On 10/14/24 08:44, Mike Rapoport wrote: >> To me, it seems like execmem itself should be excluding all HIGHMEM=y >> builds or _maybe_ all 32-bit builds because execmem has the built-in >> assumption that vmalloc memory is in the direct map. > I presume you mean execmem cache here, as execmem in general should work > everywhere. Yeah, just the new execmem cache code from the new series. > After discussion with Christoph about how to define > execmem_fill_trapping_insns(), I'm going to add a Kconfig opt-in option > that will guard the entire execmem cache implementation and have it > selected by x86_64. Seems fine to me, as long as we're clear in the changelog why it's there.
