* [PATCH] KVM: x86: hyper-v: Task srcu lock when accessing kvm_memslots()
@ 2021-05-10 8:21 Wanpeng Li
2021-05-10 8:48 ` Vitaly Kuznetsov
0 siblings, 1 reply; 4+ messages in thread
From: Wanpeng Li @ 2021-05-10 8:21 UTC (permalink / raw)
To: linux-kernel, kvm
Cc: Paolo Bonzini, Sean Christopherson, Vitaly Kuznetsov, Wanpeng Li,
Jim Mattson, Joerg Roedel
From: Wanpeng Li <wanpengli@tencent.com>
WARNING: suspicious RCU usage
5.13.0-rc1 #4 Not tainted
-----------------------------
./include/linux/kvm_host.h:710 suspicious rcu_dereference_check() usage!
other info that might help us debug this:
rcu_scheduler_active = 2, debug_locks = 1
1 lock held by hyperv_clock/8318:
#0: ffffb6b8cb05a7d8 (&hv->hv_lock){+.+.}-{3:3}, at: kvm_hv_invalidate_tsc_page+0x3e/0xa0 [kvm]
stack backtrace:
CPU: 3 PID: 8318 Comm: hyperv_clock Not tainted 5.13.0-rc1 #4
Call Trace:
dump_stack+0x87/0xb7
lockdep_rcu_suspicious+0xce/0xf0
kvm_write_guest_page+0x1c1/0x1d0 [kvm]
kvm_write_guest+0x50/0x90 [kvm]
kvm_hv_invalidate_tsc_page+0x79/0xa0 [kvm]
kvm_gen_update_masterclock+0x1d/0x110 [kvm]
kvm_arch_vm_ioctl+0x2a7/0xc50 [kvm]
kvm_vm_ioctl+0x123/0x11d0 [kvm]
__x64_sys_ioctl+0x3ed/0x9d0
do_syscall_64+0x3d/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xae
kvm_memslots() will be called by kvm_write_guest(), so we should take the srcu lock.
Fixes: e880c6ea5 (KVM: x86: hyper-v: Prevent using not-yet-updated TSC page by secondary CPUs)
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
arch/x86/kvm/hyperv.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
index f98370a3..f00830e 100644
--- a/arch/x86/kvm/hyperv.c
+++ b/arch/x86/kvm/hyperv.c
@@ -1172,6 +1172,7 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
{
struct kvm_hv *hv = to_kvm_hv(kvm);
u64 gfn;
+ int idx;
if (hv->hv_tsc_page_status == HV_TSC_PAGE_BROKEN ||
hv->hv_tsc_page_status == HV_TSC_PAGE_UNSET ||
@@ -1190,9 +1191,16 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
gfn = hv->hv_tsc_page >> HV_X64_MSR_TSC_REFERENCE_ADDRESS_SHIFT;
hv->tsc_ref.tsc_sequence = 0;
+
+ /*
+ * Take the srcu lock as memslots will be accessed to check the gfn
+ * cache generation against the memslots generation.
+ */
+ idx = srcu_read_lock(&kvm->srcu);
if (kvm_write_guest(kvm, gfn_to_gpa(gfn),
&hv->tsc_ref, sizeof(hv->tsc_ref.tsc_sequence)))
hv->hv_tsc_page_status = HV_TSC_PAGE_BROKEN;
+ srcu_read_unlock(&kvm->srcu, idx);
out_unlock:
mutex_unlock(&hv->hv_lock);
--
2.7.4
^ permalink raw reply related [flat|nested] 4+ messages in thread* Re: [PATCH] KVM: x86: hyper-v: Task srcu lock when accessing kvm_memslots()
2021-05-10 8:21 [PATCH] KVM: x86: hyper-v: Task srcu lock when accessing kvm_memslots() Wanpeng Li
@ 2021-05-10 8:48 ` Vitaly Kuznetsov
2021-05-10 8:51 ` Wanpeng Li
0 siblings, 1 reply; 4+ messages in thread
From: Vitaly Kuznetsov @ 2021-05-10 8:48 UTC (permalink / raw)
To: Wanpeng Li
Cc: Paolo Bonzini, Sean Christopherson, Wanpeng Li, Jim Mattson,
Joerg Roedel, linux-kernel, kvm
Wanpeng Li <kernellwp@gmail.com> writes:
> From: Wanpeng Li <wanpengli@tencent.com>
>
> WARNING: suspicious RCU usage
> 5.13.0-rc1 #4 Not tainted
> -----------------------------
> ./include/linux/kvm_host.h:710 suspicious rcu_dereference_check() usage!
>
> other info that might help us debug this:
>
> rcu_scheduler_active = 2, debug_locks = 1
> 1 lock held by hyperv_clock/8318:
> #0: ffffb6b8cb05a7d8 (&hv->hv_lock){+.+.}-{3:3}, at: kvm_hv_invalidate_tsc_page+0x3e/0xa0 [kvm]
>
> stack backtrace:
> CPU: 3 PID: 8318 Comm: hyperv_clock Not tainted 5.13.0-rc1 #4
> Call Trace:
> dump_stack+0x87/0xb7
> lockdep_rcu_suspicious+0xce/0xf0
> kvm_write_guest_page+0x1c1/0x1d0 [kvm]
> kvm_write_guest+0x50/0x90 [kvm]
> kvm_hv_invalidate_tsc_page+0x79/0xa0 [kvm]
> kvm_gen_update_masterclock+0x1d/0x110 [kvm]
> kvm_arch_vm_ioctl+0x2a7/0xc50 [kvm]
> kvm_vm_ioctl+0x123/0x11d0 [kvm]
> __x64_sys_ioctl+0x3ed/0x9d0
> do_syscall_64+0x3d/0x80
> entry_SYSCALL_64_after_hwframe+0x44/0xae
>
> kvm_memslots() will be called by kvm_write_guest(), so we should take the srcu lock.
>
> Fixes: e880c6ea5 (KVM: x86: hyper-v: Prevent using not-yet-updated TSC page by secondary CPUs)
> Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
> ---
> arch/x86/kvm/hyperv.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
> index f98370a3..f00830e 100644
> --- a/arch/x86/kvm/hyperv.c
> +++ b/arch/x86/kvm/hyperv.c
> @@ -1172,6 +1172,7 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
> {
> struct kvm_hv *hv = to_kvm_hv(kvm);
> u64 gfn;
> + int idx;
>
> if (hv->hv_tsc_page_status == HV_TSC_PAGE_BROKEN ||
> hv->hv_tsc_page_status == HV_TSC_PAGE_UNSET ||
> @@ -1190,9 +1191,16 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
> gfn = hv->hv_tsc_page >> HV_X64_MSR_TSC_REFERENCE_ADDRESS_SHIFT;
>
> hv->tsc_ref.tsc_sequence = 0;
> +
> + /*
> + * Take the srcu lock as memslots will be accessed to check the gfn
> + * cache generation against the memslots generation.
> + */
> + idx = srcu_read_lock(&kvm->srcu);
> if (kvm_write_guest(kvm, gfn_to_gpa(gfn),
> &hv->tsc_ref, sizeof(hv->tsc_ref.tsc_sequence)))
> hv->hv_tsc_page_status = HV_TSC_PAGE_BROKEN;
> + srcu_read_unlock(&kvm->srcu, idx);
>
> out_unlock:
> mutex_unlock(&hv->hv_lock);
Thanks!
Do we need to do the same in kvm_hv_setup_tsc_page()?
--
Vitaly
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH] KVM: x86: hyper-v: Task srcu lock when accessing kvm_memslots()
2021-05-10 8:48 ` Vitaly Kuznetsov
@ 2021-05-10 8:51 ` Wanpeng Li
2021-05-10 11:15 ` Vitaly Kuznetsov
0 siblings, 1 reply; 4+ messages in thread
From: Wanpeng Li @ 2021-05-10 8:51 UTC (permalink / raw)
To: Vitaly Kuznetsov
Cc: Paolo Bonzini, Sean Christopherson, Wanpeng Li, Jim Mattson,
Joerg Roedel, LKML, kvm
On Mon, 10 May 2021 at 16:48, Vitaly Kuznetsov <vkuznets@redhat.com> wrote:
>
> Wanpeng Li <kernellwp@gmail.com> writes:
>
> > From: Wanpeng Li <wanpengli@tencent.com>
> >
> > WARNING: suspicious RCU usage
> > 5.13.0-rc1 #4 Not tainted
> > -----------------------------
> > ./include/linux/kvm_host.h:710 suspicious rcu_dereference_check() usage!
> >
> > other info that might help us debug this:
> >
> > rcu_scheduler_active = 2, debug_locks = 1
> > 1 lock held by hyperv_clock/8318:
> > #0: ffffb6b8cb05a7d8 (&hv->hv_lock){+.+.}-{3:3}, at: kvm_hv_invalidate_tsc_page+0x3e/0xa0 [kvm]
> >
> > stack backtrace:
> > CPU: 3 PID: 8318 Comm: hyperv_clock Not tainted 5.13.0-rc1 #4
> > Call Trace:
> > dump_stack+0x87/0xb7
> > lockdep_rcu_suspicious+0xce/0xf0
> > kvm_write_guest_page+0x1c1/0x1d0 [kvm]
> > kvm_write_guest+0x50/0x90 [kvm]
> > kvm_hv_invalidate_tsc_page+0x79/0xa0 [kvm]
> > kvm_gen_update_masterclock+0x1d/0x110 [kvm]
> > kvm_arch_vm_ioctl+0x2a7/0xc50 [kvm]
> > kvm_vm_ioctl+0x123/0x11d0 [kvm]
> > __x64_sys_ioctl+0x3ed/0x9d0
> > do_syscall_64+0x3d/0x80
> > entry_SYSCALL_64_after_hwframe+0x44/0xae
> >
> > kvm_memslots() will be called by kvm_write_guest(), so we should take the srcu lock.
> >
> > Fixes: e880c6ea5 (KVM: x86: hyper-v: Prevent using not-yet-updated TSC page by secondary CPUs)
> > Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
> > ---
> > arch/x86/kvm/hyperv.c | 8 ++++++++
> > 1 file changed, 8 insertions(+)
> >
> > diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
> > index f98370a3..f00830e 100644
> > --- a/arch/x86/kvm/hyperv.c
> > +++ b/arch/x86/kvm/hyperv.c
> > @@ -1172,6 +1172,7 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
> > {
> > struct kvm_hv *hv = to_kvm_hv(kvm);
> > u64 gfn;
> > + int idx;
> >
> > if (hv->hv_tsc_page_status == HV_TSC_PAGE_BROKEN ||
> > hv->hv_tsc_page_status == HV_TSC_PAGE_UNSET ||
> > @@ -1190,9 +1191,16 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
> > gfn = hv->hv_tsc_page >> HV_X64_MSR_TSC_REFERENCE_ADDRESS_SHIFT;
> >
> > hv->tsc_ref.tsc_sequence = 0;
> > +
> > + /*
> > + * Take the srcu lock as memslots will be accessed to check the gfn
> > + * cache generation against the memslots generation.
> > + */
> > + idx = srcu_read_lock(&kvm->srcu);
> > if (kvm_write_guest(kvm, gfn_to_gpa(gfn),
> > &hv->tsc_ref, sizeof(hv->tsc_ref.tsc_sequence)))
> > hv->hv_tsc_page_status = HV_TSC_PAGE_BROKEN;
> > + srcu_read_unlock(&kvm->srcu, idx);
> >
> > out_unlock:
> > mutex_unlock(&hv->hv_lock);
>
> Thanks!
>
> Do we need to do the same in kvm_hv_setup_tsc_page()?
kvm_hv_setup_tsc_page() is called in vcpu_enter_guest() path which has
already held kvm->srcu lock.
Wanpeng
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH] KVM: x86: hyper-v: Task srcu lock when accessing kvm_memslots()
2021-05-10 8:51 ` Wanpeng Li
@ 2021-05-10 11:15 ` Vitaly Kuznetsov
0 siblings, 0 replies; 4+ messages in thread
From: Vitaly Kuznetsov @ 2021-05-10 11:15 UTC (permalink / raw)
To: Wanpeng Li
Cc: Paolo Bonzini, Sean Christopherson, Wanpeng Li, Jim Mattson,
Joerg Roedel, LKML, kvm
Wanpeng Li <kernellwp@gmail.com> writes:
> On Mon, 10 May 2021 at 16:48, Vitaly Kuznetsov <vkuznets@redhat.com> wrote:
>>
>> Wanpeng Li <kernellwp@gmail.com> writes:
>>
>> > From: Wanpeng Li <wanpengli@tencent.com>
>> >
>> > WARNING: suspicious RCU usage
>> > 5.13.0-rc1 #4 Not tainted
>> > -----------------------------
>> > ./include/linux/kvm_host.h:710 suspicious rcu_dereference_check() usage!
>> >
>> > other info that might help us debug this:
>> >
>> > rcu_scheduler_active = 2, debug_locks = 1
>> > 1 lock held by hyperv_clock/8318:
>> > #0: ffffb6b8cb05a7d8 (&hv->hv_lock){+.+.}-{3:3}, at: kvm_hv_invalidate_tsc_page+0x3e/0xa0 [kvm]
>> >
>> > stack backtrace:
>> > CPU: 3 PID: 8318 Comm: hyperv_clock Not tainted 5.13.0-rc1 #4
>> > Call Trace:
>> > dump_stack+0x87/0xb7
>> > lockdep_rcu_suspicious+0xce/0xf0
>> > kvm_write_guest_page+0x1c1/0x1d0 [kvm]
>> > kvm_write_guest+0x50/0x90 [kvm]
>> > kvm_hv_invalidate_tsc_page+0x79/0xa0 [kvm]
>> > kvm_gen_update_masterclock+0x1d/0x110 [kvm]
>> > kvm_arch_vm_ioctl+0x2a7/0xc50 [kvm]
>> > kvm_vm_ioctl+0x123/0x11d0 [kvm]
>> > __x64_sys_ioctl+0x3ed/0x9d0
>> > do_syscall_64+0x3d/0x80
>> > entry_SYSCALL_64_after_hwframe+0x44/0xae
>> >
>> > kvm_memslots() will be called by kvm_write_guest(), so we should take the srcu lock.
>> >
>> > Fixes: e880c6ea5 (KVM: x86: hyper-v: Prevent using not-yet-updated TSC page by secondary CPUs)
>> > Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
>> > ---
>> > arch/x86/kvm/hyperv.c | 8 ++++++++
>> > 1 file changed, 8 insertions(+)
>> >
>> > diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
>> > index f98370a3..f00830e 100644
>> > --- a/arch/x86/kvm/hyperv.c
>> > +++ b/arch/x86/kvm/hyperv.c
>> > @@ -1172,6 +1172,7 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
>> > {
>> > struct kvm_hv *hv = to_kvm_hv(kvm);
>> > u64 gfn;
>> > + int idx;
>> >
>> > if (hv->hv_tsc_page_status == HV_TSC_PAGE_BROKEN ||
>> > hv->hv_tsc_page_status == HV_TSC_PAGE_UNSET ||
>> > @@ -1190,9 +1191,16 @@ void kvm_hv_invalidate_tsc_page(struct kvm *kvm)
>> > gfn = hv->hv_tsc_page >> HV_X64_MSR_TSC_REFERENCE_ADDRESS_SHIFT;
>> >
>> > hv->tsc_ref.tsc_sequence = 0;
>> > +
>> > + /*
>> > + * Take the srcu lock as memslots will be accessed to check the gfn
>> > + * cache generation against the memslots generation.
>> > + */
>> > + idx = srcu_read_lock(&kvm->srcu);
>> > if (kvm_write_guest(kvm, gfn_to_gpa(gfn),
>> > &hv->tsc_ref, sizeof(hv->tsc_ref.tsc_sequence)))
>> > hv->hv_tsc_page_status = HV_TSC_PAGE_BROKEN;
>> > + srcu_read_unlock(&kvm->srcu, idx);
>> >
>> > out_unlock:
>> > mutex_unlock(&hv->hv_lock);
>>
>> Thanks!
>>
>> Do we need to do the same in kvm_hv_setup_tsc_page()?
>
> kvm_hv_setup_tsc_page() is called in vcpu_enter_guest() path which has
> already held kvm->srcu lock.
>
I, true, thanks for checking!
For the patch:
Reviewed-by: Vitaly Kuznetsov <vkuznets@redhat.com>
--
Vitaly
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2021-05-10 11:22 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-10 8:21 [PATCH] KVM: x86: hyper-v: Task srcu lock when accessing kvm_memslots() Wanpeng Li
2021-05-10 8:48 ` Vitaly Kuznetsov
2021-05-10 8:51 ` Wanpeng Li
2021-05-10 11:15 ` Vitaly Kuznetsov
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).