Hi Marc, On 2016/12/6 19:39, Marc Zyngier wrote: > On 06/12/16 06:41, Shannon Zhao wrote: >> From: Shannon Zhao <[email protected]> >> >> This is the corresponding part of commit d6400d7(KVM: arm/arm64: >> vgic-v2: Reset LRs at boot time) which is missed for new-vgic. >> >> Signed-off-by: Shannon Zhao <[email protected]> >> --- >> virt/kvm/arm/vgic/vgic-v2.c | 11 +++++++++++ >> 1 file changed, 11 insertions(+) >> >> diff --git a/virt/kvm/arm/vgic/vgic-v2.c b/virt/kvm/arm/vgic/vgic-v2.c >> index 9bab867..c636a19 100644 >> --- a/virt/kvm/arm/vgic/vgic-v2.c >> +++ b/virt/kvm/arm/vgic/vgic-v2.c >> @@ -300,6 +300,15 @@ int vgic_v2_map_resources(struct kvm *kvm) >> >> DEFINE_STATIC_KEY_FALSE(vgic_v2_cpuif_trap); >> >> +static void vgic_cpu_init_lrs(void *params) >> +{ >> + int i; >> + >> + for (i = 0; i < kvm_vgic_global_state.nr_lr; i++) >> + writel_relaxed(0, kvm_vgic_global_state.vctrl_base + >> + GICH_LR0 + (i * 4)); >> +} Since this function will use kvm_vgic_global_state which is initialized by kvm_vgic_hyp_init, if we call it in cpu_hyp_reinit/cpu_init_hyp_mode, the kvm_vgic_global_state is not initialized for now. Is that fine to move kvm_vgic_hyp_init to the first place in init_subsystems?
Thanks, -- Shannon _______________________________________________ kvmarm mailing list [email protected] https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
