On Tue, Jun 07, 2022 at 11:43:08AM +0100, Peter Maydell wrote: > So I don't have a strong view on whether these devices should > be DEVICE_NATIVE_ENDIAN or DEVICE_LITTLE_ENDIAN (except that > my impression is that a DEVICE_LITTLE_ENDIAN device on a > big-endian system is a bit weird, because it means the guest > has to byteswap everything. You see that with PCI devices because > the PCI spec mandates LE, but not often elsewhere). > > If there's an official-ish spec for how goldfish devices are > supposed to behave (does anybody have a pointer to one?) and it says > "always little-endian" then that would probably suggest that fixing > m68k would be nice if we can.
I think there are some conflicting thoughts on this. In Geert's he mentioned: Using Goldfish devices as little-endian devices should be fine. In Arnd's mail he mentions: https://lore.kernel.org/lkml/CAK8P3a1oN8NrUjkh2X8jHQbyz42Xo6GSa=5n0gd6vqcxrjm...@mail.gmail.com/#t ... the device was clearly defined as having little-endian registers, Based on that I was thinking that switching to DEVICE_LITTLE_ENDIAN would make sense. However, in a followup mail from Laurent we see: https://lore.kernel.org/lkml/cb884368-0226-e913-80d2-62d2b7b2e...@vivier.eu/ The reference document[1] doesn't define the endianness of goldfish. [1] https://android.googlesource.com/platform/external/qemu/+/master/docs/GOLDFISH-VIRTUAL-HARDWARE.TXT The documentation does not clearly specify it. So maybe maybe or1k should just be updated on the linux side and add gf_ioread32/gf_iowrite32 big-endian accessors. -Stafford
