| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <87lfkach6o.fsf@vitty.brq.redhat.com>
Date: Fri, 26 Jun 2020 11:25:19 +0200
From: Vitaly Kuznetsov <vkuznets@...hat.com>
To: Vivek Goyal <vgoyal@...hat.com>, kvm@...r.kernel.org
Cc: virtio-fs@...hat.com, pbonzini@...hat.com,
sean.j.christopherson@...el.com, linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH] kvm,x86: Exit to user space in case of page fault error
Vivek Goyal <vgoyal@...hat.com> writes:
> Page fault error handling behavior in kvm seems little inconsistent when
> page fault reports error. If we are doing fault synchronously
> then we capture error (-EFAULT) returned by __gfn_to_pfn_memslot() and
> exit to user space and qemu reports error, "error: kvm run failed Bad address".
>
> But if we are doing async page fault, then async_pf_execute() will simply
> ignore the error reported by get_user_pages_remote() or
> by kvm_mmu_do_page_fault(). It is assumed that page fault was successful
> and either a page ready event is injected in guest or guest is brought
> out of artificial halt state and run again. In both the cases when guest
> retries the instruction, it takes exit again as page fault was not
> successful in previous attempt. And then this infinite loop continues
> forever.
>
> Trying fault in a loop will make sense if error is temporary and will
> be resolved on retry. But I don't see any intention in the code to
> determine if error is temporary or not. Whether to do fault synchronously
> or asynchronously, depends on so many variables but none of the varibales
> is whether error is temporary or not. (kvm_can_do_async_pf()).
>
> And that makes it very inconsistent or unpredictable to figure out whether
> kvm will exit to qemu with error or it will just retry and go into an
> infinite loop.
>
> This patch tries to make this behavior consistent. That is instead of
> getting into infinite loop of retrying page fault, exit to user space
> and stop VM if page fault error happens.
>
> In future this can be improved by injecting errors into guest. As of
> now we don't have any race free method to inject errors in guest.
>
> When page fault error happens in async path save that pfn and when next
> time guest retries, do a sync fault instead of async fault. So that if error
> is encountered, we exit to qemu and avoid infinite loop.
>
> As of now only one error pfn is stored and that means it could be
> overwritten before next a retry from guest happens. But this is
> just a hint and if we miss it, some other time we will catch it.
> If this becomes an issue, we could maintain an array of error
> gfn later to help ease the issue.
>
> Signed-off-by: Vivek Goyal <vgoyal@...hat.com>
> ---
> arch/x86/include/asm/kvm_host.h | 1 +
> arch/x86/kvm/mmu.h | 2 +-
> arch/x86/kvm/mmu/mmu.c | 2 +-
> arch/x86/kvm/x86.c | 14 +++++++++++---
> 4 files changed, 14 insertions(+), 5 deletions(-)
>
> diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
> index be5363b21540..3c0677b9d3d5 100644
> --- a/arch/x86/include/asm/kvm_host.h
> +++ b/arch/x86/include/asm/kvm_host.h
> @@ -778,6 +778,7 @@ struct kvm_vcpu_arch {
> unsigned long nested_apf_token;
> bool delivery_as_pf_vmexit;
> bool pageready_pending;
> + gfn_t error_gfn;
> } apf;
>
> /* OSVW MSRs (AMD only) */
> diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
> index 444bb9c54548..d0a2a12c7bb6 100644
> --- a/arch/x86/kvm/mmu.h
> +++ b/arch/x86/kvm/mmu.h
> @@ -60,7 +60,7 @@ void kvm_init_mmu(struct kvm_vcpu *vcpu, bool reset_roots);
> void kvm_init_shadow_mmu(struct kvm_vcpu *vcpu, u32 cr0, u32 cr4, u32 efer);
> void kvm_init_shadow_ept_mmu(struct kvm_vcpu *vcpu, bool execonly,
> bool accessed_dirty, gpa_t new_eptp);
> -bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu);
> +bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu, gfn_t gfn);
> int kvm_handle_page_fault(struct kvm_vcpu *vcpu, u64 error_code,
> u64 fault_address, char *insn, int insn_len);
>
> diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> index 76817d13c86e..a882a6a9f7a7 100644
> --- a/arch/x86/kvm/mmu/mmu.c
> +++ b/arch/x86/kvm/mmu/mmu.c
> @@ -4078,7 +4078,7 @@ static bool try_async_pf(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn,
> if (!async)
> return false; /* *pfn has correct page already */
>
> - if (!prefault && kvm_can_do_async_pf(vcpu)) {
> + if (!prefault && kvm_can_do_async_pf(vcpu, cr2_or_gpa >> PAGE_SHIFT)) {
gpa_to_gfn(cr2_or_gpa) ?
> trace_kvm_try_async_get_page(cr2_or_gpa, gfn);
> if (kvm_find_async_pf_gfn(vcpu, gfn)) {
> trace_kvm_async_pf_doublefault(cr2_or_gpa, gfn);
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 3b92db412335..a6af7e9831b9 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -10380,7 +10380,9 @@ void kvm_arch_async_page_ready(struct kvm_vcpu *vcpu, struct kvm_async_pf *work)
> work->arch.cr3 != vcpu->arch.mmu->get_guest_pgd(vcpu))
> return;
>
> - kvm_mmu_do_page_fault(vcpu, work->cr2_or_gpa, 0, true);
> + r = kvm_mmu_do_page_fault(vcpu, work->cr2_or_gpa, 0, true);
> + if (r < 0)
> + vcpu->arch.apf.error_gfn = work->cr2_or_gpa >> PAGE_SHIFT;
> }
>
> static inline u32 kvm_async_pf_hash_fn(gfn_t gfn)
> @@ -10490,7 +10492,7 @@ static bool kvm_can_deliver_async_pf(struct kvm_vcpu *vcpu)
> return true;
> }
>
> -bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu)
> +bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu, gfn_t gfn)
> {
> if (unlikely(!lapic_in_kernel(vcpu) ||
> kvm_event_needs_reinjection(vcpu) ||
> @@ -10504,7 +10506,13 @@ bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu)
> * If interrupts are off we cannot even use an artificial
> * halt state.
> */
> - return kvm_arch_interrupt_allowed(vcpu);
> + if (!kvm_arch_interrupt_allowed(vcpu))
> + return false;
> +
> + if (vcpu->arch.apf.error_gfn == gfn)
> + return false;
> +
> + return true;
> }
>
> bool kvm_arch_async_page_not_present(struct kvm_vcpu *vcpu,
I'm a little bit afraid that a single error_gfn may not give us
deterministric behavior. E.g. when we have a lot of faulting processes
it may take many iterations to hit 'error_gfn == gfn' because we'll
always be overwriting 'error_gfn' with new values and waking up some
(random) process.
What if we just temporary disable the whole APF mechanism? That would
ensure we're making forward progress. Something like (completely
untested):
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index f8998e97457f..945b3d5a2796 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -778,6 +778,7 @@ struct kvm_vcpu_arch {
unsigned long nested_apf_token;
bool delivery_as_pf_vmexit;
bool pageready_pending;
+ bool error_pending;
} apf;
/* OSVW MSRs (AMD only) */
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index fdd05c233308..e5f04ae97e91 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -4124,8 +4124,18 @@ static int direct_page_fault(struct kvm_vcpu *vcpu, gpa_t gpa, u32 error_code,
if (try_async_pf(vcpu, prefault, gfn, gpa, &pfn, write, &map_writable))
return RET_PF_RETRY;
- if (handle_abnormal_pfn(vcpu, is_tdp ? 0 : gpa, gfn, pfn, ACC_ALL, &r))
+ if (handle_abnormal_pfn(vcpu, is_tdp ? 0 : gpa, gfn, pfn, ACC_ALL, &r)) {
+ /*
+ * In case APF mechanism was previously disabled due to an error
+ * we are ready to re-enable it here as we're about to inject an
+ * error to userspace. There is no guarantee we are handling the
+ * same GFN which failed in APF here but at least we are making
+ * forward progress.
+ */
+
+ vcpu->arch.apf.error_pending = false;
return r;
+ }
r = RET_PF_RETRY;
spin_lock(&vcpu->kvm->mmu_lock);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 00c88c2f34e4..4607cf4d5117 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -10379,7 +10379,9 @@ void kvm_arch_async_page_ready(struct kvm_vcpu *vcpu, struct kvm_async_pf *work)
work->arch.cr3 != vcpu->arch.mmu->get_guest_pgd(vcpu))
return;
- kvm_mmu_do_page_fault(vcpu, work->cr2_or_gpa, 0, true);
+ r = kvm_mmu_do_page_fault(vcpu, work->cr2_or_gpa, 0, true);
+ if (r < 0)
+ vcpu->arch.apf.error_pending = true;
}
static inline u32 kvm_async_pf_hash_fn(gfn_t gfn)
@@ -10499,6 +10501,9 @@ bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu)
if (kvm_hlt_in_guest(vcpu->kvm) && !kvm_can_deliver_async_pf(vcpu))
return false;
+ if (unlikely(vcpu->arch.apf.error_pending))
+ return false;
+
/*
* If interrupts are off we cannot even use an artificial
* halt state.
--
Vitaly
Powered by blists - more mailing lists