saga4/fix optimizable functions

[Saga4]

User description

• Add no-pr argument to CF
• Fix Optimizable Functions issue and add error logs

---

PR Type

Enhancement, Bug fix

---

Description

• Add logs around optimizable functions calls

• Save and restore optimizer.args state

• Handle missing current function errors

• Disable PRs in LSP server initialization

---

Changes diagram

flowchart LR
  Request["get_optimizable_functions request"]
  SetArgs["Set temporary optimizer.args"]
  CallGet["Call get_optimizable_functions"]
  BuildMap["Build path-to-names map"]
  Restore["Restore original optimizer.args"]
  Request --> SetArgs
  SetArgs --> CallGet
  CallGet --> BuildMap
  BuildMap --> Restore
  Restore --> Response["Return functions map"]

Loading

---

Changes walkthrough 📝

	Relevant files
Error handling	beta.py Enhance logging and state safety in beta                                  codeflash/lsp/beta.py Added info logs for optimization flow steps Saved/restored optimizer.args to prevent corruption Added error handling for missing functions Cleared function filter post-optimization +59/-8
Configuration changes	server.py Disable PR creation in LSP server                                                codeflash/lsp/server.py Set args.no_pr = True in optimizer init +1/-0

---

Need help?

Type /help how to ... in the comments thread for any questions about PR-Agent usage.

Check out the documentation for more information.

[github-actions]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 3 🔵🔵🔵⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Recommended focus areas for review State Restoration The restoration logic in get_optimizable_functions does not explicitly reset optimizer.args.file and previous_checkpoint_functions when their original values are None, potentially leaving stale state. finally: # Restore original args to prevent state corruption if original_file is not None: server.optimizer.args.file = original_file if original_function is not None: server.optimizer.args.function = original_function else: server.optimizer.args.function = None if original_checkpoint is not None: server.optimizer.args.previous_checkpoint_functions = original_checkpoint State Leakage The early exit in perform_function_optimization when no current function is set does not clear optimizer.args.function, causing it to persist across requests. if not current_function: server.show_message_log(f"No current function being optimized for {params.functionName}", "Error") return {"functionName": params.functionName, "status": "error", "message": "No function currently being optimized"} Verbose Logging A large number of show_message_log calls may flood logs and impact performance; consider reducing verbosity or batching logs. server.show_message_log(f"Getting optimizable functions for: {file_path}", "Info") # Save original args to restore later original_file = getattr(server.optimizer.args, 'file', None) original_function = getattr(server.optimizer.args, 'function', None) original_checkpoint = getattr(server.optimizer.args, 'previous_checkpoint_functions', None) server.show_message_log(f"Original args - file: {original_file}, function: {original_function}", "Info") try: # Set temporary args for this request only server.optimizer.args.file = file_path server.optimizer.args.function = None # Always get ALL functions, not just one server.optimizer.args.previous_checkpoint_functions = False server.show_message_log("Calling get_optimizable_functions...", "Info") optimizable_funcs, _ = server.optimizer.get_optimizable_functions() path_to_qualified_names = {} for path, functions in optimizable_funcs.items(): path_to_qualified_names[path.as_posix()] = [func.qualified_name for func in functions] server.show_message_log(f"Found {len(path_to_qualified_names)} files with functions: {path_to_qualified_names}", "Info") return path_to_qualified_names finally: # Restore original args to prevent state corruption if original_file is not None: server.optimizer.args.file = original_file if original_function is not None: server.optimizer.args.function = original_function else: server.optimizer.args.function = None if original_checkpoint is not None: server.optimizer.args.previous_checkpoint_functions = original_checkpoint server.show_message_log(f"Restored args - file: {server.optimizer.args.file}, function: {server.optimizer.args.function}", "Info")

[github-actions]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
Possible issue	Restore optimizer args unconditionally Always restore all three optimizer arguments unconditionally so that even None values are reapplied and no state leaks occur. Remove the conditional checks and assign back the saved originals directly. codeflash/lsp/beta.py [63-72] finally: # Restore original args to prevent state corruption - if original_file is not None: - server.optimizer.args.file = original_file - if original_function is not None: - server.optimizer.args.function = original_function - else: - server.optimizer.args.function = None - if original_checkpoint is not None: - server.optimizer.args.previous_checkpoint_functions = original_checkpoint + server.optimizer.args.file = original_file + server.optimizer.args.function = original_function + server.optimizer.args.previous_checkpoint_functions = original_checkpoint Suggestion importance[1-10]: 7 __ Why: The unconditional restoration ensures that args.previous_checkpoint_functions doesn't remain mutated when the original was None, preventing state leaks and aligning with how the other args are handled.	Medium
	Fully reset optimizer filters Also clear the file and previous_checkpoint_functions filters after optimization to fully reset optimizer state. This prevents leftover filters from affecting subsequent calls. codeflash/lsp/beta.py [272-274] -# CRITICAL: Clear the function filter after optimization to prevent state corruption +# CRITICAL: Clear optimizer filters after optimization to prevent state corruption server.optimizer.args.function = None -server.show_message_log("Cleared function filter to prevent state corruption", "Info") +server.optimizer.args.file = None +server.optimizer.args.previous_checkpoint_functions = False +server.show_message_log("Cleared optimizer filters to prevent state corruption", "Info") Suggestion importance[1-10]: 6 __ Why: Clearing file and previous_checkpoint_functions after optimization prevents residual state from affecting future calls, improving robustness in subsequent operations.	Low

[KRRT7]

what cause this fix to be needed?

[Saga4]

what cause this fix to be needed?

The LSP server was modifying shared state of the object -> server.optimizer.args, without actually restoring it. When one request set args.function to optimize a specific function, that filter remained active for all subsequent requests, causing them to return incorrect results and thats why we just starting each request with clean state right now.
The long term solution would be to fix the mutable state changes across all allocating locations in code or make each request creating own request args than sharing it across.