Simplify and optimize wasmi codegen

[Robbepop]

Due to historical reasons the wasmi codegen that converts from Wasm bytecode to its internal wasmi bytecode still performed many checks that have already been covered by the wasmparser based Wasm validation.
This PR eliminates all or most of those unnecessary duplicate checks for some reasonable speedups in translation time.

Locally I was able to see roughly 15% speedup in the translation benchmark.

In detail this PR

• Removes the emulated ValueStack during translation time and replaces it with a glorified ValueStackHeight counter.
• Simplifies the LocalsRegistry to no longer deal with local variable types so that it also became a glorified counter.
• Removes tons of type checks during translation time. These are unnecessary since the Wasm validation already performs them.

[paritytech-cicd-pr]

BENCHMARKS

	NATIVE			WASMTIME
BENCHMARK	MASTER	PR	DIFF	MASTER	PR	DIFF	WASMTIME OVERHEAD
execute/ bare_call_0	1.29ms	1.26ms	🟢 -2.13%	1.14ms	1.05ms	🟢 -8.13%	🟢 -17%
execute/ bare_call_0/typed	705.24µs	704.49µs	⚪ -0.11%	532.10µs	530.15µs	⚪ -0.42%	🟢 -25%
execute/ bare_call_1	1.40ms	1.40ms	🟢 -0.45%	1.37ms	1.33ms	🟢 -3.09%	🟢 -6%
execute/ bare_call_16	3.05ms	2.90ms	🟢 -4.91%	5.03ms	4.87ms	🟢 -3.16%	🟡 68%
execute/ bare_call_16/typed	1.29ms	1.29ms	⚪ -0.27%	2.54ms	2.50ms	🟢 -1.27%	🟡 94%
execute/ bare_call_1/typed	802.24µs	811.85µs	⚪ 1.11%	826.24µs	819.05µs	⚪ -0.90%	🟢 1%
execute/ bare_call_4	1.84ms	1.75ms	🟢 -4.02%	2.13ms	2.07ms	🟢 -2.72%	🟢 18%
execute/ bare_call_4/typed	888.86µs	888.44µs	⚪ -0.07%	996.38µs	984.87µs	⚪ -1.14%	🟢 11%
execute/ br_table	918.88µs	918.97µs	⚪ -0.01%	1.07ms	1.06ms	⚪ -1.38%	🟢 16%
execute/ count_until	887.63µs	886.63µs	⚪ -0.10%	2.27ms	2.29ms	⚪ 0.88%	🔴 159%
execute/ factorial_iterative	363.61µs	363.55µs	⚪ -0.05%	928.15µs	920.72µs	⚪ -0.91%	🔴 153%
execute/ factorial_recursive	743.94µs	773.48µs	🔴 4.07%	1.57ms	1.57ms	⚪ -0.10%	🔴 102%
execute/ fib_iterative	1.82ms	1.79ms	🟢 -1.73%	4.94ms	4.94ms	⚪ 0.00%	🔴 176%
execute/ fib_recursive	6.96ms	7.01ms	⚪ 0.59%	14.44ms	14.58ms	⚪ 1.01%	🔴 108%
execute/ global_bump	1.44ms	1.42ms	⚪ -0.77%	3.57ms	3.61ms	🔴 1.12%	🔴 154%
execute/ host_calls	32.37µs	33.43µs	⚪ 2.13%	45.06µs	44.99µs	⚪ -0.10%	🟢 35%
execute/ memory_fill	1.54ms	1.53ms	⚪ -0.07%	4.13ms	4.13ms	⚪ 0.00%	🔴 169%
execute/ memory_sum	1.52ms	1.54ms	⚪ 1.17%	4.15ms	4.16ms	⚪ -0.03%	🔴 171%
execute/ memory_vec_add	3.07ms	3.04ms	⚪ -1.12%	8.71ms	8.72ms	⚪ 0.19%	🔴 187%
execute/ recursive_is_even	1.55ms	1.33ms	🟢 -14.16%	2.60ms	2.61ms	⚪ 0.13%	🟡 96%
execute/ recursive_ok	184.16µs	175.53µs	🟢 -4.44%	358.47µs	359.00µs	⚪ 0.12%	🔴 105%
execute/ recursive_scan	237.35µs	219.92µs	🟢 -7.31%	479.66µs	465.02µs	🟢 -2.96%	🔴 111%
execute/ recursive_trap	18.71µs	17.37µs	🟢 -8.56%	35.53µs	35.90µs	⚪ 0.85%	🔴 107%
execute/ regex_redux	658.42µs	655.63µs	⚪ -0.32%	1.64ms	1.61ms	🟢 -1.26%	🔴 146%
execute/ rev_complement	598.93µs	591.38µs	⚪ -0.94%	1.61ms	1.53ms	🟢 -4.75%	🔴 159%
execute/ tiny_keccak	452.38µs	478.39µs	🔴 5.83%	1.47ms	1.39ms	🟢 -5.71%	🔴 190%
execute/ trunc_f2i	1.07ms	1.07ms	⚪ 0.19%	2.61ms	2.61ms	⚪ -0.34%	🔴 143%
instantiate/ wasm_kernel	69.99µs	71.27µs	⚪ -0.20%	94.65µs	125.47µs	🔴 32.59%	🟡 76%
translate/ wasm_kernel	4.95ms	4.14ms	🟢 -16.51%	10.28ms	8.70ms	🟢 -15.27%	🔴 110%

Link to pipeline