feat: add more lazy removal of instructions with Instruction::noop

[TomAFrench]

Description

Problem*

Followup to #6899

Summary*

Additional Context

Documentation*

Check one:

• No documentation needed.
• Documentation included in this PR.
• [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist*

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt on default settings.

[github-actions]

Execution Memory Report

Program	Peak Memory
keccak256	74.67M
workspace	123.69M
regression_4709	316.00M
ram_blowup_regression	512.61M
rollup-base-public	478.82M
rollup-base-private	325.41M
private-kernel-tail	180.42M
private-kernel-reset	245.11M
private-kernel-inner	208.50M

[github-actions]

Compilation Memory Report

Program	Peak Memory
keccak256	78.13M
workspace	123.72M
regression_4709	424.09M
ram_blowup_regression	1.58G
rollup-base-public	4.85G
rollup-base-private	1.26G
private-kernel-tail	207.14M
private-kernel-reset	669.23M
private-kernel-inner	294.37M

[github-actions]

Compilation Report

Program	Compilation Time	%
sha256_regression	1.290s	4%
regression_4709	0.828s	7%
ram_blowup_regression	18.140s	-6%
rollup-root	3.522s	5%
rollup-merge	2.114s	14%
rollup-block-merge	3.718s	11%
rollup-base-public	43.880s	5%
rollup-base-private	13.400s	0%
private-kernel-tail	0.972s	0%
private-kernel-reset	7.352s	-3%
private-kernel-inner	2.038s	2%

[github-actions]

Execution Report

Program	Execution Time	%
sha256_regression	0.051s	-2%
regression_4709	0.001s	0%
ram_blowup_regression	0.603s	-1%
rollup-root	0.105s	0%
rollup-merge	0.007s	16%
rollup-block-merge	0.106s	1%
rollup-base-public	1.234s	0%
rollup-base-private	0.452s	0%
private-kernel-tail	0.019s	0%
private-kernel-reset	0.313s	0%
private-kernel-inner	0.068s	0%

[jfecher]

Interesting this looks like it is either increasing compilation times or has no substantial effect

[TomAFrench]

Yeah in hindsight the benefits from avoiding rebuilding the instructions vec at the end of these passes are probably outweighed by the costs of having a bloated input to the next pass