Pre-load default memory bytes in InstanceCache

[Robbepop]

No description provided.

[paritytech-cicd-pr]

BENCHMARKS

	NATIVE			WASMTIME
BENCHMARK	MASTER	PR	DIFF	MASTER	PR	DIFF	WASMTIME OVERHEAD
execute/ bare_call_0	1.52ms	2.01ms	🔴 32.48%	1.10ms	1.31ms	🔴 19.22%	🟢 -35%
execute/ bare_call_0/typed	1.17ms	1.41ms	🔴 20.44%	725.84µs	953.47µs	🔴 31.34%	🟢 -32%
execute/ bare_call_1	1.59ms	2.04ms	🔴 28.59%	1.26ms	1.58ms	🔴 24.98%	🟢 -23%
execute/ bare_call_16	2.44ms	2.57ms	🔴 5.85%	3.24ms	3.47ms	🔴 7.01%	🟢 35%
execute/ bare_call_16/typed	1.55ms	1.70ms	🔴 10.19%	1.66ms	1.84ms	🔴 10.50%	🟢 8%
execute/ bare_call_1/typed	1.23ms	1.51ms	🔴 22.19%	1.04ms	1.25ms	🔴 20.60%	🟢 -17%
execute/ bare_call_4	1.75ms	2.05ms	🔴 17.78%	1.72ms	1.84ms	🔴 6.59%	🟢 -10%
execute/ bare_call_4/typed	1.20ms	1.68ms	🔴 39.91%	982.35µs	1.09ms	🔴 11.15%	🟢 -35%
execute/ br_table	1.37ms	1.50ms	🔴 9.43%	1.27ms	1.56ms	🔴 23.54%	🟢 4%
execute/ count_until	574.70µs	620.22µs	🔴 7.97%	1.36ms	1.30ms	🟢 -4.27%	🔴 110%
execute/ factorial_iterative	362.79µs	321.64µs	🟢 -11.35%	517.57µs	541.53µs	🔴 4.59%	🟡 68%
execute/ factorial_recursive	496.58µs	503.37µs	🔴 1.37%	683.55µs	757.72µs	🔴 10.86%	🟡 51%
execute/ fibonacci_iter	1.39ms	1.39ms	⚪ 0.21%	2.64ms	2.61ms	🟢 -1.12%	🟡 87%
execute/ fibonacci_rec	3.97ms	4.16ms	🔴 4.81%	6.22ms	6.49ms	🔴 4.33%	🟡 56%
execute/ fibonacci_tail	860.01µs	868.27µs	⚪ 0.80%	1.58ms	1.50ms	🟢 -5.42%	🟡 72%
execute/ global_bump	728.81µs	697.81µs	🟢 -3.94%	1.60ms	1.60ms	⚪ 0.32%	🔴 130%
execute/ global_const	681.19µs	660.29µs	🟢 -3.15%	1.39ms	1.40ms	⚪ 0.52%	🔴 111%
execute/ host_calls	36.95µs	52.71µs	🔴 43.42%	39.45µs	46.64µs	🔴 18.41%	🟢 -12%
execute/ memory_fill	1.15ms	1.15ms	⚪ -0.47%	2.31ms	2.23ms	🟢 -3.46%	🟡 94%
execute/ memory_sum	1.13ms	1.15ms	⚪ 1.05%	2.28ms	2.17ms	🟢 -4.94%	🟡 89%
execute/ memory_vec_add	2.35ms	2.31ms	🟢 -1.59%	4.81ms	4.52ms	🟢 -5.95%	🟡 96%
execute/ recursive_is_even	667.59µs	693.00µs	🔴 3.75%	990.23µs	1.06ms	🔴 7.00%	🟡 52%
execute/ recursive_ok	93.79µs	97.72µs	🔴 4.28%	142.20µs	157.48µs	🔴 10.82%	🟡 61%
execute/ recursive_scan	129.63µs	135.34µs	🔴 4.53%	198.11µs	213.92µs	🔴 7.90%	🟡 58%
execute/ recursive_trap	8.73µs	8.83µs	🔴 1.22%	13.79µs	14.84µs	🔴 7.51%	🟡 68%
execute/ regex_redux	455.78µs	471.11µs	🔴 3.33%	837.22µs	841.84µs	⚪ 0.58%	🟡 79%
execute/ rev_complement	422.93µs	422.43µs	⚪ -0.09%	811.28µs	798.48µs	🟢 -1.65%	🟡 89%
execute/ tiny_keccak	320.27µs	323.40µs	⚪ 1.27%	720.81µs	642.14µs	🟢 -10.74%	🟡 99%
execute/ trunc_f2i	728.02µs	730.94µs	⚪ 0.43%	1.53ms	1.57ms	🔴 2.06%	🔴 114%
instantiate/ wasm_kernel	55.06µs	55.71µs	⚪ 0.93%	56.75µs	53.92µs	🟢 -5.06%	🟢 -3%
translate/ erc1155	208.95µs	208.22µs	⚪ -0.36%	366.02µs	365.18µs	⚪ 0.06%	🟡 75%
translate/ erc20	103.42µs	103.56µs	⚪ 0.47%	177.09µs	176.59µs	⚪ -0.26%	🟡 71%
translate/ erc721	147.33µs	146.66µs	⚪ -0.16%	256.33µs	256.27µs	⚪ -0.08%	🟡 75%
translate/ spidermonkey	64.45ms	64.72ms	⚪ 0.35%	0.00ns	0.00ns	⚪ -0.11%	🟢 -100%
translate/ wasm_kernel	4.22ms	4.20ms	⚪ 0.23%	6.61ms	6.64ms	⚪ 0.21%	🟡 58%

Link to pipeline

[Robbepop]

Optimization experiment was indeterminate. Wasm tiny_keccak improved by 10% while call heavy benchmarks and in particular host function calls have regressed significantly in both native and Wasm.