[PATCH 0/2] two refcount and leak fixes for kvm on s390

[PATCH 0/2] two refcount and leak fixes for kvm on s390

[Christian Borntraeger]

Hello Avi,

here are two fixes for kvm on s390.

[PATCH 1/2] kvm-s390: Fix refcounting and allow module unload
[PATCH 2/2] kvm-s390: Fix memory leak of vcpu->run

None of the fixed problems is severe. So I guess this should be scheduled for 
the next merge window. Thanks

Christian

[PATCH 1/2] kvm-s390: Fix refcounting and allow module unload

[Christian Borntraeger]

From: Christian Borntraeger <borntraeger@de.ibm.com>

Currently it is impossible to unload the kvm module on s390.
This patch fixes kvm_arch_destroy_vm to release all cpus.
This make it possible to unload the module.

In addition we stop messing with the module refcount in arch code.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Carsten Otte <cotte@de.ibm.com>
---
 arch/s390/kvm/kvm-s390.c |   35 +++++++++++++++++++++--------------
 1 file changed, 21 insertions(+), 14 deletions(-)

Index: kvm/arch/s390/kvm/kvm-s390.c
===================================================================
--- kvm.orig/arch/s390/kvm/kvm-s390.c
+++ kvm/arch/s390/kvm/kvm-s390.c
@@ -185,8 +185,6 @@ struct kvm *kvm_arch_create_vm(void)
 	debug_register_view(kvm->arch.dbf, &debug_sprintf_view);
 	VM_EVENT(kvm, 3, "%s", "vm created");
 
-	try_module_get(THIS_MODULE);
-
 	return kvm;
 out_nodbf:
 	free_page((unsigned long)(kvm->arch.sca));
@@ -196,13 +194,32 @@ out_nokvm:
 	return ERR_PTR(rc);
 }
 
+void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
+{
+	VCPU_EVENT(vcpu, 3, "%s", "free cpu");
+	free_page((unsigned long)(vcpu->arch.sie_block));
+	kfree(vcpu);
+}
+
+static void kvm_free_vcpus(struct kvm *kvm)
+{
+	unsigned int i;
+
+	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+		if (kvm->vcpus[i]) {
+			kvm_arch_vcpu_destroy(kvm->vcpus[i]);
+			kvm->vcpus[i] = NULL;
+		}
+	}
+}
+
 void kvm_arch_destroy_vm(struct kvm *kvm)
 {
-	debug_unregister(kvm->arch.dbf);
+	kvm_free_vcpus(kvm);
 	kvm_free_physmem(kvm);
 	free_page((unsigned long)(kvm->arch.sca));
+	debug_unregister(kvm->arch.dbf);
 	kfree(kvm);
-	module_put(THIS_MODULE);
 }
 
 /* Section: vcpu related */
@@ -308,8 +325,6 @@ struct kvm_vcpu *kvm_arch_vcpu_create(st
 	VM_EVENT(kvm, 3, "create cpu %d at %p, sie block at %p", id, vcpu,
 		 vcpu->arch.sie_block);
 
-	try_module_get(THIS_MODULE);
-
 	return vcpu;
 out_free_cpu:
 	kfree(vcpu);
@@ -317,14 +332,6 @@ out_nomem:
 	return ERR_PTR(rc);
 }
 
-void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
-{
-	VCPU_EVENT(vcpu, 3, "%s", "destroy cpu");
-	free_page((unsigned long)(vcpu->arch.sie_block));
-	kfree(vcpu);
-	module_put(THIS_MODULE);
-}
-
 int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
 {
 	/* kvm common code refers to this, but never calls it */

[PATCH 2/2] kvm-s390: Fix memory leak of vcpu->run

[Christian Borntraeger]

From: Christian Borntraeger <borntraeger@de.ibm.com>

The s390 backend of kvm never calls kvm_vcpu_uninit. This causes
a memory leak of vcpu->run pages.
Lets call kvm_vcpu_uninit in kvm_arch_vcpu_destroy to free
the vcpu->run. 

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Carsten Otte <cotte@de.ibm.com>
---
 arch/s390/kvm/kvm-s390.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: kvm/arch/s390/kvm/kvm-s390.c
===================================================================
--- kvm.orig/arch/s390/kvm/kvm-s390.c
+++ kvm/arch/s390/kvm/kvm-s390.c
@@ -198,6 +198,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vc
 {
 	VCPU_EVENT(vcpu, 3, "%s", "free cpu");
 	free_page((unsigned long)(vcpu->arch.sie_block));
+	kvm_vcpu_uninit(vcpu);
 	kfree(vcpu);
 }
 
@@ -230,8 +231,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *
 
 void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
 {
-	/* kvm common code refers to this, but does'nt call it */
-	BUG();
+	/* Nothing todo */
 }
 
 void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)

Re: [PATCH 0/2] two refcount and leak fixes for kvm on s390

[Avi Kivity]

Christian Borntraeger wrote:
> Hello Avi,
>
> here are two fixes for kvm on s390.
>
> [PATCH 1/2] kvm-s390: Fix refcounting and allow module unload
> [PATCH 2/2] kvm-s390: Fix memory leak of vcpu->run
>
> None of the fixed problems is severe. So I guess this should be scheduled for 
> the next merge window. Thanks
>   

Applied, thanks.

-- 
error compiling committee.c: too many arguments to function