On Tue, Jan 05, 2021 at 09:56:05AM -0800, Jacob Keller wrote: > On 12/28/2020 1:33 PM, Luc Van Oostenryck wrote: > > On Mon, Dec 28, 2020 at 05:18:50PM +0000, Ramsay Jones wrote: > >> > >> Patch #16 also looks good, but I would need to study it a bit more > >> than I have time available to be totally happy. It does not seem > >> to be handling the 'lowering' of 'odd bit-sized' symbols created in > >> the previous patch (to answer my own question), so I would have to > >> apply the patches (or fetch the above branch) to study some more. > > > > Well, patch #16 doesn't contain the lowering, it kinda just enables > > the last bits to support (without out-of-bound access) packed structures > > including packed bitfields. > > > > What does the 'lowering' gain us? Or, in other words, what is still > missing after this series? As a static checker, nothing is missing and the series is now mainlined. The 'lowering' only matters if you want to somehow translate the instructions used in the IR (Intermediate Representation) into instructions for a more concrete machine, because now there are things like: 'do a load of a 5-byte word'. -- Luc