Re: [PATCH v2] KVM: X86: Grab KVM's srcu lock when accessing hv assist page

[Wanpeng Li]

On Sat, 15 Feb 2020 at 05:33, Sean Christopherson
<sean.j.christopherson@xxxxxxxxx> wrote:
>
> The subject shoud be rewritten to simply states that the vmcs12->shadow
> sync is being.  The changelog can explain the motiviation, but a one-line
> git log should clearly reveal that this affects the shadow VMCS.
>
> Subsystem should also be "KVM: nVMX:"
>
> On Fri, Feb 14, 2020 at 10:27:12AM +0100, Paolo Bonzini wrote:
> > On 14/02/20 10:16, Wanpeng Li wrote:
> > > From: wanpeng li <wanpengli@xxxxxxxxxxx>
> > >
> > > For the duration of mapping eVMCS, it derefences ->memslots without holding
> > > ->srcu or ->slots_lock when accessing hv assist page. This patch fixes it by
> > > moving nested_sync_vmcs12_to_shadow to prepare_guest_switch, where the SRCU
> > > is already taken.
> >
> > Looks good, but I'd like an extra review from Sean or Vitaly.
> >
> > We also should add a WARN_ON_ONCE that replaces the previous location of
> > the "if (vmx->nested.need_vmcs12_to_shadow_sync)", but I can do that myself.
> >
> > Thanks,
> >
> > Paolo
> >
> > > It can be reproduced by running kvm's evmcs_test selftest.
> > >
> > >   =============================
> > >   warning: suspicious rcu usage
> > >   5.6.0-rc1+ #53 tainted: g        w ioe
> > >   -----------------------------
> > >   ./include/linux/kvm_host.h:623 suspicious rcu_dereference_check() usage!
> > >
> > >   other info that might help us debug this:
> > >
> > >    rcu_scheduler_active = 2, debug_locks = 1
> > >   1 lock held by evmcs_test/8507:
> > >    #0: ffff9ddd156d00d0 (&vcpu->mutex){+.+.}, at:
> > > kvm_vcpu_ioctl+0x85/0x680 [kvm]
> > >
> > >   stack backtrace:
> > >   cpu: 6 pid: 8507 comm: evmcs_test tainted: g        w ioe     5.6.0-rc1+ #53
> > >   hardware name: dell inc. optiplex 7040/0jctf8, bios 1.4.9 09/12/2016
> > >   call trace:
> > >    dump_stack+0x68/0x9b
> > >    kvm_read_guest_cached+0x11d/0x150 [kvm]
> > >    kvm_hv_get_assist_page+0x33/0x40 [kvm]
> > >    nested_enlightened_vmentry+0x2c/0x60 [kvm_intel]
> > >    nested_vmx_handle_enlightened_vmptrld.part.52+0x32/0x1c0 [kvm_intel]
> > >    nested_sync_vmcs12_to_shadow+0x439/0x680 [kvm_intel]
> > >    vmx_vcpu_run+0x67a/0xe60 [kvm_intel]
> > >    vcpu_enter_guest+0x35e/0x1bc0 [kvm]
> > >    kvm_arch_vcpu_ioctl_run+0x40b/0x670 [kvm]
> > >    kvm_vcpu_ioctl+0x370/0x680 [kvm]
> > >    ksys_ioctl+0x235/0x850
> > >    __x64_sys_ioctl+0x16/0x20
> > >    do_syscall_64+0x77/0x780
> > >    entry_syscall_64_after_hwframe+0x49/0xbe
> > >
> > > Signed-off-by: Wanpeng Li <wanpengli@xxxxxxxxxxx>
> > > ---
> > >  arch/x86/kvm/vmx/vmx.c | 6 +++---
> > >  1 file changed, 3 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> > > index 9a66648..6bd6ca4 100644
> > > --- a/arch/x86/kvm/vmx/vmx.c
> > > +++ b/arch/x86/kvm/vmx/vmx.c
> > > @@ -1214,6 +1214,9 @@ void vmx_prepare_switch_to_guest(struct kvm_vcpu *vcpu)
> > >
> > >      vmx_set_host_fs_gs(host_state, fs_sel, gs_sel, fs_base, gs_base);
> > >      vmx->guest_state_loaded = true;
> > > +
> > > +    if (vmx->nested.need_vmcs12_to_shadow_sync)
> > > +        nested_sync_vmcs12_to_shadow(vcpu);
>
> This needs to be above the short circuit check on guest_state_loaded, e.g.:
>
>         if (vmx->nested.need_vmcs12_to_shadow_sync)
>                 nested_sync_vmcs12_to_shadow(vcpu);
>
>         if (vmx->guest_state_loaded)
>                 return;

Agreed, do it in new version.

    Wanpeng