Meta-issue: viable student projects #84
Description
This meta-issue collects issues which constitute "interesting" projects that are estimated as at most three months of work for a not-too-expert (but skilled) student, whether at Bachelor's or Master's or PhD level. They are all more than simply bug fixes or quality-of-implementation improvements: they should all be something that is evaluable and at least a bit research-interesting. They also avoid the gnarlier design-level issues that (for the moment) have to remain on my plate.
- Do binary instrumentation of allocation functions #11
- Uniqtypes should be more compact #12
- Binary instrumentation of inline allocation functions and alloca #15
- Minimise the up-front requirement to identify allocator wrappers and (simple) suballocators #20
- Derivation of runtime type information from ctf #44
- Uniqtypes should integrate with libdivide #50
- sizeofness analysis isn't sufficiently general #63
- Support debuginfod for just-in-time metadata creation #75 (marginal perhaps)
- Stack walking should use eh_elfs from frdwarf #83
- Explore how stack type metadata can be 'compiled' as code, not pre-elaborated into struct types #85
- New service: stable allocation identities #86
- Rust toolchain support #89
- C++ toolchain support #90
- Fortran toolchain support #91
- frametypes should also describe registers and callee-save stack slots #93
- Post-hoc allocation site analysis #99
- Fuse filesystem support #116