fix iperf related functions

[cagatay-y]

No description provided.

[github-actions]

Benchmark Results

Details

Benchmark	Current: f69c921	Previous: ce6a236	Performance Ratio
startup_benchmark Build Time	111.56 s	107.35 s	1.04 ❗
startup_benchmark File Size	0.75 MB	0.75 MB	1.00 ❗
Startup Time - 1 core	1.02 s (±0.06 s)	0.93 s (±0.07 s)	1.10 ❗
Startup Time - 2 cores	1.01 s (±0.03 s)	0.94 s (±0.05 s)	1.08 ❗
Startup Time - 4 cores	1.02 s (±0.04 s)	0.99 s (±0.04 s)	1.02
multithreaded_benchmark Build Time	113.92 s	110.67 s	1.03 ❗
multithreaded_benchmark File Size	0.87 MB	0.87 MB	1.00 ❗
Multithreaded Pi Efficiency - 2 Threads	88.30 % (±15.87 %)	92.42 % (±11.43 %)	0.96
Multithreaded Pi Efficiency - 4 Threads	43.23 % (±7.23 %)	45.92 % (±5.46 %)	0.94
Multithreaded Pi Efficiency - 8 Threads	23.84 % (±3.89 %)	25.34 % (±3.17 %)	0.94
micro_benchmarks Build Time	94.78 s	90.29 s	1.05 ❗
micro_benchmarks File Size	0.87 MB	0.87 MB	1.00 ❗
Scheduling time - 1 thread	75.32 ticks (±3.96 ticks)	66.89 ticks (±4.01 ticks)	1.13 ❗
Scheduling time - 2 threads	42.53 ticks (±4.38 ticks)	37.83 ticks (±4.60 ticks)	1.12
Micro - Time for syscall (getpid)	3.97 ticks (±0.22 ticks)	3.51 ticks (±0.24 ticks)	1.13 ❗
Memcpy speed - (built_in) block size 4096	73273.28 MByte/s (±50707.13 MByte/s)	81839.10 MByte/s (±56601.98 MByte/s)	0.90
Memcpy speed - (built_in) block size 1048576	29147.81 MByte/s (±23954.25 MByte/s)	30096.83 MByte/s (±24484.80 MByte/s)	0.97
Memcpy speed - (built_in) block size 16777216	23873.88 MByte/s (±19665.38 MByte/s)	26824.39 MByte/s (±22148.06 MByte/s)	0.89
Memset speed - (built_in) block size 4096	73353.95 MByte/s (±50759.04 MByte/s)	82211.20 MByte/s (±56852.70 MByte/s)	0.89
Memset speed - (built_in) block size 1048576	29947.84 MByte/s (±24420.11 MByte/s)	30820.35 MByte/s (±24892.88 MByte/s)	0.97
Memset speed - (built_in) block size 16777216	24486.04 MByte/s (±20019.36 MByte/s)	27568.62 MByte/s (±22596.81 MByte/s)	0.89
Memcpy speed - (rust) block size 4096	67101.38 MByte/s (±47015.59 MByte/s)	71055.92 MByte/s (±49922.07 MByte/s)	0.94
Memcpy speed - (rust) block size 1048576	29284.98 MByte/s (±24088.67 MByte/s)	30251.51 MByte/s (±24549.45 MByte/s)	0.97
Memcpy speed - (rust) block size 16777216	24214.39 MByte/s (±19898.21 MByte/s)	26489.85 MByte/s (±21826.15 MByte/s)	0.91
Memset speed - (rust) block size 4096	67301.02 MByte/s (±47160.97 MByte/s)	71343.28 MByte/s (±50124.43 MByte/s)	0.94
Memset speed - (rust) block size 1048576	30035.59 MByte/s (±24512.88 MByte/s)	31010.33 MByte/s (±24987.76 MByte/s)	0.97
Memset speed - (rust) block size 16777216	24858.64 MByte/s (±20292.62 MByte/s)	27257.69 MByte/s (±22311.78 MByte/s)	0.91
alloc_benchmarks Build Time	91.06 s	86.24 s	1.06 ❗
alloc_benchmarks File Size	0.83 MB	0.83 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	4031.67 Ticks (±60.54 Ticks)	4014.47 Ticks (±73.89 Ticks)	1.00
Allocations - Average Allocation time (no fail)	4031.67 Ticks (±60.54 Ticks)	4014.47 Ticks (±73.89 Ticks)	1.00
Allocations - Average Deallocation time	662.41 Ticks (±35.08 Ticks)	671.96 Ticks (±52.84 Ticks)	0.99
mutex_benchmark Build Time	93.45 s	88.98 s	1.05 ❗
mutex_benchmark File Size	0.87 MB	0.87 MB	1.00 ❗
Mutex Stress Test Average Time per Iteration - 1 Threads	13.82 ns (±0.68 ns)	13.58 ns (±0.67 ns)	1.02
Mutex Stress Test Average Time per Iteration - 2 Threads	15.66 ns (±9.51 ns)	17.56 ns (±6.54 ns)	0.89

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

[mkroening]

This is great, thanks! I tested it, and it works. I will take a look at the code later.

Two issues I noticed:

1. When Hermit is the server, it can only accept one client. After the first client is done, the next client cannot establish a connection. This does not show an error message, but might be related to the next issue.

2. When Hermit is the client, the test does not end cleanly:

   iperf3: error - unable to send control message - port may not be available, the other side may have stopped running, etc.: Function not implemented
   

These issues do not block this PR at all; I just wanted to document this.

[cagatay-y]

I am not 100% sure if this is the same problem as the one you have with multiple connections on the server, but when I run the server with the --one-off option so that it shuts off after a test is completed, it gets stuck at https://github.com/esnet/iperf/blob/8a94f44cde986106abe8c8bf06d72d327570e34e/src/iperf_server_api.c#L488. I haven't had a chance to investigate further.

I need to check what the second issue is about.

[mkroening]

Thanks, this is great!

I split up the commits a bit and refactored some tiny things. I'll merge this tomorrow. :)