feat: add generic ioctl support

[zyuiop]

This PR adds an ioctl handler to file descriptor objects, such as sockets, and a helper to register any ioctl handler at a particular file system path.

Some Linux-specific crates communicate with the kernel via the use of an ioctl file. Support of ioctls may provide an easy path when porting such libraries to work with hermit, without increasing too much the size of the ABI for too specific features.

[mkroening]

We could use bitfield-struct for these bitfield types, which is already a transitive dependency.

We should link to the reference header (this one?) in the doc comments of this struct.

[zyuiop]

@mkroening it should be good now, I rebased and my new tests pass :)

edit: not sure what the failing test is nor why it fails...

[github-actions]

Benchmark Results

Benchmark	Current: 2fb90fa	Previous: 8ed91fe	Performance Ratio
startup_benchmark Build Time	90.33 s	89.11 s	1.01
startup_benchmark File Size	0.90 MB	0.90 MB	1.00
Startup Time - 1 core	0.89 s (±0.03 s)	0.91 s (±0.02 s)	0.98
Startup Time - 2 cores	0.90 s (±0.02 s)	0.91 s (±0.02 s)	1.00
Startup Time - 4 cores	0.91 s (±0.02 s)	0.92 s (±0.02 s)	0.99
multithreaded_benchmark Build Time	91.75 s	87.92 s	1.04
multithreaded_benchmark File Size	1.00 MB	1.00 MB	1.00
Multithreaded Pi Efficiency - 2 Threads	3.41 % (±16.38 %)	3.44 % (±16.53 %)	0.99
Multithreaded Pi Efficiency - 4 Threads	1.78 % (±8.53 %)	1.56 % (±7.47 %)	1.14
Multithreaded Pi Efficiency - 8 Threads	0.98 % (±4.69 %)	0.97 % (±4.67 %)	1.00
micro_benchmarks Build Time	92.08 s	94.61 s	0.97
micro_benchmarks File Size	1.00 MB	1.00 MB	1.00
Scheduling time - 1 thread	2.15 ticks (±10.30 ticks)	2.64 ticks (±12.67 ticks)	0.81
Scheduling time - 2 threads	1.30 ticks (±6.25 ticks)	1.41 ticks (±6.78 ticks)	0.92
Micro - Time for syscall (getpid)	0.60 ticks (±2.88 ticks)	0.63 ticks (±3.03 ticks)	0.95
Memcpy speed - (built_in) block size 4096	1516.99 MByte/s (±7281.55 MByte/s)	1370.61 MByte/s (±6578.95 MByte/s)	1.11
Memcpy speed - (built_in) block size 1048576	746.66 MByte/s (±3583.96 MByte/s)	705.85 MByte/s (±3388.10 MByte/s)	1.06
Memcpy speed - (built_in) block size 16777216	219.68 MByte/s (±1054.45 MByte/s)	222.48 MByte/s (±1067.91 MByte/s)	0.99
Memset speed - (built_in) block size 4096	1846.15 MByte/s (±8861.54 MByte/s)	1818.18 MByte/s (±8727.27 MByte/s)	1.02
Memset speed - (built_in) block size 1048576	1003.56 MByte/s (±4817.10 MByte/s)	998.03 MByte/s (±4790.54 MByte/s)	1.01
Memset speed - (built_in) block size 16777216	969.86 MByte/s (±4655.34 MByte/s)	969.61 MByte/s (±4654.13 MByte/s)	1.00
Memcpy speed - (rust) block size 4096	1379.31 MByte/s (±6620.69 MByte/s)	1395.35 MByte/s (±6697.67 MByte/s)	0.99
Memcpy speed - (rust) block size 1048576	736.17 MByte/s (±3533.63 MByte/s)	760.62 MByte/s (±3650.95 MByte/s)	0.97
Memcpy speed - (rust) block size 16777216	220.62 MByte/s (±1058.96 MByte/s)	207.47 MByte/s (±995.86 MByte/s)	1.06
Memset speed - (rust) block size 4096	1818.18 MByte/s (±8727.27 MByte/s)	1558.44 MByte/s (±7480.52 MByte/s)	1.17
Memset speed - (rust) block size 1048576	990.15 MByte/s (±4752.71 MByte/s)	966.16 MByte/s (±4637.57 MByte/s)	1.02
Memset speed - (rust) block size 16777216	760.29 MByte/s (±3649.40 MByte/s)	906.89 MByte/s (±4353.08 MByte/s)	0.84
alloc_benchmarks Build Time	86.01 s	97.22 s	0.88
alloc_benchmarks File Size	0.97 MB	0.97 MB	1.00
Allocations - Allocation success	2.00 % (±13.86 %)	2.00 % (±13.86 %)	1
Allocations - Deallocation success	1.41 % (±9.74 %)	1.40 % (±9.68 %)	1.01
Allocations - Pre-fail Allocations	2.00 % (±13.86 %)	2.00 % (±13.86 %)	1
Allocations - Average Allocation time	234.19 Ticks (±1622.89 Ticks)	256.15 Ticks (±1775.01 Ticks)	0.91
Allocations - Average Allocation time (no fail)	234.19 Ticks (±1622.89 Ticks)	256.15 Ticks (±1775.01 Ticks)	0.91
Allocations - Average Deallocation time	15.33 Ticks (±106.25 Ticks)	17.50 Ticks (±121.27 Ticks)	0.88
mutex_benchmark Build Time	85.74 s	94.97 s	0.90
mutex_benchmark File Size	1.01 MB	1.01 MB	1.00
Mutex Stress Test Average Time per Iteration - 1 Threads	0.26 ns (±1.80 ns)	0.32 ns (±2.22 ns)	0.81
Mutex Stress Test Average Time per Iteration - 2 Threads	0.32 ns (±2.22 ns)	0.38 ns (±2.63 ns)	0.84

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

[mkroening]

edit: not sure what the failing test is nor why it fails...

Firecracker is a bit flaky on CI nowadays, unfortunately.

[mkroening]

This looks good to me and I am personally fine with using ioctl for this.

@stlankes, @jounathaen, are you fine with us pivoting to ioctl for controlling drivers?