tools/libxc/include/xenctrl.h | 2 ++ tools/libxc/xc_altp2m.c | 24 ++++++++++++++++++++++++ xen/arch/x86/hvm/hvm.c | 18 ++++++++++++++++++ xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/mm/hap/hap.c | 15 +++++++++++++++ xen/arch/x86/mm/p2m-ept.c | 1 + xen/arch/x86/mm/p2m.c | 7 +++++-- xen/include/asm-x86/domain.h | 1 + xen/include/public/hvm/hvm_op.h | 9 +++++++++ 9 files changed, 76 insertions(+), 3 deletions(-)
At this moment a guest can call vmfunc to change the altp2m view. This
should be limited in order to avoid any unwanted view switch.
The new xc_altp2m_set_visibility() solves this by making views invisible
to vmfunc.
This is done by having a separate arch.altp2m_working_eptp that is
populated and made invalid in the same places as altp2m_eptp. This is
written to EPTP_LIST_ADDR.
The views are made in/visible by marking them with INVALID_MFN or
copying them back from altp2m_eptp.
To have consistency the visibility also applies to
p2m_switch_domain_altp2m_by_id().
Signed-off-by: Alexandru Isaila <aisaila@bitdefender.com>
---
CC: Ian Jackson <ian.jackson@eu.citrix.com>
CC: Wei Liu <wl@xen.org>
CC: Andrew Cooper <andrew.cooper3@citrix.com>
CC: George Dunlap <George.Dunlap@eu.citrix.com>
CC: Jan Beulich <jbeulich@suse.com>
CC: Julien Grall <julien@xen.org>
CC: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
CC: Stefano Stabellini <sstabellini@kernel.org>
CC: "Roger Pau Monné" <roger.pau@citrix.com>
CC: Jun Nakajima <jun.nakajima@intel.com>
CC: Kevin Tian <kevin.tian@intel.com>
CC: George Dunlap <george.dunlap@eu.citrix.com>
---
Changes since V2:
- Drop hap_enabled() check
- Reduce the indentation depth in hvm.c
- Fix assignment indentation
- Drop pad2.
Changes since V1:
- Drop double view from title.
---
tools/libxc/include/xenctrl.h | 2 ++
tools/libxc/xc_altp2m.c | 24 ++++++++++++++++++++++++
xen/arch/x86/hvm/hvm.c | 18 ++++++++++++++++++
xen/arch/x86/hvm/vmx/vmx.c | 2 +-
xen/arch/x86/mm/hap/hap.c | 15 +++++++++++++++
xen/arch/x86/mm/p2m-ept.c | 1 +
xen/arch/x86/mm/p2m.c | 7 +++++--
xen/include/asm-x86/domain.h | 1 +
xen/include/public/hvm/hvm_op.h | 9 +++++++++
9 files changed, 76 insertions(+), 3 deletions(-)
diff --git a/tools/libxc/include/xenctrl.h b/tools/libxc/include/xenctrl.h
index 99552a5f73..414adb143b 100644
--- a/tools/libxc/include/xenctrl.h
+++ b/tools/libxc/include/xenctrl.h
@@ -1943,6 +1943,8 @@ int xc_altp2m_change_gfn(xc_interface *handle, uint32_t domid,
xen_pfn_t new_gfn);
int xc_altp2m_get_vcpu_p2m_idx(xc_interface *handle, uint32_t domid,
uint32_t vcpuid, uint16_t *p2midx);
+int xc_altp2m_set_visibility(xc_interface *handle, uint32_t domid,
+ uint16_t view_id, bool visible);
/**
* Mem paging operations.
diff --git a/tools/libxc/xc_altp2m.c b/tools/libxc/xc_altp2m.c
index 46fb725806..6987c9541f 100644
--- a/tools/libxc/xc_altp2m.c
+++ b/tools/libxc/xc_altp2m.c
@@ -410,3 +410,27 @@ int xc_altp2m_get_vcpu_p2m_idx(xc_interface *handle, uint32_t domid,
xc_hypercall_buffer_free(handle, arg);
return rc;
}
+
+int xc_altp2m_set_visibility(xc_interface *handle, uint32_t domid,
+ uint16_t view_id, bool visible)
+{
+ int rc;
+
+ DECLARE_HYPERCALL_BUFFER(xen_hvm_altp2m_op_t, arg);
+
+ arg = xc_hypercall_buffer_alloc(handle, arg, sizeof(*arg));
+ if ( arg == NULL )
+ return -1;
+
+ arg->version = HVMOP_ALTP2M_INTERFACE_VERSION;
+ arg->cmd = HVMOP_altp2m_set_visibility;
+ arg->domain = domid;
+ arg->u.set_visibility.altp2m_idx = view_id;
+ arg->u.set_visibility.visible = visible;
+
+ rc = xencall2(handle->xcall, __HYPERVISOR_hvm_op, HVMOP_altp2m,
+ HYPERCALL_BUFFER_AS_ARG(arg));
+
+ xc_hypercall_buffer_free(handle, arg);
+ return rc;
+}
diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
index 00a9e70b7c..24035261f4 100644
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -4558,6 +4558,7 @@ static int do_altp2m_op(
case HVMOP_altp2m_get_mem_access:
case HVMOP_altp2m_change_gfn:
case HVMOP_altp2m_get_p2m_idx:
+ case HVMOP_altp2m_set_visibility:
break;
default:
@@ -4835,6 +4836,23 @@ static int do_altp2m_op(
break;
}
+ case HVMOP_altp2m_set_visibility:
+ {
+ uint16_t altp2m_idx = a.u.set_visibility.altp2m_idx;
+
+ if ( a.u.set_visibility.pad )
+ rc = -EINVAL;
+ else if ( !altp2m_active(d) )
+ rc = -EOPNOTSUPP;
+ else if ( a.u.set_visibility.visible )
+ d->arch.altp2m_working_eptp[altp2m_idx] =
+ d->arch.altp2m_eptp[altp2m_idx];
+ else
+ d->arch.altp2m_working_eptp[altp2m_idx] =
+ mfn_x(INVALID_MFN);
+ break;
+ }
+
default:
ASSERT_UNREACHABLE();
}
diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c
index bc2f48bf2c..90de909fec 100644
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -2140,7 +2140,7 @@ static void vmx_vcpu_update_vmfunc_ve(struct vcpu *v)
{
v->arch.hvm.vmx.secondary_exec_control |= mask;
__vmwrite(VM_FUNCTION_CONTROL, VMX_VMFUNC_EPTP_SWITCHING);
- __vmwrite(EPTP_LIST_ADDR, virt_to_maddr(d->arch.altp2m_eptp));
+ __vmwrite(EPTP_LIST_ADDR, virt_to_maddr(d->arch.altp2m_working_eptp));
if ( cpu_has_vmx_virt_exceptions )
{
diff --git a/xen/arch/x86/mm/hap/hap.c b/xen/arch/x86/mm/hap/hap.c
index 3d93f3451c..5969ec8922 100644
--- a/xen/arch/x86/mm/hap/hap.c
+++ b/xen/arch/x86/mm/hap/hap.c
@@ -488,8 +488,17 @@ int hap_enable(struct domain *d, u32 mode)
goto out;
}
+ if ( (d->arch.altp2m_working_eptp = alloc_xenheap_page()) == NULL )
+ {
+ rv = -ENOMEM;
+ goto out;
+ }
+
for ( i = 0; i < MAX_EPTP; i++ )
+ {
d->arch.altp2m_eptp[i] = mfn_x(INVALID_MFN);
+ d->arch.altp2m_working_eptp[i] = mfn_x(INVALID_MFN);
+ }
for ( i = 0; i < MAX_ALTP2M; i++ )
{
@@ -523,6 +532,12 @@ void hap_final_teardown(struct domain *d)
d->arch.altp2m_eptp = NULL;
}
+ if ( d->arch.altp2m_working_eptp )
+ {
+ free_xenheap_page(d->arch.altp2m_working_eptp);
+ d->arch.altp2m_working_eptp = NULL;
+ }
+
for ( i = 0; i < MAX_ALTP2M; i++ )
p2m_teardown(d->arch.altp2m_p2m[i]);
}
diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c
index eb0f0edfef..6539ca619b 100644
--- a/xen/arch/x86/mm/p2m-ept.c
+++ b/xen/arch/x86/mm/p2m-ept.c
@@ -1368,6 +1368,7 @@ void p2m_init_altp2m_ept(struct domain *d, unsigned int i)
ept = &p2m->ept;
ept->mfn = pagetable_get_pfn(p2m_get_pagetable(p2m));
d->arch.altp2m_eptp[array_index_nospec(i, MAX_EPTP)] = ept->eptp;
+ d->arch.altp2m_working_eptp[array_index_nospec(i, MAX_EPTP)] = ept->eptp;
}
unsigned int p2m_find_altp2m_by_eptp(struct domain *d, uint64_t eptp)
diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c
index 9f1c29d7ef..13b96715ba 100644
--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -2519,6 +2519,7 @@ void p2m_flush_altp2m(struct domain *d)
{
p2m_reset_altp2m(d, i, ALTP2M_DEACTIVATE);
d->arch.altp2m_eptp[i] = mfn_x(INVALID_MFN);
+ d->arch.altp2m_working_eptp[i] = mfn_x(INVALID_MFN);
}
altp2m_list_unlock(d);
@@ -2638,7 +2639,9 @@ int p2m_destroy_altp2m_by_id(struct domain *d, unsigned int idx)
{
p2m_reset_altp2m(d, idx, ALTP2M_DEACTIVATE);
d->arch.altp2m_eptp[array_index_nospec(idx, MAX_EPTP)] =
- mfn_x(INVALID_MFN);
+ mfn_x(INVALID_MFN);
+ d->arch.altp2m_working_eptp[array_index_nospec(idx, MAX_EPTP)] =
+ mfn_x(INVALID_MFN);
rc = 0;
}
}
@@ -2665,7 +2668,7 @@ int p2m_switch_domain_altp2m_by_id(struct domain *d, unsigned int idx)
rc = -EINVAL;
altp2m_list_lock(d);
- if ( d->arch.altp2m_eptp[idx] != mfn_x(INVALID_MFN) )
+ if ( d->arch.altp2m_working_eptp[idx] != mfn_x(INVALID_MFN) )
{
for_each_vcpu( d, v )
if ( idx != vcpu_altp2m(v).p2midx )
diff --git a/xen/include/asm-x86/domain.h b/xen/include/asm-x86/domain.h
index 1843c76d1a..b039717778 100644
--- a/xen/include/asm-x86/domain.h
+++ b/xen/include/asm-x86/domain.h
@@ -326,6 +326,7 @@ struct arch_domain
struct p2m_domain *altp2m_p2m[MAX_ALTP2M];
mm_lock_t altp2m_list_lock;
uint64_t *altp2m_eptp;
+ uint64_t *altp2m_working_eptp;
#endif
/* NB. protected by d->event_lock and by irq_desc[irq].lock */
diff --git a/xen/include/public/hvm/hvm_op.h b/xen/include/public/hvm/hvm_op.h
index b599d3cbd0..870ec52060 100644
--- a/xen/include/public/hvm/hvm_op.h
+++ b/xen/include/public/hvm/hvm_op.h
@@ -318,6 +318,12 @@ struct xen_hvm_altp2m_get_vcpu_p2m_idx {
uint16_t altp2m_idx;
};
+struct xen_hvm_altp2m_set_visibility {
+ uint16_t altp2m_idx;
+ uint8_t visible;
+ uint8_t pad;
+};
+
struct xen_hvm_altp2m_op {
uint32_t version; /* HVMOP_ALTP2M_INTERFACE_VERSION */
uint32_t cmd;
@@ -350,6 +356,8 @@ struct xen_hvm_altp2m_op {
#define HVMOP_altp2m_get_p2m_idx 14
/* Set the "Supress #VE" bit for a range of pages */
#define HVMOP_altp2m_set_suppress_ve_multi 15
+/* Set visibility for a given altp2m view */
+#define HVMOP_altp2m_set_visibility 16
domid_t domain;
uint16_t pad1;
uint32_t pad2;
@@ -367,6 +375,7 @@ struct xen_hvm_altp2m_op {
struct xen_hvm_altp2m_suppress_ve_multi suppress_ve_multi;
struct xen_hvm_altp2m_vcpu_disable_notify disable_notify;
struct xen_hvm_altp2m_get_vcpu_p2m_idx get_vcpu_p2m_idx;
+ struct xen_hvm_altp2m_set_visibility set_visibility;
uint8_t pad[64];
} u;
};
--
2.17.1
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-develOn Wed, Feb 19, 2020 at 2:19 AM Alexandru Stefan ISAILA <aisaila@bitdefender.com> wrote: > > At this moment a guest can call vmfunc to change the altp2m view. This > should be limited in order to avoid any unwanted view switch. > > The new xc_altp2m_set_visibility() solves this by making views invisible > to vmfunc. > This is done by having a separate arch.altp2m_working_eptp that is > populated and made invalid in the same places as altp2m_eptp. This is > written to EPTP_LIST_ADDR. > The views are made in/visible by marking them with INVALID_MFN or > copying them back from altp2m_eptp. > To have consistency the visibility also applies to > p2m_switch_domain_altp2m_by_id(). I'm just wondering, what prevents the guest from calling this HVM op before doing vmfunc? This seems to only make a difference in case the altp2m mode is either set as external or limited (or have a more fine-grained XSM policy loaded). Is that correct? If so, perhaps mention that in the commit message and as a comment on the libxc function so that people don't get a false sense of security when using the mixed mode. Thanks, Tamas _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
On 19.02.2020 18:37, Tamas K Lengyel wrote: > On Wed, Feb 19, 2020 at 2:19 AM Alexandru Stefan ISAILA > <aisaila@bitdefender.com> wrote: >> >> At this moment a guest can call vmfunc to change the altp2m view. This >> should be limited in order to avoid any unwanted view switch. >> >> The new xc_altp2m_set_visibility() solves this by making views invisible >> to vmfunc. >> This is done by having a separate arch.altp2m_working_eptp that is >> populated and made invalid in the same places as altp2m_eptp. This is >> written to EPTP_LIST_ADDR. >> The views are made in/visible by marking them with INVALID_MFN or >> copying them back from altp2m_eptp. >> To have consistency the visibility also applies to >> p2m_switch_domain_altp2m_by_id(). > > I'm just wondering, what prevents the guest from calling this HVM op > before doing vmfunc? This seems to only make a difference in case the > altp2m mode is either set as external or limited (or have a more > fine-grained XSM policy loaded). Is that correct? If so, perhaps Yes, that is correct. > mention that in the commit message and as a comment on the libxc > function so that people don't get a false sense of security when using > the mixed mode. > I will add this fact in the commit message and in libxc. Good thing to point that out. Thanks, Alex _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
On 19.02.2020 10:18, Alexandru Stefan ISAILA wrote:
> @@ -4835,6 +4836,23 @@ static int do_altp2m_op(
> break;
> }
>
> + case HVMOP_altp2m_set_visibility:
> + {
> + uint16_t altp2m_idx = a.u.set_visibility.altp2m_idx;
> +
> + if ( a.u.set_visibility.pad )
> + rc = -EINVAL;
> + else if ( !altp2m_active(d) )
> + rc = -EOPNOTSUPP;
> + else if ( a.u.set_visibility.visible )
> + d->arch.altp2m_working_eptp[altp2m_idx] =
> + d->arch.altp2m_eptp[altp2m_idx];
> + else
> + d->arch.altp2m_working_eptp[altp2m_idx] =
> + mfn_x(INVALID_MFN);
Don't you need to bounds check the index before its use? And
shouldn't you return an error also for in-range ones which
aren't actually valid?
Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
On 19.02.2020 19:00, Jan Beulich wrote:
> On 19.02.2020 10:18, Alexandru Stefan ISAILA wrote:
>> @@ -4835,6 +4836,23 @@ static int do_altp2m_op(
>> break;
>> }
>>
>> + case HVMOP_altp2m_set_visibility:
>> + {
>> + uint16_t altp2m_idx = a.u.set_visibility.altp2m_idx;
>> +
>> + if ( a.u.set_visibility.pad )
>> + rc = -EINVAL;
>> + else if ( !altp2m_active(d) )
>> + rc = -EOPNOTSUPP;
>> + else if ( a.u.set_visibility.visible )
>> + d->arch.altp2m_working_eptp[altp2m_idx] =
>> + d->arch.altp2m_eptp[altp2m_idx];
>> + else
>> + d->arch.altp2m_working_eptp[altp2m_idx] =
>> + mfn_x(INVALID_MFN);
>
> Don't you need to bounds check the index before its use?
Unless we want a index out of bounds from the user. Sorry for not having
that, I will add a "altp2m_eptp[array_index_nospec(altp2m_idx,
MAX_EPTP)]" in place for the next version.
> And
> shouldn't you return an error also for in-range ones which
> aren't actually valid?
That is a good thing. Maybe -EINVAL could fit this?
And together with the bounds check it will end up something like this:
if ( !altp2m_idx < 0 || altp2m_idx >=
min(ARRAY_SIZE(d->arch.altp2m_p2m), MAX_EPTP) ||
altp2m_eptp[array_index_nospec(altp2m_idx, MAX_EPTP)] ==
mfn_x(INVALID_MFN) )
return -EINVAL;
And looking again, it could be coupled with the a.u.set_visibility.pad.
Thanks,
Alex
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
On 20.02.2020 10:59, Alexandru Stefan ISAILA wrote:
> On 19.02.2020 19:00, Jan Beulich wrote:
>> On 19.02.2020 10:18, Alexandru Stefan ISAILA wrote:
>>> @@ -4835,6 +4836,23 @@ static int do_altp2m_op(
>>> break;
>>> }
>>>
>>> + case HVMOP_altp2m_set_visibility:
>>> + {
>>> + uint16_t altp2m_idx = a.u.set_visibility.altp2m_idx;
>>> +
>>> + if ( a.u.set_visibility.pad )
>>> + rc = -EINVAL;
>>> + else if ( !altp2m_active(d) )
>>> + rc = -EOPNOTSUPP;
>>> + else if ( a.u.set_visibility.visible )
>>> + d->arch.altp2m_working_eptp[altp2m_idx] =
>>> + d->arch.altp2m_eptp[altp2m_idx];
>>> + else
>>> + d->arch.altp2m_working_eptp[altp2m_idx] =
>>> + mfn_x(INVALID_MFN);
>>
>> Don't you need to bounds check the index before its use?
>
> Unless we want a index out of bounds from the user. Sorry for not having
> that, I will add a "altp2m_eptp[array_index_nospec(altp2m_idx,
> MAX_EPTP)]" in place for the next version.
>
>> And
>> shouldn't you return an error also for in-range ones which
>> aren't actually valid?
> That is a good thing. Maybe -EINVAL could fit this?
Sure.
Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
On 2/19/20 9:18 AM, Alexandru Stefan ISAILA wrote: > At this moment a guest can call vmfunc to change the altp2m view. This > should be limited in order to avoid any unwanted view switch. > > The new xc_altp2m_set_visibility() solves this by making views invisible > to vmfunc. > This is done by having a separate arch.altp2m_working_eptp that is > populated and made invalid in the same places as altp2m_eptp. This is > written to EPTP_LIST_ADDR. > The views are made in/visible by marking them with INVALID_MFN or > copying them back from altp2m_eptp. > To have consistency the visibility also applies to > p2m_switch_domain_altp2m_by_id(). So it looks like by default the views are visible, until they're made non-visible? Also, does the last line mean that the toolstack can't change to a "non-visible" altp2m either? -George _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
On 20.02.2020 14:21, George Dunlap wrote: > On 2/19/20 9:18 AM, Alexandru Stefan ISAILA wrote: >> At this moment a guest can call vmfunc to change the altp2m view. This >> should be limited in order to avoid any unwanted view switch. >> >> The new xc_altp2m_set_visibility() solves this by making views invisible >> to vmfunc. >> This is done by having a separate arch.altp2m_working_eptp that is >> populated and made invalid in the same places as altp2m_eptp. This is >> written to EPTP_LIST_ADDR. >> The views are made in/visible by marking them with INVALID_MFN or >> copying them back from altp2m_eptp. >> To have consistency the visibility also applies to >> p2m_switch_domain_altp2m_by_id(). > > > So it looks like by default the views are visible, until they're made > non-visible? Yes, by default all the active views are visible until they're made non-visible. > > Also, does the last line mean that the toolstack can't change to a > "non-visible" altp2m either? The last line means that xc_altp2m_switch_to_view() will not be able to switch to a non-visible view. -Alex _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
© 2016 - 2024 Red Hat, Inc.