Hi, Patch 6cb7f4d86438f56a5c917ff6148c70bd237ee0e5 introduced an unconditional call to kvm_arch_do_ioperm() without protecting with an #ifdef like the remaining calls are. Cheers, Jes
Protect calls to kvm_arch_do_ioperm() introduced in 6cb7f4d86438f56a5c917ff6148c70bd237ee0e5 with #ifdef USE_KVM_DEVICE_ASSIGNMENT similar to how the other calls to this function are protected. Signed-off-by: Jes Sorensen <jes@xxxxxxx> --- qemu/qemu-kvm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) Index: kvm-userspace.git/qemu/qemu-kvm.c =================================================================== --- kvm-userspace.git.orig/qemu/qemu-kvm.c +++ kvm-userspace.git/qemu/qemu-kvm.c @@ -444,7 +444,7 @@ { CPUState *env = _env; sigset_t signals; - struct ioperm_data *data; + struct ioperm_data *data = NULL; vcpu = &vcpu_info[env->cpu_index]; vcpu->env = env; @@ -454,9 +454,11 @@ kvm_create_vcpu(kvm_context, env->cpu_index); kvm_qemu_init_env(env); +#ifdef USE_KVM_DEVICE_ASSIGNMENT /* do ioperm for io ports of assigned devices */ LIST_FOREACH(data, &ioperm_head, entries) on_vcpu(env, kvm_arch_do_ioperm, data); +#endif /* signal VCPU creation */ pthread_mutex_lock(&qemu_mutex);