feat: implement cwd, umask, access, faccessat, ftruncate, fchmod, truncate

[zyuiop]

No description provided.

[stlankes]

You have just to remove the merge conflicts.

[mkroening]

Before merging I'd like to think about whether it makes sense to implement cwd functionality implemented more natively. This is a great starting point for reference, for sure, though, and maybe it would still make sense to merge this first. 🤔

[github-actions]

Benchmark Results

Benchmark	Current: fa03443	Previous: 11432b0	Performance Ratio
startup_benchmark Build Time	90.60 s	84.25 s	1.08
startup_benchmark File Size	0.89 MB	0.87 MB	1.03
Startup Time - 1 core	0.92 s (±0.03 s)	0.86 s (±0.02 s)	1.06
Startup Time - 2 cores	0.89 s (±0.04 s)	0.86 s (±0.03 s)	1.04
Startup Time - 4 cores	0.92 s (±0.03 s)	0.86 s (±0.02 s)	1.07
multithreaded_benchmark Build Time	90.28 s	82.44 s	1.10
multithreaded_benchmark File Size	0.99 MB	0.98 MB	1.01
Multithreaded Pi Efficiency - 2 Threads	3.48 % (±16.68 %)	3.19 % (±15.29 %)	1.09
Multithreaded Pi Efficiency - 4 Threads	1.73 % (±8.31 %)	1.64 % (±7.85 %)	1.06
Multithreaded Pi Efficiency - 8 Threads	0.93 % (±4.45 %)	0.96 % (±4.59 %)	0.97
micro_benchmarks Build Time	85.87 s	81.88 s	1.05
micro_benchmarks File Size	1.00 MB	0.99 MB	1.01
Scheduling time - 1 thread	2.24 ticks (±10.75 ticks)	2.19 ticks (±10.49 ticks)	1.02
Scheduling time - 2 threads	1.18 ticks (±5.66 ticks)	1.63 ticks (±7.82 ticks)	0.72
Micro - Time for syscall (getpid)	0.55 ticks (±2.64 ticks)	0.51 ticks (±2.46 ticks)	1.07
Memcpy speed - (built_in) block size 4096	1795.98 MByte/s (±8620.69 MByte/s)	1775.57 MByte/s (±8522.73 MByte/s)	1.01
Memcpy speed - (built_in) block size 1048576	754.94 MByte/s (±3623.74 MByte/s)	783.05 MByte/s (±3758.66 MByte/s)	0.96
Memcpy speed - (built_in) block size 16777216	215.34 MByte/s (±1033.63 MByte/s)	226.06 MByte/s (±1085.08 MByte/s)	0.95
Memset speed - (built_in) block size 4096	1578.95 MByte/s (±7578.95 MByte/s)	1846.15 MByte/s (±8861.54 MByte/s)	0.86
Memset speed - (built_in) block size 1048576	994.11 MByte/s (±4771.73 MByte/s)	993.86 MByte/s (±4770.54 MByte/s)	1.00
Memset speed - (built_in) block size 16777216	985.52 MByte/s (±4730.49 MByte/s)	956.85 MByte/s (±4592.90 MByte/s)	1.03
Memcpy speed - (rust) block size 4096	1363.64 MByte/s (±6545.45 MByte/s)	1411.76 MByte/s (±6776.47 MByte/s)	0.97
Memcpy speed - (rust) block size 1048576	735.59 MByte/s (±3530.84 MByte/s)	786.63 MByte/s (±3775.81 MByte/s)	0.94
Memcpy speed - (rust) block size 16777216	222.29 MByte/s (±1067.00 MByte/s)	225.87 MByte/s (±1084.17 MByte/s)	0.98
Memset speed - (rust) block size 4096	1967.21 MByte/s (±9442.62 MByte/s)	1818.18 MByte/s (±8727.27 MByte/s)	1.08
Memset speed - (rust) block size 1048576	993.22 MByte/s (±4767.46 MByte/s)	1029.71 MByte/s (±4942.59 MByte/s)	0.96
Memset speed - (rust) block size 16777216	950.75 MByte/s (±4563.58 MByte/s)	880.76 MByte/s (±4227.63 MByte/s)	1.08
alloc_benchmarks Build Time	85.78 s	80.35 s	1.07
alloc_benchmarks File Size	0.96 MB	0.94 MB	1.02
Allocations - Allocation success	2.00 % (±13.86 %)	2.00 % (±13.86 %)	1
Allocations - Deallocation success	1.39 % (±9.65 %)	1.40 % (±9.68 %)	1.00
Allocations - Pre-fail Allocations	2.00 % (±13.86 %)	2.00 % (±13.86 %)	1
Allocations - Average Allocation time	245.92 Ticks (±1704.11 Ticks)	231.88 Ticks (±1606.84 Ticks)	1.06
Allocations - Average Allocation time (no fail)	245.92 Ticks (±1704.11 Ticks)	231.88 Ticks (±1606.84 Ticks)	1.06
Allocations - Average Deallocation time	14.92 Ticks (±103.37 Ticks)	17.14 Ticks (±118.74 Ticks)	0.87
mutex_benchmark Build Time	85.50 s	79.99 s	1.07
mutex_benchmark File Size	1.00 MB	0.98 MB	1.01
Mutex Stress Test Average Time per Iteration - 1 Threads	0.28 ns (±1.94 ns)	0.24 ns (±1.66 ns)	1.17
Mutex Stress Test Average Time per Iteration - 2 Threads	0.32 ns (±2.22 ns)	0.26 ns (±1.80 ns)	1.23

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

[mkroening]

Thanks, this looks excellent for now. I only have one request. Also, do you have newlib patches for this available? It might make sense to merge both at the same time (it would also need to revert hermit-os/newlib#171).

[mkroening]

Can we use the constants from fuse_abi::linux (already imported) on the right side here?

[zyuiop]

Good point, I have opened two PRs on newlib :)

hermit-os/newlib#172

[mkroening]

I only have one more point for discussion before going over this one last time and merging.

[mkroening]

Last round of comments :D

[zyuiop]

Here are my fixes :)

[mkroening]

I will squash the fixup commits and make sure the CI passes for each commit and merge.

Thanks a lot! :)