Add Instruction and InstructionIterator

[HalidOdat]

This PR adds a fat enum that represents a full instruction (opcode + operands) and an iterator that takes bytecode and produces instructions.

I implemented something like this in #3037, this PR just extracts and refines it.

Why is this useful/needed?

• It allows us to iterate over instructions without worrying, whether we are reading the operands correctly. Currently we do the same thing in instruction_operands() and flowgraph generation, this is duplicate code and can easily lead to mistakes (also doing it in Implement control flow graph construction #3037 )
• Extracting the operands is harder when manually accessing bytecode.
• If we have a list of instructions it's easier to do changes in it.

Note: This is not intended to replace the compact bytecode representation, that is optimized for execution and storage.

... This also makes it easier to implement an bytecode size optimization on operands, that I'm working on ;)

[github-actions]

Test262 conformance changes

Test result	main count	PR count	difference
Total	95,368	95,368	0
Passed	75,046	75,046	0
Ignored	19,302	19,302	0
Failed	1,020	1,020	0
Panics	0	0	0
Conformance	78.69%	78.69%	0.00%

[codecov]

Codecov Report

Merging #3201 (cfe700b) into main (dda713d) will decrease coverage by 0.09%.
The diff coverage is 1.88%.

@@            Coverage Diff             @@
##             main    #3201      +/-   ##
==========================================
- Coverage   50.42%   50.34%   -0.09%     
==========================================
  Files         436      436              
  Lines       42549    42618      +69     
==========================================
  Hits        21457    21457              
- Misses      21092    21161      +69     

Files Changed	Coverage Δ
boa_engine/src/vm/call_frame/mod.rs	90.47% <ø> (ø)
boa_engine/src/vm/code_block.rs	52.99% <0.00%> (-1.51%)	⬇️
boa_engine/src/vm/flowgraph/mod.rs	0.00% <0.00%> (ø)
boa_engine/src/vm/mod.rs	62.50% <ø> (ø)
boa_engine/src/vm/opcode/mod.rs	7.03% <ø> (-62.20%)	⬇️
boa_engine/src/bytecompiler/mod.rs	66.58% <100.00%> (ø)
boa_engine/src/vm/opcode/control_flow/jump.rs	92.68% <100.00%> (ø)

[jedel1043]

Really nice improvement! It simplifies our logic a LOT.
Any plans for extending it to Operation implementers or is it just not worth?

[HalidOdat]

Any plans for extending it to Operation implementers or is it just not worth?

No, this representation is intended for easier bytecode analysis, not really execution, the compact bytecode version is smaller and faster.

[nekevss]

Should we start writing implementation notes to break down macros?

[nekevss]

Overall looks good to me. Nice work! Just one thought I had while reviewing the macro.

[jasonwilliams]

Whats the affect on performance with this change?

[HalidOdat]

Whats the affect on performance with this change?

There is no change in performance, this is doesn't change the way we execute bytecode. It only implements an iterator and the result being an Instruction (opcode + operands). For easier iteration of bytecode, without worrying about the size of operands.