fix: Proper network wakers

[Gelbpunkt]

This refactors our timer interrupt handling into a separate module that all timer interrupts are configured through and handled by. It allows us to track why a timer interrupt was fired.

We make use of this to now only conditionally wake the network task's waker when necessary. This is either because we sent some network packets, a network device driver received an interrupt, or because a timer interrupt that we configured so we can poll smoltcp in the future was handled.

Closes #2126.

[github-actions]

Benchmark Results

Details

Benchmark	Current: d5f9bef	Previous: 4e2eb9e	Performance Ratio
startup_benchmark Build Time	99.72 s	98.85 s	1.01 ❗
startup_benchmark File Size	0.87 MB	0.87 MB	1.00 ❗
Startup Time - 1 core	0.93 s (±0.03 s)	0.93 s (±0.02 s)	1.00
Startup Time - 2 cores	0.94 s (±0.03 s)	0.94 s (±0.03 s)	0.99
Startup Time - 4 cores	0.97 s (±0.03 s)	0.97 s (±0.03 s)	1.00
multithreaded_benchmark Build Time	96.32 s	101.75 s	0.95 ❗
multithreaded_benchmark File Size	0.96 MB	0.97 MB	1.00 ❗
Multithreaded Pi Efficiency - 2 Threads	86.37 % (±10.70 %)	88.94 % (±9.28 %)	0.97
Multithreaded Pi Efficiency - 4 Threads	44.39 % (±3.72 %)	43.84 % (±4.40 %)	1.01
Multithreaded Pi Efficiency - 8 Threads	25.38 % (±2.39 %)	25.07 % (±1.98 %)	1.01
micro_benchmarks Build Time	93.09 s	100.40 s	0.93 ❗
micro_benchmarks File Size	0.97 MB	0.97 MB	1.00 ❗
Scheduling time - 1 thread	65.31 ticks (±3.15 ticks)	69.09 ticks (±3.48 ticks)	0.95
Scheduling time - 2 threads	36.32 ticks (±5.00 ticks)	40.35 ticks (±6.50 ticks)	0.90
Micro - Time for syscall (getpid)	3.18 ticks (±0.45 ticks)	3.35 ticks (±0.55 ticks)	0.95
Memcpy speed - (built_in) block size 4096	68886.45 MByte/s (±49215.60 MByte/s)	60564.43 MByte/s (±43274.44 MByte/s)	1.14
Memcpy speed - (built_in) block size 1048576	29769.09 MByte/s (±24441.63 MByte/s)	29279.03 MByte/s (±24148.63 MByte/s)	1.02
Memcpy speed - (built_in) block size 16777216	28839.34 MByte/s (±23981.31 MByte/s)	25925.51 MByte/s (±21792.33 MByte/s)	1.11
Memset speed - (built_in) block size 4096	70040.68 MByte/s (±49941.51 MByte/s)	61139.18 MByte/s (±43619.65 MByte/s)	1.15
Memset speed - (built_in) block size 1048576	30555.72 MByte/s (±24887.92 MByte/s)	30025.71 MByte/s (±24565.41 MByte/s)	1.02
Memset speed - (built_in) block size 16777216	29615.78 MByte/s (±24421.43 MByte/s)	26704.16 MByte/s (±22275.17 MByte/s)	1.11
Memcpy speed - (rust) block size 4096	60951.24 MByte/s (±45075.93 MByte/s)	60319.96 MByte/s (±44078.65 MByte/s)	1.01
Memcpy speed - (rust) block size 1048576	29659.45 MByte/s (±24392.31 MByte/s)	29314.22 MByte/s (±24161.97 MByte/s)	1.01
Memcpy speed - (rust) block size 16777216	28861.21 MByte/s (±24012.55 MByte/s)	24865.71 MByte/s (±20917.45 MByte/s)	1.16
Memset speed - (rust) block size 4096	62074.50 MByte/s (±45808.90 MByte/s)	60870.22 MByte/s (±44412.47 MByte/s)	1.02
Memset speed - (rust) block size 1048576	30433.79 MByte/s (±24837.14 MByte/s)	30010.85 MByte/s (±24572.25 MByte/s)	1.01
Memset speed - (rust) block size 16777216	29637.88 MByte/s (±24452.93 MByte/s)	25601.25 MByte/s (±21378.75 MByte/s)	1.16
alloc_benchmarks Build Time	91.81 s	92.98 s	0.99 ❗
alloc_benchmarks File Size	0.94 MB	0.94 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	10433.65 Ticks (±171.23 Ticks)	9692.28 Ticks (±125.85 Ticks)	1.08 ❗
Allocations - Average Allocation time (no fail)	10433.65 Ticks (±171.23 Ticks)	9692.28 Ticks (±125.85 Ticks)	1.08 ❗
Allocations - Average Deallocation time	983.04 Ticks (±354.22 Ticks)	862.97 Ticks (±109.24 Ticks)	1.14
mutex_benchmark Build Time	91.77 s	92.38 s	0.99 ❗
mutex_benchmark File Size	0.97 MB	0.97 MB	1.00 ❗
Mutex Stress Test Average Time per Iteration - 1 Threads	12.52 ns (±0.61 ns)	12.72 ns (±0.63 ns)	0.98
Mutex Stress Test Average Time per Iteration - 2 Threads	95.30 ns (±4.88 ns)	13.12 ns (±0.97 ns)	7.26 ❗

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

[mkroening]

Thanks a lot; this is looking great! :)

[mkroening]

The docs say that wakeup_time is relative. This was also done wrongly before, but maybe we should follow up to fix this.

[mkroening]

An alternative to duplicating the trace statement everywhere could be using #[track_caller] to log the caller inside WakerRegistration::wake. I am not sure about the performance implications, though.