KVM: s390: Make huge pages unavailable in ucontrol VMs

We currently do not notify all gmaps when using gmap_pmdp_xchg(), due
to locking constraints. This makes ucontrol VMs, which is the only VM
type that creates multiple gmaps, incompatible with huge pages. Also
we would need to hold the guest_table_lock of all gmaps that have this
vmaddr maped to synchronize access to the pmd.

ucontrol VMs are rather exotic and creating a new locking concept is
no easy task. Hence we return EINVAL when trying to active
KVM_CAP_S390_HPAGE_1M and report it as being not available when
checking for it.

Fixes: a4499382 ("KVM: s390: Add huge page enablement control")
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Message-Id: <20180801112508.138159-1-frankja@linux.ibm.com>
Signed-off-by: Janosch Frank <frankja@linux.ibm.com>

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index c664064..8d8a372 100644 (file)
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -4510,7 +4510,8 @@ Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits.
 Architectures: s390
 Parameters: none
 Returns: 0 on success, -EINVAL if hpage module parameter was not set
-        or cmma is enabled
+        or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL
+        flag set
 
 With this capability the KVM support for memory backing with 1m pages
 through hugetlbfs can be enabled for a VM. After the capability is

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index f69333f..ac5da6b 100644 (file)
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -481,7 +481,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
                break;
        case KVM_CAP_S390_HPAGE_1M:
                r = 0;
-               if (hpage)
+               if (hpage && !kvm_is_ucontrol(kvm))
                        r = 1;
                break;
        case KVM_CAP_S390_MEM_OP:
@@ -691,7 +691,7 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap)
                mutex_lock(&kvm->lock);
                if (kvm->created_vcpus)
                        r = -EBUSY;
-               else if (!hpage || kvm->arch.use_cmma)
+               else if (!hpage || kvm->arch.use_cmma || kvm_is_ucontrol(kvm))
                        r = -EINVAL;
                else {
                        r = 0;