Simmo Saan

Results 358 comments of Simmo Saan
trafficstars

> An easy first step is to have `Quoter` inside `Expansion_helper`: from sherlocode/ppx_universe, it seems no one is using it yet. > We cannot really integrate `Quoter` inside `Expansion_helpers` with...

> I agree that the redundancy in `Mangle.mangle` is not ideal! Then the current state of this PR achieves the sensible approach with `Expansion_helpers.mangle` and I wouldn't change that. >...

The `float128` errors come from here: https://github.com/goblint/cil/blob/51180df8d240b128098370d62a5b0c1f0c306f31/src/frontc/cabs2cil.ml#L2525-L2533 There's no explanation why that check is needed and especially why the memory alignment has anything to do with our ability to analyze...

The s390x wchar errors are related to multi-character _character_ literals. The only relevant change since 1.8.2 that I could find is when that function was converted from `int64` to `cilint`:...

> I'm not sure how relevant support for x32 and mainframes is, I'd suggest to just disable those specific tests for legacy or exotic architectures. Sure, s390x is exotic, but...

> Might also be that it his some issues that were there before? One would need to verify whether gcc can even compile the input files on this architecture before...

> Kind of impressive we never saw any unsoundness because of this... It looks like the majority of tasks in NoOverflows are LP64. A smaller proportion from bitvector, recursive and...

I added a TODO list into the original PR description with a list of things we have already cherry-picked out of here. That should help us know when at one...

One idea @vesalvojdani had was to simply compare their locations and nothing else (i.e. to ignore the bodies entirely). If an inline function is included into two files during preprocessing...

I guess this is also why testrungcc/builtin_object_size fails on macos-homebrew-ocaml-4.14-arm64 in opam-repository CI: ``` # /opt/homebrew/Cellar/gcc/12.2.0/lib/gcc/current/gcc/aarch64-apple-darwin21/12/include/stddef.h:424: Error: float128 only supported on machines where it is an alias (w.r.t. to size...