On 5 May 2014 09:07, Greg Kurz <gk...@linux.vnet.ibm.com> wrote: > POWER7, POWER7+ and POWER8 families use the ILE bit of the LPCR > special purpose register to decide the endianness to use when > entering interrupt handlers. When running a Linux guest, this > provides a hint on the endianness used by the kernel. From a > QEMU point of view, the information is needed for legacy virtio > support and crash dump support as well.
Do you care about the case of: * kernel bigendian * userspace littleendian (or vice-versa) * guest kernel passes virtio device through to guest userspace * guest userspace is doing the manipulation of the device ? (Will Deacon just suggested this as a possibility on the kvm-arm mailing list...) Also, are we documenting what the process should be for a virtio implementation to decide the endianness for a particular architecture? I assume we'd like kvmtool and QEMU to do the same thing rather than subtly different things... thanks -- PMM
