Le 11/03/2021 à 16:44, Michael S. Tsirkin a écrit : > On Tue, Mar 09, 2021 at 11:43:13PM +0100, Laurent Vivier wrote: >> read[wl]()/write[wl] already access memory in little-endian mode. > > But then they convert it to CPU right? We just convert it back ...
Yes, you're right. But there's a real problem with a big-endian guest using directly virtio-mmio without using virtio-pci. It seems there is one too many little-endian conversion somewhere, and this has not been detected before because there is no big-endian guest that uses virtio-mmio directly, and with little-endian guests the conversion is a no-op. I'm going to have a closer look to the code path... perhaps the problem is in QEMU not in the kernel. Thanks, Laurent
