On March 4, 2025 1:42:20 AM PST, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote: >On Tue, Mar 04, 2025 at 03:47:45PM +0800, Menglong Dong wrote: >> We don't have to select FUNCTION_ALIGNMENT_32B, so the >> worst case is to increase ~2.2%. >> >> What do you think? > >Well, since I don't understand what you need this for at all, I'm firmly >on the side of not doing this. > >What actual problem is being solved with this meta data nonsense? Why is >it worth blowing up our I$ footprint over. > >Also note, that if you're going to be explaining this, start from >scratch, as I have absolutely 0 clues about BPF and such. I would appreciate such information as well. The idea seems dubious on the surface.
