(opt): skip specializing calls that would be inlined

[orizi]

During the interleaved inlining/const-folding pass, a call with constant
arguments to a function that would be inlined was first specialized - paying
a full casm size estimation of the specialized function in
priv_should_specialize - and the specialized function was then immediately
inlined anyway (a specialized function inherits its base's inline
configuration, and a small base's specialization is smaller yet). Inlining
the base directly and const folding in place reaches equivalent code without
the extra estimation and specialized function lowering.

Skip specialization for such calls. As every statement is checked for
inlining right after its const-folding visit, such calls are always inlined
by the interleaved pass and never survive to the standalone const-folding
pass.

Improves full-compilation time by ~4.5% (measured on top of the size
estimation memoization). Emitted code is equivalent: e2e and contract
outputs are unchanged, and debug locations lose a redundant self-inlining
frame.

[reviewable-StarkWare]

This change is 

[orizi]

• (opt): skip specializing calls that would be inlined #10350 👈 (View in Graphite)
• (opt): fold struct destructure of a constant into member constants #10357
• (opt): reprocess replaced block ends for inlining #10349
• (opt): memoize function code size estimation #10348
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[TomerStarkware]

@TomerStarkware reviewed 7 files and all commit messages, and made 1 comment.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on eytan-starkware).

[TomerStarkware]

@TomerStarkware reviewed 1 file and all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on eytan-starkware).