Skip to contentSkip to conversation

KVM: x86: allow for more CPUID entries

PATCH In review Vitaly Kuznetsov 2 participants 9 messages Oct 1, 2020 – Oct 5, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover KVM: x86: allow for more CPUID entries

Changes since RFC:
- "KVM: x86: disconnect kvm_check_cpuid() from vcpu->arch.cpuid_entries" added to allow running kvm_check_cpuid() before vcpu->arch.cpuid_entries/ vcpu->arch.cpuid_nent are changed [Sean Christopherson]
- Shorten local variable names in kvm_vcpu_ioctl_set_cpuid[,2] [Sean Christopherson]
- Drop unneeded 'out' labels from kvm_vcpu_ioctl_set_cpuid[,2] and return directly whenever possible [Sean Christopherson]

Original description:

With QEMU and newer AMD CPUs (namely: Epyc 'Rome') the current limit for KVM_MAX_CPUID_ENTRIES(80) is reported to be hit. Last time it was raised from '40' in 2010. We can, of course, just bump it a little bit to fix the immediate issue but the report made me wonder why we need to pre- allocate vcpu->arch.cpuid_entries array instead of sizing it dynamically. This RFC is intended to feed my curiosity.

Very mildly tested with selftests/kvm-unit-tests and nothing seems to break. I also don't have access to the system where the original issue was reported but chances we're fixing it are very good IMO as just the second patch alone was reported to be sufficient.

Vitaly Kuznetsov (3):
KVM: x86: disconnect kvm_check_cpuid() from vcpu->arch.cpuid_entries
KVM: x86: allocate vcpu->arch.cpuid_entries dynamically
KVM: x86: bump KVM_MAX_CPUID_ENTRIES

 arch/x86/include/asm/kvm_host.h |   4 +-
 arch/x86/kvm/cpuid.c            | 123 +++++++++++++++++++-------------
 arch/x86/kvm/x86.c              |   1 +
 3 files changed, 75 insertions(+), 53 deletions(-)
Signature · 1 line
2.25.4
#1 lore raw

PATCH 1/3 KVM: x86: disconnect kvm_check_cpuid() from vcpu->arch.cpuid_entries

As a preparatory step to allocating vcpu->arch.cpuid_entries dynamically make kvm_check_cpuid() check work with an arbitrary 'struct kvm_cpuid_entry2' array.

Currently, when kvm_check_cpuid() fails we reset vcpu->arch.cpuid_nent to 0 and this is kind of weird, i.e. one would expect CPUIDs to remain unchanged when KVM_SET_CPUID[2] call fails.

No functional change intended. It would've been possible to move the updated kvm_check_cpuid() in kvm_vcpu_ioctl_set_cpuid2() and check the supplied input before we start updating vcpu->arch.cpuid_entries/nent but we can't do the same in kvm_vcpu_ioctl_set_cpuid() as we'll have to copy 'struct kvm_cpuid_entry' entries first. The change will be made when vcpu->arch.cpuid_entries[] array becomes allocated dynamically.

---
 arch/x86/kvm/cpuid.c | 38 +++++++++++++++++++++++---------------
 1 file changed, 23 insertions(+), 15 deletions(-)
arch/x86/kvm/cpuid.c +23 -15
index 37c3668a774f..529348ddedc1 100644
@@ -54,7 +54,24 @@ static u32 xstate_required_size(u64 xstate_bv, bool compacted)
 
 #define F feature_bit
 
-static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
+static inline struct kvm_cpuid_entry2 *cpuid_entry2_find(
+	struct kvm_cpuid_entry2 *entries, int nent, u32 function, u32 index)
+{
+	struct kvm_cpuid_entry2 *e;
+	int i;
+
+	for (i = 0; i < nent; i++) {
+		e = &entries[i];
+
+		if (e->function == function && (e->index == index ||
+		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
+			return e;
+	}
+
+	return NULL;
+}
+
+static int kvm_check_cpuid(struct kvm_cpuid_entry2 *entries, int nent)
 {
 	struct kvm_cpuid_entry2 *best;
 
@@ -62,7 +79,7 @@ static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
 	 * The existing code assumes virtual address is 48-bit or 57-bit in the
 	 * canonical address checks; exit if it is ever changed.
 	 */
-	best = kvm_find_cpuid_entry(vcpu, 0x80000008, 0);
+	best = cpuid_entry2_find(entries, nent, 0x80000008, 0);
 	if (best) {
 		int vaddr_bits = (best->eax & 0xff00) >> 8;
 
@@ -220,7 +237,7 @@ int kvm_vcpu_ioctl_set_cpuid(struct kvm_vcpu *vcpu,
 		vcpu->arch.cpuid_entries[i].padding[2] = 0;
 	}
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		kvfree(cpuid_entries);
@@ -250,7 +267,7 @@ int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu,
 			   cpuid->nent * sizeof(struct kvm_cpuid_entry2)))
 		goto out;
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		goto out;
@@ -940,17 +957,8 @@ int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid,
 struct kvm_cpuid_entry2 *kvm_find_cpuid_entry(struct kvm_vcpu *vcpu,
 					      u32 function, u32 index)
 {
-	struct kvm_cpuid_entry2 *e;
-	int i;
-
-	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
-		e = &vcpu->arch.cpuid_entries[i];
-
-		if (e->function == function && (e->index == index ||
-		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
-			return e;
-	}
-	return NULL;
+	return cpuid_entry2_find(vcpu->arch.cpuid_entries, vcpu->arch.cpuid_nent,
+				 function, index);
 }
 EXPORT_SYMBOL_GPL(kvm_find_cpuid_entry);
 
Signature · 1 line
2.25.4
Reviewed-by on v1 1/3
#2 lore raw

Re: [PATCH 1/3] KVM: x86: disconnect kvm_check_cpuid() from vcpu->arch.cpuid_entries

Quoted 7 lines · Vitaly Kuznetsov

As a preparatory step to allocating vcpu->arch.cpuid_entries dynamically make kvm_check_cpuid() check work with an arbitrary 'struct kvm_cpuid_entry2' array.

Currently, when kvm_check_cpuid() fails we reset vcpu->arch.cpuid_nent to 0 and this is kind of weird, i.e. one would expect CPUIDs to remain unchanged when KVM_SET_CPUID[2] call fails.

Since this specific patch doesn't fix this, maybe move this chunk to following patches, or to the cover letter?

Quoted 81 lines

No functional change intended. It would've been possible to move the updated kvm_check_cpuid() in kvm_vcpu_ioctl_set_cpuid2() and check the supplied input before we start updating vcpu->arch.cpuid_entries/nent but we can't do the same in kvm_vcpu_ioctl_set_cpuid() as we'll have to copy 'struct kvm_cpuid_entry' entries first. The change will be made when vcpu->arch.cpuid_entries[] array becomes allocated dynamically.

---
 arch/x86/kvm/cpuid.c | 38 +++++++++++++++++++++++---------------
 1 file changed, 23 insertions(+), 15 deletions(-)
arch/x86/kvm/cpuid.c +21 -10
index 37c3668a774f..529348ddedc1 100644
@@ -54,7 +54,24 @@ static u32 xstate_required_size(u64 xstate_bv, bool compacted)
 
 #define F feature_bit
 
-static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
+static inline struct kvm_cpuid_entry2 *cpuid_entry2_find(
+	struct kvm_cpuid_entry2 *entries, int nent, u32 function, u32 index)
+{
+	struct kvm_cpuid_entry2 *e;
+	int i;
+
+	for (i = 0; i < nent; i++) {
+		e = &entries[i];
+
+		if (e->function == function && (e->index == index ||
+		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
+			return e;
+	}
+
+	return NULL;
+}
+
+static int kvm_check_cpuid(struct kvm_cpuid_entry2 *entries, int nent)
 {
 	struct kvm_cpuid_entry2 *best;
 
@@ -62,7 +79,7 @@ static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
 	 * The existing code assumes virtual address is 48-bit or 57-bit in the
 	 * canonical address checks; exit if it is ever changed.
 	 */
-	best = kvm_find_cpuid_entry(vcpu, 0x80000008, 0);
+	best = cpuid_entry2_find(entries, nent, 0x80000008, 0);
 	if (best) {
 		int vaddr_bits = (best->eax & 0xff00) >> 8;
 
@@ -220,7 +237,7 @@ int kvm_vcpu_ioctl_set_cpuid(struct kvm_vcpu *vcpu,
 		vcpu->arch.cpuid_entries[i].padding[2] = 0;
 	}
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		kvfree(cpuid_entries);
@@ -250,7 +267,7 @@ int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu,
 			   cpuid->nent * sizeof(struct kvm_cpuid_entry2)))
 		goto out;
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		goto out;
@@ -940,17 +957,8 @@ int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid,
 struct kvm_cpuid_entry2 *kvm_find_cpuid_entry(struct kvm_vcpu *vcpu,
 					      u32 function, u32 index)
 {
-	struct kvm_cpuid_entry2 *e;
-	int i;
-
-	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
-		e = &vcpu->arch.cpuid_entries[i];
-

arch/x86/kvm/cpuid.c int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid, · line 964

Quoted 10 lines
arch/x86/kvm/cpuid.c +2 -5
-		if (e->function == function && (e->index == index ||
-		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
-			return e;
-	}
-	return NULL;
+	return cpuid_entry2_find(vcpu->arch.cpuid_entries, vcpu->arch.cpuid_nent,
+				 function, index);
 }
 EXPORT_SYMBOL_GPL(kvm_find_cpuid_entry);
 

Other than minor note to the commit message, this looks fine, so

Best regards,
Maxim Levitsky

#3 lore raw

Re: [PATCH 1/3] KVM: x86: disconnect kvm_check_cpuid() from vcpu->arch.cpuid_entries

Quoted 10 lines · Maxim Levitsky
Quoted 7 lines · Vitaly Kuznetsov

As a preparatory step to allocating vcpu->arch.cpuid_entries dynamically make kvm_check_cpuid() check work with an arbitrary 'struct kvm_cpuid_entry2' array.

Currently, when kvm_check_cpuid() fails we reset vcpu->arch.cpuid_nent to 0 and this is kind of weird, i.e. one would expect CPUIDs to remain unchanged when KVM_SET_CPUID[2] call fails.

Since this specific patch doesn't fix this, maybe move this chunk to following patches, or to the cover letter?

Basically, this kind of pairs with what's after 'No functional change intended' below: we admit the problem but don't fix it because we can't (yet) and then in PATCH3 we do two things at once. It would be great to separate these two changes but this doesn't seem to be possible without an unneeded code churn.

That said, I'm completely fine with dropping this chunk from the commit message if it sound inapropriate here.

Quoted 94 lines
Quoted 82 lines

No functional change intended. It would've been possible to move the updated kvm_check_cpuid() in kvm_vcpu_ioctl_set_cpuid2() and check the supplied input before we start updating vcpu->arch.cpuid_entries/nent but we can't do the same in kvm_vcpu_ioctl_set_cpuid() as we'll have to copy 'struct kvm_cpuid_entry' entries first. The change will be made when vcpu->arch.cpuid_entries[] array becomes allocated dynamically.

---
 arch/x86/kvm/cpuid.c | 38 +++++++++++++++++++++++---------------
 1 file changed, 23 insertions(+), 15 deletions(-)
arch/x86/kvm/cpuid.c +21 -10
index 37c3668a774f..529348ddedc1 100644
@@ -54,7 +54,24 @@ static u32 xstate_required_size(u64 xstate_bv, bool compacted)
 
 #define F feature_bit
 
-static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
+static inline struct kvm_cpuid_entry2 *cpuid_entry2_find(
+	struct kvm_cpuid_entry2 *entries, int nent, u32 function, u32 index)
+{
+	struct kvm_cpuid_entry2 *e;
+	int i;
+
+	for (i = 0; i < nent; i++) {
+		e = &entries[i];
+
+		if (e->function == function && (e->index == index ||
+		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
+			return e;
+	}
+
+	return NULL;
+}
+
+static int kvm_check_cpuid(struct kvm_cpuid_entry2 *entries, int nent)
 {
 	struct kvm_cpuid_entry2 *best;
 
@@ -62,7 +79,7 @@ static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
 	 * The existing code assumes virtual address is 48-bit or 57-bit in the
 	 * canonical address checks; exit if it is ever changed.
 	 */
-	best = kvm_find_cpuid_entry(vcpu, 0x80000008, 0);
+	best = cpuid_entry2_find(entries, nent, 0x80000008, 0);
 	if (best) {
 		int vaddr_bits = (best->eax & 0xff00) >> 8;
 
@@ -220,7 +237,7 @@ int kvm_vcpu_ioctl_set_cpuid(struct kvm_vcpu *vcpu,
 		vcpu->arch.cpuid_entries[i].padding[2] = 0;
 	}
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		kvfree(cpuid_entries);
@@ -250,7 +267,7 @@ int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu,
 			   cpuid->nent * sizeof(struct kvm_cpuid_entry2)))
 		goto out;
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		goto out;
@@ -940,17 +957,8 @@ int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid,
 struct kvm_cpuid_entry2 *kvm_find_cpuid_entry(struct kvm_vcpu *vcpu,
 					      u32 function, u32 index)
 {
-	struct kvm_cpuid_entry2 *e;
-	int i;
-
-	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
-		e = &vcpu->arch.cpuid_entries[i];
-

arch/x86/kvm/cpuid.c int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid, · line 964

Quoted 10 lines
arch/x86/kvm/cpuid.c +2 -5
-		if (e->function == function && (e->index == index ||
-		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
-			return e;
-	}
-	return NULL;
+	return cpuid_entry2_find(vcpu->arch.cpuid_entries, vcpu->arch.cpuid_nent,
+				 function, index);
 }
 EXPORT_SYMBOL_GPL(kvm_find_cpuid_entry);
 

Other than minor note to the commit message, this looks fine, so

Quoted 2 lines

Thanks!

Signature · 1 line
Vitaly
#4 lore raw

Re: [PATCH 1/3] KVM: x86: disconnect kvm_check_cpuid() from vcpu->arch.cpuid_entries

Quoted 13 lines · Vitaly Kuznetsov
Quoted 10 lines · Maxim Levitsky
Quoted 7 lines · Vitaly Kuznetsov

As a preparatory step to allocating vcpu->arch.cpuid_entries dynamically make kvm_check_cpuid() check work with an arbitrary 'struct kvm_cpuid_entry2' array.

Currently, when kvm_check_cpuid() fails we reset vcpu->arch.cpuid_nent to 0 and this is kind of weird, i.e. one would expect CPUIDs to remain unchanged when KVM_SET_CPUID[2] call fails.

Since this specific patch doesn't fix this, maybe move this chunk to following patches, or to the cover letter?

Quoted 8 lines

Basically, this kind of pairs with what's after 'No functional change intended' below: we admit the problem but don't fix it because we can't (yet) and then in PATCH3 we do two things at once. It would be great to separate these two changes but this doesn't seem to be possible without an unneeded code churn.

That said, I'm completely fine with dropping this chunk from the commit message if it sound inapropriate here.

It just threw me a bit off course while trying to understand what the patch does.

Best regards,
Maxim Levitsky

Quoted 98 lines
Quoted 93 lines
Quoted 81 lines

No functional change intended. It would've been possible to move the updated kvm_check_cpuid() in kvm_vcpu_ioctl_set_cpuid2() and check the supplied input before we start updating vcpu->arch.cpuid_entries/nent but we can't do the same in kvm_vcpu_ioctl_set_cpuid() as we'll have to copy 'struct kvm_cpuid_entry' entries first. The change will be made when vcpu->arch.cpuid_entries[] array becomes allocated dynamically.

---
 arch/x86/kvm/cpuid.c | 38 +++++++++++++++++++++++---------------
 1 file changed, 23 insertions(+), 15 deletions(-)
arch/x86/kvm/cpuid.c +21 -10
index 37c3668a774f..529348ddedc1 100644
@@ -54,7 +54,24 @@ static u32 xstate_required_size(u64 xstate_bv, bool compacted)
 
 #define F feature_bit
 
-static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
+static inline struct kvm_cpuid_entry2 *cpuid_entry2_find(
+	struct kvm_cpuid_entry2 *entries, int nent, u32 function, u32 index)
+{
+	struct kvm_cpuid_entry2 *e;
+	int i;
+
+	for (i = 0; i < nent; i++) {
+		e = &entries[i];
+
+		if (e->function == function && (e->index == index ||
+		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
+			return e;
+	}
+
+	return NULL;
+}
+
+static int kvm_check_cpuid(struct kvm_cpuid_entry2 *entries, int nent)
 {
 	struct kvm_cpuid_entry2 *best;
 
@@ -62,7 +79,7 @@ static int kvm_check_cpuid(struct kvm_vcpu *vcpu)
 	 * The existing code assumes virtual address is 48-bit or 57-bit in the
 	 * canonical address checks; exit if it is ever changed.
 	 */
-	best = kvm_find_cpuid_entry(vcpu, 0x80000008, 0);
+	best = cpuid_entry2_find(entries, nent, 0x80000008, 0);
 	if (best) {
 		int vaddr_bits = (best->eax & 0xff00) >> 8;
 
@@ -220,7 +237,7 @@ int kvm_vcpu_ioctl_set_cpuid(struct kvm_vcpu *vcpu,
 		vcpu->arch.cpuid_entries[i].padding[2] = 0;
 	}
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		kvfree(cpuid_entries);
@@ -250,7 +267,7 @@ int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu,
 			   cpuid->nent * sizeof(struct kvm_cpuid_entry2)))
 		goto out;
 	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu);
+	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
 	if (r) {
 		vcpu->arch.cpuid_nent = 0;
 		goto out;
@@ -940,17 +957,8 @@ int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid,
 struct kvm_cpuid_entry2 *kvm_find_cpuid_entry(struct kvm_vcpu *vcpu,
 					      u32 function, u32 index)
 {
-	struct kvm_cpuid_entry2 *e;
-	int i;
-
-	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
-		e = &vcpu->arch.cpuid_entries[i];
-

arch/x86/kvm/cpuid.c int kvm_dev_ioctl_get_cpuid(struct kvm_cpuid2 *cpuid, · line 964

Quoted 10 lines
arch/x86/kvm/cpuid.c +2 -5
-		if (e->function == function && (e->index == index ||
-		    !(e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX)))
-			return e;
-	}
-	return NULL;
+	return cpuid_entry2_find(vcpu->arch.cpuid_entries, vcpu->arch.cpuid_nent,
+				 function, index);
 }
 EXPORT_SYMBOL_GPL(kvm_find_cpuid_entry);
 

Other than minor note to the commit message, this looks fine, so

Quoted 2 lines

Thanks!

#5 lore raw

PATCH 2/3 KVM: x86: allocate vcpu->arch.cpuid_entries dynamically

In reply to Vitaly Kuznetsov

The current limit for guest CPUID leaves (KVM_MAX_CPUID_ENTRIES, 80) is reported to be insufficient but before we bump it let's switch to allocating vcpu->arch.cpuid_entries[] array dynamically. Currently, 'struct kvm_cpuid_entry2' is 40 bytes so vcpu->arch.cpuid_entries is 3200 bytes which accounts for 1/4 of the whole 'struct kvm_vcpu_arch' but having it pre-allocated (for all vCPUs which we also pre-allocate) gives us no real benefits.

Another plus of the dynamic allocation is that we now do kvm_check_cpuid() check before we assign anything to vcpu->arch.cpuid_nent/cpuid_entries so no changes are made in case the check fails.

Opportunistically remove unneeded 'out' labels from kvm_vcpu_ioctl_set_cpuid()/kvm_vcpu_ioctl_set_cpuid2() and return directly whenever possible.

---
 arch/x86/include/asm/kvm_host.h |  2 +-
 arch/x86/kvm/cpuid.c            | 89 +++++++++++++++++++--------------
 arch/x86/kvm/x86.c              |  1 +
 3 files changed, 53 insertions(+), 39 deletions(-)
arch/x86/include/asm/kvm_host.h +1 -1
index d0f77235da92..7d259e21ea04 100644
@@ -637,7 +637,7 @@ struct kvm_vcpu_arch {
 	int halt_request; /* real mode on Intel only */
 
 	int cpuid_nent;
-	struct kvm_cpuid_entry2 cpuid_entries[KVM_MAX_CPUID_ENTRIES];
+	struct kvm_cpuid_entry2 *cpuid_entries;
 
 	int maxphyaddr;
 	int max_tdp_level;
arch/x86/kvm/cpuid.c +51 -38

123 lines not on this page · raw message

arch/x86/kvm/x86.c +1 -0
index c4015a43cc8a..f8ed1bde18af 100644
@@ -9877,6 +9877,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 	kvm_mmu_destroy(vcpu);
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
 	free_page((unsigned long)vcpu->arch.pio_data);
+	kvfree(vcpu->arch.cpuid_entries);
 	if (!lapic_in_kernel(vcpu))
 		static_key_slow_dec(&kvm_no_apic_vcpu);
 }
Signature · 1 line
2.25.4
Reviewed-by on v1 2/3
#6 lore raw

Re: [PATCH 2/3] KVM: x86: allocate vcpu->arch.cpuid_entries dynamically

Quoted 170 lines · Vitaly Kuznetsov

The current limit for guest CPUID leaves (KVM_MAX_CPUID_ENTRIES, 80) is reported to be insufficient but before we bump it let's switch to allocating vcpu->arch.cpuid_entries[] array dynamically. Currently, 'struct kvm_cpuid_entry2' is 40 bytes so vcpu->arch.cpuid_entries is 3200 bytes which accounts for 1/4 of the whole 'struct kvm_vcpu_arch' but having it pre-allocated (for all vCPUs which we also pre-allocate) gives us no real benefits.

Another plus of the dynamic allocation is that we now do kvm_check_cpuid() check before we assign anything to vcpu->arch.cpuid_nent/cpuid_entries so no changes are made in case the check fails.

Opportunistically remove unneeded 'out' labels from kvm_vcpu_ioctl_set_cpuid()/kvm_vcpu_ioctl_set_cpuid2() and return directly whenever possible.

---
 arch/x86/include/asm/kvm_host.h |  2 +-
 arch/x86/kvm/cpuid.c            | 89 +++++++++++++++++++--------------
 arch/x86/kvm/x86.c              |  1 +
 3 files changed, 53 insertions(+), 39 deletions(-)
arch/x86/include/asm/kvm_host.h +1 -1
index d0f77235da92..7d259e21ea04 100644
@@ -637,7 +637,7 @@ struct kvm_vcpu_arch {
 	int halt_request; /* real mode on Intel only */
 
 	int cpuid_nent;
-	struct kvm_cpuid_entry2 cpuid_entries[KVM_MAX_CPUID_ENTRIES];
+	struct kvm_cpuid_entry2 *cpuid_entries;
 
 	int maxphyaddr;
 	int max_tdp_level;
arch/x86/kvm/cpuid.c +51 -38
index 529348ddedc1..3fe20c4da0a6 100644
@@ -210,46 +210,53 @@ int kvm_vcpu_ioctl_set_cpuid(struct kvm_vcpu *vcpu,
 			     struct kvm_cpuid_entry __user *entries)
 {
 	int r, i;
-	struct kvm_cpuid_entry *cpuid_entries = NULL;
+	struct kvm_cpuid_entry *e = NULL;
+	struct kvm_cpuid_entry2 *e2 = NULL;
 
-	r = -E2BIG;
 	if (cpuid->nent > KVM_MAX_CPUID_ENTRIES)
-		goto out;
+		return -E2BIG;
+
 	if (cpuid->nent) {
-		cpuid_entries = vmemdup_user(entries,
-					     array_size(sizeof(struct kvm_cpuid_entry),
-							cpuid->nent));
-		if (IS_ERR(cpuid_entries)) {
-			r = PTR_ERR(cpuid_entries);
-			goto out;
+		e = vmemdup_user(entries, array_size(sizeof(*e), cpuid->nent));
+		if (IS_ERR(e))
+			return PTR_ERR(e);
+
+		e2 = kvmalloc_array(cpuid->nent, sizeof(*e2), GFP_KERNEL_ACCOUNT);
+		if (!e2) {
+			r = -ENOMEM;
+			goto out_free_cpuid;
 		}
 	}
 	for (i = 0; i < cpuid->nent; i++) {
-		vcpu->arch.cpuid_entries[i].function = cpuid_entries[i].function;
-		vcpu->arch.cpuid_entries[i].eax = cpuid_entries[i].eax;
-		vcpu->arch.cpuid_entries[i].ebx = cpuid_entries[i].ebx;
-		vcpu->arch.cpuid_entries[i].ecx = cpuid_entries[i].ecx;
-		vcpu->arch.cpuid_entries[i].edx = cpuid_entries[i].edx;
-		vcpu->arch.cpuid_entries[i].index = 0;
-		vcpu->arch.cpuid_entries[i].flags = 0;
-		vcpu->arch.cpuid_entries[i].padding[0] = 0;
-		vcpu->arch.cpuid_entries[i].padding[1] = 0;
-		vcpu->arch.cpuid_entries[i].padding[2] = 0;
+		e2[i].function = e[i].function;
+		e2[i].eax = e[i].eax;
+		e2[i].ebx = e[i].ebx;
+		e2[i].ecx = e[i].ecx;
+		e2[i].edx = e[i].edx;
+		e2[i].index = 0;
+		e2[i].flags = 0;
+		e2[i].padding[0] = 0;
+		e2[i].padding[1] = 0;
+		e2[i].padding[2] = 0;
 	}
-	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
+
+	r = kvm_check_cpuid(e2, cpuid->nent);
 	if (r) {
-		vcpu->arch.cpuid_nent = 0;
-		kvfree(cpuid_entries);
-		goto out;
+		kvfree(e2);
+		goto out_free_cpuid;
 	}
 
+	kvfree(vcpu->arch.cpuid_entries);
+	vcpu->arch.cpuid_entries = e2;
+	vcpu->arch.cpuid_nent = cpuid->nent;
+
 	cpuid_fix_nx_cap(vcpu);
 	kvm_update_cpuid_runtime(vcpu);
 	kvm_vcpu_after_set_cpuid(vcpu);
 
-	kvfree(cpuid_entries);
-out:
+out_free_cpuid:
+	kvfree(e);
+
 	return r;
 }
 
@@ -257,26 +264,32 @@ int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu,
 			      struct kvm_cpuid2 *cpuid,
 			      struct kvm_cpuid_entry2 __user *entries)
 {
+	struct kvm_cpuid_entry2 *e2 = NULL;
 	int r;
 
-	r = -E2BIG;
 	if (cpuid->nent > KVM_MAX_CPUID_ENTRIES)
-		goto out;
-	r = -EFAULT;
-	if (copy_from_user(&vcpu->arch.cpuid_entries, entries,
-			   cpuid->nent * sizeof(struct kvm_cpuid_entry2)))
-		goto out;
-	vcpu->arch.cpuid_nent = cpuid->nent;
-	r = kvm_check_cpuid(vcpu->arch.cpuid_entries, cpuid->nent);
+		return -E2BIG;
+
+	if (cpuid->nent) {
+		e2 = vmemdup_user(entries, array_size(sizeof(*e2), cpuid->nent));
+		if (IS_ERR(e2))
+			return PTR_ERR(e2);
+	}
+
+	r = kvm_check_cpuid(e2, cpuid->nent);
 	if (r) {
-		vcpu->arch.cpuid_nent = 0;
-		goto out;
+		kvfree(e2);
+		return r;
 	}
 
+	kvfree(vcpu->arch.cpuid_entries);
+	vcpu->arch.cpuid_entries = e2;
+	vcpu->arch.cpuid_nent = cpuid->nent;
+
 	kvm_update_cpuid_runtime(vcpu);
 	kvm_vcpu_after_set_cpuid(vcpu);
-out:
-	return r;
+
+	return 0;
 }
 
 int kvm_vcpu_ioctl_get_cpuid2(struct kvm_vcpu *vcpu,
arch/x86/kvm/x86.c +0 -0
index c4015a43cc8a..f8ed1bde18af 100644

arch/x86/kvm/x86.c void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) · line 9883

Quoted 7 lines
arch/x86/kvm/x86.c +1 -0
@@ -9877,6 +9877,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 	kvm_mmu_destroy(vcpu);
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
 	free_page((unsigned long)vcpu->arch.pio_data);
+	kvfree(vcpu->arch.cpuid_entries);
 	if (!lapic_in_kernel(vcpu))
 		static_key_slow_dec(&kvm_no_apic_vcpu);
 }

Best regards,
Maxim Levitsky

#7 lore raw

PATCH 3/3 KVM: x86: bump KVM_MAX_CPUID_ENTRIES

In reply to Vitaly Kuznetsov

As vcpu->arch.cpuid_entries is now allocated dynamically, the only remaining use for KVM_MAX_CPUID_ENTRIES is to check KVM_SET_CPUID/ KVM_SET_CPUID2 input for sanity. Since it was reported that the current limit (80) is insufficient for some CPUs, bump KVM_MAX_CPUID_ENTRIES and use an arbitrary value '256' as the new limit.

---
 arch/x86/include/asm/kvm_host.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/include/asm/kvm_host.h +1 -1
index 7d259e21ea04..f6d6df64e63a 100644
@@ -133,7 +133,7 @@ static inline gfn_t gfn_to_index(gfn_t gfn, gfn_t base_gfn, int level)
 #define KVM_NUM_MMU_PAGES (1 << KVM_MMU_HASH_SHIFT)
 #define KVM_MIN_FREE_MMU_PAGES 5
 #define KVM_REFILL_PAGES 25
-#define KVM_MAX_CPUID_ENTRIES 80
+#define KVM_MAX_CPUID_ENTRIES 256
 #define KVM_NR_FIXED_MTRR_REGION 88
 #define KVM_NR_VAR_MTRR 8
 
Signature · 1 line
2.25.4
Reviewed-by on v1 3/3
#8 lore raw

Re: [PATCH 3/3] KVM: x86: bump KVM_MAX_CPUID_ENTRIES

Quoted 17 lines · Vitaly Kuznetsov

As vcpu->arch.cpuid_entries is now allocated dynamically, the only remaining use for KVM_MAX_CPUID_ENTRIES is to check KVM_SET_CPUID/ KVM_SET_CPUID2 input for sanity. Since it was reported that the current limit (80) is insufficient for some CPUs, bump KVM_MAX_CPUID_ENTRIES and use an arbitrary value '256' as the new limit.

---
 arch/x86/include/asm/kvm_host.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/include/asm/kvm_host.h +0 -0
index 7d259e21ea04..f6d6df64e63a 100644

arch/x86/include/asm/kvm_host.h static inline gfn_t gfn_to_index(gfn_t gfn, gfn_t base_gfn, int level) · line 139

Quoted 8 lines
arch/x86/include/asm/kvm_host.h +1 -1
@@ -133,7 +133,7 @@ static inline gfn_t gfn_to_index(gfn_t gfn, gfn_t base_gfn, int level)
 #define KVM_NUM_MMU_PAGES (1 << KVM_MMU_HASH_SHIFT)
 #define KVM_MIN_FREE_MMU_PAGES 5
 #define KVM_REFILL_PAGES 25
-#define KVM_MAX_CPUID_ENTRIES 80
+#define KVM_MAX_CPUID_ENTRIES 256
 #define KVM_NR_FIXED_MTRR_REGION 88
 #define KVM_NR_VAR_MTRR 8
 

Best regards,
Maxim Levitsky