Implement Wasm tail-call proposal

[Robbepop]

Closes #363.

• Passes 100% of the official Wasm spec testsuite.
• Fix interoperability bug with resumable calls.
  • Since resumable calls expect a frame on the call stack upon call resumption we cannot do a proper tail call when calling host functions in a resumable call mode.

Benchmarks show that for example tail recursive fibonacci is pretty efficient. While it is less efficient than iterative fibonacci it is miles and miles ahead of non-tail recursive fibonacci. Ideally we had more benchmarks. Tested with N = 100_000.

[paritytech-cicd-pr]

BENCHMARKS

	NATIVE			WASMTIME
BENCHMARK	MASTER	PR	DIFF	MASTER	PR	DIFF	WASMTIME OVERHEAD
execute/ bare_call_0	1.51ms	1.89ms	🔴 25.47%	1.29ms	1.47ms	🔴 13.91%	🟢 -22%
execute/ bare_call_0/typed	1.14ms	1.12ms	🟢 -2.13%	919.37µs	952.44µs	🔴 3.58%	🟢 -15%
execute/ bare_call_1	1.56ms	1.85ms	🔴 19.40%	1.38ms	1.52ms	🔴 9.99%	🟢 -18%
execute/ bare_call_16	2.54ms	2.67ms	🔴 5.21%	3.34ms	3.30ms	🟢 -1.25%	🟢 23%
execute/ bare_call_16/typed	1.81ms	1.88ms	🔴 3.89%	1.56ms	1.57ms	⚪ 0.54%	🟢 -17%
execute/ bare_call_1/typed	1.22ms	1.23ms	⚪ 0.61%	1.19ms	1.24ms	🔴 4.13%	🟢 1%
execute/ bare_call_4	1.74ms	1.77ms	🔴 2.47%	1.77ms	1.77ms	🟢 0.13%	🟢 -0%
execute/ bare_call_4/typed	1.30ms	1.27ms	⚪ -0.83%	1.22ms	1.17ms	🟢 -4.36%	🟢 -8%
execute/ br_table	1.32ms	1.29ms	🟢 -2.85%	1.36ms	1.43ms	🔴 4.89%	🟢 11%
execute/ count_until	677.45µs	635.85µs	🟢 -6.28%	1.13ms	1.11ms	🟢 -1.78%	🟡 74%
execute/ factorial_iterative	343.74µs	332.29µs	🟢 -3.47%	476.07µs	485.22µs	🔴 1.88%	🟢 46%
execute/ factorial_recursive	591.79µs	592.66µs	⚪ 0.12%	908.54µs	1.08ms	🔴 18.65%	🟡 82%
execute/ fibonacci_iter	1.47ms	1.47ms	⚪ 0.32%	2.23ms	2.06ms	🟢 -7.29%	🟢 40%
execute/ fibonacci_rec	5.33ms	4.83ms	🟢 -9.61%	8.07ms	7.99ms	⚪ -0.97%	🟡 66%
execute/ fibonacci_tail	1.48ms	1.17ms	🟢 -20.80%	1.96ms	1.70ms	🟢 -13.27%	🟢 45%
execute/ global_bump	818.60µs	824.69µs	⚪ 0.79%	1.56ms	1.56ms	⚪ 0.06%	🟡 90%
execute/ global_const	754.19µs	716.48µs	🟢 -4.96%	1.43ms	1.39ms	🟢 -2.44%	🟡 94%
execute/ host_calls	31.25µs	35.96µs	🔴 15.57%	46.38µs	47.49µs	🔴 2.33%	🟢 32%
execute/ memory_fill	1.20ms	1.21ms	⚪ 0.76%	1.94ms	1.77ms	🟢 -8.32%	🟢 46%
execute/ memory_sum	1.19ms	1.19ms	⚪ -0.44%	1.93ms	1.85ms	🟢 -4.27%	🟡 56%
execute/ memory_vec_add	2.38ms	2.38ms	⚪ -0.06%	4.58ms	4.66ms	🔴 1.84%	🟡 96%
execute/ recursive_is_even	999.76µs	1.05ms	🔴 4.63%	1.67ms	1.65ms	⚪ -0.91%	🟡 58%
execute/ recursive_ok	122.97µs	122.06µs	⚪ -0.72%	202.14µs	200.55µs	⚪ -0.76%	🟡 64%
execute/ recursive_scan	158.39µs	156.82µs	⚪ -0.96%	260.87µs	250.00µs	🟢 -3.81%	🟡 59%
execute/ recursive_trap	12.50µs	12.59µs	⚪ 0.25%	21.52µs	20.82µs	🟢 -3.36%	🟡 65%
execute/ regex_redux	521.92µs	518.59µs	⚪ -0.72%	913.04µs	916.05µs	⚪ 0.27%	🟡 77%
execute/ rev_complement	465.14µs	458.00µs	⚪ -1.26%	775.23µs	776.19µs	⚪ 0.07%	🟡 69%
execute/ tiny_keccak	331.67µs	330.80µs	⚪ 0.05%	603.16µs	576.31µs	🟢 -4.30%	🟡 74%
execute/ trunc_f2i	775.88µs	786.44µs	🔴 1.36%	1.40ms	1.38ms	🟢 -1.82%	🟡 76%
instantiate/ wasm_kernel	54.50µs	54.19µs	⚪ -1.49%	53.65µs	54.89µs	⚪ -0.32%	🟢 1%
translate/ erc1155	235.69µs	233.39µs	⚪ -1.06%	441.79µs	428.99µs	🟢 -2.93%	🟡 84%
translate/ erc20	114.17µs	113.34µs	⚪ -0.60%	212.64µs	207.07µs	🟢 -2.62%	🟡 83%
translate/ erc721	160.71µs	162.21µs	⚪ 0.61%	310.43µs	300.98µs	🟢 -3.04%	🟡 86%
translate/ spidermonkey	0.00ns	0.00ns	🔴 1.50%	0.00ns	0.00ns	🟢 -3.80%	🟢 0%
translate/ wasm_kernel	4.07ms	4.05ms	⚪ -0.61%	8.25ms	7.96ms	🟢 -3.42%	🟡 96%

Link to pipeline