Messages in this thread | | | Date | Fri, 13 Oct 2017 10:16:32 +0200 | From | Christoffer Dall <> | Subject | Re: [PATCH] KVM: remove printing of vcpu address |
| |
On Mon, Oct 09, 2017 at 05:31:41PM +1100, Tobin C. Harding wrote: > Code currently prints the address of the kvm_vcpu structure in an error > message. It is not immediately clear what value this address adds to > the error string, we can use the vcpu ID instead.
I would say it's not immediately clear how the vcpu ID is more valuable. The pointer identifies a unique vcpu across all VMs, vcpu ID does not.
> Printing unnecessary > kernel addresses to dmesg poses a security risk.
Is this really a concern we have for error messages? I understand you can get a single pointer revealed if you have broken hardware or if we have a bug, but I see numerious other examples in the kernel.
If anything, we can remote printing the vcpu pointer entirely.
Thanks, -Christoffer > > Remove the address from error message output, show vcpu ID instead. > > Signed-off-by: Tobin C. Harding <me@tobin.cc> > --- > virt/kvm/arm/arch_timer.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c > index 8e89d63005c7..ca6c331cad28 100644 > --- a/virt/kvm/arm/arch_timer.c > +++ b/virt/kvm/arm/arch_timer.c > @@ -88,7 +88,7 @@ static irqreturn_t kvm_arch_timer_handler(int irq, void *dev_id) > * interrupt at this point is a sure sign of some major > * breakage. > */ > - pr_warn("Unexpected interrupt %d on vcpu %p\n", irq, vcpu); > + pr_warn("Unexpected interrupt %d on vcpu ID %d\n", irq, vcpu->vcpu_id); > return IRQ_HANDLED; > } > > -- > 2.7.4 >
| |