2020-08-30 04:36:37

by Rustam Kovhaev

Subject: [PATCH] KVM: fix memory leak in kvm_io_bus_unregister_dev()

when kmalloc() fails in kvm_io_bus_unregister_dev(), before removing
the bus, we should iterate over all other devices linked to it and call
kvm_iodevice_destructor() for them

Reported-and-tested-by: [email protected]
Link: https://syzkaller.appspot.com/bug?extid=f196caa45793d6374707
Signed-off-by: Rustam Kovhaev <[email protected]>
---
virt/kvm/kvm_main.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 67cd0b88a6b6..646aa7b82548 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -4332,7 +4332,7 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
struct kvm_io_device *dev)
{
- int i;
+ int i, j;
struct kvm_io_bus *new_bus, *bus;

bus = kvm_get_bus(kvm, bus_idx);
@@ -4351,6 +4351,11 @@ void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
GFP_KERNEL_ACCOUNT);
if (!new_bus) {
pr_err("kvm: failed to shrink bus, removing it completely\n");
+ for (j = 0; j < bus->dev_count; j++) {
+ if (j == i)
+ continue;
+ kvm_iodevice_destructor(bus->range[j].dev);
+ }
goto broken;
}

--
2.28.0

2020-09-01 16:27:19

by Vitaly Kuznetsov

Subject: Re: [PATCH] KVM: fix memory leak in kvm_io_bus_unregister_dev()

Rustam Kovhaev <[email protected]> writes:

> when kmalloc() fails in kvm_io_bus_unregister_dev(), before removing
> the bus, we should iterate over all other devices linked to it and call
> kvm_iodevice_destructor() for them
>
> Reported-and-tested-by: [email protected]
> Link: https://syzkaller.appspot.com/bug?extid=f196caa45793d6374707
> Signed-off-by: Rustam Kovhaev <[email protected]>
> ---
> virt/kvm/kvm_main.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 67cd0b88a6b6..646aa7b82548 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -4332,7 +4332,7 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
> void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
> struct kvm_io_device *dev)
> {
> - int i;
> + int i, j;
> struct kvm_io_bus *new_bus, *bus;
>
> bus = kvm_get_bus(kvm, bus_idx);
> @@ -4351,6 +4351,11 @@ void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
> GFP_KERNEL_ACCOUNT);
> if (!new_bus) {
^^^ redundant space

> pr_err("kvm: failed to shrink bus, removing it completely\n");
> + for (j = 0; j < bus->dev_count; j++) {
> + if (j == i)
> + continue;
> + kvm_iodevice_destructor(bus->range[j].dev);
> + }
> goto broken;

The name of the label is really misleading (as it is not actually a
failure path), I'd even suggest we get rid of this goto completely,
something like

new_bus = kmalloc(struct_size(bus, range, bus->dev_count - 1),
GFP_KERNEL_ACCOUNT);
if (new_bus) {
memcpy(new_bus, bus, sizeof(*bus) + i * sizeof(struct kvm_io_range));
new_bus->dev_count--;
memcpy(new_bus->range + i, bus->range + i + 1,
(new_bus->dev_count - i) * sizeof(struct kvm_io_range));
} else {
pr_err("kvm: failed to shrink bus, removing it completely\n");
for (j = 0; j < bus->dev_count; j++) {
if (j == i)
continue;
kvm_iodevice_destructor(bus->range[j].dev);
}

rcu_assign_pointer(kvm->buses[bus_idx], new_bus);
synchronize_srcu_expedited(&kvm->srcu);
kfree(bus);
return;


> }

None of the above should block the fix IMO, so:

Reviewed-by: Vitaly Kuznetsov <[email protected]>

--
Vitaly

2020-09-01 22:47:43

by Rustam Kovhaev

Subject: Re: [PATCH] KVM: fix memory leak in kvm_io_bus_unregister_dev()

On Tue, Sep 01, 2020 at 06:25:42PM +0200, Vitaly Kuznetsov wrote:
> Rustam Kovhaev <[email protected]> writes:
>
> > when kmalloc() fails in kvm_io_bus_unregister_dev(), before removing
> > the bus, we should iterate over all other devices linked to it and call
> > kvm_iodevice_destructor() for them
> >
> > Reported-and-tested-by: [email protected]
> > Link: https://syzkaller.appspot.com/bug?extid=f196caa45793d6374707
> > Signed-off-by: Rustam Kovhaev <[email protected]>
> > ---
> > virt/kvm/kvm_main.c | 7 ++++++-
> > 1 file changed, 6 insertions(+), 1 deletion(-)
> >
> > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > index 67cd0b88a6b6..646aa7b82548 100644
> > --- a/virt/kvm/kvm_main.c
> > +++ b/virt/kvm/kvm_main.c
> > @@ -4332,7 +4332,7 @@ int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
> > void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
> > struct kvm_io_device *dev)
> > {
> > - int i;
> > + int i, j;
> > struct kvm_io_bus *new_bus, *bus;
> >
> > bus = kvm_get_bus(kvm, bus_idx);
> > @@ -4351,6 +4351,11 @@ void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
> > GFP_KERNEL_ACCOUNT);
> > if (!new_bus) {
> ^^^ redundant space
>
> > pr_err("kvm: failed to shrink bus, removing it completely\n");
> > + for (j = 0; j < bus->dev_count; j++) {
> > + if (j == i)
> > + continue;
> > + kvm_iodevice_destructor(bus->range[j].dev);
> > + }
> > goto broken;
>
> The name of the label is really misleading (as it is not actually a
> failure path), I'd even suggest we get rid of this goto completely,
> something like
>
> new_bus = kmalloc(struct_size(bus, range, bus->dev_count - 1),
> GFP_KERNEL_ACCOUNT);
> if (new_bus) {
> memcpy(new_bus, bus, sizeof(*bus) + i * sizeof(struct kvm_io_range));
> new_bus->dev_count--;
> memcpy(new_bus->range + i, bus->range + i + 1,
> (new_bus->dev_count - i) * sizeof(struct kvm_io_range));
> } else {
> pr_err("kvm: failed to shrink bus, removing it completely\n");
> for (j = 0; j < bus->dev_count; j++) {
> if (j == i)
> continue;
> kvm_iodevice_destructor(bus->range[j].dev);
> }
>
> rcu_assign_pointer(kvm->buses[bus_idx], new_bus);
> synchronize_srcu_expedited(&kvm->srcu);
> kfree(bus);
> return;
>
>
> > }
>
> None of the above should block the fix IMO, so:
>
> Reviewed-by: Vitaly Kuznetsov <[email protected]>
>
> --
> Vitaly
>
hi Vitaly, thank you for the review! i'll send the new patch