.89 MB	1.00 ❗
Multithreaded Pi Efficiency - 2 Threads	92.70 % (±8.25 %)	88.34 % (±9.63 %)	1.05
Multithreaded Pi Efficiency - 4 Threads	44.82 % (±4.09 %)	43.35 % (±2.77 %)	1.03
Multithreaded Pi Efficiency - 8 Threads	25.63 % (±2.48 %)	25.17 % (±2.41 %)	1.02
micro_benchmarks Build Time	101.63 s	100.90 s	1.01 ❗
micro_benchmarks File Size	0.90 MB	0.90 MB	1.00 ❗
Scheduling time - 1 thread	72.84 ticks (±4.23 ticks)	73.66 ticks (±4.85 ticks)	0.99
Scheduling time - 2 threads	39.68 ticks (±5.44 ticks)	38.61 ticks (±3.13 ticks)	1.03
Micro - Time for syscall (getpid)	3.00 ticks (±0.27 ticks)	3.92 ticks (±0.20 ticks)	0.76 ❗
Memcpy speed - (built_in) block size 4096	63839.54 MByte/s (±45358.10 MByte/s)	60742.27 MByte/s (±42902.76 MByte/s)	1.05
Memcpy speed - (built_in) block size 1048576	29466.91 MByte/s (±24671.87 MByte/s)	30041.63 MByte/s (±25084.98 MByte/s)	0.98
Memcpy speed - (built_in) block size 16777216	22808.82 MByte/s (±18915.94 MByte/s)	22962.35 MByte/s (±19327.03 MByte/s)	0.99
Memset speed - (built_in) block size 4096	64075.10 MByte/s (±45488.32 MByte/s)	61460.72 MByte/s (±43387.59 MByte/s)	1.04
Memset speed - (built_in) block size 1048576	30245.82 MByte/s (±25097.01 MByte/s)	30843.30 MByte/s (±25532.01 MByte/s)	0.98
Memset speed - (built_in) block size 16777216	23449.68 MByte/s (±19313.19 MByte/s)	23720.59 MByte/s (±19854.83 MByte/s)	0.99
Memcpy speed - (rust) block size 4096	56742.04 MByte/s (±42014.88 MByte/s)	58684.11 MByte/s (±43425.02 MByte/s)	0.97
Memcpy speed - (rust) block size 1048576	29412.15 MByte/s (±24620.81 MByte/s)	29918.21 MByte/s (±25119.12 MByte/s)	0.98
Memcpy speed - (rust) block size 16777216	21612.02 MByte/s (±17905.12 MByte/s)	23297.77 MByte/s (±19599.55 MByte/s)	0.93
Memset speed - (rust) block size 4096	57623.31 MByte/s (±42560.27 MByte/s)	58975.02 MByte/s (±43585.57 MByte/s)	0.98
Memset speed - (rust) block size 1048576	30215.24 MByte/s (±25060.57 MByte/s)	30705.41 MByte/s (±25550.79 MByte/s)	0.98
Memset speed - (rust) block size 16777216	22156.10 MByte/s (±18229.85 MByte/s)	24054.41 MByte/s (±20114.49 MByte/s)	0.92
alloc_benchmarks Build Time	92.46 s	98.25 s	0.94 ❗
alloc_benchmarks File Size	0.84 MB	0.85 MB	0.98 ❗
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	12351.18 Ticks (±503.53 Ticks)	9975.44 Ticks (±278.92 Ticks)	1.24 ❗
Allocations - Average Allocation time (no fail)	12351.18 Ticks (±503.53 Ticks)	9975.44 Ticks (±278.92 Ticks)	1.24 ❗
Allocations - Average Deallocation time	5410.77 Ticks (±736.45 Ticks)	1930.66 Ticks (±1176.47 Ticks)	2.80 ❗
mutex_benchmark Build Time	92.17 s	92.55 s	1.00 ❗
mutex_benchmark File Size	0.90 MB	0.90 MB	1.00 ❗
Mutex Stress Test Average Time per Iteration - 1 Threads	13.16 ns (±0.64 ns)	13.04 ns (±0.87 ns)	1.01
Mutex Stress Test Average Time per Iteration - 2 Threads	20.46 ns (±8.14 ns)	19.26 ns (±7.84 ns)	1.06

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

[jounathaen]

LGTM!

@n0toose could you maybe fix the clippy error?

[n0toose]

All Uhyve tests pass on x86_64 (apart from memory with this branch, which was fully expected): https://github.com/hermit-os/uhyve/blob/c5d573701e479e4ae1c8974bec80742ee928cff0/tests/memory.rs#L29-L30

I have not confirmed the same for macOS/aarch64 (yet), for which a small modification to account for 64-bytes was made: https://github.com/hermit-os/kernel/pull/2190/changes#diff-3f8326589ecec56a9d94f21f100834839de00873cd3eb7d653c09e962450876bR81