From: Shannon Zhao <shannon.zhao@xxxxxxxxxx> This patch set adds support for Cross type vCPU in KVM-ARM64. It allows userspace to request a different vCPU type with the physical ones and check whether the physical CPUs could support that specific vCPU. If so, KVM will trap the ID registers and return guest with the values from usersapce. This patch set is not complete since the CPU Errata is not considered and currently it only checks if the id_aa64mmfr0_el1 register value is legal. I want this as an example and need some feedback from folks if this approach is right or proper. You can test this patch set with QEMU using -cpu cortex-a53/cortex-a57/generic/cortex-a72 These patches can be fetched from: https://git.linaro.org/people/shannon.zhao/linux-mainline.git cross_vcpu_rfc You corresponding QEMU patches can be fetched from: https://git.linaro.org/people/shannon.zhao/qemu.git cross_vcpu_rfc Thanks, Shannon Shannon Zhao (7): ARM64: KVM: Add the definition of ID registers ARM64: KVM: Add reset handlers for all ID registers ARM64: KVM: Reset ID registers when creating the VCPUs ARM64: KVM: emulate accessing ID registers ARM64: KVM: Support cross type vCPU ARM64: KVM: Support heterogeneous system ARM64: KVM: Add user set handler for id_aa64mmfr0_el1 arch/arm/kvm/arm.c | 36 ++++- arch/arm64/include/asm/kvm_coproc.h | 1 + arch/arm64/include/asm/kvm_emulate.h | 3 + arch/arm64/include/asm/kvm_host.h | 49 +++++- arch/arm64/include/uapi/asm/kvm.h | 1 + arch/arm64/kvm/guest.c | 18 ++- arch/arm64/kvm/hyp/sysreg-sr.c | 2 + arch/arm64/kvm/sys_regs.c | 290 +++++++++++++++++++++++------------ include/uapi/linux/kvm.h | 2 + 9 files changed, 296 insertions(+), 106 deletions(-) -- 2.0.4 _______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
