perf: use fixed bytes hashmaps from alloy-core

[gakonst]

Follow-up to #3231 that makes use of alloy-core fixed bytes hashmaps throughout the code for Address, U256, and B256 keys.

Re-implementation of #3238 rebased on main.

[codspeed-hq]

CodSpeed Performance Report

Merging this PR will improve performance by 20.78%

_{Comparing danipopes/alloy-fixed-bytes-maps (0b1ae86) with main (087ac46)}

Summary

⚡ 130 improved benchmarks
✅ 43 untouched benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	ADDMOD_50	28.3 µs	27.2 µs	+3.75%
⚡	CALLDATASIZE_50	18.8 µs	17.8 µs	+5.58%
⚡	ADDRESS_50	18.9 µs	17.9 µs	+6.05%
⚡	ADD_50	24 µs	23 µs	+4.32%
⚡	AND_50	24 µs	23 µs	+4.18%
⚡	CALLVALUE_50	18.8 µs	17.8 µs	+5.75%
⚡	CALL_50	96 µs	90.6 µs	+5.97%
⚡	PUSH13_50	19.3 µs	18.2 µs	+5.93%
⚡	CHAINID_50	18.8 µs	17.8 µs	+5.92%
⚡	PUSH14_50	19.3 µs	18.3 µs	+5.42%
⚡	PUSH15_50	19.4 µs	18.4 µs	+5.55%
⚡	CODESIZE_50	19 µs	17.9 µs	+6.37%
⚡	PUSH16_50	19.3 µs	18.2 µs	+5.6%
⚡	BYTE_50	24.3 µs	23.3 µs	+4%
⚡	PUSH17_50	19.5 µs	18.4 µs	+5.71%
⚡	CLZ_50	24.4 µs	23.3 µs	+4.64%
⚡	COINBASE_50	18.9 µs	17.8 µs	+6.23%
⚡	PUSH1_50	18.8 µs	17.8 µs	+6.09%
⚡	PUSH18_50	19.4 µs	18.4 µs	+5.39%
⚡	PUSH20_50	19.4 µs	18.4 µs	+5.22%
...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

[rakita]

lgtm