-
Notifications
You must be signed in to change notification settings - Fork 26
Vtable Metadata #840
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Vtable Metadata #840
Conversation
Drop shims are correctly implemented with recursion to handle both Array & Tuple cases. * Initial plan * Initial analysis and planning for vtable metadata computation Co-authored-by: ssyram <[email protected]> * Initial analysis and plan for vtable metadata computation Co-authored-by: ssyram <[email protected]> * Implement basic drop case analysis in vtable metadata computation Co-authored-by: ssyram <[email protected]> * Improve drop case analysis with proper Drop trait detection Co-authored-by: ssyram <[email protected]> * Successfully implement null drop operands for types that don't need drop Co-authored-by: ssyram <[email protected]> * Refactor vtable metadata computation to use struct-based approach Co-authored-by: ssyram <[email protected]> * Changes before error encountered Co-authored-by: ssyram <[email protected]> * Changes before error encountered Co-authored-by: ssyram <[email protected]> * Improve drop shim generation with better structured naming and case distinction Co-authored-by: ssyram <[email protected]> * Implement actual drop shim function generation for vtable metadata Co-authored-by: ssyram <[email protected]> * Fix drop shim function generation with proper lifetime handling Co-authored-by: ssyram <[email protected]> * Implement proper dyn trait types and Concretize cast for drop shims Co-authored-by: ssyram <[email protected]> * Set drop shim functions as transparent for better naming display Co-authored-by: ssyram <[email protected]> * Implement generics handling for drop shim functions - first iteration Co-authored-by: ssyram <[email protected]> * Fix drop shim parameter types and implement proper dyn trait handling Co-authored-by: ssyram <[email protected]> * Implement generics handling for drop shim function references Co-authored-by: ssyram <[email protected]> * Refactor code organization and fix drop shim parameter types from pointers to references Co-authored-by: ssyram <[email protected]> * Fix generics handling with proper region variables and minimal generic parameters Co-authored-by: ssyram <[email protected]> * Complete vtable drop shim implementation with proper reference types and simplified generics Co-authored-by: ssyram <[email protected]> * re-implemented layout info * Delete vtables * minor adjust * Implement create_dyn_trait_param_type by extracting from drop field type and handle Unknown case Co-authored-by: ssyram <[email protected]> * fixd drop receiver * Add drop_glue field to TypeDecl for tracking drop implementations (#14) * Initial plan * Add drop_glue field to TypeDecl and implement translation logic Co-authored-by: ssyram <[email protected]> * Fix test case ordering after adding drop_glue field Co-authored-by: ssyram <[email protected]> * Fix panic in combine_span_iter when handling empty iterators Co-authored-by: ssyram <[email protected]> * new ml & no-enqueue --------- Co-authored-by: copilot-swe-agent[bot] <[email protected]> Co-authored-by: ssyram <[email protected]> Co-authored-by: ssyram <[email protected]> * types fixed * Implement proper generics handling for drop shim functions Co-authored-by: ssyram <[email protected]> * revised drop case * Implement comprehensive drop case analysis with array, tuple, and Box support Co-authored-by: ssyram <[email protected]> * minor * minor * Implement concretize casting for array, tuple, and Box drop shims Co-authored-by: ssyram <[email protected]> * Enhance drop shim generation with structured drop case analysis Co-authored-by: ssyram <[email protected]> * Implement tuple field access with projections for drop shims Co-authored-by: ssyram <[email protected]> * Complete comprehensive drop shim logic implementation with detailed case analysis Co-authored-by: ssyram <[email protected]> * Simplify and combine DropShimKind and DropCase enums, simplify Box handling Co-authored-by: ssyram <[email protected]> * fixed drop field * Implement complete tuple and array drop shim logic with CFG blocks and function calls Co-authored-by: ssyram <[email protected]> * Fix unused block_id assignment and add tuple/array drop test cases Co-authored-by: ssyram <[email protected]> * new example added * Complete Box drop shim generics fix with proper trait clauses Fix Box drop shim generics - provide T and A type parameters correctly Co-Authored-By: ssyram <[email protected]> * Revert "Complete Box drop shim generics fix with proper trait clauses" This reverts commit b889ebb. * Initial analysis and understanding of Box drop generics issue Co-authored-by: ssyram <[email protected]> * fixed the case * Initial analysis and planning for Box drop generics and region binder fixes Co-authored-by: ssyram <[email protected]> * Fix region binders and partial Box generics handling Co-authored-by: ssyram <[email protected]> * Changes before error encountered Co-authored-by: ssyram <[email protected]> * Fix compilation errors and add missing Box detection methods Co-authored-by: ssyram <[email protected]> * Major progress on Box generics extraction and region variable fixes Co-authored-by: ssyram <[email protected]> * Initial analysis and planning for Box drop generics and region binder fixes Fix compilation errors and add missing Box detection methods Changes before error encountered Fix region binders and partial Box generics handling Co-Authored-By: ssyram <[email protected]> * Revert "Initial analysis and planning for Box drop generics and region binder fixes" This reverts commit 2802a90. * Revert "Merge branch 'copilot/fix-95d3f204-c7e2-4956-8a71-c4a3988443ca' of https://github.com/ssyram/charon into copilot/fix-95d3f204-c7e2-4956-8a71-c4a3988443ca" This reverts commit 048ec8f, reversing changes made to eb84c07. * name & generic some fix * Implement region binder fix and Box built-in detection for drop shim generation Co-authored-by: ssyram <[email protected]> * Fix generics mismatch and improve Box built-in detection for drop shims Co-authored-by: ssyram <[email protected]> * fixed test region mismatch * Fix tuple drop shim generation to create proper field-by-field drop logic Co-authored-by: ssyram <[email protected]> * Complete array and tuple drop shim implementation with proper traversal logic Co-authored-by: ssyram <[email protected]> * init-reform * Refactor drop shim generation with proper constant expressions and utility methods - Use `new_var()` and `place_for_var()` utilities instead of manual local creation - Use `ConstantExpr` directly instead of casting to primitive types - Fix scoping issues with concrete place by using `clone()` in loops - Clean up code organization and use `return_place()` utility - Remove unused variables to eliminate warnings - All vtables tests now pass Co-authored-by: ssyram <[email protected]> * Fix array drop shim counter types to match array length expression type Co-authored-by: ssyram <[email protected]> * example changed * Implement comprehensive drop shim generation with proper array/tuple traversal and drop calls Co-authored-by: ssyram <[email protected]> * Fix Vec<T> drop detection by bypassing needs_drop for ADT types Co-authored-by: ssyram <[email protected]> * Fix recursive drop case analysis to handle primitive types correctly Co-authored-by: ssyram <[email protected]> * Simplify analyze_drop_case and fix projection issues in drop shim generation Co-authored-by: ssyram <[email protected]> * fixed box issue & new box example * Extract helper structures and simplify constant creation Co-authored-by: ssyram <[email protected]> * Refactor analyze_drop_case into smaller focused functions and organize code sections Co-authored-by: ssyram <[email protected]> * Clean up verbose comments, remove unused code, and simplify function structure Co-authored-by: ssyram <[email protected]> * new example * more cleanup * new framework * debug done with new complex example --------- Co-authored-by: copilot-swe-agent[bot] <[email protected]> Co-authored-by: ssyram <[email protected]> Co-authored-by: ssyram <[email protected]>
|
At least the Maybe I'm wrong and dismissing your PR too quickly. If that's the case, please explain in the top of your PR why you chose to do things like this, what other options you considered, and how you made sure that the logic you generated is correct as per rust semantics. |
|
Well, no, I put much time in implementing the drop logic myself -- AI generates the framework, which could not be used. So I hand implemented the recursion logic. I would quite strongly disagree with that you denied my efforts so quickly and this strongly and claimed it to be made by AI without even trying to understand. I myself tried quite hard to avoid troubles from AI and checked these parts myself. At least I made sure the test cases reflected correctly. This should show my efforts. As for the drop, the main part is about dropping the arrays and tuples. There is no logic for this right? We are now implementing the shim, which does not exist in the Rustc. So it could split the logic clearly in a pass instead of working during the translation, which requires the Charon, but actually we cannot rely on. Recall that Rustc does not have vtable method shims and we have to generate the shims ourselves. This is likewise the case for drop shims? Or, is there chances we get the drop function for the actual array or tuple types directly? If so, we can get rid of the drop logic, which essentially traverses the array or tuple types recursively and calls drop on those with drop functions. |
|
After lengthy discussion in DMs, I would like to publicly apologize for dimissing your PR out of hand, and I will also close this PR for not being acceptable for the various reasons I mentioned in DMs. #844 provides a lot of the functionality you were implementing here, except for the drops. The drop impls for arrays and tuples you add here look valuable, I'd welcome a PR that adds them in a robust way. |
Following #839 , this PR is also split from #762 . It contains the part of vtable support series for computing the meta-fields of vtable, i.e., the
size,alignanddropfields. This PR adds a new passcompute_vtable_metadatafor this purpose.The former two fields read the data from the
layoutof the type definition (possibly unavailable).The tough part is the
dropfield, where for some built-in data, e.g., arrays & tuples, the drop functions are newly implemented to recursively traverse the internal data to call the potential drop functions.