Quoting Paolo Bonzini (2018-01-09 11:34:48) > On 09/01/2018 17:02, Michael Roth wrote: > >> so you can see it's already diverged because it prints out > >> a different value for the name of the pci node and the > >> SCSI line has 8200000000000000 vs 00000000. > > FYI I actually see this with a raspberry pi 3 too > > Is this an arm32 or aarch64 install? From Peter's report, I am tempted > to blame the ARM backend (I tested 32-bit x86 and it works). But a bug > in both 32-bit ARM and aarch64 would be less likely.
In my case it's a 32-bit kernel/userspace: pi@rpi-office:~/dev/qemu-build $ uname -a Linux rpi-office 4.4.38-v7+ #938 SMP Thu Dec 15 15:22:21 GMT 2016 armv7l GNU/Linux pi@rpi-office:~/dev/qemu-build $ file ppc64-softmmu/qemu-system-ppc64 ppc64-softmmu/qemu-system-ppc64: ELF 32-bit LSB executable, ARM, EABI5 version 1 (GNU/Linux), dynamically linked, interpreter /lib/ld-linux-armhf.so.3, for GNU/Linux 2.6.32, BuildID[sha1]=4e3e0c8ddbab91a672e4c29b5bfcee9d22459cf7, not stripped > > Paolo > > > , with repeating spaces > > as well. I didn't notice it before since in my case the > > ufd_version_check() in test-migration.c was failing so the migration tests > > were reporting success but were actually just getting skipped. > > >
