The size of a memory area can never be negative.
It follows it must be defined as an unsigned value.
Let's modify the memory regions size to unsigned 128 integer
and modify accordingly the 128 bit arithmetic.
This makes memory size calculations easier and easier to understand.
I fear loud protest but really, it is a broken concept that
obfuscate the code.
Pierre Morel (7):
int128: use unsigned 128 bit arithmetic
memory: modify memory size for an unsigned 128bit int
128bit: adapt core files for unsigned 128bits
128bit: adapt sparc mmu_helper for UInt128
128bit: adapt VFIO for UInt128 arithmetic
128bit: adapt Virtio for UInt128 arithmetic
128bits: some HW components use 128bit arithmetic.
exec.c | 32 ++++----
hw/core/loader.c | 2 +-
hw/display/exynos4210_fimd.c | 4 +-
hw/display/framebuffer.c | 2 +-
hw/mem/pc-dimm.c | 6 +-
hw/pci-host/ppce500.c | 2 +-
hw/vfio/common.c | 20 +++---
hw/virtio/dataplane/vring.c | 2 +-
hw/virtio/vhost.c | 4 +-
hw/virtio/virtio-balloon.c | 2 +-
include/exec/memory.h | 4 +-
include/qemu/int128.h | 95 ++++++++++-----------
kvm-all.c | 16 ++--
memory.c | 184 +++++++++++++++++++++---------------------
memory_mapping.c | 2 +-
target-sparc/mmu_helper.c | 2 +-
16 files changed, 187 insertions(+), 192 deletions(-)
