KVM: xen: update shared_info and vcpu_info handling
Subsystems (6)
Paul Durrant
maintainerv4 cover KVM: xen: update shared_info and vcpu_info handling
From: Paul Durrant <[email protected]>
Changed in v4:
- The offset into the cache returned from get_vcpu_info_cache() was not being factored into kvm_gpc_check() or kvm_gpc_refresh() calls. Fix this.
- When transitioning from a default vcpu_info to an explicit one, copy the content across. This was previously the responsibility of the VMM.
Changed in v3:
- Patch added to make sure Xen vcpu_id is immutable once shared_info is set.
- Adjust the xen_shinfo_test selftest accordingly.
- Also have the selftest use both mechanisms to set shared_info.
- Add text to API documentation discussing copying of vcpu_info. This has been removed in v4.
- Adjust the selftest to switch from default to explicit vcpu_info part way through.
Changed in v2:
- Defer advertizement of KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA to a patch at the end of the series.
- Remove the KVM_XEN_HVM_CONFIG_DEFAULT_VCPU_INFO capability.
- Add selftests and API documentation.
Original text:
Currently we treat the shared_info page as guest memory and the VMM informs KVM of its location using a GFN. However it is not guest memory as such; it's an overlay page. So we pointlessly invalidate and re-cache a mapping to the *same page* of memory every time the guest requests that shared_info be mapped into its address space. Let's avoid doing that by modifying the pfncache code to allow activation using a fixed userspace HVA as well as a GPA.
Also, if the guest does not hypercall to explicitly set a pointer to a vcpu_info in its own memory, the default vcpu_info embedded in the shared_info page should be used. At the moment the VMM has to set up a pointer to the structure explicitly (again treating it like it's in guest memory, despite being in an overlay page). Let's also avoid the need for that. We already have a cached mapping for the shared_info page so just use that directly by default.
Paul Durrant (13):
KVM: pfncache: add a map helper function
KVM: pfncache: add a mark-dirty helper
KVM: pfncache: add a helper to get the gpa
KVM: pfncache: base offset check on khva rather than gpa
KVM: pfncache: allow a cache to be activated with a fixed (userspace)
HVA
KVM: xen: allow shared_info to be mapped by fixed HVA
KVM: xen: prepare for using 'default' vcpu_info
KVM: xen: prevent vcpu_id from changing whilst shared_info is valid
KVM: xen: automatically use the vcpu_info embedded in shared_info
KVM: selftests / xen: set KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID
KVM: selftests / xen: map shared_info using HVA rather than GFN
KVM: selftests / xen: don't explicitly set the vcpu_info address
KVM: xen: advertize the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA capability
Documentation/virt/kvm/api.rst | 52 ++-- arch/x86/include/asm/kvm_host.h | 4 + arch/x86/kvm/x86.c | 17 +- arch/x86/kvm/xen.c | 244 ++++++++++++++---- arch/x86/kvm/xen.h | 6 +- include/linux/kvm_host.h | 43 +++ include/linux/kvm_types.h | 3 +- include/uapi/linux/kvm.h | 6 +- .../selftests/kvm/x86_64/xen_shinfo_test.c | 75 +++++- virt/kvm/pfncache.c | 129 ++++++--- 10 files changed, 454 insertions(+), 125 deletions(-)
---
Cc: "H. Peter Anvin" <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Sean Christopherson <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: [email protected]
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 1/13 KVM: pfncache: add a map helper function
From: Paul Durrant <[email protected]>
We have an unmap helper but mapping is open-coded. Arguably this is fine because mapping is done in only one place, hva_to_pfn_retry(), but adding the helper does make that function more readable.
No functional change intended.
---
Cc: Sean Christopherson <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: Paolo Bonzini <[email protected]>
--- virt/kvm/pfncache.c | 43 +++++++++++++++++++++++++------------------ 1 file changed, 25 insertions(+), 18 deletions(-)
virt/kvm/pfncache.c +25 -18
@@ -96,17 +96,28 @@ bool kvm_gpc_check(struct gfn_to_pfn_cache *gpc, unsigned long len) }EXPORT_SYMBOL_GPL (kvm_gpc_check); -static void gpc_unmap_khva (kvm_pfn_t pfn,void *khva) +static void *gpc_map (kvm_pfn_t pfn) +{ +if (pfn_valid (pfn)) +return kmap (pfn_to_page (pfn)); +#ifdef CONFIG_HAS_IOMEM + else + return memremap(pfn_to_hpa(pfn), PAGE_SIZE, MEMREMAP_WB); +#endif +} + +static void gpc_unmap (kvm_pfn_t pfn,void *khva) {/* Unmap the old pfn/page if it was mapped before. */ -if (!is_error_noslot_pfn (pfn) && khva) { -if (pfn_valid (pfn)) -kunmap (pfn_to_page (pfn)); +if (is_error_noslot_pfn (pfn) || !khva) +return ; + +if (pfn_valid (pfn)) +kunmap (pfn_to_page (pfn));#ifdef CONFIG_HAS_IOMEM - else -memunmap (khva ); + else +memunmap (khva );#endif - } }static inline bool mmu_notifier_retry_cache (struct kvm *kvm,unsigned long mmu_seq)
@@ -175,7 +186,7 @@ static kvm_pfn_t hva_to_pfn_retry(struct gfn_to_pfn_cache *gpc)* the existing mapping and didn't create a new one. */ if (new_khva != old_khva) -gpc_unmap_khva (new_pfn, new_khva); +gpc_unmap (new_pfn, new_khva);kvm_release_pfn_clean (new_pfn);
@@ -193,15 +204,11 @@ static kvm_pfn_t hva_to_pfn_retry(struct gfn_to_pfn_cache *gpc)* too must be done outside of gpc->lock! */ if (gpc->usage &KVM_HOST_USES_PFN ) { -if (new_pfn == gpc->pfn) { +if (new_pfn == gpc->pfn) new_khva = old_khva; - }else if (pfn_valid (new_pfn)) { - new_khva =kmap (pfn_to_page (new_pfn)); -#ifdef CONFIG_HAS_IOMEM - }else { - new_khva =memremap (pfn_to_hpa (new_pfn),PAGE_SIZE ,MEMREMAP_WB ); -#endif - } +else + new_khva =gpc_map (new_pfn); +if (!new_khva) {kvm_release_pfn_clean (new_pfn); goto out_error;
@@ -326,7 +333,7 @@ static int __kvm_gpc_refresh(struct gfn_to_pfn_cache *gpc, gpa_t gpa,mutex_unlock (&gpc->refresh_lock);if (unmap_old) -gpc_unmap_khva (old_pfn, old_khva); +gpc_unmap (old_pfn, old_khva);return ret; }
@@ -412,7 +419,7 @@ void kvm_gpc_deactivate(struct gfn_to_pfn_cache *gpc)list_del (&gpc->list);spin_unlock (&kvm->gpc_lock); -gpc_unmap_khva (old_pfn, old_khva); +gpc_unmap (old_pfn, old_khva); } }EXPORT_SYMBOL_GPL (kvm_gpc_deactivate);
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 2/13 KVM: pfncache: add a mark-dirty helper
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
At the moment pages are marked dirty by open-coded calls to mark_page_dirty_in_slot(), directly deferefencing the gpa and memslot from the cache. After a subsequent patch these may not always be set so add a helper now so that caller will protected from the need to know about this detail.
NOTE: Pages are now marked dirty while the cache lock is held. This is
to ensure that gpa and memslot are mutually consistent.
---
Cc: David Woodhouse <[email protected]>
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: [email protected]
--- arch/x86/kvm/x86.c | 2 +- arch/x86/kvm/xen.c | 13 ++++++------- include/linux/kvm_host.h | 7 +++++++ virt/kvm/pfncache.c | 6 ++++++ 4 files changed, 20 insertions(+), 8 deletions(-)
arch/x86/kvm/x86.c +1 -1
@@ -3137,7 +3137,7 @@ static void kvm_setup_guest_pvclock(struct kvm_vcpu *v, guest_hv_clock->version = ++vcpu->hv_clock.version; -mark_page_dirty_in_slot (v->kvm, gpc->memslot, gpc->gpa >>PAGE_SHIFT ); +kvm_gpc_mark_dirty (gpc);read_unlock_irqrestore (&gpc->lock, flags);trace_kvm_pvclock_update (v->vcpu_id, &vcpu->hv_clock);
arch/x86/kvm/xen.c +6 -7
@@ -430,14 +430,13 @@ static void kvm_xen_update_runstate_guest(struct kvm_vcpu *v, bool atomic)smp_wmb (); } -if (user_len2) +if (user_len2) { +kvm_gpc_mark_dirty (gpc2);read_unlock (&gpc2->lock); + } +kvm_gpc_mark_dirty (gpc1);read_unlock_irqrestore (&gpc1->lock, flags); - -mark_page_dirty_in_slot (v->kvm, gpc1->memslot, gpc1->gpa >>PAGE_SHIFT ); -if (user_len2) -mark_page_dirty_in_slot (v->kvm, gpc2->memslot, gpc2->gpa >>PAGE_SHIFT ); }void kvm_xen_update_runstate (struct kvm_vcpu *v,int state)
@@ -543,13 +542,13 @@ void kvm_xen_inject_pending_events(struct kvm_vcpu *v) : "0" (evtchn_pending_sel32));WRITE_ONCE (vi->evtchn_upcall_pending,1 ); } + +kvm_gpc_mark_dirty (gpc);read_unlock_irqrestore (&gpc->lock, flags);/* For the per-vCPU lapic vector, deliver it as MSI. */ if (v->arch.xen.upcall_vector)kvm_xen_inject_vcpu_vector (v); - -mark_page_dirty_in_slot (v->kvm, gpc->memslot, gpc->gpa >>PAGE_SHIFT ); }int __kvm_xen_has_interrupt (struct kvm_vcpu *v)
include/linux/kvm_host.h +7 -0
@@ -1367,6 +1367,13 @@ int kvm_gpc_refresh(struct gfn_to_pfn_cache *gpc, unsigned long len);*/ void kvm_gpc_deactivate (struct gfn_to_pfn_cache *gpc); +/** +* kvm_gpc_mark_dirty - mark a cached page as dirty. +* +* @gpc: struct gfn_to_pfn_cache object. +*/ +void kvm_gpc_mark_dirty (struct gfn_to_pfn_cache *gpc); +void kvm_sigset_activate (struct kvm_vcpu *vcpu);void kvm_sigset_deactivate (struct kvm_vcpu *vcpu);
virt/kvm/pfncache.c +6 -0
@@ -386,6 +386,12 @@ int kvm_gpc_activate(struct gfn_to_pfn_cache *gpc, gpa_t gpa, unsigned long len) }EXPORT_SYMBOL_GPL (kvm_gpc_activate); +void kvm_gpc_mark_dirty (struct gfn_to_pfn_cache *gpc) +{ +mark_page_dirty_in_slot (gpc->kvm, gpc->memslot, gpc->gpa >>PAGE_SHIFT ); +} +EXPORT_SYMBOL_GPL (kvm_gpc_mark_dirty); +void kvm_gpc_deactivate (struct gfn_to_pfn_cache *gpc) {struct kvm *kvm = gpc->kvm;
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 3/13 KVM: pfncache: add a helper to get the gpa
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
A subsequent patch will rename this field since it will become overloaded. To avoid churn in places that currently retrieve the gpa, add a helper for that purpose now.
No functional change intended.
---
Cc: David Woodhouse <[email protected]>
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: [email protected]
--- arch/x86/kvm/xen.c | 15 ++++++++------- include/linux/kvm_host.h | 7 +++++++ virt/kvm/pfncache.c | 6 ++++++ 3 files changed, 21 insertions(+), 7 deletions(-)
arch/x86/kvm/xen.c +8 -7
@@ -261,8 +261,8 @@ static void kvm_xen_update_runstate_guest(struct kvm_vcpu *v, bool atomic)* alignment (and the 32-bit ABI doesn't align the 64-bit integers * anyway, even if the overall struct had been 64-bit aligned). */ -if ((gpc1->gpa & ~PAGE_MASK ) + user_len >=PAGE_SIZE ) { - user_len1 =PAGE_SIZE - (gpc1->gpa & ~PAGE_MASK ); +if ((kvm_gpc_gpa (gpc1) & ~PAGE_MASK ) + user_len >=PAGE_SIZE ) { + user_len1 =PAGE_SIZE - (kvm_gpc_gpa (gpc1) & ~PAGE_MASK ); user_len2 = user_len - user_len1; }else { user_len1 = user_len;
@@ -343,7 +343,7 @@ static void kvm_xen_update_runstate_guest(struct kvm_vcpu *v, bool atomic)* to the second page now because the guest changed to * 64-bit mode, the second GPC won't have been set up. */ -if (kvm_gpc_activate (gpc2, gpc1->gpa + user_len1, +if (kvm_gpc_activate (gpc2,kvm_gpc_gpa (gpc1) + user_len1, user_len2))return ;
@@ -677,7 +677,8 @@ int kvm_xen_hvm_get_attr(struct kvm *kvm, struct kvm_xen_hvm_attr *data)case KVM_XEN_ATTR_TYPE_SHARED_INFO :if (kvm->arch.xen.shinfo_cache.active) - data->u.shared_info.gfn =gpa_to_gfn (kvm->arch.xen.shinfo_cache.gpa); + data->u.shared_info.gfn = +gpa_to_gfn (kvm_gpc_gpa (&kvm->arch.xen.shinfo_cache));else data->u.shared_info.gfn =KVM_XEN_INVALID_GFN ; r =0 ;
@@ -955,7 +956,7 @@ int kvm_xen_vcpu_get_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)switch (data->type) {case KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO :if (vcpu->arch.xen.vcpu_info_cache.active) - data->u.gpa = vcpu->arch.xen.vcpu_info_cache.gpa; + data->u.gpa =kvm_gpc_gpa (&vcpu->arch.xen.vcpu_info_cache);else data->u.gpa =KVM_XEN_INVALID_GPA ; r =0 ;
@@ -963,7 +964,7 @@ int kvm_xen_vcpu_get_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)case KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO :if (vcpu->arch.xen.vcpu_time_info_cache.active) - data->u.gpa = vcpu->arch.xen.vcpu_time_info_cache.gpa; + data->u.gpa =kvm_gpc_gpa (&vcpu->arch.xen.vcpu_time_info_cache);else data->u.gpa =KVM_XEN_INVALID_GPA ; r =0 ;
@@ -975,7 +976,7 @@ int kvm_xen_vcpu_get_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)break ; }if (vcpu->arch.xen.runstate_cache.active) { - data->u.gpa = vcpu->arch.xen.runstate_cache.gpa; + data->u.gpa =kvm_gpc_gpa (&vcpu->arch.xen.runstate_cache); r =0 ; }break ;
include/linux/kvm_host.h +7 -0
@@ -1374,6 +1374,13 @@ void kvm_gpc_deactivate(struct gfn_to_pfn_cache *gpc);*/ void kvm_gpc_mark_dirty (struct gfn_to_pfn_cache *gpc); +/** +* kvm_gpc_gpa - retrieve the guest physical address of a cached mapping +* +* @gpc: struct gfn_to_pfn_cache object. +*/ +gpa_t kvm_gpc_gpa (struct gfn_to_pfn_cache *gpc); +void kvm_sigset_activate (struct kvm_vcpu *vcpu);void kvm_sigset_deactivate (struct kvm_vcpu *vcpu);
virt/kvm/pfncache.c +6 -0
@@ -386,6 +386,12 @@ int kvm_gpc_activate(struct gfn_to_pfn_cache *gpc, gpa_t gpa, unsigned long len) }EXPORT_SYMBOL_GPL (kvm_gpc_activate); +gpa_t kvm_gpc_gpa (struct gfn_to_pfn_cache *gpc) +{ +return gpc->gpa; +} +EXPORT_SYMBOL_GPL (kvm_gpc_gpa); +void kvm_gpc_mark_dirty (struct gfn_to_pfn_cache *gpc) {mark_page_dirty_in_slot (gpc->kvm, gpc->memslot, gpc->gpa >>PAGE_SHIFT );
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 4/13 KVM: pfncache: base offset check on khva rather than gpa
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
After a subsequent patch, the gpa may not always be set whereas khva will (as long as the cache valid flag is also set).
No functional change intended.
---
Cc: Sean Christopherson <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: Paolo Bonzini <[email protected]>
--- virt/kvm/pfncache.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)
virt/kvm/pfncache.c +5 -2
@@ -83,15 +83,18 @@ bool kvm_gpc_check(struct gfn_to_pfn_cache *gpc, unsigned long len)if (!gpc->active)return false; -if ((gpc->gpa & ~PAGE_MASK ) + len >PAGE_SIZE ) +if (gpc->generation != slots->generation)return false; -if (gpc->generation != slots->generation ||kvm_is_error_hva (gpc->uhva)) +if (kvm_is_error_hva (gpc->uhva))return false;if (!gpc->valid)return false; +if (offset_in_page (gpc->khva) + len >PAGE_SIZE ) +return false; +return true; }EXPORT_SYMBOL_GPL (kvm_gpc_check);
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 5/13 KVM: pfncache: allow a cache to be activated with a fixed (userspace) HVA
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
Some cached pages may actually be overlays on guest memory that have a fixed HVA within the VMM. It's pointless to invalidate such cached mappings if the overlay is moved so allow a cache to be activated directly with the HVA to cater for such cases. A subsequent patch will make use of this facility.
---
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: David Woodhouse <[email protected]>
--- include/linux/kvm_host.h | 29 ++++++++++++++++ include/linux/kvm_types.h | 3 +- virt/kvm/pfncache.c | 73 ++++++++++++++++++++++++++++----------- 3 files changed, 84 insertions(+), 21 deletions(-)
include/linux/kvm_host.h +29 -0
@@ -1321,6 +1321,22 @@ void kvm_gpc_init(struct gfn_to_pfn_cache *gpc, struct kvm *kvm,*/ int kvm_gpc_activate (struct gfn_to_pfn_cache *gpc,gpa_t gpa,unsigned long len); +/** +* kvm_gpc_activate_hva - prepare a cached kernel mapping and HPA for a given HVA. +* +* @gpc: struct gfn_to_pfn_cache object. +* @hva: userspace virtual address to map. +* @len: sanity check; the range being access must fit a single page. +* +* @return: 0 for success. +* -EINVAL for a mapping which would cross a page boundary. +* -EFAULT for an untranslatable guest physical address. +* +* The semantics of this function are the same as those of kvm_gpc_activate(). It +* merely bypasses a layer of address translation. +*/ +int kvm_gpc_activate_hva (struct gfn_to_pfn_cache *gpc,unsigned long hva,unsigned long len); +/** * kvm_gpc_check - check validity of a gfn_to_pfn_cache. *
@@ -1378,9 +1394,22 @@ void kvm_gpc_mark_dirty(struct gfn_to_pfn_cache *gpc);* kvm_gpc_gpa - retrieve the guest physical address of a cached mapping * * @gpc: struct gfn_to_pfn_cache object. +* +* @return: If the cache was activated with a fixed HVA then INVALID_GPA +* will be returned. */ gpa_t kvm_gpc_gpa (struct gfn_to_pfn_cache *gpc); +/** +* kvm_gpc_hva - retrieve the fixed host physical address of a cached mapping +* +* @gpc: struct gfn_to_pfn_cache object. +* +* @return: If the cache was activated with a guest physical address then +* 0 will be returned. +*/ +unsigned long kvm_gpc_hva (struct gfn_to_pfn_cache *gpc); +void kvm_sigset_activate (struct kvm_vcpu *vcpu);void kvm_sigset_deactivate (struct kvm_vcpu *vcpu);
include/linux/kvm_types.h +2 -1
@@ -64,7 +64,7 @@ struct gfn_to_hva_cache {struct gfn_to_pfn_cache {u64 generation; -gpa_t gpa; +u64 addr;unsigned long uhva;struct kvm_memory_slot *memslot;struct kvm *kvm;
@@ -77,6 +77,7 @@ struct gfn_to_pfn_cache {enum pfn_cache_usage usage;bool active;bool valid; +bool addr_is_gpa; };#ifdef KVM_ARCH_NR_OBJS_PER_MEMORY_CACHE
virt/kvm/pfncache.c +53 -20
133 lines not on this page · raw message
Signature · 1 line
2.39.2
Re: [PATCH v4 05/13] KVM: pfncache: allow a cache to be activated with a fixed (userspace) HVA
Quoted 9 lines · Paul Durrant
From: Paul Durrant <[email protected]>
Some cached pages may actually be overlays on guest memory that have a fixed HVA within the VMM. It's pointless to invalidate such cached mappings if the overlay is moved so allow a cache to be activated directly with the HVA to cater for such cases. A subsequent patch will make use of this facility.
Signed-off-by Paul Durrant <[email protected]>
I think I gave you one of these in a previous round, but
Re: [PATCH v4 05/13] KVM: pfncache: allow a cache to be activated with a fixed (userspace) HVA
Quoted 9 lines · David Woodhouse
Quoted 8 lines · Paul Durrant
From: Paul Durrant <[email protected]>
Some cached pages may actually be overlays on guest memory that have a fixed HVA within the VMM. It's pointless to invalidate such cached mappings if the overlay is moved so allow a cache to be activated directly with the HVA to cater for such cases. A subsequent patch will make use of this facility.
Quoted 6 lines
Quoted 1 line
Signed-off-by Paul Durrant <[email protected]>I think I gave you one of these in a previous round, but
Reviewed-by David Woodhouse <[email protected]>
Yes. Apologies. You gave it on v2; I definitely have it now.
Paul
Paul Durrant
maintainerPATCH v4 6/13 KVM: xen: allow shared_info to be mapped by fixed HVA
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
The shared_info page is not guest memory as such. It is a dedicated page allocated by the VMM and overlaid onto guest memory in a GFN chosen by the guest. The guest may even request that shared_info be moved from one GFN to another, but the HVA is never going to change. Thus it makes much more sense to map the shared_info page in kernel once using this fixed HVA. Hence add a new KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA attribute type for this purpose and a KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag to advertize its availability. Don't actually advertize it yet though. That will be done in a subsequent patch, which will also add tests for the new attribute type.
Also update the KVM API documentation with the new attribute and also fix it up to consistently refer to 'shared_info' (with the underscore).
NOTE: The change of the kvm_xen_hvm_attr shared_info from struct to union
is technically an ABI change but it's entirely compatible with existing users.
---
Cc: David Woodhouse <[email protected]>
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: [email protected]
v2:
- Define the new attribute and capability but don't advertize the capability yet.
- Add API documentation.
--- Documentation/virt/kvm/api.rst | 25 +++++++++++++++++++------ arch/x86/kvm/xen.c | 28 ++++++++++++++++++++++------ include/uapi/linux/kvm.h | 6 +++++- 3 files changed, 46 insertions(+), 13 deletions(-)
Documentation/virt/kvm/api.rst +19 -6
@@ -353,7 +353,7 @@ The bits in the dirty bitmap are cleared before the ioctl returns, unless KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is enabled. For more information, see the description of the capability. -Note that the Xen shared info page, if configured, shall always be assumed +Note that the Xen shared_info page, if configured, shall always be assumed to be dirty. KVM will not explicitly mark it such.
@@ -5408,8 +5408,9 @@ KVM_PV_ASYNC_CLEANUP_PERFORM __u8 long_mode; __u8 vector; __u8 runstate_update_flag; - struct { + union { __u64 gfn; + __u64 hva; } shared_info; struct { __u32 send_port;
@@ -5437,10 +5438,10 @@ type values: KVM_XEN_ATTR_TYPE_LONG_MODE Sets the ABI mode of the VM to 32-bit or 64-bit (long mode). This - determines the layout of the shared info pages exposed to the VM. + determines the layout of the shared_info page exposed to the VM. KVM_XEN_ATTR_TYPE_SHARED_INFO - Sets the guest physical frame number at which the Xen "shared info" + Sets the guest physical frame number at which the Xen shared_info page resides. Note that although Xen places vcpu_info for the first 32 vCPUs in the shared_info page, KVM does not automatically do so and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO be used
@@ -5449,7 +5450,7 @@ KVM_XEN_ATTR_TYPE_SHARED_INFO not be aware of the Xen CPU id which is used as the index into the vcpu_info[] array, so may know the correct default location. - Note that the shared info page may be constantly written to by KVM; + Note that the shared_info page may be constantly written to by KVM; it contains the event channel bitmap used to deliver interrupts to a Xen guest, amongst other things. It is exempt from dirty tracking mechanisms — KVM will not explicitly mark the page as dirty each
@@ -5458,9 +5459,21 @@ KVM_XEN_ATTR_TYPE_SHARED_INFO any vCPU has been running or any event channel interrupts can be routed to the guest. - Setting the gfn to KVM_XEN_INVALID_GFN will disable the shared info + Setting the gfn to KVM_XEN_INVALID_GFN will disable the shared_info page. +KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA + If the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag is also set in the + Xen capabilities, then this attribute may be used to set the + userspace address at which the shared_info page resides, which + will always be fixed in the VMM regardless of where it is mapped + in guest physical address space. This attribute should be used in + preference to KVM_XEN_ATTR_TYPE_SHARED_INFO as it avoids + unnecessary invalidation of an internal cache when the page is + re-mapped in guest physcial address space. + + Setting the hva to zero will disable the shared_info page. + KVM_XEN_ATTR_TYPE_UPCALL_VECTOR Sets the exception vector used to deliver Xen event channel upcalls. This is the HVM-wide vector injected directly by the hypervisor
arch/x86/kvm/xen.c +22 -6
@@ -34,24 +34,27 @@ static bool kvm_xen_hcall_evtchn_send(struct kvm_vcpu *vcpu, u64 param, u64 *r);DEFINE_STATIC_KEY_DEFERRED_FALSE (kvm_xen_enabled,HZ ); -static int kvm_xen_shared_info_init (struct kvm *kvm,gfn_t gfn) +static int kvm_xen_shared_info_init (struct kvm *kvm,u64 addr,bool addr_is_gfn) {struct gfn_to_pfn_cache *gpc = &kvm->arch.xen.shinfo_cache;struct pvclock_wall_clock *wc; -gpa_t gpa =gfn_to_gpa (gfn);u32 *wc_sec_hi;u32 wc_version;u64 wall_nsec;int ret =0 ;int idx =srcu_read_lock (&kvm->srcu); -if (gfn ==KVM_XEN_INVALID_GFN ) { +if ((addr_is_gfn && addr ==KVM_XEN_INVALID_GFN ) || + (!addr_is_gfn && addr ==0 )) {kvm_gpc_deactivate (gpc); goto out; } do { - ret = kvm_gpc_activate(gpc, gpa, PAGE_SIZE); + if (addr_is_gfn) + ret = kvm_gpc_activate(gpc, gfn_to_gpa(addr), PAGE_SIZE); + else + ret = kvm_gpc_activate_hva(gpc, addr, PAGE_SIZE); if (ret) goto out;
@@ -604,7 +607,6 @@ int kvm_xen_hvm_set_attr(struct kvm *kvm, struct kvm_xen_hvm_attr *data) {int r = -ENOENT ; -switch (data->type) {case KVM_XEN_ATTR_TYPE_LONG_MODE :if (!IS_ENABLED (CONFIG_64BIT ) && data->u.long_mode) {
@@ -619,7 +621,13 @@ int kvm_xen_hvm_set_attr(struct kvm *kvm, struct kvm_xen_hvm_attr *data)case KVM_XEN_ATTR_TYPE_SHARED_INFO :mutex_lock (&kvm->arch.xen.xen_lock); - r =kvm_xen_shared_info_init (kvm, data->u.shared_info.gfn); + r =kvm_xen_shared_info_init (kvm, data->u.shared_info.gfn, true); +mutex_unlock (&kvm->arch.xen.xen_lock); +break ; + +case KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA : +mutex_lock (&kvm->arch.xen.xen_lock); + r =kvm_xen_shared_info_init (kvm, data->u.shared_info.hva, false);mutex_unlock (&kvm->arch.xen.xen_lock);break ;
@@ -684,6 +692,14 @@ int kvm_xen_hvm_get_attr(struct kvm *kvm, struct kvm_xen_hvm_attr *data) r =0 ;break ; +case KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA : +if (kvm->arch.xen.shinfo_cache.active) + data->u.shared_info.hva =kvm_gpc_hva (&kvm->arch.xen.shinfo_cache); +else + data->u.shared_info.hva =0 ; + r =0 ; +break ; +case KVM_XEN_ATTR_TYPE_UPCALL_VECTOR : data->u.vector = kvm->arch.xen.upcall_vector; r =0 ;
include/uapi/linux/kvm.h +5 -1
@@ -1282,6 +1282,7 @@ struct kvm_x86_mce {#define KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL (1 << 4)#define KVM_XEN_HVM_CONFIG_EVTCHN_SEND (1 << 5)#define KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG (1 << 6) +#define KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA (1 << 7)struct kvm_xen_hvm_config {__u32 flags;
@@ -1793,9 +1794,10 @@ struct kvm_xen_hvm_attr {__u8 long_mode;__u8 vector;__u8 runstate_update_flag; -struct { +union {__u64 gfn;#define KVM_XEN_INVALID_GFN ((__u64)-1) +__u64 hva; } shared_info;struct {__u32 send_port;
@@ -1837,6 +1839,8 @@ struct kvm_xen_hvm_attr {#define KVM_XEN_ATTR_TYPE_XEN_VERSION 0x4/* Available with KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG */ #define KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG 0x5 +/* Available with KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA */ +#define KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA 0x6/* Per-vCPU Xen attributes */ #define KVM_XEN_VCPU_GET_ATTR _IOWR(KVMIO, 0xca, struct kvm_xen_vcpu_attr)
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 7/13 KVM: xen: prepare for using 'default' vcpu_info
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
The shared_info page contains an array of 32 vcpu_info structures which may be used by guests (with fewer than 32 vCPUs) if they don't explicitly register vcpu_info structures in their own memory, using a VCPUOP_register_vcpu_info hypercall.
Currently we rely on the VMM always registering vcpu_info structures, even if the guest doesn't make that hypercall, which is somewhat bogus as (as has been stated in the comment of a previous commit) the shared_info page is not guest memory.
Prepare to automatically use the vcpu_info info embedded in shared_info by default, by adding a get_vcpu_info_cache() helper function. This function also passes back an offset to be added to the cached khva. This is currently always zero since we're still relying on the current VMM behaviour. A subsequent patch will make proper use of it.
No functional change intended.
NOTE: To avoid leaking detail of the vcpu_info duality into the main
x86 code, a kvm_xen_guest_time_update() has also been added and use of this requires that kvm_setup_guest_pvclock() ceases to be a static function.
---
Cc: David Woodhouse <[email protected]>
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: [email protected]
v4:
- Added missing offset into calls to kvm_gpc_check() and kvm_gpc_refresh().
v2:
- Fix build warning.
--- arch/x86/include/asm/kvm_host.h | 4 +++ arch/x86/kvm/x86.c | 12 +++---- arch/x86/kvm/xen.c | 60 ++++++++++++++++++++++++--------- arch/x86/kvm/xen.h | 6 +++- 4 files changed, 58 insertions(+), 24 deletions(-)
arch/x86/include/asm/kvm_host.h +4 -0
@@ -2238,4 +2238,8 @@ int memslot_rmap_alloc(struct kvm_memory_slot *slot, unsigned long npages);*/ #define KVM_EXIT_HYPERCALL_MBZ GENMASK_ULL(31, 1) +void kvm_setup_guest_pvclock (struct kvm_vcpu *v, +struct gfn_to_pfn_cache *gpc, +unsigned int offset); +#endif /* _ASM_X86_KVM_HOST_H */
arch/x86/kvm/x86.c +4 -8
@@ -3094,9 +3094,9 @@ u64 get_kvmclock_ns(struct kvm *kvm)return data.clock; } -static void kvm_setup_guest_pvclock (struct kvm_vcpu *v, -struct gfn_to_pfn_cache *gpc, -unsigned int offset) +void kvm_setup_guest_pvclock (struct kvm_vcpu *v, +struct gfn_to_pfn_cache *gpc, +unsigned int offset) {struct kvm_vcpu_arch *vcpu = &v->arch;struct pvclock_vcpu_time_info *guest_hv_clock;
@@ -3232,11 +3232,7 @@ static int kvm_guest_time_update(struct kvm_vcpu *v)if (vcpu->pv_time.active)kvm_setup_guest_pvclock (v, &vcpu->pv_time,0 ); -if (vcpu->xen.vcpu_info_cache.active) -kvm_setup_guest_pvclock (v, &vcpu->xen.vcpu_info_cache, - offsetof(struct compat_vcpu_info , time)); -if (vcpu->xen.vcpu_time_info_cache.active) -kvm_setup_guest_pvclock (v, &vcpu->xen.vcpu_time_info_cache,0 ); +kvm_xen_guest_time_update (v);kvm_hv_setup_tsc_page (v->kvm, &vcpu->hv_clock);return 0 ; }
arch/x86/kvm/xen.c +45 -15
156 lines not on this page · raw message
arch/x86/kvm/xen.h +5 -1
@@ -16,6 +16,7 @@extern struct static_key_false_deferred kvm_xen_enabled; +void kvm_xen_guest_time_update (struct kvm_vcpu *vcpu);int __kvm_xen_has_interrupt (struct kvm_vcpu *vcpu);void kvm_xen_inject_pending_events (struct kvm_vcpu *vcpu);int kvm_xen_vcpu_set_attr (struct kvm_vcpu *vcpu,struct kvm_xen_vcpu_attr *data);
@@ -52,7 +53,6 @@ static inline bool kvm_xen_hypercall_enabled(struct kvm *kvm)static inline int kvm_xen_has_interrupt (struct kvm_vcpu *vcpu) {if (static_branch_unlikely (&kvm_xen_enabled.key) && - vcpu->arch.xen.vcpu_info_cache.active && vcpu->kvm->arch.xen.upcall_vector)return __kvm_xen_has_interrupt (vcpu);
@@ -80,6 +80,10 @@ static inline int kvm_xen_has_pending_timer(struct kvm_vcpu *vcpu)void kvm_xen_inject_timer_irqs (struct kvm_vcpu *vcpu);#else +static inline void kvm_xen_guest_time_update (struct kvm_vcpu *vcpu) +{ +} +static inline int kvm_xen_write_hypercall_page (struct kvm_vcpu *vcpu,u64 data) {return 1 ;
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 8/13 KVM: xen: prevent vcpu_id from changing whilst shared_info is valid
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
To further prepare for automatically using the vcpu_info structures embedded in the shared_info page, we need to ensure that the Xen vcpu_id cannot change under our feet. We can do this by simply returning -EBUSY to any attempt to modify the attribute while the shinfo_cache is active.
---
Cc: David Woodhouse <[email protected]>
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: [email protected]
v3:
- New in this version.
--- Documentation/virt/kvm/api.rst | 3 ++- arch/x86/kvm/xen.c | 18 ++++++++++++++---- 2 files changed, 16 insertions(+), 5 deletions(-)
Documentation/virt/kvm/api.rst +2 -1
@@ -5605,7 +5605,8 @@ KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the Xen vCPU ID of the given vCPU, to allow timer-related VCPU operations to - be intercepted by KVM. + be intercepted by KVM. Note that this must be set before the + shared_info page is set. KVM_XEN_VCPU_ATTR_TYPE_TIMER This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates
arch/x86/kvm/xen.c +14 -4
@@ -752,6 +752,18 @@ int kvm_xen_hvm_get_attr(struct kvm *kvm, struct kvm_xen_hvm_attr *data)return r; } +static int kvm_xen_set_vcpu_id (struct kvm_vcpu *vcpu,unsigned int vcpu_id) +{ +struct kvm *kvm = vcpu->kvm; +struct gfn_to_pfn_cache *gpc = &kvm->arch.xen.shinfo_cache; + +if (gpc->active) +return -EBUSY ; + + vcpu->arch.xen.vcpu_id = vcpu_id; +return 0 ; +} +int kvm_xen_vcpu_set_attr (struct kvm_vcpu *vcpu,struct kvm_xen_vcpu_attr *data) {int idx, r = -ENOENT ;
@@ -941,10 +953,8 @@ int kvm_xen_vcpu_set_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)case KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID :if (data->u.vcpu_id >=KVM_MAX_VCPUS ) r = -EINVAL ; -else { - vcpu->arch.xen.vcpu_id = data->u.vcpu_id; - r =0 ; - } +else + r =kvm_xen_set_vcpu_id (vcpu, data->u.vcpu_id);break ;case KVM_XEN_VCPU_ATTR_TYPE_TIMER :
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 9/13 KVM: xen: automatically use the vcpu_info embedded in shared_info
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
The VMM should only need to set the KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO attribute in response to a VCPUOP_register_vcpu_info hypercall. We can handle the default case internally since we already know where the shared_info page is. Modify get_vcpu_info_cache() to pass back a pointer to the shared info pfn cache (and appropriate offset) for any of the first 32 vCPUs if the attribute has not been set.
A VMM will be able to determine whether it needs to set up default vcpu_info using the previously defined KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA Xen capability flag, which will be advertized in a subsequent patch.
If the VMM subsequently sets a vcpu_info in response to the aforementioned hypercall then the content of the default vcpu_info will be copied to the new location.
Also update the KVM API documentation to describe the new behaviour.
---
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: [email protected]
v4:
- Copy the default vcpu_info content when an explicit vcpu_info is set.
- Amend API documentation.
v3:
- Add a note to the API documentation discussing vcpu_info copying.
v2:
- Dispense with the KVM_XEN_HVM_CONFIG_DEFAULT_VCPU_INFO capability.
- Add API documentation.
--- Documentation/virt/kvm/api.rst | 22 ++++--- arch/x86/kvm/xen.c | 110 ++++++++++++++++++++++++++++++--- 2 files changed, 117 insertions(+), 15 deletions(-)
Documentation/virt/kvm/api.rst +15 -7
@@ -5442,13 +5442,7 @@ KVM_XEN_ATTR_TYPE_LONG_MODE KVM_XEN_ATTR_TYPE_SHARED_INFO Sets the guest physical frame number at which the Xen shared_info - page resides. Note that although Xen places vcpu_info for the first - 32 vCPUs in the shared_info page, KVM does not automatically do so - and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO be used - explicitly even when the vcpu_info for a given vCPU resides at the - "default" location in the shared_info page. This is because KVM may - not be aware of the Xen CPU id which is used as the index into the - vcpu_info[] array, so may know the correct default location. + page resides. Note that the shared_info page may be constantly written to by KVM; it contains the event channel bitmap used to deliver interrupts to
@@ -5564,12 +5558,26 @@ type values: KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO Sets the guest physical address of the vcpu_info for a given vCPU. + The location of the shared_info page for the VM must be specified before + this attribute is set. + + If the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag is also set in the + Xen capabilities then the VMM is not required to set this default + location for any of the first 32 vCPUs; KVM will handle that + internally (using the vcpu_info embedded in the shared_info page). + Otherwise this attribute must be set for all vCPUs. + As with the shared_info page for the VM, the corresponding page may be dirtied at any time if event channel interrupt delivery is enabled, so userspace should always assume that the page is dirty without relying on dirty logging. Setting the gpa to KVM_XEN_INVALID_GPA will disable the vcpu_info. + Note that once vcpu_info is set then it may not be set again unless + it is explicitly disabled beforehand. If it is set for any of the + first 32 vCPUS then the content of the default vcpu_info will be copied + into the specified location. + KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO Sets the guest physical address of an additional pvclock structure for a given vCPU. This is typically used for guest vsyscall support.
arch/x86/kvm/xen.c +102 -8
128 lines not on this page · raw message
Signature · 1 line
2.39.2
David Woodhouse
maintainerRe: [PATCH v4 09/13] KVM: xen: automatically use the vcpu_info embedded in shared_info
Quoted 48 lines · Paul Durrant
arch/x86/kvm/xen.c +0
-0@@ -491,6 +491,21 @@ static void kvm_xen_inject_vcpu_vector(struct kvm_vcpu *v)
static structgfn_to_pfn_cache*get_vcpu_info_cache(structkvm_vcpu*v, unsigned long *offset) {
+ if (!v->arch.xen.vcpu_info_cache.active &&v->arch.xen.vcpu_id<MAX_VIRT_CPUS) { + struct kvm *kvm =v->kvm;arch/x86/kvm/xen.c +13
-0+ + if (offset) { + if (IS_ENABLED(CONFIG_64BIT) && kvm->arch.xen.long_mode) + *offset = offsetof(struct shared_info, + vcpu_info[v->arch.xen.vcpu_id]); + else + *offset = offsetof(struct compat_shared_info, + vcpu_info[v->arch.xen.vcpu_id]); + } + + return &kvm->arch.xen.shinfo_cache; + } +if (offset)
*offset = 0;arch/x86/kvm/xen.c +0
-0@@ -764,6 +779,92 @@ static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) return 0; }arch/x86/kvm/xen.c +20
-0+static int kvm_xen_set_vcpu_info (struct kvm_vcpu *vcpu,gpa_t gpa) +{ + struct kvm *kvm = vcpu->kvm; + struct gfn_to_pfn_cache *si_gpc = &kvm->arch.xen.shinfo_cache; + struct gfn_to_pfn_cache *vi_gpc = &vcpu->arch.xen.vcpu_info_cache; + unsigned long flags; + unsigned long offset; + int ret; + + if (gpa == KVM_XEN_INVALID_GPA) { + kvm_gpc_deactivate(vi_gpc); + return 0; + } + + /* +* In Xen it is not possible for an explicit vcpu_info to be set +* before the shared_info exists since the former is done in response +* to a hypercall and the latter is set up as part of domain creation. +* The first 32 vCPUs have a default vcpu_info embedded in shared_info +* the content of which is copied across when an explicit vcpu_info is
arch/x86/kvm/xen.c static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) · line 811
Quoted 10 lines
arch/x86/kvm/xen.c +10
-0+* set, which can also clearly not be done if we don't know where the +* shared_info is. Hence we need to enforce that the shared_info cache +* is active here. +*/ + if (!si_gpc->active) + return -EINVAL; + + /* Setting an explicit vcpu_info is a one-off operation */ + if (vi_gpc->active) + return -EINVAL;
Is that the errno that Xen will return to the hypercall if a guest tries it? I.e. if the VMM simply returns the errno that it gets from the kernel, is that OK?
Quoted 1 line
+ ret =
kvm_gpc_activate(vi_gpc, gpa, sizeof(struct vcpu_info));
From this moment, can't interrupts be delivered to the new vcpu_info, even though the memcpy hasn't happened yet?
I think we need to ensure that any kvm_xen_set_evtchn_fast() which happens at this point cannot proceed, and falls back to the slow path.
Can we set a flag before we activate the vcpu_info and clear it after the memcpy is done, then make kvm_xen_set_evtchn_fast() return EWOULDBLOCK whenever that flag is set?
The slow path in kvm_xen_set_evtchn() takes kvm->arch.xen.xen_lock and I think kvm_xen_vcpu_set_attr() has taken that same lock before you get to this code, so it works out nicely?
Quoted 16 lines
+ if (ret)
+ return ret;arch/x86/kvm/xen.c +14
-0+ + /* Nothing more to do if the vCPU is not among the first 32 */ + if (vcpu->arch.xen.vcpu_id >= MAX_VIRT_CPUS) + return 0; + + /* + * It's possible that the vcpu_info cache has been invalidated since + * we activated it so we need to go through the check-refresh dance. + */ + read_lock_irqsave(&vi_gpc->lock, flags); + while (!kvm_gpc_check(vi_gpc, sizeof(struct vcpu_info))) { + read_unlock_irqrestore(&vi_gpc->lock, flags); + + ret = kvm_gpc_refresh(vi_gpc, sizeof(struct vcpu_info));
arch/x86/kvm/xen.c static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) · line 839
Quoted 10 lines
arch/x86/kvm/xen.c +10
-0+ if (ret) { + kvm_gpc_deactivate(vi_gpc); + return ret; + } + + read_lock_irqsave(&vi_gpc->lock, flags); + } + + /* Now lock the shared_info cache so we can copy the vcpu_info */ + read_lock(&si_gpc->lock);
This adds a new lock ordering rule of the vcpu_info lock(s) before the shared_info lock. I don't know that it's *wrong* but it seems weird to me; I expected the shared_info to come first?
I avoided taking both at once in kvm_xen_set_evtchn_fast(), although maybe if we are going to have a rule that allows both, we could revisit that. Suspect it isn't needed.
Either way it is worth a clear comment somewhere to document the lock ordering, and I'd also like to know this has been tested with lockdep, which is often cleverer than me.
Quoted 47 lines
+ while (!kvm_gpc_check(si_gpc,
PAGE_SIZE)) { + read_unlock(&si_gpc->lock);arch/x86/kvm/xen.c +26
-0+ + ret = kvm_gpc_refresh(si_gpc, PAGE_SIZE); + if (ret) { + read_unlock_irqrestore(&vi_gpc->lock, flags); + kvm_gpc_deactivate(vi_gpc); + return ret; + } + + read_lock(&si_gpc->lock); + } + + if (IS_ENABLED(CONFIG_64BIT) && kvm->arch.xen.long_mode) + offset = offsetof(struct shared_info, + vcpu_info[vcpu->arch.xen.vcpu_id]); + else + offset = offsetof(struct compat_shared_info, + vcpu_info[vcpu->arch.xen.vcpu_id]); + + memcpy(vi_gpc->khva, si_gpc->khva + offset, sizeof(struct vcpu_info)); + + read_unlock(&si_gpc->lock); + read_unlock_irqrestore(&vi_gpc->lock, flags); + + return 0; +} +int
kvm_xen_vcpu_set_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data){
int idx, r =-ENOENT;arch/x86/kvm/xen.c +0
-0@@ -779,14 +880,7 @@ int kvm_xen_vcpu_set_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data)BUILD_BUG_ON(offsetof(struct
vcpu_info, time) !=offsetof(struct compat_vcpu_info, time));
- if (data->u.gpa==KVM_XEN_INVALID_GPA) {
- kvm_gpc_deactivate(&vcpu->arch.xen.vcpu_info_cache); - r = 0;
- break;
- }
-
- r = kvm_gpc_activate(&vcpu->arch.xen.vcpu_info_cache, -data->u.gpa,sizeof(struct vcpu_info)); + r =kvm_xen_set_vcpu_info(vcpu, data->u.gpa);
if (!r)
kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
Re: [PATCH v4 09/13] KVM: xen: automatically use the vcpu_info embedded in shared_info
Quoted 55 lines · David Woodhouse
Quoted 54 lines · Paul Durrant
arch/x86/kvm/xen.c +41
-0@@ -491,6 +491,21 @@ static void kvm_xen_inject_vcpu_vector(struct kvm_vcpu *v) static struct gfn_to_pfn_cache *get_vcpu_info_cache(struct kvm_vcpu *v, unsigned long *offset) { + if (!v->arch.xen.vcpu_info_cache.active && v->arch.xen.vcpu_id < MAX_VIRT_CPUS) { + struct kvm *kvm = v->kvm; + + if (offset) { + if (IS_ENABLED(CONFIG_64BIT) && kvm->arch.xen.long_mode) + *offset = offsetof(struct shared_info, + vcpu_info[v->arch.xen.vcpu_id]); + else + *offset = offsetof(struct compat_shared_info, + vcpu_info[v->arch.xen.vcpu_id]); + } + + return &kvm->arch.xen.shinfo_cache; + } + if (offset) *offset = 0;@@ -764,6 +779,92 @@ static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) return 0; } +static int kvm_xen_set_vcpu_info (struct kvm_vcpu *vcpu,gpa_t gpa) +{ + struct kvm *kvm = vcpu->kvm; + struct gfn_to_pfn_cache *si_gpc = &kvm->arch.xen.shinfo_cache; + struct gfn_to_pfn_cache *vi_gpc = &vcpu->arch.xen.vcpu_info_cache; + unsigned long flags; + unsigned long offset; + int ret; + + if (gpa == KVM_XEN_INVALID_GPA) { + kvm_gpc_deactivate(vi_gpc); + return 0; + } + + /* + * In Xen it is not possible for an explicit vcpu_info to be set + * before the shared_info exists since the former is done in response + * to a hypercall and the latter is set up as part of domain creation. + * The first 32 vCPUs have a default vcpu_info embedded in shared_info + * the content of which is copied across when an explicit vcpu_info is + * set, which can also clearly not be done if we don't know where the + * shared_info is. Hence we need to enforce that the shared_info cache + * is active here. + */ + if (!si_gpc->active) + return -EINVAL;
Quoted 9 lines
arch/x86/kvm/xen.c static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) · line 811
Quoted 4 lines
arch/x86/kvm/xen.c +4
-0+ + /* Setting an explicit vcpu_info is a one-off operation */ + if (vi_gpc->active) + return -EINVAL;Is that the errno that Xen will return to the hypercall if a guest tries it? I.e. if the VMM simply returns the errno that it gets from the kernel, is that OK?
Yes, I checked. Xen returns -EINVAL.
Quoted 5 lines
arch/x86/kvm/xen.c static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) · line 813
Quoted 1 line
arch/x86/kvm/xen.c +1
-0+ ret = kvm_gpc_activate(vi_gpc, gpa, sizeof(struct vcpu_info));From this moment, can't interrupts be delivered to the new
vcpu_info,
even though the memcpy hasn't happened yet?
Hmm, that's a good point. TBH it would be nice to have an 'activate and leave locked' primitive to avoid this.
Quoted 11 lines
I think we need to ensure that any
kvm_xen_set_evtchn_fast()which happens at this point cannot proceed, and falls back to the slow path.Can we set a flag before we activate the
vcpu_infoand clear it after the memcpy is done, then makekvm_xen_set_evtchn_fast()return EWOULDBLOCK whenever that flag is set?The slow path in
kvm_xen_set_evtchn()takeskvm->arch.xen.xen_lockand I thinkkvm_xen_vcpu_set_attr()has taken that same lock before you get to this code, so it works out nicely?
Yes, I think that is safe... but if we didn't have the window between activating the vcpu_info cache and doing the copy we'd also be ok I think... Or perhaps we could simply preserve evtchn_pending_sel and copy the rest of it?
Quoted 33 lines
Quoted 18 lines
arch/x86/kvm/xen.c +16
-0+ if (ret) + return ret; + + /* Nothing more to do if the vCPU is not among the first 32 */ + if (vcpu->arch.xen.vcpu_id >= MAX_VIRT_CPUS) + return 0; + + /* + * It's possible that the vcpu_info cache has been invalidated since + * we activated it so we need to go through the check-refresh dance. + */ + read_lock_irqsave(&vi_gpc->lock, flags); + while (!kvm_gpc_check(vi_gpc, sizeof(struct vcpu_info))) { + read_unlock_irqrestore(&vi_gpc->lock, flags); + + ret = kvm_gpc_refresh(vi_gpc, sizeof(struct vcpu_info));arch/x86/kvm/xen.c static int kvm_xen_set_vcpu_id(struct kvm_vcpu *vcpu, unsigned int vcpu_id) · line 839
Quoted 10 lines
arch/x86/kvm/xen.c +10
-0+ if (ret) { + kvm_gpc_deactivate(vi_gpc); + return ret; + } + + read_lock_irqsave(&vi_gpc->lock, flags); + } + + /* Now lock the shared_info cache so we can copy the vcpu_info */ + read_lock(&si_gpc->lock);This adds a new lock ordering rule of the
vcpu_infolock(s) before theshared_infolock. I don't know that it's *wrong* but it seems weird to me; I expected theshared_infoto come first?
Quoted 8 lines
I avoided taking both at once in
kvm_xen_set_evtchn_fast(), although maybe if we are going to have a rule that allows both, we could revisit that. Suspect it isn't needed.Either way it is worth a clear comment somewhere to document the lock ordering, and I'd also like to know this has been tested with lockdep, which is often cleverer than me.
Ok. I agree that shared_info before vcpu_info does seem more intuitive and maybe it would be better given the code in kvm_xen_set_evtchn_fast(). I'll seem how messy it gets in re-ordering and add a comment as you suggest.
Paul
Quoted 48 lines
Quoted 47 lines
arch/x86/kvm/xen.c +29
-8+ while (!kvm_gpc_check(si_gpc, PAGE_SIZE)) { + read_unlock(&si_gpc->lock); + + ret = kvm_gpc_refresh(si_gpc, PAGE_SIZE); + if (ret) { + read_unlock_irqrestore(&vi_gpc->lock, flags); + kvm_gpc_deactivate(vi_gpc); + return ret; + } + + read_lock(&si_gpc->lock); + } + + if (IS_ENABLED(CONFIG_64BIT) && kvm->arch.xen.long_mode) + offset = offsetof(struct shared_info, + vcpu_info[vcpu->arch.xen.vcpu_id]); + else + offset = offsetof(struct compat_shared_info, + vcpu_info[vcpu->arch.xen.vcpu_id]); + + memcpy(vi_gpc->khva, si_gpc->khva + offset, sizeof(struct vcpu_info)); + + read_unlock(&si_gpc->lock); + read_unlock_irqrestore(&vi_gpc->lock, flags); + + return 0; +} + int kvm_xen_vcpu_set_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data) { int idx, r = -ENOENT;@@ -779,14 +880,7 @@ int kvm_xen_vcpu_set_attr(struct kvm_vcpu *vcpu, struct kvm_xen_vcpu_attr *data) BUILD_BUG_ON(offsetof(struct vcpu_info, time) != offsetof(struct compat_vcpu_info, time)); - if (data->u.gpa == KVM_XEN_INVALID_GPA) { - kvm_gpc_deactivate(&vcpu->arch.xen.vcpu_info_cache); - r = 0; - break; - } - - r = kvm_gpc_activate(&vcpu->arch.xen.vcpu_info_cache, - data->u.gpa, sizeof(struct vcpu_info)); + r = kvm_xen_set_vcpu_info(vcpu, data->u.gpa); if (!r) kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
David Woodhouse
maintainerRe: [PATCH v4 09/13] KVM: xen: automatically use the vcpu_info embedded in shared_info
Quoted 8 lines · Paul Durrant
Quoted 5 lines
Quoted 1 line
+ ret =
kvm_gpc_activate(vi_gpc, gpa, sizeof(struct vcpu_info));From this moment, can't interrupts be delivered to the new
vcpu_info, even though the memcpy hasn't happened yet?Hmm, that's a good point. TBH it would be nice to have an 'activate and leave locked' primitive to avoid this.
I suppose so from the caller's point of view in this case, but I'm somewhat disinclined to add that complexity to the pfncache code.
We take the refresh_lock *mutex* in __kvm_gpc_refresh() so it's not as simple as declaring that said function is called with the gpc rwlock already held.
We also do the final gpc_unmap_khva() of the old mapping after dropping the lock; *could* we call that with a write lock held? A write lock which is going to be taken the MM notifier callbacks? Well, maybe not in the case of the first *activate* which isn't really a 'refresh' per se but the whole thing is making my skin itch. I don't like it.
Quoted 12 lines
Quoted 11 lines
I think we need to ensure that any
kvm_xen_set_evtchn_fast()which happens at this point cannot proceed, and falls back to the slow path.Can we set a flag before we activate the
vcpu_infoand clear it after the memcpy is done, then makekvm_xen_set_evtchn_fast()return EWOULDBLOCK whenever that flag is set?The slow path in
kvm_xen_set_evtchn()takeskvm->arch.xen.xen_lockand I thinkkvm_xen_vcpu_set_attr()has taken that same lock before you get to this code, so it works out nicely?
Quoted 6 lines
Yes, I think that is safe... but if we didn't have the window between activating the
vcpu_infocache and doing the copy we'd also be ok I think... Or perhaps we could simply preserveevtchn_pending_seland copy the rest of it?
I suppose you could just write the evtchn_pending_sel word in the new vcpu_info GPA to zero before setting up the pfncache for it.
When when you do the memcpy, you don't *just* memcpy the evtchn_pending_sel word; you use the bitwise OR of the old and new, so you catch any bits which got set in the new word in the interim?
But then again, who moves the vcpu_info while there are actually interrupts in-flight to the vCPU in question? Maybe we just declare that we don't care, and that interrupts may be lost in that case? Even if *Xen* wouldn't have lost them (and I don't even know that part is true).
Quoted 13 lines
Quoted 4 lines
This adds a new lock ordering rule of the
vcpu_infolock(s) before theshared_infolock. I don't know that it's *wrong* but it seems weird to me; I expected theshared_infoto come first?Quoted 8 lines
I avoided taking both at once in
kvm_xen_set_evtchn_fast(), although maybe if we are going to have a rule that allows both, we could revisit that. Suspect it isn't needed.Either way it is worth a clear comment somewhere to document the lock ordering, and I'd also like to know this has been tested with lockdep, which is often cleverer than me.
Quoted 5 lines
Ok. I agree that
shared_infobeforevcpu_infodoes seem more intuitive and maybe it would be better given the code inkvm_xen_set_evtchn_fast(). I'll seem how messy it gets in re-ordering and add a comment as you suggest.
I think they look interchangeable in this case. If we *do* take them both in kvm_xen_set_evtchn_fast() then maybe we can simplify the slow path where it set the bits in shared_info but then the vcpu_info gpc was invalid. That currently uses a kvm->arch.xen.evtchn_pending_sel shadow of the bits, and just kicks the vCPU to deliver them for itself... but maybe that whole thing could be dropped, and kvm_xen_set_evtchn_fast() can just return EWOULDBLOCK if it fails to lock *both* shared_info and vcpu_info at the same time?
I didn't do that before, because I didn't want to introduce lock ordering rules. But I'm happier to do so now. And I think we can ditch a lot of hairy asm in kvm_xen_inject_pending_events() ?
Re: [PATCH v4 09/13] KVM: xen: automatically use the vcpu_info embedded in shared_info
Quoted 73 lines · David Woodhouse
Quoted 8 lines · Paul Durrant
Quoted 5 lines
Quoted 1 line
+ ret =
kvm_gpc_activate(vi_gpc, gpa, sizeof(struct vcpu_info));From this moment, can't interrupts be delivered to the new
vcpu_info,
even though the memcpy hasn't happened yet?Hmm, that's a good point. TBH it would be nice to have an 'activate and leave locked' primitive to avoid this.
I suppose so from the caller's point of view in this case, but I'm somewhat disinclined to add that complexity to the pfncache code.
We take the
refresh_lock*mutex* in__kvm_gpc_refresh()so it's not as simple as declaring that said function is called with the gpc rwlock already held.We also do the final
gpc_unmap_khva()of the old mapping after dropping the lock; *could* we call that with a write lock held? A write lock which is going to be taken the MM notifier callbacks? Well, maybe not in the case of the first *activate* which isn't really a 'refresh' per se but the whole thing is making my skin itch. I don't like it.Quoted 12 lines
Quoted 11 lines
I think we need to ensure that any
kvm_xen_set_evtchn_fast()which happens at this point cannot proceed, and falls back to the slow path.Can we set a flag before we activate the
vcpu_infoand clear it after the memcpy is done, then makekvm_xen_set_evtchn_fast()return EWOULDBLOCK whenever that flag is set?The slow path in
kvm_xen_set_evtchn()takeskvm->arch.xen.xen_lockand I thinkkvm_xen_vcpu_set_attr()has taken that same lock before you get to this code, so it works out nicely?Quoted 6 lines
Yes, I think that is safe... but if we didn't have the window between activating the
vcpu_infocache and doing the copy we'd also be ok I think... Or perhaps we could simply preserveevtchn_pending_seland copy the rest of it?I suppose you could just write the
evtchn_pending_selword in the newvcpu_infoGPA to zero before setting up the pfncache for it.When when you do the memcpy, you don't *just* memcpy the
evtchn_pending_selword; you use the bitwise OR of the old and new, so you catch any bits which got set in the new word in the interim?But then again, who moves the
vcpu_infowhile there are actually interrupts in-flight to the vCPU in question? Maybe we just declare that we don't care, and that interrupts may be lost in that case? Even if *Xen* wouldn't have lost them (and I don't even know that part is true).Quoted 13 lines
Quoted 4 lines
This adds a new lock ordering rule of the
vcpu_infolock(s) before theshared_infolock. I don't know that it's *wrong* but it seems weird to me; I expected theshared_infoto come first?Quoted 8 lines
I avoided taking both at once in
kvm_xen_set_evtchn_fast(), although maybe if we are going to have a rule that allows both, we could revisit that. Suspect it isn't needed.Either way it is worth a clear comment somewhere to document the lock ordering, and I'd also like to know this has been tested with lockdep, which is often cleverer than me.
Quoted 5 lines
Ok. I agree that
shared_infobeforevcpu_infodoes seem more intuitive and maybe it would be better given the code inkvm_xen_set_evtchn_fast(). I'll seem how messy it gets in re-ordering and add a comment as you suggest.
Quoted 9 lines
I think they look interchangeable in this case. If we *do* take them both in
kvm_xen_set_evtchn_fast()then maybe we can simplify the slow path where it set the bits inshared_infobut then thevcpu_infogpc was invalid. That currently uses akvm->arch.xen.evtchn_pending_selshadow of the bits, and just kicks the vCPU to deliver them for itself... but maybe that whole thing could be dropped, andkvm_xen_set_evtchn_fast()can just return EWOULDBLOCK if it fails to lock *both*shared_infoandvcpu_infoat the same time?
Yes, I think that sounds like a neater approach.
Quoted 4 lines
I didn't do that before, because I didn't want to introduce lock ordering rules. But I'm happier to do so now. And I think we can ditch a lot of hairy asm in
kvm_xen_inject_pending_events()?
Messing with the asm sounds like something for a follow-up though.
Paul
David Woodhouse
maintainerRe: [PATCH v4 09/13] KVM: xen: automatically use the vcpu_info embedded in shared_info
Quoted 12 lines · Paul Durrant
Quoted 11 lines
I think they look interchangeable in this case. If we *do* take them both in
kvm_xen_set_evtchn_fast()then maybe we can simplify the slow path where it set the bits inshared_infobut then thevcpu_infogpc was invalid. That currently uses akvm->arch.xen.evtchn_pending_selshadow of the bits, and just kicks the vCPU to deliver them for itself... but maybe that whole thing could be dropped, andkvm_xen_set_evtchn_fast()can just return EWOULDBLOCK if it fails to lock *both*shared_infoandvcpu_infoat the same time?
Quoted 8 lines
Yes, I think that sounds like a neater approach.
Quoted 4 lines
I didn't do that before, because I didn't want to introduce lock ordering rules. But I'm happier to do so now. And I think we can ditch a lot of hairy asm in
kvm_xen_inject_pending_events()?Messing with the asm sounds like something for a follow-up though.
AFAICT we can just delete the whole bloody lot. But yes, it's definitely a later cleanup, enabled by the fact that we (will) now have an agreed way of taking both locks at once, which I didn't want to do in the first place.
Paul Durrant
maintainerPATCH v4 10/13 KVM: selftests / xen: set KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
If the capability (KVM_XEN_HVM_CONFIG_EVTCHN_SEND) is present then set the guest's vCPU id to match the chosen vcpu_info offset.
Also make some cosmetic fixes to the code for clarity.
---
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: David Woodhouse <[email protected]>
v3:
- Remame VCPU_ID to XEN_VCPU_ID.
- Set vcpu_id before the shared_info page is set.
v2:
- New in this version.
--- .../selftests/kvm/x86_64/xen_shinfo_test.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-)
tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c +15 -4
@@ -38,6 +38,8 @@#define VCPU_INFO_VADDR (SHINFO_REGION_GVA + 0x40)#define RUNSTATE_VADDR (SHINFO_REGION_GVA + PAGE_SIZE + PAGE_SIZE - 15) +#define XEN_VCPU_ID 1/* Must correspond to offset of VCPU_INFO_[V]ADDR */ +#define EVTCHN_VECTOR 0x10#define EVTCHN_TEST1 15
@@ -410,7 +412,7 @@ static void *juggle_shinfo_state(void *arg)struct kvm_xen_hvm_attr cache_activate = { .type =KVM_XEN_ATTR_TYPE_SHARED_INFO , - .u.shared_info.gfn =SHINFO_REGION_GPA /PAGE_SIZE + .u.shared_info.gfn =SHINFO_ADDR /PAGE_SIZE };struct kvm_xen_hvm_attr cache_deactivate = {
@@ -446,6 +448,7 @@ int main(int argc, char *argv[])bool do_runstate_flag = !!(xen_caps &KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG );bool do_eventfd_tests = !!(xen_caps &KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL );bool do_evtchn_tests = do_eventfd_tests && !!(xen_caps &KVM_XEN_HVM_CONFIG_EVTCHN_SEND ); +bool has_vcpu_id = !!(xen_caps &KVM_XEN_HVM_CONFIG_EVTCHN_SEND );clock_gettime (CLOCK_REALTIME , &min_ts);
@@ -492,9 +495,17 @@ int main(int argc, char *argv[]) "Failed to read back RUNSTATE_UPDATE_FLAG attr"); } +if (has_vcpu_id) { +struct kvm_xen_vcpu_attr vid = { + .type =KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID , + .u.vcpu_id =XEN_VCPU_ID , + }; +vcpu_ioctl (vcpu,KVM_XEN_VCPU_SET_ATTR , &vid); + } +struct kvm_xen_hvm_attr ha = { .type =KVM_XEN_ATTR_TYPE_SHARED_INFO , - .u.shared_info.gfn =SHINFO_REGION_GPA /PAGE_SIZE , + .u.shared_info.gfn =SHINFO_ADDR /PAGE_SIZE , };vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &ha);
@@ -983,8 +994,8 @@ int main(int argc, char *argv[])struct pvclock_wall_clock *wc;struct pvclock_vcpu_time_info *ti, *ti2; - wc =addr_gpa2hva (vm,SHINFO_REGION_GPA +0xc00 ); - ti =addr_gpa2hva (vm,SHINFO_REGION_GPA +0x40 +0x20 ); + wc =addr_gpa2hva (vm,SHINFO_ADDR +0xc00 ); + ti =addr_gpa2hva (vm,VCPU_INFO_ADDR +0x20 ); ti2 =addr_gpa2hva (vm,PVTIME_ADDR );if (verbose) {
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 11/13 KVM: selftests / xen: map shared_info using HVA rather than GFN
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
Using the HVA of the shared_info page is more efficient, so if the capability (KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA) is present use that method to do the mapping.
NOTE: Have the juggle_shinfo_state() thread map and unmap using both
GFN and HVA, to make sure the older mechanism is not broken.
---
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: David Woodhouse <[email protected]>
v3:
- Re-work the juggle_shinfo_state() thread
v2:
- New in this version.
--- .../selftests/kvm/x86_64/xen_shinfo_test.c | 44 +++++++++++++++---- 1 file changed, 35 insertions(+), 9 deletions(-)
tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c +35 -9
@@ -395,6 +395,7 @@ static int cmp_timespec(struct timespec *a, struct timespec *b)return 0 ; } +static struct shared_info *shinfo;static struct vcpu_info *vinfo;static struct kvm_vcpu *vcpu;
@@ -410,20 +411,38 @@ static void *juggle_shinfo_state(void *arg) {struct kvm_vm *vm = (struct kvm_vm *)arg; -struct kvm_xen_hvm_attr cache_activate = { +struct kvm_xen_hvm_attr cache_activate_gfn = { .type =KVM_XEN_ATTR_TYPE_SHARED_INFO , .u.shared_info.gfn =SHINFO_ADDR /PAGE_SIZE }; -struct kvm_xen_hvm_attr cache_deactivate = { +struct kvm_xen_hvm_attr cache_deactivate_gfn = { .type =KVM_XEN_ATTR_TYPE_SHARED_INFO , .u.shared_info.gfn =KVM_XEN_INVALID_GFN }; +struct kvm_xen_hvm_attr cache_activate_hva = { + .type =KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA , + .u.shared_info.hva = (unsigned long )shinfo + }; + +struct kvm_xen_hvm_attr cache_deactivate_hva = { + .type =KVM_XEN_ATTR_TYPE_SHARED_INFO , + .u.shared_info.hva =0 + }; + +int xen_caps =kvm_check_cap (KVM_CAP_XEN_HVM ); +for (;;) { -__vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &cache_activate); -__vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &cache_deactivate); +__vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &cache_activate_gfn);pthread_testcancel (); +__vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &cache_deactivate_gfn); + +if (xen_caps &KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA ) { +__vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &cache_activate_hva); +pthread_testcancel (); +__vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &cache_deactivate_hva); + } }return NULL ;
@@ -449,6 +468,7 @@ int main(int argc, char *argv[])bool do_eventfd_tests = !!(xen_caps &KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL );bool do_evtchn_tests = do_eventfd_tests && !!(xen_caps &KVM_XEN_HVM_CONFIG_EVTCHN_SEND );bool has_vcpu_id = !!(xen_caps &KVM_XEN_HVM_CONFIG_EVTCHN_SEND ); +bool has_shinfo_hva = !!(xen_caps &KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA );clock_gettime (CLOCK_REALTIME , &min_ts);
@@ -459,7 +479,7 @@ int main(int argc, char *argv[]) SHINFO_REGION_GPA, SHINFO_REGION_SLOT, 3, 0);virt_map (vm,SHINFO_REGION_GVA ,SHINFO_REGION_GPA ,3 ); -struct shared_info *shinfo =addr_gpa2hva (vm,SHINFO_VADDR ); + shinfo =addr_gpa2hva (vm,SHINFO_VADDR );int zero_fd =open ("/dev/zero" ,O_RDONLY );TEST_ASSERT (zero_fd !=-1 ,"Failed to open /dev/zero" );
@@ -503,10 +523,16 @@ int main(int argc, char *argv[])vcpu_ioctl (vcpu,KVM_XEN_VCPU_SET_ATTR , &vid); } -struct kvm_xen_hvm_attr ha = { - .type =KVM_XEN_ATTR_TYPE_SHARED_INFO , - .u.shared_info.gfn =SHINFO_ADDR /PAGE_SIZE , - }; +struct kvm_xen_hvm_attr ha = {}; + +if (has_shinfo_hva) { + ha.type =KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA ; + ha.u.shared_info.hva = (unsigned long )shinfo; + }else { + ha.type =KVM_XEN_ATTR_TYPE_SHARED_INFO ; + ha.u.shared_info.gfn =SHINFO_ADDR /PAGE_SIZE ; + } +vm_ioctl (vm,KVM_XEN_HVM_SET_ATTR , &ha);/*
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 12/13 KVM: selftests / xen: don't explicitly set the vcpu_info address
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
If the vCPU id is set and the shared_info is mapped using HVA then we can infer that KVM has the ability to use a default vcpu_info mapping. Hence we can stop setting the address of the vcpu_info structure.
NOTE: We still explicitly set vcpu_info half way through testing (to point
at exactly the place it already is) to make sure that setting the attribute does not fail.
---
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: David Woodhouse <[email protected]>
v3:
- Add a guest sync point to set the vcpu_info attribute
v2:
- New in this version.
--- .../testing/selftests/kvm/x86_64/xen_shinfo_test.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)
tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c +13 -1
@@ -68,6 +68,7 @@ enum {TEST_POLL_TIMEOUT ,TEST_POLL_MASKED ,TEST_POLL_WAKE , +SET_VCPU_INFO ,TEST_TIMER_PAST ,TEST_LOCKING_SEND_RACE , TEST_LOCKING_POLL_RACE,
@@ -327,6 +328,10 @@ static void guest_code(void)GUEST_SYNC (TEST_POLL_WAKE ); +/* Set the vcpu_info to point at exactly the place it already is to +* make sure the attribute is functional. */ +GUEST_SYNC (SET_VCPU_INFO ); +/* A timer wake an *unmasked* port which should wake us with an * actual interrupt, while we're polling on a different port. */ ports[0 ]++;
@@ -549,7 +554,10 @@ int main(int argc, char *argv[]) .type = KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO, .u.gpa = VCPU_INFO_ADDR, }; -vcpu_ioctl (vcpu,KVM_XEN_VCPU_SET_ATTR , &vi); + +if (!has_vcpu_id || !has_shinfo_hva) { +vcpu_ioctl (vcpu,KVM_XEN_VCPU_SET_ATTR , &vi); + }struct kvm_xen_vcpu_attr pvclock = { .type =KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO ,
@@ -903,6 +911,10 @@ int main(int argc, char *argv[])alarm (1 );break ; +case SET_VCPU_INFO : +vcpu_ioctl (vcpu,KVM_XEN_VCPU_SET_ATTR , &vi); +break ; +case TEST_TIMER_PAST :TEST_ASSERT (!evtchn_irq_expected,"Expected event channel IRQ but it didn't happen" );
Signature · 1 line
2.39.2
Paul Durrant
maintainerPATCH v4 13/13 KVM: xen: advertize the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA capability
In reply to Paul DurrantFrom: Paul Durrant <[email protected]>
Now that all relevant kernel changes and selftests are in place, enable the new capability.
---
Cc: Sean Christopherson <[email protected]>
Cc: Paolo Bonzini <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: "H. Peter Anvin" <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: [email protected]
v2:
- New in this version.
--- arch/x86/kvm/x86.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/kvm/x86.c +2 -1
@@ -4527,7 +4527,8 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL |KVM_XEN_HVM_CONFIG_SHARED_INFO |KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL | -KVM_XEN_HVM_CONFIG_EVTCHN_SEND ; +KVM_XEN_HVM_CONFIG_EVTCHN_SEND | +KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA ;if (sched_info_on ()) r |=KVM_XEN_HVM_CONFIG_RUNSTATE |KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG ;
Signature · 1 line
2.39.2