Hello, This series splits up my m68k QOM'ification patch from the qom-cpu-others RFC series. CPU models are now modelled as subclasses with their own initfns, leaving the setting of feature flags imperative.
Please review and apply. Available from: git://github.com/afaerber/qemu-cpu.git qom-cpu-m68k.v2 https://github.com/afaerber/qemu-cpu/commits/qom-cpu-m68k.v2 Regards, Andreas Cc: Paul Brook <p...@codesourcery.com> Cc: Laurent Vivier <laur...@vivier.eu> v1 -> v2: * Split off from qom-cpu-others series, rebased and split into multiple steps. * Update argument/variable naming to avoid "klass". * Instead of transforming feature flag setup into declarative M68kCPUInfo fields move the original code into initfns, following Peter's model for ARM. As a side effect, we can't move cpu_reset() into the main initfn. Andreas Färber (4): target-m68k: QOM'ify CPU target-m68k: QOM'ify CPU reset target-m68k: Start QOM'ifying CPU init target-m68k: Add QOM CPU subclasses Makefile.target | 1 + target-m68k/cpu-qom.h | 70 ++++++++++++++++++++ target-m68k/cpu.c | 170 +++++++++++++++++++++++++++++++++++++++++++++++++ target-m68k/cpu.h | 3 +- target-m68k/helper.c | 159 +++++++++++++++------------------------------- 5 files changed, 294 insertions(+), 109 deletions(-) create mode 100644 target-m68k/cpu-qom.h create mode 100644 target-m68k/cpu.c -- 1.7.7
