On Mon, 2 Mar 2020 at 14:16, Ard Biesheuvel <ard.biesheu...@linaro.org> wrote: > > On Mon, 2 Mar 2020 at 14:13, Leif Lindholm <l...@nuviainc.com> wrote: > > > > On Wed, Feb 26, 2020 at 11:03:53 +0100, Ard Biesheuvel wrote: > > > Cache maintenance on ARMv7 systems and up should be done by virtual > > > address if the purpose is to manage the cached state of contents of > > > memory. Set/way operations are only intended to maintain the caches > > > themselves, e.g., to ensure that the contents of dirty cachelines > > > are brought to main memory before the core is powered off entirely. > > > > > > UEFI on ARM is typically not involved in the latter at all, and any > > > cache maintenance it does is to ensure that the memory it occupies > > > and modifies remains in a consistent state with respect to the > > > caches. > > > > > > So let's deprecate the set/way routines now that we have removed all > > > uses of it in the core code. > > > > Does this patch simply get dropped in favour of the ASSERT variant? > > > > Yes. > > But I realised that we still need a fix for CmdRunAxf in that case :-(
As discussed on IRC, with CmdRunAxf out of the way, we can ASSERT () on any use of the set/way ops with the MMU/dcache enabled. In any case, this 6/6 will simply be dropped, and superseded by my followup series '[PATCH 0/3] ArmPkg/ArmLib: ASSERT() on misuse of set/way ops' -=-=-=-=-=-=-=-=-=-=-=- Groups.io Links: You receive all messages sent to this group. View/Reply Online (#55405): https://edk2.groups.io/g/devel/message/55405 Mute This Topic: https://groups.io/mt/71562852/21656 Group Owner: devel+ow...@edk2.groups.io Unsubscribe: https://edk2.groups.io/g/devel/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
