feat(riscv64): Add PCI support

[sbutz]

Follow-up Tasks:

• PCI Handling is very similar across architectures. Reuse code across architectures where applicable.
• Upgrade crate fdt to version 0.2 once released

[github-actions]

Benchmark Results

Details

Benchmark	Current: 009a6d8	Previous: 2e23902	Performance Ratio
startup_benchmark Build Time	217.89 s	80.34 s	2.71 ❗
startup_benchmark File Size	0.78 MB	0.80 MB	0.99 ❗
Startup Time - 1 core	2.22 s (±0.21 s)	0.75 s (±0.02 s)	2.98 ❗
Startup Time - 2 cores	2.21 s (±0.21 s)	0.74 s (±0.02 s)	3.00 ❗
Startup Time - 4 cores	2.29 s (±0.25 s)	0.74 s (±0.02 s)	3.08 ❗
multithreaded_benchmark Build Time	219.44 s	82.11 s	2.67 ❗
multithreaded_benchmark File Size	0.84 MB	0.86 MB	0.98 ❗
Multithreaded Pi Efficiency - 2 Threads	80.33 % (±13.63 %)	85.89 % (±6.61 %)	0.94
Multithreaded Pi Efficiency - 4 Threads	52.72 % (±9.68 %)	43.43 % (±2.56 %)	1.21
Multithreaded Pi Efficiency - 8 Threads	31.25 % (±5.33 %)	25.76 % (±1.53 %)	1.21
micro_benchmarks Build Time	205.57 s	80.40 s	2.56 ❗
micro_benchmarks File Size	0.85 MB	0.86 MB	0.98 ❗
Scheduling time - 1 thread	152.43 ticks (±34.03 ticks)	62.65 ticks (±4.06 ticks)	2.43 ❗
Scheduling time - 2 threads	92.54 ticks (±22.74 ticks)	34.08 ticks (±4.10 ticks)	2.72 ❗
Micro - Time for syscall (getpid)	9.59 ticks (±4.82 ticks)	3.45 ticks (±0.58 ticks)	2.78 ❗
Memcpy speed - (built_in) block size 4096	54566.20 MByte/s (±39075.43 MByte/s)	82448.38 MByte/s (±56997.13 MByte/s)	0.66
Memcpy speed - (built_in) block size 1048576	16264.68 MByte/s (±14277.99 MByte/s)	30585.98 MByte/s (±24707.84 MByte/s)	0.53
Memcpy speed - (built_in) block size 16777216	13957.74 MByte/s (±11760.94 MByte/s)	26340.06 MByte/s (±21720.96 MByte/s)	0.53
Memset speed - (built_in) block size 4096	54646.07 MByte/s (±39130.83 MByte/s)	82292.76 MByte/s (±56891.50 MByte/s)	0.66
Memset speed - (built_in) block size 1048576	16432.17 MByte/s (±14338.36 MByte/s)	31323.85 MByte/s (±25145.86 MByte/s)	0.52
Memset speed - (built_in) block size 16777216	14233.65 MByte/s (±11888.63 MByte/s)	27104.68 MByte/s (±22209.94 MByte/s)	0.53
Memcpy speed - (rust) block size 4096	54378.21 MByte/s (±38943.03 MByte/s)	74097.96 MByte/s (±51811.44 MByte/s)	0.73
Memcpy speed - (rust) block size 1048576	14580.25 MByte/s (±12514.67 MByte/s)	30361.60 MByte/s (±24602.37 MByte/s)	0.48
Memcpy speed - (rust) block size 16777216	13244.04 MByte/s (±10991.81 MByte/s)	27625.34 MByte/s (±22806.88 MByte/s)	0.48
Memset speed - (rust) block size 4096	55024.00 MByte/s (±39369.93 MByte/s)	74373.47 MByte/s (±51976.48 MByte/s)	0.74
Memset speed - (rust) block size 1048576	14833.27 MByte/s (±12620.15 MByte/s)	31110.89 MByte/s (±25033.24 MByte/s)	0.48
Memset speed - (rust) block size 16777216	13619.05 MByte/s (±11210.01 MByte/s)	28386.93 MByte/s (±23265.03 MByte/s)	0.48
alloc_benchmarks Build Time	196.99 s	74.76 s	2.63 ❗
alloc_benchmarks File Size	0.86 MB	0.87 MB	0.99 ❗
Allocations - Allocation success	91.35 %	91.31 %	1.00 ❗
Allocations - Deallocation success	100.00 %	100.00 %	1
Allocations - Pre-fail Allocations	61.55 %	61.44 %	1.00 ❗
Allocations - Average Allocation time	16295.72 Ticks (±992.36 Ticks)	5860.58 Ticks (±98.43 Ticks)	2.78 ❗
Allocations - Average Allocation time (no fail)	17337.51 Ticks (±1184.78 Ticks)	6554.81 Ticks (±92.86 Ticks)	2.65 ❗
Allocations - Average Deallocation time	4387.58 Ticks (±1245.13 Ticks)	1805.01 Ticks (±250.35 Ticks)	2.43 ❗
mutex_benchmark Build Time	200.15 s	79.82 s	2.51 ❗
mutex_benchmark File Size	0.85 MB	0.86 MB	0.98 ❗
Mutex Stress Test Average Time per Iteration - 1 Threads	31.58 ns (±7.59 ns)	12.10 ns (±0.41 ns)	2.61 ❗
Mutex Stress Test Average Time per Iteration - 2 Threads	30.70 ns (±9.23 ns)	40.26 ns (±1.68 ns)	0.76 ❗

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

[mkroening]

Thanks! Excellent work! :)

Could you enable the RISC-V PCI tests on CI that are currently disabled for RISC-V?

[mkroening]

I think CI needs ixpe-qemu on all architectures now. I have added it.

[mkroening]

RTL8139 does not work, but that is not relevant for this PR. I have excluded them for RISC-V again and opened #2636.

[mkroening]

mio seems broken too, though I am not sure if that is because of PCI, since we don't test it on MMIO.