Add bytecode generate process that removing unused function parameters

[kwonjeomsim]

This PR removes some bytecodes that generate function parameters which are not used during execution.
I add this feature in release mode only because there are some test failed in debug mode.

[clover2123]

@kwonjeomsim
Why did you disable this update in debug mode?
FYI) debug mode and Debugger are different
Debugger enabled by ESCARGOT_DEBUGGER option is one of features that escargot supports.
Using Debugger, developers could debug (investigate) their JS code like other JS debugger (Ex chrome developer mode)

[kwonjeomsim]

@kwonjeomsim Why did you disable this update in debug mode? FYI) debug mode and Debugger are different Debugger enabled by ESCARGOT_DEBUGGER option is one of features that escargot supports. Using Debugger, developers could debug (investigate) their JS code like other JS debugger (Ex chrome developer mode)

@clover2123
I fixed commit so this will disabled in debugger mode instead of debug mode. Please check again!

[github-actions]

The added if (name == "eval" || name == "arguments") replaces the original check that used scopeCtx->m_hasEval and scopeCtx->m_allowArguments. This causes the function to mark parameters as used for the names "eval" or "arguments" even when the current scope does not actually allow eval or arguments, leading to incorrect parameter usage tracking. Restore the original flag-based check or add a guard that verifies scopeCtx->m_hasEval or scopeCtx->m_allowArguments before marking the parameter as used.

[github-actions]

Replace the while loop with a for loop to make the traversal of the scope chain more idiomatic and reduce boilerplate. The for form for (auto* scope = this->currentScopeContext; scope; scope = scope->m_parent) expresses the same logic in a single line, improving readability. This change keeps the semantics identical and avoids the need for a separate assignment statement. It also aligns with common C++ style for simple pointer iteration.